Chemical Pathology BEING A DISCUSSION OF GENERAL PATH- OLOGY FROM THE STANDPOINT OF THE CHEMICAL PROCESSES INVOLVED BY H. GIDEON WELLS, Ph.D., M.D. 'I PROFESSOR OF PATHOLOGY IN THE UNIVERSITY OF CHICAGO AND IN RUSH MEDICAL COLLEGE, CHICAGO; DIRECTOR OI' THE OTHO S. A. SPRAGUB MEMORIAL INSTITUTE FOURTH EDITION. REVISED AXD RESET PHILADELPHIA AND LOXDOX W. B. SAUNDERS COMPANY 1920 Copyright, 1907, by W. B. Saunders Company. Revised, entirely reset, reprinted and recopyrighted February, 1914. Revised, entirely reset, reprinted and recopy- righted January, 1918. Revised, entirely reset, reprinted and recopyrighted July, 1920. Copyright, 1920, by W. B. Saunders Company PRINTED IN AMERICA TO ^ u 5 V 1 3 3^ e K 1 0 e n THIS BOOK IS RESPECTFULLY DEDICATED, AS A SLIGHT TOKEN OF THE GRATITUDE AND ESTEEM OF HIS PUPIL 720791 rUKFACE K) IHK FOURTH EDITION TnK rapid growth of intorost in the ohomical problems of iiiodical and Ijiological science is shown by the groat incrcas(> in the amount of mat(M'ial which must be inchided in (>aclisncceedinKe(Ution. Ahhoush this hit(>st edition has hvcu subj(>cted to extensive revision and many miiioi- alterations, yet tlu^ fieneral i)lan has not been changed. The rapidly growing iiifoiination concerning the nutritional factoi-s that are essential to growth and rej)aii-, and without which serious "Defi- ciency Diseases" may arise, has necessitated the introduction of a new" chapter to cover this subject, the importance of which has been ac- centuated by the war and its sequels. The growing bulk of material on the Reactions of Immunity reciuired a rearrangement of this material, so that a separate chajiter on Anaphylaxis and Allergy has l)een provided, foi- j)ur):)oses of convenience. Numerous sections have been entirely rewritten, and few pages have not required re- vision or addition. In order to prevent the increasing material that must be included from resulting in too cumbersome a volume, much more of the matter is printed in smaller type. It is hoped that this arrangement will achieve its aim without serious reduction in facility of use. The author recognizes fully that it w^ould be easily possible to report the existing state of knowledge on the topics covered in "Chemical Pathology" in a much ])riefer space, if only completely established evidence weie included. "With the object of serving as a guide to investigators, and with the hope of stimulating further investigations, much more than this minimal amount of existing evidence is included. It is also recog- nized that the brief discussion of the elementary principles of physical chemistry and the fundamentals of the physics and chemistry of living cells, which constitutes the introductory chapter, maj^ be out of place in a work on Pathology, and the elimination of this chapter has been seriously considered. Repeated assurances of the usefulness of such a presentation, however, have resulted in its retention, at least for the present. As with all previous editions, mj' indebtedness must be acknowl- edged to numerous colleagues who have kindly read over the sections of this book which most closely concern their own fields, and especially to the members of my Department and of the Sprague Institute who have made manj^ useful suggestions. The chapter on Diabetes is, as before, contributed by Dr. R. T. AVoodyatt, Director of the Laboratory of Clinical Research of the Otho S. A. Sprague ^lemorial Institute. H. G. W. Chicago, III.. July, 1920. PREFACE TO THE FIRST EDITION During the past score of years the subject of biological chemistry has attracted the attention and labors of a constantly increasing num- ber of investigators, many of whom have, for one reason or another, been interested in pathological conditions. Sometimes the physiolo- gist has sought for light on his problems in the evidence afforded by re- lated pathological conditions. Frequently chnicians have studied the metabohc changes and the composition of the products of disease pro- cesses. Relatively seldom, unfortunately, has the pathologist at- tacked his problems by chemical methods. From the above and other sources have come scattered fragments of information concerning the chemical changes that occur in pathological phenomena. Only when bearing upon conditions such as gout and diabetes, which concern ahke the physiologist, the clinician, and the pathologist, have the fragments been moulded together into a homogeneous whole. For the most part they still remain isolated, uncorrelated, frequently uncon- firmed items of information, scattered through medical, chemical, physiological, and physical literature. It has been the aim of the writer to collect these scattered fragments as completely as possible, and to use them as a basis for a consideration of General Pathology from the standpoint of the chemical processes which occur in pathological conditions. Owing to the diffusely scat- tered conditions of the literature on which this work is based, it cannot be claimed that all of the many contributions from which useful in- formation might be obtained have been noticed; but it is hoped that a sufficiently thorough collection of material has been made to afford a fair basis for a consideration of "Chemical Pathology." The time seems ripe for an effort of tliis nature. Within the past few years great and encouraging advances have been made in biological chem- istry, which in many instances seem to throw hght upon pathological processes. In medicine, the use of chemical methods in the study of clinical manifestations has become more general, and has yielded valuable information. Pathologists have come to feel that the op- portunities for the acquirement of knowledge by means of morphologi- cal studies have become reduced to a minimum, while the fields of pathological physiology and chemistry lie still almost unexplored. The development of research upon the subject of natural and acquired immunity has presented innumerable problems, all of which are essentially chemical. And perhaps most important of all is the general 8 PREFACE TO THE FIRST EDITION awakening of an appreciation of the importance of physiological chem- istry to mecHcal science, which has led to the introduction of laboratory courses on this subject in every medical school worthy of the name. A book on (^hemical Pathology should, therefore, seek to supply information to a varied group of readers. It should furnish collateral reading to the student who for the first time goes over the subject of General Pathology, which his text-books usually consider chief!}' from the morphological standpoint. It should e.xploit to the graduate in medicine the advances that are being made along lines that are of fundamental importance to clinical medicine. It should serve for the investigator in biological chemistry or in pathology as a source of information concerning the ground upon which the two subjects over- lap— the "Grenzgebiete" of Patholog}' and Physiological Chemistry. And, above all, it should afford a guide to the sources of our knowledge of these subjects, since nothing but direct familiarity with the original reports of the investigators themselves can give the student an im- personal view of the actual status of the questions under consideration. On account of this multiplicity of the objects in view, it has often been necessary to consider certain topics from more than one standpoint; which explains, perhaps, certain apparent irregularities in the style and manner of treatment. It has been assumed that the reader has at least an elementary knowlefige of organic and physiological chemistry. For the benefit of those whose studies in these subjects date back some years, it has seemed advisable to include in an introductory chapter an epitome of the more modem views concei-ning the chemistry of the protein mole- cule, the composition of the animal cell, and the principles of physical chemistry, in as far as they apply to biological problems. The general consideration of "Enzymes" in Chapter II is written with a similar object. In discussing these fundamental topics it has seemed advis- able to omit detailed references to the numerous original sources, — these may be found cjuoted in the special text-books cited in the foot- notes; but in presenting the more distinctly pathological topics the attempt has been made to render all the important literature available to the reader and investigator. To economize space, a complete bibli- ography has not been inserted when this exists already collected in some readily accessible review or original article; hence the references cited in the foot-notes will generally be found to include only the more recent jHiblications. These references have ])een so selected, however, that they will be found to furnish bibliographical matter sufficient to lead the investigator to all the imj)ortant literature on the topics covered in this book. As to those subjects (such as gout , diabetes, and gast ro-int(>stinal i)utrefaction) which, .because of their great practical j'linical interest, have already been discussed in available monographs at greater length than the scope of this work would permit, it has seemed appropriate merely to suiniiiai-iz(^ the most recent views and PREFACE TO THE FIRST EDITION 9 julvances, rclVniiifi; tlic leader to llic special ti'enfise.s for llie jj;eiieral and historical discussion.s. It is with the greatest pleasure tluit I acknowledge my indeljtedness to manj^ colleagues in the University of Chicago, who have kindly read the sections of my manuscript that touch upon theii-own special fields, and whose criticism and advice have been of the greatest assistance; their mimb(>r alone prevents my thanking them by name. Most par- ticularl}', however, must I express my debt to my former instructor, Professor Lafayette B. Mendel, of Yale University, whose kindly criticism and suggestions have been of inestimable value. For con- stant assistance in the preparation of the manuscript, and for the revision of the bibliography', I am ind(>bted to mv wife. H. G. W. CONTENTS CHAPTER I Paoe Introduction' 17 The Chemistry and Physics of the Cell 17 Chemistrj^ of the Essential Cell Constituents 18 Proteins 19 Fats and Lipoids (Lipins) 22 Carbohydrates 23 Inorganic Substances 24 The Physical Chemistry of the Cell and Its Constituents 24 Crystalloids and Their Properties 24 Colloids 31 The Structure of the Cell in Relation to Its Chemistry and Physics ... 39 The Nucleus 40 The Cytoplasm 42 The Cell-wall 45 CHAPTER II Enzymes 48 The Nature of Enzymes and Their Actions 49 The Principles of Enzyme Action 50 The Toxicity of Enzymes 55 Anti-enzymes 57 The Intracellular Enzymes 62 Oxidizing Enzymes 63 Lipase 70 Amylase or Diastase 73 CHAPTER III Enzymes (Continued) .- • ■ 75 Intracellular Proteases (Proteolytic Enzymes), Including a Considera- tion of Autolysis 75 Autolysis 76 Relation of Autolysis to MetaboUsm 81 Defense of the Cells Against Their Autolytic Enzymes 82 Autolysis in Pathological Processes '. 85 CHAPTER IV The Chemistry of Bacteria .\nd Their Products 101 Structure and Physical Properties 101 Chemical Composition 103 Bacterial Enzymes 109 Poisonous Bacterial Products 115 Ptomains 116 Toxins 120 Endotoxins 124 Poisonous Bacterial Proteins 125 Bacterial Pigments 126 11 12 CONTENTS CHAPTER V Page Chemistry of the Axoial Parasites 128 Protozoa 129 Cestodes 131 Nematodes 134 CHAPTER VI Phytotoxi.vs and Zootoxixs 13S Phj'totoxins 138 Zootoxins 141 Snake Venoms 141 Scorpion Poison 150 Spider Poison 152 Centipedes 153 Bee Poison 153 Poisons of Dermal Glands of Toads and Salamanders 154 Poisonous Fish 156 Eel Serum 158 CHAPTER VII Chemistry of the Immunity Reactions — Antigens, Specificity, Anti- toxins, Agglutinins, Precipitins, Opsonins, and Related Sub- jects 159 Antigens 160 Xon-Protein Antigens 161 Specificity of Immune Reactions 165 Toxins and Antitoxins 172 Chemical Nature of Antitoxins 175 Agglutinins and Agglutination 178 Precipitins 184 Opsonins 188 The Meiostagmin Reaction 1S9 The Epiphanin Reaction 190 CHAPTER VIII Chemistry of the Immunity Reactions (Continued) — Anaphyl.yxis or Allergy, Abderhalden Reaction 191 Anaphylaxis or Allergy 191 Anaphylactogens 191 Anaphylatoxin 194 Anaphvlactin 201 Hay Fever " 203 Abderhalden Reaction 204 CHAPTER IX Chemistrv of thk Immunity Reactions {Continued) — BACTKKiuLV.-iis, Hemolysis, Complement Fixation, and Serum Cytotoxins . . 206 Serum Bacteriolysis 206 Cytotoxins . . " 209 Hemolysis or Erythrocytolysis 210 Hemolysis hy Known Cheinicnl iiiid Pliysical Agencn-s 211 Hemolysis hy Serum 214 Hcmolvsis t.v liaclcriji 219 CONTENTS 13 Paoe Hemolysis by Vegetable Poisons 220 Hemolysis by Venoms 223 Hemolysis in Disease 224 Complement Fixation and Wassermann Reactions 229 Cytolysis in General 233 CHAPTER X Chkmical Means of Defensk Against Non-Antigenic Poisons .... 237 Inorganc Poisons 240 Organic Poisons 241 CHAPTER XI I.NFLAMMATION 247 Ameboid Motion and Phagocytosis 248 C'hemotaxis 248 Chemotaxis of Leucocytes 250 Phagocytosis 255 Theories of Chemotaxis and Phagocytosis 259 Artificial Imitations of Ameboid Movement 260 Relation of the Above Experiments to the Phenomena Exhibited by Leu- cocytes in Inflammation 263 Suppuration 267 Composition of Pus 269 Sputum 273 CHAPTER XII The Chemistry of Growth and Repair 276 Proliferation and Regeneration 276 Chemical Basis of Growth and Repair 279 Vitamines or Food Hormones and Deficiencv Diseases 280 Beriberi • • • ' 283 Keratomalacia or Xerophthalmia 284 Nutritional Dropsy (War Dropsy or Famine Edema) 285 Scurvy 286 Pellagra 287 Rickets 288 CHAPTER XIII Distvrbance.s of Circulation and Diseases of the Blood 290 The Composition of the Blood 290 Hemorrhage 293 Hemophilia 298 Anemia and the Specific Anemias 301 Secondary Anemias 301 Chlorosis 303 Pernicious Anemia 305 Leukemia 309 Hyperemia 312 Active Hyperemia 312 Passive Hyperemia 313 Thrombosis 315 Fibrin Formation 315 The Formation of Thrombi 323 Embolism 325 Infarction 327 14 CONTENTS CHAPTER XIV Page Edema 330 Formation of Lymph 331 Absorption of Lymph 338 The Pathogenesis of Edema 340 Special Causes of Edema 348 Composition of Edematous Fluids 352 Varieties of Edematous Fluids 358 Chemistry of Pneumothorax 365 CHAPTER XV Retrogressive Changes (Necrosis, Gangrene, Rigor Mortis, Parenchy- matous Degeneration) 367 Necrosis 307 Causes of Necrosis 371 Varieties of Necrosis 382 Fat Necrosis 387 Gangrene 391 Rigor Mortis 392 Atrophy 395 Cloudy SwelUng 396 CHAPTER XVI Retrogressive Processes {Continued), Fatty. Amyloid, Hyaline, Colloid, AND Glycogenic Infiltration and Degeneration 400 Fatty Metamorphosis 400 Physiological Formation of Fat . *. 401 Pathological Fat Accumulation 401 Pathogenesis of Fatty Metamorphosis 409 Processes Related to Fatty Metamorphosis 412 Adipocere 412 Lipemia 415 Pathological Occurrence of Fatty Acids 418 Pathological Occurrence of Cholesterol 419 Amyloid Degeneration 421 The Origin of Amyloid 425 Hyaline Degeneration 427 Colloid Degeneration 429 Mucoid Degeneration 430 Glycogen in Pathological Processes 432 Physiological Occurrence 433 Pathological Occurrence 434 CHAPTER XVII Calcification, Concretions, and Incrustations 439 Calcification . 439 Occurrence of Pathological Calcification .... . 442 Chemistry of the Process of Calcification . 443 Osteomalacia . 447 Rickets 449 Concretions 452 Hillary Calculi ,453 Urinary Calculi . 459 Corpora Aiiiylacea , 464 Other Less Common Concretions . . . 465 Pneumonokoniosis 469 CONTENTS 15 CHAPTER XVIII Paob Pathological Pigmentation 471 Melanin 471 Lipochroine 478 Blood Pigments 481 Icterus 488 UrobiUn 495 CHAPTER XIX The Chemistry of Tumors 497 A. Chemistry of Tumors in General 499 B. Chemistry of Certain Specific Tumors 516 (1) Benign Tumors 516 (2) Malignant Tumors 522 Multiple Myelomas and Myelopathic "Albumosuria" 525 CHAPTER XX Pathological Conditions Due to, or Associated with, Abnormalities IN Metabolism, Including Autointoxication 530 Uremia . 532 Toxemias of Pregnancy • 540 Eclampsia 541 Acute Yellow Atrophy of the Liver 547 Chemical Changes of Acute Yellow Atrophy 550 Acid Intoxication and Acetonuria 555 Diabetic Coma 558 Acidosis and Acetonuria in Conditions Other than Diabetes ... 563 Alkalosis 566 Acetonuria Without Marked Acidosis 567 Fatigue 569 The Poisons Produced in Superficial Bums 571 CHAPTER XXI G astro-Intestinal "Autointoxication" and Related Metabolic Disturb- ances 574 I. The Constituents of the Digestive Fluids 575 II. Products of Normal Digestion 575 III. Products of Putrefaction and Fermentation 578 A. Protein Putrefaction 578 The Pressor Bases 584 Alkaptonuria 586 Cystine and Cystinuria 590 B. Products of Fermentation of Carbohydrates 592 C. Products of the Decomposition of Fats 592 Results of Gastro-intestinal Intoxication 593 Acute Intestinal Obstruction 596 CHAPTER XXII Chemical Pathology of the Ductless Glands 597 Diseases of the Thyroid 597 The Functions of the Thyroid 597 Chemistry of the Thyroid 599 Chemistrj^ of Goiter 604 Myxedema and Cretinism 607 16 CONTENTS Page Exophthalmic Goiter 609 The Parathyroids 613 The Adrenals and Addison's Disease 615 Addison's Disease 620 The Hypophysis and Acromegaly 621 Thymus and Other Ductless Glands . 624 CHAPTER XXII 1 Uric-Acid Metabolism and Gout 626 The Chemistry of Uric Acid 626 Formation of Uric Acid 628 Destruction of Uric Acid 632 The Occurrence of Uric Acid in the Blood, Tissues and Urine 634 Gout 636 Uric-acid Infarcts 640 CHAPTER XXIV Diabetes 642 Carbohydrate Physiology 644 The Blood Sugar 649 The State of the Sugar in the Blood 650 Diose 652 Trioses 653 Tetroses 656 Pentoses 656 Chronic Pentosuria 657 Hexoses 657 Galactose 661 Levulose (Fructose) 662 Polysaccharides 663 Glycosurias 664 Phlorhizin Diabetes 666 Pancreas Diabetes and Diabetes Mellitus 671 Index 679 CHEMICAL PAIIIOLOUY CHAPTER I INTRODUCTION THE CHEMISTRY AND PHYSICS OF THE CELL Since Virchow founded modern pathology the unit of all anatom- ical considerations of disease has been the cell, and in physiology the same unit has been found equally useful. When either physiological or pathological processes are studied from a chemical standpoint, the cell is still found occupying nearly as fundamental a position, for ue can seldom go back to molecules and atoms in investigating biological problems. Although we know that within each cell are many different chemical substances, and that numerous different enzymes and other agencies are exerting their influences upon them, yet we find that the reactions are all profoundly affected by the environment in which they occur, and it is the structure of the cell that determines the environment of its chemical constituents. All chemical reactions are modified by physical influences, and an enzyme may have quite a different effect upon a substance when it acts in a test-tube from what it will have when in a living cell, whose structure permits the diffusion of one substance while preventing that of another, and where countless other substances and enzymes may participate in the changes. The cell is the structural unit of the living organism, and as l\v its physical properties it modifies chemical processes, so it becomes practically the unit in physiological and pathological chemistry. All consideration of the chemistry of disease must thus refer back to the chemistry and physics of the normal cell, and on this account a brief resume of these subjects may serve as a fitting introduction to the strictly pathological matters to follow.^ As applied to the animal tissues, the term "cell" is entirelj' a mis- nomer, for it describes accurately only such forms of "cells" as are ' Of necessity, only so much of the very extensive literature on cell structure and cell chemistry can be considered as will have direct bearing upon the subject matter to follow, referring the reader for more detailed information to such works as Wilson's "The Cell in Development aad Inheritance;" Mathews" "Physiological Chemistry;" Hammarsten's "'Physiological Chemistry;" Gurwitsch's " Morpho- logic und Biologic der Zelle;" Hober's " Physikalische Chemie der Zelle und der Gevvebe;" Hamburger's "Osmotischer Druck und lonenlehre;" Loeb's "Dynamics of Living Matter;" Oppenheimer's "Handbuch der Biochemie;" and Bottazzi, "Handbuch der vergl. Physiologic," Vol. I, for general discussion, and to the most important monographs for treatment of special points. 2 17 ■18 • '-' • tk^'CHEMISTRY AND PHYSICS OF THE CELL ■found in ■ plants,' iti which the prominent feature is the hmiting wall, forming a cell to enclose a fluid content. In most instances the "cell" answers better to the definition, "a mass of protoplasm;" but usage makes language, and no possible confusion can arise from the prevailing universal use of the original term, except, perhaps, that the term is prone to carry with it the thought of the walls of the cell being much more prominent than they really are. This is not so unfortunate a result, perhaps, for, as we shall see later, the limiting surfaces of the cell, even when too thin to be readily demonstrable, may plaj^ a much more important part in cell chemistry than their appearance indicates. The morphological division of the cell into cell wall, cytoplasm, nucleus, and nucleolus can hardly be followed out chemically, for if we surmount to some extent the difficulties in the way of studying the different portions separately, we find that the differences between them are rather quantitative than qualitative. And, furthermore, how- ever different the cells of one organ or tissue may appear from those of another organ or tissue under the microscope, when analj^zed by the chemical methods at present at our cUsposal we find the differences very slight indeed. Certain substances are found in every living cell, and in quantities usually not greatly dissimilar; hence they are as- sumed to be the most important constituents of protoplasm, and are sometimes called the primary constituents of the cell. Manj^ other secondary constituents may also be present, some of which are so nearly universal that we are not sure but that they really are primary cell components; such are fat and glycogen. Others are characteris- tics of certain cells, such as melanin and keratin, or specific products of cell metabolism, such as mucin and the specific enzymes. The great histological and chemical differences existing between different tis- sues depend often on these secondary products, as in fat tissue and squamous epithelium; or upon the intercellular substance, as with connective tissue, cartilage, bone, etc., which may be looked upon as products of cell activity. Protoplasm, as the term is generally used, includes the various primary constituents with the fluids permeating or dissolving them, but does not include the more conspicuous secondary constituents, such as fat droplets, pigment granules, etc., nor the cell membrane when such exists. Evidently it is a vcrj'' indefinite term, to be avoided as much as possible, particularly because of the confu.sion as to whether it includes the nucleus or not, different authors differing in this respect in their usage of the word. CHEMISTRY OF THE ESSENTIAL CELL CONSTITUENTS To enumerate the jirimary or essential constituents of the cell ab- solutely is not possible, for the rapid advances in chemistry may alter all classifications without warning, but i)i;u'tica11y th(\v may be CHEMISTRY OF riiOTEINS 10 grouped iiiulcr tho h('a(liii|!;.s of i)rotc'iiis, lipins, carbohydrates, salts, and water, and no attempt will i)o made to give here more than the most essential features concerning each. Proteins - In the last few years we have obtained something approaching a scientific uuderstandinp; of the chemical nature of this great group of the most highly com- ])lcx bodies known to chemistry. Our information has been obtained almost e.\- clusively tiu-ough studies of the products obtained by splitting up the protein molecule, for as yet relatively little has been accomplished through synthesis. Proteins can be decomposed by the action upon them of acids or alkalies in various concentrations, by sui)erhcatcd steam, by digestive ferments, and by bacteria. The products obtained in these different waj's arc not all the same, for some sub- stances may be formed by oxidation, reduction, decomposition, combination, or condensation of the various products of simple cleavage, and it is necessary to distinguish between the primary cleavage products (those which exist as radicals within the molecule) and the secondary products (those not existing preformed in the molecule but formed by transformation of the primary products). This can usually be done, and it is found that so far as the primary products are concerned, it makes little difference which method of cleavage (or hydrolysis, since in the si)litting, water is combined with the organic substances) is used. At first the proteins split up into compounds still possessing many of the fea- tures of the typical protein molecule, such as albumoses and peptones, and these bodies are then further resolved into simpler substances, which are not aggregates of several smaller molecules as are the proteins, and which can be obtained in pure crystalline form. No matter which method is used we find the process going through these stages, and, as before mentioned, the primary crj'stalline products obtained are practically the same quantitatively as well as qualitatively. Some methods, e. g., bacterial decomposition, however, lead in the end to more profound or different decomposition of the cleavage products into secondary substances. The similarity of the results obtained in these different ways indicates that there are definite lines of cleavage in the protein molecule along which separation takes place, independent of the nature of the agency at work, and that the substances obtained represent the "building stones" of the entire molecide. These substances all have in common one important point: each one is an acid, which has a NH2 group substituted for a hydrogen atom on the carbon nearest the acid radical (the a-position). It makes no difference what the rest of the radicals are, whether they are simple chains (leucine), or members of the cyclic or aromatic series (tyrosine), or sulphur-containing bodies (cystine), without exception this relation of a NH2 group to an acid radical is constant, as in this formula: NH2 / R— CH— COOH. Through this arrangement every one of the constituents of the protein mole- cule is provided with a group with a strong basic character arid a group with a strong acid character, and hence it is possible for them to unite with one another in indefinite numbers, and, because of the great variety of individuals, in practi- cally an infinite number of combinations. It is believed that it is in just this way that the protein molecule is built up. By artificially uniting various cleavage products Emil Fischer succeeded in producing large molecules made up of several amino-acid radicals (called by him "polypeptids")^ which show some of the char- acteristics of the peptones, and this is the nearest that investigators have yet come to synthesizing a protein molecule. The union is accomplished by the splitting off of water, corresponding to the addition of water that occurs when the protein - For the complete literature of this subject see "The Chemical Constitution of the Proteins," PHmmer, London, 1917; "The General Character of the Proteins," Schryver, London, 1912; "The Vegetable Proteins," Osborne, London, 1910 (all in the series of "Monographs on Biochemistrv"). Also "The Physical Chemistry of the Proteins," T. B. Robertson, New York, 1918. 3 Reviewed by Fischer, in Ber. deut. Chem. Gesell., 1906 (39), 530. 20 THE CHEMISTRY AND PHYSICS OF THE CELL molecule undergoes cleavage. It ms^y be illustratetl Ijy showing tlie formation of the simplest polypeptid, (jlycy I glycine. N'H2 0 XH2 O i CH2 — C— I OH + H I HN — CH2— COOH = CH: — C — HN — CH>— COOH + HjO. (glycine) (glycine) (glycylglycine) For these reasons it is believed that tlie protein molec^ile consists of great numbers of ami no-acid groups, combined with one another through their basic and acid radicals, and that the various proteins are different from one another because they contain different numbers or varieties or orders of combination of amino-acids. For example, the globin of hemoglobin yields no glycine on hydrolysis, while gelatin yields IG.o per cent. On the other hand, gelatin is free from tyrosine. Some of the prolamins (proteins obtained chiefly from spermatozoa) yield as high as 58 to 84 per cent, of arginine, while the simpler amino-acids with but one N (mono- amino-acids) are scanty, and most varieties are lacking. It will be noticed that when two amino-acids unite, as seen in the formation of glycylglycine, an acid radical and a basic radical are still left free. In this may be seen the e.xplanation of the peculiar amphoteric nature of proteins. As long a? these two groups are free the proteins can combine with either acids or bases, a? they are well known to do, and hence they react as either acids or bases under dif- ferent conditions. It must not be imagined that the structiu-e of the complete molecule is simply a long straight chain of amino-acids joined only in tlie same way as are the com- ponents of glycylglycine. The existence of the diamiuo-acids. of the benzene rings, of hydroxyl groups, (as in serine or t\'rosine), of ring compounds, (as pyrroli- dine carboxylic acid), of substances with two acid groups, (as glutaminic and aspartic acid), adds complications to the formation until it is impossible to estimate just how all the various building stones may be arranged. We must bear in mind the size of the protein molecule, which Hofmeister has estimated (for serum albumin) as having a molecular weight of 10,1GG, and for hemoglobin the molecular weight has been estimated at 16,669. Within such a "giant molecule" there is room for variety almost beyond comixitation. The Proteins of the Cell. — By physiological chemists proteins are clas.sified into simple proteins, of which egg and serum albumin are types; and compound proteins, which are characterized by having some special non-protein group which can be split off, leaving behind a characteristic protein residue, e. g., nucleo-proteins, glyco-proteins. As primary cell constituents the following varieties of proteins may be mentioned; albumin, globulin, nucleo-protein, nucleo-albiunin or phospho-protein, and insoluble pi'oteins. At one time it was thought that cytoplasm consisted chiefly of albumin, like white of egg. but we now know that this forms but a small part of the cell proteins, often occurring only as traces. It is held by some that true albumin OcciH-s only as a building or intermediate cleavage product of the more complicated forms of cellular proteins, and is itself of relatively slight importance in cell life, not participating in cliemical changes except as a food-stuff. Albumins arc characterized chiefly by tiieir greater solubility in water, and in being less easily j)rccipitated tluin most proteins. Th(>y seem to be a fundnmcntal type of proteins. The tlircc forms of albumin that ha\(' ix'cii dc^i-rilu'd in animal tissues or jjroducts arc egg-albumin, lactalbumiu of milk, aiul scrum albumin; pro- hal>ly cell albumin is most closely related to the last, and what has been described as cell albumin is perhaps in many cases but .serum albumin that has been imper- fect I v icmoved. CIII'MlSriy'Y OF I'h'oTKIXS ' 21 Globulins also oocur in nil cells, hut in small amounts in most animal ceils except llie muscles, whose cliief proteins helonp; to tliis or a closely related Kroup. Tiie globulins are quite similar to the albumins, so that there is really no sliarp line between the two j!;roups. Their insolubility in water separates them from albu- mins, and their solubility in dilute lunitral salt solutions from the more complex proteins. An important feature of tli(> f^loljulins is the low temperature at which they coagulate — some so low that IIallil)urton' believes it possible that they may be coagulated within the cells during high fevers. Hammarsten has long maintained that simple proteins form a relatively in- significant part of the cytoplasm, in opjiosition tf) the once-prevalent view that the nucleo-proteins were limited to tlie nucleus, and that the cytoplasm was chiefly albumin and globulin. 'Plie general trend of opinion as influenced by the result.s of researches has been favorable to his contentions, and we shall jjrobably not be far wTong in accepting his statement that — "The chief ma.ss of the protein sub- stances of the cells does not consist of proteins in tlie ordinary sen.se, but consists of more complex phospliorized l^odies, and that the globulins and albumins are to be considered as nutritive materials for the cells or as destructive products in the chemical transformation of the protoplasm."^ Nucleo-proteins are considered to be the most important constituents of the cell, botli in quantity and in relation to cell activity. In structure the nucleo- proteins are very complex, as indicated by the difTerent products yielded on hydro- lytic cleavage of the molecule. Furthermore, there are many varieties, depending both upon the nature and proportions of the component parts. They may be described as consisting of two primary constituents — (1) nucleic acid and (2) a protein body, in chemical combination with each other like a salt. The term itiicleic acid covers a large group of substances, which are characterized, on the one hand, by their frequent occurrence bound with proteins, and, on the other hand, by their yielding phosphoric acid and purine bases, pyrimidines and pentoses or hexoses on cleavage. Diagranimatically the manner of cleavage of the nucleo- proteins may be indicated as follows: * Nucleo-protein . /\ nuclein^ protein nucleic acid protein phosphoric acid purine bases, pyrimidines and carbohydrates. In the cell the nucleo-proteins probably exist partly as solid structures, c. (j., the chromatin framework of the nucleus, and partly dissolved in the plasma. An interesting phenomenon is the alteration in the chromatin nucleo-proteins during cell division, when they seem to lose part of the combined protein and approach more. nearly pure nucleic acid — just as inorganic salts occur with the acids and bases saturating each other more or less incompletely, e. g., mono-, di-, and tri- basic phosphates. In this we have a chemical explanation of the intensity of the staining of dividing nuclei by basic dyes." Phosphoproteins resemble nucleo-proteins to the extent that they also yield phosphoric acid, and are somewhat similar in solubility and digestibility. They are essentially different, however, in that they do not yield nucleic acid or purine bases on cleavage. Probably members of this group are also constant components of cells. Glycoproteins (or gluco-prnl(in>i) and phof^pho-c/lycnproteins are also believed to occur frequently or constantly in protoplasm. They are compounds of proteins with a sugar or sugar-like group, which probably usually contains nitrogen, thus differing from the ordinary hexoses and pentoses. ' Halliburton and jNIott, Archives of Neurology. 1903 (2), 727; also .see Halli- burton's "Chemistry of Muscle and Nerve." 'See Kossel, Munch, med. Woch.. 1911 (58), 65. * Probably nuclein should be considered as merely one variety of nudeoprotcin, with less protein than the other varieties. ' The chemistry of the nucleo-proteins is discussed in the chapter on Trie Acid Metabolism and Gout, Chap, xxiii. 22 THE CHEMISTRY AND PHYSICS OF THE CELL Insoluble proteins, or bodies resembling the coagulated proteins in their lack of solubility in various fluids, are left behind after the other proteins have been ex- tracted from the cells. Their significance is not known : whether to a large extent artificially produced or whether a normal structural element of the cell. Fats and Lipoids (Lipins)^ Lipoids is a term in common use but of indefinite significance; most usually it comprehends the intracellular substances which are soluble in ordinary fat solvents, but which are not simple fats or fatty acids, lecithin and cholesterol being the most important of the lipoids. For the entire group of fats and lipoids the term lipins has been pro- posed by Gies and Rosenbloom.^ Lipoids and ordinary fats, that is, lipins, occur in all cells, but their demonstration is not always readily possible. The microscopic appearance of a cell, even when special stains for fat are used, gives no correct idea of the amount of lipins actually present. Thus normal kidneys contain 15 to 18 per cent, in their dry substance, but none of this can be detected readily with the microscope. A kidney which seems microscopically the site of marked fatty degeneration may show no more fat when examined chem- ically than a normal kidney, which in section appears to be quite free from fat. This is because some of the intracellular fat is so bound, chemically or physically, with the proteins that it cannot be seen, nor can it be stained by the dyes ordinarily used for that purpose; only when degenerative changes of certain kinds have liberated it from com- bination does it become visible and stainable bj^ ordinaiy methods (Rosenfeld). By the special fixation method of Ciaccio the fatty com- pounds of even normal cells may be made stainable (Bell),^° showing that the so-called masked fat is really in a not altogether invisible form. Whether the intracellular fat has any function other than that of serving as a food-stuff is not known, but there can be no question of the importance of the phosphorized fats, or phospholipins. Phospholipins are primary cell-constituents and are probably important both in metabolism and physically. Hammarsteu regards them as concerned in the ijuilding up of the nucleus. As will be shown later, manj- of the most essential physical properties of the living cell depend upon the presence in it of lipoids, of which phosphatids are apparently the chief. Of the ether-soluble substances in the heart, for example, 09 to 70 per cent, are phosphatids, 8 per cent, of the dry weight of the myocardium. Many different substances have been described as phosphatids, but the chemical identity of but few is sufficiently established. Of these the most imjjortant are lecithin and ccphalin, which are most intimately associated. There are several possible varieties of lecithin, depending upon the fatty * Full discussion in "Lecithin and Allied Substances (The Lipins)," by Hugh MacLean, Monographs on iJiociliemistry, London, 1918. " MacLean uses "li))in" to include "subslances of a fat-like nature yielding on hydrolysis fatty acids or derivatives of fatty acids and containing in their nu)lecule either N, or N and 1'. As there is need for a term covering the fats, phospliatids, cholesterols and related Ixxlies, the suggestion of C.ie.s .-uid lUi.senbloom is followed for tlie i)resent in tliis i)ook, and tin* word lipin used witli the broader significance. "> Jour. Med. lies., 1911 (1-9), SiW. FATS AND LI poms 23 acid radical they contain. The assumed structural formulaof one lecithin, stearyl- olcyl lecitliin, is as follows: CH2 — O — C18 — HssO I CH— O— Ci,— H,30 I CH2— 0— PO— OH I 0— CHo— CH2— N = (CHa),. OH It differs from ordinary fats, therefore, in having two special groups, one the phos- phoric acid, the other the choline radical, w^hich last may be of some importance in pathological processes. In its phj^sical properties it is quite similar to the ordinary fats, although it forms even finer emiilsions in water, which are practically colloidal solutions (W. Koch). Cephalin diiTers in having for the base amino-ethyl alcohol (XH0CH2CH2OH) instead of choline, and is probably as widely spread in the tissues as lecithin. 1^ It has been held by some that there are many phospholipins, which may be speci- fic for different cells, tissues and species, but it seems more probable that these supposed specific lipoidal substances are merely mixtures of lecithin, cephalin and their derivatives in varying proportions (Levene).^^ Cholesterol, which is another lipoid, is nearly as universally present as leci- thin, it exists both free and in combination ^\•ith fatt}' acids, for cholesterol is an alcohol and not at all similar to the fats chemically, although very similar physically. The empirical formula is C27H4SOH or C27H46OH, and it is related to the terpenes. It seems to be relatively inert chemically, and therefore is probably important only because of its effect on the phj'sical properties of the ceUs. By some it is considered to be a decomposition or cleavage product of the proteins, which is in accordance 'ndth its abundance in masses of old necrotic tissue, e. g., atheromatous masses, old infarcts, and old exudates. Doubly Refractive Lipoids and Myelins. '^ — In practically aU normal tissues there are present droplets of lipoid nature which are characterized bj' sho-ndng prominent crosses when examined with crossed Nicol prisms (anisotropic), the adrenal and corpus luteum containing them most abundantly. Chemically they seem to be mixtures of various lipoids in inconstant proportions, but probably the anisotropic character is most usuallj' dependent upon the presence of cholesterol esters. The term myelin was first appUed by Virchow to pecuHar fatty substances found in various normal and pathological tissues, because they showed physical characters similar to those of the myeUn substance of nerves, but as many of these substances are doubly refractive, or can be easily made so, some authors use the term myelin as if it were synonymous with doubly refractive Hpoids. There are, however, myelins which are not always doubly refractive, and also doubly refractive hpoids w^hich do not swell up in water to form myeUn figures, etc., as is characteristic of true myeUns. Chemically, however, the mj^ehns and doubly refractive substances are probably related, consisting of mixtures of cholesterol, cholesterol esters, phospholipins and perhaps soaps, in varying proportion. They will be considered further in discussing Fatty Metamorphosis, Chap. XVI. Carbohydrates The third great class of food-stuffs, the carbohydrates, is represented in the ceU by pentoses and hexoses combined with proteins and with lipoids, and also by glycogen, which exists free. Glycogen is a difficult substance to isolate in minute quantities and, therefore, although it is not found in all cells by our present methods, yet it may well be that it is a constant constituent of the protoplasm. There is no e\a- " Koch and Woods, Jour. Biol. Chem., 1905 (1), 203. 12 Jour. Biol. Chem., 1919 (39), S3. "See Adami, Join:. Amer. Med. Assoc, 1907 (48), 463; Karwdcka, Ziegler's Beitr., 1911 (50), 437; Schultze, Ergebnisse Pathol.,-1909 (13, pt. 2),»253; Bang, Ergebnisse Physiol., 1907 (6), 131; 1909 (8), 463. 21 THE CHEMISTRY AMJ PHYSICS OF THE CELL dence, however, that it is anything more than a source of heat and energy to the cell. Its properties and occurrence will be considered more fully in the discus- sion of glycogenic infiltration. Since glycogen is formed from dextrose and is constantly breaking down into dextrose, it is probable that the latter is also con- stantly present in the cells. Inorganic Substances Up to this point the substanf es of the cytoplasm that have been discussed have all been organic compounds which do not naturally exist independent from living or once living cells, yet the inorganic substances of the protoplasm are also of vital importance. As Mann says, "so-called pure ash-free proteins are chemically inert, and, in the true sense of the word, dead bodies. What puts life into them is the presence of electrolytes." The various salts of potassium, sodium, calcium, magnesium, and iron which all cells contain do not exist merely dissolved in the water of the cell, but in part they are comlnned with the organic ccmstituents of the protoplasm. They are not combined as simple additions of the salts to the proteins; but io)is, both anions and cations, are united in chemical combina- tion to the large protein molecule (ion-proteins)." Possibly the proteins partici- pate in vital chemical processes only as ion compounds with inorganic elements. It is extremely difficult, indeed almost impossible, to secure proteins entirely free from inorganic substances (ash-free proteins). The fact that inorganic substances are held in the cells cliemically rather than by simple diffusion into them from the surrounding fluids is shown by the great difference in the proportions of various salts in the cells and in the extra-cellular fluids. Thus potassium is nearly always much more abundant in the cells than in the tissue fluids, while sodium is more abundant in the fluids. Piiosphoric acid is also more abundant in the cells, and chlorin in the plasma. In cells iron seems to exist chiefly in combination with the nucleo-protcins. '^ THE PHYSICAL CHEMISTRY OF THE CELL AND ITS CONSTITUENTS'" From the standpoint of physical chemist ly the cell consists of a collection of colloids and crystalloids, electrolytes and non-electrolytes, dissolved in water, in lipoids, and in each other, surrounded by a semi- permeable membrane, and perhaps subdivided by similar membranes or surfaces. Physical chemical jjrocesses, as we shall see later, play an all-important part in the life phenomena of the cell, and there- fore some space may be occupied profitabl}^ in explaining the nature of these changes and of the .substances that participate in them. Crystalloids and their Properties Crystalloids, or substances that tend under favorable conditions to form crystals, and which diffuse readily through most diffusion membranes, form a relatively small part of the total mass of the cell, but they are fully as essential as the colloids. The chief representa- tives of this groui) that are found usually or constantly in the cell are the inorganic salts, sugar, and the innumerable decomposition products of the proteins, including particularly urea, creatine, purine bases, amino-acids, etc. Most of these are by no means so characteristic of living things as are the colloids, sometimes occurring (juite inde- "SeeJ. Loci), Science, 191't (SO), 4:{<). " See Macallum on Microchemistry, lMgei)nisse Piiysiol., 1908 (7), 552. "See Hayliss, "Principles of (Jciieral Physiology," London, 1915, for a more extensive discussion of these topics. CRYSTALWIDS 25 peiHlciif ly ot" a (•cUiilai' origin, which the proteins never do. Tlie inoi- ganic salts in i)arti('iihir seoni (luilc i'oi-eifj;n to H\inji processes, and as they enter and leave the body i)ractieally unchanged they are evidently not a source of energy through chemical change. Their importance to the cell lies almost entirely in their physical or physico- cliemical properties. The organic crystalloids, although of nutri- tional value, also have ])hysical jiropei-ties in some respects similar to those of the inorganic crystalloids, and therefore to this extent they exert similar influences, but the essential difference between the organic and the inorganic crystalloids is that all the latter are electrolytes, while many of the organic crystalloids that occur in cells are non- electrolytes. The importance of this distinction lies not in the utility or non-utility of these substances as conductors of electrical currents in the ordinary sense, but rather on the existence of those properties which determine their conductive al)ility. Electrical conductivity is an index of ionization, and upon ionization depends the chief influence of the electrolytes upon vital activities. The importance of this process of dissociation or ionization lies in the fact that with most substances no chemical reaction can occur while the substance is in the non-ionized state. The chemical properties of ionizable substances are produced largely by the ions they liberate on dissociation. As a consequence, the physiological effects of electrolytes are due to their ionic condition, and through the ions that are present in the cell many of its various chemical processes are brought about. Not all substances ionize with the same readiness, which causes a great difference in their properties. The reason that acetic acid is a weaker acid than hydro- chloric acid is that it does not ionize to such an extent, and so a cor- responding ciuantity does not introduce as large a number of hydrogen ions into a solution. Larger molecules, as a rule, ionize less than smal- ler ones of similar nature, e. g., stearic acid ionizes less than acetic acid and therefore is a weaker acid. Likewise the properties of a substance which depend upon its ions will be less marked when it is in a solvent that produces little ionization. For example, bichloride of mercury owes its antiseptic properties to the Hg ions that it sets free when in solution. It is well known that solutions of mercury, and for that matter most other antiseptics, are much less actively germicidal in alcohol than when in water, because their ionization is less in alcohol; and the germicidal properties decrease as the proportion of alcohol increases, until the germicidal effect of the mixture is no greater than that of alcohol alone in the same strength. If we had no electrolytes in the cell, electric charges could not be carried about in it, and hence chemical reactions could not occur. It is this fact that makes the inorganic salts of such vital importance to the cell life. To repeat Mann's words, it is the electrolytes that put life into the proteins. Water itself is almost absolutely non-disso- ciated, and proteins so little that for some time it was doubted if they 26 THE CHEMISTRY AND PHYSICS OF THE CELL really did ionize. Probably all soluble substances do dissociate to a certain minimal degree, but it is so slight for most of the constituents of the cell except the inorganic salts (the organic acids and alkalies, and a few dissociable organic products of protein metabolism, occur in such insignificant amounts as to be almost negligible) that without them there would be little chemical activit}^ possible, and hence life would be absent or at a very low ebb indeed. As before mentioned, the inorganic salts probably exist in the cell not only as salts, but also, and perhaps chiefly, as ions and ionic compounds with the cell proteins. Many applications of the facts and theories of ionization have been made in physiology and a few applications have also been made in pathology, especially the relation of ions to edema, to diuresis and glycosuria, and also to problems of immunity. No attempt will be made here to go further into the observations and theories concerning ionization or its role in physiology, but for more extensive information as well as for the complete bibliographj^ the works mentioned below may be referred to.^' The applications in pathology will be brought out as the subject under discussion in subsequent chapters necessitates and it is largely to facilitate the understanding of such reference that this brief summary of the subject of ionization has been introduced. In the same spirit we take up the subjects of diffusion and osmosis. Diffusion and Osmosis. — Although the non-electrolytes do not ionize to any considerable extent, and therefore are relatively inactive chemically, the crystalloidal non-electrolytes, of which sugar and urea are the two chief examples among the cell constituents, possess in common with the electrolytes the important property of diffusion. By this process the exchange of chemical substances between the blood and the cell is brought about, by it the chemical composition of the different parts of the cell and between different cells is equalized, and without it chemical change would be practically impossible. Dif- fusion occurs most simply between two solutions of unlike nature, or between a solution of a substance and the solvent alone, when placed directly in contact with one another. If we place in the bot- tom of a cylindrical vessel a solution of copper sulphate and above it some water, carefully avoiding mixing, it will be found after some time that the fluid has become equally blue throughout. This is brought about by the movement of the dissolved particles which gradually carries them through the entire mass of iluid, and as their migration is against the force of gravity, they evidently accomplish " "Physical Chemistry in the Service of Medicine," Wolfgang Pauli, transla- tion by M. H. Fischer, New York, 1907. " Physikalische Chemie dor Zelle und der Gewehe," Hober, Leipzig, 1915. "Osmotisohe Druck und lonenlehre in den medicinischen Wissenscliaften," Hamburger, Wiesbaden. "Studies in General Physiology," Loci), University of Chicago Press, 1905. "Dynamics of Living Matter," Locb, Columbia University Press, New York, 1906. Spiro and J. Loeb Oppenheimer's "Handbuch der Biochemie," 1908 (2), 1-141. "Physical Chem- istry of Vital Piienomena," McCIendon, Princeton Univ. Press, 1917. DIFFCSION AND OSMOSIS 27 work. 'Hiis process is not dependent upon ionization, for a s(jluti(ni of cane-sugar or of urea will show the same diffusion. A sohition of l)rotein or other colloid does so nuich more slowly, however, indeed, ([uite imperceptibly. If we were to introduce a piece of filter-paper between the water and the copper sulphate solution, the diffusion would go on the same, the pores of the paper permitting the passage of the molecules with- out hindrance. If, instead of filter-paper, there were introduced a sheet of some substance free from pores, the diffusion would be much more affected. If the septum was of such a nature that the sub- stances in solution were insoluble in it (e. g., glass), diffusion would of necessity stop; but if it were something in which the solvent or the solute was soluble, such as a gelatin plate, then these substances would dissolve in it, and diffusing through its substance escape into the fluid on the other side. The last example indicates the conditions afforded in the animal cell, and also in the usual laboratory diffusion experiments when the membrane is generally either an animal mem- brane or a parchment paper, both of which are composed of colloids. Crj^stalloids are generally soluble in colloids and hence pass through such diffusion membranes; colloids dissolve but slightlj' in colloids, and hence they do not pass through a diffusion membrane readily, and are, therefore, but very slightly diffusible. The process of diffusion, if uninterrupted, always continues until the solution is of exactly the same composition throughout. If on one side of the diffusion membrane there is a substance that passes through the membrane rapidly, and on the other a substance that passes through slowlj' or not at all, there will soon be an unequal condition on the two sides of the membrane, for the diffusible substance would ac- cumulate in equal amounts on each side, while the non-diffusible would remain where it was. On one side there would then be more material exerting osmotic pressure than on the other, and if the mem- brane were flexible, it would bulge toward the opposite side. The pressure is supposed to be due to the bombardment of the containing walls by molecules or ions of the substances in solution, and hence the more molecules and ions in solution, the more pressure. When equal numbers of particles are on each side of the partition, the pressure is equalized. It is quite possible to have membranes readily permeable to one substance and almost entirelj^ impermeable to another; such membranes are called semipermeable. To produce osmotic pressure it is not necessary that the membrane be absolutely impermeable to any of the substances — it may only be relatively less permeable for the solute than for the solvent. If, for example, we fill a parchment bag with concentrated sugar solution, tie up the top tightly and throw it into water, it will swell up rapidly and eventually burst. But if the parchment is in the form of a tube, open at the top, and the lower end is placed in water, the amount of fluid inside the tube will in- 28 THE CHEMISTRY AND PHYSIOS OF THE CELL crease at first, but eventually the sugar will diifuse out to such an extent that the solution is of the same concentration inside and out- side of the tube, and the column of fluid will again become of equal height on both sides. These results indicate that the water passes through the membrane more rapidly than does the sugar, but that eventuall}^ tlu; sugar can all pass through. Exactly similar conditions exist in cells, particularly plant cells. The typical cell of plant tissue consists of a distinct wall, usually cellulose, lined internally by a layer of protoplasm which incloses a mass of aqueous solution, the cell sap. containing sugar and various other solutes. The outer wall is readil}- permeable b}' water and by most solutes, whereas the protoplasmic layer inside it behaves like a semipermeable membrane, which permits water to pass through readily but hinders greatly the passage of most solutes; that it is somewhat permeable is attested by the fact that the cell sap contains solutes derived from the external fluids. As a result of this arrange- ment there is a constant tendency for the cavity of the cell to be distended by water and for the solutes within it to exert their con- siderable pressure upon the cell wall. Because of the strength of the cellulose layer the cell can withstand great pressures that Avould tear apart the tender protoplasmic layer that really determines the osmotic pressure that causes the rigidity or turgor of plant (-('lis, and explains the ability of a tender green shoot to hold itself up- right or horizontal in the air; and it is the force that enables growing roots to lift great stones or tear apart rocks in whose clefts they grow. If certain plant cells are placed in distilled water, the pressure may rise to such an extent that the cells burst, and it was through studying this phenomenon that Pfeffer worked out the basis of our pn^sent knowledge of osmotic pressure. If the cell is placed in a solution of greater concentration than its cell sap, the pressure outside will be greater than that inside and the protoplasmic meml>rane will be forced away from the cellulose wall, while its central cavity shrinks and i)erhaps disappears entirely, the protoplasm forming a ball in the center. This is practically what occurs when a plant stem is cut and it "wilts" — the water is removed by evaporation, the osmotic pressure outside the cells becomes greater than that inside, and the water passes out. Likewise when a plant c(>ll dies the turgor is lost because the membrane becomes ))ermeable, and so pi'(>ssure soon be- conuis the sanu^ on both sides of the cell wall. In animal cells the wall is not so highly developed as in plants, nor is it backed up b}^ a rigid material like celluIos(>; indeed for many animal cells there is no well-defined wall and the piotojilasm appears to be naked. Nevertlu^le.ss the behavior of tlu> animal cells indicates that they do possess what resiMubles a ('(>11 wall, in that they behave when in solutions as if they wei-e surrounded by a dif- fusion membi'aTie. The degi-ee to which piieiioineii;! ot' this iiatui'e DIFFISIOX AM) OSMOSIS 29 aic sluiwn vniic's witli (liflciciit cells; with red coiijusclcs. I'oi- example, the osmotic pressure influences arc very marked, as shown by the wrinkling or crenation of the corpuscles when they are placed in fluitls of higher concentration than the blood plasma, and by their swelling and disintegration with escape of the hemoglobin {hemoly- sin) when they are put into distilled water or solutions of less con- centration than the plasma. Other tissue cells seem to undergo more or less alteration from changes in the osmotic pressure in the fluids surrounding tiieni. The dift"usion membrane that surrounds the cell is generally nut well defincul, and for most cells seems to be but a surface condensation of the protoplasm, perhaps formed through the effects of surface tension. It seems probable that this surface dif- fusion membrane contains a large proportion of cell lipoids, i.e., cholesterol and phospholipins (for the red corpuscles this is practically certain); hence substances soluble in lipoids penetrate the cell read- ily, while to many substances insoluble in lipoids the cell is nearly or quite impermeable (Overton). Probably the wall of the animal cell is not so nearly semipermeable as is that of the plant cell, for nowhere in the animal bod}- do we get such turgor in the cells as we see in plant tissues. Lacking a cellulose wall, animal cells could not develop such an internal pressure without rupturing and such a process of rupturing {plasmorrhexis, plasmoptysis) does not seem to be a normal occurrence in animal tissues. We shall be most nearly correct, probably, if we look upon the animal cell as possess- ing a delicate diffusion membrane at its surface, through which water passes more readil}^ than do most crj'stalloids, and through which colloids pass almost not at all, but the exclusion of each of these types of substances is merel}^ relative and not absolute. AVithin the cell, also, the colloids probably exist as a more or less well-developed emulsion, so that we have here a practically limitless amount of surface formation all through the protoplasm; such a structure could permit the endless number of reactions of a living cell to go on side by side in the same cell. Stuches b}' G. L. Kite^^ seem to show that all of the protoplasm has much the same relation to solutions as does the external layer or cell membrane, for he found that if drops of solutions which can penetrate a cell from outside be injected directly into a cell they diffuse through it, but substances which cannot pene- trate from outside are also unable to diffuse through the cell after they have been injected into it. Since osmotic pressure, exactly like gas pressure, is presumably produced by the bombarding of the w^alls of the container by parti- cles in the solution, the amount of pressure will vary in proportion to the number of particles present. With non-electroh^es, such as sugar and urea, the moving particles seem to be mole- cules, and so a solution of sugar or urea will produce an osmotic 's Amer. Jour. Physiol., 1915 (37), 282. 30 THE CHEMISTRY AND PHYSICS OF THE CELL pressure directly proportional to the number of molecules it con- tains. In the case of the electrolytes, however, the ions produce pressure as well as the molecules, and hence an electrolyte in solution will produce a relatively high osmotic pressure as compared with an equivalent solution of a non-electrolyte, since each molecule may jaeld two or more ions. Colloids, however, exert so slight an osmotic pressure that it is difficult of detection; this probably depends on the great size and slight motility of their molecules. In the many and important osmotic processes of the animal organism, therefore, the colloids take no part except in helping to form the diffusion mem- brane, and in preventing the diffusion of one another. ^^ It is interest- ing to consider also that colloids under ordinary conditions do not greatly modify the diffusion of crystalloids through a solution con- taining both classes of matter. The fact that a cell is full of dis- solved colloids does not seriously affect the osmotic properties of the intracellular crystalloids, provided the colloids are not condensed in such a way as to form diffusion membranes. But as all the cleav- age products of proteins after they have passed the peptone stage are crystalloids, by decomposition of the intracellular proteins the os- motic pressure may be greatly raised. As long as the cell is living there can be no constancy in composition, for metabolic processes, by producing from proteins that have no osmotic pressure crystal- loidal substances that do have osmotic pressure, cause intracellular osmotic conditions to be continually varying. As a result, streams of diffusing particles are moving about in every direction, setting up new chemical reactions and consequent new osmotic currents. The greater the difference in osmotic pressure between a cell and its environment, and between the different parts of the same cell, the more powerful the osmotic effects, and as a result the greater the capacity for accomplishing work. Indeed, we may look upon cell life as a constant attempt at the establishment of equilibrium, both chemical and osmotic, ichich is nether achieved because the move towards one sort of equilibrium is always against the other. All the food-stuffs — -fats, carbohjairates and pro- teins— are characterized by being colloids when intact and crystalloids when disintegrated, thus: colloidal proteins «=^ crystjilloidal aiiiino acids colloidal f!;lycoG;en <=i crystalloidal siijiar nondilTiisil)Ie fats ^ dilTusihlo soaps and {glycerol. In consequence of this, if the crystalloids difTuse from the blood into a cell there is at once an excess of this end of the equation, and, hastened by the intraccillular enzymes, partial syntliesis to the colloid soon '" Under experimental conditions it is found that the nature of the nienibrane greatly modifies the osmotic jiressure; for if a ^iven colloid is soluV^le in a cer- tain mend)rane and a certain crystalloid is not, the colloid will diffuse through the membrane wliile the crystalloid is held back. (Kaldenberj!;, ,Iour. Physical Chem., 1900 (10), 111.) COLLOIDS 31 occurs to establish chemical ('quili])riuin. Chemical changes in the crystalloids, by oxidation, nuhiclion or hych'olysis, upset this chem- ical ecjuihbrium, and hence further diffusion, synthesis and hydrolysis continue, one upsetting the other continuously. If equilibrium were established we should have no further reactions, and the cells would be inactive. The constant upsetting of the equUihrium is what con- stitutes cell life. The relation of osmotic pressure and osmosis to physiological prob- lems is only beginning to be studied. It is apparent that they must be of essential importance in absorption from the ahmcntary canal, in absorption and oxciction betAveen the cells and the blood stream, and in secretion by glandular organs; but it is also certain that they are no less important in all the less obvious chemical and physical processes of the cell. 2° In pathological processes osmotic pressure may play an equally important role, and the facts discussed in the prececUng paragraphs will be alluded to frequently in subsequent chapters. COLLOIDS-i Since Graham in 1861 studied the differences between the sub- stances that chd or did not diffuse reachly through animal or parch- ment membranes, soluble substances have been classified in the two main groups of colloids and crystalloids, which distinction Graham believed separated two entirely different classes of matter. Although at the present time the differences between the two classes do not seem so great, yet the same division is found useful in classification. By colloids Graham indicated those substances which were dissolved to the extent of showing no visible particles in suspension, but which either did not pass through diffusion membranes at all, or did so very slowly indeed, as compared to the crystalloid substances. Under cer- tain conditions they tended to assume a sticky, glue-like nature, hence the name. (Many substances are now known which have the chief properties of the colloids and are therefore classified among them, but never are glue-hke, e. g., the colloidal metals, so that the name has lost some of its original significance.) The phj^sical prop- erty which Graham particularly noted in the colloids, besides their non-difTusibility was the tendency to assume various states of solidity. 2° For further consideration of the subject of osmotic pressure in these rela- tions, see: Livingston, "The Role of Diffusion and Osmotic Pressure in Plants," University of Chicago Press, Chicago, 1903; Czapek, "Biochemie der Pflanzen," Jena. Also, Spiro, Pauli and Hober, all previously cited. ^' For full discussions of the nature of colloids see: Hober, " Physikalische Chemie der Zelle," Leipzig, 1914; Pauli, Ergebnisse der Physiologie, 1907 (6), 105; Bechhold, "Colloids in Biology and Medicine," translated by J. G. M. Bullowa, 1919; Wo. Ostwald, "Grundriss der Kolloidchemie," and "Theoretical and Applied Colloid Chemistry," both translated by M. H. Fischer. A go9d brief discussion of colloids is given by Young in Zinsser's "Infection and Resis- tance." 32 THE CHEMISTRY AND PHYSICS OF THE CELL Not only can they be in solution, when he called them "sols" (when the solvent was water, " hydrosols ") , but they can become quite firfn although containing much water (then called "gels" or "hydrogels"). The gels may assume a firm, coagulated condition, the so-called "pec- tous" state, which state is permanent in that the gel form cannot be reobtained from the pectous modification. Finally the colloid can be in a dry, solid state, quite free from water, and then not a gel at all. Included in the great class of colloids are all forms of proteins, and also gums, starch, dextrin, glycogen, tannin, probably the en- zymes, and also the greater number of organic dyes; also there are in- organic colloids, such as silicic acid, arsenic sulphide, hydrated oxide of iron, and many other similar compounds, besides the elements themselves, especially the noble metals, which may exist in colloidal form. It will be seen at once that the chief constituents of the cells, in fact nearly all the primary constituents except the inorganic salts, are organic colloids, and therefore the properties of the cells are largely dependent upon the properties of the colloids. In considering the characteristics of the colloids we at once meet the question — What distinguishes the colloids from the crystalloids, on the one side, and from suspensions or emulsions on the other? The sum and substance of our present conception of the natiu-e of colloidal solution may be brieflj'' sunmiarized as follows: It is possible for solid substances to be so divided among the par- ticles of a solvent that they remain permanentlj^ in this condition, • neither aggregating into masses nor separating out through the action of gravity. With some substances, as sugar, for example, the solid seems to divide up into its molecular form, each molecule being free from all others of its kind except during occasional contacts. Some other substances, as salt, go still further, and the molecule cUvides into two or more parts, which have different electric charges {ionization). The first of these classes of substances forms a solution which contains no particles visible by any known means, does not contain particles large enough to reflect light imioingiiig upon them, exerts a large osmotic pressure, but does not conduct electricity. The other, in which ionization has occurred, differs solely in its capacity to conduct elec- tricity readily. Both are true solutions of crystalloids; the one which does not ionize is a n.on-electroh/tc; the other, by virtue of its ionization, is an electrolyte, the ions carrj'ing electric charges through the solution. At the other end of the scale we have substances which are quite insoluble when in masses, but which, when very finely divided by me- chanical means, can l)e sus])ended and uniformly distributiul through a fluid without having any maiked tendency to aggregate or settles out. Such suspensions or emulsions contain particles visible under the microscope, usually appe;ii Imhid, refract light, are non-diffusible, exert no osmotic pressure, and do not transmit electricity. .Such mixtur(!s are obviously veiy dilTca'cnt from the true solutions above COLLOIDS 33 described. Between these two extremes stand the colloids, which vary- in tlieir properties so that they approach sometimes the suspensions (e. g., lecithin, or coagulated egg-albumin in colloidal suspension), and sometimes more nearly the true solutions (e. g., dextrin). No sharp boundaries can be drawn between any of the members of the series. Indeed, one substance may present all the different stages under different conditions, some agreeing with the properties of the typical suspensions, and some with the properties of the true solutions. The colloids stand in an intermediary position, differing quantitatively in one way or another from the true solutions, but yet approaching them closely and sometimes almost indistinguishably resembling them. For the most part, however, they show characteristics decided enough to entitle them to separate classification, and to make any confusion with the crystalloids impossible. The Characteristics of Colloids. — The chief properties of the colloids 'are, then, as follows: - s^ Amorphous Form. — This, like almost all other "colloidal properties," is not absolute, for in egg-albumin, hemoglobin, and various globuUns we have proteins which in every respect are typical colloids, yet they form crystals readily and abundantly. Oxyhemoglobin, the molecular weight of which is calculated at about 14,000. exhibits Tyndall's phenomenon and will not pass through a very fine porcelain filter, and therefore resembles the colloids decidedly, yet it forms beautiful crystals. The very fact that crystals are formed, Spiro points out, is proof that when in solution the individual molecules must have been free and separate, for other\\dse they could scarcely unite in the definite spatial relations necessary to produce crystalline forms. With these few exceptions, however, the colloids do not present any typical structure, and are not crystalUne under anj^ visible condi- tion. But when they are made insoluble by chemical means they may, under certain conditions, produce rather characteristic non-crystalline structures, a matter that ^\•ill be discussed in a subsequent paragraph. Solubility. — Although we speak of "colloidal solutions," this terra does not commit us to the theory of the identity of the solution of colloids ^^^th that of crystalloids. We have above stated wliat seems to be a fair view of the matter as shown by many methods of experimentation. Most colloids seem to be, in fact, suspensions of masses of molecules, or perhaps of very large single molecules, and a true solution is Likewise a suspension of single molecules or of ions. When the aggregations of molecules are sufficiently large, we have an ordinarj^ sus- pension; but a single protein molecule is as large as a very great number of mole- cules of such substances as sugar (crystalloid); or tannin, C14H10O9 (colloid); or calcium carbonate (insoluble, suspension); and it would be strange if a true solution of a protein did not behave in many particulars Uke a suspension of mo- lecular aggregates of dimensions similar to the dimensions of protein molecules. Nearly all colloidal solutions show Tj^ndall's phenomenon, \yhich demonstrates the existence of particles in saspension large enough to reflect Ught from their sur- faces." Most of the colloids are held back by very fine filters to a greater or less degree; some are almost entirely retained by a hardened paper filter, while others pass through the finest-pored clay filters.' Furthermore, the metallic colloids, such as those of platinum, gold, and silver, are unquestionably suspensions of finely di\dded particles of metal, yet they exhibit all the typical phenomena of colloids, passing through many sorts of filters, and even inducing the same hydro- lytic changes as many enzymes. " It is highly probable, however, that Tyndall's phenomenon when exhibited by true colloidal solutions (e. g., soluble proteins), depends on the presence of aggre- gates and not properly on the dissolved colloids. (See McClendon and Prender- gast. Jour. Biol. Chem., 1919 (38), 549.) 3 34 THE CHEMISTRY AND PHYSICS OF THE CELL It must also be mentioned that the solvent is probably an important factor in determining the colloidal or non-colloidal nature of a substance; e. g., soaps form true solutions in alcohol and colloidal solutions in water; gelatin forms colloidal solutions in water but not in ether, whereas rubber forms colloidal solutions in ether but not in water. Closely related to solubility is the phenomenon of imbibition, which may be defined as the taking up of a fluid by a solid body wathout chemical change. Not all colloids possess this property, but it is shown by most of the organic colloids, particularly the proteins. Fick distinguishes capillary, osmotic, and molecular imbibition, the latter of which is the form exhibited by colloids, and it occurs in- dependent of the existence of pores or other preformed spaces in the imbibing body. The imbibition of water by colloids is more than a simple mechanical process, for it is accompanied by a contraction in the total volume of solid and water, and by the evolution of heat. The forces developed are far greater than those of osmotic pressure; e. g., to prevent imbil>ition of water by starcli requires a pressure of over 2500 atmospheres. On the otlier hand, the physical properties of an aqueous colloidal solution show that the colloid is not chemically combined in the form of a hydrate. To describe this peculiar relation Hofmeister and Oswald recommend the term "mechanical affinity." Hardy has shown that water held in a gelatin jelly cannot be removed by great pressures (400 pounds to the square inch), but after the nature of the jelly is so changed by formalin that it is no longer liquefiable by heat, the water can be easily expressed from the loose meshwork that is formed. It would seem from this that the imbibition and retention of water by colloids may be closely related to surface phenomena. Hofmeister has shown that organized animal tissues obey the same laws of imbibition as do simple gelatin plates, and probably this phenomenon of colloids is very important in physiological and patho- logical processes. Non-diffusibility. — The lack of power to pass through animal and parchment membranes, which was Graham's starting-point in the study of colloids, is also only a relative condition. This is shown by the following figures, giving the rela- tive time required by the same amount of different substances to pass through a certain diffusion membrane: Sodium chloride 2 . 33 Sugar 7.00 Magnesium sulphate 7 . 00 Protein 49.00 Caramel 98.00 This difference of time is so great, however, as to permit of separation of salts from proteins, etc., by dialysis, a process in constant use. Primarily the ability to diffuse through a given membrane requires that the diffusing substance be soluble in the membrane. Diffusion menil)ranes are always composed of colloids, e. g., animal bladders, or parchment, whicli is a colloidal cellulose. Crystalloids are generally solul>le in colloids, while colloids are little or not at all soluble in other colloids, and hence do not dilTuse through one another readily and permeate diffusion membranes very slowly. For example, if a stick of agar jelly be jilaced in a solution of ammoniated copper sulphate (a crystalloid), and another be placed in a solution of Prussian blue (a colloid), it will be found that the copper solution penetrates the agar completely before the colloidal solution of Pru.ssian blue has penetrated it at all. This property is of great importance, undoubtedly, in keep- ing different colloidal constituents of the cell in given localities witliin its proto- plasm, c. g., tl)e colloidal glycogen remains wliere it is formcil in tlie cytoplasm, unaljle to escape from tlie cell, whereas the crystalloidal sugar from which it is formed and into which it is converted, diffuses raiiitlly into or out of the cell. The osmotic pressure of tiie colloids is extremely small. The closely related phencjinena of dijliision, (U'prc.ss'ion of freezing-point, and ehrnlion of boiling-point, are also exhibited by colloids to l)ut an extremely slight degree. For example, in one experiment, tlie di.s.solving of from 14 per cent, to 44 ]K'r cent, of egg-albumin in water lowered the freezing-point but 0.02° to 0.00°; and some other colloids have even less effect. The results of the latest and best experiments seem to in- dicate tiiat the trifling ericcts of colloids upon osmotic pressure and upon freezing- and l)oiling-i)oin1s oiiserved in colloidal solulions are due to the colloids tliemselves, ratlier than to included inii)urities, although it may possibly be th.at some oi COLLOIDS 35 these effocts are due to the high surfiice tension and cohesion afTnity of the rolloids. In all cellular processes accompanied by manifestations of osmotic pressure or difTtision, liowcner, the crystalloids may be considered as almost entirely responsible. Electrical Phenomena. — As colloids do not separate freely into ions when di.s- solved, they do not con(hict electricity apprecial)ly. However, when an electric current is passed throunh water containing colloids in sf)lution, the colloidal par- ticles tend to pass to one ])ole or the other. Most colloids move toward the anode. This phenomenon, cdln phoresis^, is also generally exhibited by suspensions, and hence in this particiUar the colloids resemble suspensions rather than solutions. Ilelmholtz has explained the movement of the suspended particles as due to the accumulation of electrical charges upon the surfaces of two heterogeneous media when in contact. The nature of the charge depends upon both the suspended substance and the fluid; e. g., sulphur or graphite particles siispended in water assume a negative charge and move toward the anode, but when suspended in oil of t\n-pentinc they become positively charged and move toward the cathode. \N'ater has such a high dielectric co7istant that most substances suspended in water become negatively charged as compared with the water, and move toward the positive pole or anode. Hardy has observed that colloidal solutions of coagulated proteins move toward the anode when in alkaline solution, and toward the cathode when in acid solu- tion.-'-' This peculiar i)roperty of proteins suggests that perhaps simple surface phenomena do not suffice to account for the electrification of all colloid particles. Knowing the peculiar amphoteric character of proteins, which is probably due to the presence of both NHo and COOH groups in the molecule, we can readily under- stand that in an acid solution the NH2 radicles are combined with the acid,leaving the COOH radicles free. The molecule would then have acid properties, and could dissociate into an acid H ion and a basic or electrically positive colloid ion. The colloid ion would then go toward the negative pole slowly, because of its great size. When a suitable concentration of both ions is produced the proteins will move towards both poles, this concentration being, in the case of serum albumin, H = 10~^ (Michaelis). Living protoplasm behaves in most instances, as if the proteins were acids bound to inorganic cations (Robertson), and is usually stimu- lated at the cathode on the "make" of the current. It is permeable to ions, and the vitality of a tissue is so dependent on the maintenance of normal permeability that the permeability may be employed as a sensitive and reliable indicator of its vitality (Osterhout-^). This maj^ be done by determining the electrical resis- tance of the cells, which is lowered by anything that lowers their vitality. Surface tension,-^ which may be described as the force ivilh which a fluid is striving to reduce its free surface to a minimum, is highly exhibited by colloids as compared with crystalloids. The formation of emulsions and the spreading out of oil upon the surface of water depend upon surface tension. Ameboid movement may be attributed to changes in surface tension, as also may phagocytosis. (The relation of surface tension to these processes will be considered under the subject of Inflammation.) The effect of colloids upon chemical processes going on within their solutions or gels is surprisingly small. Salts in solution in a thick gel of agar or gelatin will diffuse almost as rapidly as in water; they will also ionize as rapidly as in watery solutions, and chemical reactions occur with nearly the same speed and completeness as if the colloids were absent. Furthermore it makes little difference whether these processes are measured in a colloidal solution that is liquid, or after it has set in the gel form. These facts merely indicate that the colloids do not greatly impede the movements of molecules or ions in "According to Field and Teague (Jom-. Exper. Med., 1907 (9), 222), native proteins in serum move towards the cathode, no matter what the reaction. -' Science, 1914 (40), 488. ^^ See article on "Sm-face Tension and Vital Phenomena," by Macallum, Ergeb- nissed. Physiol., 1911 (11), G02. 36 THE CHEMISTRY AND PHYSICS OF THE CELL solutions. On the other hand, as before mentioned, colloids diffuse very slowly into each other. Hence, in the cell the colloids are quite fixed in their positions, whereas the crystalloids may wander about freely, and tliis arrangement is certainly of great importance in bio- logic processes. Pauli suggests the probability that the fixation of the colloid causes the cell to have different properties in different parts, and so various reactions may occur independently in different areas of the cytoplasm. The possibility of the correctness of this view is increased when we consider that the enzymes are colloids, for there is much evidence to show that thej^ are distributed in just such an uneven manner within the cells. Although colloids permit the passage of dissolved crystalloids through them, they greatly interfere with the movement of larger particles. This property accounts for the ability of colloids to hold many insoluble substances in such extremely fine suspensions that they seem superficially to be in true solution. If, for example, sodium phosphate is added to a solution of casein in lime-water, the calcium phosphate formed does not precipitate. It is not in solution, how- ever, but rather exists as a suspension of ver}^ finely divided particles of the salt which the colloid keeps from aggregating into particles large enough to be visible or to overcome the viscosity of the fluid and sink to the bottom. Probably in this way many substances, in- cluding calcium salts, are carried in the blood, held in permanent suspension by the proteins. Substances thus finely chvided w^ill have extremely large surface area for reactions, and, therefore, will undoubt- edly undergo changes with considerable rapidity and facility, although not in solution. Precipitation and Coagulation of Colloids. — Because of the slender margin by which the colloids are separated from the suspen- sions, their persistence in solution is generally in a precarious con- dition. Relatively slight changes suffice to throw the colloids out of solution, and when once precipitated, they are often incapable of again dissolving in the same solvent. Solutions of albumin may under- go spontaneous coagulation on standing for some time, and agitation rapidly produces the same effect in many protein solutions. Some inorganic colloids are as readily coagulated as the proteins. A com- paratively small rise in temperature, less than to 50° C. with some proteins, renders the protein perfectly insoluble. Furthermore, we have coagulation of protein solutions by enzyme action. The inor- ganic "colloidal suspensions" may be precipitated by the addition of very small quantities of electrolytes. Colloidal solutions of the tj'pe of the proteins are not so readily jireciintaled by most clectrolj'tes, but if to the solution large quantities of crystalloids are added, the protein molecules are practically crowded out of solution, as in the "salting-out" process used in separating proteins by ammonium sul- phate and other salts. The effect of heat upon different colloids is COLLOIDS 37 peculiar, in that sonic varieties, as silicic acid, aluininiuni hydrate, and many proteins are rendered so insoluble that they cannot again be dissolved in any fluid without first being modified in some way; where- as colloids of the type of gelatin and agar are made more soluble by heat. The change of colloids into insoluble forms, the "pedous" condition of Graham, requires the presence of water, for the dry col- loids may be heated to relativelj'' high temperatures without losing their solubility. On the other hand, dehydration of colloids while in solution will result in their precipitation and coagulation, as occurs in protein solutions when alcohol is added. If solutions of two oppositely charged colloids are brought together they may precipitate, but if either is present in excess the precipita- tion may be incomplete, or even completely absent. This inhibition of precipitation is of particular interest because it so closely resembles the phenomenon observed in the precipitin reaction, whereby an excess of the antigenic protein will prevent precipitation. Also cer- tain colloids will prevent the precipitation of other colloids by elec- trolytes, which fact is the basis of the Lange reaction of spinal fluid with colloidal gold. Colloids are precipitated by many electrolytes, apparently through the formation of true ion compounds, one or both of the ions of the electrolytes uniting with the colloid ion; although some writers, as Spiro, believe that the combination is merely an additive one between entire molecules. Mathews^^ has advanced the theory that the solu- tion tension of the ions is an important factor in determining the pre- cipitation of colloids by electrolytes. In general, precipitation of colloids results from the reduction of the surface in proportion to the mass, because of an aggregation of the particles; this may be brought about by changing the surface electrical conditions, by uniting the molecules chemically, or by reducing the amount of the solvent. The Structure of Colloids and of Protoplasm. ^^ — Two very different sorts of substances are usually included under the term colloid, because they show the essential features of colloids in most respects; but as in many other respects they are quite unlike each other, it may be well to distinguish between them in some way. As a type of one class we may take gelatin; of the other, such a substance as colloidal arsenious sulphide. Gelatin solutions form gels upon cooling or evap- oration, and redissolve when heated or when more solvent is added. Arsenious sulphide does not form gels upon cooling, and when solidified in any way, does not redissolve. In addition, the gelatin type is very viscous, and is not coagulated by the presence of salts unless these are added in large amounts; while the other type does not render the fluid in which it is dissolved appreciably more viscid, and it forms a precipi- tate immediately if minute amounts of electrolytes are introduced. ^* American Journal of Physiology, 1905 (14), 203. " Review by Harper, Amer. Jour. Botany, 1919 (6), 273. 38 THE CHEMISTRY AND PHYSICS OF THE CELL As the former type resembles in many details the true solutions, while the latter approaches more closely to the suspensions, it has been proposed to distinguish them by the terms "colloidal solution" and "colloidal suspension."-'^ Of the two types, the colloidal solutions are by far the more important in biological considerations, since the colloidal suspensions are usually prepared artificially and seldom occur in nature, e. g., Bredig's colloidal suspensions of the noble metals. The colloidal solutions of proteins are of two tj^pes — one, such as albumin, forms a coagulum when heated, which under ordinary conditions is not reversible; that is, it does not again go into solution. Gelatin, however, becomes more fluid when heated, and when cooled it forms a gel which is readily reversible to the soluble form under the influence of heat. Within the cell, as far as we know, occur only the first type, the proteins that form non-reversible coagula. An extensive study of the physical structure of the colloids has been made by Hardy. ^^ As long as the colloid is in solution it is structureless, although, as before mentioned, the existence of free solid particles can be demonstrated by certain optical methods. The solution is homogeneous, and although perhaps viscid, still it is a typ- ical solution. Such solutions can become solid, either by the effect of temperature, of certain chemical fixing agents, or physical means. It was found by Hardy that in undergoing this solidification there oc-' curs a separation of the solid from the liquid, the solid particles adhering to form a framework holding the liquid within its interstices. Heat-reversible gels show no structure until they are made irreversible by hardening agents, etc.; e. g., a jelly of gelatin appears structui'e'ess, but when treated with formalin or other fixing agent, the structural appearances described below appear. The figures formed by the framework vary according to the nature and concentration of the colloid and of the solvent, and also with the fixing agent used, the temperature, and the presence or absence of extraneous substances. In general, however, the figures obtained in the solidification of pro- tein solutions by fixing agents, such as bichloride of mercury or formalin, bear a striking resemblance to the finer structures of protoplasm as described by cytologists. There is produced an open network structure with spherical masses at the nodal points, or minute vesicles hollowed out in a solid mass, or a honej'comb appearance, or, when the concentration of the colloid is very slight, perhaps there is only a precipitation of fine granules of protein such as we often see in histo- logical preparations of edematous cells and tissues. All these forms seem to depend chiefly uj^on the concentration of the colloid. The important fact is that when the chemicals ordinarily used as fixatives of cells for histological purposes act upon solutions of colloids that are perfectly homogeneous, they produce very constant and charac- ^* Noyes, American Cliemical Journal, 1905 (27), 85. 2» Journal of Physiology, 1899 (24), 158. CELL STRUCTURE 39 tcristic formations wliicli recall at once the structures found in the protoplasm of hardened colls. Moreover, the use of different fixing agents, such as osmic acid, formalin, and bichloride of mercury, pro- duces just the same differences in the structure of colloidal solutions that they produce in the protoplasm of cells hardened by tliom. Neither are the appearances seen in unfixed specimens reliable indi- cations of the true structure of the living protoplasm. Granules of secretion may disappear after or during the death of the cell (e. (j., glycogen) or they may swell up (e. g., mucin granules), thus giving the aj^pcarance of a network or honeycomb which is then incorrectly ascribed to the protoplasm itself. Death of the cells, even when not produced by external influences, seems to be accompanied by coagula- tion of some parts of the cell constituents, and hence a cell examined in anything but its normal living condition, an extremely difficult matter, will not present a true idea of how it appears and is composed while in that condition. By microdissection with the Barber pipette method it is possible to study the properties of unaltered living cyto- plasm, and Seifriz^" concludes from his studies that protoplasm is a homogeneous structureless solution, probably an emulsion hydrosol, i. e., a colloid in which both phases are liquid, one of them, the disper- sion medium, being water. Normal cytoplasm is at all times non- miscible in water, but readily degenerates into a condition in which it is miscible. If, with these facts in mind, we consider the theories of morpholo- gists as to the finer structure of the cell protoplasm based upon stud- ies of cells fixed in various hardening agents, it becomes evident that the possibility that the "foam structure" advocated by Biitschli, or the "thread," "reticular," and "pseudo-alveolar" structures of Fro- mann, Arnold, Reinke, and others, are all simply the effect of fixatives upon colloid solutions, is very real. The objection always advanced to these theories of protoplasmic structure, namely, that the structures described were at least in part artificial productions, not present in the normal living cell, and variously described and interpreted by differ- ent investigators, because each worked with a chfferent hardening fluid or different technic, is strongly supported by these observations upon colloids. This matter will receive further consideration in the next section. THE STRUCTURE OF THE CELL IN RELATION TO ITS CHEMISTRY AND PHYSICS" It is obviously impossible to separate nuclei, nucleoh, cytoplasm, and cell membranes from each other (except with sperm heads) and to isolate them in quantities sufficient for analysis, and therefore we =» Biol. Bull., 1918 (34), 307. '^ Reviews of the significance of cell structure for pathology are given by Benda and Ernst in Zentrlbl. allg. Path., 1914, Bd. 25, Ergiinzungsheft. 40 THE CHEMISTRY AND PHYSICS OF THE CELL are still quite uncertain as to just the chemical differences that exist between them. That there are differences is certain, and by means of micro-chemical reactions, by comparing analyses of cells in which nucleus or cytoplasm predominate, and by studying their phj'sico- chemical relations to one another, we have arrived at more or less tangible ideas on the question of the relation of the structural elements of the cell to its compositi m. The Nucleus 32 Although the nucleus presents morphologically a sharp isolation from the cytoplasm, and displays equally sharp tinctorial differences, it is probable that chemically the differences between nucleus and c^^to- plasm are quantitative rather than qualitative. The characteristic affinity of certain elements of the nucleus for basic stains depends upon the presence in the nucleus of nucleoproteins in large proportion, and to a limited degree nucleoproteins are characteristic of nuclei. Their affinity for basic dyes depends upon the nucleic acid radical. ^^ For example, the heads of spermatozoa contain nucleic acid bound to simple proteins in such a way that it readily forms a salt or salt-like combination with basic dyes, and so the sperm heads appear intensely stained by alum-hematoxylin, etc. Ordinary chromatin threads of nuclei appear to contain somewhat more firmly bound protein in their nucleoprotein molecules, and hence stain less intensely than do the spermatozoa heads, except when in karyokinesis, when the chromatin nucleoprotein seems to approach that of the spermatozoa in avidity for basic dyes. We also have nucleoproteins with the nucleic acid so thoroughly saturated by protein that they do not stain at all by basic dyes, and these seem to exist principally in the cytoplasm, and also to form the ground-substance of the nuclei, occupying the spaces between the chromatin particles (this achromatic substance of the nuclei is called linin or plastin by some cytologists) . Besides the chromatin and the nucleoli, there is a peculiar chromatophile substance, suspended in the finer part of the nuclear structure in the same manner as the chro- matin itself is in the coarser portions; this was called lanthanin by Heidenhain,^'' and is probably similar to the substances also described as parachromatin and paralinin. Undoubtedly the other forms of pro- teins found in the cell, such as globulin, albumin, and nucleoalbumin, exist both in the nucleoplasm and in the cytoplasm, the essential dif- ference being that the proportion of nucleoprotein is nuich greater in the nucleus. As nucleoproteins arc little alTectcd bj^ peptic digestion, 32 Earlier literature bv Albrecht, "Pathologic der Zelle," Lubarsch-Ostertag, Ergeb. der. allg. Pathol., 1899 (6), 1900: see also Kossel, Miinch. mcd. Woch., 1911 (58), 05. -■'■' Herwerdon (Arch. Zellforsch., 1913 (19), 431) found that the basophilic gnmuUvs are disiutcgrated specifically by nuclease, supporting the view that they are nucleic acid compounds. ■>* Festschr. f. Kollikcr, 1892, p. 128. CELL STRUCTURE 41 it is possible to isolate nuclear elements, especially the chromatin, for analytic purposes, and it has been demonstrated by this means also that nuclein is the chief constituent of the staining elements. The distribution in the nucleus of the other primary constituents of the cytoplasm, such as lecithin, cholesterol, and inorganic salts has not yet been worked out, except that Macalluni"''^ found that nuclei contain no chloride, as indicated by their not staining with silver nitrate, and also no potassium, ^^ but the chromatin contains firmly bound iron. Nucleoli, which not all varieties of nuclei possess, differ from the'other nuclear structures in having an affinity for acid rather than for basic dyes," at least in fixed tissues. Their chemical composition lias not been ascertained. Zacharias con- siders the nucleoli as composed of nuclein well saturated with protein, because of its staining reactions and its relative insolubility in alkalies, and classes it with plastin or linin, which forms the achromatic part of the nucleus and is also present in the cytoplasm. Macallum'* found that they reacted for organic phosphorus microchemically, but less strongly than did chromatin fibers. The- nuclear membrane is an imcertain structure, at times dense and staining as if formed of a layer of chromatin, in other cells staining like the cytoplasm with which it seems to be continuous, in most cells disappearing during karyokinesis, and in some protozoa being entirely absent. Naturally the composition of the nuclear membrane is unknown, but it is probable that it acts as a diffusion mern- brane of partially semipermeable character, maintaining different conditions in nucleus and cytoplasm. Functionall}^ the nucleus is the essential element of the cell; an isolated nucleus with but a minimum of cytoplasm may be able to de- velop new cytoplasm, but isolated cytoplasm soon disintegrates, al- though it may manifest Hfe for some time by movement and chemical activities. It has been frequently suggested that the nucleus controls oxidative processes, and there is some microchemical evidence for this.^^ Lynch-^" calls attention to the improbability that the part of the cell most removed from the oxygen should be the organ of oxida- tion, and finds evidence that the function of the nucleus is that of organic synthesis. An enucleated cell may move, respire, digest, respond to stimuli and exhibit any activity which is dependent solely upon catabolic or destructive processes of protoplasm. The group of phenomena which it never shows are those of growth, regeneration and division, i. e., those depending on synthetic activities. It should be mentioned that certain cells, such as bacteria and algse, 'seem to have no true nuclei, but Macallum^^ found that the forms he examined gave reactions for phosphorus and iron in a similar way to the nucleoproteins of a nucleus, suggesting that in such cells the nuclear elements are diffused through the cell rather than differen- tiated. To* quote Wilson: "The term 'nucleus' and 'cell body' 35 Proceedings of the Roval Society, 1905 (76), 217. se Jour, of Physiol., 1905 (32), 95. 3^ Nucleoli of nerve-cells are an exception, being basophilic. =58 Proc. of the Roval Societv, 1898 (63), 467. 33 See Osterhaut, Science, 1917 (46), 367. « Amer. Jour. Physiol., 1919 (48), 258. *^ "Studies from the University of Toronto," 1900. 42 THE CHEMISTRY AND PHYSICS OF THE CELL should probably be regarded as only topographical expressions, de- noting two differentiated areas in a common structural basis." Because of the relative acidity of the nuclei they are electrically negative to the cytoplasm, particularly when in karyokinesis, and the chromatic elements of the nucleus can be shown to carrj'" a negative electric charge."^- Sperm-heads in isotonic cane-sugar solution move rapidly — ^2000 microns a minute — ^toward the anode, when a current is passed through the solution; and leucocytes also go toward the anode under the same conditions, the rate depending upon the pro- portion of nucleoplasm and cytoplasm, large leucocytes sometimes even going slowly toward the cathode. The SertoH cells of the testi- cle, which have a round mass of cytoplasm with a number of minia- ture spermatozoa heads at one side, orient themselves in the current so that the side or end containing the spermatozoa drags the mass of cytoplasm toward the positive pole. The Cytoplasm The cytoplasm, as before mentioned, contains all the primary cellular constituents, and also such secondary constituents as the par- ticular cell possesses. Nucleoproteins are undoubtedly present in unknown proportions, but with the nucleic acid well saturated by proteins, and perhaps also to a large extent combined with carbohy- drates to form the glyconucleoproteins. Sometimes the nucleoproteins of the cytoplasm may be partly of the unsaturated class, and show an affinity for basic stains, as in the case of the Nissl bodies of the nerve-cells, the basophilic granules of mast cells,^^ and perhaps also the cytoplasm of plasma cells. The great question concerning the cytoplasm is its structure — whether homogeneous, alveolar, areolar, fibrillar, foam-like, or granular. On a previous page have been men- tioned the experiments of Hardy, which show that homogeneous solu- tions of protein, when fixed by the same reagents as are used in the customary fixation of histological materials, may show quite the same microscopical structures as are shown by the cytoplasm of cells. Network, foam, and alveolar structures are produced in albumin and gelatin solutions when they are hardened by bichloride of mercury, osmic acid, formalin, etc., and the same characteristic differences that are produced in cells by these different reagents are likewise produced in the hardened protein solution. Protein struc- tures hardened under strain form radiating structures resembling centrosomes and the radiating threads seen in cells. If elder pith is saturated with protein solutions and then hardened, sectioned, and stained by the usual methods, appearances resembling closely the structure of a hardened cell may be found in the spaces of the « Pentamalli, Arch. Entwick. u. Org., 1912 (34), 444; McClendon, Proc. Soc. Exp. Biol, and Med., l'.)l() (7), 111; Hardy, .lour. Pliysiol., 1913 (47), lOS. CELL STRUCTURE 43 l)ith — even a (-(Mitral, iiuclcus-likc mass may be suspencJcd in a net- work of anastomosing threads. These and many other experi- ments indicate that much of tlie work done on cell structure by means of studies of hardened cells cannot be considered of value in deciding the structure of living cells; but, nevertheless, the fact remains that man}^ cells that can be observed while alive and uninjured under the microscope are seen to have a definite structure in the cytoplasm, e. g., sea-urchin eggs, which show a characteristic alveolar structure. A compromise view of the structure of protoplasm (and cytoplasm in particular) which takes account of what appear to be facts brought out on both sides of the question, is that while in some cells definite structural arrangements of the cytoplasm exist, in most cells the proteins are chiefly in a homogeneous solution; most of the structures seen in fixed cells, except the mitochondria, chromatin threads, nuclear membrane, nucleoli, and centrosomes, are produced by the coagulation of the proteins, and are not present during life. When a framework does exist, it is a fair inference, by analogy with the cell mem- brane and the stroma of the red corpuscles, that the cell lipoids are largely responsible for its formation, and that they form a prominent part of its composition. This question of the presence or absence of structure in the cytoplasm is of more importance than as a mere mor- phological problem, for if the cytoplasm is subdivided into innumer- able little chambers, each surrounded by a membrane, it is probable that processes of diffusion and conditions of osmotic pressure will be very different from what they would be if the cytoplasm were a simple homogeneous colloid solution, like a lump of semisolid gelatin or agar. In such colloidal masses diffusion and osmosis go on almost as if there were no colloids in the solvent at all, whereas most membrane struc- tures that are found in living tissues seem to have a decidedly semi- permeable character. From what we know at the present time of intracellular physics and chemistry there is no necessity for assuming that semipermeable septa exist within the cell. All the intracellular processes with which we are familiar could go on without such structures. It is not neces- sary to assume a compartment structure to explain the possibility of different chemical reactions going on in different parts of the cell at the same time, for most of the cell reactions seem to depend on enzymes, which we know are not readily diffusible in solutions of col- loids, and, therefore, might remain fixed without requiring any en- closing walls or retaining framework. Certainly, many cells are free from structural cytoplasm, for we see particles of solid matter moving about within them quite freely. In some cells the nuclei migrate about in the cell, as also do digestive and excretory vacuoles, which motion would seem to be rather destructive if the protoplasm had a structure at all permanent. When a portion of the cytoplasm is cut free from the body of 44 THE CHEMISTRY AND PHYSICS OF THE CELL certain cells it at once forms a round drop, just as any insoluble fluid would do in another of different surface tension, and not at all as if it were bound into a fixed structure by a framework. Other cells, however, retain their form under the same conditions. The structure of even so evidently complicated a cytoplasm as that of striated muscle fibers is in doubt; a classical observation on this point is the passage of a minute worm through the substance of a muscle- cell, its progress being as unimpeded as if there were no such things as disks, bands, rods, and striae in the cell. Many features of ame- boid movement also seem to indicate that the cytoplasm follows much the same laws as a drop of fluid in a heterogeneous medium, for we can make a drop of mercury or of chloroform in water, or of oil in weak alcohol, react to various stimuli in much the same waj' that an ameba would. If we look upon the cytoplasm as a drop of emulsion colloid, the surfaces of the particles in the emulsion furnish of them- selves adequate explanation of many of the phenomena of isolation of chemical reactions, etc., without lacking in harmony with the evi- dences of structural homogeneity. This hypothesis fits all sides of the problem and has many supporters at the present time.*^ The question of structure in the nucleus is quite a different matter, in so far as the chromatin threads and the nucleolus are concerned. In ameboid movement the nucleus seems to play a passive role and to be dragged about by the cytoplasm, indicating quite a high degree of rigidity. It is probable, however, that the achromatic portion between the chromatin threads and granules has much the same structure or lack of structure as the cytoplasm. The various secretory granules, fat-droplets, pigment-granules, glycogen gran- ules, keratin, etc., that may lie in the cytoplasm, are inconstant constituents, varying with different cells, and under varying conditions in the same cells, and lie beyond the scope of our discussion of the general composition of the cell. Ac- cording to Ruzicka'*^ there is contained in all cells, both in nucleus and cytoplasm, an insoluble substance which corresponds structurally to the "plastin" of the cytologists, and chemically is related to the reticulins and other albuminoids; this he looks upon as the ground substance of the cells, corresponding to the albu- minoid ground substance or stroma of organized tissues. Certain of the granulations observed in the cj^toplasm of cells seem to be de- finite, constant structures of the living protoplasm, and these are now called mito- chondria, which term includes many forms of granules described \mder various names." Their solubility and staining reactions suggest that they contain phos- pholipins, perhaps associated with proteins. Their functional importance is in- dicated by the fact that usually their number varies directly with the metabolic activity of the cells, and they may lie related to histogenesis. Other histological cellular structures also permit of more or less satisfactory identification by microchemical methods, and Unna""' especially has contributed to this field. By staining sections with dyes of varying reaction, after extracting the sections with various solvents, he has obtained evidence of the chemical nature "An excellent discussion of this question is given bv Alslierg, Science, 1911 (34), 97. ** Arch. f. Zellforsch., 1908 (1), 5S7. « Sec review bv Cowdry, Amer. .lour. Anat., 191() (19), 423; Carnegie Inst. Publ., No. 2."), 1918. " See review by Gans, Dcut. med.Woch., 1913 (39), 1944. CELL STRUCTURE 45 of some of the cell structures, although it is by no means certain that the conclu- sions drawn will all be verified. In the nucleolus he finds a substance resembling glol)uliu, the i^runuloplasiu of tlic cell body lie rej^ards as an albumose, the spongio- plasni as histone, mast cell granules as mucin or mucoid substances. Nissl bodies he holds to be albumose, altho otiiers have beUeved them to be nucleins." The Cell-wall** The cell membrane in most animal cells is inconspicuous struc- turally, but in discussing osmosis it was shown that it is of the greatest biological importance. There is no direct chemical or microscopical evidence at hand showing the composition of the animal cell mem- brane, but by observations on its behavior when the cells are in solu- tions of different sorts, facts have been collected indicating that phospholipins and cholesterol, and probabl}^ alhed fat -like bodies, are prominent constituents. The substances that difTuse through most cell walls are just the substances that are soluble in or dissolve these lipoids, e. g., alcohol, chloroform, ether, etc., and it is probable that the anesthetic effects of many of these substances depend in some way on their fat-dissolving power and the large proportion of lipoids in nerve-cells. These observations were made first by Overton^^ and Meyer, ^"^ and led to the now prominent but disputed hypothesis that the permeability of cells is determined by the lipoids. Of particular in- terest for our purpose are Overton's observations on the effects of dyes on living cells. The best known vital stains (z. e., stains that will enter the living cell without requiring or causing injury to it) are neu- tral red, methylene blue, toluidin blue, thionin, and safranin. If uninjured cells, e. g., frog eggs, are placed in watery solutions of these dyes they soon become filled with the coloring-matter, which seems to penetrate the cell menbrane quite uniformly at all points; if the dyed eggs are then placed in clear water, the stain diffuses out again, showing it to be simply absorbed, rather than chemicallj^ combined. In contrast to these stains the sulphonic acid dyes, such as indigo car- mine and water-soluble indulin, nigrosin, and anilin blue, do not pene- trate the living cell at all. Overton tested the solubilit j^ of dyes which are not vital stains and found them all insoluble in oils, fats, and fatty acids; but the dyes staining living cells were readily soluble in lecithin, cholesterol, "protagon," and cerebrin, the so-called cell lipoids. Fur- thermore, if crumbs of lecithin, "protagon," or cerebrin were placed in very dilute watery solutions of these dyes, they were found to absorb from the water the vital stains, but not the others, which indicates that stains that penetrate living cells are more soluble in lipoids than they are in water. " See Unna, Berl. klin. Woch., 1914 (51), 444; Muhlmann, Arch. mikr. Anat., 1914 (85^ ,361. •** See Zangger, "Ueber Membranen und Membranenfunktionen," Ergebnisse d. Physiol., 1908 (7), 99; also R. S. Lillie, "The Role of Membranes in Cell Pro- cesses," Popular ScienceMonthlv, Feb., 1913. " Jahrb. f. wissentschaftl. Botanik, 1900 (34), 669. " Arch. f. exp. Path. u. Pharm., 1899 (42), 109. 46 THE CHEMISTRY AND PHYSICS OF THE CELL Many exceptions to this rule of the fat-solubihty of dyes which can penetrate Hving cells have been found, especially by Ruhland,^^ and the universal applicabihty of the Overton-Meyer hypothesis has been questioned. It is at once evident that the common foodstuffs which enter the cell, such as water, sugar, amino-acids, and salts are not lipoid-soluble, hence it has been suggested that the cell membranes must have a "mosaic" structure, some of the blocks being lipoids or lipoid compounds, and others proteins without lipoids. (Robertson*^ suggests that there is a superficial film of concentrated protein about the cells, underlaid by a discontinuous lipoid layer.) There is, fur- thermore, evidence that the entire cell substance has a profound effect upon diffusion within the cell, so that it is at present impossible to say whether the osmotic phenomena of cells depend upon a cell mem- brane or upon the entire cell substance. ^^ It may be that there are membranes or surfaces within the cell, as postulated in the foam structure hypothesis of protoplasm, or that a homogeneous protoplasm develops surfaces where in contact with substances entering from the outside. Many facts indicate that either the delicate external membrane of animal cells or the entire cytoplasm has the features of a semi- permeable membrane, to the extent of permitting certain substances to diffuse through and not others. Had they the property of some of the artificial semipermeable membranes, of letting water pass through but holding back almost absolutely all crystalloids, the re- sult would be the development of an enormous disproportion in the pressure between the inside and the outside of the cell. Furthermore, the exchange of nutritive material and excretion products between the blood and the cells would be impossible. But permitting some sub- stances to pass into the cell results in their accumulation within the cell, until they are in sufficient concentration to neutralize the osmotic pressure exerted on the outside of the cell. As evidence of this elec- tive permeability we have the fact that the proportion of certain salts within the cell is quite different from what it is in the fluids bathing them; e. g., animal cells generally contain more potassium and less sodium than the fluids surrounding them. The inorganic constituents of red cells are different from those of the plasma, the corpuscles not containing any calcium at all, while the magnesium seems to enter them freely; in other words, the red corpuscle seems to be impermeable to calcium and permeable to magnesium. If the salts in a corpuscle are in smaller proportions than in the surrounding fluid, it indicates that the cell membrane is not freely permeable for them; if in greater proportion, that some constituent of the cell is holding them in combination, possibly as ion-protein compounds. Probably 6' Jahrb. f. Wisscnschaft. Botanik, 1912 (51), 376. '•■'Jour. Biol. CluMn., 1908 (4), 1. " Sec Kite, Aincr. Jour. Thysiol., 1915 (37), 282; Chambers, ibid., 1917 (-13), 1. CELL STRUCTURE 47 inorganic salts are present in the cell by virtue of both physical and chemical influences, some simply (Hffusiii^; in and out, others com- bining with the proteins and being held chemically. Bechhold summarizes his conception of cell walls as follows: "Every cell at its surface possesses a membrane which is dependent upon the composition of the interior of the cell. This membrane may be visible and may have been formed through the gelatinizal ion of the cell protoplasm at the periphery. It may, on the other hand, be so thin as to be invisible, being formed by the concentration and spread- ing out of such albuminous and fatty colloids as diminish the surface tension of the cell content at the interface. The cell membranes, de- veloping as a result of the gelatinization of cell protoplasm, are at first, in youth, expansile and elastic; with increasing age these mem- brane colloids, depending upon their environment and upon chemical influences, or as a result of mere colloid aging phenomena, become poor in water and lose their elasticity." The intercellular substance varies greatlj^ in different tissues. In the case of the supportive tissues it is the important element, and the cells seem to exist chiefly for the purpose of forming and keeping it in repair as it is worn out. In the epi- thelial and secreting tissues, however, the intercellular substance is reduced to a minimum, except in so far as a cement substance is required, and the cells generally lie in almost immediate apposition. It is probable that there is a greater or less amount of cement substance, even between the most closely applied cells, and this substance seems to be related to mucin. It can generally be brought out by stain- ing with silver nitrate, and Macallum^^ points out that this reaction is merely a micro-chemical test for chlorides, and indicates that the cement substance con- tains them in larger proportion than does the cytoplasm. " Proceedings of the Royal Society, 1905 (76), 217. CHAPTER II ENZYMES Every cell is constantly accomplishing an enormous number of chemical reactions of varied natures, at one and the same time; how many we do not know, but the score or more that we do know to be constantly going on in the liver cell, for example, are probably but a part of the whole. Furthermore, reactions take place between sub- stances that show no inclination to affect each other outside the body, and proceed in directions that we find it difficult to make them take in the laboratory. Proteins are being continually broken down into urea, carbon dioxide, and water; yet to split proteins even as far as the amino-acid stage requires prolonged action of concentrated acids or alkalies, or super-heated steam under great pressure. But all the time in the cell innumerable equally difficult changes are going on at once, within its tiny mass, always keeping the resulting heat within a frac- tion of a degree of constant, and the resulting products within narrow limits of concentration. We have already indicated the means used to keep the concentration of the cell products within safe limits; namely, the processes of diffusion and osmosis and their modification by the cell structure. The forces that bring about the chemical reac- tions reside, we say, in enzymes, although in so doing we only shift the attribute formerly conceded to the cell, to certain constituents of the cell whose nature and manner of action are equallj^ unknown. When the only enzymes that were known were limited to those se- creted from the cell, and found free in fluids, such as pepsin and tryp- sin, the chemical changes that went on in the cell were ascribed to its "vital activity." Buchner, by devising a method to crush yeast cells, and finding the expressed cell contents able to produce the same changes in carbohydrates that the cells themselves did, proved the ex- istence within living cells of enzymes similar to those excreted by cer- tain cells, and substantiated the belief of their existence that had become general before it was thus finally corroborated. Growing out from this and subsequent experiments has come a larger and larger amount of evidence that many of the chemical activities of the cells are due to the enzymes they contain, until now the point is reached where one may rightfully ask if cell life is not entirely a matter of enzyme activity. There are certain facts, however, which seem to in- dicate that there are some essential differences between cells and 48 NATURE OF ENZYMES 40 enzymes. One of the most important of these is the difference in the susceptibility to poisons of enzymes and cells. ^ Strengths of certain antiseptics that will either destroy or inhibit the action of living cells, such as alcohol, ether, salicyhc acid, thymol, chloroform and toluene, will harm free enzymes in solution little or not at all. This fact has been of great assistance in distinguishing between the action of en- zymes and of possible contaminating bacteria in experimental work. Although this difference between enzymes and cells is characteristic, it does not finally decide that the cell actions are not enzyme actions, for it may well be that the poisons act chiefly by altering the physical conditions of the cell so that diffusion is interfered with, thus seriously interfering with the exchange of cleavage products between different parts of the cell, and checking intracellular enzyme action, which we shall see later requires free diffusion of the products for its continuance. At the very least, however, we may look upon the intracellular en- zymes as the most important known agents of cell metabolism, and consequently of all life manifestations, and the changes they undergo or produce in pathological conditions must be fully as fundamentally important as is their relation to physiological processes. It therefore becomes necessary for us to consider carefully — 'THE NATURE OF ENZYMES AND THEIR ACTIONS Since up to the present time no ferment has been isolated in an absolutely pure condition we are entirely unfamiliar with their chemical characters, and conse- quently are obUged to recognize them solely by their action. As far as we know, true enzymes never occur except as the result of cell life — they are produced with- in the cell, and increased in amount b}^ each new cell that is formed, and, further- more, they are present in every living cell without exception. As the same facts are equally true of the proteins it is natural to associate the enzymes with pro- teins, and so explain the importance of the proteins for ceU life.^ If enzymes are obtained in any of the usual ways from animal cells or secretions they are always found to give the reactions for proteins, even if repurified many times. But it is well known that whenever proteins are precipitated the other substances in the solution tend to be dragged down by the colloids, and it is possible that the en- zymes are merely associated with the proteins in this way. Furthermore, enzymes are known to become so closely attached to stringy protein masses, such as fibrin and silk, that they cannot be removed by washing. Some have claimed that they have secured active preparations of pepsin and invertase that did not give protein reactions and contained very little or no ash or carbohydrate ; but it has so far been impossible to secure trypsin free from protein, and diastase seems to be certainly of protein nature. Davis and Merker^ find that the more pepsin is purified the more 1 See discussion by Vernon, Ergebnisse d. Physiol.. 1910 (9), 234. - It would not be profitable to discuss fully all the various theories and hypotheses that have been advanced, but the reader is referred to the following chief compilations of the entire subject: Oppenheimer, "Die Fermente und ihre Wirkungen," Leipzig; BayHss, "The Nature of Enzyme Action," Monographs on Biochemistry, London; Stern, "Physico-chemical Basis of Ferment Action," in Oppenheimer's "Handbuch d. Biochemie," Vol. 4, pt. 2; Samueley, "Animal Fer- ments," ibid, Vol. I; A. E. Taylor, "On Fermentation," Univ. of California I\ibUca- tions; Euler, "General Chemistry of the Enzymes," translated by T. H. Pope, New York, 1912. ' Another important point is that the closest imitation of enzymes, Bredig's 'inorganic ferments," seem to owe their action to their colloidal nature. * Jour. Amer. Chem. Soc, 1919 (41), 221. 4 50 ENZYMES it approaches the character of a protein, possibly a glycoprotein, with increasing proteolytic activity.^ Analyses of enzymes purified as completely as possible do not have great worth, for the "purified" enzymes are probably far from pure; however, it is of some importance that the.y vary greatly in the proportions of carbon, hydrogen, and nitrogen which they contain, indicating that possibly dif- ferent enzymes may be of very different nature. The enzjanes have been found to possess definite electrical charges; in neutr-il solutions trypsin is negative or amphoteric, pepsin and invertase negative (Michaelis).^ Macallum has shown microchemically that phosphorus is closely associated with the formation of zymogen granules in cells, which seem to be started in the nucleus; and there are many other observations suggesting that certain ferments are closely related to the nucleo-proteins. This is particularly true of the oxidases, which seem also to contain iron and inanganese. A final point of importance in support of the protein nature of enzymes is that pepsin destroj^s trypsin and diastase, while trypsin destroys pepsin.' So uncertain, however, is our information concerning the chemical nature of the enzymes, that it has become possible for an hypothesis to be developed urg- ing that enzymes are immaterial, that the actions we consider as characterizing enzymes are the result of physical forces which may reside in many substances, and perhaps even free from visible matter, but the weight of evidence at present available is entirely in favor of the view that enzymes are specific colloidal sub- stances, although perhaps of widely differing chemical nature. A valuable piece of evidence of the material existence of enzymes is their specific nature, lipase affecting only fats, and trypsin only proteins, indicating chemical individuality. They are true secretions, formed within the cell by recognizable steps; and, furthermore, when injected into the body of an animal, they give rise to the forma- tion of specific immune bodies that antagonize their action. Emil Fischer's work with the sugar-splitting enzymes, moreover, indicates that they owe their action to their stereochemical configuration. He prepared two sets of sugar de- rivatives which differed from each other solely in the arrangement of their atoms in space (i. e., isomers) and found that one specific enzyme would split members of only one of the varieties, while another enzyme would act only on the variety with the opposite isomeric form. These experiments make it very probable that there must be a certain relation of geometrical structiu-e between an enzyme and the substances it acts upon, and leaves little question of its material nature. Bredig has found that colloidal solutions of metals have many of the properties of true enzymes, accomplishing many of the decompositions produced by en- zymes, being affected by temperatures of nearly the same degree, and even being "poisoned" by substances that destroy or check enzj^mes.^ The only possible explanation of these observations seems to be that the enzyme effects are brought about by surface -phenomena. A colloidal solution of platinum, as far as is known, differs from a piece of metallic platinum solely in the enormously great amount of surface it offers in proportion to its weight, and it is well known that surfaces may affect chemical action. Hence we have the possibility that some enzyme actions, at least, may depend upon the existence of a very large surface, and since by no means all colloids are enzymes, that this surface must bear a certain relation in form to the surface of the body that is to be acted upon. The Principles of Enzyme Action The effects produced by enzymes, which at one time were con- sidered quite unique and remarkable, have no\v been made compara- tively plain, chiefly through the observations of Ostwald on related 5 Bokorny (Biochem. Zeit., 1919 (94), G9) finds -that the amount of formalde- hyde fixed by emulsin supports the hypothesis that this enzyme is a protein. " Bioclieni. Zeit., 1909 (l(j), 81 and 480; (17), 231. ' Falk lias ol)tained evidence that ester-splitting enzymes may be proteins owing their activity to the presence in the molecule of active groupings, perhaps of enol-lactim structure, — C(OII) = N — . B.y treating pure proteins with alkali, which favors the fornuition of enol-lactim groupings, the proteins were made to acquire esterase properties. (See Science, 1918 (47), 423.) " See also Fischer and Hooker, J. Lab. Clin. Mini., 1918 (3), 373. ENZYME ACTION 51 c;ieniical reactions; ami by the investigations of Croft Hill, Kastle and Loevenhart, and others, on enzymes, which show that enzyme action is in no way different from chemical action observed independ- ent of enzymes. The fundamental consideration is that chemical re- actions are reversible, that is, that their tendency is to establish an equilibrium, and that the change may be from either side of the equa- tion.^ The action of enzymes is similar to that of all catalytic agents, that is, they increase the speed of reaction. In the case of such a reaction as that of NaOH and HCl, the reaction is so rapid that the effect of catalyzers could hardly be noticed; but with many other substances the reaction is very slow, and without the presence of catalyzers it would go on almost or quite imperceptibly. For ex- ample, ethyl butj'rate saponifies on the addition of water according to the following equation: C.H5 - O - OC - C3H7 + H2O ^ C0H5OH + HOOC - C3H7. On the other hand, if ethjd alcohol and butj-ric acid, the products of this reaction, are placed together, they will combine to form ethyl butyrate; in other words, the reaction is reversible, as indicated by the arrows in the equation. In any event, however, the reaction is not complete, but continues only until there exists a certain definite pro- portion of ethyl alcohol, butyric acid, eth3'l butyrate, and water, when the change will stop, i. e., equilibrium is established. The time that would be required for this reaction to occur at room temperature would be extremely long, the change being hardly noticeable, but in the presence of a catalytic agent the reaction goes on much more rapidly. Catalytic agents, therefore, merely hasten reactions which would go on without them, and they do not initiate or change the na- ture of chemical reactions at all. When equilibrium is established, the reaction stops and the enzyme has nothing more to do. Furthermore, enzj^mes will hasten synthesis just as well as they hasten catalysis. Croft Hill first showed that maltase would synthesize glucose intc maltose; Kastle and Loevenhart soon after estabhshed the synthesis of ethyl butyrate under the influence of lipase. Taylor^ ° first syn- thesized one of the normal body fats, triolein, by the action of hpase (from the castor-oil bean) upon oleic acid and glycerol. Successful synthesis of fats by pancreatic lipase is described by Lombroso.^^ It may seem improbable at first sight that the synthesis of proteins can be accomplished by enzymes, as is the relatively very simple synthesis of carbohydrates and fats, but the improbability disappears when we recall that the products of protein cleavage are reconverted into body proteins after absorption from the intestines. Proteins manifestly are synthesized and we have not a little reason to believe 9 See Taylor, Arch. Int. Med., 190S (2), 148. i» Univ. of California Publications (Pathology), 1904 (1), 33. 11 .\rch. di farmacol., 1912 (14), 429. 52 ENZYMES that this is accompUshed by enzymes, presumably by a reversal of their action in the establishment of equilibrium. Abderhalden^^ has obtained some evidence of protein formation in mixtures of amino acids derived from autolyzing tissue when acted upon by ferment- containing extracts of the same tissue. Taylor^^ was able to synthesize protamin, one of the simplest proteins, by the action of trypsin upon its cleavage products, and it has been found that the addition of proteolytic enzymes to solutions of pure albumose leads to the forma- tion of a jelly-like, insoluble protein substance, " plastein," which seems to be the effect of a reversed action on the part of the enzymes. ^^ Another well known synthetic action that seems to be due to reversible ferment action is the formation of hippuric acid from benzoic acid and glycine in the Iddney; the formation of glucose into glycogen and its reformation are also probably both accomplished by one and the same enzyme acting reversibly. Other reversible reactions less closely related to animal cells have also been described. The reversible nature of enzyme action explains many problems of metabolism, and makes the whole field much clearer. The following consideration of the newer understanding of fat metabolism on this basis may explain the manner in which chemical changes are believed to occur in the cells and fluids of the body:^^ In the intestines fat is split by lipase into a mixture of fat, fatty acid, and glycerol; but as the fatty acid and glycerol are diffusible, while the fat is not, they are separated from the fat by absorption into the wall of the intestine. Hence an equilibrium is not reached in the intestine, so the splitting continues until practically all the fat has been decomposed and the products absorbed. When this mixture of fatty acid and glycerol first enters the epithelial cells Uning the intestines there is no equilibrium, for there is no fat absorbed with them as such. Therefore the lipase, which Kastle and Loevenhart showed was present in these cells, sets about to establish equilibrium by combining them. As a result we have in the cell a mixture of fat, fatty acid, and glycerol, which will attain equilibrium only when new additions of the two last substances cease to enter the cell. Now another factor also appears, for on the other side of the cell is the tissue fluid, containing relatively little fatty acid and glycerol. Into this the diffusible contents of the cell will tend to pass to establish an osmotic equilibrium, which is quite independent of the chemical equilibrium. Tliis abstraction of part of the cell contents tends to again overthrow chemical equilibrium, there now being an excess of fat in the cell. Of course, the lipase will, under this condition, ex- hibit the reverse action and split the fat it has just built into fatty acid and glycerol. It is evident that these processes are all going on together, and that, as the composi- tion of the contents of the intestines and of the blood-vessels varies, the direction of the enzyme action mil also vary. In the blood-serum, and also in the lym- phatic fluid, there is also lipasC; which will unite part of the fatty acid and glycerol, '2 Fenuoritforschung, 1914 (1), 47. " Jour. Biol. Chem., 1909 (5), 381. 1^ See Michcli, Arch. ital. biol, 190G (46), 185;Levene and Van Slvke, Biochem. Zeit, 1908 (13), 458; Taylor, Jour. Biol Chem., 1909 (5), 399; Gav and Robert- son, ibid. 1912 (12), 233; Alxlerhalden, IVrnientforsch., 1914 (1), 47; v. Knaffl- Lenz and Pick, Arcli. exp. Path., 1913 (71), 29(), 407. '5 See Loevenhart, Amer. Jour, of Fliysiol., 1902 (0), 331; Wells, Journal Amer. Med. Assoc, 1902 (38), 220. The discrepancies between the action of lipase in the tissues and in vitro are well explained by Taylor, Jour. Biol. Chem., 1906 (2), 103. ENZYME ACTION 53 and by rcinovinf:; tlicm from the fluid about the cells favor osmotic diffusion from tlie intestinal epithelium, thus facilitating absorption. Quite similar must be the process that takes place in the tissue cells through- out the body. In the blood-serum bathing the cells is a mixture of fat and. its constituents, probably nearly in equilibrium, since lipase accompanies them. If the diffusible substances enter a cell containing lipase, c. r/., a liver cell, the process of building and splitting will be quite the same as in the intestinal epithelium. The only difference is that here the fatty acid maj' be removed from the cell by being utilized by oxidation or some other chemical transformation.'* To summarize, it may be stated that thiouglK)iit the body there is constantly taking place both splitting and building of fat. Fat enters the cells, leaves them, and is utilized only in the form of its acid and alcohol, never as the fat itself. Fat constitutes a resting stage in its own metabolism. If proteolytic enzymes also act reversibly, then the phenomena of protein metabolism are similarly explained, for there is no doubt that every cell and body fluid contains proteolytic enzymes. All metabolism, then, may be considered as a continuous attempt at establishment of equilibrium by enzymes, perpetuated by prevention of attainment of actual equilibrium through destruction of some of the participating substances by oxidation or other chemical processes, or by removal from the cell or entrance into it of materials which overbalance one side of the equation. In just what manner the enzymes accomplish their catalytic effect is yet unknown. 1^ A favorite idea is that they form loose compounds with the substance to be split and with water; the resulting compound being unstable and breaking down, the water remains attached to the components of the substance. Enzymes do not act catalytically on all substances by any means, but show a decidedly specific nature. They affect only organic sub- stances, and the actions are limited to two processes — hydrolj'sis and oxidation, or the reverse processes of dehydration and reduction.'^ The most essential difference between the enzymes and the chemicals that can accomplish hydrolysis or oxidation is this: the ordinary chemical reagents produce their effects on many sorts of substances, whereas the enzymes are specific; thus hydrochloric acid will hydrolyze starch or protein with equal facility, but pepsin will not affect starch at all. The very specific nature of the enzymes, their activation by other body products, the fact that they seem to be bound to the substance 16 Bradley (Jour Biol. Chem., 1910 (S), 251; 1913 (13), 407-439) calls atten- tion to the great concentration necessary for fat synthesis by lipase in vitro, and the lack of correspondence between the amount of fat and of lipase in various tissues, questioning the importance of lipase for fat synthesis in the living tissues as well as the significance of reversed enzyme reaction for biological processes in general. '^ See Euler, "Chemical Dynamics of Enzyme Reactions," Ergebnisse d. Physiol. 1910 (9), 241. 18 Alcoholic fermentation may be an exception, the change being CeHnOs = 2C2H6OH -|- 2CO2, but it is very possibly an intramolecular oxidation. 54 ENZYMES upon which they act, that they are susceptible to heat, and that they produce immune bodies when injected into experimental animals, all suggest the probability of a relationshij) between enzymes and toxins. This matter will be discussed more fully in considering the chemistry of immunity against enzjanes. General Properties of Enzymes. — Other properties of enzymes may be briefly mentioned. The speed of reaction they produce increases with the amount of enzymes present. Very dihite acids favor the action of nearly all ferments, and alkalies are unfavorable for all but trypsin, ptyalin, and a few others. Weak salt solutions also are more favorable than distilled water. (These facts suggest strongly the possibility that ions play an important role in the process.) Water and dilute glycerol dissolve enzymes, which form always colloidal solutions that are very slightly dialyzable; and they may be precipitated from solution by alcohol, and redissolved again with but sUght impairment of strength. Filtra- tion through porcelain filters is not complete, from 10 to 25 per cent, of most en- zymes being lost in each filtration and enzymes are subject to great absorption by surfaces, e. g., charcoal, kaolin.^ As before mentioned, many chemicals poison- ous to bacteria have little influence on most enzymes, but nearly all substances when concentrated are injurious or destructive, and some enzymes are kno^^Ti that are more susceptible to antiseptics than are the cells that contain them. Formaldehyde is very destructive to most enzj^mes, even when dilute. The effect of protein-coagulating antiseptics upon enzymes is, of course, greatly modified by the amount of protein substances mingled with the enzymes ; and the effects of heat and other injurious influences are greatly decreased by the presence of proteins and other impurities. All enzymes are most active between 35° and 45° C, and it is interesting to note that tvobert found this equally true for enzymes derived from cold-blooded animals.'" Although enzymes can stand temperatures of 100° C. or more when dry, in water they are generally destroyed somewhat below 70° C. Low tempera- ture, even — 190° C. (liquid air), does not destroy them. The loss of power through heating occurs gradiially, and there is no sharp line at which their action disappears. Sunlight is harmful to enzymes in solution, but only in the presence of oxygen; this effect is augmented by the presence of fluorescent substances. Nascent oxygen is destructive to enzymes. '-' Radium and .x-rays seem to have a deleterious effect upon most enzymes, and retard their rate of action ; but apparently, autolytic enzymes (Neuberg"-) and tyrosinase (Willcock-'O are not injured by these agencies. -^ Ultra violet rays are also injurious to enzymes,-^ and they can be destroyed by violent shaking (Shaklee and Meltzer.-*^). Labile as enzymes are, their persistence when dry is remarkable; Kobert found active trypsin in the bodies of spiders that had been in the Nuremberg Museum for 150 years, and Seiirt'-^ found that the muscle tissue of mummies contained active glvcolytic ferment. All enzymes as ordinarily prepared have the property of decomposing hydro- gen peroxide, a property possessed by substances of varied nature; this effect is prevented by CNH, which does not prevent other enzyme manifestations, indi- cating that this property is due to an associated enzyme, catalase. The retardation of enzyme action by accumulation of the products of their '9 See Hedin, Ergebnisse d. Physiol., 1910 (9). 433. 20 However, Hosaka (Mitt. med. Gesell. Tokio, 1017 (31), 1) states that frog pancreatic diastase is most active between 5° and 37°, whereas for guinea pig pancreatic diastase the optimum temperature is 27°-55°. Activity begins to be inhibited at 45° and 65° respectively. •^1 See Burge, Amer. Jour. Pliysiol., 1914 (34), 140. " Berl. klin. Woch., 1904 (41), 1081. " Jour, of Physiol., 1906 (34), 207. 24 Gudzent (Zeit. Strahlenther., 1914 (4), 666) denies that radium acts on enzymes. " Agulhon, Ann. Inst. Pasteur, 1912 (26), 38; Burge etal, Amer. Jour. Phvsiol., 1916 (40), 42C.. 2« Amer. .lour. Physiol., 1909 (25), 81. " Berl. kliii. Woch., 1904 (41), 497. TOXICITY or ENZYMES 55 action is simply explained as being due to establishment of equilibrium ; in some instances, however, the substances produced are of themselves harmful to the enzymes, e. g., alcohol and acetic acid. ('han^f'S in reaction, fixation of the enzyme by cleavage products, and other side reactions may also be at least partly responsi- ble. There is a periodicity in enzyme action which makes quantitative results uncertain.'-'* Activation of Enzymes. — Within the cell, the enzymes — at least those that are excreted, such as trypsin and pepsin — exist with few exceptions in an inactive form, tiie zymogen. Their activation appears to take place normally only after they have been discharged from the cell, but after the death of an organ it may result from the decomposition products that are formed. Under physiological conditions this activation appears to he brought about by special activating substances. In the case of the pancreas it is the cnterokinase, which is furnished by the epithelial cells of the intestine. Enterokinase appears to unite with tr3'psinogen to form an active enzyme, which reminds one of the way that comple- ment and the intermediarj- body unite to form hemolytic and bacteriolytic substances.-^ Kinnses, having the same action as enterokinase upon the trypsinogen, are found in various tissues and organs, but generally much less active than the enterokinase. It is verj^ probable that it is through this mechanism that the rate of enzyme action is modified, and perhaps it is a means of defense of the body against its own enzymes; as the prozymes are more resistant to harmful agencies than the enzymes, it also may be a method of storage. The activity of various enzymes is greatly increased by certain more or less specific substances, referred to usually as "coenz^-mes;" thus bile-salts act as co-enzymes for lipase (Loevenhart). The Toxicity of Enzymes Although present normally in greater or less amounts in all the cells in the body, when artificially isolated and injected directly into animals nearly all enzymes seem to be extremely toxic. As foreign proteins, especially extracts of tissues, are generally more or less toxic, it is difhcult to state how much of the toxicity of a given enzyme- containing solution depends on the enzj^me and how much on the admixt proteins. The following statements are taken at the face value placed on them hy the several investigators quoted, and are subject to discount until the enzymes have been isolated and investigated in a 'pure condition, if such a thing shall ever become possible. The first thorough study of the toxicity of enzymes was made by Hildebrandt,^" who found that pepsin, invertase, diastase, emulsin, mja-osin, and rennin were all toxic. The symptoms produced in dogs were trembling, uneasiness, difficulty in walking, and finally coma. The anatomical changes observed were: numerous hemorrhages throughout the body, fatty degeneration of the liver and myocardium, renal congestion, and numerous thromboses. Considerable fever re- sults, and Mayer considers this responsible for the relative harmless- ness of rennin, the action of which is impaired above 40°. That these effects are due to the enzymes themselves rather than to contaminating =8 GroU, Nederl. Tijdschr. v. Geneesk., 1918 (1), 1085. 2^ BayUss and StarUng (Jour, of Physiol., 1905 (32), 129), question the anal- ogy of zymogen-kinase combinations to complement-amboceptor combination. Walker, however, finds evidence that many enzymes consist of a specific ambo- ceptor and a non-specific complement or kinase (Jour, of Physiol., 1906 (33), p. xxi.) 30 Virchow's Archiv, 1890 (121), 1. 56 ENZYMES bacteria is shown bj^ Kionka and by Achalme^^ who obtained similar results with enzymes made sterile by filtration through porcelain. "Wago^2 obtained also an amyloid-like degeneration widely spread in animals injected with filtered solutions of commercial trj'psin, pan- creatin and amylopsin. Achalme found that such sterile prepara- tions of pancreatic juice injected subcutaneously into guinea-pigs produce a marked local pink gelatinous edema, followed by gangrene; if the animal dies, the blood is non-coagulable. Apparently cells of nearly all types can be destroyed by trypsin, which may cause necrosis in one-fourth hour; however, spermatozoa and surface epithelium resist strong trypsin solutions. Intravenous injections cause death with lesions in the heart muscle and severe hemorrhages. After recovery from one injection of trypsin the animal is temporarily somewhat more resistant to another injection, and there are other resemblances to anaphylactic intoxication (Kirchheim^^). Fiquet^^ also observed that trypsin and pepsin rendered the blood incoagulable, but after some time the coagulability of the blood is increased and thrombosis is frequent. Wells^^ found that pancreatic extracts containing very active trypsin and lipase, injected intraperi- toneally, produced an acute inflammatory reaction, but no fat necrosis. Extracts containing active lipase and inactive trypsin were less toxic, but produced fat necrosis. Extracts of liver and blood serum, rich in lipase, were almost without effect on dogs and cats. Pa-pain was found to be much more toxic than any animal enzyme, causing violent local hemorrhagic inflammation. Schepilewsky'^ also found papain much more toxic than rennin and pancreatin; repeated injection of the two latter caused amyloidosis in rabbits. Active immunity does not follow repeated injections of papain. ^"^ Lombroso^^ found that inactive pancreatic juice was much less toxic than the activated, showing that it is the trypsin that is the important toxic agent. He also found that succus entericus in doses of 1 to 5 c.c. is toxic, but not lethal for dogs. Pancreatic lipase is hemolj'tic (Noguchi^^) if activated by fats, which suggests that when this enzj^me gets into the blood it may cause hemolysis. Urease has a definite toxicity because it decomposes the urea in the blood and tissues, fatal intoxication from NH3 poisoning resulting.*" Hildebrandt*' observed that enzymes were positively chcmotactic, but it is probable " Ann. d. I'lnst. Pasteur, 1901 CIS), 737. 32 Arch. Int. Med., 1919 (23), 251. " Arch. exp. Path. u. Pharm., 1911 (66), 352; 1914 f78), 99; 1913 (74), 374. ^' Arch. d. M('d. Exper., 1899 (11), 145. 36 Jour. Med. Research, 1903 (9), 92. '8 Cent. f. Bakt., 1899 (25), 849. " Stenitzer, Biochem. Zeit., 1908 (9), 382. 38 Abstract in Biochem. Centralblatt, 1903 (1), 712. 3» Biochem. Zeit., 1907 (6), 185. " See Carnot and Gerard, Compt. Rend. Acad. Sci., 1919 (169), 88. <» Virchow's Arch., 1893 (131), 5. ANTI-ENZYMES 57 that the prcxkicts of their action on the tissues are the chief chemo- tactic agents. The enzymes that are secreted into the gastro-intcstinal tract seem to be chiefly destroyed, but part is ehminated in the feces, and part that is absorbed apparently reappears in the urine in very small quantities. *2 Pepsin, diastase, and rerinin all have been found in- nor- mal urine; but trypsin is present chiefly as trypsinogen, especially abundant after a meat diet.*^ Pepsin and rennin enter the urine as the zjmiogens, in quantities in proportion to the amount in the stom- ach, and are absent in gastric carcinoma (Fuld and Hirayama**). During resolution of pneumonia, leucocytic protease may appear in the urine (Bittorf*^). Ferments injected subcutaneously are said seldom to be eliminated in any considerable amounts in the urine, but Opie^^ has demonstrated the presence of lipase in the urine in pancreatitis with fat necrosis, and Wago^^ found that injected trypsin is excreted rapidly and abundantly. Hildebrandt was able to prove that emulsin remained active for at least six hours after it was injected into animals subcutaneously, by its splitting amj^gdahn which was then injected, the CNH hberated by the cleavage of the amygdalin causing death. Anti-enzymes Injection of enzymes into animals leads to the appearance of sub- stance^ in the serum of the animals that antagonize the action of the enzymes.*^ The principles involved are quite the same as in the im- munization of animals against bacterial toxins or against foreign proteins. This seems to have been first observed by Hildebrandt, and it has been taken up extensively in recent years in the study of the problems of immunity\ An interesting observation that was made rather early in these studies was that normal blood-serum possesses a marked resistance against the action of proteolytic enzymes, not being at all digested by dilutions of enzymes that will rapidly digest a serum that has been heated. This property seems to be shared by egg-white*^ and by the tissues and organs of the body (Levene and Stookey^"). The anti-enzyme action is easily destroyed by heat of about 70°, by the action of dilute acids, and even by prolonged standing. It is ^2 Falk and KoUeb, Zeit. klin. Med., 1909 (68), 156. "^v. Schoenborn, Zeit. f. Biol., 1910 (53), 386. " Berl. klin. Woch., 1910 (47), 1062. « Deut. Arch. kUn. Med., 1907 (91), 212. « Johns Hopkins Hosp. BuU., 1902 (13), 117. " Jour. Immunol., 1919 (4), 19. *8 According to Porter (Quart. Jour. Exper. Physiol., 1910 (3), 375) enzymes in contact mth various membranes are inactivated, and substances appear which are strongly inhibitive to the enzymes; it is possible that this effect depends largely on zymoids, which unite -wath the substrate and deviate the enzymes. " Sugimoto, Arch. exp. Path., 1913 (74), 14. " Jour. Medical Research, 1903 (10), 217. 58 ENZYMES exerted not only against the secreted proteolytic enzymes, pepsin and trypsin, but also against the intracellular enzymes of various organs. We therefore distinguish between normal and immune anti-enzymes. It seems highly probable that the resistance of the body tissues to digestion by their own enzymes and by the enzymes of one another depends in some way upon the presence of anti-enzymes in the cells and tissue fluids, for self-digestion of tissues is greatly impeded by serum. ^^ Weiland^^ h^g demonstrated that certain intestinal worms contain a strong antitrypsin, to wliich he attributes their ability to live bathed in pancreatic juice without being digested. ^^ Similar properties have been ascribed by other observers to the cells of the mucosa of the stomach^'* and intestine, and to the mucus itself (de Klug),^^ but the work of Bensley and Harvey^^ indicates that the absence of free acid in the gland cells and lumen is perhaps the chief protection of the stomach from pepsin, Kirchheim^^ holds that the intestines are protected less by anti-enzymes than by rapid absorption and removal of the enzymes, which are really not present in any con- siderable excess in the intestinal contents. The anti-enzjmies seem only to inhibit enzyme action, and not to destroy the enzjane itself.^* Normal anti-enzymes do not seem to be at all specific, according to V. Eisler;^^ that is, human serum is no more resistant to human tryp- sin than is pig serum — ^indeed, it is less so.^" Cathcart^^ found that normal antitrypsin is connected with the "albumin fraction" of the serum, i. e., the fraction precipitated between half and full saturation with ammonium sulphate. Globulins do not possess this action, but they are not easily digested. Antitrypsin is found in all varieties of serum, and is little or not at all specific. It is destroyed by Q5-70°C.^^ for one-half hour, but retains its anti-en zj^matic activity after drying, and is equally effective against all sorts of pro- teins. The normal anti-tryptic activity decreases during fasting and " Wells, Jour. Med. Research, 1906 (10), 149. *'■* Zeit. f. Biol. 1903 (44), 45; see also Dastre and Stassano, Compt. Rend. Soc. Biol., 1903 (55), 130 and 254; and Hamill, Jour, of Physiol., 1900 (33), 479. ^^ Burge (Jour. Parasitol., 1915 (1), 179) suggests that the protection of para- sites, and perhaps of the alimentary epithelium, depends on the active oxidative properties of these tissues destroying the enzymes. 6' See Blum and Fuld, Zeit. klin. Med., 1906 (58), 505; Langenskiold, Skand. Arch. Physiol., 1914 (31), 1. " Arch, internat. d. physiol., 1907 (5), 297. " Biological Bulletin, 1912 (23), 225. " Arch. exp. Path. u. Pharm., 1912 (71), 1. 68 Bayliss and Starling (Jour, of Physiol., 1905 (32), 129; and Meyer, Biochem. Zeit., 1909 (23), 68, oppose the view of Delezenne that the antitryptic action of the blood is due to an antikinase, and believe the antibody acts upon trypsin. " Ber. d. Wien. Akad., 1905 (104), 119. «" This is contradicted by Glaessner, Hofmeister's Beitriige, 1903 (4), 79. 8' Jour, of Physiol., 1904 (31), 497; also see Kiimmerer and Aubry, Biochem. Zeit., 1913 (48), 247. *- Unless otherwise specified, all temperatures are given'according to the Centi- grade scale. ANTI-ENZYMES 59 increases during digestion (Rosenthal^^) ; it is increased during preg- nancy"'* and the blood of the fetus shows less than that of the mother. Normal antitrypsin unites with trypsin according to the law of mul- tiple proportions (Meyer) and the reaction is not reversible (Ronfloni). It is found in the urine, and in infianunatory exudates, but not in normal serous fluids, and it resists putrefaction. Normal serum does not seem to inhibit the enzymes which act upon purines. Fuld and Spiro"* found that the natural antirennin of normal horse serum is in the pseudoglobulin fraction. Since acids destroy the anti-enzyme property of the serum, it is not effective against pepsin-HCl mixtures. Against trypsin, however, it is very effective. Zunz^*^ states that nor- mal serum acts more upon enterokinase than upon trypsin, and believes that the inhibition depends upon colloids which modify surface ten- sion and adhere to the proteins. Red corpuscles and living unicellular organisms, including bacteria, are likewise resistant to trypsin, and normal serum also seems to contain an antirennin. ^^ Oppenheimer and Aron^* consider it probable that the re- sistance of normal serum to trypsin digestion depends upon the con- figuration of the protein molecules, which perhaps, when in fresh, uninjured condition, present no suitable surfaces for attack by the ferment. Hedin attributes antitrj^ptic action to adsorption of the enzyme by some constituent of the serum, much as charcoal inhibits tryptic digestion. Fresh and inactivated serum will prevent pepsin from digesting protein, but this is not due to a true antipepsin, according to Ham- burger."^ Jobling^" and his co-workers have advanced evidence that the nor- mal antiprotoase action of serum depends on the lipoids of the senuii,^^ which var}^ in activity directly with the degree of unsaturation; there- fore they were able to decrease the antiferment action of serum by extracting the lipoids with fat solvents (and to restore the activity by replacing the lipoids), or by saturating the double bonds of the fatty acids with halogens, or by modifying the degree of dispersion 8' Folia Serologica, 1910 (6), 285; also Franz and Jarisch, Wien. klin. Woch., 1912 (25), 1441. " See Franz, Arch. f. Gyn., 1914 (102), 579. «5Zeit. f. physiol. Chem., 1900 (31), 132. *8 Mem. Acad. roy. med. Belgique, 1909 (20), fasc. 5. " Czapek (Ber. Deut. botan. Gesell., 1903 (21), 229) states that anti-oxidases occur normally in certain plants, strongly specific against the oxidase of the same plant species. °8 Hofmeister's Beitrage, 1903, (4), 279. «3 Jour. Exper. Med., 1911 (14). 535; Arch. Int. Med., 1915 (16), 356. There seems to be no relation between the antipeptic and antitrvptic powers of sera (Rubinstein, Ann. Inst. Pasteur., 1913 (27), 1074). I 1 '" Series of articles in Jour. Exper. Aled. ; also review in Jour. Lab. and Clin. Med., 1915 (1), 172. See also Zeit. Immunitat., 1914 (23), 71. " Yamakawa (Jour. Exp. Med., 1918 (27), 689), liowever, does not believe that the antienzyme which prevents autolysis of serum itself is of lipoidal nature. 60 ENZYMES of the lipoids. Soaps of saturated fatty acids do not inhibit serum protease. Opie^2 has found that the serum of inflammatory exudates con- tains an anti-enzymatic substance, destroyed at 75° and by acids; it is not present in normal cerebrospinal fluid, but appears here as in other serous cavities during- inflammation (Dochez).''^ Antitrypsin has also been found in pathological urines (v. Schoenborn).^'* The power of the blood serum to inhibit the activity of trypsin and leucocytic protease has been found to vary greatly in disease, and, as having diagnostic possibilities, this property has been considerably investigated.''^ It is especially increased in conditions associated with cell destruction, such as pneumonia and cancer, which suggests that the increased antitryptic activity results from the formation of specific antibodies for the intracellular proteases liberated during the disease, but as yet this has not been satisfactorily established, so we do not know whether the "antitrypsin reaction" depends upon an antibody for trypsin or upon some entirely different factor. In cachexia the inhibiting effect of the serum is especially marked and it is therefore usually pronounced in cancer, but the increased inhibition is some- times absent in cancer (10 per cent, of all cases) and often present in other conditions, so that the positive diagnostic value is slight. It may also be present without cachexia and often seems to parallel the number of leucocytes in the circulating blood. Sarcoma shows it less than carcinoma, while in exophthalmic goitre and tuberculosis an antitryptic increase is said to be quite constant (Waelli).''® In pregnancy there is usually an increase demonstrable after the fourth to sixth months, continuing until two weeks after delivery, and highest in cases of pregnancy toxemias (Ecalle)." Severe traumatism may also cause an increase.''^ As normal serum contains a tryptic enzyme as well as a substance inhibiting trypsin, the antitryptic activity is at most but a measure of the difference between these (Weil), and might depend on either lowered trypsin or increased antitrypsin content. Doblin^^ and many others believe with Jobling that the active agent is not a true immune antibody, but as yet general agreement has not been reached on this point (see Meyer). Kirchheim*" has found that the union of trypsin and antitrypsin does not follow the physico-chemical laws to a true antigen-antibody reaction. Rosenthal has advanced evidence ^2 Jour. Exp. Med.; 1905 (7), 316. "Jour. Expcr. Med., 1909 (11), 718. ''* Zeit. f. Biol, 1910 (53), 386. ^^ For literature and review see Wiens, Ergebnisse Phj'siol., 1911 (15), 1; Weil, Arch. Int. Med., 1910 (5), 109; Meyer, Folia Serologica, 1911 (7), 471. '« Mitt. Grenz. Med. u. Chir., 1912 (25), 184. " Arch. Mens. Obst. Gvn., 1917 (6), 97. " Zunz and Govaerts, C. R. Soc. Biol., 1918 (81), 146. '» Zeit. f. Immunitilt, 1909 (4), 229 80 Arch. exp. Path., 1913 (73), 139. ANTI-ENZYMES 61 to support the hypothesis that the presence of ])ro(hicts of protein cleavage in the serum is responsible for the antitryptic action, but this has not been confirmed. Attempts have been made to regulate sup- purative processes by the introduction of either leucocytic proteases, or antiprotease in the form of active serum (see Wiens^^). Whether antiprotease can be specifically developed by immunizing with leuco- protease is a matter of disagreement,^^ but no increase of antiprotease follows the enormous destruction of leucocytes caused by injecting thorium- A'. ^2 The anti-enzymatic property obtained in the serum by injecting enzj'mes into animals differs from that normall}'" present in the serum in many ways. It may be made much stronger than it ever is in normal serum, and against many varieties of enzymes for which an anti-enzjaiie does not naturally exist. Especiall}^ important is the fact that it is highly specific (v. Eisler); serum of an animal immu- nized against dog trypsin will show a much greater ef5"ect against dog trypsin than it does against trypsin from other animals. This fact permits us to distinguish between enzymes of apparently similar nature but of different origin, and proves that they have a struc- ture at least in some respects different from one another, since they are combined by different antibodies. Apparently that element of the enzymes which determines their action on specific substances is involved in their antigenic properties, since antiproteases will not inhibit diastase or lipase. This specificity is limited, however, for the anti-enzymes for leucocytic and pancreatic proteases are said to be identical. ^^ Artificial immune serum is said to have been obtained against trj^psin, pepsin, ^^ lipase, emulsin,^'^ autoh-tic enzymes, laccase, amylase, invertin, diastase, tjTOsinase, urease,^^ rennin, catalase, and fibrin ferment. ^^ By immunization against bacteria an immunity against their proteolytic enzymes is also obtained, ^^ which is inde- pendent of and different from antitr3^psin, being especially in the globulin fraction, while the antibody for pancreatic trypsin is chiefly in the albumin (Kammerer^^). From the work of Kirchheim and Reinicke^^ it seems probable that the increased resistance following 81 See Bradley, Jour. Hyg., 1910 (12), 209. *- G. Rosenow and Farber, Zeit. exp. Med., 1914 (3), 377. 8' Jochmann and Kantorowicz, Mlinch. med. "Woch., 1908 (55), 728. «^ Bayliss (Jour, of Physiol., 1912 (43), 455) was unable to obtain antiemulsin, and Pozerski (Ann. Inst. Pasteur, 1909 (23), 205) failed to obtain antipapain, but positive results are reported by v. Stenitzer (Biochem. Zeit.. 1908 (9), 382). ** Jacoby says that the disappearance of urease from the blood after repeated injection does not depend on the formation of an antienzyme (Biochem. Zeit., 1916 (74), 97). 8* For a review of much of the earlier literature on this subject see Schiitze, Deut. med. Woch., 1904 (30), 308. 8' Dungern, Miinch. med. Wochenschr., 1898 (45), 1040; Bertiau, Cent, f . Bact., 1914 (74), 374. s8 Deut. .\rch. klin. Med., 1911 (103), 341. «9 Arch. exp. Path., 1914 (77), 412. 62 ENZYMES immunization with trypsin is simply an increase in nonspecific resist- ance, such as follows injection of peptone and man}' other poisonous substances. Wago*^ was able to demonstrate precipitins and com- plement fixing antibodies in antitryptic sera that were not strongly antienzymatic, and Young^° was unable to produce antitryptic sera by immunizing with trypsin-, in spite of the presence of active precipitins for the injected trypsin solutions. There is, indeed, a growing suspicion that much of the evidence of specific antibodj^ formation for enzymes must be revised. Resemblances of Enzymes and Toxins. — As can be seen from the al)ove state- ments, the enzymes behave in many respects lilce the toxins, both in tlieir manner of acting upon other substances and in the reaction they produce when introduced into the bodies of animals. As Oppenheimer says, "the bonds between enzymes and toxins are drawing closer and closer." According to some experiments, the enzymes behave much as if they possessed a haptophore and a toxophore group, the former of which combines with the substance that is to be acted upon; and immunity appears to be produced by the development of receptors that combine the haptophore groups, these receptors constituting the antiferments. There is abundant evidence of a toxin-like structure in enzymes, from the numerous ob- servations on the formation of "zymoids" which can neutralize anti-enzymes or combine with the substrate, although no longer active as enzymes. The oxidizing enzymes especially, with their complex relationship of substrate, com- bining body (peroxides) and enzyme, present striking analogies to immune reac- tions (Moore^')? and the proteolytic substances of the blood resemble the lysins in certain respects (Dick).'^ Enzymes and toxins also resemble one another in being readily absorbed by membranes, precipitates, and highly developed sur- faces in general.^' Finally, there is much reason to believe that the hemolytic toxin of cobra venom is a lipase, which acts by splitting lecithin into hemolytic substances (Coca).^^ THE INTRACELLULAR ENZYMES « Until a recent time our knowledge of enzymes in the animal body was limited to those present in the digestive secretions. With few exceptions these are without influence in pathological processes, since they seem to be but little absorbed, and rarely enter the blood or tissues in any other way. But with the more recently disclosed intra- cellular enzymes, many of which are present in every cell,^^ the rela- tion to pathology is very intimate. These intracellular enzymes, as we now know them, and their chief properties, are as follows: s" Biochem. Jour., 1918 (12), 499. 9' Biochem. Jour., 1909 (4), 165. »2 Jour. Infectious Diseases, 1911 (9), 282. 9' See Porter, C,)uart. Jour. Exp. Physiol., 1910 (3), 375. 9' Jour. Infect. Dis., 1915 fl7), 351. ^'' Sv.G Vernon, Ergebnissc d. Pliysiol., 1910 (9), 138; also liis monograph, "Intracellular Enzvmes," London, 1908. 9« Hcrlitzka (Arch. ital. biol., 1907 (48), 119) and others have shown that the diiferent enzymes appear one by one in the development of tiie ovum. Their activity is modified considerably by infections (Siel)er, Hiochem. Zeit., 1911 (32), 108) aiid other diseases ((irossinanh, ibid., 1912 (11), ISl). OXIDIZING ENZYMES ()3 OXIDIZING ENZYMES" Although oxidation of organic compounds is the chief sour(;c of energy in the animal body, yet the way in which it is accomplished is very little understood. We only know that it is brought about within the cells, and that substances that outside the body are oxidized with difficulty, are completely oxidized to car})on dioxide and water within the cells, and that this is done with just such a degree of rapid- ity that the heat produced is in exactly the amount necessary for the wants of the bod3^ There can be little question that this oxida- tion is accomplished through catalytic agents acting within the cells, and certain of them have been placed in a condition permitting of study. As yet their exact relations to intracellular oxidation are not clearly defined, but for the present they may be grouped pro- visionally as oxidizing enzymes. That some of them arc highly specific is shown by those disorders, such as alkaptonuria and diabetes, in which the body loses the power to oxidize a certain chemical substance while retaining the normal power to oxidize innumerable other sub- stances. According to Lillie^^ the oxidative processes in cells take place most actively in relation to the membrane surfaces (or phase boundaries) of the cells. Of the oxidizing enzymes as yet identified none can be considered as of importance in the energy-producing oxidations of the body (Battelli and Stern), all the enzymes of this class yet known being apparently concerned with less essential oxidiz- ing processes; it is indeed possible that the essential oxidation of food-stuffs ma}^ not be dependent on enz3'mes (Engler and Herzog).^^ An agent accelerating the essential oxidizing activities of the tissues has been described by Battelli and Stern' under the name of pnein, and an anti-pneumin which holds it in check. Closely related to the oxidiz- ing enzj^mes is — • Catalase. — It has long been known that most enzymes possess the power of decomposing hydrogen peroxide, with liberation of oxygen; but it was not until 1901 that it was finally demonstrated by Loew that this property was due to a separate enzyme and was inde- pendent of the specific properties of the various other enzymes. This ferment is very wide-spread, and so is generally obtained along with the other enzymes when attempts are made to isolate them from the cell. It was named catalase by Loew, and he described two forms, a- 9^ Complete bibliography and exhaustive discussion by Kastle, Bull. Hygienic Lab.,- No. 59; by Loele, Ergeb. allg. Path., 1912 (16, Pt. 2), 760; and by BattelU and Stern, Ergebnisse d. Phj-siol., 1912 (12), 96. Concerning the chemistry of vital oxidations see Dakin, "Oxidations and Reductions in the Animal Body," Monographs on Biochemistry, London, 1912. Good review by v. Fiirth. "Chem- istry of Metabolism," Chaps. 22 and 23; translated by A. J. Smith, Philadelphia, 1916. 38 Jour. Biol. Chem., 1913 (15), 237. 93 Zeit. phvsiol. Chem., 1909 ^59), 327. iBiochem. Zeit., 1911 (33), 315; 1911 (36), 114. 64 ENZYMES catalase, which was thought to be a nucleoprotein,'^ and ^-catalase, which has the properties of an albumose. It has been demonstrated by Bach and Chodat that peroxides are contained in plant cells, and they also occur in animal cells. According to Golodetz and Unna^ the catalases are held in the cytoplasm of the cells while the peroxida- ses are in the nucleus. Just what function the catalase performs is at present merely a matter of speculation, but that it serves an important purpose is indicated by the observation of Burge* that the amount of catalase in tissues varies directly with their activity. He also as- cribes the specific dynamic action of proteins to their causing an increase in blood catalase. Becht,^ however, questions the validity of the evidence so far brought forward in support of the hypothesis that catalase is essentilly responsible for tissue oxidation.^ Loew considers that it destroys peroxides formed in metabolism, which|are]very poisonous to cell life. Shaffer has found evidence that under the influence of catalase the oxygen liberated is in the molecular form, O2, and therefore relatively inert; whereas when peroxides spontaneously decompose, they liberate atomic oxygen which is an active oxidizing agent. He found that uric acid is oxidized by per- oxide of hydrogen, but when catalase is present, this oxidation is prevented. According to this the function of catalase is rather to prevent dangerous forms of oxidation than to help in normal oxi- dative proceses. For the present, however, nothing can be said positively on this subject. Occurrence of Catalase under Normal and Pathological Conditions.'' — Battelli and Stern found that the catalytic power of the tissues endures many hours after death. Its abundance is different for different organs of the same animal, but remarkably constant for the same organ in the same species. In general the order in'^decreasing strength is: liver, kidney, blood, spleen, gastro-intestinal mucosa, salivary glands, lung, pancreas, testicle, heart, muscle, brain; but this order varies in different species. Catalase is abundant even in the early embryo (^lendel and Leavenworth) and in sea urchin eggs it increases rapidly after they are fertilized (Lyon). 8 Leucocytes contain little, most of that in the blood being in the stroma of the red blood-corpuscles. The body fluids contain little or none. Injected intravenously, catalase (of the liver) is destroyed rapidly, and docs not appear in the'-urine; it does not cause any toxic effects, nor does it increase resistance to poisoning by venoms. The tissues also contain anti-catalases, and still further a substance which protects the catalase froni the anti-catalase; this protective sub- stance is called the pliilocatalase by Battelli and Stern. The gas evolved by the action of pus on H2O2 was found by Marshall' to be pure oxygen, each c.c. of a certain sample of pus examined liberating 133.9 c.c. of gas. The active constituent of pus, he states, is contained in the serum and not in the 2 Not corroborated by Waentig and Gierisch, Fermentforsch., 1914 (1;, 165. ' Berl. klin. Woch., 1912 (49), 1134. •' vVmer. Jour. I'liysiol., 191G (41), 153; 1917 (42), 373; 1919 (48). 133. See also Alvarez and Starkweather, ibid.; 1918 (47), GO; Dutcher, Jour. Biol. Chem., 1918 (36), 63. 5 Amer. Jour. Physiol., 1919 (48), 171. 8 See also Stelilc, Jour. Biol. Chem., 1919 (39), 403. ^ Concerning the catalase of lower animals see Ziegcr, Biochem. Zeit., 1915 (69), 39. 8 Amer. Jour. Physiol., 1909 (25), 199. » Univ. of Pcnn. Med. Bull., 1902 (15), 366. OXIDIZING ENZYMES 65 corpuscles. Catalase is abundant in the tissues of lower animal forms, e. g., Ascaris.^" Substances decomposinp; HoOj have been found also in bacterial cul- tures, first by Gottstein, and later in the cell juices expressed from tubercle bacilli by Hahn. Locwenstein" found an enzyme aRreeinp; with catalase in filtered bouillon cultures of diphtheria bacilli and staphylococci, l)ut not frorn tetanus, typhoid, and colon bacilli or cholera vibrios; the catalase is quite distinct from the toxin. He also found that the addition of H2O2 to a diphtheria toxin-antitoxin mixture destroyed the toxin, leaving the antitoxin free. A similar destruction of tetanus toxin l)y peroxides, first demonstrated by Sieber, can occur without the catalase. Winternitz^- and his associates have made extensive studies of the catalase activity of the blood and tissues in disease. They found that all tissues have re- duced catalase activity in chronic nephritis, in proportion to the severity of the condition, and experimental nephritis in animals has the same effect; the blood shows great reduction in catalase in vu-emia, and a less reduction with less severe nephritic manifestations. Eclampsia shows little or much reduction of catalase in the blood in proportion to the amount of renal involvement; normal pregnancy and labor have no elTect. Anemia is associated with irregular decrease in catalase, including primary anemias and the secondary anemias of typhoid and pneumonia; cardiac disease has no effect if the kidneys are normal. Acute peritonitis causes a rise in blood catalase; diabetes, leukemia and jaundice were w-ithout effect. In hyperthj-reosis the catalase tends to increase, in hypothyreosis to decrease; com- plete removal of the thyroid causes a decrease which disappears on feeding thy- roid. Intravenous injection of salts, acids and alkalies decreases the catalytic activity of the blood. In shock, blood catalase is decreased." Normal indi- viduals show considerable variations in the catalase activity of the blood, but for each individual it is remarkably constant; age has very little influence. In the tissues post mortem change causes but slight reduction in catalase. Extirpation of large amounts of kidney or liver tissue has little effect, but removal of the spleen, ovaries or testicles causes a transient decrease in the catalase of the blood. In diabetes and starvation, tissue catalase is said to be decreased.'^ If the red cor- puscles are prevented from laking, the catalase activity manifested by the blood in vitro is reduced (Strauss)'* and iodides increase the catalase activity of the blood. Catalase and anticatalase have been found in pathological urine, in both acute and chronic nephritis (Primavera).'^ Kahn and Brim'^ also found traces of cata- lase in normal urine, greatly increased in urine containing blood, bile or acetone, normal in cancer, high in diabetic acidosis, Hodgkin's disease, septic infections and typhoid. Grossman'* foimd that bacterial poisons generally increase the catalase content of the various viscera, and Rosenthal'^ observed a great decrease in the liver and blood of mice receiving intraperitoneal inoculations of cancer. The catalase activity of the non-cancerous organs of cancer patients is not affected, except slightly lowered by cachexia (Colwell) ;'° however, the liver tissue between cancer nodules may show less catalase than normal liver. •^' In phosphorus poi- soning the catalase content of the liver, heart and blood is decreased (Burge).'-' But it is to be borne in mind that the questionable accuracy of our existing methods of determining quantitatively the amount or activity of catalase in tis- sues makes the foregoing statements of uncertain value. True Oxidizing Enzymes. — While it is by no means certain that catalase is active in causing intracellular oxidations, there are other '0 Magath, Jour. Biol. Chem. 1918 (33), 395. " Wien. klin. Woch., 1903 (16), 1393. 12 Review in Arch. Int. Med., 1911 (7), 624. »3 Burge and Neill, .\mer. Jour. Physiol., 1918 (45), 286. 1* Burge, Science, 1918 (47), 347. '6 Bull. Johns Hopkins Hosp., 1912 (23), 120. '6 Riforma Med., 1906 (12), 1266. " Amer. Jour. Obst., 1915 (71), 39. 18 Biochem. Zeit., 1912 (41), 181. 19 Deut. med. Woch., 1912 (38), 2270. 20 Arch. Middlesex Hosp., 1910 (19), 64. 21 Blumenthal and Brahn, Zeit. Krebsforsch., 1910 (8), 436. " Amer. Jour. Phvsiol., 1917 (43), 545. 66 ENZYMES enzymes or enzyme-like substances that come more properly under the head of oxidases or oxidizing enzymes. Battelli and Stern contend that the only real oxidases which have yet been completely established are: 1. Polyphenoloxidases (oxidizing phenols and their amino com- pounds, but not tyrosine); 2. Tyrosinase; 3. Alcohol oxidase; 4. Xanthine oxidase; 5. Uricase. Chodat and Bach believe that the enzymes which are designated above as polyphenoloxidases have a com- plex structure, consisting of peroxidase and oxygenase. ^^ Mathews^* holds that "under the term oxidases there have been confused two classes of substances, one which activates the oxygen; the other the more important class, which activates, by dissociation, the reducing substances. The latter are specific, the former not." This view has received support by Bach. Peroxidase. — This name is given to an enzyme that is believed to cause oxida- tion by activating peroxides, and is quite distinct from catalase and from the other oxidases. The peroxide on wluch it chiefly acts in the cell is supposed to be the so-called "oxygenase." Oxygenase. — This can also act as an oxidizer independent of the peroxidase, in the presence of certain manganese compounds. Loevenhart and Kastle ques- tion the true enzyme nature of this and other "oxidases," which they look upon as organic peroxides, behaving like other peroxides rather than as catalyzers. Prac- tically tlie existence of these bodies is demonstrated by their power to turn tinc- ture of guaiac blue, and they are, therefore, present in pus. Von Fiirth^^ sums up the situation in these words: "In the tissues active cata- lytic agents, the peroxidases, are widely distributed; which seem, just like the coloring matter of the blood, to be capable of conveying the ox3^gen from peroxides to very readily oxidizable substances. We find too in the statements bearing upon the oxygenases, the aldehydases and indophenoloxidases, occasion for assiun- ing that there are substances in the tissues charged with oxygen which are able to give this off to easily oxidizable matter; and these we may in a measure regard as peroxides. But that is all. We do not know whether the peroxidases are ferments or not." By their conception of oxygenase and peroxidase Chodat and Bach would displace entirely the idea of enzymes oxidizing directly, the true "oxidases," which they consider mixtures of oxygenase and peroxidase. There have been, in any event, a number of ferments described that seem to possess distinct oxidative powers. As each is quite specific in its action, oxidizing but one substance, or one group of related substances, they are generally designated by the name of the substances upon which thej^ act. Most studied of these are aUlehytlase and tyrosinase. Aldehydase,^^ which is characterized l)y oxiilizing aldehydes, particularly salicyl-aldehydc. According to Jacquct, this enzyme is so intinuitely bouutl with the cell that it cannot l)e olitained in extracts until after the cells are dead, but is present in expressed cell-juices. It can be isolated by the usual metliods, is de- stroyed by boiling, acts best when no free oxygen is present, and its action is in- hibited 1j3' CNH. It lias been demonstrated in nearly all organs and tissues except pancreas, nuiscle, marrow, and mammary gland; it is present in the blood in small amounts, but not at all in the bile. It is most abundant in the liver-*^ and spleen, and is present in jjig embryos, 9 cm. long, but not in tliose 2-3 cm. long. Jacoby has obtained a body with the properties of aldehydase which did not give protein reactions. It is a true enzyme, since it o.xidizes aldeliydes without itself being ^' See also Onslow, Biochcm. Jour. l'.)l'.) (13), 1. ^-i Jour. Hiol. Chem., 190), 1. ■■'^ Battelli and Stern do not include aldeliydase among the oxidizing enzymes, on the ground tliat its action is not oxidative but liydrolj-tic. ^o BatteUi and Stern, Biochem. Zeit., 1910 (29)," 130. OXIDIZING ENZYMES 67 used up. Its rariRo of actiou is limited, for Jacoby found it mthout effect upon acetic acid and stearic acid. Tyrosinase. — Tliis cnzyine, which is found both in animal and plant tissues, is particularly iiitcrostinu; in relation to the formation of pigments. Bcrtrand found tliat the traiisforiuatiou of the juice of lac-yicldin^!; plants into the black lacquer was liroufiht al)()ul by the action of an oxi), 76. 8' Also discussed under "Diabetes," chap. xxiv. As glycolysis by blood and tissues can occur witliout oxygen, Battclli and .Stern exclude tlie glycolytic from the oxidizing enzvmes. •^'^ Zeit. physiol. ('hem., 1903 (39), 336; also see Simpson, Hiochcin. .lour., 1910, (5), 126. "' For literature ou lii)ase see Connstein, Ergcbnis.se Physiol., 190 I (3, Abt. 1), 194; concerning the l)ehavi()r of lipase sec Tavlor, Jour. BioL ('hem., 1906 (2), 103; Palk, Proc. Natl. Acad., 1915 (1), 136; Science, 1918 (47), 423. LIPASE 71 to the enzyme splitting fats, the triu; lipase, is not yet known. Much of the work so far reported on the occurrence of lipase in tissues is of questionable value, especially as to quantitative results, because of faulty methods. SaxP* points out and avoids some of these errors, and finds that during autolysis of tissues the splitting of the natural fats present in the cells is but slight; simple esters are attacked more, especially amyl-salicylate ; muscle and blood are the least active tissues. Most authors agree that lymphoid cells are especially rich in lipolytic enzymes. ^^ In the serum of normal individuals the esterase content seems to be quite constant,^'' and Quinan" found the tissue content also constant, the liver containing about twice as much as ttie kidnej' and over three times as much as the muscle. He states that different parts of the brain have characteristic lipase activity (butyrase).*^^ Thiele^^ has found that blood, chyle, and various tissues also contain an enzyme which can hydrolyze lecithin, but except in the pancreas i does not hydrolyze neutral fats. The brain contains enzymes hydrolyzing mono- and triacetin, lecithin and cephalin.'"' Little is known about the part played by lipase in pathological con- ditions. According to Achard and Clerc,^' the amount of spUtting of ethyl butyrate by the blood-serum is lessened in most diseases, and in- creases and decreases with the health of the patient; accorchng to Pribram^2 a^id SagaP^ it is increased in the blood during fevers. Clerc^^ found that acute arsenic, phosphorus and diphtheria-toxin poisoning increased this property of the serum, while chronic poison- ing and staphylococcus intoxication lowered it. Somewhat similar results were obtained by Grossmann,''* but Saxl found no increased activity in phosphorus poisoning. Using the ethyl butyrate test, Winternitz and Meloy^^ found that the more nearly normal an organ is the more cleavage of the ester; lipolytic activity is low at birth, increases rapidly during the first few days of life, and does not de- crease in old age. There is a decline in activity of tissues in diabetes, tuberculosis, and the toxemia of pregnancy, in the livers of passive congestion and fatty degeneration, in the pneumonic lung and the cirrhotic liver. After taking food there is a slight increase in esterase, e^Biochem. Zeit., 1908 (12), 343. ^^ The distribution of lipases in different species of.'animals and their^various organs has been investigated by Porter, Miinch. med. Woch., 1914 (61), 1774. « Sagal, .Jour. Med. Res., 1916 (34), 231. «^ Ibid., 1915 (32), 45. " Ibid., 1916 (35), 79. «^ Biochem. Jour., 1913 (7), 275. " English and MacArthur (.Jour. Amer. Chem. Soc, 1915 (37), 653), who have also found in sheep brain, erepsin, amylase, catalase, enzymes decomposing arbutin and salol, probably pepsin and trypsin, but not peroxidase, oxidase, reductase, guanase, urease or rennin. " Compt. Rend. Soc. Biol., 1902 (54), 1144. " Cent. inn. Med., 1908 (29), 81. "Compt. Rend. Soc. Biol., 1901 (53), 1131. '* Biochem. Zeit., 1912 (41), 181. 'Uour. Med. Res., 1910 (22), 107. 7.2 ENZYMES reaching a maximum in three hours. '^^ Whipple''^ finds the blood Hpase (butyrase) increased whenever there is injury to the liver, such as in chloroform anesthesia and puerperal eclampsia; it is lowered in cirrhosis. Poulain^^ found that the butyric-splitting power of lymph-glands draining infected areas was decreased. Fischcr^^ observed, in a case of extreme lipemia in diabetes, that the lipolytic power of the blood was absent. The lipase of lipomas presents no demonstrable difference from that of ordinary fatty areolar tissues.*" Lipase has also been demonstrated in pus by a number of ob- servers,*^ who agree that there is more in exudates than in transu- dates. Zeri*2 found lipase in the urine only when pus or blood was also present, but Pribram and Loewy*^ found it in nephritis, con- gestion, polyuria and other conditions. Lorenzini,** however, re- ports that in albuminuria the lipase content of the urine is reduced, in common with other enzymes, there being a simultaneous accumula- tion of enzymes in the blood. Fiessinger and Marie*^ contend that the lymphocytes of exudates are the chief source of lipase, and suggest that this may be of effect in defense against the fatty tubercle bacilli. Toxins were found by Pesci*® to increase the butyrase but not the other lipases of liver tissue. In syphilis the lipolytic activity of the serum is increased,*' which may be related to Bergell's** observation on the origin of lipase in lymphocytes (corroborating Fiessinger and Marie). Jobling and Bull** state that a specific serum lipase increase occurs in animals immunized to red corpuscles, and that this lipase has to do with hemolysis; but MendeP** found no evidence that hemolj'sis by ricin is related to lipase. Abderhalden and Rona*^ found that excess feeding of fats leads to an increase in the lipase of the blood. The part played by lipase in fatty degeneration must be of great importance, but as yet it has been little considered, except that Loeven- hart, and Duccheschi and Almagia*^ found no appreciable difference '" Jobling et al, Jour. Exp. Med., 1915 (22), 129. " Whipple et al, Bull. Johns Hopkins Hosp., 1913 (2-4), 207 and 357. '« Comp. Rend. Soc. Biol., 1901 (53), 786. '^Virchow's Arch., 1903 (172), 218. '" Wells, Arch. Int. Med., 1912 (10), 297. " Achalme, Comp. Rend. Soc. Biol., 1899 (51), 5GS; Zeri, II Policlinico, 1903 (10), 433; Memmi, Clin. Med. Ital., 1905 (44), 129. 82 II PoUclinico, 1905 (12), 733. 8' Zeit. phvsiol. Chem., 1912 (76), 489. "'Policlinico, 1915 (22), 358. " Compt. Rend. Soc. Biol., 1909 (68), 177. See also Resell, Dout. Arrh. klin. Med., 1915 (118), 179. 8« PatholoM;ica, 1912 (3), 207. 8' Citron and Reicher, B(>rl. klin. Woch., 1908 (45), 1398. 88 Miinch. nicd. Wocli., 1909 (56), 64. 89 Jour. Kxp. Med.. l',M2 (16), 483. «<• Arch. Fisiol., 1<)09 (7), 1()S. "Zeit. plivsiol. Chem., 1911 (75), 30. "2 Arch. Ital. Biol., 1903 (39), 29. AMYLASE OR DIASTASE 73 in the lipase eontoiit of normal and pliosphoius-poisoned livers, but in chloroform poisoning Quinan''' found a decrease in the butyrase of the liver, although it was increased in the kidneys and muscles. This question will be considered more fully in discussing fatty meta- morphosis. An improved method of testing for lipase action has been devised by Rona and ]\Iichaclis, by measuring the change in surface tension caused by hydrolysis of a soluble ester, usually tributyrin. Using this, Bauer'* found that every human serum contains fat-splitting enzymes, which are greatly decreased in carcinoma and advanced phthisis, some- what decreased in syphilis and exophthalmic goitre, and increased in early pulmonary tuberculosis. Caro'^ found a decrease in all cases of cachexia, but there was no relation between the lipolytic enzyme and the blood picture. The blood contains no thermostable antilipase analogous to the antitrypsin. Red corpuscles are said to contain an enzyme sphtting cholesterol esters, "cholesterase."^^ In leucocytes a "lipoidase" has been found by Fiessinger and Clogne" that splits choline out of lecithin. Fat necrosis, resulting from the escape of pancreatic juice into the peripancreatic tissues and abdominal cavitj^ undoubtedly is largely the result of lipase action. (See "Fat Necrosis," Chapter xv, for complete consideration.) Amylase or Diastase"" Although under ordinary conditions starch is not supposed to enter the blood stream and tissues, yet all tissues and body fluids are capable of hydrolyzing starch. Apparently the amylase is derived from the pancreas and salivary glands, and possibly from many or all other tissues (King), but it is not quantitatively related to the amount of carbohydrate in the diet of a species or an individual (Carlson and Luckhardt). In the blood it occurs in the albumin fraction.'* There is disagreement in the literature as to the variations in amount of amylase in the blood during disease, and little information concerning its distribution in the tissues. Normally the kidneys and Uver seem to be most active and Winslow says that all glycogen-containing organs produce diastase. During acute infections the blood amylase is increased, presumably coming from the leucocytes (King). It is greatly increased when the pancreas is acutely inflamed or injured 93 Jour. Med. Res., 1915 (32), 73. sMVien. klin. Woch., 1912 (25), 1376 (bibliography). " Zeit. klin. Med., 1913 (7S), 286. 3«See Cvtronberg, Biochem. Zeit., 1912 (45), 281. 9^ Compt. Rend. Acad. Sci., 1917 (165), 730. "" Literature given by Watanabe, Anier. Jour. Physiol., 1917 (45), 30; Geyelin, Arch. Int. Med., 1914 (13), 96; Stocks, Quart. Jour. Med., 1916 (9), 216; McClure and Pratt, Arch. Int. Med., 1917 (19), 568; Winslow, Hospitalstidende, 1918 (61) 832 ""sSatta, Arch. Sci. Mt'd., 1915 (.39), 46. 74 ENZYMES (Stocks). In diabetes it is ordinarily increased, but not in syphilitic diabetes. ^^ Intravenous or subcutaneous injection of starch is said to increase the blood amylase, presumably as a defensive reaction (Abderhalden), but the amylase ordinarily in the blood seems to be a waste substance on its way to excretion, rather than a functionating enzyme of the blood. There appears to be no normal antiamylase in the blood. Starch granules taken up by phagocytes show a glycogen reaction after some time, suggesting that these cells have intracellular diastases.^ Because of possible diagnostic significance, the amylolytic activity of the urine has been particularly studied, and found normally to be approximately constant for 24 hour specimens of the same individual. - Anything impairing the excretory capacity of the kidney decreases the urinary amylase, although sometimes when the urine contains blood, pus, or much albumen there may be an increased amylase excretion in spite of diminished functional activity. There may be an increase in the amylase in the blood when the urinary amylase is decreased, but with normal kidneys increase of the blood amylase causes an increase in the urine; hence, acute pancreatic diseases cause an increased urinary amylase (Stocks), but this is not constant (McClure and Pratt). In diabetic urine it is said to be usually decreased, but this is mostly accounted for by the dilution of the urine. Parenteral in- jection of starch causes a marked increase in the amount of diastase in the urine (King).^ 99 De Niord and Schreiner, Arch. Int. Med., 1919 (23), 484. > Okazaki, Sei-I-Kwai Med. Jour., 1917 (36), 101. ^ In infants the urine amylase is low (McClure and Chancellor, Zeit. Kinder- heilk., 1914 (11), 483. Fetal blood contains much less than the maternal blood (Kito, Amer. Jour. Physiol., 1919 (48), 481). 3 Proc. Soc.^Exp. Biol., 1917 (15), 101. CHAPTER III ENYZMES (Continued) INTRACELLULAR PROTEASES' (PROTEOLYTIC ENZYMES), INCLUDING A CONSIDERATION OF AUTOLYSIS To what extent synthesis of proteins goes on in the body is still a problem; still more uncertain is the part played by reversible action of proteases. There is evidence enough that somewhere in the body the amino-acids can be rebuilt into protein, for several investigators have succeeded in keeping animals in nitrogenous equilibrium by feed- ing them products of proteolysis that contained no proteins whatever, and as the proteins of the animal body are being broken down in- cessantly, it must be that they were replaced by synthesis of the non- protein material fed to the animals. In addition, it has long been questioned whether amino-acids absorbed from the intestines are not resynthesized into proteins while passing through the intestinal wall. Cohnheim found that in the intestinal epithelium there is an enzyme, erepsin, capable of splitting albumoses and peptones into the amino- acids, which enzyme presumably exists for the purpose of securing complete cleavage of all ingested proteins into their ultimate "build- ing stones." This may be looked upon as a provision to reduce all varieties of proteins to their common elements, so that the body by quantitative selection can resynthesize them into its own types of protein, for, as is well known, foreign proteins {e. g. egg-albumin) introduced directly into the blood stream cannot be utilized, but are excreted unaltered in the urine. ^ As was shown for lipase, the as- sumption that such synthesis occurs as a normal physiological process by reverse enzyme action, requires that the proper enzymes be present in the cells throughout the body, and within recent years it has been abundantly demonstrated that such is the case. For over half a century it has been known that amebse digest solid proteins within their bodies, but it is only within a few years that proteolytic enzymes have been definitely isolated from them. It has been much the same with the intracellular proteases of the higher organisms. In 1871 Hoppe-Seyler referred to the liquefaction of ^ As the possibility exists that ferments which digest proteins may be able to perform a certain amount of synthesis of proteins, the term "proteolytic enzyme" seems to be less suitable than the term "'protease," which merely means an enzyme acting on proteins, and does not compel us to accept any particular view as to what the action is. ^ According to Austin and Eisenbrey (Arch Int. Med., 1912 (10), 305), dogs on a nitrogen-free diet can utilize horse serum injected intravenously. 75 76 ENZYMES dead tissues within the body which occurred without putrefaction, and, as he noted, resembled the effects of the digestive ferments. It was nearly twenty years later that Salkowski^ showed definitely that this softening of dead tissues was really brought about through a true digestion by intracellular enzymes, which produced the same splitting products that were at that time considered characteristic for tryptic digestion (leucine and tyrosine). The process he named "autodiges- tion." This important observation remained almost unnoticed for ten years more, when Jacoby,^ in 1900, took up the investigation of this matter of cellular self-digestion, and after this the importance of the principles involved became for the first time generally appreciated. Jacoby rechristened the process "autolysis," by which name it is now commonly known. AUTOLYSIS^ Autolysis is generally studied by the method used bj' Salkowski, which depends upon the difference in the susceptibility of bacteria and of enzymes to antiseptics. The organs are ground to a pulp, placed in flasks with or without the addition of water or dilute acids, and bacterial action is prevented by the addition of antiseptics that are not poisonous to enzymes — 'toluene and chloroform are most com- monly used. It is possible also to secure organs in an aseptic con- dition and to permit them to undergo autolysis without the use of antiseptics, but the practical difficulties are such that this method is seldom used — it is sometimes designated as "aseptic autolysis, in contradistinction to antiseptic autolysis by the Salkowski method. In a short time it can be seen that digestive changes have taken place, particularly if comparisons are made with control flasks in which the enzymes have been destroyed by boiling. To determine the rate of autolysis the amount of nitrogen that remains in the form of coagu- lable compounds, and that which is converted into soluble, non- coagulable compounds (albumoses, peptones, ammonia compounds, amino-acids, etc.), is compared. The method may be illustrated by a concrete example: A given specimen of emulsionized liver tissue was permitted to digest itself for twenty-two days. At the end of that time 39.4 per cent, of the nitrogen was still contained in the com- pounds that remained insoluble or became so after the autolysis was stopped by boihng; while 60.6 per cent, of the nitrogen was in a soluble form. A control specimen from the same liver was boiled while fresh to kill the enzymes, and then let stand under the same 3 Zeit. f. klin. Med., 1890, supplement to Bd. 17, p. 77. •" Zeit. f. physiol. Cheni., 1900 (30), 149. 5 Resviiiu'; of literature l)v Salkow.ski, Deut.schc Klinik, 1903 (11), 147; also see Schlesinger, HofnieistxT's lieitriiKe, 1903 (4), 87; Oswald. Biochem. Centr., 1905 (3), 365; Levene, Jour. Ainer. Med. Assoc., 1906 (46). 77t); Nicolle, Ann. Inst. Pasteur. 1913 (27), 97; von Fiirth, " Chemistry of Metal)olisiii," Ainer. Transl., 1916. PRINCIPLES OF AUTOLYSIS 77 conditions. In this specimen 90.4 per cent, of tlie nitrogen was in an insoluble form, and 9.6 per cent, was soluble. Therefore, over half of all the protein of the liver had been changed into non- coagulable substances in the course of about three weeks (at 37° C). Complete disintegration of the proteins with liberation of all the amino-acid complexes is probably never reached. Of 45.8 grams of amino-acids present in 100 grams of liver, in ten days' autolysis there had been set free but 1.85 gm., after 30 daj^s 10.1 gm., and after 50 days but 29.1 gm. (Abderhaldcn and Prym.'') Bj' determining the freezing point and conductivitj'' of autolyzing mixtures, valuable evidence can be obtained as to the rate of change, which, in some cases, is much more significant than the usual estimation of soluble and in- soluble nitrogen (Benson and Wells^). Titration of the free amino- acids b}' the formaldehyde method, together with the estimation of proteose and peptone nitrogen, also furnish valuable information, while the Van Slyke method of determining free amino-acids is especially useful for this purpose. Since Jacoby's paper appeared, the field has been invaded by many workers, who have examined practically everj^ tissue in the body, and found that all possess the power of self-digestion; or, in other words, proteases are present in every cell in the hodij} The rate of digestion is very different in different organs, however, liver digesting rapidly while brain and muscle tissue digest much more slowly, aud the auto- lytic activity varies under different conditions;^ thus, fever causes a great increase in the proteolytic activity of the muscles.^*' The char- acter of the antiseptic used modifies greatly the rate, salicylic and ben- zoic acids giving the most rapid autolysis, while of non-acid antiseptics toluene is perhaps the least inhibitory. One of the most important factors in modifying the rate of autolysis is the H-ion cencen- tration developing in the tissues. ^^ Acidity acts, partly, at least, by so modifying the substrate that the enzj^mes can attack it, and a very small excess of acid will destroy the enzymes; Bradley^'- estimates this destructive acidity at about that concentration of H-ions which is indicated by methyl orange and Congo red, the maximum" autolysis being obtained with an acidity at about pH = 6.00. A reaction approximating that of blood (pH = 7.4 — 7.8) reduces autolysis to a minimum. A latent period has been observed before autolysis in «Zeit. physiol. Chem., 1907 (53), 320. ^ Jour. Biol. Chem., 1910 (8), 61. ^ Except, perhaps, the red corpuscles (Pincussohn and Roques, Biochem. Zeit., 1914 (64), 1). * Concerning autolysis of skin, see Sexsmith and Petersen, Jour. Exp. Med. 1917 (27), 273. '" Aronsohn and Blumenthal, Zeit. klin. Med., 1908 (65), 1. Striated muscle autolyzes much less rapidly than cardiac and unstriated. (Bradley, Proc. Am. Soc. Biol. Chem., 1918 (33), xi). " See Morse, Jour. Biol. Chem., 1916 (24), 163. 'Uour. Biol. Chem., 1915 (22), 113; 1916 (25), 261. 78 ENZYMES vitro seems to begin, part of which time may be occupied in the develop- ment of sufficient acidity to permit of autolysis, although Bradley's^^ results indicate that it can be accounted for largely by the time re- quired for proteolysis to proceed far enough to be detected by chemical means. Dernby^'* finds that in several tissues studied, including leucocytes, there are two intracellular proteases, one resembling pepsin in carrying digestion only to the peptone stage and in requiring an acid medium, optimum pH = 3.5; the other resembling ereptase, splitting only peptones and peptids into amino-acids, with optimum reaction pH = 7.8, and inhibited by acid reaction. Autolysis of tis- sues proceeds furthest in a pH range between 5 and 6, presumably because in this condition both enzymes can act. From these facts it is evident that quantitative studies of rates of autolysis are valueless if the H-ion Tconcentration is not taken into consider atio7i. The cleavage products resulting from tissue autolysis seem to contain a much larger proportion of the nitrogen in the form of ammonia and its compounds than is the case with simple tryptic digestion, because of the presence of deaminizing enzymes which split the NH2 groups out of the amino-acids and purines. According to Bostock^^ the greater the acidity the less NH3 is formed. It is quite probable that in tissue autolysis several intracellular enzymes are in action which may not be present in pancreatic or gastric juice; for ex- ample, in the liver is an enzyme, arginase, which splits the urea radical out of the arginine of the proteins (Kossel and Dakin^''), and the enzymes which disintegrate purines are also absent from the digestive juices. On the whole, however, the products are quite similar to those obtained by tryptic digestion. To give a concrete example, Dakin^^ detected in the products of autolysis by the kidney in acid solution, the following substances: Ammonia, alanine, a-aminovalcrianic acid, leu- cine, a-pyrollidine carboxylic acid, phenylalanine, tyrosine, lysine, histi- dine, cystine, hypoxanthine, and indole derivatives, including probably tryptophane.^^ The cleavage of simple peptids by different tissues shows characteristic differences, the distribution of the enzj-me which splits glycyl-tryptophane having been most studied. During life the cells retain this enzyme, and hence it appears in the body fluids only when the tissues are being rapidly disintegrated (Mandelbaum).^^ During autolysis the changes are by no means limited to the pro- teins. Glycogen is split into glucose very early, and the sugar under- goes further changes. Fats are also split by the lipase, fatty acids " Jour, liiol. Chcm., 191G (25), 3G3. !■• .lour. Biol. Chem., 1918 (35), 179. >5 Biochein. Jour., 1912 (6), 388. >6 Zeit. physiol. Chcm., 1901 (42), 181. iMour. of PhysioloKY 1903 (30), 84. 18 The results of autolysis by (iilTcrcnt tissues are said to be quite dissimilar. See Kashi\val)ara, Zeit. i)hysiol. Chcm., 1913 (85), IGl. '9 Miinch. med. Woch., 1914 (01), 401. PRINCIPLES OF AUTOLYSIS 79 being found in iiutolyzod organs. Reducing substances appear, and as before mentioned, numerous volatile fatty acids are said to be produced. ^Vluch doubt exists concerning the supposed formation of volatile fatty acids and gasses during autolysis since it was shown by Wolbach, Saiki and Jackson-" that anaerobic bacteria are almost in- variably present in aseptically removed dog livers, for control of auto- lysis by anaerobic cultures has seldom been carried out. However, there is much evidence that lactic acid is formed, and perhaps par- tially destroyed, in autolysis (Tiirkel,^^ Ssobolcw^^). Carefully con- trolled experiments by Lindcmann-^ seem to show that even in the absence of bacteria, autolyzing liver and heart can produce volatile acids, CO2 and hydrogen. The increase in fat described by some authors is probably only apparent, and due rather to the liberation of the fat from its combination with the proteins so that it is free and not "masked," as in normal organs.^'' Lecithin is decomposed, yielding choline, but cholesterol remains unchanged except for some hydrolysis of cholesterol esters. ^^ Creatine is changed to creatinine in autolyzing muscle, and apparently both are formed in autolysis of blood and liver. -^ The nucleo-proteins seem to be attacked by the autolytic enzymes, as the purine bases are prominent among the products of autolysis, and in quite different proportions from those obtaining in digestion of the same tissues by other means. Apparently autolytic enzymes, like trypsin, attack the protein group of the nucleoproteins, liberating the nucleic acids. These in turn are attacked by specij&c enzymes, nucleases,'^'' which liberate the purine bases, which are further decom- posed by specific enzymes, guanase, adenase, etc. (See Chap, xxiii). It is improbable that the intracellular enzymes are merelj^ pan- creatic enzymes taken out of the blood by the cells, because of the differences previously cited; furthermore, Matthes^^ found that the liver retained its autolytic power after the pancreas had been extir- pated (in dogs), and that the autolytic degeneration of cut peripheral nerves went on just the same, indicating that the autolytic enzj^mes do not owe their origin to the pancreas. Whenever tissues are disintegrated in any considerable quantities, as after extensive burns, peptolytic enzymes become demonstrable in 20 Jour. Med. Res., 1909 (21), 267. 2» Biochem. Zeit., 1909 (20), 431. " Ibid., 1912 (47), 367. See also v. Stein and Salkowski, Biochem. Zeit., 1913 (40), 486. "Zeit. f. Biol, 1910 (55), 36. 2^ See Krontowski and Poleff, Beitr. Path. Anat., 1914 (58), 407. "Corper, Jour. Biol. Chem., 1912 (11), 37; Kondo, Biochem. Zeit., 1910 (27), 427. 26 Myers and Fine, Jour. Biol. Chem., 1915 (21), 583; Hoagland and McBrvde, Jour. Agric. Res., 1916 (6), 535. "Sachs, Zeit. physiol. Chem., 1905 (46), 337; Jones, ibid., 1903 (41), 101, and 1906 (48), 110. 28 Arch. f. exp. Path. u. Pharm., 1904 (51), 442. 80 ENZYMES the blood and urine, and presumably these are related to the cell autolysis. 2^ They are noticeably increased in most infectious diseases in which the reaction between the body defenses and the infecting organism takes place in the blood stream (Falls). ^^ Also in the pre- mortal state a similar increase in peptolytic enzyme in the serum is associated with a high non-protein nitrogen figure for the serum. ^^ The relation of the autolytic enzymes to the increased proteolytic power of serum in pregnancy, as evidenced in the Abderhalden reaction {q.v.) has not yet been determined, ^^ ^^t Falls finds evidence of their correlation.^" Blood proteases are also increased in pregnancy. They bear no constant relation to the leucocyte count. Autodigestion of serum is normally prevented by the presence of a specific antienzyme, which latter can be inhibited by chloroform and various saturated monovalent ketones and alcohols (Yamakawa).^^ Influence of Chemicals on Autolysis. — As a general rule the addition of anti- septics to tissues to prevent bacterial action reduces the rate of autolysis, but as most of the results of "aseptic" autolysis so far reported are open to question, there is a reasonable doubt as to just how much depression of autolysis there is. Yoshimoto^^ finds that of the antiseptics ordinarily used, salicjdic acid, boric acid, and mustard oil (one-eighth saturated solution) permit the greatest auto- lysisj but it is probable that the acidity of the first two aiitiseptics plays an im- portant part, hence the value of the results obtained in autolysis with these acids is questionable. However, sodium salicylate and benzoate are said to favor autolysis (Laqueur).^^ Toluene seems to interfere much less with autolj'sis than chloroform or thymol (Benson and Wells^*^), and bromides are less harmful than toluene (Laqueur). Toluene vapor, acting on solid aseptic tissues, seems to cause more depression of autolysis than is usually observed in autolysis in solution. ^^ Dorothy Court^* found the only satisfactory antiseptics to be chloro- form, formaldehyde, benzoic and salicj'lic acids, and HNC; she emphasizes the fact that for different sorts of materials the different antiseptics give variable re- sults, so that the antiseptic used should be selected with reference to the material. Autolysis proceeds rapidly in weak ethyl alcohol, 5 per cent, being the minimum strength that will prevent putrefaction; for complete suppression of autolysis by alcohol the strength must be at least 90 per cent, net, after allowing for the water content of the tissues (Wells and Caldwell). ^^ Certain inorganic substances in proper concentrations have been reported as increasing the rate of autolysis [mercury^" and silver, ■'^ (colloidal^- or salts)], 29 See Pfeiffer, Miinch. med. Woch., 1914 (61), 1099, 1329. 3" Jour. Infect. Dis., 1915 (16), 466; also Petersen and Short, Jour. Infect. Dis., 1918 (22), 147. " See Schulz, Miinch. med. Woch., 1913 (60), 2512; Mandelbaum, ibid., 1914 (61), 461. 32 See Sloan, Amer. Jour. Physiol., 1915 (39), 9. 33 Jour. Exp. Med., 1918 (27), 689. 34 Zeit. physiol. Chem., 1908 (58), 341. 36Zeit. physiol. Chem., 1912 (79), 38 and 65. 38 Jour. Biol. Chem., 1910 (8), 61. 3' Cruickshank, Jour. Path, and Bact., 1911 (16), 167. 38Proc. Roy. Soc, Edinburgh, 1912 (32), 251. 39 Jour. Biol. Chem., 1914 (19), 57. "Truffi, Biochem. Zeit., 1910 (23), 270. « Izar, ibid., 1909 (20), 249. ••2 The accelerating influence of colloidal metals is denied by Bradley, Proc. Amer. Soc. Biol. Chem., 1918 (33), xi. PRINCIPLES Op AUTOLYSIS SI yellow pho.splioius," iodides/' arsenic,'" CaClo/'"' salts of Kc, Mr, and cobalt/' as well as salts of selenium, tclhiriuin," and manganese, ■*' colloidal sulfur'" but not colloidal carbon." The favorable concentrations of these metals are very low; thus the optimum proportion of arsenic is 0.007 milli^;rams i)er 1 Km. tissue, while 0.04 mg. inhibits autolj'sis. CO2 increases and oxyncn decreases autolysis" in vitro (Laqueur). There is disagreement as to whether radium rays augment autolysis.*' Injection of iodids into animals is said to increase the postmortem autolysis of their tissues (Stookey, Kepinow), as also do iron salts,'"' while large doses of salicylates decrease it (Laqueur). Morse" attributes the accelerating action of iodin and bromin to increased acidity from formation of halogen acids, and Bradley*^ finds evidence that most inorganic salts that stimulate autolysis act by increasing Fl-ion concentration. Addition of tuberculin to tissues at first delays and tiien increases the autolysis (Pesci^"*), and diphtheria toxin in small amounts increases autolysis (Barlocco," Bertolini**), neutralization by anti- toxin not preventing this effect. Lipoids also accelerate autolysis (Satta and Fasiani"). According to Soula*" narcotic poisons decrease, and convulsive poisons increase the rate of autolysis of nervous tissue. Glucose in one per cent, concentration decreases autolysis, and this may be related to the ''protein-sparing action of carbohydrates."*' E.xtracts of various ductless glands, or removal of these glands from animals, seem to have but slight effect on autolysis.*- In considering the foregoing statements allowance must be made for the fact that in most of the work cited there has been no proper consideration of H-ion concentration in the autolyzing mixtures. Relation of Autolysis to Metabolism It having been shown that proteases are present in all cells, the next question to be considered is, do they act only to destroy tissues after death, or are they of importance in metabolism? Since it is presumably necessary for proteins to be split into diffusible and easily oxidized forms in order that they may enter the cell, and be built up into the cell proteins, or be decomposed with the liberation of energy, the autolytic proteases ma}^ be assumed to be of prime importance in protein metabolism; but to prove it is another matter. Jacoby found " Saxl, Hofmeister's Beitr.. 1907 (10). 447; Virchow's Arch., 1910 (202), 149. ^ Kepinow^ Biochem. Zeit., 1911 (37), 238. Kaschiwabara, Zeit. phvsiol. Chem., 1912 (82), 425. Not confirmed by Albrecht, Jour. Biol. Chem., 1919 (41), 111. ^5 Izar, Biochem. Zeit., 1909 (21), 46; Laqueur and Ettinger, Zeit. physiol. Chem., 1912 (79) 1. «Briill, Biochem. Zeit., 1910 (29), 408. ^"Preti, Zeit. phvsiol. Chem., 1909 (GO),' 317; PoUini, Biochem. Zeit., 1912 (47), 396. " Fasiani, Arch. sci. med., 1912 (36), 436. ^9 Bradley, Jour. Biol. Chem., 1915 (21), 209; 1915 (22), 113. 5" Faginoli, Biochem. Zeit., 1913 (56), 291. *' Izar and Patane, ibid., p. 307. ^' M. Morse found oxvgen without effect on autolysis. Biochem. Bullet., 1915 (5), 143. " See Loewenthal and Edelstein, Biochem. Zeit., 1908 (14), 485; Brown, Arch. Int. Med., 1912 (10), 405. •■*< Kottmann, Zeit. exp. Path., 1912 (11), 355. "Jour. Biol. Chem., 1915 (22), 125. 5«Cent. f. Bakt., 1911 (59), 71 and 186. "Cent. f. Bakt., 1911 (60), 43. s' Biochem. Zeit., 1913 (48), 448. 59Berl. klin. Woch 1910 (47), 1.500. ^oCompt. Rend. Soc. Biol., 1913 (73), 297. " Shaffer, Proc. Soc. Biol. Chem., 1915 (8), .xlii. *\Izar and Fagiuoli, Sperimentale, 1916 (70), 265. 6 82 ENZYMES that if he hgated off a portion of the hver and let it remain in situ in the animal the necrotic tissues showed an accumulation of leucine, tyrosine, and other cleavage products of the proteins, which suggested that these same bodies are being formed in the liver constantly, but that they are as constantly removed from the normal organs by the circulating blood, or are undergoing further alterations which cease when the circulation is checked. The influence of various chemicals upon nitrogen elimination seems to correspond to their effect on auto- lysis (Izar,^^ Laqueur^^). Also, the histologic changes of starvation are similar in many respects to those of autolysis (Casa-Bianchi^^). Among other observations possibly bearing on the same question are those of Hildebrandt,^^ who found that autolysis in the functionating mammary gland is much more active than in the resting gland; and of Schlesingeir,^^ who found that autolysis was at its maximum (in rabbits) in new-born animals, decreasing rapidly in the first few months of life, and that in conditions associated with emaciation the rate of autolysis varied directly with the degree of emaciation. Wells^^ sought for a possible influence on autolysis by thyroid extract, which increases protein metabolism, but could demonstrate none in vitro; Schryver,^^ however, reported that autolysis was more rapid in the liver of dogs fed thyroid extract for some days before death than it was in control animals. The results of the former observer, but not those of the latter, have been confirmed by Morse. ^^ Defense of the Cells Against their Autolytic Enzymes The question of why the autolytic ferments do not destroy the cells until after death is a revival of the old problem of ''why the stomach does not digest itself," and the answer that satisfies some is that dead protoplasm is essentially different from living protoplasm. More specific replies are suggested by Wiener's studies on the relation of the reaction of the tissues to their autolysis. He found that auto- lysis does not begin in an organ until the original alkalinity is neutra- lized by the acids which are formed in all dead and djnng cells. ^' If enough alkali is added to the material from time to time to neutralize the acidity as it develops, autolysis docs not take place. Although «'Internat. Beitr. Erniihrungstor., 1910 (1), 287. " Zeit. physiol. Cheni., 1912 (79), 1 et seq. " Frankfurter Zeit. Pathol., 1909 (3), 723. "" Hofmeister's Beitrilge, 1904 (5), 463; see also Grinimcr, Hiochcni. Zeit., 1913 (53), 429. " Hofmeister's Reitr., 1903 (4), 87. "* Amer. Jour, of Physiol., 1904 (11), 351; eorrohorated bv Kottiiuinn, Zeit. klin. Med., 1910 (71), '•.m). «» Jour, of Physiol., 1905 (32), 159. 'ojour. Biol. (Jhem., 1915 (22), 125. " ()])io (toe cU.) found, liowevcr, llial avitolysis of leucocytes was more rapid in an alkaline inediuiii. Doeliez (.lour. \<]\]). Med., 1910 (,12), tHKi) stnt(>s that liver also contains an enzyme active in an alkaline meiliiun, hut which exists as an inactive zymogen until activated by acids. See also Dernhy." DEFENSE AGAINST AUTOLYSIS 83 the spleen contains an enzyme digest inj!; in alkaline solution," and another which acts best in weak acids, the latter appears more proniinontly under ordinary conditions because the spleen and the blood contain antibodies which check the enzyme that acts in alka- line solutions, while acids destroy this antibody (Hedin).^^ Organic acids are formed in autolysis of the tissues, and the latent period be- tween the time of the removal of an organ from the body and the ap- pearance of autolysis may be explained partly by the time recjuired for the neutralization or alkalinity. Bradley'^ has also obtained evi- dence that the acid renders the substrate susceptible to digestion by the proteases. Dernby's^^ demonstration of the existence of pepsin-like and erepsin-like enzymes suggests that there must be developed enough acidity to permit the peptase to form peptones before the ereptases can begin their further cleavage. Maximum autolysis is known to occur when tissues are first made acid and then neutralized or slightly alkalinized (Hedin) . The old observation that rigor mortis disappears most rapidly in muscles that have been exhausted just before death is probably explained by the greater amount of acid in such muscles. If we imagine that autolysis is limited to periods when the cells have an acid reaction, however, we limit the range of usefulness in the living cell to a minimum, since during life the tissue fluids, and presumably the cell contents, are preponderatingly alka- line. The control of autolysis by maintenance of a low H-ion concen- tration is undoubtedly an important factor, for Bradley found that a reaction equal to that of blood almost completely inhibits autolysis, while the degree of increased H-ion concentration that may develop in local asph3^xia, or after death, produces optimum conditions for autolysis. Still another possible defense of the Hving cells may be found in the existence of specific antienzymes. Just as the serum contains anti- trypsin, so it seems to contain substances antagonistic to the autolytic enzymes. Levene and Stookey found that tissue juices show a resist- ance to digestion, Yamakawa^^ found that serum autolysis is prevented by an antienzyme, and Opie found that the serum of inflammatory exudates retards the action of the autolytic enzymes that are con- tained within the leucocytes. Serum also inhibits autolysis of the tissues, so it is probable that continuance of the circulation may pro- vide antibodies to the tissues to hold the intracellular enzymes in check, possibly without interfering with their action on other pro- teins than those of the cell structure.''* (See Antienzymes.) It is ^2 Morse (Jour. Biol. Chem., 1917 (31), 303) considers this enzyme to be heter- olj'tic, derived from the white cells. " Festschrift f. Hammarsten, Upsala, 1906. '••According to Guggenheimer (Deut. Arch. khn. Med., 1913 (112), 248; Dent, med. Woch., 1914 (40), 63),- the serum in various diseases has a characteristic stimulating or inhibiting effect on in vitro tissue autolj'sis, but the conditions of such experiments are so complex as to make their significance doubtful. 84 . ENZYMES highly probable that serum cheeks autolysis at least in part by virtue of its "buffer" function, which interferes with the development of acidity. Lack of oxygen cannot be held solely responsible for auto- lysis, according to the stuches of Morse, ^^ who found that autolysis occurs in muscles with divided nerves but intact blood supply. Never- theless, reduced blood supply results in increased H-ion concentration which greatly facilitates autolysis, and it cannot be denied that auto- lysis is observed chiefly if not solely in asphyxiated tissues. There can be no question that the supply of food-stuff is of essential importance in determining autolytic changes, for it has been found by Conradi,^^ Rettger,^'' and Effront'* that bacteria and yeasts begin to undergo autolysis when they are placed in distilled water or salt solution, which they do not do, to any such extent at least, when in nutrient media. (In this way it has been found possible to obtain the intracellular poisons of such bacteria as tj^phoid and cholera.) Autolysis is not marked so long as the bacteria are supplied with nourishment, but when nutrient material is lacking, autolytic decom- position is no longer repaired and the bacteria disintegrate. Pre- sumably the changes are the same in tissue cells, and anemic necrosis may be explained in this wa}'. Tissue enzymes are also capable of digesting bacteria (Turro^^). Another direction in which the key to the action of these enzymes may be sought has been indicated by Jacoby,^*^ who found that to a certain degree the autolytic enzymes of each organ are specific for that organ. Liver extract will not split lung tissue, although it will split the proteoses that are formed in lung autolysis, possibly because these proteoses are less specific than the proteins from which they arise, or perhaps because of the erepsin the extract contains (Vernon). Leuco- cytic proteases, however, seem capable of splitting foreign proteins of all sorts. Richet^^ states that the protease of liver tissue does not attack either muscle tissue or liver tissue that has been coagulated. Another hypothesis has been advanced by Fermi,*- who suggests that the protoplasm of living cells is not digested because its structural configuration is such that the enzymes cannot unite with it, an attract- ive but practically undemonstrable idea. Lastly, it must be considered that at least to some extent the en- zymes exist in the cells in their inactive zymogen form, and per- haps are changed into the active form as needed, and inhibited or changed back again wlien their work is temporarily finished. A " Amer. Jour. Phvsiol., lUlf) (liO), 147. '» Deut. mod. Woi-li., l<)():} (2«)), 20. "Jour. Med. Hcsciiroh. n)04 (Ui), 79. '8 Bull. Soc. Cliiiii., HH).") (;};{), 847. '»Cont. f. Bakt., 1902 (:V2), 105. »« llofineistcr's Boitr., 190:} (;}), 440. 8' (Joiiipt. Boiul. Soc. Biol., 190:} {'•>'■>), tiot). 82 Cent. 1'. Bakl., 1910 (■>(>), .1."). AUTOLYSIS ].\ JWTIIOIAHIKAL I' HOC ESSES 85 rhythmical chaiijic of tliis nalui'c might be imajiiiicd as occuniMg and accounting for interaction by the enzymes, particiihirly since; rliythmi- cal changes in metabolism are known to occur {e. g.,) rhythmical pro- duction of carbon dioxide (Lyon^^), and enzyme action in vitro may show rhythmic variations (Groll).^'* Autolysis in Pathological Processes All absor];)tion of dead or injured tissues, and of organic foreign bodies, seems to be accomplished by means of digestion by the enzymes of the cells and tissue fluids. We may distinguish between the diges- tion brought about bj^ the enzymes of the digested tissue itself, or autolysis, and digestion by enzymes from other cells or tissue fluids, or hetcrolysis (Jacoby). Heterolysis is accomplished particularly by the lecucocytes, which contain ferments capable of digesting not only leucocytic proteins but apparentlj^ every other sort,^'' from serum- albumin to catgut ligatures. The heterolysis may be intracellular when the material to be digested has first been taken up by the cells (phagocytosis); or extra-cellular, either by enzymes normally con- tained in the blood plasma and tissue fluids, or by enzymes liberated by the leucocytes and fixed tissue cells. On death and dissolution of a cell the intracellular enzymes are released, ^^ but it is not known to what extent the enzymes may be secreted from intact living cells. As far as pathological processes show, the amount of liberation of en- Z3'mes from normal cells is very slight, if any, and the digestive en- zymes of the blood plasma seem to be very feeble, but this is perhaps because they are largely held in check by the anti-enzymatic substances of the serum. Pathological autolysis and heterolysis, therefore, are brought about chieflj'^ bj' enzymes liberated from dead or injured cells. Bacteria, however, can multiply upon a medium of coagulated protein, which suggests that they also secrete proteolytic substances. In pathological conditions digestion of degenerated tissues seems usuallj'^ to be the result of both autolysis and heterolysis. An infarct softens because the intracellular enzymes digest the dead cells, exactly as «^ Science, 1904 (19), 350. 8^ Nederl. Tijdschr. v. Geneesk., 1918 (1), 1085. 8^ Manj^ authors suggest that the leucocytes merely carry enzymes from one organ, particularly the pancreas, to another, and that these enzymes are not formed by the leucocyte itself. Opie (Jour. Exp. Med., 1905 (7), 759) has shown, however, that the bone-marrow contains proteolytic enzymes which are like those of the leucocytes in that they act best in an alkaline medium, whereas the auto- lytic enzymes of the lymphatic glands and most other tissues act best in an acid medium. This leaves little room for doubt that the leucocj'tes are equipped with their characteristics enzjmies when they leave the bone-marrow, and that they are not obtained later in the pancreas or elsewhere. More recently, however, van Calcar (Pfliiger's Archiv., 1912 (148), 257) has revived the idea of the origin of leucocj'tic enzymes in the digestive glands. 8^ Peptolytic enzymes appear in the urine after severe superficial burning, pre- sumably coming from the disintegrated cells. (Pfeiffer, Miinch. med. ^^'och., 1914 (61), 1329.) 86 ENZYMES they do when the tissue is removed from the body, ground up, and put in the incubator under toluene. In addition leucocytes wander in, disintegrate, and their liberated enzymes help in the process, as also do, to a less degree, the enzymes of the blood plasma. It is because of the heterolysis by leucocytic enzymes that a septic infarct becomes softened so much more rapidly than does a sterile infarct, and by comparing the rate of softening in septic and aseptic infarcts we see that the cellular autolysis is a very slow process as compared to the heterolysis accomplished by the leucocytes. The explanation of this may lie in the fact that most intracellular proteases act best in an acid medium (Wiener), while leucocytic proteases act best in an alkaline medium (Opie), and the infarcts of small size are seeped through by alkaline blood fluids. When an infarct is large, we find it undergoing central softening while the periphery remains firm; this corroborates our hypothesis, for acids are developed during autolysis (Magnus-Levy), which at the periphery are neutralized by the blood plasma, so that only at the center is autolysis active. The inhibit- ing action of the serum also has a similar effect, limiting autolvsis at the periphery. Necrotic areas of any kind are absorbed b}'' similar processes. Apparently all varieties of cells are subject to autolysis or heterolysis whenever they are killed or sufficiently injured. Involution of the uterus probably depends upon autolysis, which is much more active in the puerperal uterus (Ferroni^^), and creatine is found in the urine when such autolysis occurs, ^^ although A. Morse^^ considers this to be independent of the uterine autolysis. Atrophy may be looked upon as an autolysis in the normal course of catabolism, not met by a corres- ponding building up of the proteins, but M. IVIorse^^ could find no evidence that the atrophy and involution of the tadpole tail is ac- companied by an accelerated autolysis. The solution of fibrin by tissues, fibrinolysis, is considered to be distinct from tissue autolysis by Fleisher and Loeb.^'' In atrophic cirrhosis the fibrinolytic activity of the blood is increased, which may explain the hemorrhagic tendencj' of this disease. ^^ In the case of septic softening the action of the bacteria needs also to be taken into consideration, since they produce proteolytic ferments, but their effect seems to be relatively small as compared with leucocytic digestion.^- Intracellular digestion of *' Arm. di Ostetrica e Ginecol., 1906 (2), 553; see also Sleinons, Bull. Johns Hopkins IIosp., 1914 (25), 195; Arthur Morse, Jour. Amer. Med., Assoc, 1915 (05), l()i:i. 8« Shaffer, Amer. .Jour. Physiol. 1908 (23), 1. 89 Max Morse, Am. Jour, i'hysiol., 1915 (30), 145. 90 .Jour. Biol. Chcm., 1915 (21), 477. 91 Goodi)asture, iiull. .Johns Hopkins JTosp., 1914 (25), 330. 92 Tlic enzymes of staphylococcus are nmcli more strongly proteolytic tlian those of streptococcals (Knapj), Zeit. f. Heilk. (Chir.), 1902 (23), 230.) which may he one reason why tlie latter so much more frequently produces lesions without suj)- puration tlian does the former. AUTOLYSIS IN PATHOLOGICAL PROCESSES 87 necrotic tissue by leucocytes seems also to be relatively unimportant. Suppuration, therefore, must be considered as the result of digestion of dead tissue by enzymes derived from the leucocytes, the plasma, the bacteria, and the destroyed cells themselves. A tubercle does not ordinarily suppurate, because the tubercle bacillus and the sub- stances it produces are not strongly chemotactic, and hence not enough leucocytes enter the necrotic area to produce a digestive softening. The products of autolysis may of themselves be toxic; albumoses and peptones certainly are, and the other cleavage products are prob- ably not altogether innocuous. (See "Autointoxication.") Some of the symptoms of suppuration, particularly the fever and chills, have been ascribed to the autolytic products rather than to the bacterial poisons, particularly as aseptic suppuration is accompanied by fever. Jochmann^^ has found evidence that the protease of leucocytes can cause fever and also reduce the coagulability of the blood. The work of Vaughan and other recent students of the reaction to foreign pro- teins, shows that typical fevers can be produced by the enzymatic disintegration of proteins in the body.^'* Degenerative changes in nervous tissue are associated with autolytic decomposition of the lecithin (NolP^) and the liberated choline, or its more toxic derivatives, may be a source of intoxication.^^ In all conditions associated with autolysis, such as resolving pneumonic exudates, large abscesses, soften- ing tumors, etc., albumoses (and peptones?) may appear in the urine. Autolytic products may also be hemolytic (Levaditi^^), and they may prevent clotting of the blood (Conradi^^). It is probable that among the products of autolysis are bactericidal substances, ^^ although it is doubtful if the concentration is often sufficient for them to be of influence except in well walled areas. There is also much evidence that after extensive traumatism, especially as observed in war, the products of the tissue autolysis may be responsible for serious intoxication, and possibly for conditions interpreted at times as shock. ^ The observations made in experimental anaphylaxis suggest that it is especially the slightly altered proteins, perhaps only changed in their colloidal properties, that are most likely to be responsible for these shock-like intoxications. However, it is also possible that amines derived from the aminoacids may be of 9'Virchow's Arch., 1908 (194), 342. 9^ See Vaughan, "Protein Split Products," Philadelphia, 1913. « Zeit. phvsiol. Chemie, 1899 (27), 380. 9« See Halliburton, Ergebnisse der Physiol., 1904 (4), 24. " Ann. d. I'lnst. Pasteur, 1903 (17), 187; also Fukuhara, Zeit. f. exp. Path. u. Pharm., 1907 (4), 658. 98 Hofmeister's Beitr., 1901 (1), 136. 99 See Bilancioni, Arch, farmacol., 1911 (11), 491. 1 Delbet, Bull. Acad. Med., 1918 (80), 13; Cannon, Compt. Rend. Soc. Biol., 1918 (81), 850; Turck, Med. Record, June 1, 1918. 88 ENZYMES importance in producing shock whenever tissues are injured.^ Methyl guanidine may also be formed from disintegrating tissues and has con- siderable toxicity. Work has been reported upon autolytic processes in a number of pathological conditions, which may be discussed briefly as follows: Exudates. — The presence of leucine, tyrosine, proteoses, and pep- tones in pus has been known for many years, and the reason for their appearance is now clear. Muller,^ many years ago, observed that purulent sputum digested fibrin, but that non-purulent sputum did not have this property. Achalme^ found that pus would dissolve gelatin, fibrin, and egg-albumen. Ascoli and Mareschi^ detected autolysis in sterile exudates obtained experimentally. Umber^ found that ascitic fluid exhibited autolytic changes, which observation could not be confirmed by Schiitz^ in pleural exudates and ascitic fluids. Zak^ found that autolysis was inconstant in various exudates. The differences in these results are explained by Opie's^ observation that in experimental inflammatory exudates the leucocytes are capable of marked autolysis, whereas the serum contains an antibody which holds this autolysis in check; if the antibody is destroyed by heat, then the serum proteins are also digested by the leucocytic enzymes. This antibody seems to be contained normally in the albumin of the blood-serum. In old exudates the antibodies are decreased, and auto- lysis then occurs, explaining the variable results of Umber, Schiitz and Zak. The intracellular proteases of the polynuclear leucocytes act best in an alkaline medium; those of the mononuclears in acid medium. If the proportion of serum to leucocytes is high, then there is no autolysis, as in serous exudates; but if the leucocj^tes are abun- dant, then the antibody is overcome and we get autoh'sis, as in ordinary suppurative exudates. Animals with but little protease in their leucocytes (e. g., rabbits), do not ordinarily produce a liquid pus (Opie). Exudates produced by bacterial infection also seem to possess the properties above described. Galdi^" found autolysis greater in exu- dates than in transudates, but observed no constant relation between the number of leucocytes, or the amount of chlorides, and the rate of autolysis. All exudates, according to Lenk and Pollak,^^ contain enzymes splitting glyeyl-glycinc (peptolytic enzymes); the most active 2 See Abel and Kubota, Jour. Pharm. Exp. Ther., 1919 (13), 243. 3 Kossel, Zeit. f. klin. Med., 18S8 (13), 149. ^ Compt. Rend. Soc. Biol., 1899 (51), 568. 6 See Maly's Jahresbericlit, 1902 (32), 5(38. 6 Munch, nied. Woch., 1902 (49), 11(59. 'Cent. f. inn. Med., 1902 (23), ll(il. 8 Wien. klin. Woch., 1905 (18), 37(j. 9 Jour, of Exper. Med., 1905 (7), 310 and 759; 190(i (8), 410 and 530; 1907 (9), 207, 391 and 414; also a full review in Arcii. Int. Med., 1910 (5), 541. '0 See Folia Ilcinat., 1905 (2), 529. '» Dcut. Arcli. klin. Med., 1913 (109), 350; See also Wiener, liiocliein. Zeit., 1912 (41), 149; Al!iiid<'ll)auiii, Miinch. lued. Woch., 1914 ((il), 4()1. AUTOLYSIS IN PATHOLOGICAL PROCESSES 89 exudates are tliose of cancer and tuberculosis, tlie least active are passive congestion fluids; pleural exudates contain more active enzymes than peritoneal exudates of .similar character. Knapp^- holds that in pus the cocci and the enzymes they produce are responsible for much of the digestion. Pus cells alone do not undergo digestion so rapidly as when bacteria are present, and di- gestion is more rapid if the bacteria are alive than when inhibited or killed by antiseptics. Streptococcus is almost inactive, staphylococcus is quite active, and B. coli still more so. However, pus corpuscles free from bacteria are highly proteolytic, causing digestion in serum plates in dilutions of 1-700 (Jochmann). Knapp could find no rela- tion between the autolytic power of the pus and the severity of the in- fection from which it resulted. A constant constituent of pus is d-lactic acid, ^2 ^nd it increases during autolysis; this may well modify the rate of autolysis of pus. (See also the discussion of the "Chem- istry of Pus," Chap, xi.) Proteolytic Enzymes of the Leucocytes.^^ — By the introduction of the plate method of testing the proteolytic activitj^ of leucocytes, Miiller and Jochmann brought the study of this particular vital activity into the range of clinical laboratories, and aroused much general interest in what had previously concerned only a few pathol- ogists, especially E. L. Opie. The principle is that of permitting the leucocj^tes or other cells to act upon a blood serum plate at a tem- perature of 55°, which prevents bacterial action but permits the pro- teolytic enzymes of the cells to digest the coagulated serum, forming depressions in the surface ("Dellbildung"). This proteolytic activity is, of course, heterolysis rather than autolysis. Many modifica- tions of this method have been introduced (such as using casein- agar), but the principle involved is the same, and they are fully explained and discussed in the article by Wiens. Normal blood does not contain enough leucocytes to cause observable cUgestion, but my- elogenous leukemia blood causes distinct chgcstion while lymphatic leukemia does not, showing that it is the polynuclears and myelocytes that are responsible. Other observations fasten the proteolytic activ- ity upon the neutrophile granules. Leucocytes of normal human blood will, if concentrated enough, cause digestion of serum plates, as also, of course, will pus. The leucocytes of rabbits, guinea pigs, and practically all animals except man, apes and monkej^s, are de- void of proteolytic activity demonstrable by the plate method. Nor- mal serum, both homologous and heterologous, exercises a strong inhibition on this digestion, so that it is necessary to have an excess of leucocytes present to obtain the reaction. The activity of leucocytic autolysis is indicated by the observation that in drawn cerebrospinal 12 Ito, Jour. Biol. Chem., 1916 (26), 173. '^ Full bibliography by Wiens, Ergebnisse Phj-siol., 1911 (15), 1; Jochmann, Kolle and Wassermann's Handbuch, 1912 (2), 1301. 90 ENZYMES fluid the leucocytes all disappear in from three to sixty-three days, and in 24 hours the count has been observed to drop from 392 to 6.^* The leucocytic enzymes seem to be very resistant against chemicals, especially against formaldehyde, so that museum specimens of leuk- emic tissues preserved in formalin for years are still proteolytic. Liver tissue is but slightly proteolytic by this test, spleen more so, and leuco- cyte-containing fluids, such as saliva and colostrum, are quite active. Pancreas tissue, has, of course, strong proteolytic action, but it is shown to be distinct from the leucocytic protease bj^ being inhibited by certain sera that do not inhibit the leucocytic protease. In gen- eral, tissues do not cause much proteolysis of serum plates unless they are invaded by many leucocytes, which applies also to tumors, including multiple myelomas. Besides proteases, leucocytes contain other enzymes. ^^ To quote the summary by Morris and Boggs^^ "it has been shown that the normal and pathological neutrophile leucocytes and myeloblasts contain an oxidase and probably a lipase and an amylase; myeloblasts contain an amylase. In lymphoid tissues two proteases and a lipase have been shown to exist. In leukemia leukoprotease has been demon- strated in the myeloid variety of the disease, while it has not been found in chronic lymphoid leukemia. Lipase has been demonstrated in two cases of myeloid leukemia, and oxidase in all mj^-eloid cases observed in which the neutrophilic cells were present in excess." Jobling and Strouse,^^ confirming Opie's observation of two distinct proteases in leucocytes, find also evidence of an ereptic enzyme acting in either acid or alkaline fluids. ^^ Pneiunonia. — In the stage of resolution lobar pneumonia presents a striking example of autolysis. The often-remarked phenomenon that the lung tissue itself is not in the least affected, while the dense contents of the alveoli are rapidly dissolved and removed is explained by the invariable immunity of living cells to digestive enzymes. Ex- cept for some slight possible assistance by the alveolar epithelium and the enzymes of the serum, the enormous and rapid digestion of pneumonic exudates is accomplished by the leucocytic enzymes. The rapid rate of digestion may be accounted for by the absence of circu- lation within the alveolar contents, which permits the leucocytes to " Bonaba, Anales de le Facultad de Med., 1919 (4), 111. '^ According to Tschernoruzki (Zeit. physiol. Chem., 1911 (75), 216) amylase, diastase, catalase, peroxidase, and nuclease, but not lipase. I also found uricase absent from dog leucocytes (.Jour. Biol. Cnem., 1909 (0), 321). Fiessinger and Marie (Compt. Rend. Soc. Biol., 1909 (67), 177) state that the lymphocytes contain lipase, although myeloid cells do not. (See also Resch, Deut. Arch. klin. Med., 191.5 (118), 179). Leucocytes are also said to contain a "lipoidase" s])lit- ting choline from lecithin (Fiessinger and Clogne, Compt. Rend. Acad. Sci., 1917 (165), 730. i« Arch. Int. Med., 1911 (8), 806. "Jour. Exj). Med., 1912 (16), 269. 1* Concerning enzymes of normal leucocytes see also Fiessinger and Clogne, Ann. de M^-d., 1917 (4), 445; Parker and Franke, Jour. Med. Res., 1917 (37), 345. AUTOLYSIS IN PNEUMONIA 91 act unimpeded by the anti-bodies of the blood plasma. Digestion of the exudate continues after death, accounting for the marked dif- fuse softening observed in pneumonic lungs in bodies kept some days before autopsy. As long ago as 1888, Kossel^^ mentioned that Fr. Miiller had found that glycerol extracts of purulent sputum exhib- ited a digestive action upon fibrin and coagulated protein, whereas non-purulent sputum did not possess this property. In 1877 Filehne extracted ferments in the same way from the sputum in gangrene of the lung; Stolniknow, in 1878, found a similar ferment in pneu- monic sputa, and Escherichin 1885 showed that the proteolytic action of tuberculous sputum was independent of putrefaction. Other early observations of similar nature are reviewed by Simon, 2" who demon- strated the presence of leucine and tyrosine in the autolyzed lungs. In a later work ^Miiller reports finding three grams of leucine and tyrosine in a pneumonic lung, as well as lysine, histidine, and purine bases from the decomposed nucleoproteins. The appearance of free purines during autolysis of pneumonic lungs has been investigated by Mayeda,2i Long and Wells. ^^ Boehm^^ isolated histidine and arginine from the same material. Rietschel and Langstein^'* found 0.32 gm. leucine in the urine of a pneumonic child. Flexner^' noted that autolysis, while very rapid in the gray stage, is but slight in the red stage (because of paucity of leucocytes) and also in unresolved pneumonia, which he considers as due to some inter- ference with autolysis. Silvestrini^^ found that in gray hepatization the reaction was strongly acid, in red faintly so; the graj"- hepatization showed more peptone, and leucine and lactic acid were both demon- strable. A fibrin-digesting enzjmie was isolated, and milk was coagu- lated. Weiss^^ has reported finding a toxic albumose in gray pneumonic lungs. Lord-* has found in pneumonic lungs a proteolytic enzyme active at pH 7.3 to 6.7, but inactive at higher acidity; also an enzjine sphtting peptone to amino acids and active at pH 8.0 to 4.8, but most active at 6.3 to 5.2. He, therefore, pictures resolution of pneu- monic exudates as occurring in two stages: First, proteolysis while the reaction is nearly neutral, and later as the acidity increases the cleavage of the peptone increases. He also finds that the pneumo- cocci cannot long survive a reaction more acid than pH 6.8, and their dissolution takes place at reaction from 6.0 to 5.0, which is of 19 Zeit. f. klin. Med., 1888 (13), 149. 20 Deut. Arch. klin. Med., 1901 (70), 604. "1 Deut. Arch. klin. Med., 1910 (98), 587. 22 /Wd, 1914 (115), 377. 23 Ibid., 1910 (98), 583. 24 Biochem. Zeit., 1906 (1\ 75. 25 Univ. of Penn. Med. Bull., 1903 (16), 185. -^ Bull. del. Soc. Eustachiana, 1903, abst. in Biochem. Centralbl., 1903 (1), 713. " Arch. Int. Med., 1919 (23), 395. 28 Jour. ExD. Med., 1919 (30), 379. 92 ENZYMES significance in view of the observation that pneumonic lungs are more acid than normal organs, acidity as high as pH 6.0 to 5.4 having been found. 2^ Rzentkowski^" found an increase of non-coagulable nitrogen in the blood of pneumonics, probably resulting from autolysis in the exu- date. According to Dick^^ the blood serum after the crisis contains an enzyme which acts specifically on the pneumococcus proteins. Petersen and Short^^ found an increase of serum ereptase in the blood preceding or accompanying crisis or lysis, and suggest that it may have a function in attacking the toxic protein fragments. In the liver during experimental pneumococcus septicemia, autolysis is said to be increased in rate..^^ Almagia** suggests that the bactericidal action of the prod- ucts of fibrinolysis in pneumonia may be of importance in checking the disease. Necrotic Areas. — Jacoby^^ found that if a portion of a dog's liver was ligated oF and the animal kept alive for some time, the necrotic tissue contained the same products that he had obtained in experi- mental autolysis. The absorption of necrotic tissues generallj^ is ascribable to either autolysis or heterolysis. Presumably there is no great difference in the self-digestion of an organ which is necrotic because its blood supply is cut ofT, and of a similar organ removed from the body aseptically and allowed to undergo aseptic autolysis in an incubator. At the periphery there might be some effects pro- duced in vivo by the inhibitive action of the serum or the digestive action of the leucocytes, but beyond that no marked differences are to be expected. In both cases asphyxia is present, leading to increased acidity, without which little if any autolysis can occur. It has been found that in experimental infarction of the kidney there develops sufficient acidity to permit of autolysis, and the hydrolysis of the proteins increases with the development of acidity (Straus and Morse.^®) A study of the relation of autolysis to the histological changes that occur in necrotic areas by Wells^'^ gave evidence that there occurs early a decomposition .of the nucleoproteins of the nuclei, which is probably brought about by the intracellular autolytic enzymes. The liberation of the nucleic acid and the reduction in the bulk of nuclear material through the digestion away of the protein is probably the cause of the pycnosis observed in necrotic areas. Later the nucleic acids are further decomposed through the special enzymes described by Jones, Sachs, and others, the "nucleases." This is prosuniably -3 Jour. Amer. Med. Assoc, 1919 (72), 1364. 30 Virchovv's Arch., 190.5 (179), 405. 31 Jour. Infect Dis., 1912 (10), 383. '2 Jour. Infect. Dis., 1918 (22), 147. " Medigrcceanu. Jour. Exp. Med., 1914 (19), 31. 3^ Festschr. for Celli, Torino, 1913, p. 459. ^oZcit. physiol. Chein., 1900 (30), 149. '« Proc. Soc. Exp. Biol. Med., 1917 (14). 171. " Jour. Med. Research, 1900 (15), 149. AUTOLYSIS IN NECROTIC TISSUES 93 the cause of the h).ss of nueleai- staining so characteristic of necrosis. That these changes are due to the intracellular enzymes was shown by implanting in animals pieces of sterile tissues, the enzymes of which had been destroyed by heating; these were found to undergo altera- tions only after several weeks, and then as the result of the action upon them of invading leucocytes. The slow rate of autolysis that occurs in infarcts and other aseptic areas is presumably due in part to the action of the antibodies of the serum, for it was found, experimen- tally, that the histological changes of autolysis when the tissues are placed in heated serum proceed about twice as rapidly as when they are placed in fresh serum. Chemotactic substances do not seem to be formed in aseptic dead tissues, but the slow absorption of such tissues is, however, finalh^ accomplished by the leucocytes acting from the periphery, there being little actual autolysis of the dead cells b}' their own enzj'mes. The rapidit}' with which autolytic changes occur in different organs, as indicated by the disappearance of nuclear staining, seems to be about as follows: (1) Liver, kidney (epithelium of convoluted tubules); (2) spleen, pancreas; (3) kidney (collecting tubules, straight tubules, glomerules); (4) lung (alveolar and bronchial epithelium); (5) thyroid; (6) myocardium; (7) volun- tary muscle; (8) skin (epithelium); (9) brain (cortical cells). Stroma cells seem to be attacked chiefly by enzymes from the parenchyma cells. Of all cellular elements, the endothelium of the vessels seems to have the greatest resistance to both autolysis and heterolysis. The finer structural changes of aseptic autolysis of liver in salt so- lution, have been carefuly studied by Launoy,^^ who notes a period of relative latency (20 to 24 hours at 38°) , followed by rapid changes in both cytoplasm and nucleus, associated with the appearance of myelin forms. Dyson^^ describes loss of the Altmann's granules in autolyzing cells. Cruickshank*" states that when aseptic autolysis of tissues kept in a moist chamber is observed microscopically the changes are slower, and there is less solution of the cytoplasm, but in general the results are much the same. No. fat could be found by special stains. Fetuses that have undergone aseptic autolysis in the uterus show complete loss of nuclei in 5 to 6 days, a stage correspond- ing to 8 to 15 days autolysis in the moist chamber. In experimental nephritis Simons^^ observed a decreased autolysis of the kidneys. Degenerated nervous tissue also undergoes a slow autolj^sis which, according to Noll,*^ results in the splitting of "protagon" with hbera- tion of lecithin. ]\Iott, Halliburton, ■'•^ Donath, and others have shown '* Ann. Inst. Pasteur, 1909 (23), 1. "9 Jour. Path, and Bact., 1912 (17), 12; also Aschoff, Verb. deut. Path. Gesellsch, 1914 (17), 109. " Jour. Path, and Bact., 1911 (16), 167. "Biochem. Zeit., 1S14 (67), 483. " Zeit. physiol. Cheiu., 1899 (27), 390. *^ General resume in Ergebnisse der Physiol., 1904 (4), 24. 94 ENZYMES that in nerve destruction lecithin is spht up with-hberation of cho- line (see "Choline")- Koch and Goodson^* found that degenerated nervous tissue is characterized, chemically, by containing a relatively increased amount of nucleo-proteins, with an absolute decrease in solid constituents, while the lecithins are greatly altered. In caseation autolysis is very slight, as is shown by the persistence of the caseous material for long periods of time without absorption. Presumably the toxin of tuberculosis destroys the autolytic ferments of the cells it kills, ^^ and as there is little chemotactic influence, leuco- cytes do not enter the caseous area. Jobling and Petersen^^ find evidence that the soaps of unsaturated fatty acids present in tubercles are responsible for the inhibition of digestion. Spiethoff^'' found that pure caseous material is usually free from even traces of albumose and peptone, but the caseous material at the periphery mixed with tissue elements contains them in very small quantities, suggesting that at the periphery of caseous areas some slight autolysis does occur. The fact that B. tuberculosis is, itself, very poor in proteolytic enzymes as compared with most other bacteria may be another factor. When leucocytes are attracted into a tuberculous focus softening goes on rapidly, showing that there is no loss of digestibility of the caseous material, but merely a lack of enzymes. Pus from a cold tuberculous abscess will not digest fibrin, but if iodoform is injected, leucocytes enter in great numbers, softening is rapid, and the pus will then di- gest fibrin (Heile^^). On serum plates tuberculous pus produces no digestion unless a secondary infection or other cause has resulted in a local accumulation of leucocytes. ^^ Tuberculous material contains, like the lymphocytes, an enzyme which is proteolytic in acid media and which is inhibited by normal serum (Opie and Barker'*^). Correlation of Histological and Chemical Changes. — A careful study of the relationship of the chemical chan}.';es produced by autolysis, to the histological changes of necrosis and autolysis, has been made by H. J. Corper,^" and colored plates published together with analytical figures make it possible to correlate at a glance the structural and chemical changes of necrobiosis. Corper found that in the early stages, characterized by a high grade of pycnosis but no further nu- clear changes, the nucleins are still intact; but with well developed karyorrhexis and beginning karyolysis, some ten per cent, of the nuclein nitrogen has become soluble in the form of purine bases. When karyolysis is completed so that no more nuclei remain in a stainable condition, only twenty-eight ])er cent, of the nucleo- proteins was found to have been dccom])oseil to free purine bases,^' the remaining seventy-two per cent, being intact although unstainable. This rather surpris- ing observation indicates that the stainable chromatin rejjresents but about one- fourth of the nucleins of the cell, which is in accord with the views of Hamnuirsten «Amer. .Jour. Physiol., 1906 (15), 272. " However, Pesci (Pathologica, 1912 (3), 144) states that tuberculin increases autolysis in vitro. "Uour. Exp. Med., 1914 (19), 383. ■" Cent. f. inn. Med., 1904 (25), 481. "8 Zeit. klin. Med., 1904 (55), 508. "Jour. Exper. Med., 1909 (11), 686. ^"Jour. Exper. Med., 1912 (15), 429. " Marshall (Jour. Biol. Chem., 1913 (15), 81) has also found that much of the nucleic acid remains unaltered in autolysis of thymus. AUTOLYSIS OF THE LIVER 95 and others. The lecithin disintegrates somewhat more completely, about one- half or two-thirds beinp disintegrated by the time nuclear destruction is complete, after whicli this and all other autolytic change is slow. The change from coagul- able to non-coagulable forms of nitrogen was as follows: Normal spleen, non- coagulable nitrogen, 5.7 per cent, of the total; stage of marked pycnosis, without rhexis or lysis, 7.4 per cent.; stage of karyonhexis and early karyolysis, 2(j.5 jjer cent. ; stage of complete karyolysis, ;^0.3 per cent. That is, when nuclear structures in the spleen have lost their staining properties entirely through autolysis, about 72 per cent, of the nuclein nitrogen, 50 per cent, of the insoluble phosphorus com- pounds. 70 per cent, of the coagulable nitrogen, and about two-thirds of the lecithin are still intact. Liver Degenerations. — ^Thc relation of the disintegration observed in phosphorus-poisoning and acute yellow atrophy to the experimental autolysis of the liver has been the object of much study. Salkowski originally pointed out that the same products were found in the blood, urine, and liver tissue in acute yellow atrophy as are produced in autolysis. Jacoby^^ found that the livers of dogs, taken just as the animals were dying of phosphorus-poisoning, contained free leucine and tyrosine; also, he found that the rate of autolysis of such livers after removal from the body was much greater than in normal livers. The oxidizing ferments (aldehydase) are not destroyed by the proc- ess. He found that addition of minute amounts of phosphorus to liver enzymes did not increase their proteolytic power; nevertheless, he seems inclined to assume that in phosphorus-poisoning alteration in the autolytic enzj^mes is an important factor in the liver degen- eration. It would seem much more probable that phosphorus is a poison that kills cells and does not destroy their autolytic enzymes, hence favoring autolysis. The liver degeneration following chloro- form poisoning may, perhaps, be explained in a similar way, the cells behaving exactly as bacteria would do under the same conditions. Taylor^^ has analyzed several livers in degenerative conditions for amino-acids and found them only in one liver, which showed necrosis probabl}^ due to chloroform poisoning, and which was from a case clinically resembling acute yellow atrophy. Here he obtained 4 gm. of leucine, 2.2 gm. of tyrosine, and 2.3 gm. of arginine nitrate. Wald- vogel and Tintemann,^* in phosphorus livers, found an increase in protagon, jecorin, fatty acids, cholesterol, and neutral fat, while lecithin w^as decreased. Wakeman" found arginine, histidine, and ly- sine decreased in phosphorus livers in proportion to the total nitro- gen, indicating that the protein-splitting enzyme in this condition either picks out certain varieties of proteins first, or removes the nitrogen-rich constituents most rapidly. ^^ 52 Zeit. f. physiol. Chem., 1900 (30), 174. " Univ. of Calif. Public, (pathol.), 1904 (1), 43. " Cent. f. Path., 1904 (15), 97. " Berl. klin. Woch., 1904 (41), 1067. 5^ Considerable quantities of amino-acids of various sorts have been isolated from the liver in acute vellow atrophv and chloroform necrosis bv Wells (Jour. Exper. Med., 1907 (9), 627; Jour. Biol. Chem., 190S (5), 1-29); but the value of these figures is questionable because it is possible that the alcohol in which the tissues were kept before analysis was not strong enough entirely to prevent au- tolysis (Wells and Caldwell. Jour. Biol. Chem., 1914 (19), 57). 96 ENZYMES " It is probable that many poisons may injure the liver cells to such an extent that they cannot maintain their normal chemical equili- brium, but without destroying the autolytic enzymes. When this occurs, the liver undergoes autolysis, and we get marked degenerative changes with appearance of amino-acids in the blood and urine, reduction in coagulability of the blood and numerous hemorrhages, giving a picture both clinically and anatomically more or less like that of typical acute yellow atrophy. Chloroform is a poison that stops cell activities without destroying the proteolytic enzymes, hence the cells undergo autolysis, and, as a result, we have many cases of what appears to be acute yellow atrophy following chloroform anes- thesia. The liberation of HCl in the liver cells during chloroform poisoning, as demonstrated by Evarts Graham, ^'^ may be largely re- sponsible for the rapid disintegration of the liver in this condition. ^^ (See ''Acute Yellow Atrophy, " Chap, xx.) Probably the liver changes in puerperal eclampsia, and in streptococcus and other septicemias are of a similar nature. ^^ Autolysis of fatty livers in tuberculosis is said to yield more lactic acid than the livers from other conditions (Yous- souf).60 Postmortem changes are undoubtedly due to two factors, bac- terial action and autolysis. In tissues kept at a low enough tempera- ture to exclude bacterial action, but not so low as absolutely to stop enzyme action, ^^ there occurs a slow autolysis; this constitutes the "ripening" process of meat. Fish flesh may also ripen when made sterile in saturated salt solutions, as Schmidt-Nielsen^"^ has shown occurs with salted herrings, oxy-acids and xanthine bases being promi- nent among the products. The softening of muscles in rigor mortis is probably also an autolytic manifestation, as muscles contain proteases acting best in acid medium, and the muscle is known to become in- creasingly acid after circulation ceases within it. The short duration of rigor mortis when the body is kept warm, and its early disappear- ance when death has been preceded by muscular exhaustion (which increases the acidity), agree with this view. The oarly postmortem softening of many organs in pathological conditions is also probably an autolytic manifestation. Flexner-'^ has called attention to this in relation to the softening of the parenchymatous organs in acute infectious diseases, such as typhoid and septicemia. Scluunm noted great autolytic activity in a swollen spleen from a case of perityphlitis. " Jour. Exp. Med., 1915 (22), 48. 68 Quinan (.Tour. Med. Res. 1915 (.32), 73) found no ehanfre in the rate of in tritro autolysis of liver tissue from experimental cliloroform poisoning. It was found increased by phlorhizin (Satta and Fasiani, Arch. di. Fisiol., 1913 (11), 391). f-a Wells, Jour. Amer. Med. Assoc, 190G (40), 341. 6" Virchow's Arch., 1912 (207), 374. *' Some chemical chaufre may taivc i)lac.e at temiieratures as low as —2 and — 14 (Costantino, Arch. farm, sper., 1917 (24), 255). «Mrofmeister's Heitra^rc, l«K)3 (3). 2(57. AUTOLYSIS IN INFECTION 97 Histological changes are produced by autolysis in the organs after death that are, as might be expected, much like those seen in necrotic areas. ^^ At first the changes resemble those of parenchymatous de- generation (cloudy swelling), and often there is an apparent increase in fat, which is probabl}' due to liberation of masked fat through the destruction of the protein.*'' Nuclear staining is lost (karyolysis) , and eventually even cell forms become indistinguishable. (See p. 94 on structural changes of postmortem autolysis.) • Still-born children that have been carried for some time after death usually show considerable disintegration of the viscera, especially the liver. This is undoubtedly due to autolysis, which Schlesinger^^ has shown can begin before birth if the fetus dies in utero. Autolysis in Relation to Infection. — According to Conradi^Hhe substances produced in tissue autolysis have a decided inhibiting effect upon bacteria, which apparently depends upon the antiseptic proper- ties of the aromatic derivatives that are split out of the protein mole- cule in autolysis. This action is manifested not only in vitro, but the autolytic products will also render harmless lethal doses of certain bacteria if they are injected simultaneously with the bacteria into an animal. One specific class of products of autolysis which is strongly bactericidal is the soaps. ^^ It may well be questioned, however, whether enough of these substances ever accumulates in infected tissues during intra vitam autolysis to have much affect upon the in- fecting bacteria; yet this property may possibly explain the steriliza- tion of old pus collections and similar infected localized accumulations within the body. The bacteria themselves also produce autolytic products that are pow^erfully bactericidal. (See "Bacteria, "Chap. iv.) Blum®^ says that the autolytic products of Ij^mph-glands neutra- lize tetanus toxin, but are inactive against diphtheria toxin and cobra venom. Products from other autolyzed organs and from fresh lymph- glands were without influence on the tetanus toxin. The antitoxic principles of the autolytic product were destroyed by heating, weak- ened by acids and alkalies, and in other respects showed prop- erties strikingly like those of true antitoxin. It is quite possible that bacterial toxins may be destroyed by autolytic enzymes, for Baldwin and Levene^^ have shown that trypsin, pepsin, and papain destroy tetanus and diphtheria toxin, while tuberculin is destroyed by trypsin, but not readily by pepsin, possibly because it is of a nucleoprotein «^ More fully discussed bj' Wells, Jour. Med. Research, 1906 (15), 149. ^* Siegert (Hofmeister's Beitr., 1901 (1), 114) found no actual increase in fats and fatty acids in autolysis even when an increase was apparent histologically , although ether-soluble materials of other nature than fat mav be increased. See also Hess and Saxl, Virchow's Arch., 1910 (202), 149. " Hofmeister's Beitr., 1903 (4), 87. «« Hofmeister's Beitr., 1901 (1), 193. See also Bilancioni" and Almagia.^* 8' See Lamar, Jour. Exp. Med., 1911 (13), 1. 88 Hofmeister's Beitr., 1904 (5), 142. " Jour. Med. Research, 1901 (6), 120. ^ 7 98 ENZYMES nature. The leucocytic proteases, however, seem not to attack either toxins or Uving bacteria (Jochmann). Bertohni'° states that auto- lyzing hver will destroy diphtheria toxin. On the other hand, there are many pathogenic bacteria which do not secrete their toxic materials, but store them up within the cell body, e. g., typhoid, cholera, and, indeed, the majority of pathogenic forms. These endotoxins are probably hberated from the bacteria only through digestion of their cells, either by their own autolytic enzymes or by the enzymes of the infected tissues and leucocytes. Leukemia. — The abundant elimination of uric acid and other pu- rine bodies in the urine in leukemia testifies to the great amount of destruction of nucleoprotein that is going on during the disease, and these are probably derived from the autolysis of leucocytes, which per- haps depends on the relatively large proportion of leucocytes to serum. Schumm'^^ has found that leukemic spleens and bone marrow autolyze rapidly and completely, and he isolated many of the cleavage products of protein digestion from such autolysates. Leucocytes from myeloid leukemia hquefy alkaline gelatin vigor- ously, but those from lymphatic leukemia do not; the liquefaction is inhibited by normal serum (Stern and Eppenstein).''^ By the serum plate method this observation has been much extended, and the hetero- lytic action of the leucocytes has been found limited to the neutro- phile granules. In neutral media evidence is obtained of the presence of protease in the lymphocytes of chronic lymphatic leukemia and the leucocytes of acute and chronic myeloid leukemia; maltase, lipase and amylase are found in both types of cells, and oxidase in the granular cells derived from the marrow (Morris and Boggs).'^^ v. Jaksch,"^ Erben,'^^ and others have noted the occurrence of peptones and albu- moses in leukemic blood, particularly if removed postmortem. The improvement in leukemia that follows .r-ray treatment is associated with an increased nitrogen elimination, probably due to autolysis of disintegrating cells, '^^ although a;-rays have no appreciable effect upon the leucocytic proteases in vitro (Miiller and Jochmann). (See also "Leukemia," Chap, xiii.) Tumors. — Probably because of the great amount of necrosis that is constantly going on in all malignant growths, with subsequent di- gestion of the dead cells, autolytic products are present in them in very considerable amounts. This was first demonstrated by Petry," '"' Biochem. Zeit., 1913 (48), 448. ^' Hofmeister's Beitr., 1903 (3), 576; 1905 (7), 175. ^^ See discussion of leucocytic enzymes, p. 89. Longcope and Donhauser (Jour. Exper. Med., 1908 (10), G18) found proteases in the large lymphocytes in acute leukemia, which were most active in an alkaline medium. " Arch. Int. Med., 1911 (8), 806. '^ Zeit. f. physiol. Chem., 1892 (16), 243. '^Zeit. f. klin. Med., 1900 (40), 282; Zeit. f. Heilkunde, 1903 (24), 70; Ilof- meiter's Beitr., 1904 (5), 461. '6 Musser and Edsall, Univ. Penn. Med. Bull, 1905 (18), 174. " Zeit. f. physiol. Chem., 1899.(27), 398; Hofmeister's Beitr., 1902 (2), 94. AUTOLYSIS IN TUMORS 99 who fouiul that cai-ciiionias of the breast contained much of their nitrogen in compounds not coagulated by heat, while in the normal gland practically all is coagulable. He also demonstrated an autolytic property in tumor tissue, showing that tumor cells do not difTcr in this respect from normal cells. Beebe'** found products of autolysis con- stantly present in several tumors; namely, a carcinoma of the broad ligament, a hypernephroma, an angiosarcoma, and a round-cell sarcoma. Ncuberg^'-^ found that while, according to other observers, most enzymes, as well as bacteria, are very susceptible to the action of radium rays, the autolytic enzymes of cancer cells are an exception, for cancer tissue exposed to radium undergoes autolysis much faster than cancer tissue not exposed to radium; x-rays are less active in this respect. He attributes the effects of radium on cancer to its deleterious effects on the oxidizing and other enzymes of the cells, destroying their activities, which results in destruction of the cells by the autolytic enzymes.^" A cancer of the stomach was found to contain autolytic enzymes capable of digesting lung tissue (pepsin was excluded) and autolyzed cancers yielded much pentose. Blu- raenthal and Wolf ^^ believe that tumor tissues have particularly active autolytic enzymes, since liver tissue added to tumor tissue underwent autolysis much more rapidly than normal; but tumors do not cause digestion of serum plates unless many leucocytes are present (Mliller and Kolaczek).^- Cancer extracts digest peptids in ways cUfferent from normal tissues, which seems to indicate some fundamental ab- normality in their metabolism (Abderhalden,^^ Neuberg).^^ The al- most constant presence in gastric juice of patients with carcinoma of the stomach, of ereptases hydrolyzing proteoses and peptids, is generally attributed to the disintegration of the cancer with libera- tion of these enzymes. ^^ Tumors also contain nuclease^Ho disintegrate their nucleic acid, and the same outfit of purine-splitting enzymes as normal tissues," so that in regard to the nucleoproteins of tumors, - autolysis follows the same course as in normal tissues. '8 Amer. Jour. Physiol., 1904 (11), 139. ^^Zeit. f. Krebsforschung, 1904 (2), 171; Berl. klin. Woch., 190-4 (41), 1081; ibid., 1905 (42), 118; Arb. Path. Inst. Berlin, 1906, p. 593. 8«Wohlgeinuth, Berl. klin. Woch., 1904 (41), 704, found that autolysis in tuberculous lung tissue was three or four times more rapid when exposed to radium rays. ^Heile (Arch. klin. Chir., 1905 (77), 107) looks upon the favorable effects of x-rays as partly produced by their liberation of autolytic enzymes from the leucocytes. «' Med. Klinik.', 1905 (1), No. 7. ^■- Miiller and Kolaczek, Miinch. med. Woch., 1907 (54), 354; Hess and Saxl }\ifn. khn. Woch., 1908 (21), 1183; Kepinow, Zeit. f. Krebsforsch., 1909 (7)' 517. ' "Zeit. physiol. Chem., 1910 (66), 277. 8* Biochem. Zeit., 1910 (26), 344. soggp Jacques and Woodyatt, Arch. Int. Med., 1912 (10), 5G0; Hambureer Jour. Amer. Med. Assoc, 1912 (59), 847. ' ^^ Goodman, Jour. Exp. Med., 1912 (15), 477. " Wells and Long, Zeit. Krebforsch., 1913 (12), 598. 100 ENZYMES The non-cancerous livers of cancerous patients were found by Yous- souf^° to produce more lactic acid during antiseptic autolysis than did livers in other conditions. Autolysis of organs of cancer patients is about as rapid as normal (ColwelP^). Several observations have suggested that tumor tissues might contain proteolytic enzymes dif- fering from those of normal tissues especially in their ability to digest heterologous normal tissues, but at present this work needs confirma- tion and amplification before it can carry the weight of speculation which has been heaped upon it.^^ Micheli and Donati^° attribute the hemolytic properties possessed by extracts of malignant tumors to the products of autolysis that are present, which Petry has also demonstrated to produce hemolj'sis. Emerson^i attributes the disappearance of HCl from the gastric juice in carcinoma of the stomach to neutralization by basic products of autolysis, a hypothesis that may well be questioned. (See also "Tumors," Chap, xix.) Various other intracellular enzymes have been described, which for the most part have as yet no significance in pathology. An exception is fibrin ferment, ■which will be considered fully in discussing thrombosis. Ferments coagulating milk seem 'to be widely spread in the tissues. The precipitation of plastein from proteose solution by organ extracts (Niirnberg) may be either the effect of a •coagulating ferment or due to reverse action of the proteases. Ferments split- ting specifically maltose, lactose, sucrose, glucosides, and nucleoproteins have been described, and the glycogenolytic ferment is probably nearly imiversally pres- ent. Other enzymes decomposing amino-acids into ammonium compounds may also exist. The enzymes acting specifically upon the nucleic acids and the purine bodies are discussed in Chapter xxiii. 88 Arch. Middlesex Hosp., 1910 (19), 55. 83 See, for example, Rulf, Zeit. Krebsforsch., 1906 (4), 417; Muller, Cent. inn. Med., 1909 (30), 89. 90 Riforma med., 1903 (19), 1037. " Deut. Arch. klin. Med., 1902 (72), 415. CHAPTER IV THE CHEMISTRY OF BACTERIA AND THEIR PRODUCTS STRUCTURE AND PHYSICAL PROPERTIES^ In structure, as in nearl}' all other respects, bacterial cells stand intermediate between the cells of ordinary plant and animal tissues. Their cell wall seems to be generally more highly developed than that of animal cells, and less so than the wall of most plant cells. The much vexed question as to the existence or non-existence of a nucleus seems to be best answered by Zettnow, who considers that the portion of the bacterial cell usually made evident bj^ ordinary staining methods con- sists of a mixture of nuclear substance (chromatin) with non-chromatic substance {end o plasm) ; the outer membrane, which requires special methods for its satisfactory demonstration, consists of a modified cytoplasm {ectoplasm). Some bacteria consist chiefly of chromatin (e. g., vibrios), but the proportion of the different elements varies greatly, not only in different varieties, but also in the same variety under different conditions. The fact that the chromatin is not aggre- gated into the usual nuclear form may be ascribed to the low stage of development reached by bacteria in the scale of evolution; or, as Vejdovosky has suggested, to the extremely rapid rate of cell division in the bacteria which prevents the chromatin from appearing in the resting stage which a nucleus constitutes. Finer structures within the bacterial cell have as yet been only imperfectly discerned. The thickness of the ectoplasm varies greatly even in the same species, being generally greatest in older cultures. In some forms the ectoplasm may constitute one-half of the total mass of the cells. The capsule seems to arise through a swelling of the ectoplasm, and is probably present in at least a rudimentary stage in all bacteria (Migula). Plasmolysis and Plasmoptysis. — Under conditions of altered osmotic pressure the bacterial cell behaves quite similarlj^ to the plant cell. 2 If placed suddenly in a solution of higher osmotic pressure than 1 In this chapter references will not generally be given that can be found by consulting Kolle and Wassermann's Handbuch. A general consideration of the Biology of the Bacteria, including references to the effects of light, heat, osmotic pressure, etc., is given by Miiller, Ergb. der Physiol., 190-1 (4), 138; concerning their chemistry see H. Fischer, Lafar's Handbuch der Technischen Mykologie, 1908 (1), 222. ''Literature, see Gotschlich, Kolle and Wassermann's Handbuch, vol. 1. 101 102 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS the one in which it has been, the cell contents shrink away from the cell wall {plasmolysis) indicating that there exists a semipermeable mem- brane through which water passes more rapidly than salts. If the change in osmotic pressure is gradual, the bacteria accomodate them- selves to it by the slow diffusion of the salts through the cell membrane, indicating that it is not absolutely semipermeable. Different bac- teria behave differently, some bacteria not being plasmolyzed by solutions that plasmolyze others. As a rule, old bacteria plas- molyze more rapidly than young, and in some varieties there seems to be a spontaneous plasmolysis, to which has been attributed the irregular staining of diphtheria and tubercle bacilli, the polar staining of plague bacilli, etc. Plasmolysis occurs only in living bacilli, but does not nec- essarily cause death. The Gram-staining bacteria cannot generally be plasmolyzed, and contain more water. ^ When bacteria pass from solutions of higher osmotic concentration into solutions of lower concentration, the phenomenon of plasmoptysis is produced. The cell contents swell until the cell wall gives way at some point, and then exude as glistening drops, which may become detached from the wall and escape free into the fluid. Plasmoptysis is shown best by bacteria that have been grown on salt-rich media before being placed in the salt-free fluid. Not all varieties of bacteria can be made to undergo this change, depending probably upon the degree of permeability of their cell membranes for salts. The ex- posure of the naked cell contents to the hypotonic fluid outside the cells makes plasmoptysis more serious for bacterial life than plasmo- lysis, but how often, if ever, either process plays a part in the resistance of infected animals against bacteria is unknown. The resistance of bacteria to direct pressure is striking; spore bearers may not be killed under direct pressure of 12,000 atmospheres for 14 hours, and non- spored bacteria resist 3,000 but not 6,000 atmospheres.^ Chemotaxis.^ — Just as with unicellular animal organisms, bacteria respond to chemotactic influences, in general being attracted bj'' sub- stances favorable for food, such as peptone, amino acids, ^ dilute potas- sium salts, etc., and being repelled by harmful substances, such as strong acids and alkalies. Attempts have been made to separate different organisms in mixed cultures by means of their response to chemotaxis, but without striking success. It is possible that chemo- taxis may play a part in the localization of bacteria from the blood stream in favorable localities, just as leucocytes arc attractd to points of injury, but this has not been demonstrated. (The chemotactic influence of bacteria upon leucocytes is discussed in Chapter xi.) 3 Nioolle and Alilairc, Ann. Inst. Pasteur, 1009 (23), 547. *Larfc'on, Hartzell and Diehl, Jour. Infect. I)is., lOlS (22), 271. ^ Concerning the adsorption of l)acteria sec Bechhold, Kolloid-Zcitsrhr., 1918 (23), 35. " Pringsheiin and Ernst, Zcit. pliysiol. Clieni., 19U) (97), 17(). COMPOSITION OF BACTERIA 103 CHEMICAL COMPOSITION This varies greatly, not only between different species, but even in the same species grown on different media ;^ in this respect bacteria are much more modified by their environment than are higher or- ganisms. On the other hand, they can develop in solutions containing only a few of the simplest organic and inorganic compounds and syn- thesize the complex components of their cells, as well as enzymes, toxins, pigments.^ They usually contain between 80 and 90 per cent, of water. Grown on a salt-rich medium they yield much ash; grown on a peptone-rich medium they contain much protein; grown on a fat- rich medium they contain much material solulile in ether. Cholera vibrios grown on a bouillon medium contained 69.25 per cent, of pro- tein, and 25.87 per cent, of ash, whereas the same organism grown on Uschinsky's medium, which contains no proteins but only various simple chemical compounds, contained but 35.75 per cent, of protein and 13.7 per cent, of ash (Cramer). Even in the same medium two different strains of the same organism may show equally great dif- ferences: Two strains of cholera vibrios grown on the same medium showed respectively 65.63 per cent, and 34.37 per cent, of protein. It is evident, therefore, that quantitative analyses of bacteria show nothing as to- their nature, and on account of the extreme limits of their variation are practically valueless. The specific gravity of bac- teria, generally between 1.12 and 1.345, also varies wdth media and age.^ In an electric field they move towards the anode. ^"^ Qualitatively the variations are not so great — all bacteria contain proteins, lipoid substances, and salts, of which phosphates are most prominent in the ash. The character of the proteins and fats of bacteria grown on peptone bouillon is quite the same as when they are grown on protein-free media.'' The older analyses of bacterial constituents are of little value. Recent studies prove that the chief constituent of the cell contents is a true nucleoprotein (Iwanoff^^) con- taining some sulphur and iron; probably many of the "pyogenetic pro- teins," "bacterial toxalbumins," "bacterial caseins" of earlier investi- gators are true nucleoproteins.^^ The stainable substance of anthrax bacilli behaves as if it were a chromatin, while the spores resemble ^ See Dawson, Jour. Bact., 1919 (4), 133. ^ Concerning fundamentals of nutrition of bacilli see Koser and Rettger, Jour. Infect. Dis., 1919 (24), 301; Long, Amer. Rev. Tuberc, 1919 (3), 86. 9 Stigell, Cent. f. Bakt., 1907 (43), 487. '" Buxton, Zeit. physikal. Chem., 1906 (57), 47; Girard and Audubert, Compt. Rend. Acad. Sci., 1918 (167), 301. Concerning the decirical condxiclivity of bacteria see Thornton, Proc. Royal Soc, London, Sec. B., 1913 (85), 331. " Tamura, Zeit. physiol. Chem., 1913 (88), 190. '- Hofmeister's Beitr., 1902 (1), 524; bibliography by Lustig, KoUe and Wasser- mann'8 Handbuch, 1913 (ii), 1362. ' ' The purity of many of the preparations worked with as bacterial nucleopro- teins, is very doubtful. (See Wells, Zeit. Immunitat., 1913 (19), 599.) 104 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS linin (Rozicka).^^ The predominance of nuelein compounds is shown by Ruppel's summary of the composition of dried tubercle baciUi, namely, in per cent.,tuberculonucleic acid, 8.5; nucleo-protamine, 24.5; nucleo-protein, 26.5; fatty matter, 26.5; inorganic, 9.2; insoluble "proteinoid" residue, 8.3. In a water bacillus Nishimura found xanthine, guanine, and adenine, indicating the presence of nucleo- protein; others have found that bacterial nucleoproteins split off pentoses, as do the nucleoproteins of higher cells. If it is true that bacterial nucleo-proteins contain pentose it ranks them with the plant nucleo-proteins, for animal nucleic acids contain hexose. On the other hand, Levene found in bacterial nucleic acid the pyrimidines, thj-mine and uracil, which are respectively characteristic of animal and vege- table nucleic acids. Mary Leach^^ found evidence that the colon bacillus is largely made up of nuelein or glyco-nucleoproteins, but contains no cellulose. Other proteins, namely, globulins and nucleo- albumins, have also been described as constituents of the bacterial plasma. The^complete amino-acid content of bacterial protein does not seem to have been worked out, although the workers in Vaughan's labora- tory have identified many of the usual amino-acids of proteins among the products of hydrolysis of bacteria. ^^ Analysis of B. mesentericus shows it to be deficient in diamino-acids, tyrosine, glycine, and to contain 16.6 per cent, of glutamic acid." Tamura^^ found phenyl- alanine and valine high in tubercle bacilli and very low in B. diph- therice, in which tyrosine is more abundant. In an azobacterium, lysine has been found especially abundant.'^ Cystine has been lacking in several analyses. Tamura^° also found that bacteria can synthesize from simple nonprotein media the purines, phosphatids and the typical proteins containing the aromatic amino-acids. This syn- thetic activity of bacteria, in view of the large quantity of bacterial substances in feces, may possibly be of importance in metabolism studies, leading to erroneous conclusions as to utilization or sjmthesis of proteins by the subject.^' In common with other forms of cellular life, bacteria require certain specific substances, "vitamins," to permit of their growth ;22 also they produce substances with the value of vitamins.-^ 1^ Arch. Entwicklungsmk., 1906 (21), 306. /^ Jour. Biol. Chem., 1906 (1), 463. Full bibliography on Chemistry of Bac- teria. See also Vaughan, "Protein Split Products in Relation to Immunity and Disease," Philadelphia, 1913. '« See Wheeler Jour. Biol. Chem., 1909 (6), 509. '^ Horowitz- Wiassowa, Arch. Sci. Biologique, 1910 (15), 40. '8 Zeit. phvsiol. Chem., 1913 (87), 85; 1914 (89), 289. 19 Omelian.sky and Sieber, Zeit. physiol. Chem., 1913 (88), 445. 20 Zeit. physiol. Chem., 1913 (88), 190. 2' Osborne and Mendel, Jour. Biol. Chem., 1913 (18), 177. "See Davis, Jour. Infect. Dis., 1917 (21), 392; Kligler, Jour. Exp. Med., 1919 (30), 31. " Pacini and Russell, Jour. Biol. Chem., 1918 (34), 43. BACTERIAL CARBOHYDRATES Ai\D LII'IXS 105 The slimy material produced in cultures by some varieties of bac- teria is, at least for certain forms, a body closely related to or identi- cal with true mucin, ^^ but in certain cases {B. radicicola) it is a gum related to the dextrans and free from nitrogen (Buchanan). ^^ Tu- bercle bacilli grown for many years on artificial media may produce a true mucin (Weleminsky).-^ Heim" considers that anthrax bacilli also produce mucin. Some nonpathogenic bacteria contain granules of sulfur in their protoplasm, and others have noteworthy quantities of iron in the sheath. Bacterial Carbohydrates. — -The earlier descriptions of cellulose or hemicellulose in the cell membrane of bacteria have been contested. ^^ Numerous investigators have reported that the insoluble bacterial cell wall consists chiefly of chitin, which on being split with acids yields 80 to 90 per cent, of the nitrogenous carbohydrate, glucosamin.^'^ The distinction is a verj^ important one, since cellulose is a typically vege- table product, while chitin is equally typically animal in origin, being found chiefly in the shells of lobsters and crabs, the wings and cover- ings of flies, beetles, etc. Chitin seems to be a polymeric form of glucose-amine,^" an amino-carbohydrate, just as cellulose is a polymer of a simpler carbohydrate. Other carbohydrates seem to be scanty in the bacterial cell, but Tamura^^ does not accept the chitinous nature of bacterial carbohydrate, finding in tubercle and diphtheria bacilli a hemicellulose, apparently a pentosan yielding 1-arabinose on hydro- lysis. Wester^^ found no chitin in several varieties of bacteria, and cellulose only in B. xylinum; he therefore considers it probable that bacterial cell walls do not alw^ays consist of the same substance. Cra- mer could find no glucose in any variety, although there are some bac- teria that contain material reacting like starch with iodin. Levene,^^ however, found in B. tuberculosis a substance with some of the properties of gl3^cogen. Bacterial Fats. — By staining methods, fats have been recognized in many species, and by extraction with fat solvents lecithin, cholesterol, simple fats, and specific bacterial fats have been isolated; this is par- ticulary true of B. tuberculosis.^^ Numerous studies of these fats of -* Rettger, Jour. Med. Research, 1903 (10), 101. " Cent. f. Bakt., II Abt., 1909 (22), 371. -^ Berl. klin. Woch., 1912 (49), 1320. " Miinch. med. Woch., 1904 (51), 426. 28 However, Dreyer (Zeit. ges. Brauw., 1913 (36), 201) states that the cell wall of yeasts contains a hemicellulose and a manno-dextran. See also Kozniewski, Zeit. physiol. Chem., 1914 (90), 208. =9 See Viehofer. Ber. Deut. Chem. Ges., 1912 (30), 443. '" Morgulis states that chitin consists of two parts, one containing'all the glucose and amino groups, the other being a stable nitrogenous compound yielding no glucose. (Science, 1916 (44), S66.) " Zeit. phvsiol. Chem., 1914 (89), 304. 22 Pharm. Weekblad, 1916 (53), 1183. "Jour. Med. Research, 1901 (6), 135. ^* See Camus and Pagniez, Compt. Rend. Soc. Biol., 1905 (59), 701. 106 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS B. tuberculosis have been made^^ and by using different extractives, from 20 to 40 per cent, of the entire weight of the bacilh has been found sokible in fat solvents. Kreshng found that the substance soluble in chloroform had the following composition: Free fatty acid 14.38 per cent. Neutral fats and fatty acid esters 77.25 per cent. Alcohols obtained from fatty acid esters 39.10 per cent. Lecithin 0. 16 per cent. Substances soluble in water 0.73 per cent. Bulloch and Macleod found that ethereal extracts did not contain the acid-fast substance which they consider to be a wax-like alcohol, soluble in hot, but insoluble in cold absolute alcohol or in ether. The simple fats seem to be formed by oleic, isocetinic, and myristinic acids, and there is some lauric acid in the form of a soap. Kozniewski'® obtained what seemed to be a lauric acid ester of a dodecyl-alcohol, and Biirger^^ attributes the odor of tubercle bacilli to the presence of salicylic aldchj^de. Cholesterol could not be found in tubercle, diphtheria and other bacteria examined by Tamura, although there probably are lipochromes giving the cultures their color. ^^ There is still much disagreement as to whether the acid fastness of tubercle bacilli depends upon waxes, alcohols, fatty acids, or lipoid-protein compounds. ^^ It must be admitted that a high content of fatty ma- terials is regularly present in acid-fast bacilli; thus, in an acid-fast bacillus isolated from leprous lesions, 34.7 per cent, of fats, fatty acids and cholesterol, and 1.7 per cent, of lecithin were found by Gurd and Denis. ^^ Miller"*! attributes the unstained, spore-like areas of tubercle bacilli to oleins, as bacilli grown on olive and sperm oil show a marked decrease in acid fast areas. Tamura*^ states that the phosphatids of B. tuberculosis and a saphrophyte exaniinetl by him were not lecithin but a diaminophos- phatid, although diphtheria bacilli seemed to contain lecithin.'*^ He found in both a high molecular alcohol, "mykol," to which he ascribes acid- and Gram-fastness. In a Gram-negative bacillus** he found lecithin, but no cholesterol or mykol. Apparently the fats of tubercle bacilli resemble in character and complexity the "waxes" of j)lants (Burger), ^^ which are called "ccrolii)oids" b}^ Czajiok. By growing tubercle bacilli on suitable media they can l)o made 1o lose their acid- "^ For literature see BuUocli and Maclood, .Tour, of lly^fiene, 1004 (4), 1. ^* An/eifzicr d. Akad. Wiss. KraUau, Matli.-naturwiss Kl., 1912, p. 942. " liiocliein. Zeit., 191(> (7S), 155. " I'anzer (Zcil. jjliysiol. CIhmii., 1912 (78), 414) could not demonstrate cho- lesterol in tiihorcle Invcilli l)ut did lind a small amount of some substance imitinp; with (lij:;it()nin. "•Sec Camus and Pnnniez, Pre.sse Med., 1907 (15), 05; Deyke, Munch, mcd. Woch., 1910 (57), (VM. "Jour. J';\i)er. Med., 1911 (11), COG. ■" .Jour, i'utii. and linel., 191(1 (21), 11. ^^Zcit. i)iivsi(>l. Clieiii., 1913 (S7), S5. "//;i(/., 1911 (S9), 2,S9. ** Ilnd., 1914 (90), 2S0. BACTERIAL LII'INS 107 fast property, although still Gram-positive (Wherry).''^ The observa- tion of Miss Sherman, ^^ that tubercle bacilli are almost absolutely impermeable to fat-soluble dyes which stain their isolated fats well, and her corroboration of Benians' demonstration that acid-fastness depends on the integrity of the bacillary envelope, make the role of the fatty substances uncertain. The high content in unsaturated fatty acids gives acid-fast bacteria a high antitryptic power, which may be concerned in the defense of the bacteria and also in the persistence of caseous material in tubercles (Jobling and Petersen).*^ The oily material obtained by extracting tubercle bacilli with cold ether is non-toxic, while the waxy material extracted with hot alcohol produces foreign body tubercles (Morse and Stott).*^ By staining with Sudan III, Sata^^ demonstrated fats, not only in the acid-fast bacilli, but also in anthrax, Staphylococcus aureus, B. mucosus, and actinomyces; but not in diphtheria, pseudo-diphtheria, plague, cholera, and chicken cholera bacilli, or in members of the colon group. ^^ Only a few bacteria form fat on agar free from gly- cerol, but potato is a favorable medium. Ritchie^^ obtained positive fat staining in B. diphtherice and anthracis, but not in S. pyogenes aureus or M. tetragenus, although these last forms contain chemically demonstrable lipins. Analyses of different bacteria show a relatively low content of lipins as compared with tubercle bacilli, varying from 1.7 per cent, in B. subtilis to 8.5 per cent, in staphylococci (Jobling and Petersen). ^2 However, the degree of unsaturation of the fatty acids is less with tubercle bacilli than with other bacteria examined by these authors. Extensive studies of bacterial fat stains are reported by Eisenberg,^^'' but practically nothing is known of the character of the fatty or lipoid constituents of bacteria outside the acid-fast group. Spores differ from their parent bacteria in containing a much greater propor- tion of the soUd constituents and less water. In molds Drymont found that the spores contained over 60 per cent, of dry substance, and almost all the water was so held as to resist drj'ing by temperatures below boiling; the drj' substance is very rich in protein and poor in salts. As the spores may lose their chromatin content without loss of capacity to propagate, it would seem that this is not a nuclear chromatin but merely a reserve food supply.^' The wall of the spore con- sists of a ''cellulose-like " substance and a very hygroscopic extractive matter. The great resistance of spores to drying and to heat can be readily understood in view of these facts. They contain, and perhaps secrete, active enzymes (Effront).*' Flagella also seem to be composed of a relativelj' condensed protein. "Jour. Infect. Dis., 191.3 (13), 144. " Jour. Infect. Dis., 1913 (12), 249. ^' Jour. Exp. Med., 1914 (19), 239. « Jour. Lab. Chn. Med., 1916 (2), 159. « Cent. f. allg. Path., 1900 (11), 97. ^^ Auclair (Arch. M;'d. Exper., 1903 (15), 725) contends that the ether and chloroform extracts of many pathogenic bacteria contain important toxic sub- stances. Holmes (Guy's Hosp. Reports, 1905 (59), 155) states that injection of fatty acids from tubercle bacilli into rabbits causes a lymphocytosis. 6' Jour. Pathol, and Bact., 1905 (10), 334. "Jour. Exp. Med., 1914 (.20), 456. 52" Virchow's Archiv., 1910 (199), 502. " Ruzicka, Cent. f. Bakt., 1914 (41), 641. " Mon. sc. Quesneville, 1907, p. 81. 108 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS Staining Reactions. — The staining reactions of bacterial cells are much as if the bacteria consisted entirely of chromatin, so that at one time the theory prevailed that bacteria consisted merely of a nucleus and a cell wall, without any true cytoplasm. The demon- stration of abundant nuclcoprotein in the contents of bacterial cells explains their staining affinity for basic anilin dyes. Owing to some unknown differences in composition, not all bacteria are stained equally well by the same basic dyes." Although the staining of bac- teria depends upon a chemical reaction between the nucleoproteins and the basic dye, yet the combination is not usually a firm one, being readily broken by weak acids in most cases. That the decolorization of bacteria depends upon dissociation of the dye-protein compound is shown by the fact that absolutely water-free alcohol will not de- colorize dry bacteria, nor do water-free alcoholic solutions of dyes stain dehydrated bacteria. There seems to be a marked difference in the accessibility of dead and living bacteria to stains; thus, only dead bacteria stain with AgNOs." Gram's Method^' of staining has been ascribed to the formation of an iodin- pararosanilin-protein compound which is not easily dissociated by water in the case of bacteria that stain by this method, and which is readily dissociated and dissolved out in the case of bacteria that do not retain the stain. Only para- rosanilin dyes (gentian violet, methyl violet, victoria blue) form such combinations, the rosanilin dyes not being suitable.^'* It is probable, especially from the obser- vations of Deussen, that the nucleoproteins are the essential cell constituents, and other cells than bacteria {'.<;.. sperm) may be Gram-positive. The relation of bacterial protein to Gram staining is shown by the fact that trypsin will digest killed bacteria which are Gram-negative, but not Gram-positive forms; gastric juice attacks only a few Gram-positive bacteria. ^'^ They are also more resistant to alkalies, 1 per cent. KOH dissolving only the Gram-negative bacteria. Brundy'" considers that they are more permeable to iodin, so that a more central iodin-dye precipitate occurs, and Eisenberg*'' suggests that lipoid- protein compounds in the surface are important, in support of which is the obser- vation that ether extraction of staphylococci renders them negative to Gram's method, while colon bacilli treated with lecithin become positive.*^'- Jobling and Petersen^^ have also found the lipoids of Gram-positive bacteria more resistant to extraction by fat solvents than lipoids of Gram-negative bacteria, and Tamura^' found that the lipoid extract contains the bacterial element responsiVile for Gram staining. The first-named authors suggest a relation between the high content in unsaturated fatty acids, with the higli affinity for iodin, and the positive Gram staining. On the other hand, Ilottinger''' attributes (!ram staining solely to the degree of disjjersion of the nucleo-proteins, which he believes \o be higher in the '* The presence of serum interferes with staining, probably from protective colloid action (Fleisher, .lour. Med. Res., 1017 (3()), 31.) '« Nyfeldt, Nordiskt Med. Arkiv, 1917 (^O), 1S4. " Full review by Deussen, Zeit. llyg., 191S (Sf)), 23r). '* Any metallic iodid may be substituted for Kl (Leidv, .lour. Lab. Clin. Med., 1919 (4), 3.04). '" Hingers, Schermann and Schrciber, Zeit. f. Ilyg., 1911 (10), 119; ^^'einkoff, Zeit. Imuiunilat., 1912 (11), 1. ""Cent. f. iiakt., ii Abt., 190S (21), 62. «' Cent. f. Hakt., 1910 (.')(>), 193. «Mour. rath, and Hact., 1911 (Hi), 140. ''Zeit. i)livsiol. Ghem., 1914 (S9), 2S9. «' Cent. f. iiakt., 1910 (7(1), 3t)7. BACTERIAL ENZYMES 109 Gram-nepativc forms. Benians*^ has found that crushed Gram-positive bacteria are proni])tly decolorized, indicating that the dye and the cell contents do not form an insoluble comjiound, but that the l)ac"terial cell \vall is the chief factor in detennininK CIram jiositiveness; presumably the iodin renders the cell membrane impermeable to alcohol. This important contribution has been confirmed, as far as the staining of tubercle bacilli is concerned, by Hope Sherman,*" who corrob- orates the finding of Benians that if the bacilli are not intact they are neither acid fast nor (Irani ])ositive. The same is true of yeast cells (Henrici)/' but Dcus^en states that press juice from yeast (Buchncr's zymase) contains Gram-positive £;ranules. Bacterial Enzymes^^ The metabolic processes of bacteria seem to be closely dependent upon enzyme action, just as with higher cells. Liquefaction of gelatin is a familiar example of the enzyme action of bacteria; and since the filtered cultures of liquefactive bacteria are also capable of digesting gelatin, the enzymes are evidently excreted from the cells. Dead bacteria, killed by thymol or by other antiseptics that do not destroy proteolytic enzymes, will also digest gelatin. Numer- ous investigations have established the wide-spread occurence of many soluble enzymes both in bacteria and in their secretions, indi- cating that bacterial cells are as dependent on enzymes for the pro- duction of their metabolic activities as are higher types of cells, and that these enzj^mes are not only present as intracellular constituents, but that they also escape from the cells. Even the spores contain active enzymes. ^^ A striking property of bacteria is their reducing power, which has led to the introduction of selenium and tellurium salts, which are reduced to the metals, as an index of bacterial life and activity (Gosio). The diffusion method of Wijsman, or, as it is more frequently called, auxanogmphic method of Beijerinck, offers a relatively simple means of detecting the presence of extracellular bacterial enzj-mes. Eijkman^" in particular has used this method, which consists of mix- ing agar with milk, or starch, or whatever material is to serve as the indicator of the enzyme action; the agar is then inoculated with bac- teria and plated (or else the bacteria are inoculated as a streak on the surface of the agar) . About each colony there will appear a zone of clearing in the medium if it produces enzymes digesting the admixed substance. By this means Eijkman found that all bacteria that produce enzymes digesting gelatin also digest casein, and those that do not digest gelatin are equally without effect on casein; therefore, it is probably the same enzyme that digests both. As the hemolytic 65 Jour. Path, and Bact., 1912 (17), 199. 66 Jour. Infec. Dis., 1913 (12), 249. 6^ Jour. Med. Res., 1914 (30), 409. 68 See Fuhrmann (" Vorlesungen iiber Bakterienenzyme," Jena, 1907) for com- plete bibliography to that date. 65 Effront, Mon. sc. Quesneville, 1907, p. 81. " Cent. f. Bakt., 1901 (29), 841. 110 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS action of bacteria is not constantly related to their gelatin-dissolving property, the hemolysis probably is produced by other means than the proteolytic enzymes. ^^ A few pathogenic bacteria (anthrax, cholera and some strains of hemolytic streptococci^^) digest starch" and B. pyocyaneus, StapJnjlococcus pyogenes aureus, and B. prodigiosus all produce fat-splitting enzymes demonstrable by tliis method."'* B. pyocyaneus, Eijkman found, digested elastic tissue readily," as also did a bacillus resembling B. subtilis obtained from the tissue of a gan- grenous lung. Rennin is produced by many bacteria, as is shown by their coagu- lating milk, independent of any acid reaction," and protease from pyocyaneus causes "plastein" formation in albumose solutions (Zak),^^ Bacteria which give negative results by the plate method may con- tain active lipase demonstrable in killed bacteria by direct action upon fats and esters, these lipases behaving exactly like the lipase of animal tissues (Wells and Corper);" staphylococcus and pyocyaneus are more actively lipolytic than B. coli, B. dysenierice and B. tuber- culosis. Urease seems to be widespread." Tubercle bacilli contain enzymes resembling lipase, trypsin, erepsin, nuclease and urease, but not amylase,- elastase or invertase.^° Schmailowitsch^^ stated that the amount and nature of enzymes produced by bacteria is modified by the amount and nature of their food, but Jordan found that gelatinase is produced by bacteria grow- ing on non-protein media; he failed entirely to support the statement of Abbott and Gildersleeve^^ that bacteria grown on gelatin produce much more active gelatin-dissolving enzyme than do bacteria grown on bouillon. Diehl^^ found that bacteria grown on media containing no organic nitrogen produce no proteolytic enzymes, and the enzyme content of bacteria is much modified by the composition of the media, depending on the character of the amino-acids present rather than the proteins themselves. Jacoby'*'* has made extensive studies on the ^' See Jordan, Biol. Studies by the pupils of W. T. Sedgwick, 1906, p. 124. " Tongs, Jour. Aincr. Med. A.ssoc., 1919 (73), 1277. '•' In relation to carbohydrate enzymes, the extensive studies of Kendall (Jour. Biol. Chem., 1912, vol. 12) should be consulted. He emphasizes especially that as a rule bacteria ferment carboliydrates in preference to attacking proteins when both foodstuffs are available. " See Buxton (American Med., 1903 (0), 137) concerning enzymes of numerous Dfl.ctiGri&i "> Cent. f. Bakt.. 1903 (35), I. " Contradicted by DcWacle, Cent. f. Bakt., 1905 (39), 353. " llofmeister's Beitr., 1907 (10), 287. "Jour, infect. Dis., 1912 (11), 388; literature on bacterial lipases. See also Kendall." '" See Jacoby, Biochem. Zeit., 1917 (80), 357. 80 Corper and Sweany, Jour, liact., 1918 (3), 129. «' Wratschel)naja Cazetta, 1902, p. 52. «2.Jour. Med. Re.seareli, 1903 (10), 42. «Mour. Infect. Dis., 1919 (21), 347. «♦ Biochem. Zeit., 1917 (83), 74. BACTERIAL ENZYMES 111 requirements for the production of urease on Uschinski's medium, and finds that while bacteria will grow if the sodium asparaginate is present, no urease is formed unless leucine is also added. There does not seem to be any important relation between enzyme production and pathogenicity.*^ In general, bacterial proteolytic enzymes resemble trypsin more closely than they do pepsin, acting best in an alkaline medium; but the enzymes extracted from bacterial cultures are very feeble as com- pared with pancreatic trypsin. It is probable that there are several distinct proteolytic enzymes in bacterial cells, gelatinase being a dis- tinct protease (Jordan). ^"^ Abbott and Gildersleeve found that the gelatin-dissolving enzyme of bacteria resists a temperature of 100° C. for as long as fifteen to thirty minutes, but Jordan found that the reaction of the medium modifies greatly this heat resistance. Schmailo- witsch*^ states that some bacteria produce an enzyme acting in acid medium upon gelatin but not upon albumin, and this enzyme carries the digestion only as far as the gelatin-peptone stage, whereas the enzymes acting in an alkaline medium carry the splitting through to leucine, tyrosine, etc. Kendall and Walker*^ state that the proteolytic enzymes of B. proteus are not formed when the bacteria have enough carbo- hj^drate supplied so that they need not depend on proteins for their energy requirements; deaminization is independent of proteolysis and represents intracellular enzjTne action. Plenge^^ suggests that there is a special enzyme digesting nucleoproteins (nuclease). Bac- teria are able to split nucleic acids and to convert amino-purines into oxypurines, but they do not carry the oxidation to uric acid; putre- factive bacteria can slowdy destroy uric acid (Schittenhelm),^'' and B. coll destroys purines. ^^ Cacace^- investigated the cleavage products of gelatin and coagu- lated blood when digested by B. anthracis, Staph, pyogenes aureus, and Sarcina aurantiaca, and found that proteoses and peptone are produced, which disappear in the later stages of digestion. Rettger^^ found leucine, tyrosine, tryptophane, as well as phenols, skatole, indole, aromatic oxy-acids, and mercaptan, among the products of bac- terial decomposition of egg-albumen and meat; proteoses and pep- tones appear in the early stages, but later disappear, as also eventually do the leucine, tj-rosine, etc. Choline has also been found in the 85 Rosenthal and Patai, Cent. f. Bakt., 1914 (73), 406; (74), 369. 8« Corroborated by Bertiau, Cent. f. Bakt., 1914 (74), 374. " Abst. in Biochem. Centr., 1903 (1), 230; see also DeWaele, Cent. f. Bakt., 1905 (39), 353. 8* Jour. Infect. Dis., 1915 (17), 442. See also Berman and Rettger, Jour. Bact., 1918 (3), 367. 89 Zeit. f. physiol. Chem., 1903 (39), 190. 90 Zeit. physiol. Chem., 1908 (57), 21. »' Siven, Zeit. phvsiol. Chem., 1914 (91), 336. 92 Cent. f. Bakt., 1901 (30), 244. 93 Amer. Jour, of Physiol., 1903 (8). 284 112 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS products of autolysis.^* MoUiard^^ reports that Isaria densa produces large masses of crystals of glycine, even when grown on proteins that contain little or no glycine. The digestive power of the filtrates of cultures and of killed bacteria is far less than that of the living bacteria (Knapp).^^ Streptococci digest proteins of exudates feebly, staphjdococci more rapidl}-, and colon bacilli are still more active. He could find no relation between the proteolytic power of the bacteria and the severity of the infection from which they came. Staphylococci can cause coagulation of plasma and then dissolve the coagulum, showing the presence of two enzymes, staphylokinase and fibrinolysi7i (KMnschmidt) .^"^ Sperry and Rettger,^^ however, found that even the most actively putrefactive bacteria are unable to attack or grow upon carefullj^ purified proteins, although the presence of small amounts of amino acids or other available nu- trient makes the proteins available to the bacteria; apparently they must have some nutrient more available than intact protein molecules to enable them to grow sufficiently to produce enough free enzymes to attack the proteins. By virtue of their proteolytic enzymes, filtrates of bacteria that liquefy gelatin also can digest hardened liver, kidney and other tissue elements in vitro the changes resembling those of necrobiosis.^^ Oxidizing Enzymes. — Catalase is demonstrable in bacteria, the anaerobic forms showing the least activity (Rywosch),' but practically no species is entirely inactive (.Joi'nsj ;' it may exist as either endo- or ecto-enzyme. B. pi-oteus synthe- sizes catalase even when grown on a simple sj^nthetic medium containing, besides inorganic salts, sodium lactate and alanine or aspartic acid (Jacob^-).^ Certain bacteria and actinomyces exhibit oxidative effects, resembling tyrosi7iase, but such an enzyme could not be extracted by Lehmann and Sano.'' Tsudji,* however, not only observed oxidation of tyrosine, but states furthermore that proteus pro- duces always a d-oxyacid product and subtilis a 1-oxyacid type, regardless of whether they have oxidized d-, 1-, or dl-tyrosine. Immunity against bacterial enzymes may be secured as it is against other enzymes. Abbott and Gildcrsleeve^- found that by injections into animals of proteolytic bacterial filtrates which were only slightly toxic, the serum of the animals acquired a slight but specific increase in resistance to the proteolytic enzymes of the filtrates.^ Normal serum contains a certain amount of enzyme-resisting substance. »^ Kutscher and Lohmann, Zcit. phvsiol. Chcm., 1903 (39), 313. »''Compt. Rend. Acad. Sci., 1918 (167), 7S(i "oZeit. f. Ileilk. (Chir. Abt.), 1902 (23), 230. 0' Zcit. Immunitiit., 1909 (3), 510. •■"•.four. Biol. Chem., 1915 (20), 445; Jour. Hact., 1910 (1), 15. »'■' Bittrolff, Zicgler's Beitr., 1915 (00), 337. • Cent. f. Bakt., 1907 (44), 295. 2 Arcli. f. Ilvg., 190.S (07), 134. =• Biocliem. Zcit., 191S (SS), 35 and (89), 350. < Arcli. f. llyg., 190S (()7), 99. ' Acta Schoiac Med. Tniv. Kioto, 191S (2), 115. • Antigehitiiiase has also been obtained by Bertiau, Cent. f. Bakt., 1914 (74) 374. AUTOLYSIS OF BACTERIA 113 Other observers have found that immunization against Hving or dead bacteria leads to the production of substances antagonistic to their enzymes, but the degree of resistance acquired is never great, v. Dun- gem" found that the serum of animals infected with various bacteria prevented digestion of gelatin by the enzymes obtained from cultures of the same species of bacteria. He applied this fact to the diagnosis of infectious conditions, finding that the serum of a patient with osteomyelitis was over twenty times as strongly inhibitory to staphy- lococcus enzymes as was serum of normal persons. The reaction is specific, cholera vibrio enzymes not being inhibited to any correspond- ing degree. Kantorowicz^ and de Waele^ state that bacteria contain an intra- cellular anti-protease which, with most bacteria, holds in check the proteolytic action; only with the liquefying bacteria are the proteases in excess. Bacteria grow well in strong solutions of enzymes, and with- out destroying the enzymes (Fermi).'" After Gram-negative bacteria have been heated to 80° they are readily digested by trypsin, pepsin or leucocytic proteases; but Gram-positive bacteria are resistant even after heating. This is ascribed by Jobhng and Petersen" to the un- saturated fatty acids, which are present in greater amounts in Gram- positive bacteria. Autolysis of Bacteria. — Autolysis occurs also in bacteria, their pro- teolytic enzymes digesting the cell substance whenever the organisms are killed by agents (chloroform, toluene, etc.) that do not destroj' these enzjaiies, and which, being fat solvents, may facilitate digestion by removing the inhibitory lipoids. Even the absence of food leads to autolysis, presumably because the normally existing autolj'tic processes are not counteracted by synthesis of new protein material; hence, autolysis occurs when bacteria are placed in salt solution or distilled water. Although it had been known for many years that yeast cells digest one another when there is nothing else for them to live upon, the first definite study of bacterial autolysis seems to have been made by Levy and Pfersdorff'^ ^nd Conradi.'^ The former digested anthrax bacilli (in whose bodies are contained rennin, lipase and protease) under toluene for several weeks and obtained a slightly toxic product. Conradi permitted dj'sentery bacilli and typhoid bacilli to digest themselves in normal salt solution for twenty-four to forty-eight hours at 37° C., and obtained in this way the soluble, highly poisonous endotoxins of the bacteria, which are liberated by the de- struction of the bacterial structure by the autolytic enzymes. Longer ^ Miinch. med. Woch., 1898 (45), 10-10. 8 Miinch. med. Woch., 1909 (56), 897. 9 Cent. f. Bakt., 1909 (50), 40. 1" Arch. FarmacoL, 1909 (8), 481. "Jour. Exp. Med., 1914 (20), 321. 1- Deut. med. Woch., 1902 (28), 879. 13 Ibid., 1903 (29), 26. 114 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS autolysis results in the destruction by the enzymes of the endotoxins themselves. Rettger^^ found among the autolj^ic products of bac- teria, leucine, tyrosine, basic substances, and phosporic acid. Under favorable conditions complete autolysis can occur in Iwo to ten days. Brieger and Mayer'^ found that at room temperature (15° C.) practically no autolysis occurs with typhoid bacilli in distilled water, and the soluble products thus obtained are quite non-toxic, although if injected into animals they give rise to the production of agglu- tinins and bacteriolysins. Bertarelli'® has used the products of au- tolysis of cholera vibrios successfully in the production of immunitj^ and states that the products of autolysis consist largely of nucleins. It is probable that in every culture bacteria are constantly being destroyed, either by their own enzymes or by the proteolytic enzymes of the other bacteria. Some bacteria are much more rapidly auto- lyzed than others, cholera vibrios, colon, typhoid, and dj^sentery bacilli being rapidly digested, while streptococci, staphylococci and tubercle bacilli are very little and slowly autolyzed. In general, the Gram-positive organisms resist autolysis longest, but pneumococci autolyze readil}'. Conradi,'^ who has shown that certain products of autolysis of tis- sues are bactericidal, believes that also in cultures powcrfulh' bacteri- cidal substances are produced through autoh'sis of the bacteria. This he thinks, accounts for the decrease in numbers of living bacteria that always sets in after a short period of growth on artificial media; but there is much doubt as to these substances being of any considerable importance in the body.^^ It has been found by Turro'^ that ex- tracts from various tissues containing autolytic enzymes can digest bacterial cells. ^^ It is very possible that the endotoxins contained within such pathogenic bacteria as typhoid and cholera are liberated through digestion of the bacteria, either by autolysis or by the en- zymes of the leucocytes and tissues of the organism that they have infected. These, and a number of other bacteria, produce no soluble toxins that diffuse from the cells as do diphtheria and tetanus toxin, and it is difficult to explain liie toxic effects these bacteria produce without assuming that their iiitrnccllular toxins are liberated in some such way. It is also cjuite probable that th(> enzymes found in fil- trates from bacterial cultures arc liberated from tlu> liacterial cells 1^ Jour. Mod. Research, 1904 (13), 79. " Deut. med. Woch., 1904 (30), 980. "Cent. f. Hiikt., l'.)()r> (3S), 5SI. " Miinch. mod. Woohon.sclir., 1905 (ry2) 1701. '8Soo iMJkinnii, ('out. f. Hiikt., 190() (41), 3(17; I'assini, AVion. klin. Woch., 190<) (19), (127. '" Cent. f. liukt., 1902 (32), 10."). *" Si^;wurl, (Arl). ii. d. i'liMi. Inst. Tiil)inK<'n, 1902 (3), 277) found tl.at trypsin and poi),siri (witliout acid) do not injuro liviufi; antluax bacilli. POISONOUS BACTERIAL PRODUCTS Ho only when these have been uutolyzed.-' With the possibh; exception just mentioued, there is httle evidence that the bacterial enzymes play any important role in infectious diseases. They may be a slight factor in the digestion of tissue and exudates in suppuration, but as compared with the leucocytic enzymes their influence is probably mi- nute; beyond this they have no apparent influence upon their host, and are chiefly concerned in the metabolism of the bacteria. The proteoses and peptones produced by bacterial action and isolated from cultures do not seem to be any more toxic than those produced by pepsin and trypsin, but violent poisons may be liberated from bacteria during autolysis, as Rosenow^^ has shown for the pneumo- coccus and other bacteria; these poisons seem similar to or identical with the so-called anaphjjlatoxin which is supposedly formed by the digestion of bacteria with serum complement, and presumably they are proteoses or polypeptids, but their exact nature is not known. (See Anaphylatoxin, Chap, vii.) POISONOUS BACTERIAL PRODUCTS Almost without exception all the harm that bacteria do is brought about by means of the chemical substances produced in one vray or another by their metabolic processes. Animal parasites may do harm mechanically, but with the possible exception of the effects of capil- lary emboli (especially with anthrax), bacteria produce all their ef- fects through chemical means. The poisonous chemical substances produced by bacteria are commonly grouped into four classes: I. Products of the decomposition of the media upon which the bacteria are growing; among these the best known are the ptomams. II. Soluble poisons manufactured by the bacteria, and secreted from the cell into its surrounding media — the true toxins. III. Poisons manufactured by the bacteria which do not escape from the normal cell but which are as specific in their poisonous prop- erties as the true toxins; because of their intracellular situation they are called endotoxins. IV. Poisonous protein constitutents of the bacterial cell which form part of the cell protoplasm, but which are not soluble, and the poisonous effects of which are not specific and not usually responsible for the disease; these are called bacterial proteins. 21 Emmerich and Loew (Zeitschr. f. Hyp., 1899 (31), 1), having found that pyocyanase is capable of destroying and digesting other bacteria than pyocy- aneus, suggested that it might be a potent factor in producing artificial immu- nity. Their rather remarkable hypotheses have been much contested, and are of questionable value. (See Petrie, Jour, of Pathol, and Bacteriol., 1903 (8), 200; also, Rettger, Jour. Infectious Diseases, 1905 (2), 5G2; Emmerich, Miinch. med. Woch., 1907 (54), 2217). " Jour. Infect. Dis., 1912 (10), 113; (11), 94, 235 and 480. 116 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS Ptomains Ptomains, the soluble basic nitrogenous substances that are found in the medium in which bacteria have been growing, were the first bacterial products that were recognized, and for some time it was believed that it was through the production of such alkaloid-like sub- stances that bacteria caused disease, just as poisonous plants owe their effects to poisonous alkaloids. It was soon found, however, that the ptomains that could be isolated from cultures of pathogenic bac- teria were insufficient by themselves to cause the poisonous effects that such cultures produced when injected into animals. The isolated ptomains were not onlj^ far less poisonous than the original culture, but furthermore they did not produce the symptoms and anatomical changes characteristic of the diseases that the pathogenic organism caused. Moreover, the majority of ptomains are not very poisonous, and highly poisonous ptomains may be produced by non-pathogenic bacteria. As a result, the work on ptomains, which once occupied many laboratories and promised to reveal the entire chemistrj^ of bac- terial intoxication, has now been almost completely dropped. The interest in ptomains is by no means entirely historical, however, for it is possible that poisonous ptomains at times do enter the body and cause illness, perhaps even death. The close chemical resemblance to vegetable alkaloids of some of the ptomains that may arise in decom- posing corpses, makes them of great importance to chemists searching for the cause of death in cases of supposed poisoning. Therefore the most essential features of the ptomains and their chief known rela- tions to intoxications will be briefly discussed, referring the reader for a full consideration to Vaughan and Novy's "Cellular Toxins" and Barger's "The Simpler Natural Bases." The ptomains owe their basic character to nitrogen-containing radicals, principally amino-groups, and hence arc formed from ni- trogenous substances, chiefly proteins, which contain their nitrogen in the amino form. Probably most ptomains arise from the decompo- sition of the protein medium upon which the bacteria grow, although undoubtedly part of the ptomains is also formed from the destruc- tion of the bacterial cells themselves; how large a i)art of the pto- mains is formed by intracellular bacterial processes and how much by cleavage of the proteins of the media by extracellular bacterial enzymes is unknown. The structure of the ptomains shows them to be V(!ry closely related to the amino-acids obtained ])y cleavage of the prot(;in molecule by enzymes antl other hj'drolytic agencies; and the determination of the composition of the several amino-acids of the proteins lias quite cleared up the problem of the origin of the pto- mains. Presumably these secondary changes result from the action of special enzynu^s upon th(> aniiiio-acids. Most of the jitomains are fre(! from or ]nn)v in oxygen, lieiicc^ rcihiction processes, or lack of sufli- ciciit oxygen foi- oxidation, are prob;il>ly ini|)ortaiit in t heir proihiction. PTOMAlNS 117 The poisonous ptomains, which are decidedly in the minority among the entire group, are themselves subject to decomposition, being most abundant in the cultures after a certain period of time, and then decreasing in amount. Very old cultures show almost none of the higher molecular forms of nitrogen, such as ptomains, these substances having been changed into anmionium and nitrate compounds. In sharp contradistinction to the toxins, the yiomalns are by no means specific. No matter upon what medium diphtheria bacilli grow, the toxin produced has qualitatively the same properties, whereas the nature of the ptomains depends not only upon the nature of the bac- teria producing them, but also even more upon the sort of soil upon which the bacteria are grown, the temperature, the duration of the process, and the quantity of oxygen furnished. The same organism may produce totally different ptomains when grown on different media or under different conditions. Another essential difference is that we cannot obtain an immune serum, antagonizing the action of ptomains, by injecting ptomains into animals. If ptomains do cause intoxications presumably it is when they are taken in with food in which they have been produced b}- bacterial de- composition. Besides this food poisoning, it is also possible that pto- mains may be formed by putrefaction within the gastrointestinal tract. Another possible source of ptomains is furnished by decom- posing tissues in gangrene. It is doubtful if ptomains are produced in sufF.cient quantities by pathogenic bacteria infecting living tissue to be of any importance. Food poisoning is by no means uncommon, but we do not know how often it is due to ptomains; it may be the result of poisonous materials contained abnormally in the food, that are not ptomains, e. g., botuhsm; or it may be due to an infection of the animal from which the meat came with pathogenic organisms, particularly the B. enteritidis of Gaertner and other bacteria related to the colon-t3'phoid group; or in other ways food ordinarily wholesome may become poisonous.-^ The commonest sources of ptomain poison- ing are supposed to be imperfectly preserved canned meats, sausages, decomposing fish, cheese, ice-cream, and milk.-^^ Chemical Composition of Ptomains. — To indicate the composition and nature of ptomains a few of the more important ones will be described. As illustrative of the simpler forms may be mentioned: Methvl amine, CHs-NH.. Di-methyl amine, CH3 - XH - CH3. Tri-methyl amine, CH3 - N - CH3. I CH3. These bodies, which are commonly found in decomposing proteins are but very slightly toxic, and of little pathological importance. 23 See Jordan, "Food Poisoning," University of Chicago Press, 1917. -•' All these matters are discussed at length by Vaughan and Xovy, to whose book the reader is referred. 118 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS The source of the ptomains in the various amino-acids is usuall}' easily traced through their chemical structure, and Ackermann and Kutscher" have classified them in this relation under the name " aporrhegma." When we examine the structural formulae of some of the larger ptomaln mole- cules and compare them with the formula" of the amino-acids that form the protein molecule, the relation is apparent, e. g., compare iso-amylamine with leucine. CHav CH3\ ^CH-CHj-CHo-NH, ^CH-CHo-CH-XH, CH/ . CH3 aeucine) ^COOH. (iso-amylamine) Putrescine, C4H12N2, structural formula, NH2-CH,-CH2-CH2-CH2-XH2, and cadaverine, C5H14N2, structural formula, XH2 - CH2 - CH2 - CH2 - CH2 - CH2 - XHo, are of interest because they have been found in the intestinal contents, arising from putrefaction of proteins, and also are sometimes present in the urine in cystinuriaJ^ They are closely related to the diamino-acids, lysine and ornithine. Thej' are but slightly toxic, although capable of causing local necrosis Avhen injected sub- cutaneously. (See further discu.ssion on these and the Pressor Bases in Chap, xxi.) The Choline Group. — Another group of ptomains, including cho- line and closely related substances, is also of interest. These ptomains are: Chohne, CH2OH— CH2— X(CH3)3— OH Xeurine, CH2=CH— X(CH3)3— OH Muscarine, CH(0H)2— CH>— X (CH3).-r-0H Betaine, COOH— CH,— X(CH3)3— OH The first point of importance is that choline is present in every cell normally, forming the nitrogenous portion of the lecithin mole- cule. Its source in putrefaction of tissues is, therefore, plain. It is possible that choline is liberated from nerve tissues when they break down in the body during hfe," and there is a considerable literature on the supposed finding of choline in the blood and cerebrospinal fluid in diseases of the central nervous system and experimental lesions in nervous tissues. At present it seems probable that these observations depend upon faulty methods of analj'sis, and it is ex- tremely doubtful if enough choline is ever set free at one time from even severe acute nervous lesions to be detected in the body fluids by chemical means. ^^ Hunt^^ has devised a physiological test that per- mits of the detection of as little as 0.00001 mg.. but he was unable to " Zeit. physiol. Chem., 1910 (69), 2G5. 2«Udrdnsky and Baumann, Zeit. physiol. Choiii., 1889 (13), 562; 1889 (15) 77. " Coriat (Amer. Jour, of Physiol., 1904 (12), 353) has studied the conditions under which choline may be produced from lecithin. Putrefaction of lecithin or lecithin-ridi tissues liberates choline as also does aut()ly.-;is of brain tissue; neither pcjisin nor trypsin, however, splits it from the lecitliin. In brain tissue, therefore;, there seems to be an enzyme different from trypsin, which splits choline out of tlie lecithin molecule. -"Sec Wehstcr, liioclicm. .lour., 1900 (4), 123 ; Kajiura. (.luart. Jour. Exper. Physiol.. 190.S (1), 291; llandel.sniuiin, Deu). Zeit. Nervenheilk., 1908 (35), 428; Dori'c and (Jolla, Biochem. Jour., 1910 (5), 306. -"Jour. I'liurmacol., 1915 (7), 301. PTOMA'iNS 119 obtain evidence that choline is of any sip;nificancc in either physiologi- cal or patholofj;i('al processes. Normally the largest amounts by far are obtained from the adrenals, which also seem to contain choline deriva- tives of much greater physiological activity. Choline itself is some- what toxic, but the closely related body, neurine, into which it may be transformed, is highly j)oisonous, which makes chf)line an important indirect source of intoxication. It is possible, for example, that lecithin taken in the food splits off choline in the gastro-intestinal tract, and this being converted into neurine gives rise to intoxication which may be ascribed to food intoxication. Likewise it has been suggested that the intoxication of fatigue may be due, at least in part, to choline and neurine produced from lecithin decomposed during the period of cellular activity. The close structural relation to choline and neurine, of the mushroom poison, muscarine, which produces physiological effects very similar to those of neurine, indicates the close relationship of the putrefactive ptomains and the vegetable alkaloids. Indeed a muscarine apparently identical with that of the mushroom has been found in decomposing flesh, and neurine, presum- abl}^ derived from lecithin, may be found in human urine. ^" Betaine, the fourth member of the group, which has but slight toxicity, is particularly well known as a constituent of plant tissues. Both neurine and muscarine are extremely poisonous and quite similar in their effects. Subcutaneous injection of but 1 to 3 mg. of muscarine in man produces salivation, rapid pulse, reddening of the face, weakness, depression, profuse sweating, vomiting, and di- arrhoea. Neurine, likewise, causes salivation, lachrymation, vomiting, and diarrhoea. In fatal poisoning respiration ceases before the heart stops. Both poisons resemble physostigmine in their stimulation of secretion and are equally well counteracted by atropine. The toxicity of these substances is so great that not a large amount would need to be formed by oxidation of choline to produce severe symptoms, although it is not known that this actually occurs in the body. When introduced by mouth, the lethal dose of neurine is ten times as great as when injected subcutaneously, indicating that chemical changes in the gastro-intestinal tract or liver offer some protection against in- toxication by these substances when taken in tainted food. Choline, although by no means so poisonous as neurine, has a similar action when administered in sufficiently large doses. According to Brieger, it is about one-tenth to one-twentieth as toxic as neurine. ^^ Choline '" Kutscher and Lohmann, Zeit. physiol. Chem., 1906 (48), 1. '' Halliburton, "Chem. of Muscle and Nerve," 1904, p. 119, states that choline produces a fall in blood pressure by dilating the peripheral vessels, whereas neur- ine constricts the peripheral vessels; he uses this difference in phj'siological ef- fect as a means of distinguishing the two substances. Injected into animals, choline causes a considerable but transient decrease in the number of leucocj'tes in the blood, followed later by an increase (Werner and Lichtenberg, Deut. med. Woch., 1906 (32), 22). 120 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS seems to be rapidlj^ destroyed in the body, not appearing in the urine^^ but forming formic acid and perhaps glyoxyhc acid. Donath^^ found that choline injected directly into the cortex or under the dura is extremely toxic, causing severe tonic and clonic convulsions, and believes that choline may be responsible for epileptic convulsions. This view has been opposed, and properly so, by Handelsmann-^ and others. The attempt to ascribe importance to choline as a cause of either toxic or therapeutic effect of x-rays seems also to be entitled to but slight consideration.^'* It is probably a factor in the lowering of blood pressure which results from injection of extracts of various tis- sues, in which it is commonly present in minute amounts,^' for very minute amounts of choline will produce a decided fall in blood pres- sure.^^ The Pressor Bases. — By decarboxylation of amino acids, amines are obtained, and some of them, notablj^ those derived from leucine, tyrosine, phem-lalanine and histidine, have a marked effect on non-striated muscle. These are discussed in Chapter xxi. Toxins Certain bacteria produce soluble poisons by sj^nthetic processes, which poisons are secreted into the surrounding medium and repre- sent the chief poisonous products of the bacteria, being capable of causing most or all of the symptoms attributed to infection by the specific bacteria that have manufactured them. To this class of solu- ble poisons the term toxin has now become limited (for reasons that will be mentioned below), including not only toxins of bacterial ori- gin, but also poisons of similar nature produced by animals (snake venoms, eel serum, etc.) and by plants (ricin, abrin, crotin). The chief bacteria secreting true toxins are B. diphtherice, B. tetani, B. pyoajaneus, and B. botulinus. Dysentery bacilli, the anaerobes of gas gangrene, and perhaps a few other pathogens also secrete a toxin. Pick considers the active constituent of tuberculin to be a true toxin, or closely related thereto. Also the hemoljiic poisons produced by many bacteria seem to be true toxins. It will be seen that the term toxin has been greatly narrowed since the time when all ptomains and other poisonous bacterial products were called toxins, until now it has come to include the specific poisons of but a few of the great group of pathogenic bacteria. Chemical Properties of Toxins. — The chemical nature of the toxins is entirely unknown. By various precipitation methods they 32 V. Iloosslin, Hofmeister's Beitr., 1906 (S), 271. ^•' Zeit. f. physiol. Chciii., 190:5 (159), 52(5; also see Med. News, 1905 (86), 107, for literature and incthods of analysis. ^«Sce Schenk, Deut. med. Wocli., 1910 CMt), 1130. " Schwarz and Ledercr, I'lliigcr's Arch., 1908 (124), 353; Kinoshita, ibid., 1910 (132), 607. •'• Mendel cl al. Jour. I'harin. and Exj). Ther., 1912 (3), 648; Hunt and Taveau. liulletin 73, llyg. Lab. U. H. V. U. Service. TOXINS 121 may be carried down, but incliukMl with them arc masses of impurities, chiefly proteins. They behave hke electro-positive colloids," but diffuse faster than proteins. It is not certain that toxins are not proteins, for although certain investigators report that by purification processes very active toxins have been obtained that did not give the protein reactions, yet the toxins are attacked by proteolytic en- zymes, and, like proteins, are precipitated by nucleic acid (Kossel). Furthermore, accumulating experience with immunological processes adds increasing doubt as to the possibility of antibody formation being incited by anything but proteins. Oppenheimer says of the toxins, "we must be contented to assume that they are large mo- lecular complexes, probably related to the proteins, corresponding to them in certain properties, but standing even nearer to the equally mysterious enzymes with whose properties they show the most ex- tended analogies both in their reactions and in their activities." These similarities between toxins and enzymes are very striking, and in discussing the nature of the enzymes we have mentioned the reasons for considering them related to the toxins; we may now^ take up the other side of the question and consider the relation of the toxins to the enzymes. Resemblance to Enzymes. — First of all w^e meet the same diffi- culty in isolating toxins that we do in isolating enzymes. "A pure toxin is as unknown as a pure enzyme" (Oppenheimer). At first both were believed to be proteins; now both are considered by many not to be proteins, but molecular complexes of nearly equally great dimensions. That toxins, like enzymes, are colloids, has been abun- dantly demonstrated.^* Both pass through porcelain filters, but both lose much of their strength in the process, and they are almost en- tirely held back by dialyzing membranes. They behave similarly as regards adsorption by suspensions,^^ and have similar effects on the physical properties of their solutions (Zunz)."*" Neither will with- stand boiling, and most forms are destroyed at 80° instantly or in a very short time; on the whole, however, toxins are more susceptible to heat, as well as to most other injurious agencies. Both stand dry heat over 100°, and extremely low temperature, without much injury. Left standing in solution for some time they gradually lose their specific properties, and in each case this seems to be due to an altera- tion in the portion of the molecule that produces the destructive effects {toxophore or zymophore group), while the portion of the mole- cule that unites with the substance that is to be attacked (haptophore group) remains uninjured, the toxin becoming a toxoid, the enzyme " Field and Teague, Jour. Exper. Med., 1907 (9), 86. '8 See Zangger, Cent. f. Bakt. (ref.), 1905 (36), 239. '^ By fiocculation of the colloids bearing adsorbed toxins it may be possible to secure them in comparatively pure condition (London, Compt. Rend. Soc. Biol., 1917 (80), 756. "Arch, di Fisiol., 1909 (7), 137. 122 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS a fermentoid. Enzymes as well as toxins are poisonous when injected into animals, and the animals react to each by producing substances (antibodies) that render each inert, probably in the same way. On the other hand, enzymes and toxins seem to produce their effects ac- cording to different laws: — A small amount of enzyme can in course of time produce an almost indefinite amount of effect, whereas toxins act more nearly quantitatively. It seems as if the enzyme were bound to the body upon which it acts, as is the toxin, but that after it has destroyed this body it is set free in a still active form, ready to accom- plish further work, whereas the toxin is either not set free, or it be- comes inactive after it has once bepn combined. Agencies Destroying or Modifying Toxins. — Toxins are very sus- ceptible to light, direct sunlight soon destroying the power of toxin solutions. Fluorescent substances destroy toxins both in vitro and in the body.'*^ They are generally destroyed by moist heat of 80°, but resist 100° when dry. Oxygen, even dilute as in air, is harmful; and all oxidizing agents, including oxidizing enzymes, destroy them quicld3^■*- Like enzymes, they withstand such antiseptics as chloroform, tol- uene, etc., and are precipitated by the heavy metals. Some agencies seem to attack only the toxophore portion of the molecule, e. g., iodin, carbon disulphid (Ehrlich). Certain toxins (diphtheria, dysentery) can be converted into non-toxic modifications by acids, the original toxicity being restored by bases (Docrr),''^ which fact, Pick maintains, is in support of the protein nature of toxins. Salts of monovalent metals have no effect on toxins, but bivalent and trivalent salts are injurious to them, tetanus toxin being more sensitive than diphtheria toxin. X-rays are said to weaken them.^'* 'Introduced into the gastro-intestinal tract, most bacterial toxins are not absorbed (botulinus toxin excepted), cause no symptoms, and do not reappear in the feces; they are therefore destroyed by the contents of the tract, pepsin, pancreatic juice, and bile all being capa- ble of destroying toxins. ^^ They maj% however, when injected sub- cutaneously, circulate unimpaired in the blood of non-susceptible animals, gradually disappearing, more through slow processes of de- struction than by elimination. When injected into susceptible animals, they soon disappear from the blood, being fixed in the organs that they attack. Toxins are also bound ])y lipoids, fats and similar substances, which accounts, at least in part, for the affinity of tetanus ^' Literature Kiven by Not^uchi, Jour. Kxppr. Med., 1900 (S), 263. ** According!; to Pitini (liiocheni. Zeit., 11)10 (2.')), 257) toxins cause their harm- ful effcot.s 1)V reihiciiiK tlie oxidizing capacity of the tissues. <' Wien. klin. Wocli., 1907 (20), T). <' (lerlKirtz, Herl. klin. Wocli., 1909 (40), 1800. •"' IJuldwin and Levone (.lour. Mi-d. lie.soarch, 1901 (()), 120) found that diph- tlioria ami tetanus toxin are l)otli destroyed, ai)parently througli digestion, by pci)sin, trypsin and jiajiain acting for several days. Ueview of Literature by Lust, Ifofnieister's lieitr., 1901 (ii), 132. See Vincent, Ann. Inst. Pasteur, 1908 (22), 341. TOXINS 123 toxin for nervous tissues.""' In comnion witli other colloids they are adsorbed by surfaces, such as charcoal, kaolin, etc.; such ad.sorption is accompanied by little change in any of the physical properties of the solution, except an increase in surface tension (Zunz). Differences from Ptomains. — While ptomains are formed by cleavage processes from the medium upon which the bacteria grow, and the same ptomains can be produced by several different kinds of bacteria, the toxins are synthetic 'products of absolutely specific iiaturc. However, the toxins seem to be produced little if at all by growing the bacteria on Uschinsky's or similar media, which contain no proteins, carbohydrates, or fats, but merely simple organic and inorganic salts of known composition.^' Nevertheless diphtheria toxin is essentially the same no matter on what sort of medium the l)acteria are grown, whereas ptomains vary with the nature of the siilistance from wliich they are produced. Toxins arc true Sgcretions of bacterial cells, just as trypsin is of pancreat'c c.pIIs, or tli,Y.roiodin of thyroid cells. SvEi-bodies caiTTKf produced agamsttoxins, but not against ptomains. — ' ^ " ~ Ehrlich's Conception of the Nature of Toxins. — Chemical studies of toxins being impossible, we have been obliged to study them through their physiological effects, just as we have obtained informa- tion concerning enzymes through their specific actions. In this way Ehrlich developed well-crj'stallized ideas concerning the structure of toxins, as well as the manner in which they act, which may be briefly summarized as follows: Each toxin molecule consists of a large num- ber of organic complexes, grouped, as in other organic compounds, as side-chains about a central chain or radical. One or more of these complexes has a chemical affinity for certain chemical constituents of the tissues of susceptible animals, with which the toxin molecule unites; this binding group is called the haptophore (meaning "bear- ing a bond"). Another side-chain or group of side-chains exerts the injurious effects upon the tissue to which the molecule has been bound by the haptophore, and cannot produce these injurious effects unless it has been so bound. This injury-working group is called the toxo- pJiore. An animal is susceptible to a toxin only when its cells con- tain substances which possess a chemical aflB.nity for the haptophore groups of the toxin, and also substances which can be harmfully influ- enced b}^ the toxophore groups. Tetanus toxin, for example, owes its effects to the fact that nervous tissues contain chemical substances having a strong affinit}' for the haptophore group of tetanus toxin, and also substances that can be attacked with serious results by the toxophore group of the toxin. The nature of the changes brought about by the toxophore groups of toxins is not understood; there are many resemblances to the action of enzymes, but the analogy is by no means complete. We find perhaps the closest analogy to the en- zymes in the toxic substances that destroy- red corpuscles and bacteria (hemolysins, bacteriolysins) , which will be considered in another place. The immunity against toxins and enzymes seems to be produced by ^«Loewe, Biochem. Zeit., 1911 (33), 225, and (34), 495. "See Rettgerand Robinson, Jour. Med. Res., 1917 (38), 357. 124 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS identical processes, which consist in an overproduction of the cellular constituents (receptors) which bind the haptophore groups to the cells, these excessive receptors being secreted into the blood, where they combine with the toxin or enzyme so that it cannot enter into combi- nation" with the cells. This "side chain theory" of Ehrhch has been a useful working hypothesis, although it is becoming highly probable that it does not picture the exact method of toxin and antitoxin action. ^^ Immune substances cannot be produced against ptoma'ins, or for that matter against the vegetable alkaloids, or against any chemical bodies of known constitution. Another difference between the action of toxins and simpler chemical poisons is, that while with the latter the effects are produced in a very short time after injection, there is a latent period of several hours before symptoms appear after injecting toxins. What occurs during this latent period is not fully known, but that there is a latent period suggests a resemblance to enzyme action. An alkaloidal or other chemical poison enters the cell, and its harm is done at once. A toxin combines with the cell, and then, if it produces its effects by an enz3'matic alteration of the cellular struc- ture, some time must elapse before the changes are great enough to cause the appearance of symptoms. Endotoxins" By 'far the greater number of pathogenic bacteria do not secrete their poisons as toxins into the surrounding medium, aUhough they manifestly cause disease by poisoning their host. Among them are such organisms as the typhoid bacilhis. pneumococcus, the pus cocci, cholera vibrios, and many others. If cultures of these organisms are filtered, the filtrate will be found to be but slightly toxic (except for the hemolytic i)oisons). although the bodies of the bacteria after they have been killed by chloroform or other antiseptics are highly poisonous if injected into an animal. These bacteria, then, produce poisons which do not escape from the cells into the culture-medium, but are firndy held within them. By using various means these intracellular toxins, or endotoxins, can be obtained independent of the bacterial cells. One of these is to grind up the cells, which can be particu- larly well done if they are first made brittle by freezing at the temperature of liquid air (MacFadyen's method). By very great ])ressure in the Buchner press the cellular contents can he expressed. They may also be obtainctl by letting the bacteria autolyze themselves for a short time in non-nutrient fluids (Conradi,^" et ai). Endotoxins obtained in this way are soluble and highly poisonous, and it is undoubtedly through their action that the characteristic diseases are produced by the bacteria that contain them. Presumably the endotoxins are liberated in the body either by autolysis, or by heterolysis by the enzymes of the body cells and fluids, and there is some (luestion as to whether they are preformed specific con- stituents of the bacteria, or merely the jjoisonous ])roduct of enzymatic disintegra- tion of the ba(!terial i)roteins, similar to the "anaphylatoxins."^' Endotoxins differ from the true toxins in one imj)ortant respect: namely, it is difficuU or impossible to obtain an antitoxin for endotoxins by immunization of <« See Coca, .lour. Infect. Dis., 1915 (17), ;J51. *• See general r(>view by I'feifTer, Jahresber. d. Immunitatsforsch., 1910 (ti), 13. "0 Deut. med. Woch., \\)0A (29), 2ti. "See Dold and llanau, Zeit. Immunitat., 1913 (19), 31; Zinsser, "Infection and Resistance," N. W, 1911, ("Iim]). xvii. POISONOUS BACTERIAL PROTEINS 125 animals.''- Animals immunized against endotoxins develop in their serum sub- stances that arc bactericidal and agglutinative to the bacteria from which the poisons are derived, but the serum will not neutralize the endotoxins. As a re- sult, we are unable to perform experiments indicating whether endotoxins have the same structure as the true toxins, i. e., a haptophore and a toxophore group, but presumablj- their nature is different in some essential particular. The chemical nature of the endotoxins is also unknown, for they are always obtained mixed with the other constituents of the bacteria.*' Tuberculin, once supposed to be an albumose, is produced even when the bacilli are grown on a protein-free medium, and in the active solution no albumose or other protein is then found. Hence it seems probable that tuberculin is of the nature of a polypeptid, which gives no biuret reaction but fs destroyed by pepsin and trypsin, according to Loevenstein and Pick,*^ but not by erepsin (Pfeiffer).** Whether tuoerculin should be considered an endotoxin liberated by the disintegra- tion of the bacilli in the cultures is unknown; Pick looks upon it as a secretion of the bacilli, and closely related to the true toxins. Since far more bacterial diseases are brought about by endotoxins than by true toxins, the failure to secure antitoxins for these substances has been a great check in the progress of serum therapj', and the problem of the endotoxins is one of the most important in the entire field of immunity. Poisonous Bacterial Proteins If we filter a Iwuillon culture of diphtheria bacilli through porcelain, wash thoroughly with salt solution the bacteria remaining, and collect them thus freed from their secretion products, it will be found that extracts of the bacterial subtance or the bodies of the killed bacteria themselves are quite free from the typical toxin. This indicates that the toxin is eliminated from the bacteria as fast as it is formed, and no considerable quantity is retained within the cell. The bacterial substance, however, or proteins isolated from it, is found to produce severe local changes when injected into the bodies of animals, necrosis and a strong inflammatory reac- tion with pus-formation being the chief features. This local effect is not a specific property of the diphtheria bacillus, for other bacterial proteins, including proteins from non-pathogenic bacteria, will produce the same changes; indeed, many proteins that are derived from vegetable and animal sources have equally marked pyogenic properties. All foreign proteins when introduced into the circulation of animals are more or less toxic, and the toxic effects of the bacterial proteins are, for the most part, neither specific nor particularly striking. There are a few pathogenic organisms, however, which seem to produce neither true toxins nor endotoxins, notably the tubercle bacillus and the anthrax bacillus, and with these there may be a relation between their protein constituents and their specific effects. Numerous protein substances have been extracted from bacterial cells, partic- ularly nucleoproteins, but also proteins resembling albvmiins, nucleo-albumin, and globulins. In all probability the chief proteins of the bacterial cell are nuclein compounds, which is indicated both by their nuclear staining and by the anah'ses " Positive results are claimed by Besredka (Ann. Inst. Pasteur, 1906 (20), 304), and some others; see Kraus, Wien. klin. Woch., 1906 (19), 655; Zeit. Im- munitiit., 1909 (3), 6-16. It is suggested by Wasisermann (KoUe and Wasser- mann's Handbuch, 1912 (2), 246) that this difficulty in obtaining antiendotoxins depends on the large size of the molecule, — the small diffusible toxin molecule is so altered in its physical condition through union with the antibody that its properties are much altered, whereas the large endotoxin molecule must be di- gested by complement before its toxicitj- is destroyed. *^ The Aggressins of Bail, to which he ascribes the pathogenicity of bacteria, are too little established to permit of a discussion from the chemical standpoint. By many they are believed to be nothing more than endotoxins. (Literature given by Miiller, Oppenheimer's Handb. d. Biochem., 1909 (II (1) ), 681; Dud- geon, Lancet, 1912 (182), 1673). According to Ingravelle (Ann. d' ig. sperim., 1910 (20), 483), tvphoid aggressins are found in the albumins. 5^ Biochem. Zeit., 1911 (31), 142. " Wien. khn. Woch., 1911 (24), 1115; see also Lockmann, Zeit. phvsiol. Chem., 1911 (73), 389. 126 CHEMISTRY OF BACTERIA AND THEIR PRODUCTS of Iwanoff;^^ and many of the nucleoproteins, both of bacterial and non-bacterial origin, cause considerable local inflammatory reaction when injected into animals. Tiberti''' claims that vaccination with non-lethal doses of the nucleoproteins of anthrax bacilli will protect animals against inoculations of virulent anthrax bacilli. Some of the earlier observations on the toxicity of bacterial proteins were erroneous because impure proteins, containing toxins, endo-toxins, and ptomains were used. Schittenhelm and Wcichardt*^ have found, however, that bacterial protein.s are much more toxic than any ordinaiy proteins, as indicated by loss of nitrogen, temperature changes and alterations in the leucocytes of injected animals. Furthermore, there are few other proteins that produce so much inflammatory reaction as the bacterial proteins. \'aughan and his students have been able to split off from the bodies of various pathogenic bacteria toxic materials which are stated to resemble in some respects the protamins,''^ a though they do not all give a satisfactory biuret test. These toxic materials are evidently quite different from either the true soluble toxins or the endotoxins, since they resist heating for ten minutes, at 110"^ in the autoclave with 1 per cent, sulphuric acid, this being a method used for securing the substance. Since the sarcime and B. prodigionus also yield similar toxic products, t;hey cannot be considered as the specific toxic substances of the pathogenic bacteria, but ap- parently are common to all proteins of whatever origin. With .-ome bacteria the splitting process with sulphuric acid sei)arates completely the toxic from the non- toxic insoluble bacterial substance,"" e. s when in contact with eosinopliiles.* PROTOZOA These uuicellular I'orins possess all the chemical characters of the cells of higher forms, even to the more specialized constituents. Thus it has been demonstrated that protozoa contain proteolytic enzymes,'" and that they secrete an acid into their digestive vacuoles. '^ On the other hand, Amoeba coll does not seem to digest the red corpuscles and the bacteria that it takes up.'^ Whether the Amceha coli produces any toxic materials, specific or non-specific, has not yet been determined, but the necrosis that it produces in liver abscesses, when bactewal cooperation can often be excluded by culture, strongly indicates the production of necrogenic substances. Apparently these sub- stances are not chemotactic, in view of the absence of leucocytic ac- cumulation in the lesions of amebic dysentery'. There is also no evidence, clinical or experimental, that amebic infection causes the formation of anti-substances of any kind in the body of the host. The spontaneous recovery from amebic and other protozoan infections, however, may be considered as indicating the development of an immunity against these organisms.'^ Numerous observers have sug- gested the possibility of obtaining artificial immunity against pro- tozoa, and Rossle'^ has obtained immune sera against infusoria. The serum of rabbits immunized against amoebae was foimd b}' Sellards'^ to be cytolytic for the same amoebae, but no antibodies could be foimd in the blood of patients with amebic dysentery. The serum of persons infected with bilharzia is said to give specific com- plement fixation reactions (Fairley) . ^^ Novy '^ has obtained immunity against trypanosomes, but the serum of immune animals will not confer passive immunity. Braun and Teichmann,'^ however, claim ^ Arch. Int. Med., 1913 (11), 165. » Supported by Paulian, Presse Med., 1915 (23), 403. ^ Weinberg and Seguin, Ann. Inst. Pasteur, 1910 (30), 323. "> Mouton, Compt. Rend. Soc. Biol., 1901 (53), 801. " Le Dantec, Ann. Inst. Pasteur, 1890 (4), 776; Greenwood and Saunders, Jour, of Phy,siol., 1894 (16), 441. '- Musgrave and Clegg, Bureau of Gov't. Laboratories, Manila, 1904, No. 18, p. 38. '^ Concerning immunity to protozoanjinfections see Schilling, KoUe and Wasser- mann's Handbuch, 1913 (7), 566. '^ Arch. f. Hyg., 1905 (54), 1; full review of this topic. '5 Philippine Jour. Sci., 1911 (6). 281. '» Jour. Kov. Armv Med. Corps, 1919 (32). 449. '"Jour. Infec. Dis., 1912 (11), 411. '« Zeit. Immunitat., Ref., 1912 (6), 465. 9 130 CHEMISTRY OF THE ANIMAL PARASITES positive results with immune serum from rabbits; they found no poison- ous agent in trypanosome substance. ^^ The fact that trypanosomes themselves readily become immune to various trypanocidal chemicals has been demonstrated and extensively studied in Ehrhch's laboratory. Gonder^o j^^s made the interesting observation that trypanosomes which can be stained by certain vital stains, become unstainable while alive if immune to arsenic compounds, suggesting that this immunity is associated with considerable structural or chemical changes. Plasmodium malariae undoubtedly produces toxic substances, which seem to be of such a nature that they do not diffuse from the red corpuscle, but are only liberated when the corpuscle breaks up on the maturation of the parasite. In this way the characteristic par- oxj^smal manifestations of the disease are produced. The nature of the poison or poisons is unknown, but we have evidence that it is hemolytic, since malarial serum may hemolyze normal corpuscles, ^^ and extracts of the parasites are stronglj^ hemolytic (Brem-^); prob- ably the malarial hemoglobinuria is caused by this hemolj-sis. Pre- sumably malarial poisons are not extremely toxic for parenchymatous cells, since the parenchymatous lesions in malaria seem to be relatively slight as compared with the intensity and duration of the intoxication. Some authors state that the toxicity of the urine is increased after the paroxysm,-^ which, however, does not necessarily indicate that a poison formed by the parasites is excreted in the urine. Immunity seems to be seldom developed against the malarial poison or against the parasite itself, although some persons seem to be naturally im- mune, w^hile some acquire immunity through previous infection.-^ The blood of persons with malaria seems to contain no antibodies for the parasite (Ferrannini),^^ although it seems to have some antihemo- lytic power (Brem). (Concerning the pigment present in the ma- larial parasites see "Pigmentation," (Chap, xviii). Sarcosporidia of sheep yield aqueous and glycerol extracts that are highly toxic for rabbits (Pfeiffer), the poisonous constituent of which was called sarcocystin by Laveran and Mesnil.-^ This is so highly toxic that 0.0001 gm. is fatal to rabbits (per kilo), other ani- mals being less susceptible. It loses its toxicitj'' on heating at 85° for twenty minutes, and is impaired at 55-57° for two hours. It 1" Hintze (Zeit. f. Hyp;., 1915 (80), 377) obtained little immunity with T. brucei, hut Scliiiling and Kondoni (Zeit. Imiminitat., 1913 (IS). t)51) obtained'a poison from Nagana trypanosomes which produced active imnumity in mice. When trypanosomes are killed by weak electric currents they may liberate an active poison (Uhlenhutli and Seyderhelni, Zeit. Immimitjit., 1914 (21), 366). "Zeit. Immunitilt., 1913 (15), 257. " See Regnault, Kevue de MM., 1903 (23), 729. "Arch. Int. Med., 1912 (9). 129. "(Quoted from HIanchurd, Arch. d. Parasitol., 1905 (10), 83; this article gives a r6sum(' of tlie siibj(!c( of the iox'w. substances produced by the animal parasites. " See Celli, Cent. f. Hakt., 1900 (27), 107. " Kiforma Med., 1911 (27), 177. =» Compt. Rend. Soc. Bio!., 1899 (51), 311. CESTODES 1 '4 1 produces pruritis ami otluu- aiiapliylactic symptoms, and altliouf^h the serum of sheep with this parasite does not confer passive anaphylaxis to sarcosporidia, yet it does give positive complement fixation." That it is a true toxin is shown by Tcichmann and Braun,-** who produced an effective antitoxin by immunizing rabbits; only rabbits seem to be susceptible to the toxin. The sarcosporidia contain also a distinct thermostable agglutinin. The lethal dose of dried substance of sar- cosporidia is, for rabbits, but 0.0002 gm., and the poison seems to unite with the lipoids of the nervous system (Teichmann).^^ It is probable that the pathogenic protozoa, at least in some instances, have a semi- permeable membrane about them, for GoebeP'' found that trypano- somes are very susceptible to changes in osmotic conditions. rCESTODES] Taenia echinococcus has been by far the most studied, its abundant fluid content furnishing suitable material for investigation. That this fluid is toxic has been repeatedly observed when, through rup- ture or puncture, the fluid has escaped into the body cavities; such accidents are often followed by violent intoxication, sometimes by death. 3^ As long as the cyst is unopened no toxic manifestations are observed. The most constant symptoms are local irritation and in- flammation, accompanied by urticaria, which may also be produced experimentally in man if the cyst contents are injected subcutaneously. The symptoms are so strikingly similar to those of anaphylactic intoxication, that it is now generally believed that they are the result of such a reaction in a person sensitized by absorption of antigenic substances from the cyst.^^ Carriers of echinococcus cysts have been found to have in their blood antibodies giving precipitins^ and complement fixations'* reactions with extracts of echinococcus, and sometimes with other taenia. ^^ The antigen of the echinococcus is be- lieved by some to be a lipoid ;S6 in the case of Taenia saginata, at least, it seems to be associated with the lecithin (JMeyer^^). Graetz," however, states that the protein of the hydatid cyst is derived from the host, and that it is therefore incapable of causing anaphylaxis in that host, but it may undergo alterations in the cyst so that it is toxic after the order of anaphylatoxins (q. v.) . The complement fixa- " McGowan, Jour. Path, and Bact., 1913 (18), 125. 28 Arch. f. Protistenk., 1911 (22), 351. " Ibid., 1910 (20), 96; see also Knebel, Cent. f. Bakt., 1912 (66), 523. 30 Ann. Soc. Mrd. d. le Gand, 1906 (86), 11. 3' See Achard, Arch. gen. de Med., 1887 (22), 410 and 572. ^^ See Boidin and Laroche, Presse Med., 1910 (18), 329; Ghedini and Zamorani, Cent. f. Bakt., 1910 (55), 49. " Welch, et a/., Lancet, 1909, Apr. 17. " Kreuter, Miinch. med. Woch., 1909 (56), 1828; Weinberg, Ann. Inst. Pasteur 1909 (23), 472. 3^ Meyer, Berl. klin. Woch., 1910 (47), 1316; Zeit. Immiinit.nt., 1910 (7), 732. 3« Israel, Zeit. Hyg., 1910 (66), 487; Meyer, Zeit. Immunitat., 1911 (9), 530. 3' Zeit. Immunitat., 1912 (15), 60; general review. 132 CHEMISTRY OF THE AMMAL PARASITES tion reaction with echinococcus fluid has been found quite rehable in the cHnic, 93 per cent, of positive reactions having been obtained in 500 cases collected by Zapelloni,'^ while controls were always negative. The fluid of the echinococcus cysts has generally a specific gravity of 1005-1015, and contains 1.4-2 per cent, of solids. Most abundant are sodium chloride, about 0.8 per cent., and sugar, 0.25 per cent., the latter presumably coming from the glycogen contained in the wall. Cholesterol is often abundant, while inosite, creatin, and suc- cinic acid are often found. Clerc has found traces of lipase, but other enzj'^mes seem to be absent or in very small amounts. Proteins are present only in traces, unless inflammation has occurred. Schil- ling''^ found the molecular concentration of the cyst fluid to be quite the same as that of the patient's blood. The fluid is said not to be toxic to laboratory animals.*" The cyst wall consists of a hyaline substance which seems to stand between chitin and protein, and probably consists of a mixture of both. Because of the chitin it yields about 50 per cent, of a reducing, sugar-like body when boiled with acid. Glj'cogen is also usually present, but it is limited to the germinating membrane."'^ Other cestodes, when in the cystic form, contain fluids which are more or less toxic. Thus Moursou and SchlagdenhaufTen*- found a "leucoma'in" in the Cy.sticercus tenuicoUis, the larva of Taenia 'mar- ginata, which causes urticaria and other toxic sjaiiptoms when in- jected into animals (thus resembling histamine). The fluids of Cysticercus pisifonnis (the common cestode of rabbits) have been found toxic for frogs, and Vaullegeard*^ has determined the presence of an "alkaloid" and a "ferment toxin" in this fluid. The fluids of the cysts of Ccenurus cerebralis, Ccenurus serialis, and Echinococcus polymorphus have all been found toxic, and it is probable that this is a general rule with the cestodes,'''' but human forms other than the echi- nococcus seem not to have been investigated;'''' according to Jnmmes and Mandoul, extracts of taenia are bactericidal.^" Dibothriocephalus latus ficqucntly causes anemia, which has been attributed to a poison liJK'rated by the parasite when it undergoes disintegration, and possibly as a secretion of the living worm.'*^ All the intestinal cestodes iire equipped with a well-developed excretory apparatus, and it is easy to imagine that their excretory prochicts may be toxic to tlic animal into whose intestine they ai'e exereteil. 3» Policlinicu, Suim;., liKf) (22j, \us. ti 11. 3»Cent. inn. iMid., 1904 (2.'3), HXi. ■"•Uraetz. Cent. f. IJukt., 1910 (.Wj, 2:U; Zi-it. Iniiiuinitat. 1912 (15), 60. ^' Hrault und Loupcr, .K)ur. I'livs. et. Tatli. m''n., H'Ol (ti), 295. ^■-Coinpt. Hcnd. S(.c. Bio!,, I8S2 (95), 791. ■••' Bull. S(K'. liniu'cniic (li? Nurinaiulie, 1901 (l), SI. " HhuK-liaid., Iin- ril.-' ■"• Scniaiai- iiu'd., 1905 (25j, .55. ^« See alo .Joycu.x, .\rcli. d. I'aiasitol., 1907 (ID, 409. *'' Litt'ia lire l)\- Kl;iiu-liai(l, lor <•//.-■' CESTODES 133 Talhivist^"' has made oxIcMisivc studies of Ix)! liiioccplialns, whicli show that the active hemolytic agent is contaiiuMl In the lii)oids of the parasites, presumably as a cholesterol ester of oleic acid."*" The j)roglottidcs contain a proteolytic enzyme, which apparently digests the substance of dead segments, liberating the hemolytic lipoid, which constitutes about ten per cent, of the solids of the parasite. There is also a hemagglutinin, which, unlike the hemolytic substance, is ther- molabile, and causes the appearance of an antibody in immunized animals. In common with other parasites, antitryptic and antijieptic effects are exhibited by extracts. Rosenqvist'" has studied the metabolism of twenty-one cases of bothriocephalus anemia, and found evidence in nearly all of a toxo- genic destruction of protein, which ceases promptly when the worms are removed. He has found that these worms produce a poison which is globulicidal, and probably also generally cytotoxic, since in the anemias that they produce, the elimination of purine bodies of tissue origin (endogenous purine) is increased. The nitrogenous metabolism is quite the same in pernicious anemia and in ])othriocephalus anemia. Isaac and v. d. Velden"' state that the blood of patients infected with this parasite gives a predpitin reaction with autolytic fluid obtained from bothriocephalus, and that rabbits immunized with such autolytic fluids developed a precipitin. Complement fixation relictions may be demonstrated in human infections with bothriocephalus or other taenia (Jerlov).^^" Other Taenia. — There is much less evidence that other forms of tffuia produce toxic substances which injure their host, although the clinical manifestations ob- served in persons harboring tajnia are often of such a nature as to indicate strongly an intoxication. Jammes and Mandoul*- found no toxic manifestations produced by extracts of Tcenia saginata, which negative finding is supported by Cao,^' Tall- qvist and Boycott," using various sorts of taenia. These results contradict the earlier positive findings of Messineo and Calamida,^^ who found extracts of taenia from dogs to be hemolytic, chemotactic (especiallj^ for eosinophiles), and to cause local fatty degeneration in the liver. Extracts of T. perjoliata and plicata (of the horse) were found highly toxic for guinea-pigs by Pomella,^' the liematopoietic organs being greatly stimulated. Bedson" found that extracts of all sorts of helminths produced similar effects on guinea-pigs, the chief lesions being in the adrenals and thyroid. Possibly these differences in results are due to the fact that different parasites were studied by different investigators; furthermore, tests of toxicity of human parasites upon rabbits and guinea-pigs can haidly be con- sidered conclusive. Le Dantec did not find a precipitin for T(snia saginata ex- tracts in the blood of persons harboring this parasite, and negative results with « Zeit. klin. Med., 1907 (61), 427. *^ Faust and Tallqvist, Arch. exp. Path. u. Pliarm., 1907 (57), 307. ^o Zeit. klin. Med., 1903 (49), 193. 5> Dent. med. Woch., 1904 (30), 982. s'" Zeit. Immunitat., 1919 (28), 489. "Conipt. Rend. Acad. Sci., 1904 (138), 1734. " Riforma med., 1901 (3), 795. ^ Jour. Pathol, and Bacteriol., 1905 (10), 383. « Cent. f. Bakt., 1901 (30), 346 and 374. « Compt. Rend. Soc. Biol., 1912 (73), 445. " Ann. Inst. Pasteur, 1913 (27), 682. 134 CHEMISTRY OF THE ANIMAL PARASITES several other taenia were obtained by Langer,^* but complement fixation reactions may be given/* Picou and Ramond^" state that tirnia extracts undergo putrefaction very slowly, and attribute this to a bactericidal property, which was observed with several forms of ttmia by AUesandrini. Weinland^' has found that many intes- tinal parasites exhibit antitnjplic properties, ^^ but in a study of the histological changes of autolysis I observed a t enia in the intestine of a dog undergo more rapid karyolytic changes than did the intestinal epithelium. Dastre and Stessano*' state that extracts of Taenia serrata act upon enterokinase rather than on tryp- sinogen. NEMATODES Ascaris. — The toxicity of members of this group has been a matter of dispute, although, as with the Taenia, there have been observed in patients symptoms that ^vere more easily explained as due to chemical substances than as due to mechanical irritation. Miram, while studying Ascaris megaloceyhala, suffered from attacks of sneezing, lachrymation, itching, and swelling of the fingers, v. Linstow suffered from a severe attack of conjunctivitis with chemosis after touching his eye with a finger that had been in contact with one of these worms. Others have had similar experiences, and it has been found that the fluid from these worms is toxic to rabbits. In man it seems to affect especially those who have been sensitized by previous poisoning, some persons being entirely insusceptible. An extensive investigation of ascaris from both the chemical and toxicological standpoint has been made by Flury,^* which indicates the source and nature of these toxic substances. Because of the practically anaerobic conditions under which the worms live, Flury believes, the products of their metabolism are charactetized by being incompletely oxidized, and resemble the products of anaerobic bac- teria. Most important of these are volatile aldehj'des and fatty acids, especially valerianic and butyric acids, in less quantities formic, acrylic and propionic acids. The toxicologic action of these volatile substances is of such a character as fully to explain the severe irrita- tion of skin and mucous membranes observed in persons handling these parasites; aldehydes are notoriously inclined to produce con- ditions of hypersensitivenoss, c. r/., formaldehyde. It is quite jiossi- ble that the severe constitutional symptoms observed occasionally in persons infected with ascaris, are produced by these substances or by poisonous substances set free through disintegration of worms which have dietl and renuiined in the ])()wcl. A capillarv poison re- sembling scpsin, poisonous bases acting lik(> atropine and coniine, " Munch, mod. Woch., 1905 (.W), 1(H)5. '» Mover, Zoit. Immunit;it., 1 ami FloLscher, Jour. Infec. i)i.s., 1910 (7), ()25. " Miinch. incd. Woch., 1908 (55), 436. '"Jour. K\\). Med., 1909 (11), 331. " .Vrch. Parasitol., 1910 (14), 5; see also Ashcrolt, C'onipt. Jientl. Soc. Biol., 1914 (77), 442. NEMATODES 137 and that the parasite prochices inucli vohitilo fatty acids, especially butyric; both lecithin and cholesterol were absent. The dermatitis produced by uncinaria larvie is ascribed by C. A. Smith^** to an alco- hol-soluble substance. Watery extracts of Sclerostoma were found by Grosso^* to cause but slight chemotaxis without eosinophilia. Filaria seem not to produce any apj)reciabl(! amount of toxic ma- terial, if we may judge by the slight evidence of intoxication shown by infected incUviduals. Jin exception may be made in the case of the guinea-worm (Dracunculus or F. medinensis). This parasite causes chiefly mechanical injury unless its bodj^ is ruptured, which may happen in attempting to remove it forcibly; this accident is fol- lowed by violent local inflammation or gangrene, which indicates that some powerfully irritant substance is liberated from the torn body of the worm.**" "8 .lour. Amer. Med. Assoc, 1906 (47), 1693. " Folia Hematol., 1912 (14), 18. 8" Earthworms are said by Yagi (Arch, internat. pharmacodyn., 1911 (21), 105) to contain a hemoljrtic substance, "lumbricin," the properties of which he describes. Nukada and Tenaka (Mitt. med. FakuU., Tokio, 1915 (14), 1), found an antipyretic agent which seems to be derived from tyrosine. CHAPTER VI PHYTOTOXINS AND ZOOTOXINS The production of substances possessing the essential features of true toxins is by no means limited to the bacterial cell. In the plant kingdom such substances are formed, and called phytotoxins. Of these, the best known are ricin, abrin, crotin, and robin.' Among the toxins of animal origin, zootoxins, are the venoms of poisonous snakes, lizards, spiders and scorpions, and the serum of eels and snakes. PHYTOTOXINS 2 The chief phytotoxins are as follows: Ricin, from the castor-oil bean {Ricinus communis). Abrin, from the seeds of Ahrus precatorius. Crotin, from the seeds of Croton tiglium. Robin, from the leaves and bark of the locust, Rohinia pseudoacacia. Curcin, from the seeds of Jatropha curcus. In their general properties all these substances are very similar and may be considered together. They resemble proteins in many re- spects, for they can be salted out of solutions in definite fractions of the precipitate, are precipitated by alcohol, and are slowly de- stroyed by proteolytic enzymes. For some time they were referred to in the literature as toxalbumins, until Jacoby stated that, by com- bining the salting-out method with trypsin digestion, he was able to secure preparations of ricin and abrin that did not give the pro- tein reactions. This seemed to place them in the same category with bacterial toxins and enzymes, i. e., large molecular colloids, closely resembling the proteins with which they are associated, but still not giving the usual protein reactions. Because of their great similarity to bacterial toxins this seemed a very probable description, and it has been generally accepted. More recent work by Osborne, Mendel, and Harris,-* however, does not support Jacoby's contention. They found the toxic properties of ricin associated inseparably with the coagulable albumin of the castor beans, and were able to isolate it in such purity that one one-thousandth of a milligram (0.000001 gram) was fatal per kilo of rabbit, and solutions of 0.001 per cent, would agglutinate red 1 The poison in certain i>oas, especially Lnthijrns saluris, which causes severe periphenil i)aialvsis, ('(ilhi/ri.sm) is believed to he an alkaloid. (liXill discussion by Stockman, I'^diiih. Med. Jour., 1917 (10), 277). 2 Ji.'suini' of literature by Ford, Cent. f. Hakt., 1913 (58), 129; Jacoby, Kolle and Wass(M-inann's IIundbu(;Ii, 1913 (2), M.')3. •' Atiier. Jour, of IMiysiol., 190.") (14), 259. 13S IMMUNITY AGAINST PIIYTOTOXINS 139 corpuscles. The toxicity was also impaired or destroyed by tryptic digestion. They consider that probably, because of its extremely great toxicity, Jacoby was able to get active preparations that con- tained too little active substance to give the protein reactions. As they remark: "If one-thousandth of a milligram of a compound giving on analysis every indication of being a relatively pure protein, is physiologically active in the degree characterized by our experi- ments, the toxicity of any impurity must be infinitely greater than that of any known toxins." Against the claim that the toxic principle is simply carried down with the protein is the fact that it does not come down in the first fraction that is precipitated, the globulin, which usu- ally carries down all ipipurities. All the ricin comes down betv\een the limits of one-fifth and one-third saturation with ammonium sul- phate, exactly as does the albumin. During germination of the castor bean the ricin disappears with the albumin.^ Field^ has found evi- dence that the agglutinin and toxin of pure ricin are separable, but Reid believes them identical. Of 21 varieties of ricinus seeds ex- amined by Agulhon,^ all yielded hemagglutinins. Ricin agglutinates not only corpuscles, but tissue cells of all sorts, and causes precipitates in normal serum. '^ Curcin alone seems to have no hemagglutinative action.'* Immunity. — The phytotoxins have been very serviceable in the study of immunity, since thej^ obey the same laws as bacterial toxins and can be handled in more definite quantities. By their use Ehrlich first determined that toxin and antitoxin act quantitatively. They seem to possess haptophore and toxophore groups, and immunity is readily obtained against them, not only by subcutaneous injection, but by dropping into the conjunctival sac, and also by feeding, show- ing their direct absorbability and their resistance to digestion. The antitoxin is present in the milk of the immunized mother and im- munizes the suckling; but little is carried through the placenta into the fetal blood. The immunity is specific, ricin antitoxin, for exam- ple, not protecting against abrin (although it is said to protect against robin). Roemer found that in animals immunized by conjunctival application the eye so used became immune to the local action of the poison before the other eye did, indicating a local development of immune substance. In general immunization the immune substance appears first in the spleen and bone-marrow. Normal serum gives a precipitate with ricin, but immune serum gives a much heavier one. Antiricin, like other antitoxins, is inseparable from the proteins of the serum. ^ Agulhon, Ann. Inst. Paste-ir, 1915 (29), 237. 5 Jour. Exper. Med., 1910 (12), 551; Reid, Landwirtsch. Versuchst., 1913 (82), 393. « Ann. Inst. Pasteur, 1914 (28), 819. ' Michaelis and Steindorff, Biochem. Zeit., 1906 (2), 43. « Felke, Landwirts. Versuchst., 1913 (82), 427. 140 PHYTOTOXINS AXD ZOOTOXINS Physiological Action. — Their poisonous action is manifold, most prominent being agglutination of the erj'throcytes, local cellular de- struction, and, to a less extent, hemolysis. Jacoby believes that in ricin there are several toxic substances differing in physiological prop- erties, similar to Ehrlich's findings in diphtheria toxin (toxones, etc.). B}'- long action of pcpsin-HCl upon ricin, he secured a prepa- ration with all the other properties of ricin except that it was inactive against erythrocytes; the same result could not be obtained with abrin. Heating to 65° or 70° does not destroy the toxicity of phy- totoxins, but boiling does. There is a latent period of several hours after injection of the poison, the onset of sj'mptoms being sudden; death rarely occurs in less than fifteen to eighteen hours (Osborne et al.). Flexner^ has studied particularly the histological changes pro- duced by ricin and abrin poisoning in animals. Both act alike, af- fecting the tissues much as bacterial toxins do (diphtheria). Fever, albuminuria and convulsions are followed by exhaustion and lowered temperature. Punctiform hemorrhages are found beneath the serous surfaces, with fluid in the peritoneal cavity. At least in the case of ricin the hemorrhages are not due to blood changes, but to a spe- cial toxin destroying the endothelial cells. ^^ There occur a general lymphatic enlargement and marked changes in the intestinal mucosa, with swelling of the Peyer's patches. The spleen is swollen and dark in color, as also is the liver, which shows much focal necrosis. The glycogen content of the liver is decreased in abrin poisoning. ^^ Subcutaneous injection causes local edematous inflammation without suppuration. Histologically, in the most affected organs are found much cellular necrosis and disintegration, especially of lymphoid and epithelial cells. Changes in the capillary endothelium, fibrinous thrombi, and abundant hemorrhagic extravasations are widespread. Probably agglutinative thrombosis by red corpuscles plays an im- portant part in these intoxications (Ehrlich), but Aschof^- ascribes the thrombosis to the fragments of disintegrated marrow and blood cells. The great amount of intestinal injury probably depends upon the fact that these poisons are largely eliminated through the intestinal mucosa. There are also severe changes in the bone marrow, accom- panied by the appearance of micleated (M ytlnocytes in the blood. ^•■' Mushroom Poisons.'^ — The jjoisons of the three chief poisonou.s imishrooms, Avinnitd viuscari'i, IlcivcUa escitlrnta, and Amnnita ph nil aides, differ from one another quite essentially. The poisonous principles of tiie first and second, muscarine and helvellic acid, are non-protein substances, of known chemical com- » Jour. Kxper. Med., 1897 (2), 197. '" Amer. Jour. Med. Sei., 190.3 (12(>), 200. " Doyon, Coinpt. Ifend. Soc. liiol., 1909 (t)7), ^50. '= Arch. Int. Med., ]\i\A (12), iiOA. '•' HuntiuK, .Jour. Kxper. Med., 190() (8), t)2'). 'Miesume l)y Mtirner, TTpsala L:ikaref. Korh.. I9I9 (21). 1. l'atli(»l()Kical anatomy described I)y TryMi, xirchows .\rehiv.. 1919 (22(1), 229. SNAKE VEXOMS J 11 position, which :iio discussed elsewhere; hut the A7nanila phalloides, the most important of tho tiuoo, owes its toxic proijcrtics to at least two poisonous con- stituents. One is powerfully hemolytic, is desiioyod hy heating thirty minutes at 0.')°, and acts directly upon red corpuscles without the presence of serum. "^ The studies of Ford'" and his associates have shown that this hemolysin is a glueoside, j'ielding on hydrolysis pentose and volatile bases, and yet capable of actinfi as an antigen, since actively antihemolytic sera can })e produced by im- munizing animals. This substance corresponds to the pfinllin of Kobert. Prolj- ably this hemolytic poison is not the important agent in poisoning by Amanita, as it is easily destroyed by heat and the digestive fluids. The thermostable poi.son, A7/i(iuita-t(>.rin, gives no reactions for either glucosides or proteins,'^ and does not confer any considerable antito.xic jiroperty on the blood of immunized animals. The toxin kills acutely, the animals dj-ing in 24 — 48 hours, and showing no changes beyond a fatty degeneration of the internal organs. The hemolysin kills slowly in 3 — 10 days, causing local edema and hemoglobinuria. Amanita muscaria contains a heat-resistant agglutinin which also seems to be a glucoside, but it is not toxic nor antigenic. An extensive study of many fungi by Ford'* led him to classify the toxic action in three groups: (1) nerve poisons, e. g., muscarine; (2) those causing structural changes in the viscera, e. g., A. phallnidcs, causing fatty degeneration; (3) gastro- intestinal irritants, c. g., Lactariits torjuinosus. The poison of Rhus toxicodendron has also been found bj' Acree and Synie" to be a glucoside,'-" and the same is true of the poison oak, Rhus diversiloba, which has no antigenic properties. 2' (The effects of the phytotoxins on the blood are discussed under "Hemolysis" in Chapter ix. Vegetable hemolytic poisons that do not resemble the toxins, e. g., glucosides, etc., will also be found dis- cussed under the same heading.) ZOOTOXINS" Snake Venoms -^ This important class of poisons, first thoroughly investigated by Weir Mitchell (I860), and Mitchell and Reichert (1883), has re- cently aroused great interest through its relations to bacterial toxins and the problems of immunity. The poisons of different species of snakes seem to have much in common with one another, whether de- rived from the Elaiperine snakes (cobras and numerous other Indian and Australian snakes), or Viperidce (including most poisonous Amer- ican snakes), or Hijdrophince (the poisonous sea-snakes), although very characteristic ditferences exist between each. '* The hemagglutinin of Agaricus cam,pestris is precipitated at a H-ion concen- tration of 2.G X 10-^ (Brossa, Arch. sci. med., 1915 (39), 241). '6 See Jour, of Pharm., 1910 (2), 145; 1913 (4), 235, 241, and 321. '^ Rabe (Zeit. exp. Path., 1911 (9), 352) considers it to be an alkaloid. '» Jour, of Pharm., 1911 (2), 285. '9 Jour. Biol. Cheni., 1907 (2), 547. -"Questioned by McNair, Jour. Amer. Chem. Soc, 1916 (38). 1417. ^' Adelung, Arch. Int. Med., 1913 (11), 148. -- Full review and literature given by Faust, "Die tieris.ihen Gifte," Braun- schweig, 1906; also in Abderhalden's Handbuch, Vol. II. iSachs, Kolle and Was- sermann's Handbuch, 1913 (2), 1407. -^ Elaborate review and bii)Iiography given by Noguchi, Carnegie Institution Publications, 1909 No. Ill; also by Calmette, "Les venins. les animaux venimaux et la serotherapie antivenimeuse," Paris, Masson, 1907; Calmette. Kolle and "VVas- sermann's Handbuch, Vol. II, p. 1381; with reference to North American snakes, see Prentiss Wilson, Arch. Int. Aled., 1908 (1), 516. 142 PHYTOTOXINS AND ZOOTOXINS The essential anatomical differences between the different classes of snakes are as follows: Colubrickp, whicli include all the non-poisonous snakes, have no mechanism for injecting poisons into their victims. Colubridce venenoscB are venomous snakes resembling in many particulars the harmless Colubrines. but having short poison fangs, firmly fastened to the maxilla in an erect position; in this class are included the cobra and the venomous snakes of Australia. Vi- peridce, or vipers, are characterized by a highly specialized apparatus for in- jecting the poison; their poison fangs are very long, and the maxillary bone, to which they are fastened, is so articulated that it rotates about a quarter of a circle when the snake strikes, bringing the fangs into an erect position. The fangs are canalized and pointed at the end like a hypodermic needle, and the poison is forced through them under considerable pressure bj^ a large muscle that contracts over the salivary gland. Accessory fangs in various stages of develop- ment are also present to replace any fang lost in action. All the poisonous snakes of Morth America, with one insignificant exception, belong to the vipers, and to a special class known as the "pit vipers," because of the presence of a deep pit of unknown function above the maxilla. The exception mentioned is the "coral snake" found on the coast of Florida, around the Gulf of Mexico and in the south- eastern states; it is a member of the colubrine poisonous snakes, of small size, and seldom causes serious poisoning. The poisonous vipers are the rattlesnakes (Crotahis), of which there are some ten to twelve or more species, and Sistrurus of which there are two species ; the copperhead adder {Andstrodon coniroirix) and the water mocassin {Andstrodon piscivorus). The classification used above is the one followed in most publications on poisonous snakes; a more modern classification divides the snakes {Ophidia) into several .series, one of these including all poisonous snakes under the title of Pro- teroglypha, and dividing this series into the three families: (1) Elapimv, including cobras, coral snakes, etc.; (2) Hydrophince, the poisonous sea-snakes; (3) Viperidce, including all snakes with erectile fangs. -^ The source of the venom is probably in part the blood, since snake blood has been found to contain poisons very similar to some of those in the venom; therefore these are presumably simplj^ filtered out by the venom glands, and not manufactured by them.^^ Other poison- ous constituents of venom are not found in snake serum, and there- fore are probably manufactured by the venom gland. Apparently many of the harmless snakes produce a poisonous saliva, since extracts of their glands are said by Blanchard^^ to possess the properties of the venoms, and if so these snakes are harmless chiefly because they lack an apparatus for injecting the poison. As a rule, however, the venom glands are much more highly developed in the poisonous snakes, and are connected with a specialized injection apparatus; in structure they are compound racemose glands. Properties of Venom. — As ejected, the venom is weakly acid or neutral in reaction, and free from bacteria, contrary to earlier ideas (Langmann). Its specific gravity is 1030 to 1077, and it contains a large amount of solids, generally 20 to 40 per cent, by weight. These are precipitated by alcohol, ether, tannin, and iodin, but do not ad- ^* For a full discussion of the characteristics of the poisonous snakes of North America, see the monograph with that title by Stejneger, Report of U. S. Na- tional IVIuseum, 1S<.).'{, Wasliingtou. A good summary is also given by Langmann, Jlefcreiic(! Handbook of Medical Sciences, ('oiicerniiig })oisonous sca-snakos, Ili/droplniliu, see Hoiilanger, Natural Science, 1S92 (1), 41. The poisonous snakes of India are descril)ed by Fayrer, in "The Thanatopliidia of India," JiOndon, 1874. '^'' Contradicted by Arthus, Arch, internat. phvsiol, 1912 (12), 102. ^» Com])t. Uend. Soc. Biol., 1894 (40), 35. SNAKE VENOMS 143 here to precipitates of phosphates as do enzymes and toxins (Cal- mette). They do not diffuse through dialyzing membranes. When dried, the venom can be kept almost indefinitely without losing its strength, specimens over twenty years old having been found unim- paired. Glycerol and alcohol also seem not to injure it, but oxidiz- ing agents of all kinds arc very destructive. Light impairs the power of venoms, as also does radium (Phisalix).^^ Eosin and erythrosin also reduce the power of venom through their photodynamic action, affecting the neurotoxic properties less than the hcmatotoxic compo- nents (Noguchi).^^ Cobra venom withstands even 100° for a short time, but crotaline venoms are destroyed at 80-85°. Much work has been done upon the nature of the constituents of venom. As early as 1843 Prince Lucien Bonaparte found that there were proteins in the venom, which was corroborated by Mitchell in 1861. In 1883 Mitchell and Reichert described two poisonous pro- tein constituents of venom, one of which was coagulable by heat and seemed to be a globulin; the other resembled the proteoses (they called it "peptone," according to the nomenclature of that time). To the globulin they ascribed the local, irritating properties of venom • to the albumose, the systemic intoxication. Corresponding to their action, venoms of different serpents were found to vary greatly in the proportions of these proteins. Cobra venom, which acts chiefly sys- temically, contains 98 per cent, of albumose and but 2 per cent, of globulin; rattlesnake venom, with its marked local effects, contains 25 per cent, of the irritating globulin; moccasin venom contains 8 per cent, of globulin. Several other observers soon corroborated the main facts of Mitchell and Reichert's report; but, as has been seen in connection with the consideration of the composition of enzymes, tox- ins, etc., the fact that a substance is carried down with a protein is no proof that it is itself a protein. What has been established is merely that the irritating component of venom can be destroyed by heat, and is removed with the globulin in fractional separation; while there remains a substance not destroyed by boihng, which comes down at least in part with the albumoses of the venom, and causes chiefly systemic manifestations. Since venoms act as antigens and stimulate the formation of spe- cific antibodies, it is to be presumed that the poisonous principles are proteins, or toxalbumins, although this conclusion does not neces- sarily follow. Faust^^ believes the poison of venoms not to be proteins, but glucosides, free from nitrogen, resembling very much quillajic acid, and therefore belonging to the saponin group of hemolytic agents. He has isolated such a substance from cobra venom, which he calls ophiotoxin (C17H26O10), and from rattlesnake venom a sub- " Compt. Rend. Soc. Biol., 1904 (56), 327. =« Jour. Exper. Med., 1906 (8), 252. -9 Arch. exp. Path. u. Pharni., 1907 (56), 236; 19111(64), 244 144 PHYTOTOXINS AND ZOOTOXINS stance which seems to be a polymer of the ophiotoxin, (C34H54O21). Possibly these glucosides are bound to proteins, forming compound proteins which act as specific antigens. According to this work the snake venoms and the dermal poisons of toads and frogs are all closely related substances. Enzymes in Venoms. — .\s venom causes rapid liquefaction of tissues into which it is injected, Ploxner and Noguchi^" tested crotalus and cobra venom for proteases, and found that the\^ digested muscle rapidlj', and also gelatin and unboiled fibrin; whereas boiled fibiin and boiled egg-albumen were undigested.^' Kinases and nucleases are also present in venoms (Delezenne).*- \\'ehrmann'^ found that venom digests fibrin and inverts saccharose, but does not digest starch. Martin^^ found fibrin ferments in various venoms, which are probably important agents in causing thrombosis. There are also active lipases in venom=, to which many of the effects, especially hemolj'sis and fatty degeneration of the tissues, may be at least partly due (Noguchi), and the hemolysin of cobia venom seems to be a lipase that splits lecithin into hemolytic substances (Coca).^* Delezenne'- found zinc always present in venom and attributes to it a relation to the enzyme activity. Toxicity. — Calmette has determined the toxicity' of several ven- oms, and gives the following figures: 1 gm. cobra or aspis kills 4000 kgm. of rabbit. 1 gm. hoplocephaius kills 3450 kgm. of rabbit. 1 gm. fer de lance or pseudechis kills 800 kgm. of rabbit. 1 gm. Crotalus horrid us kills 600 kgm. of rabbit. 1 gm. Pelias berus kills 250 kgrn. of rabbit. The danger of the bite depends not only upon the difference in the strength of the venom of different varieties of serpents, but also upon the size of the snake, the time of year and condition of hunger or plenty, and particularly whether the entire discharge is injected suc- cessfully or not. The fatal dose of cobra venom for an adult man is variously estimated at from 0.01 to 0.03 gm., while the venom of Hydrophince is about ten times as toxic; for crotalus venom the lethal dose is probably 0.15 to 0.3 gm. (Noguchi). Probably in the major- ity of strikes, by no means all the fluid ejected by both fangs is in- jected beneath the skin of the victim. A large diamond rattler may eject as much as a half teaspoonful of venom at one discharge and such a dose would usuallj' l)e fatal. Repeated ejections decrease the strength of the venom rapidly, until it may have almost no toxicity. In general, venom is most active in warm weather and immediately after the snake has fed; in winter its toxicity is slight. The mortality in America from snake-l)ites is very hard to ascer- tain, various authors giving figures at w'ulo variance. The extensive »« Univ. of i'uuM. iMed. liull., 1902 (15), 300. '• See also lloussay and Xegrete, Revista de I'insl. l)act. Buenos Aires, 1918 (1), 431. 32 Anil. lust. Paslcur, 1919 (33), ()8. " Ann. (1. I'ln.st. I'ustcur, 1,S9S (12), 510. ^* Jour, of IMiysiol., 1905 (32), 207. ^''Jour. Infect. Dis., 1915 (17), 351. SNAKE VENOMS 1 »•') studies of Willsou'"' show about. t(Mi per cent . mortality from all venom- ous snakc-bitcs in this country, the (Uffercnt species giving figures as follows: Coral snakes, twenty to fifty per cent.; water moccasins, seventeen per cent.; large rattlesnakes, eleven to twelve per cent.; copperheads and ground rattlers, no mortality excej)! in children or in cases of complications. The mortality in children is at least double that in adults. Many deaths from snake-bites of all kinds are due to the treatment rather than to the bite. The poisonous snakes of Australia, although numerous, are not very virulent, and the mortality is given as about seven per cent. A full charge of venom from the cobra and many other Indian snakes is inevitably fatal (Fayrer). The crotaline snakes of the tropics are more venomous than those of the north, Lacheris laticeolatus of Central America and Mexico being nearly as dangerous as the cobra. When venom is taken into the stomach in the intervals of diges- tion, enough may be absorbed to produce death, especially in the case of those venoms which contain a large proportion of the albumose, which is dialyzable; but during active digestion the venom undergoes alteration and is rendered harmless. It has been found experiment- ally in animals that cobra venom placed in the stomach causes ordi- narily no harm whatever, but if a loop of the intestine is isolated, a fistula established and allowed to heal, venom introduced through this opening always produces death. It is probably not so much the pepsin and hydrochloric acid that destroys the venom, as the trypsin. If the bile-duct is ligated, the venom is destroyed just the same. Much of the venom seems to be eliminated into the stom- ach, no matter how it is introduced into the system, and apparently it is also partly excreted by the kidneys. Rattlesnake venom seems not to be absorbed through mucous membranes. Physiological Action. — As indicated in the preceding paragraph, the effects of the bites of different classes of snakes arc quite differ- ent. Langmann describes the symptoms as follows: Cobra Poisoning. — "Within an liour, on an average, the first constitutional symptoms ai)pear: a pronounced vertigo, quickly followed by weakness of the legs, which is increased to paraplegia, ptosis, falling of the jaw with paralysis of the tongue and epiglottis; at the same time there exists an inability to speak and swallow, with fully preserved sensorium. The symptoms thus resemble those of an acute bulbar paralysis. The pulse is of moderate strength until a few minutes after the cessation of respiration; the latter becomes slower, labored, and more and more superficial until it dies out almost impe.ceptibly. Death occurs at the latest within fifteen hours; in 32 per cent, of all cases in three hours. There are very few local changes." Cushny" finds that cobra venom produces paralysis of the motor nerve terminations of .muscle, resembling the action of curare; the central nervous system is not directlj^ involved. Death recults from failure of the moto ■ nerve ends in the respiratory muscles to transmit impulses to the muscles. Alkaloids that are antagonistic to curare (physostigmine, guani- dine) are not effective in cobra poisoning, but are them,selves rendered inactive. "« Arch. Int. Med., 1908 (1), 516. " Trans. Roy. Soc, London (B), 191(5 (208), 1. 10 146 PHYTOTOXINS AND ZOOTOXINS Viper Poisoning. — "After the bite of a viper the local changes are mo>t pro- nounced; there are violent pains in the bleeding wound, hemorrhagic discolora- tion of its surroundings, bloody exudations on all the mucous membranes, and hemoglobinuria. Usually somewhat later than in cobra poisoning constitutional symptoms develop; viz., great prostration with nausea and vomiting, blood pres- sure falls continuously, and respiration grows slow and stertorous. After a tem- porar}'- increase in reflexes, paresis supervenes, with paraplegia of the lower extremities, extending in an upward direction and ending in a complete paralysis. It therefore resembles an acute ascending spinal paralysis. If the patient re- covers from the paralysi-s, a septic fever may develop; not rarely there remain suppurating gangrenous wounds, which heal poorl}^" It will be noticed that there is lacking the usual period of incuba- tion that follows injection of bacterial toxins, and if it happens that the venom has been injected directly into one of the veins, death may occur within a few minutes. When recovery occurs, the disappear- ance of symptoms is remarkably abrupt, within a few hours a des- perately sick person becoming almost entirely free from all evidences of the intoxication. Pathological Anatomy. — Postmortem examination shows changes varying with the nature of the poisonous snake that has caused death. In the case of a cobra bite, according to Martin, the areolar tissue about the wound is infiltrated with pinkish fluid; the blood is often fluid; the veins of the pia are congested, and the ventricles often contain turbid fluid; the kidneys may show much congestion. When death occurs in a few minutes, enormous general intravascular clotting is found, which seems to be the cause of death. After death from a viper bite the site of the wound is the seat of intense edema and extrava-ation of blood; if in the muscles, these are much softened and disorganized. Hemorrhages are found in all organs and in the intestinal tract. If death occurs after several days it is generally because of sepsis, and shows the usual changes of this condition; in addition, as a rule, to marked gangrenous, ulcerative, and sloughing processes at the site of the bite. Histologically there are found, in addition to innumerable hemorrhages in nearly all the organs, many vessels plugged with thrombi composed of more or less hemolyzed, agglutinated erythrocytes. The changes produced in the nervous tissue by the Australian tiger snake are described by Kilvington,'* who found marked chromatolysis, the Xissl bodies breaking into dust-like particles, and eventually all stainable substance disappearing from tlie cytoplasm; the nucleus retains its central 7)osition, but often loses its outline and may disappear. The cells around the central canal of the cord are most affected. There are no inflammatory changes in the nervous sj'stem, and if death occurs very quickly there may be no microscopic alterations. Hunter'^ found similar changes in the Xissl bodies in both krait and cobra poisoning; in the mcdullated fibers he found the myelin sheath converted into ordinary fat. The venom of sea snakes {Enhydrina vn'aka- dien) has a severe action on the nervous tissues, while Dnboia has none (Lamb and Hunter^"). Nowak" studied experimental animals, and found much fatty change in the livers, even if death occurrcil one-half hour after iH)isoning; also focal necrosis in the liver, acute parenchymatous alterations in the kidney, and pneumonic patches in the lungs. Effects on the Blood. — There has been much discussion concerning the part played by the abundant and prominent intravascular clotting in causing death after snake-l)ite. Lamb" states that when venoms are slowly absorbed tiie coagulahility of the blood is decreased and it is found fluid after death, but when a fatal do.se of venom (vi!)er) is rapidly absorbed, clotting is increased and thmni- bosis is the chief cause of death. Martin has demonstrated very lu-tivo fibrin " Jour, of Physiol., 1902 (28), 426. " Glasgow Med. Jour., 190.'} (59), 98. ^"Lancet, 1907 (ii), 1017. *' Ann. d. I'Inst. I'asteur, 1898 (12), 3(59. " Indian Medical Gazette, Dec, 1901. SNAKK VhWOMS 1 17 forinciits in snake venom (Inc. rit.). It is highly proliahl ■, iiowever, thit many of tlie thrombi of venom poisoning are not prixhiced by coagiihition of fibrin, but by agpihttination of tlie r(>(i corpusflos, wliich Floxncr'^ has .sliown can cause largo clots in the heart and great vessels, as well as "hyalin" thrombi in the small vessels, lloussay"" states that most snake venoms destroy the cytozyme (which combines with serozyme and calcium to form thrombin), so that the blood becomes incoagulable. The Argentine crotalus and lachesis venoms, however, coagulate even <'it rated blood. Nature of Venoms. — The varied effects produced by venoms have been found to be chic to a number of poisonous elements which they contain, and Avhich have been chstinguished and separated from one another by Flexner and Noguchi.^"^ These are hemotoxins [hemoly- sins and hemagglutinins) , leucocytolysins, neurotoxins, and endothel- iotoxins {hemorrhagin) , but it must be taken into consideration that Fausf*^ beheves that the single glucosidal poison which he has found in rattlesnake venom is responsible for all the effects of the venom, except the hemagglutination. [In another place (see "Hemolysis") the nature of the hemolytic agent is discussed.] Venom agglutinin is quite independent of the hemolysin, for it is destroyed by heating to 75°-80°, whereas the hemolysin is destroyed only partly at 100°. Agglutinin acts in the absence of serum complement, and therefore is not an amboceptor; it is apparentlj^ more like the toxins in its na- ture. The agglutination of the corpuscles does not interfere with their subsequent hemolysis. Michel states that the agglutinin of cobra venom can be separated from the hemolysin and the toxin by means of ultrafiltration through collodion membranes, as the agglutinin exists in larger molecular aggregates. "^^ ' The leucocytotoxins were found by Flexner and Noguchi to be quite distinct from the hemolysins, for after saturating all the hemoly- sin with red corpuscles, the venom still shows its effects on the leucocytes, which effects consist in cessation of motility and disintegra- tion, affecting particularly^ the granular cells. The leucocytotoxin, however, resembles the hemolysin in that it appears to be an ambo- ceptor. Leucocytes are also agglutinated by venom, possibly by the same agglutinin that acts on the red corpuscles. Serum complement is inactivated in vitro by cobra venom through changes in the globulins brought about by the venoms.'*'^ By saturating venom with either red corpuscles or nerve-cells it was found by Flexner and No- guchi that the toxic principle for each is distinct and separate. ^^ Other sorts of cells, however, are able to combine, or at least remove some parts of the toxic elements, but to a much less degree. The neurotoxin, like the hemolysin, resembles an amboceptor, and since *^ Univ. of Penn. Med. Bull., 1902 (15), 324. "" Prensa Med. Argentina, 1919 (6), 133. *' Jour. Exp. Med., 1903 (9), 257; l^niv. Penn. Med. Bull.. 1902 (15), 345. ••s Arch. Exper. Path. u. Pharm., 1911 (64), 244. « Compt. Rend. Soc. Biol.. 1916 (77), 150. " Hirschfeld and Klinger, Biochem. Zeit., 1915 (70), 398. 148 PHYTOTOXINS AND ZOOTOXINS venom contains no coniplonient, the neurotoxin has first to be supplied with complement by the victim's blood or tissues before it can harm the cells. The venoms are not only toxic for mammalian cells, but also for simple unicellular organisms, including; bacteria; tadpoles are paralj'zed in solutions containing one part of cobra venom per million. ^'^ The pronounced hemorrhage-producing property of serums, par- ticularly that of the rattlesnake, was also found to be due to a specific toxin acting on the endothelium of the capillaries and small veins, and not to the changes in the blood itself, as had formerly been thought. This endotheliotoxin, which Flexner and Noguchi call "hemorrhagin," is quite distinct from the other toxic substances, being destroyed at 75°, a temperature that leaves the neurotoxin and hemolysin unin- jured. Its endotheliolytic action is show in the glomerular capil- laries, where it causes hemorrhage and hematuria (Pearce).^^ Variations in Venoms. — In distribution among the various poi- sonous reptiles these toxins seem also quite distinct from one another, which explains the difference in the effects of bites of snakes of various kinds. Cobra venom contains chiefly neurotoxin, hence the symp- toms of cobra bite are largely of nervous origin, with but little local tissue change. Rattlesnake venom owes its effects chiefly to hemor- rhagin, hence the marked local necrosis and extravasations of the blood, and the generalized hemorrhages; the nervous effects following- viper bite are probably, in part, due to hemorrhages in the nervous tissue. Cobra venom produces great hemolysis and little agglutina- tion. Rattlesnake venom has relatively little agglutinative or hemo- lytic power. Water moccasin and copperhead venoms are more agglutinative than either, and intermediate in heniolytic strength; they cause much local tissue destruction. The exact action of cobra venom on various centers and organs has been studied by ElUot.^" It raises blood pressure when in dihition of 1:10.000,000. by contractinfi; vessels and stiinulatint;; the heart; low lethal doses kill by i)ara- lyzing the respiratory center. Krait (Bungarus coeridues) venom acts similarly, but loss {xiwerfully, and cannot be neutralized by Calmette's antivenin.'^i Sea-snake venoms are by far the most poisonous of all. For Enhi/drina valaLd- dieu the lethal dose for rabbits is O.OOOOti ^ram per kilo body weight. It acts by vagus stimulation and paralysis of respiratory- centers ami oi motor nerve- endings.''-' Hussell's viper {Ddhtda Hussrllii) owes its effects chiefly to intravascular dot- ting, according to Laiiil) and llauMa,'"' and conlnins no neurotoxin. It is not neutralized by Calmette's antivenin. 'I'iie dots are ilue to agglutination anil con- tain no fibrin (Flexner). ••* Bang and Overton, (Biochem. Zeit., 1911(31), 243) state that corpuscles can take up the Mcurotoxin, wliicli is s()lul)le in fats and lipoitls. ^».)()ur. Mxpcr. iMcd., l'.U)'.> (11), .'):VJ. "Lancet, l!»Ot (i), 715. "'Elliot, Siljjir, and ('ariiiidi:iel, b:iiicft, I'.KM (ii), IfJ. '- Fraser and J<:ili()t, Lancet, 1004 (ii), 141; a'so Rogers, .lour, of Phy.siol., 1903 (;}0), iv. 'I'lie al)ove are also given completdv in the Philosophical 'hans- actions of the iloyal Societv, 1904-,'), vol. 1S7. '•'.Jour, of I'a'tli. and liact., 1902 (S), 1. .S.V.I A' A' VK.XOMS 14<) The "(lila Munst(M" {Ilrlotlcrmn suspect urn) sclddrii (■.•iii.s(>s .serious i)()is()iiiiin ill man, l)ut may kill small animals, suoli as frops.'' Its poison is only slightly iKMnolytic. hut prochicos (Icucncnitivc clianpcs in the nervous system fLanRinann). The liemohsin is activated by lecithin (Cooke and Loeb). An elaborate series of studies bv Leo Loeb and his associates give all the known facts concerning the Gila Monster." Loss of Bactericidal Powers. — The frequency of inurketl and persistent sloughing and suppuration at the site of snake-bites, particu- larly from the vipers, and the common termination in sepsis, was attributed by Welch and Ewing^^ to a loss of bactericidal power of the blood, which they found followed experimental venom poisoning. This has been ascribed by Flexner and Noguchi to saturation of serum complement by the numerous amboceptors of the venoms, so that no complement is left for the serum to use against the bacteria. In serum whose complements do not combine with the venom ambocep- tors (e. g., Necturus) the normal bactericidal powers are not in the least impaired by the addition of venom. Morgenroth and Kaya ascribe the loss of complement to a destruction by some agent in the venom. Snake Serum. — The serum of serpents is also toxic for other animals," even when the serpent is not a venomous one; c. g., the harmless pine snake (Pitijophis cateniferis). The toxicity of snake serum seems to depend chiefly uj)on its hemo- toxic etTects (hemagglutination and hemolysis), the toxic substances resembling amboceptors and similar to, but not altogether identical with, the amboceptor of the venoms. Crotalus tissues also produce poisoning in proportion to the blood they contain, but are without toxic effects of their own (Flexner and Noguchi). Antivenin. — Snake venom has the essential property of all true to.xins of immunizing, with the appearance of an antitoxin in the blood. The first successful immunizations seem to have been made by Sewall,^* but the practical production of antitoxic serum was first accomplished by Calmette^^ and by Fraser.''° At first it was be- lieved that cobra' antivenin neutralizes the neurotoxins and hemoly- sins of venoms of any origin, and also of snake serums, and, therefore, should be quite effective against cobra and similar venoms which pro- duce chief!}- neurotoxic and hemolj^tic changes. This implies that these toxic substances are of identical nature in all snakes, no matter how dissimilar the snakes may be, but various investigators, especially Lamb, have found sufficient specificit}' exhibited by different venoms and antivenoms to indicate the necessity of employing the specific antiserum in each case of snake bite. A special antitoxin against rattlesnake venom and its hemorrhagic to.xin has been successfully '* Thorough study by Van Denburgh and Wright, Amer. Jour, of Physiol.. 1900 (4), 209. ' " Carnegie Inst. Publication No. 177, 1913. *« Lancet, 1894 (1), 1236; Ewing, Med. Record, 1894 (45), 663. °' Questioned by Welker and Marshall, Jour. Pharmacol., 1915 (6), 563. '' Jour, of Physiol., 1887 (8), 203. " Ann. d. I'lnst. Pasteur, 1894 (6), 275: also subsequent articles in 1897 (11). 214; 1898 (12), 343. «« British Med. Jour., 1895 (i), 1309. 150 PHYTOTOXINS AND ZOOTOXINS prepared by Noguclii.*'' This crotalus antivenin also neutralizes hemolysins of various venoms, and also of snake serums. Presumably antivenin neutralizes venoms in exactly the same way that antitoxin neutralizes toxins; i. e., cell receptors are thrown off from the injured cells during immunization, which combine with venom amboceptors in the blood, and thus prevent their combining with the cells. Antivenin also prevents the inliibiting action of venom on bactericidal serum, indicating that it prevents the venom ambo- ceptors from binding the serum complement. The reaction of venom and antivenin is certainly a chemical one, being likened bj' Kyes^- to that of strong acids upon strong bases. The serum of animals immunized to venoms contains precipitins for the proteins of these venoms, and, to some extent, for the serum proteins of the same species of snakes. These precipitins are highly but not absolutely specific, and they bear no exact quantitative rela- tion to the other antibodies present in the same sera.^^ As is well known, snakes are nearly or quite insusceptible to snake venom. Cunningham^* found that serum of cobras was devoid of antitoxic property, so the immunity of snakes must be ascribed to an absence of cell receptors in their tissues, with which their venom am- boceptor can combine. The reputed immunity of the mongoose and hedgehog depends partly on a relatively low susceptibility, but prob- ably more on the agility of the mongoose and the defensive spines of the hedgehog. Platypus Venom. — The only mammal with a venomous secretion is that strange freak, the duck-billed platypus {Omithorhynchus paradoxus). The males have a hollow movable spur on each hind foot, communicating like a fang with the venom gland, which secretes a venom with properties resembling the venoms of the Australian snakes, but much weaker. Scorpion Poison " This poison is secreted by a pair of specialized glands in the pos- terior segment of the elongated abdomen, surrounded by a firm cap- sule with a sharp apex through which the poison is discharged. Its effect on man is usually confined to local pain, swelling, and occa- sionally phlegmonous inflammation with constitutional symptoms after bites from the largest species. In Africa a large scorpion {An- drocionus) exists, that is reputed frequently to cause fatal poisoning, especially in children. Manchuriaii scorpions {Buihus martcnsi, Karchi) seem to be less toxic than this or Mexican scorpions {Centi'urus «i Univ. of Penn. Med. Bull, 1901 (17), 154; Jour. Expcr. Med., 1906 (8), 614. «- liorl. kliii. Woch., 190 t (11), 494. *^ See lloussay and Negrcte, Hev. inst. bact. Buenos Aires, 1918 (1), 15. «' Nature, 1S9() (55), 139. *"* A comi)l('to discussion of the literature on i)ois()nous invertebrates, etc., is given by v. Fiirth, " N'crglcicliende chemische IMiysioldgic dcr niedcren Tiere," Jena, 190.'}; and by Faust, "Die (icrisclicn (Jifte," liraunschweig, 190(). Con- cerning scorjjions see Kubota, .Jour. I'li.irriiacol., 191S (11), 447. SCORPION POISON 151 exlicauda, Wood). In Korea, however, of 81 cases collected by Mori,'^ four were fatal. The majority of serious results following scorpion bites, as well as bites of poisonous insects to be considered later, are, however, due to infection of the wound, which occurs readily because of local necrosis and hemorrhages, and also because of the unfavorable conditions existing in tropical climates. Apparently these bites favor local infection much as do those of vipers. When general symptoms do occur, they are described as resembling strychnine poisoning, with trismus, stiffness of the neck and eventu- ally of the respiratory muscles, which seems to be the chief cause of death (Cavorez). Thompson," however, observed only seldom severe symptoms, consisting of general paralysis that passed off in a few hours. Most experimenters with scorpion poison describe it as chiefly a nerve-tissue poison, and it also seems to act as a hemolysin and ag- glutinin (Bellesme and Sanarelli), but Todd*^^ found it without ac- tion on corpuscles and not capable of combining with nervous tissues. Houssay^^ states that scorpion poison (B. quinquestriatus) is above all a muscular poison of the veratrine type, and a powerful peripheral excitant of the salivary and lachrymal secretions. Calmette^° gives the lethal dose for a guinea-pig as 0.5 milligram, while Phisalix and Varigny put it at 0.1 milhgram and state that scorpion blood is also poisonous. Wilson^ ^ found the toxicity of the venom equal to 0.1 gram per million, that is, one gram of poison will kill 10,000,000 grams of guinea-pig, ^^ hence it is much stronger than cobra venom. It is quite stable, and keeps many months in an ice chest; is not affected by heating to 100° for a brief period, but is de- stroyed after 12 or 13 minutes' heating. The active constituents are precipitated by saturating with ammonium sulphate, or by an excess of alcohol. They are destroyed by either pepsin or trypsin (Kubota).^^ The average amount of toxin in an Egyptian scorpion {Buthus quin- questriatus) is sufficient to kill about 35 kilos, which agrees with the fact that fatal poisoning by this scorpion is rare in adults, but reaches 60 per cent, in children. The venom is harmless when taken into the stomach, and is said to be made inactive by ammonia, calcium hypo- chlorite, and peroxide of hydrogen. Calmette claims that antivenin for cobra in part neutralizes scorpion poison, a statement which could not be corroborated by Todd, who succeeded, however, in preparing an efficient antiserum by immunizing horses with scorpion venom. ^^ 5^ Korean Med. Soc. Jour. (Chosen Igakukai-Zasshi), 1917, p. 47. " Proc. Acad. Nat. Sci. of Philadelphia, 1886, p. 299. «8 Jour, of Hygiene, 1909 (9), 69. "Jour, phvsiol. path, gon., 1919 (18), 305. '"> Ann. Inst. Pasteur, 1895 (9), 232. '' Records of Egyptian Gov't., School of Med., 1904; abst. in Jour, of Phvsiol., 1904 (31), p. xlviii. '^ Exactl}^ the same toxicity is shown by Korean scorpions (Mori).''^ " A successful serum has also been prepared in Brazil (see Brazil-Medico, 1918 (32), 161). 152 PHYTOTOXINS AND ZOOTOXINS Houssay^* also describes the antiscorpion serum as strictly specific. A large number of naturalists and raconteurs have furnished interest- ing tales of suicide by scorpions, which are more than improbable in the light of our present knowledge concerning natural immunity. Many animals seem to possess more or less immunity to scorpions (Wilson) J especially such wild animals as are much exposed to them. Spider Poison The poison apparatus of the spiders consists of two long pouches lying in the thorax and extending into the jaws, at the apex of which the poison is discharged. Some of the larger members of the family are very poisonous, e. g., the Malmignatte (Latrodectes tredecim- guttatas), of the vicinity of the lower Volga in southern Russia, is said to have destroyed 70,000 cattle in one year, the bite being fatal in 12 per cent, of all cases, although rarely killing man. Other members of this species in Chili, Madagascar, and other countries are not much less venomous. Kobert^'* has studied the poison of Mal- mignatte and found it distributed throughout the body of the spider, even in the eggs, and resembling in nature the snake venoms. It is destroyed by heating, and seems to be of protein nature; the chief effect is upon the nervous system and heart. "^ A number of common spiders investigated by Kobert were ap- parently not poisonous for mammals, except the "cross spider" (Epe- ira diadema) , which has since been thoroughly studied by him and by Sachs.''** Walbum" states that the chief poison of these spiders is found in the ovaries, the salivary poison being much weaker, and the hemolysin is found chiefly in the albumin fraction. Epeiratoxin re- sembles the snake venoms strikingly, according to Sachs, for it con- tains a powerful hemolysin which he calls "arachnolysin," acting very differently with different sorts of blood, and destroyed by heating at 70°-72° for forty minutes, and it behaves with lecithin and cholesterol like cobra venom. ^^ By imnmnizing a guinea-pig Sachs succeeded in securing an antitoxin of some strength. The agglutinin is quite dis- tinct from the hemolysin. ^^ Only such blood is hemolyzed as is able to bind the poison in the stroma of the red corpuscles. The discovery of this hemolysin explains Robert's observation of hemoglobin, methemoglobin, etc., in the urine of persons bitten by spiders. Sjiider hemolysins have been studied extensively by Houssay,**" who finds '^ "BeitrJige zur Kentnisse der Giftspinnen," Stuttgart. 1901. "• In western America and South America is foiind a sj>ider {Latrodectes uiae- lans) the bite of wliich is capable of causing severe spasm of the ah(h)minal musch\s. according Ui At wood (Southern Californ. Pract., \'ols. 10, 12 and 10). Kellogg and (4)lcman (.lour, of Para>itoI., 1915 (1), 107), found extracts of the i)oison glands of this spider to he iiitililv toxic. '« Ilofmeister's Heitr.. 1902 (2), 125. " Zeit. immunitiit., 1915 (2:5), 02."^. '*Pini, 1! I'oliciinico (Se/,. Med.), 1909 (10), 20S. '"v. S/.ily. Zeit. Immunitat.. 1910 (5), •^S(). «»Comp. Rend. Soc. Biol., 1910 (79), OaS. INSECT POISONS 15:^ that spiders without liciiiolNsius poison flics oxactly as those with hemolysins. Von Fiirth considers that the bite of the historically famous Italian tarantula is able to cause no more than local inflammation, and Ko- l)ert found that the entire extract of six Russian tarantulas (which are supposed to be more poisonous than the Italian) caused no symp- toms when injected into a cat. An antitoxin is said to have been secured against the Russian tarantula.^-'" In all probal)ility the other poisonous spiders possess toxic sub- stances allied to those of the venoms, with hemolytic, agglutinative, and neurotoxic products, Sachs' studies indicating the general sim- ilarity of all tlie zootoxins. Centipedes Undoubtedh' the severity of centipede poisoning has been greatly exaggerated, the results being usually limited to local inflammation, frequently spreading some distance in an erysipelas-like manner. An authentic case of fatal poisoning of a child four years old by a centipede (Scolopendra heros) has been reported from Texas by G. Linceicum,*^ death resulting five to six hours after the bite was re- ceived. Besides the local pain and inflammation, vomiting w^as marked, occurring also in five other non-fatal cases. Centipedes secrete their poison in relatively large glands, which discharge at the apices of a pair of specialized claws that take the place of the first pair of legs. The nature of this poison seems not to have been investigated. Numerous chemical substances are de- scribed as secreted bj' other glands of these animals, including prus- sic acid and a camphor-like matter (see v. Fiirth). Bee Poison Bee poison has been better studied than mcst insect poisons, begin- ning with the work of Paul Bert (1865). It is secreted by the glands into a small poison sac, and stored up until ejected. Cloez found that bee poison was precipitated by ammonia, tannin, and platinic chloride, and Langer proved it to be a non-volatile organic base. As excreted, it is acid, contains 30 per cent, of solids, and one honey-bee secretes 0.0003-0.0004 gm. It contains formic acid and much pro- tein, but it has been stated that the poison is protein-free, and is not destroyed by heat (100°), weak acids, or alkalies. On the other hand, it is said to be destroj-ed by proteolytic enz3'mes, which would indicate that it is of protein nature. Arthus**- believes the evidence indicates that the bee venom is a proteotoxin. How^ever, there are many points of resemblance between the effects of insect stings and the local «"« Konstanzoff, Russky Wratsch., 1907, Xo. 17. " .\mer. Jour. Med. Sci., 1866 (52), 575. 82 Jour. Pharm. Chim., 1919 (20), 41. 154 PHYTOTOXINS AND ZOOTOXINS effects of histamine injection.*^ Hemolysis is produced both in vitro and in vivo with all sorts of blood, but to very different degrees, thus resembling spider toxin. The hemolytic action is greatly increased by the presence of lecithin, forming a toxolecithid like "cobra lecithid."^^ Locally bee poison causes necrosis, with marked h3'-peremia and edema. A 4500 gm. dog was killed by intravenous injection of 6 c.c. of a 1.5 per cent, solution of pure poison (Langer).^^ Immunity is undoubtedly possible, for bee-keepers frequently show a great decrease in susceptibility. On the other hand, abnormally great susceptibility is frequently seen, some cases of fatal poisoning having been observed. ^^ Dold***^" was unable to secure experimental immunity to bee poison. Wasps and Hornets presumably produce poison-s similar to those of the bees. A study by Bertarelli and Tedeschi*'' establishes this for a species of wasp (Vespa crabro L.). Ants also produce formic acid, a fact so well known that it has come to be considered that this is the source of their toxicity.*^" Von Fiirth, however, sug- gests the probability that ant poison, like that of the bees, owes its chief effects to other more complex, unknown poisons.^* Lice. — Persons bitten by large numbers of lice maj'- exhibit a distinct intoxica- tion, accompanied by an eruption resembling measles.^* The nature of the poison is not known, but it does not produce a severe local urticaria like the sting of bees and wasps. Poisons of Dermal Glands of Toads and Salamanders It has been known for centuries that toads produce poisonous sub- stances, Pare in 1575 having discoursed interestingly, if inaccurately, on this topic. Numerous studies have been made of these poisons, which are secreted by the dermal glands and therefore cannot be used for poisoning either prey or enemies (except those that feed upon them) ; the most extensive study being that of Faust. ^° He isolated two constituents, apparently the same in different species of toads; one, which he called bufotalin, is very active, resembling the digitalis group; the other, hufonin, is much less active. Bufonin is neutral in reaction, soluble in warm alcohol, but slightly in cold. Analj^sis in- 83 See Eppinger, Wien. klin. Woch., 1913 (26), 1413; SoUmann and Pilcher, Jour. Pharm., 1917 (9), 309. 8^ Morgenroth and Carpi, Berl. klin. Woch., 1906 (43), 1424. 85 Arch. exp. Path. u. Pharm., 1896 (38), 381; Arch, internat. Pharniac. et Ther., 1899 (6), 181. 8» irospitaLstideude, 1905, No. 27. 8«" Zeit. Iiniiuinitiit., 1917 (26), 284. 8' Cent. f. Jiakt., 1913 (68), 309. 8'"The sting of nettles is said to be due to formic acid. (See Dobbin, Xatur Sept. 18, 1919.) 88 An attempt by Barratt ( Ann. Trop. Mi-d. and I'arasitol., 1910 (4), 177) to obtain a poison from culex mosciuitos was unsuccessful. The l)odies of "black files" contain an aiitive poi.son that could not !>(> ithMitified by Stokes (Jour. Cut. Di.s., 191 I (32), 830), Ijcyond thai it is insolul)l(' in alcohol, which does not in- activate it, and tiiat it is destroyed by trypsin. 8" IJirschleldcr and Moore, .Vrcli. Int.'lMed., 1919 (23), 119. »» Arch. f. exp. Path. u. Pharm., 1902 (47), 279. Complete Ijihliugraphy and review. POISONS OF TOADS 155 dicates an empirical I'oiiuuhi of C34H54O2. It probably is the cause of the milky appearance of the dermal secretion. Bufotahn seems to be C34H46O10, is acid in reaction, soluble in chloroform and alcohol, but not in petroleum ether. Subcutaneous injection of 2.G mg. bufo- talin killed a dog (weighing 4 kg.) in four to five hours; given by mouth it causes much vomiting and diarrhea, so that large doses are not fatal. It causes much local irritation when applied to mucous membranes, but produces no marked change at the site of injection. The effects on the circulation resemble in all respects those of the digi- talis group; bufonin acting similarly but much weaker than bufotalin. Bufotalin seems to be derived from bufonin by oxidation, and the latter is quite similar to cholesterol, apparently having the following formula: HO-H26C17-C17H26-OH. An important consideration is that Faust has also isolated from the venom of cobra and crotalus, poisons which seem related to these toad poisons, the cobra poison being assigned an empirical formula of Ci7H260io, and the crotalus poison C34H54O21. Fiihner^^ considers bufotalin to be more closely related to the saponins. Phisalix and Betrand^- have found poison in the blood of toads similar to that of the glands. The hemolytic property observed by Pugliese^^ may be due to the acidity of the dermal secretion. The poisons of chfferent species seem to be quite the same in all (Faust). From the dermalsecretion of the large tropical toad, Bufo agiia, Abel and Macht^* have isolated two distinct active substances; one identical with epinephrine, which constitutes nearly seven per cent, of the crude venom; the other, which makes up 36 per cent., is called hvfagin, has a composition indicated by the formula C1SH24O0, and therefore is presumably related to the rest of this group which arises from choles- terol. In physiological action bufagin resembles digitalis, and it is ex- tremely active. The toad is relatively immune to bufagin, but not at all to the epinephrine. A Chinese drug derived from toad skins has been found to contain similar ingredients (Shimizu^^), as well as a substance resembling picrotoxin in action. Salamanders also produce poisonous secretions in their dermal glands, which have been studied especially by Faust,^" and earlier by Zalesky,^^ who isolated an organic base which he named samandarin. Faust describes samandarin as first stimulating and then paralyzing the automatic centers in the medulla. The poison resembles the alkaloids, having the formula C26H40N2O, and produces death in doses of 0.7-0.9 mg. per kilo (dogs) with respiratory failure. Immuniza- tion of rabbits was practically impossible. A second alkaloid, samandaridin (C-.uH.nXO) is also present in even greater quantities than the samandarin, and differs only in being weaker. Frogs also have similar poisons in their skins, extracts of Rana esculenta skin 91 Arch. exp. Path. u. Pharm., 1910 (63), 374. 9- Arch. d. physiol. norm, et path., 1893 (5), 511. 9' Archivo di farm. e. terap., 1894 (2), 321; Arch, ital de Biol., 1895 (22), 79. "Jour. Amer. Med. Assoc. 1911 (5lJ), 1531; Jour, of Pharm., 1912 (3), 319. 95 Jour. Pharmacol., 1916 (8), 347. 9«Arch. e.xper. Path. u. Pharm. 1898 (41), 229 (literature); 1900 (43), 84. 9' Hoppe-Seyler's Med. Chem. Untersuch., 1866, p. 85. 156 PHYrOTOXJX.s AXD ZOOTOXINS being highly toxic. '^* The dermal secretions of most of the an?phibians are poison- ous, not only for mammals, but also for reptiles, and in large doses for the animals producing them (Phisalixj."'-' Bert' and also Dutartre- have described a digitalis- like poison in the secretion of the dermal glands of frogs. It is evident that all these poisons are quite distinct from the venoms, and from the true toxins, apparenth' being simple chemical compounds not related to the proteins and not capable of causing im- munization. Poisonous Fish ^ There are numerous fish, especiall}' in tropical waters, which defend themselves by injecting poisons into their enemies. This is ac- complished by spines, to which are attached poison glands.^ Dunbar- Brunton^ has described two such fish {Trachinis draco and Scorpoenn scorpha) of Mediterranean waters. Wounds by these spines cause in animals intense local irritation and edema and paralj'sis of the part, followed by gangrene about the site of the wound; in fatal poisoning death occurs in from one to sixteen hours, with general paralysis. The sufferings of persons so poisoned are said to be extreme, and death may occur either directly from the poison or later from sci:)sis following the local gangrene. Presumabl}' this poison is not dissimilar to that of the snakes; it probably is not an alkaloid, as Dunbar- Brunton suggests. It affects chiefly the heart, according to Pohl," and contains a hemolytic principle which behaves like the venom hemolysins in that it is activated by serum (Evan)." Several other fish secrete poison in glands attached to long spines, one of the most poisonous being Synanceia brachio, which is known to have caused fatal intoxication in several instances. Only the Murw- nidce seem capable of poisoning by biting; they have a well-developed poison apparatus on the gums, but nothing is known concerning the poisons they produce. Many fish develop poisonous decomposition products remarkably soon after death, especially in tropical climates, so that a fish that is perfectly wholesome if eaten immediately after being caught may be very poisonous if kept but a few hours. There is a decided differ- ence in fish of different varieties in this respect, so that some cannot be safely marketed. Some of the poisonous products of the decomposition of fish seem to be, early products of protein cU^avage, of liigli inoh'cular »» Caspari and Loewv, Med. Klinik, 19II (7), 1204. "9 Jour. Phvs. et Path, gen., 1910 (12), 325. ' Compt. Rend. Soc. Biol., 1885, p. 524. 2 Ibid., 1890, p. 199. ^ Full discu.ssion and literature given by Faust, "Tierische Giftc," p. 134. * For a list of fLsh with poison glands see Pawlowskv, Zool. Jahrb., 1912 (31), 529. » Lancet, 189(i (ii), (iOO. • Prager med. Woch., 1893 (18), 31. ' British Meeculation in this direction is not justifiable. The attempts to produce antitoxin against cantharidin have not yielded con- vincing results/" nor against einnephrine.^' De Angelis^* claimed that he had produced specific precipitins for various natural and sj'nthetic dyes, but this work has, as was to be expected, failed of confirmation.^^ Elschnig and Salus^" state that melanin from the eye is antigenic, producing complement-fixing antibodies specific for melanin but not for the si)ecies. Woods^' has corroborated this and also demonstrated anaphylactic sensitization. We know too little concerning the composition of melanin to interpret these observations. '^ In general terms, therefore, antigens ai'e protein molecules, and the reactions of immunity are reactions against proteins foreign to the body of the host, and manifested by the presence in the blood of the reacting animal of substances which combine with and cause recognizable changes in the foreign protein.^- These changes are recognized in many ways, such as precipitation, agglutination, com- plement-fixation, etc., and the question at once arises as to whether these different manifestations depend each upon a separate antibody, or if several or all of them are not caused by a single antibody, the action of which is indicated by the different reactions which are made manifest by different procedures in each case.^^ This question will be discussed further in later paragraphs. Knowing that the antigens are merely foreign proteins which have been introduced into the body of an animal, there naturalh' occurs the thought that the animal body is continually receiving in its food foreign proteins, and against which it defends itself in the alimentarj- canal by enzymatic action, which disintegrates these proteins until they have lost their colloidal character.^-* Logically following this comes the idea that perhaps the reactions of immunitj' are simply the same or similar disintegrative enzymatic actions, carried on within the blood and tissues to protect the body in the same way against foreign proteins which the alimentary digestive apparatus has not had the opportunity to destroy. This conception of the nature of im- mune reactions to antigens has been especially advanced and in- « See Manoilov, Wicn. klin. Woch., 1912 (25), 1701. « Champy, Compt. Rend. Soc. Biol., 1907 (G2), 1128. " PoUak, Zeit. physiol. Chem., 1910 (GS), C9. " Ann. di Ig. Spcrim., 1909 (19), 33. " Takemura, Zeit. Immunitiit., 1910 (5), 697. *" Graefe's Arch., 1911 (79), 428. " Jour. Immunol., 1918 (3), 75. ^^ Drew lias found no evidence of antibody formation l\v immunizing molluscs and ecliinodcrms (Jour, of Hyg., 1911 (11), 188), from which he concludes that the reaction to foreign jjroteins is not a universal projierty of protoplasm; a sweep- ing generalization which requires more extensive invest iagt ion for its establish- ment. C'antacuzene (C'oni])t. Kend. Soc. Hiol., 1913 (74), 111) obtained precipitins by immunizing J'/i(illusiii nunnilhita with mammalian l)K)i)d, l)ut no hemolysins with tiiis or AjthrodUe (iculrald and Elcdone inoschata. "See Dean, Lancet, Jan. 13, 1917. " Carrel and Ingebrigsten (Jour. Exp. Med., 1912 (15). 287) have found that tissues growing in vitro witli foreign blood produce iiemolytic antil>odies for that blood, indicating that isolated cells can react to antigens l)v anlilxuly production. SPECIFICITY 1G5 vcstifiiatecl by \'ic'tor (\ \'auji;liair'' and l)is co-workers, and l)y ImiiII Abdcrhalden, who has demonstrated in various ways an increased proteolytic power in the blood of animals which have received pa- renteral injections of foreign proteins.^" Thus, if the antiserum re- acts on the specific proteins within a dialyzin receptors of a cell are coni])ined by toxin molecules that have lost their toxophore *^ Ehrlich has used certain diaKrams to illustrate these various groups and their relations to the cells and to one another, which arc generally used in exi)lairiiiig his thef)ry. From a teaching standpoint they have seemed to be umicsiralilc, in that the student soon comes to ascril)e piiysical properties and ai)i)caranc('s to what should he considered as oliemical coml)inations. The toxopliorc grouj) Ix'comes "the hlack fringed end of the toxin." etc. To one accustomed to thinking in chemical terms there is no diiiiculty in following the literature and understanding the reactions as chemical reactions, which they are. TOXINS AND ANTITOXINS 173 p;r()up (toxoid is the muuc given to such altei-ed toxins), tiie cell can- not then be injured by the corresponding; active toxin, showing that the toxin nuist first become united to a cell receptor by its haplophorc group before the toxophore group can cause an injury. Animals that are naturally immune to toxins may owe their im- munity to the fact that their vital tissues contain no substances with a chemical affinity for the toxin, and hence the toxin cannot unite with them to cause harm. (In Ehrlich's terminology, the cells con- tain no receptors for the toxin.) The toxin may not combine with any tissue element at all in such immune animals, and may circulate for some time harmlessly in the blood, or it ma}^ combine with some organ where it does little harm, e. g., tetanus toxin is said to combine chiefly in the liver of some animals, and therefore it does not harm their nervous system. According to this theory, the antitoxin consists of cell receptors that have been produced in excess and secreted by the cells into the blood. In the blood they combine with any toxin that 'may have been intro- duced, and by saturating its affinities render it incapable of uniting with the cells. As the toxin harms cells only after it has been chemi- cally united to them, it is rendered harmless when its affinities for the cell (the haptophore groups) are saturated by cell receptors in the blood stream. The process of immunization consists in injuring the body cells to such a degree that they are stimulated to regenerate the receptor groups with which the toxin combines; these receptor groups are produced in excess, and not only replace those combined by the toxins, but the excessive groups escape free into the blood. Hence the serum of an immunized animal is antitoxic because it contains free cell receptors that can unite with the toxin. An important point is that the receptors liberated by all animals which have been immunized wdth a given toxin seem to be the same — horse serum, or sheep serum, or goat serum will neutralize diphtheria toxin if the animals have been made immune to this toxin; and, furthermore, their serum when intro- duced into the body of an entirely different animal, e. {7., a guinea-pig, will neutralize diphtheria toxin within its body. Equally important is the fact that the antitoxin for one toxin will not neutralize any other toxin; e. g., diphtheria antitoxin will not neutralize tetanus toxin, or conversely. This means that diphtheria toxin is attached to chemical groups of the body cells (receptors) which are quite difTerent from the groups to which tetanus toxin unites, and hence different receptors are thrown out in immunizing against each. True toxins have been designated monovalent antigens^ since animals immunized with a puri- fied toxin produce onl}^ the one antibody, the antitoxin, whereas many protein antigens produce precipitins, Ij^sins, agglutinins and other antibodies; presumably this is because of the relatively small size of the toxin molecule, which limits the number of its antigenic radicals (Pick). Or it may well be that the immune body for antitoxin is 174 CHEMISTRY OF THE IMMUNITY REACTIONS quite different from the antibody or antibodies resulting from immuni- zation with non-toxic protein antigens, for there is some reason to believe that the several types of reactions that may be accomplished with the serum of animals immunized to foreign proteins or cells all depend on one single antibody, which accomplishes the destruction of the antigen by sensitizing it to the enzymes of the blood and tissues. The neutralization of toxin by antitoxin is believed by many in- vestigators to be a chemical process, which occurs as well in the test- tube as in the body. It seems to occur according to the laws of definite proportion, sl given amount of antitoxin neutralizing a proportionate amount of toxin under equal conditions (hence the toxin is not de- stroyed by antitoxin through a ferment action, as was at first suggested). .Neither the toxin nor the antitoxin is destroj^ed in the process of neu- tralization, as has been proved by suitable experiments, but they appear /to be chemically united to each other, as any two large molecules may be. Pick and Schwarz believe that the union of toxin and antitoxin takes place in two steps — first, colloidal adsorption, and then the specific reaction. ^^ There is some question as to whether the union with anti- toxin completes the neutralization of the toxin, or whether there is then necessary a further destruction of the toxin in the body. But whether necessary or not, such further destruction does take place. Neutrali- j zation occurs more rapidly under the influence of warmth, and more slowly in the cold; and it is more rapid in concentrated than in dilute solutions, just as with ordinary chemical reactions. It is said that it requires two hours for tetanus toxin to be completelj^ combined with the corresponding quantity of antitoxin at 37°. According to Arrhe- nius and Madsen, reaction of antitoxin upon toxin is accompanied by the liberation of much heat — 6600 calories per gram molecule, or about half as much as is set free by the action of a strong acid upon a strong^ base.^^ Union of toxin and antitoxin causes no change in the surface tension of the fluid in which the reaction occurs (Zunz),*^ and the neutral toxin-antitoxin compound (diphtheria) is not absorbed by ani- mal charcoal, which absorbs each of the constituents when free. The physico-chemical studies of the reaction between tetanolysin and its- antibody gave results which led Arrhenius to conclude that in the reaction there are formed from one molecule of toxin and one molecule of antitoxin, two molecules of the reaction products (analogous to the reaction between alcohol and acid which yields one molecule of ester and one of water). In general, the union of toxin and antitoxin is "Also von Krogh (Zeit. f. Hyg., 1911 (68), 251). Bordet, Biltz, and others look upon the neutralization of toxin as an adsorption process entirely. g ** Literature of chemical and jihysical reactions of toxin and antitoxin given ■ by Zangger, Cent. f. Bakt. (ref.), 1905 (80), 238; Arrhenius, "Ininiuno-cheni- listry," 1907 and "(Quantitative Laws in Biological Chemistry," London, 1915;. I also review in Zeit. Chemother., Kef., 1914 (3), 157; Oppenheimer and IMichaelis, iHandbuc'li der Biochemie, Vol. II (1). ^ 85 Bull. Acad. Royal i\Ied. Belg., 1911; also Bertolini, Biochem. Zeit., 1910 (28), 60. ANTITOXINS 175 dissociated by acids. ''^ On dilution of a neutral toxin-antitoxin mixture, a certain amount of dissociation seems to occur, but there is opposition to the view that the law of mass action applies to the re- action between toxin and antitoxin. If toxin is added to antitoxin in several fractions, with some interval of time between each addition, the final mixture is much more toxic than if the same quantities of toxin and antitoxin were put together at one time. This phenomenon is commonly referred to as the Danysz effect, and indicates that the toxin-antitoxin union is physical rather than chemical, for it seems to be quite analogous to such a phenomenon as the taking up of more dye by several pieces of blotting paper added in series to a dye solution, than by the same amount of paper added in one piece. There is no relation between antitoxins and enzymes. The anti- toxin acts quantitatively, and produces no detectable alteration in the toxin, or in any other substance, as far as we know. It also has but one functioning group (haptophore), the one with which it combines with the toxin; whereas both toxins and enzymes seem to have two functionating groups, one which unites with the cell or substance that is to be attacked, the other which produces the chemical changes. But there is evidence that union with antitoxin or fixed receptors pre- pares the toxin for its disintegration, which, presumably, is then accomplished by enzymatic action as with other antigens. Chemical Nature of Antitoxins*^ This is as entirely unknown as is the nature of the toxins. In- vestigation of antitoxic serum (principally diphtheria antitoxin) has shown that the antitoxic properties are closely related to the serum globulin, which, however, by no means proves that antitoxin is serum globulin or any other sort of protein. According to Ehrlich's theory, antitoxin consists of free cell receptors, and these receptors are pre- sumably simple chemical groups which may be but a part of a larger molecule, or they may be entire protein molecules. In any event they behave as colloids; moving toward the cathode in an electrical field, ^^ diffusing little or not at all, their reaction curve resembling more an absorption curve than the reaction curves of crystalloids, and being influenced by all conditions that influence colloids. Whether the receptor groups are secreted in a free condition in antitoxin*^ for- mation, or combined in a large molecule, is unknown. By saturating serum with magnesium sulphate, or half saturation with ammonium sulphate, three chief groups of proteins can be pre- 8« Morgenroth and Ascher, Cent. f. Bakt., 1911 (59), 510. '^ Review and bibliography given by Crawford and Foster, Amer. Jour. Phar- macy, 1918 (90), 765. 88 According to Field and Teague (Jour. Exper. Med., 1907 (9), 86) both \ toxin and antitoxin move towards the cathode, which is opposed to the theory that this reaction is simply one of oppositely charged colloids. (See also Bechhold, I Munch, med. Woch., 1907 (54), 1921.) ' 176 CHEMISTRY OF THE IMMUNITY REACTIONS cipitated and isolated. ^^ These are fibrinogen, euglohulin (true glob- ulin), and -pseudo-globulin (soluble in water). Pick^" found that the precipitate obtained by 36 per cent, volume saturation with am- monium sulphate contained no antitoxin; the antitoxin came down in the precipitate obtained on raising the strength from above 38 per cent, to 46 per cent." According to Pick, in horse serum the anti- toxin is associated with the pseudo-globulin, ^^ and Gibson and Banzhaf found that the blood of horses immunized to either diphtheria or tetanus toxin shows a marked increase (40 to 114 per cent.) in serum globulin, varying somewhat according to the antitoxin content, the more solu- ble globulins being most increased. At the same time the serum al- bumin and euglobulin content decreases in proportion, while the fibrinogen shows no characteristic alterations.^^ Mej^er^"* and his colleagues, however, find in their study of the blood proteins during immunization, that the proportion of globulins increases according to the severity of the intoxication, and not in any definite relation to the degree of immunity or antitoxin production. The average antitoxic horse serum contains 12 per cent, albumin, 78 per cent, of soluble glob- ulin containing antitoxin, 10 per cent, euglobulin. By heating 12 hours at 57° a considerable part of the soluble globulin becomes insoluble, without a corresponding loss of antitoxin (Banzhaf) . The relation of antitoxins to proteins has also been investigated by permitting digestive enzymes to act on antitoxic serum. Pick di- gested the antitoxin-containing globulin of horse serum for several days with trypsin; after five days, when part of the protein was still not digested, the antitoxin was but little impaired in strength; after nine days, when most of the protein was digested, the antitoxin had lost two-thirds of its strength. This indicates a considerable resist- ance of antitoxin to trypsin, but also shows that it is affected in much the same way as the globulin (which is itself very resistant to trypsin) and therefore is presumably of similar nature. Antitoxin seemed to be much more rapidly destroyed by pepsin-HCl digestion than by trypsin, in which respect it again resembles the serum globulin. ^^ *^ See rcsum6 by Gibson, Jour. Biol. Chem., 1905 (1), 161; Gibson and Banzhaf, Jour. Exper. Med., 1910 (12), 411. 90 Hofmeister's Beitr., 1901 (1), 351. 91 Gibson and Collins (Jour. Biol. Chem., 1907 (3), 233) question the reliabil- ity of some of Pick's results, and repudiate the salt fractionation method of clas- sifying proteins. 92 Miss Homer found tetanus and diphtheria antitoxin associated with the pseudoKlobulins, but the antibodies in antidj^sentcry and antimeningococcus serum were chiefly in the euglobulin fraction (Jour. Physiol., 1918 (52), xxxiii). 93 During immunization the antitryplic power of the horse serum increases with the antitoxin increase (Krause and Klug. Berl. klin. Woch., 190S (45), 1454.) 9' Jour. Exp. Med., 1916 (24), 515; 1917 (25), 231; Jour. Infect. Dis., 1918 (22), 1. 9'' P,erg and Kclser (Jour. Agric. Ri-s., 1918 (13), 471) found that trypsin and pepsin destroy tlie antitoxin and scrum proteins at about the same rate, and their failure to observn; "significant cliemical ciuinges" in the proteins of serum acted upon by weak acid or alkali that slowly inactivates antitoxin, does not seem to warrant tlieir deduction that antitoxin is non-protein. See also Crawford and Andrus, Amer. Jour. Pharm., 1917 (89), 158. ANTITOXINS 111 In favor of tho view that antitoxin is a definite protein body is also the fact that it is not carried down in indifferent precipitates, as are the enzymes, but conies down always in a certain fraction of the protein precipitates, e. g., we can precipitate all the serum albumin from an antitoxic serum, and it does not carry down with it any of the antitoxin. Another important point has been brought out by Ai-rhenius and IMadsen,^''' who determined approximate^^ the molecu- lar weight of toxin and antitoxin by means of their rate of diffu- sion, and found that the toxin (diphtheria toxin and tetanolysin) diffused ten or more times as rapidh' as the corresponding antitoxin. Gelatin filters also hold back antitoxin and let toxin pass through, and toxins diffuse into cells which seem to be impermeable for the anti- toxin. This indicates that the antitoxin molecules are much larger than the toxin molecules, agreeing with the idea that antitoxin is of protein nature and that toxin either is not protein or is smaller than most protein molecules. Taken altogether, the evidence indicates a closer resemblance of antitoxins to proteins than has been show^n for the toxins, and all attempts to separate antitoxins from proteins have so far failed. Antitoxins are retained to greater or less extent by porcelain filters, do not pass through dialyzing membranes readily, and are in general easily destroyed by chemical and physical agencies, although much less so than are most toxins. Heating to 60°-70° injures, and boiling quickly destroys them, although like the enzymes and the pro- teins, they resist dry heat to 140°, and also extremely low temperature, without change. Putrefaction of the serum destroys the antitoxins (Brieger)." They can- be preserved for a very long time when dried completely, but in the serum they gradually disappear, especially if exposed to light and au*. Acids and alkahes destroy antitoxins, acids being the more harmful in low concentrations. Like the enzymes, antitoxins are destroyed hj ultra-violet ra3-s. They are destroyed in the alimentar}' tract, without appreciable absorption, except in the case of new-born animals sucking mothers whose blood and milk contain antitoxin. ^^ When subcutaneousl}'- injected, antitoxin soon disappears from the blood; part may be bound to the tissues, part may be destroyed, since only traces appear in the urine. It resists autolysis. ^^ " Festskrift Statens Serum Institut, 1902. " Behring states that tetanus antitoxin resists putrefaction. 98 Romer and Much, Jahrb. f. Kinderheilk., 1906 (63), 684; McCIintock and King (Jour. Infect. Dis., 1906 (3), 701) found appreciable absorption of antitoxin when digestion was impaired bj^ drugs. Full review of literature on transmission of antibodies from mother to offspring given bv Famulener, Jour. Infect. Dis., 1912 (10), 332; Heurlin, Arch. Mens. Obs. et Gvn., 1912 (1), 497. " Wolff-Eisner and Rosenbaum, Berl. klin. Woch., 1906 (43), 945. 12 178 CHEMISTRY OF THE IMMUNITY REACTIONS AGGLUTININS AND AGGLUTINATION^ The relation of agglutination of bacteria by the serum of immunized animals to their immunity is not known, for it is not established that agglutination helps in the defensive reaction. ^ Agglutinated bacteria seem not to be severelj' injured by the process, and can grow vigorously in agglutinative serum. Possibly agglutination favors phagoc3'tosis and lessens dissemination of the infecting organisms, but it is not generally considered that the influence on the com-se of infection is great. ^ Agglutination, therefore, may be looked upon as an incident in the infection, rather than as a definite method of resistance, and it is equally well produced by immunizing with foreign cells or any foreign protein masses of suitable size which contain soluble antigens. For the production of agglutination it is necessary that the cell contain an antigen (agglutinogen) which has an affinity for the specific constituent of the serum, agglutinin. Normal serum may contain agglutinin;'* e. g., typhoid bacilli are sometimes agglutinated by normal serum, even when it is diluted thirty times, but by immunization this property can be greatly increased until agglutination may be obtained with dilutions as high as one to a million. Whether normal agglutinins are essential^ different from immune agglutinins is not known. ° Many protein solutions, especially extracts of plant tissues and leguminous seeds, cause marked non-specific hemagglutination." Likewise, bac- terial extracts may agglutinate red corpuscles.^ In immunization the agglutinogen, which is probably an intracellular protein, acts as a stimu- lator to the formation of the specific agglutinin. Hence, when we inject either extracts of cells or entire cells, we secure agglutinins, for the agglutinogens are liberated from the cells upon their disintegra- tion. In erythrocytes ^he agglutinogen seems to be in the stroma.^ We can obtain agglutinins against nearly all bacteria, including non-pathogenic forms, but in varying strengths. Agglutinins are found in the blood stream in the highest concentrations, but they are ^ Bibliography given by Miiller, Oppenheimer's Handbuch der Biochemie, 1909 (II (1), 592: Landsteiner, ibid., p. 428; Paltauf, Kolle and Wassermarin's Hand- buch., 1913 (II), 483. ^ Bull, however, would ascribe much importance to agglutination of bacteria for their removal from the circulation (Jour. Exj). Med., 1915 (22), 48-4). P\ijimoto (Jour. Immunol., 1919 (4), G7) also attributes to agghitinins the jKiwer to impair the glucolytic action of B. colt, but there is no evidence in his experiments that it is agglutinin rather than some other serum component that is responsible. On the otiicr hand Zironi (Atti accad. Lincei, 1917 (2G), 19) found that agglutination does not modify reproductive or biochemical activities of bacteria. ^ IJiaizot (0. R. Soc. Biol., 1918 (81), 350) states that it is possible to modify typhoid baciUi l)v treating them with nitric acid, hydroquinone or bj'' heat, so that they will jjrodiicf! immunitj' without producing agglutinins. * Even cold blooded animals may have normal agglutinins for bacteria and mammalian corjjuscles (.see Takeiiouchi, Jour. Inf. Dis., 1918 (23), 393, 415. *.See Andrejew, Arb. kaiserl. (iesuntlhtsamt., 1910 (33), 84. « Mendel, Arch. Fisiol., 1909 (7), KiS. ' Fiikuhara, Zeit. Immunitat., 1909 (2), 313. 8 Chyosa, Arch. f. Ilyg., 1910 (72), 191. AGGLUTINATION 17'.) also present in the various organs, and to greater or less extent in the other body fluids, excepting usually the spinal fluid (Greer and liecht).''* The place of their founation is unknown, but they can be formed by spleen tissue grown in artificial cultures.'" vSince bacteria contained within a collodion sac iinplantcfl in an animal give rise to the production of agglutinins, it is evident that the agglutinogens are diffusible to some extent, at least, through collodion. Old cul- tures of bacteria contain free agglutinogens, probably liberated from disintegrated cells, and filtrates of such cultures will neutralize ag- glutinins, showing both that the agglutinogens arc filtei-al)le, and that the reaction of agglutination is a chemical one and not dependent upon the presence of cells. Agglutinogens are said to pass through dialyzing membranes, while agglutinins do not, so it is evident that the agglutinogen is of smaller molecular dimensions than the agglutinin, just as toxin molecules are smaller than antitoxin molecules. Agglu- tinogens are not destroyed by formalin, heat, or ultraviolet rays in concentrations sufficient to kill the bacteria containing them.'' Just what constituent of the bacteria acts as the stimulus to the production of the agglutinin is unknown. Apparently', there are at least two bacterial substances with this property, one of which seems not to be a protein, since it is soluble in alcohol, gives no biuret reaction and resists temperatures up to 165°. The other gives all protein reactions, and is destroyed by heating to 62°. We consider, therefore, that there are two agglutinogens in the bacterial cell, one, thermostable, the other, thermolabile. The difference in the func- tion of these two agglutinogens is still a matter of dispute. Likewise, the question as to whether they occur in the membrane or within the bacterial cell is still open, but Craw found that the insoluble residue of crushed typhoid bacilli, after being washed free of all soluble con- stituents, was but slightly agglutinated by active serum; therefore, the agglutinogens are probably soluble intracellular substances. Stuber holds that bacterial agglutinogens are lipins.'^ Properties of Agglutinins. — Like most of the other immune substances, agglu- tinins are precipitated out of the serum in the globulin fraction. All attempts to separate them from proteins have been unsuccessful. Stark'^ found that trypsin does not attack the agglutinins readily, corresponding to the resistance of the serum globulins to this enzyme; alkaline papayotin solution destroys them slowly, while pepsin acts more rapidly. Alkalies are destructive even when quite dilute, while acids are much less harmful. The temperature resistance of agglu- tinins seems to be variable, plague agglutinin being destroyed at 56°, while purified typhoid agglutinin may resist 80°-90°; most agglutinin serums lose their activity at 60°-65°. The rate of reaction of agglutinins increases with the temperature, as long as this is not high enough to injure the reacting substances.^'* They are not precipitated by specific precipitins, but are readily absorbed by charcoal. 3 Jour. Infect. Dis., 1910 (7), 127. >" Pryzgode, Wien. klin. Woch., 1913 (26), 841. ^1 Stassano and Lematte, Compt. Rend. Acad. Sci., 1911 (152), 623. »2 Biochem. Zeit., 1916 (77), 388; also Bauer, Biochem. Zeit., 1917 (83), 120. 1' Inaug. Dissert., Wiirzburg, 1905. 1* Madsen, et al, Jour. Exper. Med., 1906 (8), 337. "^ 180 CHEMISTRY OF THE IMMUNITY REACTIONS The observations of Bond'^ suggest that they may become physically bound to other colloids within the body. The structure of the agglutinins (in the Ehrlich theory) is similar to that of the toxin; i. e., there is a haptophore group by which they combine with the aggluti- nogen, and a toxophore group by which they produce the changes that cause agglu- tination. The agglutinogen is probably related to the antitoxins in structure, having a single haptophore to unite with the agglutinin. By degeneration of the toxophorous group of the agglutinin, agghitinoids may be formed. It is believed that agglutinins are cell receptors, which have a group with a chemical affinity for the agglutinogen of the bacterial protoplasm, and also another group which brings about the agglutination. Thej' are, therefore, more complex than the simple receptors that unite with toxins, and are called receptors of the second order. According to Ohno^® the reaction of agglutinin and antigen is in constant propor- tions, and seems to be a chemical rather than a physical reaction. Coplans^^ finds this reaction associated with an increase in conductivitj- in the solutions, but whether this depends upon the agglutinin reaction itself, or upon associated processes, is questionable. Agglutinated bacteria can be again separated from one another by the action of organic and inorganic acids, alkalies, acid salts, and bj"- heating to 70'' or 75°. and after once being separated they cannot be reagglutinated by fresh serum. i* The Mechanism of Agglutination. — This has been a fruitful field of research, in which the application of physical chemistry has been very profitable. At first it was believed that the clumping was brought about by loss of motility, until it was found that non-motile bacilli were equally affected. Similarly, the hypothesis of adhesion of the fiagellse was disposed of. Gruber^^ and others supposed that a sticky substance, " glabrificin," was absorbed from the serum by the bacilli, which caused them to adhere on contact with one another; but this does not explain the flocking together of non-motile bacilli. Paltauf considered that the specific precipitin (see next section) pro- duced by immunization carried the bacilli down in the precipitate formed, and there is reason to believe that this reaction is of im- portance, but it does not explain all the facts of agglutination, nor is the relation between agglutinating and precipitating power of im- mune serums a constant one. In support of this hj^pothesis is the observation of Scheller^" that mixtures of typhoid bacilli and agglu- tinating serum lose their agglutinability by vigorous shaking, which may be interpreted as the result of disintegration of the agglutinating precipitate. Shaking of either bacteria or serum alone is without' effect. Neisser and Friedemann^^ found that if the bacterial cells were saturated with lead acetate, washed in water until all soluble lead was removed, and then treated with H2S, they were promptly agglutinated and precipitated, supporting other observations that indicate that precipitation within the bacterial cells can lead to agglutination. This sort of agglutination is related to the process of formation of "Brit. Med. Jour., June 14, 1919. " Philippine Jour. Sci., 1908 (3), 47. "Jour. Path, and liact., 1912 (17), 130. '8 Eisenberg and V^)lk, Zeit. f. Infektionskr., 1902 (40), 192. "^'8 For comploto bibliography, sec Craw, Jour, of Hygiene, 1905 (5), 113. 20 Cent. f. Bakt., 1910 (54). 150. 2' Miinch. nied. Wocli., 1904 (51). 4(15 and S27. AGGLUTINATION 181 coarse floceuli in soJiitioiis, and probably depends uj)()n alterations in surface tension. Bordet-^ made the important observation that agglutination does \ not occur if both the bacterial suspension and the agglutinating serum are dialyzed free from salts before^ mixing; but if, to such mixtures, a small amount of NaCl is added, agglutination and precipitation of the bacteria occur at once. This observation brought the phenomenon of bacterial agglutination into close relation with the precipitation of colloids by electrolytes, Bordet comparing it to the precipitation of particles of inorganic matter suspended in the fresh water of rivers that occurs when the fresh water meets the salt water of the ocean. He found that the agglutinin combined with the bacteria in the ab- sence of the salts, and the resulting compound was precipitated by the addition of minute amounts of electrolytes, ^'^ which alone did not precipitate or agglutinate the bacteria or the serum. This indicates that the agglutinins cause a change in the bacteria which brings them, under the same physical laws as the inorganic colloidal suspensions, which are characterized by being precipitated by the addition of traces of electrolytes."'* This precipitation is undoubted!}^ due to changes in solution tension and surface tension (see "Precipitation of Colloids, " introductory chapter). Before the agglutinin combines with the bacteria they behave like the colloidal solutions of organic colloids, being precipitated only by the salts of heav}- metals, alcohol, formalin, etc., or by great concentrations of neutral salts. Field and Teague-^ have found that agglutinins carry positive charges while bacteria are negative, and that b}' the electric current agglutinins can be separated from bacteria with which they have combined; this shows that the agglutinin is not destroyed in the reaction. Teague and Buxton^^ consider that neutralization of the electric charge of the bacteria is not, however, the only important factor in agglutination. According to Bechhold^^ normal bacteria behave like inorganic suspensions that have each particle protected by an albumin-like membrane, which prevents them from being thrown out of suspension by solutions of alkali salts, etc. After being acted on by agglutinin they are so altered that they behave like the unprotected inorganic suspensions, and are precipitated by salts and other electrolytes. This suggests the possibility that the agglutinin makes the bacteria 22 Ann. d. I'Inst. Pasteur, 1899 (13), 225. 23 Corroborated for sensitized red corpuscles by Eisner and Friedemann, Zeit. Immunitat., 1914 (21), 520. 2* Arrhenius (Zeit. physikal. Chem., 1903 (46), 415) has attempted to show that the gas laws are applicable to the partition of agglutinin between the bacteria and the medium, which he compares to the partition of iodin between water and carbon disulphid. This idea is not accepted Vyy Craw {loc. cit.), nor by Drej'er and Douglas, Proc. Roval Soc, 1910 (82), 185. 25 Jour. Exper. Med\, 1907 (9), 86. 2« Zeit. phvsikal. Chem., 1907 (57), 76. 2' Zeit. f. physikal. Chem., 1904 (48), 385. 182 CHEMISTRY OF THE IMMUNITY REACTIONS permeable for these electrolytes. Buxton and Shaffer^** also found that bacteria which have been acted upon by agglutinin behave as if their proteins had been so changed that they are more capable of absorbing or combining with salts than when in their normal condition. Strong salt solutions inhibit agglutination by preventing the binding of the agglutinin.-^ Tulloch^" observed that in the presence of salts of mono- and di-valent cations, unsensitiz.ed bacteria do not readily precipitate or agglutinate, but sensitized bacteria, as Bordet showed, agglutinate with small quantities of salts. In this respect unsensit- ized bacteria behave like "non-rigid colloids, " such as fresh egg white, while sensitized bacteria resemble "rigid colloids," such as denatured egg white. Hence he advances the hypothesis that the process of sensitization is akin to that of denaturation of proteins, the specificity perhaps depending on different degrees of denaturation. Mansfeld^^ would bring agglutination into line with other serological reactions as a protein digestion process, by his hypothesis that bacteria are held in suspension by protective colloids which are digested by an enzyme, the agglutinin. He finds in favor of this hypothesis that the tempera- ture and reaction curves correspond to enzyme actions, that agglu- tinating serum contains an enzyme digesting protein extracted from bacteria, and that during agglutination the agglutinogen is destroyed. Agglutination obeys the same laws as other similar physical phenomena; the rate of agglutination depends upon the concentration of the suspension and of the electrolytes, and varies with the valence of the cations. Although bacteria in an electric stream move toward the anode like all suspensions, after being acted on by agglutinin they are agglutinated by the current between the poles ;^2 this indicates the importance of the electrical charges of the bacterial surfaces in their agglutination reactions. In all respects the behavior of bacteria and agglutinin resembles the behavior of colloidal mixtures in suspension (Neisser and Friede- mann)'*'' which form an electrically amphoteric colloidal suspension, so that the ions of electrolytes or the electric currents, by discharging them unequally, cause precipitation. Physico-chemical researches, however, have yet failed to explain the specific character of the ag- glutinins for specific bacteria, but Michaelis^'* has developed an inter- esting analogy in the specific agglutination of bacteria by acids. This 28 Zeit. pliy.sikal. Chem., 1907 (57), 47. 29 Landstoinor and Ht. VVclocki, Zoit. Iiuiiiunitiit., 1910 (8), 397. 5" Biocheni. Jour., 1914 (8), 293. 3' Zeit. Iiiununitut., 1918 (27), 197. " Bocliliold; liowover, liuxton and Teague (Kolloid Zcitschr., 1908, II, Suppl. 2) 8tat(! tliat agglutinin bacteria do move towards tlie anode, but slower tli.'in normal bat-teria. " Miincli. med. Woch., 1904 (51), 405 and 827; see also ( uranl-Mangin and llemi, (.'onipt. Rend. Soc. I'iol., 1904, vol. 5(); ami Zangger, Cent. f. Hakt. (ref.), 1905 (3()), 225. ^* Folia Serologica, 1911 (7), 1010; akso IJeniasch, Zeit. Iiumunitat., 1912 (12), 208. AGGLUTINATION 183 is based on the fact that the optimum concentration of H-ions which precipitates proteins from solution is characteristic and constant for each protein, and the same is true for the aRglutination of bacteria by acids, the a^shitination by acids beinp; even more sharply specific in some cases than the agglutination l)y immune sera; e. g., typhoid and paratyphoid bacilli are readily distinguished because the former are agglutinated by a concentration of H-ions from 4 to8X10~^, while paratyphoids require 16 to 32X10"^, and colon bacilli are not agglu- tinated at all by acids. The acid agglutination, however, does not al- ways affect all strains in the .same way, some strains which are not readily agglutinable by antisera also resisting acid agglutination.^' According to Arkwright,^'' typhoid bacilli contain two extractable proteins that are agglutinated by acids, one at 3.6X10"^ and the other at 1.1 X 10"^; the former seems to be related to, if not identical with, the substance that is precipitated by immune serum. Apparently acid agglutination of bacteria belongs to the same class of reactions as the coagulation by H-ions of amphoteric colloids of preponderatingly acid character. Bacteria which have been sensitized by serum are more sensitive to acid agglutination than are normal bacteria." Alterations in the agglutinability of bacteria are marked, e. g., strains of typhoid bacilH freshly cultivated from human infections may be practically inagglutinable even by active serum, but after pro- longed cultivation on media they may or may not develop agglutina- bilit}^ This phenomenon has not yet been satisfactorily explained, but it may depend on an active immunity of the bacteria against the agglutinins. Such bacteria injected into rabbits produce antisera that will agglutinate ordinary agglutinable strains, but not themselves; hence they do not lack agglutinogens. They give normal complement fixation reactions, and hence do not lack receptors, and they agglu- tinate with acids and chemicals much the same as ordinary agglutin- able strains. ^^ Moreover, identical strains of bacteria grown on media of different composition may show considerable variations in agglu- tinability (Dawson). ^^ Conglutination. — Under this term Bordet and Gay described the observation that in ox serum there is a substance wliich combines with corpuscles (or bacteria) that have been acted upon by agglutinating sera, and augments the agglutina- tion.'"' Dean finds that, in general, agglutination requires two agents, one being the specific antibody, and the other a precipitable substance, probably a globulin. When cells have combined with the antibody the precipitable substance is aggre- gated on their surfaces, and, presumably, determines the agglutination. Co-agglu- tination, described by Bordet and Gengou as the agglutination bj' an antigen and the homologous antibody, of the corpuscles of another animal, is probabh- closely related to these phenomena (Dean). " See Kemper, Jour. Inf. Dis., 1916 (18), 209. 3« Zeit. Immunitat., 1914 (22), 396; Jour. Hyg., 1914 (14), 261. 'Mvrumwiede and Pratt, Zeit. Immunitat., 1913 (16), 517. =8 Mcintosh and McQueen, Jour Hyg., 1914 (12), 409. " Jour. Bact., 1919 (4), 133. "Literature given bv Dean, Proc. Royal Soc. (B), 1911 (84), 416; Hall, Univ. CaUf. Publ., Pathol., 19i3 (2), 111. 184 CHEMISTRY OF THE IMMUNITY REACTIONS PRECIPITINS" If to a solution containing proteins we add in proper proportions the serum of an animal immunized against the same protein, a pre- cipitate will soon form. While not absolutely specific, the quantitative specificity of the precipitin reaction is sufficiently characteristic to be of great value in biological, bacteriological, and medicolegal work, and it is of importance to the phj^siological chemist, since it furnishes a means of distinguishing between closel}^ related forms of proteins, more delicate by far than any known chemical reagent. The serum reactions also prove that there are sometimes essential differences be- tween the proteins of , different species of animals, even when by all other methods these proteins seem to be practically identical; e. g., lactalbumin of cow's milk is in some respect different from lactal- bumin of goat's milk since it produces a different precipitin. Medi- colegally they offer an accurate method of determining the origin of blood and serum stains, no matter how old the stain may be; thus Hansemann'*'- found that material obtained from a mummy 5000 years old gave the precipitin reaction.'*^ Production of Precipitins. — For the production of the precipi- tation reaction it is necessary to have in the substance used for immu- nization a certain group, the precipitinogen, which when injected gives rise to production of precipitin by the animal. Apparonth' almost any protein may act as a precipitinogen if injected into the proper ani- mal, but it must he a foreign protein; rabbit serum will not pioduce precipitins if injected into a rabbit, "*•* probably because it is normally present in the, blood of the rabbit and therefore does not stimulate any reaction; but certain chemical alterations in the proteins of an animal, such as hegiting, iodizing, or partial digestion, may render them so different from the normal proteins of the same animal that they will act as an antigen when present in the blood of that animal, or another of the same species, from which they were derived. Of the natural proteins of serum the globulins arc much more active precipitinogens than the albumins. In general the more foreign the protein, the greater the amount of precipitin; closely related animals, e. g., rabbit and guinea-pig, produce little precipitin for one another's proteins. This indicates distinctly that difference in species depends upon orj is associated with difference in chemical composition of the proteins. Different species of animals have very different capacity for produc- ing precipitins, rabbits producing active sera, while guinea-pigs can ^' For complete l)ibliogiaphy of the suliject of "Precipitins" see the r('>suin(' by Michaclis, Ujjpenheinier's llaiull). il. Hiochcinic. I'.IOK, II (1), bi>'2; Kraus, KoUc and Wassiirinann's llandb., 1913, II; Llilenliiitli and Stel't'enliagen, ibid., Ill, 257; Zinsser, "Infection and Resistance." « Munch, nied. Woch., 1904 (30), 572. " Not corroborated by Schnudt, Zeit. allg. Pliysiol.. 1907 (7), 309. ^* Rarely a slight reaction against homologous i)roteins has been obtained {iso- precipitins). PRECIPITINS 185 produce hut fcobl}' precipitating sera. Cantacuzene''" believes that precipitins are formed chiefly in the lymphoid tissues and bone marrow, and that the mononuclear macrophages are most active in their for- mation." This view is supported by the observations of Hektoen,''^ that anj'' agent that injures the bone marrow and lymphoid tissues (e. (J., Roentgen rays), tends to interfere with antibody production. Apparently only proteins can pioduee precipitins; when split to the peptone stage they lose this property, but the proteins of serum resist tryptic digestion a long time before losing their precipitinogcnic property,""* which is destroyed much more quickly by pepsin-HCl mixtures. The precipitate itself is very resistant to disintegrative agencies, including putrefaction (Friedbergor),^^ but is soluble in dilute acids and alkalies. It has the power of binding complement (Gay)*'' and if the complement causes solution of the precipitate, poisonous substances are formed (Friedberger). Excess of antigen prevents the formation of precipitate, or redissolves it, but excess of antiserum has no effect. Since both reacting substances are colloids they follow the laws governing other mutually precipitating colloids, and precipitation occurs only when they are brought together in concentrations that lie within definite zones of relative proportions. It is, of course, perfectly possible to have a union of precipitin and antigen without any visible ))recipitate occurring, since the product of the reaction is not neces- saiily insoluble under all conditions; in this case the occurrence of a reaction must be demonstrated by some other method, e. g., the com- plement fixation reaction. At present it is not established that pre- cipitins can be secured against lipoids or other non-protein substances. Possibly precipitins can be produced for closely related substances with molecules approximating in size the protein molecule, e. g., certain substances present in supposedly protein-free filtrates of bacterial cultures. As with the agglutinin reaction, electrolytes must be present or precipitation will not occur. Neither the precipitin nor the antigen seems to be altered appreciably by the reaction, since when either is separated from the precipitate it retains its original properties. Since precipitation of colloids is accompanied by or dependent upon an aggregation of their particles, the precipitin reaction is closely related to the agglutination reaction. The amount of precip- itation obtained is much modified b}^ the amount of inorganic salts present, and, according to Friedemann,*^ there is a general resem- ^5 Ann. Inst. Pasteur, 190S (22), 54. *^ Spleen tissue cultivated artificially in the presence of horse serum produces specific precipitins for horse serum, and tissue from the spleen of a guinea pig that has received injections of horse serum also develops precipitins for horse serum when grown in cultures (Prvzgode, Wien. klin. Woch., 1914 (27), 201). ^^ Jour. Infect. Dis., 1915 (17), 415; 1918 (22), 28. *^ Fleischmann, Zcit. klin. Med., 1906 (59), 515. " Cent. f. Bakt., 1907 (43), 490. 50 See Univ. of Calif. Publ. Pathol., 1911 (2), 1. 51 Arch. f. Hyg., 1906 (55), 361. U 186 CHEMISTRY OF THE IMMUNITY REACTIONS blance between the precipitin reactions and the precipitations occur- ring when colloids precipitate one another; i. e., when an amphoteric colloid reacts with either an acid or a basic colloid. ^^ So far, however, attempts to interpret the precipitin reaction, as Arrhenius has tried to do, on the basis of the laws of physical chemistry, have not met with much success (Michaelis). We prefer the attitude of Krogh,^^ who states that the colloidal chemical part of immunological reactions is to be looked upon as only a preliminary step to the real chemical process that completes the reaction and gives it the specific characters. As mentioned in the preceding section, agglutination of bacteria is be- lieved to be independent of the precipitins, although very probably influenced by them. As wdth all the other substances of this class, the precipitins have a haptophore group by which they unite to the protein molecule, and another group by which they produce the change re- sulting in precipitation. When the latter group is destroyed by heating to 72°, the precipitin is converted into a precipitoid, which possesses the property of preventing the precipitation of unheated precipitin by the specific antigen.^'* The immune serum contains the precipitin, which is the passive reagent that is thrown down by a trace of the immunizing material (precipitinogen). The resulting precipitate is the insoluble modifica- tion of the previously dissolved precipitin, and originates chiefly or entirely in the proteins of the immune serum, ^'^ according to the work of Welsh and Chapman, especially. But as the precipitate is able to sensitize anaphylactically, both actively and passively, it would seem that it must contain both the antibody (which confers passive sensi- tization) and antigen, to cause active sensitization (Weil).^^ The precipitate may, when of maximum amount, contain more nitrogen than corresponds to the entire euglobulin of the immune serum, and the euglobulin contains all the precipitin, so it seems probable that the precipitate consists of more than the precipitin alone; it maj^ be added that the precipitate is always less in amount than the total globulin of the antiserum. ^^ It is always greater when the reaction is between homologous antiserum and antigen, than with even closely related but heterologous antigens,^** so that the quantitative measurement of the amount of precipitate is of value in applying this reaction to deter- mine the nature of protein solutions. The dilution of the reacting solutions is of influence, however, for if in too dilute solutions weak " See Friedemann and Friedenthal, Zeit. exp. Path. u. Ther., 190G (3) 73; Iscovesco, Compt. Rend. Soc. Biol., 1906, Vol. 61, and subsequent volumes. " Jour. Infect. Dis., 191G (19), 452. ^* Precijntinogens are relatively resistant to moderate heating, and heated extracts of bacteria are used for precipitin tests under tlie name thermoprccipiiiiis. See review by A. Ascoli, Virchow's Arcli., 1913 (213), 182. " Moll, Zoit., exp. Path. u. Ther 1900 (3), 325; Welsh and Chapman, Proc. Royal Soc, B., 190ling and Petersen have suggested that the antigen-antibody combination may adsorb or bind the antiproteases of the serum, so that the normal protease digests the serum proteins. Or it niay be that union of antigen and antibody activates the complement, or binds it to the antibodj' so that it digests either the antibody or other proteins of the serum. It also is suggested that enzymes are set free from the tissues injured by the specific protein, or bj' disease, which digest the foreign protein or the cellular proteins that may have escaped from the tissues into the l)lood stream. The reaction possesses a certain specificity, but just the degree of this specificity has not been agreed upon. The claim of Abderhalden*^ and his followers, that it is by far the most specific of immunity reactions, whereby disintegration of small amounts of any given organ of an individual can be determined by specific reactions between his serum and that organ, with such refinement that even cerebral localiza- tion is possible,*^ is diflficult to accept. There are so many possible sources of error in the original technic that even with great care the charge of incorrect results from incorrect technic cannot be escaped, and therefore, those who do not accept the doctrine of its specificity are always on the defensive. Nevertheless, so many care- ful and experienced investigators have found the original Abderhalden reaction to give at times absoluteh^ non-specific and hopelessly paradoxical results, that its diagnostic value for either clinical or scientific purposes must be considered at present as unproved,*^ whatever the final decision as to its standing as a specific reaction may be. Serum treated with various inert, finelj' divided particles, such as kaolin, starch, silicates, etc., may acquire the property of giving positive reactions. This is another point of resemblance to anaphylatoxin formation, and against the speci- ficity of the reaction, indicating that the antigen merelj^ acts as a non-specific adsorbent. By far the most satisfactory results have been recorded in the diagnosis of pregnancy by means of placental antigen. This may be explained by the fact that the protease activity of the serum seems to be increased in pregnancy,*^ and hence the reaction with placenta is more marked than with the serum of non- pregnant individuals. But simply shaking normal serum with kaolin or other foreign substances may cause it to give strong reactions with placenta antigen (Wallis). *- Supported by Smith and Cook, Jour. Infect. Dis., 1916 (18), 14. De Waele states that it is the serum globulin that is digested (Compt. Rend. Soc. Biol., 1914 (76), 627). *' A reply to numerous criticisms is given by Abderhalden, Ferment forschung, 1916 (1), 351; this and other numbers of this journal also consist largely of articles on the Abderhalden reaction. 8^ See Retinger, Arch. Int. Med., 1918 (22), 234. *^ O. J. Elsesser (Jour. Infect. Dis., 1916 (19), 655), working in my labora- tory with the purified vegetable proteins of Osborne, found that at best the speci- ficitj' of the reaction was less than that of the anaphylaxis reaction, and there were many absolutely non-specific and irrational reactions. As these pure proteins furnish a much more appropriate material for studying specificity than the tissues or sera commonly used, it would seem that the results thus obtained are excellent proof of the uncertainty and unreliability of the reaction. Careful quantitative studies of the setting free of amino-acids by serum incubated with placenta, by Van Slyke and his associates, also showed a complete lack of specific proteolysis by pregnancv serum (Arch. Int. Med., 1917 (19), 56; Jour. Biol. Chem., 1915 (23), 377; see" also Hulton, ibid., 1916 (25), 163). «6 See Sloan, .Amer. Jour. Physiol., 1915 (39), 9. CHAPTER IX CHEMISTRY OF THE IMMUNITY REACTIONS (Continued)— BACTERIOLYSIS, HEMOLYSIS, COMPLEMENT FIXATION, AND SERUM CYTOTOXINS SERUM BACTERIOLYSISi The bactericidal property of serum may be shown by its destruc- tion of the life manifestations of bacteria without marked alteration in their structure, or it may be accompanied by dissolution of the bacterial cell (bacteriolysis). How much of the bacteriotytic process is performed by the serum itself, or how much by the autolytic en- zymes of the bacterial cell, is unknown, but the latter is probably a factor. The bactericidal property of immune serum has been shown to be quite independent of the antitoxic properties and also to have quite a different mechanism. This last is shown in the following manner : If we heat bactericidal serum made by immunizing an animal against bacteria, say the cholera vibrio, at 55° for fifteen minutes, it will be found to have lost its power of destroying these organisms. ^ Normal serum of non-immunized animals is equally without effect upon the vibrios. If however, we add to the inactivated heated serum an equal quantity of inactive normal serum, the mixture will be found to be as actively bactericidal as the original unheated immune serum. This phenomenon is interpreted to mean that, by immuniza- tion, some new substance has been developed which, although b}- itself incapable of destroying bacteria, is able, when united with some sub- stance present in normal serum, to destroy bacteria readily. The substance present in normal serum is also incapable of affecting bac- teria by itself, but needs the presence of the substance developed by immunizing to render it bactericidal. Hence the bactericidal prop- erty in this case depends on two siibstances acting together: one, de- veloped during immunization and therefore called the immune body, is specific for the variet}^ of bacteria used in immunization, and is not destroyed by heating at 55°. The other, present in normal serum, is not increased during immunization, is not (altogether) specific in character, and is destroyed by heating at 55°; as its action is com- ^ Review and bibliography by Miiller, Oppenheimer's Handb. d. Biochcni ., 1909 11(0,629. ^ Normal human serum often exhil)its some jiower to destroy Ijacteria, even after heating to 55°. The nature of this thermostable bactericidal agent is unknown. (See Salter, Zeit. Hyg., 1918 (86), 313). 206 AMBOCEPTOR AND COMPLEMENT 207 plenientary to that of the specific iininune body, it is called tlic com- plement.^ It is believed that the action of these substances is as follows: The ininiunc body is, like antitoxin, a cell receptor which unites the bac- teria to the cell. It differs from the antitoxin, however, in that it has two affinities, one for the complement and the other for the bacterial substance. On account of the existence of the two affinities it is called an amhoceptor. Some serums contain such amboceptors for certain bacteria without previous immunization, hence the term im^mune amboceptor is reserved for amboceptors developed by imnmnization. Amboceptor and Complement.'* — The function of the ambo- ceptor is to unite the bacterial protoplasm, to which it is attached by one affinity, to the complement which it holds by its other affinity, or, to put it in a more strictl}^ chemical way, the addition of the ambocep- tors to the bacteria gives them a chemical affinity for complement. It is, therefore, an intermediary body, uniting the complement to the bacterial protoplasm. The complement^ is the substance that actually destroys the bacteria, in which respect, as well as in its susceptibility to heat, it resembles the enzymes. Complement is present in normal serums, and, as it is not increased in amount during immunization it may not be sufficient to satisfy all the amboceptors, hence it may be impossible to secure marked bactericidal effects even when many amboceptors have been formed. If the complement in an immune serum has been destroyed by heating, it may be replaced by adding normal serum from another animal, even of some other species; indi- cating either that the complement is not absolutely specific in its nature, or that quite the same complement may be present in the blood of many different animals. The origin of the complement is unknown, but it has been urged that the leucocj^tes are an important source of /this substance, if not its chief one;^ there is evidence, how- ever, that various organs and cells may also produce complement.'^ Its most important characteristics are its extreme susceptibility to heat, and the resemblance of its action to the action of enzymes.^ Hektoen^ found that it could be made to unite with Mg, Ca, Ba, Sr, and SO4 ions, which rendered the complement (for typhoid bacilli and red corpuscles) inactive. Man waring ^° found that these ions could be ^ The polynuclear leucocytes also contain bacteriolj^tic agents, " endolysins, " of a similar complex structure, but quite distinct from the serum bacteriolvsins (See Kling, Zeit. Immunitat., 1910 (7), 1). ■• See also Hemolysis, Chapter X. ° Review and bibliography by Noguchi, Biochem. Zeit., 1907 (6), 327. ^ Cholera antiserum will produce the Pfeiffer phenomenon of lysis of cholera vibrios in animals made leucocj'te-free with thorium. (Lippmann, Zeit. Immuni- tat., 1915 (24), 107.) ' See Dick, Jour. Infect. Dis., 1913 (12), HI; and Lippmann and Plesch, Zeit. Immunitat., 1913 (17), 548. « See Walker, Jour, of Physiol., 1906 (33), p. xxi. 3 Trans. Chicago Path. Soc, 1903 (5), 303. 10 Jour. Infectious Diseases, 1904 (1), 112. 208 CHEMISTRY OF THE IMMUNITY REACTIONS separated again from the complement by simple chemical precipita- tion. Acids stronger than CO2 and of the higher saturated or un- saturated fatty acid series, inactivate complement in strengths greater than n/40, and alkalies are equally inhibitive.^' Ultraviolet rays destroy complement. ^^ Sherwood^^ has made a study of various sub- stances that may be present in the blood in excessive amounts during pathological conditions, such as CO2, lactic acid, acetone, etc., and finds that they interfere seriously with the action of complement, which suggests that they may favor infection or interfere with recovery from infection. Presumably the complement is a protein, for it has antigenic prop- erties, so that immunization with sera containing either complement or complementoid causes anticomplement activity in the blood of the immune animal. Also, it is destroyed by trypsin free from lipase, ^^ and, like other colloids, is readily adsorbed by surfaces; hke enzymes, complement is destroyed by shaking, ^^ and gradually disappears on standing. There are some striking resemblances between the be- havior of complement and of certain compounds of protein with soaps and lipoids, as pointed out especially by Noguchi, but that these are identical with true complement is doubtful. (See Hemolysis.) Its colloid nature is attested by the large loss when complement is filtered through Berkefeld filters. ^*^ A careful review of the evidence has led Liefmann^'^ to the conclu- sion that the reaction of complement to sensitized corpuscles is more hke that of ferment to substrate than of antigen to antibod3^ In its effect of dissolving bacteria (and also other cells against which animals may have been immunized) co77iplement resembles the enzymes, and by many it is looked upon as related to them, but the changes it pro- duces do not resemble those produced by proteolj'tic enzymes in all details.^* In particular, complement seems to participate in reactions according to the law of definite proportions, unlike the enzymes. ^^ In certain immune reactions, colloids (lecithin, silicic acid)-" can play the role of complement and immune body, but these reactions are pro- bably quite different from those of bacteriolysis by immune serum. Structure of Complement. — According to the Ehrlich theory, complement, like toxins and enzymes, possesses at least two groups: one, the haptophore, by 11 Noguchi, Biochem. Zeit., 1907 (6), 172. 12 Courmont et al, C. R. Soc. Biol., 1913 (74), 1152. " Jour. Infect. Dis., 1917 (20), 185. 1^ Michaelis and Skwirsky, Zeit. Immunitat,. 1910 (7), 497. lii Noguchi and Bronfenbrenner, Jour. Exp. Med., 1910 (13), 229; Ritz, Zeit. Immunitiit., 1912 (15), 145. i» See Schmidt, Arch. f. Hyg., 1912 (70), 284; Jour. Ilyg., 1914 (14), 437. 1^ Zoit. Immunitiit., 1913 (16), 503. " The curve of complement action resembles that of enzyme action. (Thicle and Emblcton, Jour. Path, and Bact., 1915 (19), 372.) i» See Liebermann, Dcut. med. Woch., 190(5 (32), 249. 20 Landsteinor and Jagic, Wien. klin. Woch., 1904 (17), 03; Munch, med. Woch., 1901 (51), 1185. CYTOTOXINS 209 which it unites witli tho .aiiihocrnptor; the other, the toxophoro (or zymophorr, because of its enzynie-lilce action), which attacks the bacterial protoplasm. It may def^enerate and lose its toxoi)hore f^roup while retaining the pow(!r to combine by means of its ha])topiiore )j;i"'Jiil>» thus formiiip; a romplruicntoid. (Complement and aiubocejitor exist side l)y side in the serum, not uniting with one another until the ambocei)tor has become attached to the liactcrial protoplasm. It is generally stated that if serum containing complement be so treated as to separate the globulins from the albumin, it is found that the complement has been divided into two parts, one i)resent in each of the protein fractions. The globulin fraction of the complement will unite to am])oceptor which is fixed to cells, and hence is called the mid-pircc of the complement, for it will unite also with the end-piece of the complement contained in the all)umin fraction, and then cytolysis can take jjlace. Without the intervention of the globulin mid-piece the albumin end-piece cannot unite with the amboceptor, while in the ab.sence of end-piece the amboceptor mid-piece complex can cause no cytolysis. Both fractions of the complement are destroyed by heat, but if the mid-piece is bound to the ambo- ceptor it resists heating. The mid-piece corresponds to Ehrlich's haptophore, the end-piece to the toxophore group, and this complex structure is common to both bacteriolytic and hemolytic complement. Bronfenbrenner and Xoguchi,^* however, contend that the supposed cleavage of complement is merely an inactiva- tion by the agencies employed, all the complement being in the albumin fraction in a condition capable of reactivation, not only by globulin but by simple ampho- teric substances, a view which has not been generally accepted. ^- Amboceptors are formed, according to Wassermann, and Pfeiffer and Marx, in the spleen and hemopoietic organs, since in immunization they can be demonstrated in these organs before the)^ appear in the circulating blood. The stability of the amboceptors is very considerable: serum prepared in 1895 by PfeifTer against cholera vibrios was found to have lost almost none of its activity after eight years in an ice-box (Friedberger). Heating twenty hours at 60° scarcely injures them, but 70° for one hour destroys them almost completely, and heating the serum to 100° destroys all the immune bodies. They are quite resistant to putrefaction, and, like the antitoxins, do not dialj'ze. Strong salt solutions will prevent the union of complement and amboceptor in vitro, and probably to greater or less degree in the animal body, but the union of antigen and amboceptor is not pre- vented by salt.^' Alkalies may prevent the union of amboceptor with the cells, or extract it from the cell to which it has united; and they maj"- also inhibit the union of amboceptor and complement. Amboceptors are not inactivated by shaking, as is complement, but they are destroyed alike by ultraviolet rays, and both resist x-rays.^'* According to Pfeiffer and Proskauer,^^ digestion of the globulin precipitate, in which amboceptors are carried down, does not destroy their activity completely even when all the proteins are thus removed. Removal of the nucleo-albumin or nuclein does not remove the amboceptors from the serum. Immune serum kept three months in alcohol yielded an extract with distilled water that was rich in immune bodies, but almost free from protein. Pick, Rhodain, and Fuhrmann found that immune bodies are precipitated entirely in the euglobulin fraction of the serum protein. From these experiments it has been thought bj^ some that the bacteriolytic amboceptor is not itself a protein, although closely associated with the serum globulins.-^ CYTOTOXINS Just as precipitins can be obtained for proteins derived from other sources than bacterial cells, so also upon immunizing an animal -^ Jour. Exp. Med., 1912 (5), 598; good review of literature. " See Leschlev, Zeit. Immunitat., 1916 (25), 44. " Angerer, Zeit. Immunitiit., 1909 (4), 243. " Scaffidi, Biochem. Zeit., 1915 (69), 162. " Cent. f. Bakt., 1896 (19), 191. ^® -\scoli found that the active substance of anthracidal serum, which is not an amboceptor, is contained in the pseudo-globulin fraction of asses' serum, but in goat's serum part is in the euglobulin fraction. (Biochem. Centr., 1906 (5), 458.) 14 210 CHEMISTRY OF THE IMMUNITY REACTIONS against various types of cells other than bacteria, substances appear in its serum that exercise a destructive effect upon the tj^pe of cells injected. In other words, the reactions of animals to infection are not specially devised for combating bacteria and their products, but can be equally exerted against non-bacterial cells and their products. In the case of soluble proteins, as before mentioned, the antibodies show their effects by precipitating them, with agglutination of the par- ticles into flocculi and perhaps a subsequent digestion; in the case of cells, whether bacterial or tissue cells, the antibodies cause agglutina- tion and loss or impairment of vitality. This injury may be mani- fested by loss of motion in the motile cells (bacteria, spermatozoa, ciliated epithelium) or by solution of their contents (bacteriolysis, erythrocytolysis, leucocytolysis, etc.), or by cell death without marked morphological alterations (B. typhosus, spermatozoa). If we inject red corpuscles, leucocytes, spermatozoa, renal epithelium, or anj^ other foreign cell, the reaction is as specific as it is if we inject bacteria, and of exactly the same nature. Therefore, all that has been said pre- viously concerning bactericidal substances and agglutinins can be transposed to apply to immunity against tissue cells. As a matter of fact, however, the transposition is generally made in the other direc- tion, for red corpuscles are much easier cells to study than bacteria, because their laking gives prompt and readily recognized evidence that the toxic serum has brought about changes. Much of our knowl- edge of bactericidal serum has been obtained through studies of the mechanism of erythrocytolysis, the results of which have then been applied to the subject of bacteriolysis. Both on this account, there- fore, and because solution of red corpuscles is of itself an important process in many intoxications and diseases, the subject is of great theoretical and practical importance. Hemolysis-' or Erythrocytolysis In hemolysis the essential phenomenon consists in the escape of the hemoglobin from the stroma of the corpuscles into the surroimd- ing fluid. As it is not exactly known in what way the stroma holds the hemoglobin normally, whether purely physically or in part chem- ically, or whether the stroma consists of a spongioplasm or of sac-like membranes, or both, the ultimate processes that permit the escape of the hemoglobin are not finally solved. However, the agents bj' which the escape is brought about are well known and extensively studied, and they are found to be of extremely various natures. Thej' may be roughh' classified as: (1) known physical and chemical agents; (2) unknown constituents of blood-serum; (3) bacterial products; (4) certain vegetable poisons; (5) snake venoms. ^' Through usage this term has been limited to the sohition of the rod corpus- cles, which is more accuratelv described hv the term crythron/toli/xis. For liihli- ography see Sachs, Ergel)nisse der Pathol.", 1<»02 (7), 7i4; Ht'Oti (11), 515; Kolle and Wassermann's Handbuch, 1913 (II), 793; Landsteiner, Handbuchd. Biochem., 1909 (II (D), 395. MKCIIAMSM OF IIKMOI.YSIS 211 Hemolysis by Known Chemic\l and Physical Agencies The Mechanism of Hemolysis — -If distilled water is added to corpuscles of any kind, osmotic changes arc l)ound to occur, since within tlic cells arc abundant salts, soluble in water, whicii will begin to diffuse outward in an attempt to establish osmotic equilibrium be- tween the corpuscles and the surrounding fluid. Conversely, water enters the corpuscles at the same time, and accumulating there leads to swelling until such injury has been produced as permits the hemo- globin to escape and enter the surrounding fluid. Before this oc- curs the fluid is opaque because of the obstruction to light offered by the red cells, but on the completion of hemolysis the fluid becomes transparent. The stroma now settles to the bottom, while the hemo- globin diffuses into the fluid, making it red, but perfectly transparent. This process has long been known as the "laking" of blood, and is essentially the condition present in all forms of hemolysis. That the hemoglobin escapes only through inj\ny of the stroma and not through simple osmotic diffusion, is shown by the fact that if salt solution of the same concentration as normal serum is used instead of distilled water, no such escape of hemoglobin occurs. As hemoglobin is per- fectly soluble in salt solution, it should pass out if it diffused as do the salts. Since there is no escape of hemoglobin in such a salt solution, it is evident either that the stroma is not permeable to hemoglobin, or else the hemoglobin is in some way attached to or combined with the stroma. Again, if the corpuscles are placed in a solution of salt more concentrated than their own fluids, water escapes and the cor- puscles shrink; as no hemoglobin escapes with the water, it is evident that the stroma is not permeable to hemoglobin when intact. Because of the resemblance of the process of hemolysis to the rupture of plant cells with escape of their contents when they are placed in distilled water, it might be assumed that hemolysis is largely a phj'sical matter, but if a red corpuscle in an isotonic solution is cut into pieces, the hemoglobin does not escape, indicating that its structure is quite dissimilar to that of the simple vegetable cell and that there is some union of stroma and of hemoglobin, either physical or chemical.-^ Physico-chemical studies also indicate that there is no true covering membrane to red corpuscles, for the absorption of ions by hemoglobin is the same as the absorption by corpuscles.-^ M. H. Fischer'"^ interprets hemolysis as a separation of lipoid-protein stroma and ad- 28 Stewart (Jour, of Physiol, 1899 (24), 211) found that in hemolysis by physical means or under the influence of serums, there is no marked increase in the electrical conductivity, but hemolysis by saponin and by water causes an increase of conductivity, presumably because of the escape of electrol^'tes ; cor- roborated by A. Woelfel, Biochem. Jour., 1908 (3), 146; see also Moore and Roaf, tuid D S ^ " Rohonyi, Kolloid-chem. Beihefte, 1916 (8), J37, 391. Knaffel-Lenz (Arch, ges. Physiol., 1918 (171), 51) also finds evidence that there is no limiting lipoid membrane about red cells. 30 Kolloid Zeit., 1909 (5), 146. 212 CHEMISTRY OF THE IMMUNITY REACTIONS sorbed hemoglobin, which process can be dupHcated experimentally with a combination consisting of a corresponding solid hydrophilic colloid, fibrin, and a hydrophobic colloid dye, carmin^e; this artificial combination behaves exactly like a corpuscle to simple hemol^'tic agents. ^^ Repeated alternate freezing and thawing is another physical means of bringing on hemolysis. Heating to 62°-64° C. causes hemolysis of mammalian corpuscles; in cold-blooded animals this seems to occur at a slightly lower temperature. Hypertonic solutions produce hemo- lysis, and it may be that freezing and desiccating cause hemolysis through the resulting hypertonicity.^- Some chemical agents are capable of liberating hemoglobin, even when the corpuscles are in isotonic solutions. The ordinar}' salts of serum, of course, do not have this property, but ammonium salts are strongly hemolytic. The chemical agents that dissolve red cor- puscles seem to be those that have the power of penetrating the stroma. Ammonium salts and urea penetrate the corpuscles freel}^ and cause hemolysis. Sugar and NaCl seem not to penetrate the corpuscles, and therefore do not produce hemolysis. Of the perme- ating substances, there seem to be two types: one, like urea, does not produce hemolysis when in a solution of NaCl isotonic with the serum; the other, like ammonium chloride, is not prevented from producing hemolysis by the presence of NaCl.^^ All these agents seem to effect hemolysis by acting on the stroma, for when the stroma of corpuscles hardened in formalin has its leci- thin and cholesterol removed with ether, saponin, a powerfully hemo- lytic substance, seems to have no effect. The action of saponin and of many other hemolytic agents can be prevented b}' the presence of cholesterol in excess, suggesting that it is this constituent of the stroma that is affected.^* By studying hemolysis under dark field illumination Dietrich^^ found that in water hemolysis a diffusion of hemoglobin ^^ Concerning the influence of H-ion concentration on hemolysis see Walbum, Biochem. Zeit., 1914 (63), 221. ^2 Guthrie, Jour. Lab. Clin. Med., 1917 (3), 87. ^^ Hamburger, in his book, "Osmotischer Druck und lonenlehre," reviews ex- haustively the physical chemistry of hemolysis. The following is his summary of the permeability of red corpuscles by various substances: Organic Substances. — (a) Impermeable for sugars; namely, cane-sugar, dextrose, lactose, also arabit and mannit. (6) Permeable for alcohols, in inverse proportion to the number of hydroxyl groups that they contain; also for aldohj'des (except paraldehyde), ketones, ethers, esters, antipyrin, amides, urea, urethan, l)ile acids and their salts, (c) Slightly permeable for neutral amino-acids (glycocoll, aspar- agin, etc.). Inorganic substances, not including the salts of the fixed alkalies, (a) Com- pletely impermeable for the cations Ca, Sr, Ba, Mg. {b) Pcnncable for NHj ions, for free acids and alkalies. ^■•Ransom, Deut. med. Woch., 1901 (27), 194; Koliert, "Saponinsubstanzen'' Stuttgart, 1904; Abdcrhalden and Le Count, Zeit. exp. Path. u. Ther., 1905 (2), 199. Noguchi (Univ. of Penn. Med. Bull., 1902 (15), 327) found lecithin without this proi)erty. "Verb. Deul. Patli. (icscll., 190S (12), 202. MECHANISM OF HEMOLYSIS 213 takes place tluoufih the coipusculur substance, which is not visibly altered; in serum hemolysis there is first a precipitate formed in the outer layer, which swells. There is no evidence that the erythrocj^tes contain proteolytic enzymes of their own that might disintegrate them.'*' The fact that chloroform, ether, bile salts, soaps, and amjd alcohol will cause hiking is probably intimately connected with the fact that lecithin and cholesterol, important constituents of the stroma, are both soluble in these substances.''' In general it can be said that iiemolytic agents dissolve lipoids or hydrolyze proteins or lipoids, thus destroying the power of the stroma to retain the hemoglobin.^* Nearly all the non-specific hemolytic agents are inhibited to greater or less degree by the serum, in which inhibition both the proteins and cholesterol are concerned. '^^ Cholesterol also influences many other inununity reactions, inhibiting some and stimulating others.*" The resistance of the corpuscles to hemolj'sis by various agents differs greatly in disease, although fairly constant in normal blood, the dif- ferences being caused in some cases by changes in the permeability of the corpuscles, and sometimes by changes in the environment of the corpuscle or the presence of protective substances in either the cor- puscles or the plasma. Arseniuretted hydrogen, when inhaled, causes intravascular hemo- lysis, and there are many other drugs and chemicals with the same property, among which may be mentioned nitrobenzol, nitroglycerin and the nitrites, guaiacol, pyrogallol, acetanilid, and numerous aniline compounds. Probably the hemolysis produced by autolytic products belongs in this category."*^ Alcoholic extracts of tissues are com- monly hemolytic; these extracts when added to serum take on prop- erties which cause them to resemble closely hemolytic complement (Noguchi), and the soaps seem to be the active constituents of the extracts. AsHs, although strongly hemolytic in the living body, does not hemolyze corpuscles in the test tube (Heffter), and this is true of some other poisons, which probablj' produce their effects through tissue changes. •^^ The bile acids and their salts will also produce hemolysis, as seen in jaundice. Sodium bicarbonate solutions of one or two per cent, are hemolytic for some varieties of corpuscles, but 0.1 per cent. Na2C03 and NaHCOs do not cause hemolysis. A study of the hemotytic properties of one class of lipolytic hemolytic agents, 38 Von Roques, Biochem. Zeit., 1914 (64), 1. " See Koeppe, Pfiiiger's Arch., 1903 (99), 33; Peskind, Amer. Jour. Phys., 1904 (12), 184; Moore, Brit. Med. Jour., 1909 (ii), 684. 38 See Herzfeld and Klinger, Biochem. Zeit., 1918 (87), 36. " See V. Eisler, Zeit. exp. Path., 1906 (3), 296. *° Walbum, Zeit. Iminunitiit., 1910 (7), 544; Dewey and Nuzum, Jour. Infect. Dis., 1914 (15), 472. " Concerning hemolysis b.y alcohols, ketones, etc., organic acids, and essences see Vandevelde, Bull. Soc. chim. de Belgique, 1905 (19), 288. " Friedberger and Brossa, Zeit. Immunitat., 1912 (15), 506. 214 CHEMISTRY OF THE IMMUNITY REACTIONS the terpenes, shows that their hemolytic activity varies much accord- ing to their physical properties, generally decreasing directly with in- crease in the solubilitj' in water (Ishizaka).-*^ Leucocytes are dissolved by some of these agents, particularly the bile salts although they are affected by no means so rapidh^ or so much as are the erythro- cytes. There seems to be no relation between the erythroh^tic and leucolytic powers of these substances. Water causes swelling, with solution of the granules in time, and the same is true of ammonium-chloride solutions. Various chemicals cause morphological alterations in the leucocytes, and of bacterial products the toxins of pyocyaneus and diphtheria seem to be particularly leucocidal, causing a striking karj'orrhexis (Schiirmann)." Hemolysis by Serum Normal blood-serum of many animals causes hemolysis to greater or less degree when mixed with red corpuscles of another species of animal, and this property can be greatly increased by immunizing the animal with red corpuscles in the usual way. This hemolysis oc- curs both in the test-tube and in the body, in the latter case causing severe anatomical changes or even death. In all respects the mech- anism of hemolysis by serum seems to he identical with that of bac- teriolysis. Two substances are concerned, one the amboceptor, which resists heat and which is increased by immunizing;*^ the other, com- plement, which is destroyed at 55° and which is present in normal serum. In this case the substances may be referred to as hemolytic amboceptors and hemolytic complements. In spite of the availability of these particular cytolytic substances for study, very little has been learned of their exact nature and prop- erties. It is known that amboceptor is combined with the red cells in a certain sense quantitatively, a definite amount being re- quired to saturate a given amount of corpuscles so that they will all be hemolyzed when complement is added; and that this reaction is complete in less than fifteen minutes at 45°. What change this addi- tion of amboceptor brings about in the corpuscles is unknown. It has also been shown that at 0° the affinity between the amboceptor and the corpuscle is greater than it is between amboceptor and com- plement, so that it is possible at this temperature to remove all the amboceptor from a serum by treating it with red corpuscles, and thus we can obtain complement free from amboceptor. This experi- ment also shows that the two bodies exist side by side in the serum without combining, and that combination occurs only after the ambo- ceptor has become united to the erythrocyte. Moreover the hemoly- tic amboceptor can be separated from the antigen to which it has been " Arch. exp. Path., 1914 (75), 195. ** Cent. f. Pathol., 1910 (21), 337. *^ In an extensive study of the hemolytic antiI)ody, Thiele and Embleton (Zeit. Imnuinitiit., 1913 (20), 1) describe its formation as in several steps, at first being tlicrmolabile and uniting with the corpuscle only when warmed. They also find complt^ment to have sev(>ral components. This is not coufiriued by Siier- man, Jour. Infect. Dis., 1918 (22), 534. HEMOLYTIC AMBOCEl'TOR 215 combined.'"' Hemolysis hy iimmine sera takes place l)est in a medium with a reaction corresponding to that of the blood, acids being more iiarmful than alkalies; with unfavorable reaction the complement does not unite with the aml^oceptor, although the latter unites with the corpuscle.'^ The Amboceptor. — Amboceptor is, a.s a rule, destroyed hy heating to 70" or liifiiher.'^ Its place of origin is unknown. Metchnikolt holds that it is derived chiefly from the leucocytes, in support of which view is the fact that leucocytes dissolve red corpuscles after ingesting them; however, other phagocytic cells have the same power, particularly endothelial cells, and it is an open question whether the intracellular digestion of engulfed cells is the same process as extracel- lular hemolysis; prol)ably it is not, for there seem to be more disintegrative changes in intracellular digestion than in hemolysis. Quinan^^ found that the diffusible constituents of hemolytic serum played no role beyond that of maintaining os- motic pressure. He was unable, however, to localize the immune body in any of the protein constituents, and Liebermann and Fenyves.sy^" believe that they obtained the amboceptor in a protein-free condition, in which it behaves like a weak acid. Amboceptors are insoluble in lipoids or lipoid solvents (Meyer),*' and they move towards the cathode in an electric field, as do other antibodies. *- The amboceptor complement reaction resembles a bimolecular reaction wliich is accelerated bj' its end products (v. Krogh).*' Many of the effects of hemolytic amboceptors can be duplicated with silicic acid;^^ and a dye, brilliant green, may in minute quantities sensitize corpuscles so that they are hemolyzed by very small amounts of normal serum, or by lecithin.^^ The amboceptors of normally hemolytic serum seem to be no different from those in immune serum, and amboceptors of one animal can combine with comple- ment furnished by the serum of an entirely different animal. It is the amboceptor alone that gives the specific nature to the reaction, and, as is the case with all other immunizations, it is very difficult to secure antibodies by immunizing an animal with blood from another animal of its own species, isohemolysiris. The place of origin of hemolysins is unknown, as with other antibodies, but that it is not in the blood seems to have been established conclusively by Hektoen and Carlson.** Immune hemolysins cannot pass from the mother to the fetus before birth*" but they can be transmitted through the colostrum (Famulener).*^ Although Ehrlich held that the union between cell and amboceptor is pureh' chemical and follows ordinary chemical laws, especially the law of multiple pro- portions, Bordet and other French observers have claimed that the union between amboceptor and corpuscle is physical and not chemical.*^ Probably the union is *'' Kosakai, Jour. Immunol., 1918 (3), 109. *'' Michaehs and Skwirsk}', Zeit. Immunitat., 1909 (4), 357. ^^ Ultraviolet light destroys immune hemolysin (Stines and Abelin, Zeit. Immunitat., 1914 (20), 598). " Hofmeister's Beitr., 1904 (5), 95. *° Jahresber. d. Immunitat., 1911 (7), 2. *' Ibid., 1909, Vol. 3. " Teague and Buxton, Jour. Exper. Med., 1907 (9), 254. " Biochem. Zeit., 1909 (22), 132. " Landsteiner and Rock, Zeit. Immunitat., 1912 (14), 14. ** Browning and Mackie, Zeit. Immunitat., 1914 (21), 422. *« Jour. Infect. Dis., 1910 (7), 319. " See Sherman concerning normal antibodies in the fetus. Ibid., 1918 (22), 534). "76id, 1912 (10), 332. ^' Bang and Forssmann (Hofmeister's Beitr., 1906 (8), 238) suggest that the amboceptor merely renders the corpuscle permeable for the complement, perhaps through action on the lipoid membrane; the complement then acts directly upon some constituent of the corpuscle, without the amboceptor acting as a combining substance in any way. They found that the substance in blood which stimulates the antibody formation in the case of hemolysin formation, is chemically separable from the substance in blood which unites with these antibodies; therefore, they conclude, the "receptors" of cells are not identical with the antibodies. (See con- troversy with Ehrlich in IMiinch. nied. Woch., "N'ols. 56 and 57.) 216 CHEMISTRY OF THE IMMUNITY REACTIONS with the stroma rather than with the hemoglobin, and the result of the union is to render the stroma permeable to the hemoglobin, or to separate the bonds that unite the hemoglobin to the stroma, f^" There are grounds for believing that the amboceptor not only binds the complement, but that it also produces changes in the corpuscles (Muir). Mathes^* contends that red corpuscles cannot be dissolved by hemolytic serum or by pancreatic juice until after they have been killed; as heated serum does not kill them, this is presumably done by the complement. Corpuscles that have been killed can then be dissolved in their own serum. Le- vene^^ tried to produce hemolytic serums by immunizing with different consti- tuents of corpuscles, using — (1) pure crystalline hemoglobin; (2) proteins of the stroma soluble in salt solutions; (3) an extract with alcohol-ether; and (4) an extract in 1.5 per cent, sodium bicarbonate. Only the last gave positive results, and the serum was almost devoid of agglutinative properties. Injection with corpuscles that had been digested with trypsin gave about the same results as alkaline ex- tracts; corpuscles digested by pepsin gave a much weaker serum; in neither was agglutination obtained. According to Bang and Forssmann*'^ and others ethereal extracts of red corpuscles give rise to production of hemolysins on immunization, and this "lysinogen" substance can be precipitated with acetone, is insoluble in alcohol, is not destroyed by boiling, and gives rise to no agglutinin. Numerous other olDservers, however, have failed to confirm these findings. Ford and Halsey^* ol)tained serum with both lytic and agglutinative powers by injecting either the stroma or the laked blood free of stroma. Stewart^^ obtained similar results by immunizing with corpuscles laked by physical means, by serums, or by saponin. Pure hemoglobin itself is not antigenic. ^^ According to Guerrini,*^^ nucleoprotein obtained from dog's blood engenders specific hemolysins, and Beebe states that nucleoproteins from visceral organs do not have this effect. Levene's alkaline erythrocyte extracts probably also contained nucleoproteins. Vedder,^* was unable to produce hemolysins when he used ether extracts of corpuscles as antigen, or with globulin from stroma, but the protein extract left after removing the globulin, presumably albumin, as well as lipoid-free stroma, produced hemolj-sin. The henolysin itself seems to be a globulin. On the other hand, Bennett and Schmidt^' obtained hemolysin by immunizing with the globulin precipitated from hemolyzed erythrocytes by CO2. Immunization with extracts of tissues and cells of various sorts, even when entirely free from blood (e. g., spermatozoa), may produce hemolytic sera. The fact that various tissues from many different species of animals, when used as antigen, may give rise to hemolysin for sheep corpuscles, is an interesting but so far unexplained phenomenon, which is discussed under "Specificity" (Chapter vii). The Complement. — Hemolytic complement possesses the same properties as bacteriolytic complement, resembling enzymes to the extent that it is susceptible to heat, causes a disintegration of cells, and is largely retained by Berkefeld filters.'" The joint action of amboceptor and complement is strikingly like the activation of trypsinogen by kinase. On the other hand, hemolysis by serum is quite different from the effect of trypsin on corpuscles, as trj-psin completely dis- organizes the hemoglobin and destroys the stroma, while in hemolysis the stroma and hemoglobin seem to be merely separated from one another but not chemically altered. Again, hemolysin acts quantitatively, although that may be due to a ^" Corpuscles treated with osmic acid will unite with hemolysins of diverse origin, but when used for immunizing they engender no hemolysins (Coca; also V. Szily, Zeit. Immunitiit., 1909 (3), 451). Heating corpuscle stroma alters greatly the reactivity (Landstciner and Frasek, ibid., 1912 (13), 403). " iVIiinch. med. Woch., 1902 (49), 8. " Jour. Med. Research, 1904 (12), 191. " Hofmeister's Beitr., 1906 (8) 238. " Jour. Med. Research., 1904 (11), 403. ^' Amer. Jour, of Physiol., 1904 (11), 250. 8" .Schmidt and liennctt, Jour. Infect. Dis., 1919-(25), 207. " Riv. crit. di clin. mod., 1903 (4), 561. "8 Jour. Immunol., 1919 (4), 141. " Jour. Immunol., 1919 (4), 29. '0 Muir and Browning, Jour. Path, and Bact., 1909 (13), 232. HEMOLYTIC COMPLEMENT 217 difFercnce in tlie waj' the biiidinp; to the coll occurs, rather than in the nieth(Ki of action of the complement. Landsteiner and others have suRKested that a lij)oir|al complement dissolves the corpuscle lipoids, lil)eratinK the hemoRlohin, while Neiiher^ and others have supported the hypothesis that conipieiuent is virtually a lipase which splits the lipnids out of the corj)Uscles. Hordct believes that the hemolysin causes a lesion of the stroma which chanses the resistance to osmotic influences. Complement is present in the plasma in about the same amounts as in the corresponding sera, so it is not a substance set free only by coaf^uiation of the blood (W'atanabe).'" Dick'^ has found evidence that the complement is a ferment formed in the liver, and that it causes actual proteol5'tic chanp;es. Job- ling" associates the serum lipase with the hemolytic complement.'* Ohta'* ol)served no increase in non-coagulable nitrogen during hemolysis, but Dick found an increase in the free amino acids; therefore, as yet agreement has not been reached as to whether hemolysis depends in any way upon proteolysis or lipolysis in the corpuscle stroma. Although the serum of one animal may complement the immune bodies in serum of several other varieties, and also produce lysis of many sorts of cells, it may be that not one complement does all the complementing; Ehrlich and others have asserted that one serum may contain several complements of slightly differing natures. Noguchi,'" Liebermann and Fenyvessy, and others have pointed out the striking resemblance between hemolytic complement and certain compounds of soaps or lipoids with serum proteins, and it is possible that such compounds are of importance in serum hemolysis; but there seems also to be evidence of the existence of distinct protein complements, entirely different from these,"' and it is possible that the protein complements are the important agents in specific hemo- lysis by immune sera.'* Antibodies can be obtained for both complement and hemolytic amboceptor by immunizing against serum containing them, and in manj^ serums antihemolysins exist normally. Against certain vegetable hemolysins this antihemolytic action is very strong (Kobert). Antihemolysins are generally anticomplements, but in a number of instances anti-amboceptors have been obtained. The existence of im- mune bodies specific for hemolytic amboceptor and complement, supports the view that both of these agents are proteins. In hemolysis as in bacteriolysis the complement exhibits two function^, corres- ponding to the "end-piece" and" mid-piece" fractions. Herzfeld and Klinger'^o con- sider the mid-piece to be a globulin which renders the surface of the corpuscles more "' Jour. Immunol., 1919 (4), 77. "2 Jour. Infect. Dis., 1913 (12), 111. " Jobling and Bull, Jour. Exper. Med., 1913 (17), 61; also Bergel, Deut. Arch, klin. Med., 1912 (106), 47. ~* Thiele and Embleton, however, state that hemolysin is not a lipase, and that the hemolytic power of serum has no relation to its lipolytic power (Jour. Path, and Bact., 1914 (19), 349). " Biochem. Zeit., 1912 (46), 247; see also McNeil and Kahn, Jour. Immunol., 1918 (3), 295. '« Biochem. Zeit., 1907 (6), 172 and 327; Jour. Exper. Med., 1907 (9), 436. '' SeeLiefmann, el al., Zeit. Immunitiit., 1912 (13), 150. ^'Liebermann and Fenyvessy (loc cit.y"" believe that serum hemolysis takes place as follows: First, the amboceptor acts on the corpuscle, injuring it so that it becomes less resistant; second, this combination acts upon the comple- ment (a soap compound) and frees the soap so that it can unite with the ambo- ceptor-corpuscle system; third, the soap causes hemolysis; fourth (as a separate step), the escape of the hemoglobin from the corpuscles. Tissot ascribes import- ance to the fatty acids of the plasma (Compt. Rend. Acad. Sci., 1919 (168), 1283). Bergel (Zeit Immunitiit., 1918 (27), 441) supports the hypothesis that immune hemolysis and agglutination depend on a solution of the lipoids of the cells. In this reaction the lipoids act as antigen, the new-formed amboceptor is formed by the lipoids of the lymphocj-tes as a zymogen which is activated by serum comple- ment, and is specifically bound by the lipoid antigens of the corpuscles. That is, the lipoids are the haptophore groups of the antigen; they bind the receptor of the thermostabile lipase zymogen, which is activated by the non-specific complement. '8" Biochem. Zeit., 1918 (87), 36. 218 CHEMISTRY OF THE IMMUNITY REACTIONS capable of taking up certain disintegration products contained in the serum (per- sensitization) which constitute the so-called end-piece, and which produce hemo- lysis by direct hydrolysis or solution of the stroma elements. Hemagglutinin. — Agglutination of red corpuscles occurs under the influence of immune serum as well as under the influence of some normal serums. In all respects the principles seem to be the same as those described for bacterial agglutination. The hemagglutinating antibody behaves like the other antibodies and proteins under the in- fluence of chemical and physical agencies, but Landstciner and Jagic have obtained strong agglutinating solutions containing very little protein. BergeF^ contends that hemagglutination is produced by lipase from the lymphocytes, which alters the lipoid membranes of the erythrocytes. Agglutination occurs at much lower temperatures than hemolysis, and also is not checked by heating the serum to 55° ; hence it is possible to observe hemagglutination independent of hemo- lysis. Serums may contain hemagglutinins and not be hemolytic; the'reverse is also true. The conglutmin effect of beef serum (Bordet and Gay) is also observed with corpuscles as with bacteria. As agglu- tination occurs in corpuscles that have been fixed in formalin or sub- limate' it is probably not the proteins that are affected, but some other of the ingredients of the stroma, of which lecithin and cholesterol seem to be the chief. Certain vegetable poisons produce agglutination of red corpuscles, especially ricin, abrin, and crotin, and the fact that ricin has httle or no hemolytic action shows the independence of the processes. Anti- sera for these vegetable poisons are also antiaggiutinative, acting, as Ehrlich showed, on the poison and not on the corpuscles. The seeds of many non-poisonous leguminous plants, and also of Solanaceoe, yield extracts that are strongly agglutinative for red corpuscles; in Phaseolus multiflorus the active substance is found in the proteose of the seed, and seems to be a part of the stored food (Schneider).^" It is not present in other parts of the plant. Snake venojus contain agglutinins, destroyed by heating to 75°; their agglutinating power being in inverse ratio to their hemolytic power. Corpuscles aggluti- nated by venoms may be again separated by potassium permanganate solutions. ^^ Silicic acid and certain other colloids may act as agglu- tinins, their effects bearing a relation to the effects of electrical charges upon agglutination of bacteria or of colloids {q. v.).^^ Corpuscles that have been sensitized by hemolytic amboceptors are much more readily agglutinated by salts of heavy metals, especially copper and zinc, presumably because of quantitative alterations in the electrical charge of the corpuscles induced by the antibody.'** " Zoit. Iininunitat., 1912 (14), 255; 1913 (17), 109. 8»,J()ur. liiol. Chom., 1912 (11), 47; l)il)li()grapliv. »' .See FlexntT, irniv. of Pmn. Mod. Bull., 1902 (15), .T24 and :J»)1. 82 S(>e J/indstrinor aiul .JaKJc. Aliiiich. nicd. \\'ocli., 1901 (51), 1185. 83 lOisiicr and T'lirdcinaiin, Zcit. Iiuniuiiitat., 1914 (21), 520. HEMOLYSIS HY JiACTKRIA 210 Agglutination of the corpuscles during life may bo of great patho- logical importance, for such masses of agglutinated corpuscles may readily produce capillary thrombi and emboli, which, if wide-spread, may create much disturbance. Somethmes the serum of one indi- vidual of a species agglutinates the corpuscles of another individual of the same species (isoagglutination) j^^" a fact which must be taken into account in performing transfusion of blood, lest dangerous ag- glutination take place. Agglutination of an individual's corpuscles by his own serum {autoagglutination) , may also be observed under experimental, and perhaps under pathological conditions (Land- steiner),**^ this pathological autoagglutination probably occurring especially at temperatures below 37°. (See Paroxysmal Hemoglobin- uria.) Many bacteria produce substances that are agglutinative for human red corpuscles, among them being B. typhosus, pyocyaneus, and staphylococcus. Flexner^^ has found in typhoid fever thrombi that seemed to be composed of agglutinated red corpuscles, almost free from fibrin and leucocytes. Probably many of the so-called "hyaline thrombi" found frequently in infectious diseases are really composed of agglutinated, partly hemolyzed red corpuscles (see "Thrombosis," Chap. xiii). Hemolysis by Bacteria^^ Both pathogenic and non-pathogenic bacteria produce hemolytic substances that are excreted into the fluids in which they grow. Dur- ing many infectious diseases marked hemolysis occurs, especially in those diseases accompanied by septicemia. After death the hemo- globin of the blood goes into solution, and the resulting staining of the walls of the blood-vessels, and later of the tissues everywhere, is generally familiar. In the post-mortem hemolysis probably the pu- trefactive organisms are chiefly concerned, although it is marked a very short time after death in many cases of septicemia, particularly when the infecting organism is the streptococcus, and here probably the pathogenic organism is the chief cause of the hemolysis. The hemolytic action of bacteria can be studied both in vitro and in vivo. Among the best known hemolytic bacterial toxins are tetanolysin, pyocyanolysin, typholysin, staphylolysin,^'' and streptocolysin, as they have been termed. Of tliese, the case of pyocyanolysin is question- "« Review bv Happ, Jour. Exp. Med., 1920 (31) 313. s^ See also Clough and Richter, Bull. Johns Hop. Hosp., 1918 (29), 86; Rous and Robertson, Jour. Exp. Med., 1917 (27), 509. " Univ. of Penn. Med. Bull., 1902 (15), 324; Amer. Jour. Med. Sci., 1903 (126), 202. *^ See Pribram, Kolle and Wassermann's Handbuch., 1913 (II), 1328. *" Analysis of staphj-loh^sin by Burkhardt (Arch. exp. Path, und Pharm., 1910 (63), 107), showed it to be dialyzable, protein- and biuret-free, thermolabile and soluble in ether. From B. puiidum he isolated a hemolj'tic substance which seems to be a derivative bj' oxidation of erucacic acid (oxj-diniethylthiolerucacic acid). 220 CHEMISTRY OF THE IMMUNITY REACTIONS able, because it has been described as resisting heat above the boil- ing-point, and Jordan'^^ seems to have proved that the hemolysis is ascribable to the alkalinity that this organism produces in culture- media. Other bacterial hemolysins are, however, destroyed by heat at 70° or less for two hours; but they are altogether different from ordinary immune hemolysins. Apparently streptocolysin is simph' a toxin for red cel,ls,^^ and unites directly to the cell receptois without the intervention of any intermediary body. As a similar structure has been shown for staphylolysin and tetanolysin, it is probable that the bacterial hemolysins are all merely toxins with a particular affinity for red cells, and against some of these bacterial hemotoxins antitoxic sera are obtainable, although there is usually some question as to how much of the antagonistic effect depends on true antitoxins and how much upon the cholesterol in the serum. However, a- strong antiserum has been obtained against the hemotoxin of B. Welchii.^^ Of course bacteria may also form many non-specific hemolytic substances as products of their metabolism, such as acids and bases. Secondary anemia occurring in the infectious diseases is probably to be explained largely by this hemol3^tic property of bacterial toxins. Hemoglobinuria may also be produced in the same way in some in- stances. Intravenous injections of filtrates of the saprophyte, B. megatherium, will produce hemoglobinuria in guinea-pigs, hence hemo- lysis is not an exclusive property of pathogenic bacteria, and with streptococci LyalP^ found that the hemolysin titer did not afford a criterion of virulence. No immunity to streptococci is produced in animals immunized with streptococcus hemolysin.^- Pneumococci produce an intracellular hemolytic toxin which is very labile and antigenic; living pneumococci convert hemoglobin into methemoglobin, but this the hemolytic extracts of pneumococci cannot do (Cole).^^ Streptococcus viridans has the same propertj^,^* which may play a part in the effects of infections with these organisms, von Hellens^* states that streptocolysin is ether soluble and heat resistant. Hemolysis by Vegetable Poisons A number of plant poisons are strongly homolj'tic, and some of them owe much of their to.xicity to their effect on the erj'throcytes. One group consists of the bodies often called "vegetable toxalbu- mins," because they seem to be proteins, and includes ricin, abrin, crotin, curcin and robin. ^^ Of these, crotin and curcin are particu- «8 Jour. Medical Research, 1903 (10), 31. «9 Jour. Amer. Med. Assoc, 1903 (41), 962; Jour. Infect. Dis., 1907 (4), 277. s" Ford and Williams, Jour. Immunol., 1919 (4), 385. »' Jour. Med. Kes., 1914 (30), 515. »2 McLcod and McXeo, .lour. Path, and Bact., 1913 (17), 524. »3 Jour. J']xper. Med., 1914 (20), 347, 3ti3. »' liiakc, Jour. Expcr. Mod., 1910 (24), 315. . »" Cent. f. Bakt., 1913 ((iS), 002. ^^ The sap of (Cotyledon Schcidcckeri contains hemolytic substances of peculiar character. (See Kritchevvski, Jour. Exp. Med., 1917 (20), 069.) HEMOLYSIS BY SAPONINS 221 larly actively licuiolytic, while riein, abiin, and robin are more marked by their agglutinating action, hemolysis being produced only by relatively large doses. Their effects vary greatly, however, according to the species of animals whose blood is used. They resemljle the bacterial toxins, in that immunity can be secured against them, and the immune serum will prevent their hemolytic action. Heating the toxalbumins to 65° or 70° does not destroy the hemolytic or agglutinating action except with phallin, but 100° does. The action of these substances is not like that of the enzymes, in that it is quantitative, a given amount acting on a given amount of cor- puscles to which it is bound. Madsen and Walbum" observed that red corpuscles had the power of dissociating neutral mixtures of ricin and antiricin, the ricin entering the corpuscles from which it could be recovered. ^'^ Ford and Abel believe the hemolytic agent of amanita to be a glucoside. (The general nature and other properties of these substances are considered under the heading of " Phytotoxins," in Chap, vi.) Saponin Group. — Another quite distinct group of vegetable hemolyzing agents consists of the "saponin substances. "^^ These are a closely related group of glucosides, found in at least 46 differ- ent families of plants, and they are strong protoplasmic as well as hemolytic poisons. They differ altogether from the true toxins, be- ing heat resistant, having no resemblance to proteins, and not giving rise to antibodies on immunization of animals.^ The degree of their toxicity is not directly proportional to their hemolytic activity; they seem to injure chiefly the nerve-cells. Apparently hemolysis is brought about by action upon the lipoids of the red corpuscles, for addition of cholesterol to saponin prevents its hemolytic effect;^ leci- thin does not have the same property.^ Both cholesterol and leci- thin combine with saponin, the cholesterol compound being quite inert, whereas the lecithin compound is both hemolytic and toxic. The compound formed between a typical saponin, digitonin, and cholesterol, is so insoluble that it has been found useful in the quan- titative analysis of cholesterol.'* Normal serum seems to contain " Cent. f. Bakt., 1904 (36), 242. ^^ According to Pascucci (Hofmeister's Beitr., 1905 (7), 457), ricin combines directly with lecithin, the compound being strongly hemolytic. ^^ Complete literature on saponin given by Kobert, "Die Saponinsubstanzen," Stuttgart, 1904; also Kunkel, "Handbuch der Toxokologie," Jena. 1 Saponins are characterized by their ready solubility in water and the foam- ing, soapy character possessed by the solution; hence their technical applications as soap bark, etc. Heated with dilute acids they split off sugar; also when acted on by glucoside-splitting enzymes (from spiders), according to Kobert. Saponin from Quillaja (soap-bark) has the formula C19H30O10 (Stiitz). Most are colloids, but some crystallize. 2 Ransom, Deut. med. Woch., 1901 (27), 194; Madsen and Xoguchi, Cent. f. Bakt., 1905 (37), 367; Pascucci, Hofmeister's Beitr., 1905 (6), 543. 3 Noguchi, Univ. of Penn. Med. Bull., 1902 (15), 327; Meyer, Hofmeister's Beitr., 1908 (11), 357. ^ Windaus, Chem. Berichto, 1909 (42), 238. 222 CHEMISTRY OF THE IMMUNITY REACTIONS an antihcmolysin for saponin, and therefore hemoglobinuria is not produced by all saponins on intravenous injection. Careful immu- nization leads to a slight increase in this antihemolytic action of the serum, possibly due to an increased formation of cholesterol (Ko- bert). The resistance of corpuscles to saponin hemolysis varies in disease, being especially low in jaundice (M'Neil).^ A study of the toxicity of the members of this group by Kobert^ shows that in general they have similar properties, but that minor differences exist between them. All cause hemolysis, some in dilu- tion as great as 1:100,000. Some produce hemoglobinuria when in- jected intravenously, others do not. All paral3^ze the heart, but the injuries to the central nervous system are the chief cause of death. Marked local changes are produced at the site of injection, but the leucocytes are apparently not injured, although sterile suppuration is produced. There is a period of latency after intravenous injection of small doses — twenty-four hours or more — before the tappearance of symptoms. Sapotoxin is one of the most actively toxic and hemolytic products of quillaja. Cyclamin is also a member of this group (derived from Cyclamen), and is said to be the most active of all as a hemolytic agent (Tufanow). SoLANiN^ is obtained from all parts of the potato plant, combined with malic acid; it is found particularly in young sjjrouts, but not in any considerable amounts in normal potatoes.^ Its formula is unknown ,but as it splits up into an alkaloid {solanidin) and sugar it is called a gly co-alkaloid. In its action it resembles the saponins, being a powerful protoplasmic poison, killing bacteria, and hemolyzing blood in very great dilutions." A great number of hemolytic poisons are obtained from poisonous mushrooms. Best known of these is: Helvellic Acid, from Helvella esculenta, which has the empiric formula CiaHiiiO?.'" Intravenously injected it produces hemoglobinuria and icterus, with hemoglobin infarcts in the kidneys (Bostrocm)." Phallin, or Amanita hemolysin, described by Robert as a toxalbumin, has been found iDy Abel and Ford to be a glucoside, and thus belongs to the saponin group. (See Chap. vi. for further discussion.) In the leaves of the ivy, Hedera helix, a hemolytic glucoside has been found by Moore. i- It is of interest that Faust believes the poisonous agent of cobra venom to be a glucoside, closely re- sembling sai^otoxin. As will be seen, all these last-mentioned vegetable hemolytic agents are essentially different from either the bacterial or scrum hemol.ysins, or from the abrin, ricin, crotin, or robin group, in that the}' are of relatively simple chemical composition, and quite unlike proteins, en- zymes, or toxins. The manner in which they cause hemol^'sis is unknown, but from their relation to saponin it is probable that, like '•> Jour. Path, and Bact., 1910 (15), 56. 6 Arch. exp. Path. u. Pharm., 1SS7 (23), 233. 'Literature, see Meyer and Schmiedeberg, Arcii. f. exi). Path. u. Pharm., ISOo (36), 361; Perles, ibid.,'\mO (2{)), SS. * See Kunkel, "Handbuch der Toxokologie," p. 873. " Concerning human solanin jjoisoning see liothe, Zeit. f. Hyg., l*.)19 (88), 1. lOBoehm and Kiilz, Arch. exp. Path. u. Pliarm., 1S85 (19), -103. 11 Deut. Arch, kliii. Med., 1883 (32), 209. 12 Jour. Pharnuicol., 1913 (4), 263. . HEMOLYSIS BY VENOMS 223 it, they cause injury by coniljinin^ witli or dissolvinji; the lijjoids of the stroma of the corpuscles. Extracts of Morchella esculenta do not hemolyze corpuscles in vitro, although powerfully heiiioh'tic when injected into animals, and causing severe hemoglobinuria; so that it is probable that they cause their hemolj'tic effects indirectly through the changes which they produce in the tissues of the poisoned animal.'-^ Hemolysis by Venoms'^ The laking of blood-corpuscles by venoms is of peculiar interest from the standpoint of immunity phenomena, since it was demon- strated by Flexner and Noguchi that the hemolytic principle of the venoms resembles an amboceptor, in that some substance behaving like complement has to be furnished by the blood. Kyes found that this complementing agent is lecithin,^" and was able to produce what he considers to be compounds of the hemolysin with lecithin, called "lecithids." The hemolytic activity of these lecithids is very great, and they seem to be free from the neurotoxic principle of the venoms. Whether they represent true compounds of a hemoh'tic amboceptor with lecithin, or are simply actively hemolytic products of the cleavage of lecithin by an enzymatic activity of the venom, is at present unsettled ;^^ it seems probable, however, that the hemolysin of cobra venom is a lipase that splits lecithin into two hemoh'tic components, oleic acid and " desoleolecithin " (Coca).'^ Noguchi suggests that not only lecithin, but also soaps, especially of unsaturated fatty acids, and probably protein compounds of soaps and lecithin, may act as the hemolytic "complement" which activates venoms. The hemolytic agents of venom seem to be secreted by the salivarj- glands of the reptiles from their blood, which contains almost identical amboceptors, differing chiefly in that they can be activated onl}' by agents contained in snake blood, while the amboceptors of venom can be activated b}- nearlj' all sorts of blood. Venoms from cobra, rattle- snake, moccasin, and copperhead possess in each a varietj' of inter- mediary bodies (amboceptors) that seem to be at least partly identical in nature, although they may varj^ in quantity. In order of decreasing hemolytic power for mammalian corpuscles come venoms from cobra, water moccasin, copperhead, and rattlesnake. These venoms are also agglutinative for all corpuscles tried, and agglutination will occur at 0° C. Exposure for thirty minutes at 75°-80° C. destroys the agglutinating property. In general, the hemolytic power of the 1' Friedberger and Brossa, Zeit. Immunitat., 1912 (15), 506. '^ General review of literature on the hemolytic properties of animal poisons given bv Sachs, Biochem. Centralblatt. 1906 (5), 257; Noguchi, Jour. Exp. IMed., 1907 (9), 436. 15 Cruickshank also found that other lipoids than lecithin may activate cobra venom (Jour. Path, and Bact., 1913 (17), 619). 16 See Kyes, Jour. Infect. Dis., 1910 (7), 181; v. Dungern and Coca, ibid., 1912 (10), 57; Manwaring, Zeit. Immunitat, 1910 (6), 513; Bang, iUd., 1910 (8), 202; Coca, Jour. Infect. Dis., 1915 (17), 351. 224 CHEMISTRY' OF THE IMMUNITY REACTIONS venoms for different sorts of corpuscles varies in inverse propor- tion to their agglutinative power. The hemolytic intermediary bodies are resistant to heat, suffering but slight loss of power at 100° C. Red corpuscles of the frog are not hemolyzed by venom, and those of necturus (mud pupp}') but slightlj', agreeing with the known resist- ance of cold-blooded animals to snake-bites. The erythrocj^tes of different individuals show considerable varia- tions in their resistance to hemolytic agents, perhaps depending upon tile amount or upon the manner of fixation of the lipoids in the cor- puscles; thus the corpuscles of syphilitics show a heightened resist- ance to hemolysis by cobra venom (Weil)'^ except in the earliest stages, when they are hypersensitive. Also, the serum of persons suf- fering from various diseases, especially mental diseases, inhibits the hemolysis of human corpuscles by cobra venom. '^ After splenectomy there is an increased resistance to venom hemolysis.'^ Eel serum is remarkably hemolytic, so much so that a quantity' of 0. 1 c.c. per kilogram of body weight will kill a rabbit or guinea-pig in three minutes when injected intravenously. Heating at 54° C. for fifteen minutes destroys the hemo- lytic action, and, unlike ordinary serum hemolysins the addition of complement does not restore its activity. Animals can be immunized against this serum. In- troduced into the stomach in ordinary quantities eel serum is not to.\ic. It can be dried and redissolved without losing its activity, but acids and alkalies readilj' destroy it. Mosso, who first discovered the toxicity of eel serum, called the un known active principle ichthyotoxin. It is found chiefly in the albumin fraction of the eel serum. ^^ Many other animals produce hemolytic poisons (e. g., spiders, bees) which are discussed under Zootoxins, Chapter vi. Hemolysis in Disease During health there is alwaj^s going on a certain amount of de- struction of red corpuscles that have outlived their usefulness; hence in disease we may have to deal with either an alteration in the nor- mal processes of blood destruction or the introduction of entirely new processes. Although the place and manner of normal red corpuscle destruction is'inot completel}^ known, yet it seems probable that there is relatively little hemolysis within the circulating blood. "When a red corpuscle becomes damaged, it seems to become more susceptible to phagocytosis, and it is then picked out of the blood, chiefl}' by the endothelial cells of the sinuses of the liver, spleen, hemolymph glands, and bone-marrow. Within these cells it apparently undergoes hemo- lysis. Eventually, the resulting pigment is split up b}' the liver, the non-ferruginous portion forming the bile-pigments, while the iron seems to be mostly withheld to be worked over into new hemoglobin.-' >' Jour. Infect. Dis., 1909 (6), 688; Stone and Schottstaedt, Arch. Int. Med., 1912 (10), 8. *" See articles on this subject in the Miinch. med. Woch., 1909, Vol. 56. 1" Kolmer, Jour. Exp. Med., 1917 (25), 195. 2»Sato, Kyoto .Jour. Med. Sci., 1917 (14), 36. 21 Muir and Dunn (.lour. Path, and Bact., 1915 (20), 41), find that after acute hemolytic anemia in ral)bits the excess iron stored in the organs lias been nearly all aljsorbed by the time regeiuMatioii of tlu> blood is complete. HEMOLYSIS l.\ DISKASE 225 (Seo "Pisincntutioii," (hap. xviii.) Whoiicvcr (luring disca^; red corpuscles arc more rapidly injured than they are under normal condi- tions, these processes of normal hemolysis are exaggerated and we not only find the phagocytic cells of the spleen and glands packed with corpuscles, but endothelial cells elsewhere, and also leucocytes, take on the hemolytic function. At the same tinie there results an exces- sive production of bile-pigment from the destroyed red corpuscles, which has an etiological relation to the so-called "hemato-hepatogen- ous" jaundice. If hemolysis is very excessive, the blood pigment accumulates in other organs than the liver and spleen. According to Pearcc^^ and his associates, when the blood contains at one time more than 0.06 gm. of free hemoglobin per kilo of body weight, it begins to be excreted by the kidneys; smaller amounts are cared for chiefly by the liver, and even when much larger amounts of hemoglobin are present in the blood the liver takes care of most of it, only a rela- tively small proportion, 17 to 36 per cent, being excreted in the urine. Hence it is possible to have hemolytic jaundice without hemoglobin- uria. Part of the pigment is converted into urobilin, and the amount of this pigment in the stool is an index of the amount of hemolysis.-^ In persons with hemolytic hemoglobinemia, intravenous injection of hemoglobin will produce hemoglobinuria with smaller dosage than in normal persons, who require at least 17 c.c. of laked corpuscles to pro- duce hemoglobinuria.-^ It is possible that the globin, which is quite toxic when free,'^ may play a part in the symptomatology of hemolytic poisons. The stroma of the erythrocytes also seems to be toxic. ^^ The hemolj^sis of the acute febrile diseases is readily explained by the demonstrable hemolytic property of the products of the organisms that cause them, such as streptocolysin, staphylolysin, etc. Perhaps at the same time products of altered metabolism may also play a part, but it does not seem probable from experimental results that the thermic condition per se has much effect. In malaria, although the parasites enter and destroy the corpuscles in which they live, yet this alone does not account for all the blood destruction of the disease, for the amount of anemia is quite without relation to the number of parasites to be found. There is good reason to believe that the Plasmodia produce hemolytic substances that are chschargcd into the serum. In the primary anemias hemolj^sis seems to be the essential process, although the agents involved are at present unknown. Absorption of hemolytic products of intestinal putrefaction or infection has always come in for much suspicion, without ever becoming completely "Jour. Exp. Med., 1912 (16), several articles. " See Robertson, Arch. Int. Med., 1915 (15), 1072. " Sellards and Minot, .Jour. Med. Res., 1916 (34), 469. " Schittenhelin and Weichardt, Miinch. med. Woch., 1912 (59), 1089. -« Barratt and Yorke, Brit. Med. Jour., Jan. 31, 1914. 226 CHEMISTRY OF THE IMMUNITY REACTIONS established. Here also the hemolysis seems to take place in the endo- thelial cells rather than in the vessels. In such a disease as pernicious anemia there is much reason to assume that defective or abnormal hematogenesis is an important factor. Probably the anemia of nephritis is at least partly the result of hemolytic action of the retained products of metabolism, in which connection the hemolytic properties of ammonium compounds may be recalled. In some diseases associated with anemia it has been found that the blood-serum of the patient is cUstinctly isohemolytic, although isoagglutination seems to be more frequent. The fluids that can be obtained from cancers have been found to be hemolytic, while antihemolysin has been found in ascitic and pleural effusions. Autolytic disintegration of hver, and presumably other tissues, may also cause the presence of hemolytic substances in the blood. ^^ Arseniuretted hydrogen may produce hemolysis in some such way, since it causes no hemolj^sis in the test tube (Heffter). The very great hemolytic action of soaps and free fatty acids, which varies directly with the number of unsaturated carbon atoms they contain (Moore^^), makes it possible that these substances play a part in the hemolysis of disease, especially since the fatty acids of the liver are characterized by their high content of free fatty acids. Bile is strongly hemolytic, and in icterus this is an im- portant consideration. In many forms of poisoning hemolysis is a prominent feature; in some it seems to be the chief effect of the poison, e. g., potassium chlorate and arseniuretted hydrogen. In severe extensive burns there may occur hemolysis, and hemoglobinuria may also result. The hemo- globinuria of "blackwater fever" has been the cause of much discus- sion as to whether the malarial parasite or the quinine is the cause, with a divided opinion resulting, although, undoubtedly, cases do occur in malaria without administration of quinine. The studies of Brem-® indicate that the hemolysis is produced by a hemolysin coming from the Plasmodium, and that the quinine influences the condition by preventing the action of an antihemolytic substance present in the blood. After removal of the spleen, hemolysis by the hemolymph glands exceeds that of the primitive spleen, causing an excessive destruction of red corpuscles (Warthin^"). This suggests that the spleen may normally dispose of some hemolytic agent which acts either by stimu- " Maidoni, Biochoin. Zeit. ,1912 (45), 328. lieinolytic lipoids are believed by some to l)e liheruted from injured tissues (see Kirsche, Hiochem. Zeit., 1913 (55), 169), but McPhedran (Jour. Exp. Med., 1913 (18), 527) could find no evidence that any particularly hemolytic fatty ucid, more active than oleic acid, can be isolated from either normal or diseased tissues. "^^ Brit. Med. Jour., 1909 (ii), 684; sec al.so Lamar, Jour. Exper. Med., 1911 (13), 380. 2» Arch. Int. Med., 1912 (9), 129. a" Jour. Med. Research, 1902 (7), 435. PAROXYSMAL HEMOGLOBINURIA 227 lating phagocytosis or by so altering the red cells that they are particularly susceptible to phagocytosis. This idea is not substanti- ated by the work of Pearce,^^ who found the anemia of splenectomy accompanied by an increased resistance of the corpuscles to hemolysis, and no hemolytic agent was present in the blood. There also occurs the group of anemias associated with great enlargement of the spleen, and in which removal of the spleen may result in a return to nofmal blood conditions; a fact suggesting, among other possibilities, that there may be poisons which stimulate directly the hemolytic action of the spleen independent of the natural stimulation of splenic hemoly- sis which comes from the presence in the splenic blood of injured red corpuscles. ^- Resistance to hemolysis varies greatly in disease conditions'^ and often specific- ally,— i.e., resistance may be increased to one agent, decreased for another, and normal with a third. Attempts have been made to use this resistance as a diag- nostic or prognostic index, but not with great success in most cases. Apparently changes in the plasma lead to alterations in the permeability of the corpuscles, which determines their behavior with hemolytic agents; also changes in the pro- portion of lipoids and hemoglobin may modify hemolysis. As an example of this condition may be cited observations on hemolysis by cobra venom, the cor- puscles having been found less resistant in dementia precox, more resistant in carcinoma and syphilis. Butler" states that fragility of the corpuscles is abnor- mally high in exophthalmic goiter, cancer, syphilis, tabes, anemia and malaria. In obstructive jaundice the corpuscles show an increased resistance to hemolysis by hypotonic salt solution, but in congenital hemolytic jaundice the resistance is decreased.'^ Using saponin hemolysis, Bigland'^ found the resistance greatly decreased in icterus, although the serum had an increased protective action because of antagonism between the saponin and the bile salts; in all anemias resistance was found increased, except pernicious anemia, which showed normal or slightly subnormal resistance; high temperature decreases resistance. As will be seen from the few examples cited, the resistance to different hemolj'tic agents may vary with the same corpuscles.'^ Paroxysmal Hemoglobiniiria.^^ — This condition seems to depend upon the presence in the serum of a hemolytic amboceptor (an auto- hemolysin) , which will combine with the corpuscles of the same indi- vidual and sensitize them for his own complement (Donath and Landsteiner, Eason) . This au^ ohemolysin can react with the corpuscles only at low temperature, such as may be furnished in the peripheral vessels by exposure to cold, and the complement unites when the tem- perature of these corpuscles again reaches 37° in other parts of the '^ Pearce ct al, Jour. Exp. Med., 1912, ^'ol. 16. See also Roccavilla, Arch. Med. Exp., 1915 (26). 508. " See Banti. Semain Med., 1913 (33), 313. '3 Keview bv Paltauf, Krehl and Marchand's Handb. allg. Pathol., 1912 (II (1),) 83. '* Quarterly Jour. Med., 1913 (6), 145. " See Richards and Johnson, Jour. Amer. Med. Assoc, 1913 (51), 1586 Giffin and Sanford. Jour. Lab. Clin. Med., 1919 (4), 465. '6 Quart. Jour. Med., 1914 (7), 370. " Bibliography by Krasny, Folia Hematol., 1913 (16), 353. '' Landsteiner, Handbuch d. Biochem., Vol. 2 (1), p. 492; Meyer and Emmerich, Deut. Arch. klin. Med., 1909 (96), 287. 228 CHEMISTRY OF THE IMMVXITY REACTIOXS body. In susceptible persons attacks of hemoglobinuria may be brought on merely by holding the hands in cold water, and their blood serum will sensitize to hemolysis human corpuscles (even of normal individuals),^^ m vitro at low temperatures.""* Certain infections, es- pecially syphilis, ^^ predispose to paroxysmal hemoglobinuria. Not only the hemolytic amboceptors, but also an auto-opsonin is present (Eason) and the resistance of the red corpuscles is decreased to various harmful agencies, including COo and other weak acids.*- The cor- puscles of three cases studied by Moss'*^ showed an increased resistance to hypotonic NaCl solutions. Just before the rigor, hemolysins may be found in the blood, disappearing after the hemoglobinuria (Rob- erts).** In a case studied by Dennie and Robertson,*^ hematuria resulted from destruction of only 6.3 c.c. of the patient's blood, and 90 per cent, of the liberated hemoglobin was excreted within two hours. There also occur conditions in which aiito-aggluti7iati on occurs without hemolysis when the blood is cooled.*'' Pathological Anatomy in Hemolysis. — The lesions produced in the organs of animals poisoned with hemolytic agents are usually pronounced and quite char- acteristic. There is often a subcutaneous edema, which is usually- blood-stained, and similar fluid may be present in the serous cavities. The fat is yellowish, and the muscles are darker in color than is normal. The spleen is usually much swollen, soft, friable, and very dark in color. The liver is usually swollen and mottled with red areas in a yellow background. The renal cortex is dark in color, even chocolate-colored, and the pyramids are comparatively light; hemo- globin is frequentlj'^ present in the urine. In the lungs are often found hemor- rhages or areas resembling small infarcts. The blood may be thin and even distinctly transparent. Microscopically the red corpuscles are found in all condi- tions of degeneration, and often fused together. In the liver, besides patches of congestion, fatty changes are present if the animal lives long enough. Large phagocytic cells packed with red corpuscles are abundant in the spleen and lymph- glands, as well as diffuse accumulations of the blood-cells, which are often fused; and much pigment is also present, both free and in the cells. Pigment also accum- ulates in the renal epithelium, which often shows much disintegration; congestion is prominent, and hemorrhages into both interstitial tissue and glomerules are frequent. Some of the lesions are due to the hemolysis, and some to the associated agglutination of corpuscles, which form hyaline thrombi. Pearce"" lias found that agglutinative serum when injected into dogs causes widespread necrosis in the liver, which is followed by proliferation of connective tissue and the production of changes re.seml)ling cirrhosis. There is a marked decrease in the glycogen content of the liver, and of its lipolytic activity' (Andrea).'" 39 See Lorant, Deut. Arch. klin. Med., 1918 (126), 148. *" Widal looks upon paro.\ysmal hemoglobinuria as an autoanajjliviaxis (Semain M6d., 1913 (33), 013). *' Matsuo, Arch. f. klin. Med., 1912 (107) 335. « lierghausen, Arch. Int. Med., 1912 (9), 137. " Folia Serologica, 1911 (7), 1117. ** Brit. Med. Jour., 1915 (2), 398. *' Arch. Int. Med., 1915 (lb), 205. '"' See Rous and Robertson, Jour. Exp. Med., 191N (27^, .")()3; C'lough and Richter, Hull. Johns Hop. Hosp., 191S (29), 80. "".Jour. Kx]). Med., 1900 (8), 04; Jour. Med. IJesearcli. 1900 (14). 541. *' Arch, internal. i)harmacodyn., 1905 (14), 177. COMPLEMENT FIXAT/(>.\ liKM'TlOS 229 COMPLEMENT FIXATION"' AND WASSERMANN REACTIONS'' The original principle involved in these reactions was first demon- strated by Bordet and Gengoii, and is essentially as follows: If a specific antigen and amboceptor unite in the presence of complement, the complement is then united to the amboceptor-antigen compound to complete the reaction. When sufficient amounts of amboceptor and antigen are present the entire quantity of available complement may be thus fixed, and, consequently, the mixture contains no more complement for further reactions. As complement does not ordinarily unite with amboceptors except when the amboceptors are united with their specific antigens the fact that in a given system of complement + amboceptor + antigen \ there is no free complement, is evidence of a reaction between ambocep- tor and antigen; in consequence of which this reaction can be used to determine the presence of a specific amboceptor in a serum^ by using the corresponding antigen; or conversely, with a scrum containing a known amboceptor we can detect the presence in a solution of the specific antigen. The indicator of the presence or absence of comple- ment which is in universal use, is the ability of the mixture to hemolyze erythrocytes in the presence of the specific hemolytic amboceptor. Thus, if typhoid bacilli and a typhoid antiserum which contains both complement and specific amboceptor, are mixed in proper proportions and incubated for a short time, the complement will be bound to the bacilli. If we then add this mixture to sheep corpuscles which have been acted upon by an antisheep-corpuscle serum, from which the complement had been previously removed by heating, no hemoh'sis will occur, for we have added no free complement. But if our original mixture had contained dysentery bacilli instead of typhoid bacilli the complement would not have been fixed, and the addition of this mixture, containing free complement, to the sensitized sheep corpuscles would cause prompt hemolysis. This reaction was at first used for the detection of antibodies in ■•^ The reaction of ''complement fixation" must not be confused with the eu- tirelj' distinct reaction of ''complement deviation," a mistake very likely to ha])pen because of the careless but erroneous use by some writers of the latter term in describing the first-named reaction. Complement deviation (or Neisser-Wechs- berg phenomenon) is produced when an excess of amboceptors is present together with antigen and a limited amount of complement, which results in absence of complement activitv. The mechanism of this reaction lias not been satisfactorily explained. Thjotta (Norsk. Mag. Laegevid., 1919 (80), 1051) believes it to de- pend on some special substance, distinct from the known antibodies, which adsorbs complement. ■"Literature given by Meier, Jahresbor. d. Immunitatsforsch., 1909 (4), 58; Sachs and Altmann, Kolle and Wassermann's Handbuch, Ergiinzungsbd.. 2, 1909, p. 476; Noguchi, "Serum Diagnosis of Svphilis and Luetin Reaction,"" I'hiladel- phia, 1912. 230 CHEMISTRY OF THE IMMUNITY REACTIONS sera,^° and for the identification of bacteria, and was found to be ex- quisitely delicate, detecting most minute amounts of antigens with the sharpest specificity limits of any of the immunity reactions. On account of the delicacy of this reaction it can be used to determine the presence in tissues of specific organisms which cannot be culti- vated; thus, it has been possible to demonstrate the existence of a specific scarlatinal virus^^ in the tissues during this disease, although the actual organism cannot be isolated. This fact led Wassermann to use extracts of the livers of congenital sj'-philitic fetuses, which contain great quantities of spirochetes, as an antigen for complement fixation reactions, whereby it should be possible to determine in a given serum the presence of specific amboceptors for the virus of syphilis, such amboceptors being present in persons infected with syphilis as a result of the reaction to the infection. As originally in- troduced, then, the Wassermann reaction was supposed to be simply a specific reaction between syphilitic antigen, specific syphilitic am- boceptors, and non-specific complement. It was soon learned, how- ever, that the reaction as it occurred in syphihs was decidedly different from the original complement fixation reaction of Bordet and Gengou, for it was found possible to substitute in the reaction for extracts of tissues containing syphilitic virus (spirochetes), the most varied sorts of tissue extracts, coming from tissues certainly free from spirochetes (e. g., ox heart). Noguchi and Bronfenbrcnncr^^ summarize the present state of the matter in these words: "We know merely this: that complement in the presence of syphilitic antigen may be rendered inactive by one or more substances in the body fluids of a syphilitic or parasyphilitic patient." Extended investigation of these non-specific antigens which give specific complement fixation with S3^philitic sera, has shown them to be related to the lipoids, especially the lecithins, as indicated by the fact that the most efficient antigens contain the accton-insolublo frac- tion of the tissue lipoids. The antigenic value of this fraction of different liver extracts varies nearly directly with its power to com- bine with iodin*^ (Noguchi and Bronfenbrenncr), which indicates that the unsaturated fatty acids arc important in the reaction.^'' 'S" Accoi-fling to Gay (Univ. of Calif. Publ., Piithol., 1911 (2), 1, full discus- sion) conipleinent fixation i.s produced by an antincn-antihody I'oinjjlex distinct from precipitinof^en-precipitin, but Dean (Zeit. f. Ininumitat., 1912 (13), 81) be- lieves that they represent two phases or stages of the same reaction. Thiele and Embleton (Zeit. luuaunitat., 1913 (10), 430) consider tliat in syphilis it is not a specific antibody, but an anti-complementary substance which arises from the disintegrating tissiies. ^' Koesslerand Koessler, Jour. Infec. Dis., 1912 (9), 3GG. " Jour. Exp. Med., 1911 (13), 43. '^ Not corroborated by Hrowning, Cruickshank and (iilmour.'''' ^* An interesting observation made l)y Noguchi and Hrt)nfenbrenner, is that ex- tracts from fatty livers are alnut.st devoid of antigenic projuMties; but So (Cent. f. Bakt., 1912 (03). 43.S) found that the extract from fatty hearts of guinea-pigs was more active than from normal hearts. COMPLEMENT FIXATION REACTION 231 Cnule lecithins from different sources vary in efficiency, heart lecithin being more active than liver lecithin, brain and egg yolk lecithin follow- ing. Pure lecithin is not effective, the activity of lipoid solutions depending upon some other substance which is difficult to separate from lecithin (MacLean).^^ Addition of cholesterol to the lipoid solutions increases greatly their activity. ^^ An acetone-precipitated "antigen" of this class is not a true antigen, however, for fixation antibodies are not developed in animals injected with such a lipoid which has been shown to be entirely efficient in the Wassermann reaction. ^^ As for the substance in the syphilitic serum which participates in the Wassermann reaction, it would seem to be related to the globulins, which are decidedly increased in the blood and spinal fluid^'' of syphili- tics,^** especially the euglobulin.^'' P. Schmidt^' ascribes the reaction to the physico-chemical properties of the globulins of the syphilitic serum, which, he believes, possess a greater affinity for the colloids of the antigen than normal globulins; this affinity is held in check in normal serum by the albumins of the serum, which are relatively or absolutely decreased. That physico-chemical factors do play a part is evidenced by the common observation that the turbidity of the antigen suspension is closely related to its efficiency, clear solutions being less active. Slight changes in H-ion concentration will change a reaction from negative to positive, or reverse; and neutral salts can change a negative to a positive reaction, but not the reverse (Gumm- ing.) "^^ The lipoids in syphilitic sera are said by Peritz^^ to be increased, but the lipoid content and the antibody titer do not show any constant relation (Bauer and Skutezky).''^ The cholesterol con- tent of syphilitic blood shows no evidence of a quantitative relation tb the Wassermann reaction. ^^ Friedemann^^ believes that a globulin- soap compound is the active substance in syphilitic sera. Mcintosh" says that the active component differs from typical antibodies in not ^^ Monographs on Biochemistry, "Lecithin and Allied Substances," London, 1918. 5" Browning et al, Zeit. Immunitat., 1912 (14), 284; Jour. Pathol, and Bact., 1911 (16), 135 and 225. Klein and Fraenkel believe the "antigen" of ox heart extracts to be a combination of lecithin with cholesterol and small amounts of a soap-like substance similar to jecorin (Miinch. med. Woch., 1914 (61), 651.) " Fitzgerald and Leathes, Univ. of Calif. Publ., Path., 1912 (2), 39. «8 Pfeiffer, Kober and Field, Proc. Soc. Exp. Biol., 1915 (12), 153. " See Rowe, Arch. Int. Med., 1916 (18), 455. «« MuUer and Hough, Wien., klin. Woch., 1911 (24), 167. " Zeit. f. Hyg., 1911 (69), 513. See also Hirschfeld and Klinger, Zeit. Immuni- tat., 1914 (21), 40. " Jour. Infect. Dis., 1916 (18), 151. " Zeit. exp. Path., 1910 (8), 255. «* Wien. klin. Woch., 1913 (26), 830. «5 Weston, Jour. Med. Res., 1914 (30), 377; Stein, Zeit. exp. Med., 1914 (3), 309. « Zeit. f. Hyg., 1910 (67), 279. 6^ Zeit. Immunitat., 1910 (5), 76. 232 CHEMISTRY OF THE IMMIMTY REACTIONS passing through collodion or porcelain filters, and there are many who hold that the reacting substance is a product of tissue disintegration. Wassermann^^ has found evidence that the antibody is derived from the lymphocytes, at least in the spinal fluid of syphilitics. Whether true antibodies are concerned in the Wassermann reaction is a question. In favor of this view is the fact that the serum of rabbits immunized with congenital syphilis livers contains an anti- body giving the Wassermann reaction, exactly like the serum of syphilitics.'^^ On the other hand, the actual substance of pure cultures of spirochetes does not ordinarily act as antigen with syphilitic sera in the Wassermann reaction (Noguchi). It is highly probable that wdien syphilitic liver extracts are used as antigen in the Wassermann reaction, we have a true Bordet-Gengou reaction of complement fixa- tion with the syphiUtic substance present in this extract, in addition to the reaction which is accomplished by the lipoids. Whether the complement is destroyed by enzymes, '^'^ or is inhibited by anti-complement present in syphilitic serum, or is destroyed by some toxic substance in the serum^^ are matters still under discussion. A favorite interpretation of the Wassermann reaction, which seems to harmonize with the known facts, is that there is a precipitation of serum globulin by the lipoidal colloids of the antigen, and adsorption of the complement by this precipitate. Apparently the globulins of the serum in syphilis are altered in some specific but as yet unknown way, whereby they acquire in greatly increased degree the capacity to form this adsorbent precipitate."- Alterations in the lipoids also seem to exist, for it is known that con- ditions that modify the serum lipoids also modify the reaction. There seems little doubt that the reaction is not chemical but physical, and the union of complement to antibody follows essentially the laws of adsorption. Also its intimate relation to the precipitin reaction seems to be established (Dean)."^ The changes in cliaracter of the blood serum in sypliilis are sufficient to give not only immunological but also frank chemical or physico-chemical manifesta- tions. For example, Bruck'^ states that the precipitate obtained when nitric acid is added to syphilitic serum is more abundant and of a characteristic gelatin- ous appearance. "Platinum cloride also produces a heavier precii>itate in syphilitic sera (Brown and Iyengar).''' The globulin responsible for the Wassermann reaction is said to precipitate more readily l)y ammonium sulphate and other re- "* Wa.ssermann and Lange, Berl. klin. W'och.. 1914 (51), 527. "" Citron and -Munk, Deut. med. Woch., 1910 (30). 15(30; Eiken, Zeit. Imnuini- tiit., 1915 (24). ISS. ■» Manwaring. Zeit. f. Immunitiil., MtOi) CM, .^09. " Kiss. (7>iV/., 1910(4). 70:i. "See Nathan, Zeit. Iminunitat., 191S (27>, 219; Walker, ,Iour. Path, Bact., 1917 (21), 184. "Lancet. .Jan. 13, 1917. '< Miinch. med. Wocli., 1917 (04), 25. See al.s. Physiol., 1915 (lt)2), 54. ^ So(! Myers, jour. I'harmacol.. 1910 (8), 417. lIowevcM-, Biberfeld finds mor- phine tolerance to be specific (Biocnem. Zcit., 1910 (77), 283). TOLERANCE TO POISONS 239 stimulus of the poison, just as they are in the case of immunization with toxins, with the difference that the combining substances are not thrown off into the blood. For example, it has been claimed that arsenic is ordinarily combined and held in the liver by a nuclcoprotein, and the suggestion has been made that in arsenic habituds this nuclco- protein is increased in amount. Again, saponin seems to act upon the cholesterol of the red corpuscles, and Kobert observed increased resist- ance to the action of saponin exhibited by the serum of immunized animals, which he attributes to an increased amount of cholesterol, perhaiis liberated by the corpuscles decomposed by the injected poison, or perhaps produced in excess by the tissues. Wohlgemuth^ has also suggested that in the case of poisoning with large amounts of sub- stances which combine with glycuronic acid (e. g., lysol), excessive quantities of this substance are formed by the cells and excreted into the blood, where they neutralize the poisons in much the same manner as the antitoxins neutralize toxins. But besides these scanty examples of tolerance to poisons, the body possesses a number of methods for opposing manj'- other poisons with more or less success; and, poisons invariably acting chemically, the defenses are in turn largely chemical. We have elsewhere referred to the destructive action of the enzymes of the digestive tract upon bac- terial and similar poisons; this means of defense cannot apply to non-protein chemical substances except possibly glucosides and toxic lipoids. But the acidity of the gastric juice, the alkalinity of the bile and pancreatic juice, and the precipitating effect of the hydrogen sulphide formed in intestinal putrefaction are all factors that help to neutralize or prevent the absorption of certain poisons, their total efficiency, however, being on the whole very slight. After absorption of a poison a large series of chemical reactions and physiological proc- esses is brought into play, and there are few poisons indeed whose harmful influence is not more or less decreased by these means. Ko- bert" classifies these protective processes as follows: 1. Rapid elimination, either before absorption by means of diarrhea and vomiting, or by the same means after absorption in case the poisons are excreted into the digestive tract (e. g., morphine, venoms, antimony, and many other metals). Many poisons are very rapidly eliminated by other routes (e. g., anes- thetics, curare), in some instances causing harm, particularly to the eliminating organ (e. g., kidneys in phenol poisoning, intestines in ricin poisoning). The routes and conditions of elimination of poisons have been fully discussed by Lewin." 2. Deposition and Fixation in Single Organs or Tissues. — In this respect the liver is especially important, probably because of its location and function as a filter for all the blood coming fresh from the alimentary tract." The manner and » Biochem. Zeitschr., 1906 (1), 134. ^ "Lehrbuch der Intoxikationen," Stuttgart. 10 Deut. med. Woch., 1906 (32), 169; see also Mendel et al., Amer. Jour. Physiol., 1904 (11), 5; 1906 (16), 147 and 152. " Concerning the detoxicating function of the liver see Woronzow, Dissertation, Dorpat, 1910; Rothberger and Winterberg, Arch, internat. Pharmacodyn., 1905 (15), 339. 240 DEFENSE AGAINST NON-ANTIGENIC POISONS means by which this fixation is brought about are unknown. It is possible that the power of the tissues to bind poisons may become increased bj' repeated doses, lead- ing to "specific acquired tolerance. '"^^ According to Slowtzoff^' arsenic is fixed by the nucleus in a very firm combination;'* mercury by globulins in a less stable combination; copper by the nucleins, but less firmly than the arsenic. Other poisons, chiefly alkaloids, are probably combined with bile acids. Possibly some poisons combine with glycogen. These compounds are but slowly broken up, and thus the poison reaches the more susceptible and more important tissues in a rela- tively diluted condition. The bones seem to hold in harmless form poisonous fluorides, and to less extent arsenic, barium, and tungsten, which persist in the bones for a great length of time. Leucocytes are possibly important binders of poisons, perhaps through combination with their nucleins,'^ but storage in these labile cells is necessarily of relatively brief duration. 3. Combination with substances formed or contained in the tissues; the result- ing substance being less toxic than the poison alone. Under this heading may be included both chemical combination and physical absorption or solution, such as the deviation of the lipoid-soluble narcotics from the central nervous system by excessive tissue fats, or by fats therapeutically introduced.'* Many poisons combine with the inorganic constituents of the tissues; e. g., btyium and various aromatic substances with SO4; silver with CI, etc. 4. Chemical alteration, with or without subsequent combination with other sub- stances, by such means as oxidation, reduction, hydrolysis, and neutralization. 5. Impaired absorption should also be considered as a means of defense against poisons. This may depend upon the injury to the gastro-intestinal tract produced either by the poison itself or by some independent pathological condition. Cloetta considers impaired absorption important in acquired immunity to arsenic (see below) and it may also modify the effects of other poisons." The chemical reactions employed in defense against simple chemical poisons have been particularly considered bj^ E. Fromm,^^ whose out- line is here partially followed, and to which the reader is referred for bibliography. INORGANIC POISONS MetaUic poisons, such as lead, silver, mercury, and arsenic, are made insoluble, particularly by forming compounds with proteins in the alimentary tract, intestinal walls, blood, or internal organs; also by forming sulphides with the H2S of the intestinal contents. Accord- ing to Cloetta^^ immunization against arsenic depends entirely upon a reduction of absorption in the intestine, for the longer arsenic is taken, the less appears in the urine and the more appears in the feces. ^^ At the same time the resistance to arsenic injected subcutaneouslj^ is not increased at all, and no increase in resistance can be obtained '2 Santesson, Skand. Arch. Physiol., 1911 (25), 28. '■' Hofmeister's Beitr., 1901 (1), 281; 1902 (2), 307. " Denied by Heff'ter. (Arch, intcrnat. de Pharmacodyn., 1905 (15), 399), who considers it more a i)hysico-clK'mical process. '* Stessano, Comjjt. Rend. Acad. Sci., 1900 (131), 72. '«See (Iraham, Jour. Inf. Dis., 1911 (8), 147. "v. Lhota, .Vrch. internal, pharmacodyn., 1912 (22), (il. '* "Die cheniischen Schiitzmittel des Tierkorpers bei Vergiftungen," Strassburg, Karl Triibner, 1903. See also r6sum6 by EUinger, Deut. med. Woch., 1900 (26), 580. '* Arch. exp. Path. u. Pharm., 1906 (54), 196; Corresponbl. Schweizer .\oTzte, 1911 (41). 737. ^" Not accepted bv lluusmann, Krgel)nisse Physiol., 1907 (6), 58; or Joachi- nioulu, .Vrch. cxi). Path.. 1916 (79). 119. DEFENSE AOAIXST ISOliGANIC I'OISOSS 241 by repeated siihcutaiicous injecdoiis of sul)letluil doses, 'riicic is, however, reason to question the authenticity of the reputed toku-ance of habituds to arsenic (Joachimoglu).-" Antimony does not produce tolerance in experimental animals (Cloetta).-' The manner in which various inorganic ions antagonize the physiological action of one another (e. g., sodium and potassium, calcium and nuignesium) is still an important problem. -'- Free acids and alkalies are partly neutralized by the alkaline and acid contents of the gastro-intestinal tract, partly by forming com- pounds with the proteins, and partly by the alkalies and carbonic acid of the blood stream. (See "Acid Intoxication," Chap, xx.) Phos- phorus^^ and sulphides arc oxidized after absorption into phosphoric and sulphuric acid, which are in turn neutralized by the alkalinity of the blood^and tissues. Lillie-^ has called attention to the close, palisade arrangement of the nuclei of the epithelium lining the ali- mentary tract, which makes it necessary for all substances absorbed to pass through the zone of their active oxidative influence, a fact uncloubtedly of great importance in the defense of the body. Reduction of iodic acid, chloric acid, hj-pochlorous acid, and their salts occurs in the body, resulting in their conversion into the much less toxic iodides and chlorides. Tellurium compounds are also re- duced and rendered insoluble. This reaction occurs to some extent in the intestines; how much in other organs is unknown. Methylation, the addition of CH3 groups, is observed in poisoning by tellurium, which is eliminated in the breath as methyl telluride, and also in the sweat and feces. ^^ Selenium, pyridine, and some other substances also combine with methane. The source of the methane is possibly in the xanthine molecule. Summary. — There are, therefore, three chief reactions used against inorganic poisons in the body, oxidation, reduction, and splitting off of water; neutralization of acids or alkalies and the formation of al- buminates and sulphides being included under the last heading, since in these reactions the splitting off of water is an essential step. ORGANIC POISONS In the case of organic poisons an ecjually small number of primary reactions is emploj-ed in their detoxication, but in more complicated manners and combinations corresponding with the complexity of organic compounds. -' Arch. exp. Path. u. Phann., 1911 (64), 352. •" See Osterhout, Proc. Phil. Soc, 1916 (55), 533. -' Increased tolerance to phosphorus maj' be obtained by repeated small doses, but it lasts only while the poison is being given continuously (Oppel, Ziegler's Beitr., 1910 (49), 543). Accompanying the tolerance are structural changes in the liver cells to which are ascribed some significance bv Oppel. -* Amer. Jour. Physiol.. 1902 (7), 412. " See Mead and Gies, Amer. Jour. Phvsiol., 1901 (5), 105. Caffein may be demethylated in the liver. Kotake, Zeit., physiol. Chem., 1908 (57), 378. 16 242 DEFENSE AGAINST NON-ANTIGENIC POISONS Oxidation, which has abeady been mentioned as a means of de- struction of bacterial toxins, is naturally one of the most effective agents in the destruction of simpler organic substances, since the ordinary decomposition of all organic food-stuffs is through oxidation. There are numbers of specific examples of the conversion of a poisonous into a less poisonous or non-poisonous substance by oxidation. All acids of the fatty acid series are oxidized vigorously in the body, eventually into CO2 and H2O; and pathologically produced acetic and lactic acids are destroyed in this way. The hver contains an oxidase destroying alcohol, which is not increased in the livers of animals made tolerant to alcohol (J. Hirsch).-^ Uric acid is oxidized vigorously by many organs (except in man), as are other members of the purine series, such as caffeine and theobromine. Presumably oxi- dation of organic poisons as well as of food-stuffs is brought about by the oxidizing enzymes of the cells, as shown by Ehrlich's indophenol reaction, which consists of the oxidation of paraphenylene diamine and a-naphthol, with a resulting synthesis. This reaction is said by Lillie^'^ to occur principally in and about the cell nuclei or cell membranes. Combination, with or without Preliminary Oxidation. — Oxi- dation is also an essential preliminary step to many of the protect- ing combinations, in which a cell constituent is united to an organic poison. The most important of these combining substances are: 1. Sulphuric Acid. — One of the earliest and most important observa- tions on the protective action of sulphuric acid was made by Baumann and Herter,-- who showed that phenol is eliminated as a potassium salt of the sulphuric acid derivative, as follows: CeHsOH + HO-SO3K = C^HsO-SOsK + HoO, a reaction that has been put to practical use in treating phenol poison- ing. As phenol and cresols are produced constantly in intestinal de- composition, this reaction is undoubtedly of great service, since the salt formed is relatively harmless. Indole and skatole are similarly de- toxicated by being converted into corresponding salts, but onh' after a preliminary oxidation into indoxyl and skatoxijl, according to the following reaction : CH C(OH) C6H4<^^CH + 0 = C6H4^^CH. NH NH (indole) (indoxyl) C(OH) C-O-SO2OK. CaH4<^^CH + HO— SOjOK =C«hZ^CH + H2O. NH NH (indoxyl) (indican) " Biochem. Zeit., 1916 (77), 129. " Zdt. physiol. Cliem., 1877 (1), 247. DEFENSE AGAINST ORGANIC POISONS 243 A host of other aromatic organic substances are similarly combined with sulphuric acid,^* with or without preliminary oxidation, includ- ing all substances resembhng phenol or which through oxidation are changed into phenols, such as cresol, thymol, anilin, naphthalin, pyro- gallol, and tannin. By this means a poisonous substance is converted into a relatively harmless one, which is readily soluble and rapidly eliminated. 2. Glycuronic acid occupies the same position as sulphuric acid, com- bining particularly with naphthol, thymol, camphor, chloral hydrate, and but}'! chloral. Sometimes a substance may appear in the urine combined in part with sulphuric, in part with glycuronic acid, show- ing the similarity of their function. Apparently when there is not sufficient sulphuric acid in the body to combine with all the poison, the excess unites with glycuronic acid,^^ although combination between glycuronic acid and the aromatic substance begins to occur before all the sulphuric acid is exhausted. ^"^ Glycuronic acid represents merely a first step in the oxidation of glucose, as follows : OHC-(CHOH)4-CH20H + 00 = 0HC-(CH0H)4-C00H + H2O. (glucose) (glycuronic acid) This oxidation occurs after the aldehyde group of the glucose has been combined by some other substance; hence the aldehyde group escapes oxidation, although ordinarily more easily oxidized than the alcohol group. Just as with the addition of sulphuric acid, oxidation may be a preliminary step to the addition of glycuronic acid; e.g., naphthalin is oxidized into a-naphthol, before uniting to glycuronic acid, as fol- lows: H H /C=C\ H H H HC4 ^C-C^ /C = C\ OH ^C-Cf >CH + O = HC< >C-C^ H \C = C/ ^C-Cf >CH H H H \C = C/ H H (naphthalin) (a-naphthol) The same is the case with many camphors and terpenes. Reduction may be the preliminary step, as with chloral hydrate, which is first reduced to trichlor-ethyl-alcohol. In still other cases splitting off of water is the chief preliminary step. 3. Glycine is one of the longest known combining substances, the observation of the combination of glycine with benzoic acid to form hippuric acid being the first proof of synthesis in the animal body dis- covered by Wohler (1824). The reaction is as follows: CeHsCOOH + H2N-CH2-COOH = C6H5CO-HN-CH2-COOH + H2O. (benzoic acid) (glycine) (hippuric acid) 28 3gg Hammarsten's Text-book (fourth American ed.). P- 542. 28 See Austin and Barron, Boston Me^. and Surg. Jour., 1905 (152), 269. Wohlgemuth has observed a case in which all the sulphuric acid of the urine was in organic combination (Berl. klin. Woch., 1906 (43), 508). 30 See Salkowski, Zeit. physiol. Chem., 1904 (42), 230. 244 DEFENSE AGAINST NON-ANTIGENIC POISONS A special enzyme has Ijeen found in kidney substance vvliicli can bring about this reaction outside the body. Normally this enzyme occurs chiefly in the kidney l)ut may also occur in other organs. Man}' other aromatic compounds also combine with glycine before elimination, e. g., salicylic acid. Some are first altered to a suitable form by oxidation; e. g., toluene is oxidized to benzoic acid, xylene to toluic acid, nitro-benzaldehyde to nitro-benzoic acid. Many of the sub- stances that can be made to combine with glycine in the body are of such a foreign nature that they never could need neutralization under any other than experimental conditions, but here, as with the sul- phuric and glycuronic acid reactions, combination occurs whenever a suitable substance is present in the blood, glycine alwa^'s being abun- dant as a cleavage product of the proteins. 4. Urea ma}' also be a means of defense, forming salts with organic acids which are rapidly eliminated; e. g., amido-benzoic acid and nitro- hippuric acid. 5. Methane. — Methylation, which occurs also with tellurium, is observed on administration of pyridine, as shown by the following equation: H H H H HC/ ^N + CH, + O = Kcf S^/ H H H H (pyridine) This reaction is of special importance, because many alkaloids contam a pyridine group; and the resulting methyl compound may be less toxic than the original alkaloid — e. g., methyl morphine. 6. Sulphur split off from proteins may combine with CNH and CNK, converting them into the much less toxic sulphocyanides.^' 7. Bile Acids. — All the above mentioned reactions are protective largely because the substances formed are soluble and rapidly elim- inated, as well as being less toxic than the original poison. Com- pounds of many poisons are formed with bile acids which are insoluble, and therefore only slowly dissolve or decompose, thus protecting the body from overwhelming doses of the poison. Such compounds are formed, not only with inorganic poisons, but also with alkaloids, espe- cially strychnine, brucine, and quinine. They ar(^ then tlepositetl in the livei', to be slowly dissolved and eliminated. ((Occasionally acetic acid and cysteine have been observed to act as combining substances. Calcium may be considered a defensive agent against certain poisons [oxalic and citric acids] with. which it forms insoluble* compounds, although it is probable that the toxicity of oxa- lates depends largely upon their robbing the cells of calcium.''-) "' See Meuriee, Arch. int. I'liariiuicodyii., 1900 (7), 11. ■"' See Robert son .•nil I Miniicl I, .lour. I'harniiicol., IIHL' (:>), «);5.'). Mi-rriioDs or defksse '1\') Neutralization of oi'^aiiic acids entering llic Ixxly or Ioi'iikmI in iiictabolisin is accomplished by the sodium carbonate of the blood when ill small amounts; if excessive in quantity {e. g., diabetic coma), a portion is combined with ammonia and appears as an ammonium salt in the urine. Mafjiiesium and calcium salts may also help in the neutralization, prolKil)ly at the expense of the bone tissue.''' (See "Acid Intoxication," Chap. xx). Dehydration, which plays a prominent part in a numlier of the abovc-nuMitioned syntheses, is particularly important in the change of ammonium carbonate into urea: NH.— Ov NH2. >C0 = >CO + 21120 NH4— O^ nh/ As ammonium salts of all sorts are very toxic, especially hemolytic, while urea is not, this process is probably one of the most important detoxicating reactions of the body because of the great amount of ammonium compounds that is constantly being formed in nitrogenous metabolism. Summary. — As Fromm points out, the variety of reactions and the variety of defensive substances are both remarkably small in num- ber. The reactions are: oxidation and reduction, h^'dration and de- hydration, and perhaps simple addition (meth^lation). The chief knov.-n protective substances are the alkalies of the blood, proteins, hydrogen sulphide, sulphuric acid, glycine, urea, cysteine, bile acids, glycuronic acid, and acetic acid. All these substances are normally present in the body, and none of them is specific against any one poison, but each combines with several poisons. This last fact is interesting in comparison with the highlj^ specific nature of the immune substances against bacteria and their products. As far as we know, no particular increase in the neutralizing sub- stances results from the administration of inorganic or organic poisons. The body does not appear to produce any excessive amounts of sul- phuric acid in carbolic-acid poisoning, or of glycine when benzoic acid is administered. Both substances are present in the body norm- ally, and as much as is available combines with the poison; if there is not enough, the remaining poison combines with something else, or goes uncombined. In other words, the neutralizing substances des- cribed above do not appear to be the result of any special adaptation to meet a pathological condition. They are present in the body as a result of normal metabolism; they have an affinity for various chemical substances, some of which happen to be poisons; if these poisons happen to enter the body, they may be combined and neutral- ized to some extent, but, as a rule, very incompletely-. There appears '^ In this connection it may he iiuMitioncd tliat the bactericidal power of the Mood is increased if the blood is inoie alkaline, rlocreased if it is less alkaline, than usual. 246 DEFENSE AGAINST NON-ANTIGENIC POISONS to be no elaborate process of defense against the chemically simple poisons, such as seems to be called into action by bacterial infection, and hence a degree of resistance or immunity similar to that present after an attack of scarlet fever or smallpox does not exist for strychnine or phosphorus. It is also of interest to consider that unicellular organisms may show a marked capacity to increase their resistance to poisons, as shown especially by Ehrlich's studies on trypanosomes, which readily become immune to various trypanocidal drugs, including arsenic •compounds, and which transmit this acquired immunity through suc- ceeding generations. Yeasts and bacteria can also exhibit increased tolerance to antiseptics, and Effront found that yeasts owe their aug- mented tolerance to fluorides to an increased content of calcium, which precipitates the fluoride which enters the cells; this tolerance is also transmitted to new generations of yeasts. The acquired tolerance is specific in all these cases, and may, indeed, be accompanied by a decreased resistance to other poisons; thus, protozoa acclimated to alcohol may be more susceptible to other chemicals.^'* Paramecia made immune to antimony are not immune to arsenic, and this specific immunity is transmitted to succeeding generations (Neuhaus).^^ 3^ Daniel, Jour. Exper. Zool., 1909 (6), 571. 36 Arch. Internat. Pharmacoydn., 1910 (20), 393. CHAPTER XI INFLAMMATION' Although morphological alterations are prominent features of the reaction of the tissues to local injury and infection, yet at the bottom the processes of inflammation are brought about by and result in chemical alterations. The causes of inflammation are in nearly all cases chemically active substances, but for the most part their nature is too little known to permit of speculation as to what chemical char- acteristic or characteristics a substance must possess to exhibit the power of causing an inflammatory reaction. Even in the case of in- flammation due to mechanical, thermal, and electrical injuries, it seems probable that most of the features of the inflammatory reaction are brought about by the action of chemical substances produced by alterations in the tissue constituents at the point of injury,^ for tissue proteins that have been altered in necrosis are chemotactic,^ as also are extracts of tissues. The essential features of inflammation, namely, local hyperemia and related vascular disturbances, exudation of plasma, migration of leucocytes and their phagocytic action, all may be caused by the action of chemical substances upon the vessels and leucocytes. Active hy- peremia in the case of inflammation is due to stimulation of the vaso- dilator nerves or paralysis of the vaso-constrictors, or direct par- alysis of the muscular fibers of the arterioles; these may result from mechanical, thermal, or electrical stimuli, but in local infection the cause is usually chemical products of bacterial growth or of tissue disintegration. The escape of blood plasma (inflammatory edema) appears to depend upon a number of factors (discussed more fully under "Edema," Chap, xiv) of which the most important seem to be: (1) injury to the capillary walls, produced largely by the chemical 1 For extensive reviews and bibliography see Adami, in Allbutt's System of Medicine; reprinted also as a naonograph, "Inflaninuition," 1909; also Opie, Arch. Int. Med., 1910 (5), 541. Some interesting ideas are advanced by Klemensiewicz, "Die Entziindung," G. Fischer, Jena, 1908. 2 Schlaepfer (Zeit. e.xp. Path., 1910 (8), 181) finds that the reduction of methj-- lene blue is decreased in inflammatory areas, and advances the hypothesis that inflammatory stimulants are o.xidation stimulants, inflammation occurring only when the amount of oxidation aroused by the stimulant is insufficient. In accord with this is the observation of Amberg (Zeit. exp. Med., 1913 (2), 19) that sub- stances facilitating oxidation reduce inflammatory reactions. (See also WooUey, Jour. Amer. Med. A.ssoc., 1914 (63), 2279.) Another observation of similar sig- nificance is that phagocytosis is stimulated by H2O2, and that phagocytes react to HNC in the same wav as the respiratory center (Hamburger, Internat. Zeit phys.-chem. Biol., 1915 (2), 245-264). 3 Burger and Dold, Zeit. Immunitat., 1914 (21), 378. 247 248 INFLAMMATIOX causes or products of the inflammation; (2) increased osmotic pres- sure in the tissues, due to increased or abnormal formation of crystal- loidal substances with high osmotic pressure from large molecular compounds, many of which are colloids (proteins) without apprecia- ble osmotic pressure; (3) alterations in the hydration capacity of the colloids, whereby, through decrease in salts or increase in acidity, they come to possess a greater affinity for water (M. H. luscher). By far the most characteristic feature of inflammation, however, is the behavior of the leucocytes — -their increase in number in the blood, their migration from the vessels and accumulation about the point of injury, and their engulfing and destroying various solid particles, such as bacteria and degenerating tissue elements. These processes, which seem to indicate something approaching independent volition on the part of the leucocytes may, however, be well explained by ap- plication of known laws of chemistry and physics, without passing into the realms of the metaphysical. This will be attempted under the heading of: AMEBOID MOTION AND PHAGOCYTOSIS The accumulation of leucocytes at a given point in the body indi- cates that some means of communication must exist between this point and the leucocytes in the circulating blood. No direct com- munication by the nervous system or other structural method existing, the only possible explanation is that the communication is through the fluids of the body, and depends upon changes in their chemical composition or physical condition. As the latter generally depends upon the former, the communication is considered to be accomplished by chemical agencies, and called chcmotaxis. Chemotaxis Changes in the chemical composition of a fluid have been shown frequently to affect the motion of living organisms suspended in it. One of the earliest observations was that of Engehnann,'' who noticed that Bacteriurn ternio suspended in water tended to accunuilate about a l)ubble of oxj^gen in the water. Pfeffer^ discovered that the sper- matozoids of certain ferns were attracted powerfully by very dilute solutions of malic acid, which is contained in the female sperm cell, inchcating that the migi-ation of the sperm cells in the proper direc- tion depends on a chemical conuinuiication, and he projiostvl tlie t(>rni ('h(uuotaxis foi" this phenomenon. Strong sohitions of malic acid, <>ii the other hand, repelled spermatozoids. Cane-sugar was found to at- tract the spermatozoids of a certain foliaceous moss. In the case of the mali(; acid, it seems to be the anion that jircxhices the effect, since salts of malic acid have exactly the same propei'ty. ^ HotiiniHoho ZcitiiiiK, 1S81 (39), 441. <- lTi)f('r,sii(!li. iuis (leiu Hot. Iiistitut in 'riil)iii^;i'ii. ISSI-INSS, \U\. 1 uiul J. ClIEMOTAXIS AM) TUOPISMS 249 Stahl's'^ oxpcriiucnl with a laij;(' jelly-like ijjasmodiuin (Aethal- ium septicum) growing on l)aik in wet places, has become classical. He found that if the Plasmodium was placed on a moist surface, and nearl)y was placetl a drop of an infusion of oak hark, the organism moved by tiie process of protoplasmic streaming toward and into the infusion. If a piece of oak bark was placed in the water, plasmodial arms were stretched out to it and the piece of bark was soon com- pletely surrounded by the organism. These movements were found to occur in any direction, even exactly against the force of gravity. Other substances, as acids or strong solutions of salt or sugar, wercT found to cause the plasmocUum to move away from them, although when sufficiently dilute they exerted an attraction. A Plasmodium might, however, move into a strong sugar solution if kept with a scanty supply of moisture for some time, and after it had lived in such a strong solution (2 per cent.) for some time, a weaker solution or pure water was as injurious as the concentrated sugar solution previousl}' had been. Temperature was also found to exert a marked ihermofactic effect. If a Plasmodium was placed on a filter-paper, one end of which was in water at 7°, and the other in water at 30°, it would move toward the warmer end. The Theory of Tropisms. — Ciliated protozoa, which can move about freely in water, show very distinct reactions to stimuU of all sorts. The first step in their change of direction of movement is considered by many obser\'ers to be an orien- tation of the organism until it is headed in the axis along which it is to move. This is ascribed by J. Loeb^ to the existence of a certain degree of equality of irritability of symmetrical parts of the body. The stimulant, whether it be rays of light, or diffusing chemicals, or heat-waves, moves along definite lineb, and the organism receives at first unequal stimuli on symmetrical parts of the body, unless the axis of the organism is parallel to the lines of motion of the stimulant. As long as the stimulant acts on symmetrical parts of the body unequally, these parts will react unequally until at length the body is swung into a position where the stimulation is equal, when it will stay in this position and move along a line parallel to the line taken by the stimulant. Not only protozoa, Init much higher forms, including some vertebrates are believed to react in this way to stimuli — e. g., the maintenance by fish of a position heading up stream. The above con- stitutes the so-called ^^ theory of tropisms," and we have such reactions to stimuli of all sorts, not only chemotropism and thermotropism, but also heliotropism (reac- tion to light); geotropism (to gravitjO, eledropism (to electricityj, thig-motropism (reaction to contact), etc. The work done upon tropisms applies particularly to ciliated, freely motile organisms, and interests us less in connection with leucocytes than do the obser- vations on such forms as Amoeha.^ In passing may be mentioned the thigmotaxis or thigmotropism (reaction to mechanical stimuli) shown by spermatozoa, which explains their apparently difficult feat of advancing in opposition to the cilia of the epithelial lining of the female generative tract. It may also be noted that the nature of reactions of organisms to various stimuli is not con.'^tant for even the same organism. Copepods (minute Crustacea) may be negatively heliotropic in the day and go away from the bright surface of the water, whereas at night « Botanische Zeitung, 1884 (42), 145 and 161. ^ Comparative Physiology of the Brain. New York, 1900, p. 7. * For full details see Jennings (Publication No. 16, Carnegie Institute. Wash- ington, 1904; also J. Loeb, "Studies in General Physiology." 250 INFLAMMATION the same animals are positively heliotropic and swarm to the surface illuminated brightly by a lantern. Variations in heliotropism may, in some cases, be explained as due to chemical changes that occur in the organism, which explanation is made more probable by J. Loeb's experiments, which show that change in composition in the fluid in which animals are suspended may cause a complete reversal in their reaction to a constant stimulus. Motile bacteria seem to behave much like cili- ated protozoa in their reaction to stimuli. Chemotaxis of Leucocytes^ That leucocytes come to the site of an infection because of chemical substances produced by bacteria at this point, that is to say, through chemotaxis, was first clearly pointed out by Leber^° in 1879, who likened the attraction of such substances for leucocytes to the effect of malic acid upon spermatozoids as shown by Pfeffer. He found that in keratitis, leucocytes invaded the avascular cornea from the dis- tant vessels, not in an irregular manner, but all moved directly toward the point of infection, where they collected. As dead cultures of staphylococci produced a similar, although less marked, accumulation of leucocytes, he sought the chemotactic substance in their bodies, and isolated a crystalline, heat-resisting substance, phlogosin, which at- tracted leucocytes in animal tissues. He also observed that capillary tubes filled with phlogosin or with staphylococci were soon invaded by masses of leucocytes. Since Leber's experiments, many other investigations have been made showing that chemical substances of many different origins other than bacterial exert a chemotactic influence on leucocytes. Some sub- stances are indifferent in effect, most are positive, while some are be- lieved to repel leucocytes; i. e., are negatively chemotactic. Negative Chemotaxis. — Probably the substances that repel leuco- cytes are few in number; Kanthack, indeed, doubted the existence of really negative chemotactic action upon leucocytes. Verigo" also considers that as yet no actual negative chemotaxis has been satisfac- torily demonstrated; but, by analogy with the effects of chemicals on amebae, ciliata, and plasmodial forms, which all show a decided nega- tive chemotaxis under certain influences, it would seem most prob- able that leucocytes also should be repelled as well as attracted by chemicals.'^ N on -bacterial Chemotactic Substances. — One of the earliest significant studies of the effects of non-bacterial substances upon chem- otaxis was made by Massart and Bordet,'^ who showed that products " Review of literature on leucocytes by llelly, Ergeb. allg. Pathol., 1914(17(n), 1. '0 Fortschritte der Med., 1888 (6), 460. " Arch. d. M('d. oxper., 1901 (13), 585. "* Salomonsen's observation (Festskrift vcd iiulviclscMi af Statrns Scrum Iii- stitut, Kopenhagen, 1902, Art. XII), that ciliated infusoria wlu-ii killed show a strong negative effect on other ciliates, is of much interest, particularly as it seems to he tlu; ojjposite of the positively chemotactic effect of dead upon living leucocytes. Tlu; negative reaction of different ciliata was specific for their own kind quantitatively, but not qualitatively. '•^ Ann. d. I'lnst. Pasteur, 1891 (5), 417. CHEMOTAXIS OF LEUCOCYTES 251 of the disintegration of leucocytes and otlujr cells had a strong posi- tive chemotactic influence. They also corroborated the statement of Vaillard and Vincent that lactic acid is an actively repellant substance, for thoy found that tubes containing a pyocyaneus culture, which ordinarily became filled with leucocytes rapidly, did not become in- vaded at all if lactic acid was also added in a strength of 1 : 500, although leucocytes did enter when the dilution was 1 : 1000. Gabritchevskyi^ studied the chemical influence of a large number of substances on leucocytes and divided them into three groups: I. Substances exerting "negative chemotaxis, " including those that at- tracted only a few leucocytes. ^^ II. Substances with "indifferent chemotaxis" which attracted moderate numbers of leucocytes. III. Substances with positive chemotaxis. If we correct the groupings made by Gabritchevsky we have the following classification : I. Substances negatively chemotactic or indifferent: (a) Concentrated solutions of sodium and potassium salts; (6) Lactic acid in all concentrations; (c) quinine (0.5 per cent.); {d) alcohol (10 per cent.); (e) chloroform in watery solution; (/) jequirity (2 per cent., passed through Chamberland filter); {g) glycerol (10 per cent, to 1 per cent.); {h) bile; (i) B. cholerae gallinarium. 11. Substances with feeble chemotaxis : (o) Distilled water; (6) dilute solutions of sodium and potassium salts (1-0.1 per cent.); (c) phenol; {d) antipyrin; (e) phloridzin; (/) papayotin (in frog); (g) glycogen; {h) peptone; (i) bouillon; 0) blood and aqueous humor; {k) carmine. III. Substances with strong positive chemotaxis: (a) Papayotin (in rabbits); {h) sterilized living cultures of bacteria, whether pathogenic or non-pathogenic. These results can only be considered as suggestive and not as accu- rate findings, in view of other contradictory results. Buchner^^ obtained from the pneumohacillus of Friedlander, a protein which ex- erted a strong chemotactic influence, thus showing the chemical nature of the attraction of leucocytes by bacteria, and he isolated other similar proteins from other bacteria. He also obtained a "glutin- casein" from grain which was related chemically to the bacterial pro- teins, and which was equally chemotactic. The metabolic products of bacteria, however, he found to be negatively chemotactic. Alkali albuminate and hemi-albumose were strongly positive, but peptone was not. Glycine and leucine were found to be chemotactic, but urea, ammonium urate, skatole, tyrosine, and trimethylamine were not. It was also observed that if the positively chemotactic sub- stances were injected subcutaneously, they produced general as well as local leucocytosis. The products of the action of serum on bacteria, " anaphylatoxin, " produce inflammatory reactions, and probably are '^ Ann. d. I'Inst. Pasteur, 1890 (4), 346. ^* Evidently these substances were not all negatively chemotactic, but were relatively slightly chemotactic or indifferent; yet in the literature generally these experiments have been cited as indicating a negative chemotactic influence of the substances studied. '« Berl. klin. Wochenschr., 1890 (27), 1084. 252 IXFLAMMATIOX important factors in pathology; the products of tissue disintegration have simihir effects.'' Certain drugs (notably quinine, morphine and chloral) when injected subcutaneously seem to reduce the amount of leucocj'tic emigration at a point of local injury (Ikeda).'^ In gas gangrene negative chemotaxis is striking, possibly depending on the abundant organic acids produced by gas bacilli.'" V. Sicherer-° found that chemotaxis of leucocytes may be observed outside the body. If a tube containing positively chemotactic sub- stances (dead beer-j^east cells and dead staphylococci were the strong- est) is placed with one end in a leucocyte-containing exudate, the leucocytes pass up into it against gravity. Bloch^^ demonstrated that carbol-gh'cerol extracts made from each of the different viscera and tissues exerted a positive chemotaxis, discrediting the statements of Goldscheider and Jacob that only extracts of hematogenetic tissues showed positive chemotaxis. Egg- albumen, gelatine, albumen-peptone, and alkali albuminate w'ere also positive, carbohydrates feebly so, and fat not at all. Metallic copper, iron, mercury, and their salts have also been found to be chemotactic substances, but it is very probable that thej- act in part through destrojang the tissues in their vicinity, which give rise to decom- position-products having a positive effect. Adler,'-- however, found that bichloride of mercury as dilute as 1 : 3000 caused more leucocytic invasion of a piece of saturated elder pith than did even cultures of pyogenic bacteria. ^^ Aletchnikoff observed that leucocytes might, after a time, be at- tracted toward substances that at first seemed to repel them. If the blood is full of toxins, the subcutaneous introduction of bacteria does not lead to a local accumulation of leucocytes, presumably because the difference in chemotaxis between the blood and the tissue fluids is not great enough to cause emigration; in this connection should be men- tioned Pfeffer's observation that B. tcrmo in a peptone solution will not migrate toward another stronger peptone solution, unless the lat- ter is at least five times as strong as the former. Leucocytes will migrate freely toward substances that kill them; of the bacterial prod- ucts the toxins of pyocyaneus and diphtheiia l)acilli l)eing especially destructive and causing typical karj'orriiexis.^^ Substances soluble in lipoids are said by Hamburger-' to increase phagocytic activity when in extreme dilutions, although stronger concentrations are highl}- " See Dold, Arb. Path. Inst. Tiihingoii, 1014 («»). .SO. '».Iour. Phamuicol., 1916 (8), 1.37. '■'Sec Kinivs-Kohcits ;iiul CowoU, .lour. Putli. I^act.. 1!)17 fJl i. 473. =» VAiwi. f. iiakt., 1,S99 (2()), 3()0. -' Cent. f. allK. Path., 1K9() (7), 785. -- Festschr. for -V. Jacolii. 1900, New "\ oik. -•H'onccrnin^ the ctTects of iotlin and iodide.s upon the ItMic.ocvtos. see TIeiiiz, Vircliow's Aicli., 1S99 (l.W), 44. 2* Schiirniann, Cent. f. Pathol., 1910 (21), 337. "Arch. NY>erland., 1912 (III, H), 134; Brit. M<<.|. ,luur.. 191(1 a\ 37. niKMOTAXIS OF LEUCOCYTES 253 toxic lor Icucocylcs. It ;iii clcclric cunciit is pnsscMl tliioii^li two fiii- goivs there will be foiiiul more leucocytes in the tissues of the catiiodc finger than in the anode finger, prosuniahly because the OH-i(jns increase ameboid movement. ^^ Man}'' substanc(!s have b(>en used to inciease the numbcjr of leuco- cytes in the circulating blootl in the hojjc of increasing resistance to infections, a result that does not seem to follow artificial leucocytosis with any recognizable uniformit3\ A compilation of the literature on this subject In- Gehrig'-^ shows such contiadictory findings as to indicate that most of the recorded work is of little value. He was unable to corroborate the current statement that antii)yretic drugs increase the number of leucocytes in the blood. Nucleinic acid and tissue extracts seem to increase circulating leucocytes with considerable regularity, while with thorium-X and benzol they can be reduced to almost complete extinction. The behavior of inflammatory processes in animals thus deprived of available leucocytes has considerable experimental interest. ^^^ If less than 1000 leucocytes per cubic mm. are present in the blood, no leucocytic exudate can be produced,-^ although the other features of inflammation occur as usual. Relation of Cell Types to Migration. — Of the leucocytes, the most actively affected by chemotaxis is the polymorphonuclear vari- ety, but not all substances affect each variety of leucocyte in the same way; for example, infections with most animal parasites result in both local and general increase in the eosinophilous forms, and similar effects have been obtained by the injection of extracts of animal para- sites. Lymphocytes are much less active, presumabl}^ because they contain less of the mobile cytoplasm and consist chiefl}^ of the struc- turally fixed nuclear substance. Undoubtedly many of the cells in so-called lymphocytic accumulations seen in certain conditions, such as tuberculosis, are not lymphocytes from the blood, but are newly divided cells of the tissue.^" The experimental evidence concerning lymphocytic emigration is very contradictory. Fauconnet^' has found that tuberculin injections cause in man general increase in leucocytes, but only of the polymorphonuclear form. Long-continued intoxication of animals, however, may result in lymphocytic increase, but local introduction of the toxin leads to accumulation of polymor- phonuclear cells and not lymphocj^tes. Wolff^- claims that tetanus and diphtheria toxins produce lymphocytosis in experimental animals. Wlassow and Sepp-^^ state that lymphocj'tes are not capable of ameboid -^ Schwyzer, Biochem. Zeit., 1914 (60), 454. 2- Zeit. exp. Path., 1915 (17), 161. "" See G. Rosenow, Zeit. exp. Med., 1914 (3i, 42. 2' Camp and Baunigartnpr, Jour. Exp. Med., 1915 (22), 174. 30 See resume bv Pappenlieim, Folia Hematol.. 1905 {2), 815; 1906 (3j, 129. " Deut. Arch, kliri. Med., 1904 (82), 1(17. ^= Berl. klin. Woi-h., 1904 (41), 1273. 33 Virchow's Arch.. 190t (176), 185. 254 INFLAMMATION movement or phagocytosis in the body, although after heating to 44° they may become motile for a short time. Particularly significant is the experiment of Reckzeh^* who found that in lymphatic leukemia with the lymphocytes greatly exceeding the polymorphonuclear forms in the blood, the pus from an acne pustule or from cantharides bhsters contains practically no 13'mphocytes, but is composed of the usual polynuclear forms. Experiments on the nature of the leucocytes attracted by different chemotactic agents have been made by Borissow^^ and Adler.^^ Both agree in stating that none of the substances tested shows any special affinity for any single type of leucocytes. Zieler^'^ observed that in the skin of rabbits exposed to the Finsen light, active migration of lymphocytes takes a prominent part in the reaction. General Ij^m- phocytosis may be produced by certain substances (pilocarpine, mus- carine, BaClo) which cause contraction of the smooth muscles and force these cells out of the spleen (Harvey), ^^ but such a process has no relation to chemotaxis. It is notorious that infections with animal parasites cause both local and general increase in eosinophiles, and we may even have local mast-cell leucoc3^tosis.^^ Tissue cells were found by Alder to migrate far into blocks of elder pith, apparently rather later than the leucocytes. As they showed changes of form indicating ameboid motions he considers their migra- tion to be an active process. The existence of the polymorphonuclear forms in the pith seems to be very transient. The position taken by the young blood-vessels and cells in granula- tion tissue, at right angles to the surface, possiblj^ also depends on chemotaxis determining the direction in which the new cells shall pro- liferate. Thermotaxis of Leucocytes. — Heat seems to affect leucocj^tes much as it does ameba^, moderate temperatures being positively thermotactic. Mendelssohn^" states that the thermotaxis is most pronounced at a temperature of 36°-39° C. (97°-102° F.), but is still marked as low as 20° C. Temperatures higher than 39° C. (102° F.) do not seem to attract them. Wlassow and Sepp-" state that motility of leucocytes is increased by warming to 40° C, and that temperature of 42°-46° C. causes the movements to become very irregular, with feeble power of contraction. Lymphocytes are not motile at ordinary temperature, but at 44° they begin to move, and once motile, they continue their motion when cooled as low as 35°; this motility is considered to be entirely abnormal and only the result of degenerative changes. Murphy^- and his colleagues have found tliat exjiosure of animals to suitable degrees of overheating, leads to marked lymphocj'tosis. »^ Zeit. f. klin. Med., 1903 (50), 51. " Ziegler's Beitrilge, 1894 (1(5), 432. 36 PY^stschrift f. A. Jacobi, New York, 1900. " Cent. f. Pathol., 1907 (18), 289. "Jour, of Physiol., 1900 (35), 115; see also Rous, Jour. Exper. Med.. 1908 (10), 238. "See Milchener, Zeit. klin. Med., 1899 (37), 194; Massaglia, Cent. f. Path. 1910 (21), 534. "Roussky Vratch 1903. <' Virchow's Archiv., 1904 (176), 185. "Jour. Exp. Med., 1919 (29), 1. PHAGOCYTOSIS 205 If mixtures of leucocytes and bacteria sensitized with oj)Sonins are kept at low- temperature, the bacteria become attached to the surface of the leucocytes, not being ingested until the mixture is wanned.^' This indicates that two separate processes are involved in phagocytosis. Temperature probably plays but a minor part in attracting leucocytes in pathological processes, however. The local heat of an inflamed area is due chiefly to the accumulation of blood in the part, and would not influence the leucocytes to migrate from the still warmer blood into the ti.ssues. Segale,^* hciwever, has demonstrated tiiat there is some actual heat production through increased metab- olism in inflamed tissues, which may have .some slight effect. By increasing motility the temperature of fever may favor migration and phagocytosis, and local application of heat to inflamed areas may induce local leucocytic accumula- tion. In burns the duration of the period of excessive temperature is usually too brief to account for the attraction of leucocytes that results; this accumu- lation is undoubtedly due to the products of the resulting cell degenerations. The influence of light, mechanical stimulation, and gravity upon leucocytes seems not to have been studied. The phagocytosis of insoluble non-nutritive particles has been ascribed to tactile stimulation, but the details of the operation of such stimuli are unknown, and the entire question of tactile stimulation is unsettled. In experiments with elder pith it has been observed that leucocytes penetrate to the center, even when the pith contains only physiological salt solution. As Adler remarks, it is difficult to explain such migration as due to tactile stimuli; but on the other hand, no other explanation has been offered. Phagocytosis *= The engulfing of bacteria, cells, tissue products, etc., by leucocytes seems to be but an extension of the phenomenon of chemotaxis. When the substance toward which the leucocyte is drawn is small enough, the leucocyte simply continues its motion until it has flowed entirely about the particle. Later the particle becomes, as a rule, more or less altered within the cell, unless it is a perfectly insoluble substance, such as a bit of coal-dust. This action upon the engulfed object is un- doubtedly due to the action of intracellular enzymes. ^*^ Protozoa take their food into a specialized digesting vacuole which has been shown by Le Dantec*^ (in Stentor, Paramoecium, and some other varie- ties) to contain a strongly acid fluid. Miss Greenwood*^ has also demonstrated acid in several forms of protozoa, which is formed under stimulation of injected particles, whether nutritious or not. Mouton'*^ ^^Ledingham, Proc. Royal Soc, 1908 '80), 188; Sawtchenko, Arch. sci. biol. 1910 (15), 145. " Jour. Exp. Med., 1919 (29), 235. ** See review by Metschnikoff, Kolle and Wassermann's Handb. d. Path. Mik- roorganismen, 1913 (II), 655; also H. J. Hamburger. " Physikalisch-chemische Untersuchungen iiber Phagocyten," Bergmann, Wiesbaden, 1912, where is given a full account of the author's important researches on the principles of phagocytic behavior. « See Opie, Jour. Exp. Med., 1906 (8), 410. " Ann. d. I'Inst. Pasteur, 1890 (4), 776. " Jour, of Physiol., 1894 (16), 441. " C. R. Acad, des Sciences, 1901 (133), 244. 256 IXFLAMMATIOX has been able to extract from the bodies of protozoa (rhizopods) a feebly proteolytic enzyme. This "atnibodiastase," as he calls it, is active in alkaline, and faintly acid media, and digests colon bacilli that have been killed by heat, but not living bacilli. This last fact is highly suggestive in connection with the important question of whether leucocytes engulf and destroy virulent bacteria or only those that have been previously injured by the tissue fluid. It was impossible to se- cure either invertase or lipase in extracts of protozoa. Whether bac- teria are digested in leucocytes by the same enzymes that digest the leucocytes themselves after they are killed (?'. e., the autolytic fer- ments), or by some specialized enzyme is not known. Metchnikoff, however, has noted the localized production of acid in the cytoplasm of leucocytes of the larva of Triton taeniatus. The eventual excre- tion of the remains of the bacteria or other foreign bodies by the phagoc3^tes is ascribed by Rhumbler to changes in the composition of the particles through digestion, so that they have a greater surface affinity for the surrounding fluids than for the protoplasm of the cell. Calcium and magnesium salts increase phagocytosis and leucocytic migration, ^<^ while changes in osmotic pressure decrease these activi- ties, as also does quinine even in dilutions of 0.001 per cent. Phago- cytosis cannot take place in the absence of elect rol3'tes, according to Sawtchenko.^^ Fat-soluble substances in general increase phagocyto- sis (Hamburger),^- but cholesterol inhibits phagocytosis.^^ (its ef- fects being suppressed by lecithin) ^^ acting apparent I3' by virtue of its OH group. Agents facilitating oxidation favor phagocj'tosis (Arkin).^^ Maximum phagocytosis occurs at the normal bodj^ tem- perature of the animal furnishing the leucocytes (Madsen and Wulf' .'^ Phagocytosis cannot be readil}- ascribed to chemotaxis, however, in the case of phagocj^tosis of perfectly' insoluble, chemicalh' inert par- ticles, such as coal-dust. The leucocytes seem to take up foreign bodies without reference to their nutritive value, absorbing India-ink granules and bacteria impartially when they are injected together, and loading themselves so full of carmine granules that they cannot take up bacteria subsequent!}- injected. It is possible that foreign bodies first become coated with a layer of altered protein which then leads to phagocytosis, but there is no sufficient evidence for this sur- mise. Kite and Wherry^^ state that leucocytes take up car])on parti- *" Hamburger, Biochem. Zeit., 1910 (26), GO; Eggers, Jour. Infect. Diseases, 1909 (6), 662. According to Radsma (Arcli. neerl. pliysiol., 191S (2), .301) cal- cium salts only favor phagocytosis in leucocytes that have previously had their calcium l)oun(l by citrate or oxalate in the ])rocess of isolation. "Arch. .sci. Inol. 8t. IVtershurg. 1911 (1(1), UH ; 1912 (17), 12S. " Hamburger and de llaan, Arch. .\na(. und Phv.sio!., 19i;J, I'hys. .Vbt., p. 77 " Dewey and Nuzum, .Jour. Infect. I)is., 1914 (if)), 72. '■'Htuber, liiochem. Zeit., 1913 (51), 211; 1914 ('>;{), 493. "Jour. Infect. Dis., 1913 (13), 41S. "Over, Dan.ske Vid. Selsk. Forh., 191t) ((>), 339. " Jour. Infect. Dis.. 191.") (Ki), 109. PHAGOCYTOSIS 257 cles anclTsimilar substances because thc^ leucocytes are "sticky," which presumably is correct, but what constitutes the "stickiness" and why it varies under the influence of scrum is not indicated. Presuma- bly it represents an altered viscosity, which is known to be increased by increased acid content such as niiglit be produced by local asphyxia.^* The nature of mechanical stimulation of cells is explained by Ostor- hout*" as a chemical reaction to rupture of semipermeable cellular surfaces, and there is evidence from plant cells supporting this hy- pothesis, but its applicability to animal cells has not been investigated. The experiments of Schaeffer*^" seem to show that amcba exhibit positive chemotaxis towards such insoluble substances as carbon parti- cles and glass fragments, even at a distance, although the mechanism is unexplained. Similar investigations have not been made with leucocytes. Not only leucocytes but tissue cells are capa})le of moving and per- forming phagocytosis when properly stimulated, and apparently all or nearly all fixed cells may act as phagocytes under some conditions. Their power of independent movement is much less than their phagocytic power. Endothelial cells are particularly active in pha- gocytosis, as also are the new mesodermal cells produced in inflamma- tion. Apparently they obey the same laws as the leucocytes, and not only take up bacteria, but also fragments of cells and tissues, red corpuscles, and even intact leucocytes and other cells. Brodie''^ considers that phagocytosis by endothelial cells in lymph-glands is the natural end of the leucocytes, and red corpuscles seem to have a similar fate. Phagocytosis is usually accomphshed solely by the cytoplasm of the cells, the nuclei maintaining a passive role; but, according to Detre and Selli,^- the phagocytosis of particles of lecithin is accomphshed by the nuclei, which seem to have a specific affinity for this substance. Giant-cell formation may also be considered as the result of chemo- taxis, the cells moving toward the attracting particle, and when the particle is larger than the cells they spread out upon its surface, their cytoplasm flowing together because of altered surface tension. The peripheral disposition of the nuclei probably depends on the fact that in ameboid motion the nucleus of the cell plays an entirely passive r61e, being dragged along by the cytoplasm, and hence it is located most remotely from the attracting particle. Digestion of materials taken into a giant-cell seems to go on as in the individual cells that compose it.®^ ^8 See Woolley, Jour. Amer. Med. Assoc, 1914 (63), 2279. *' Proc. Natl. Acad. Sci., 1916 (2), 237. «» Biol. Bull., 1916 (31), 303. 6' Jour, of Anat. and Phvsiol., 1901 (35), 142. 6- Berl. klin. Woch., 1905 (42), 940. «3 See Faber., Jour, of Path, and Bact., 1893 (1), 349. 258 INFLAMMATION Influence of the Serum on Phagocytosis (Opsonins). — Phagocytosis of bac- teria by leucocytes seems not to be merely a reaction between the leucocytes and the bacteria. Wright and Douglas have demonstrated that certain substances in the blood-serum are necessary to prepare the bacteria for phagocytosis, these substances being termed by them "opsonins." If leucocytes are washed free from serum with salt solution and let stand in a test-tube with such bacteria as Streptococcus pyogenes, Staphylococcus pyogenes, B. typhosus, B. coli, B. tuberculosis, and various other organisms, no phagocytosis occurs. If, however, some serum from a normal or an immunized animal is added to the mixture, active phago- cytosis soon takes place. The action of opsonins is also involved in phagocytosis by endothelium." The character and properties of the opsonins are further considered among the reactions of immunity (Chapter vii). Results of Phagocytosis. — After phagocytosis has been accom- phshed, the fate of the engulfed object depends upon its nature. If digestible by the intracellular enzymes it is soon destroyed, but in the case of engulfed living cells, it seems probable that they must be first killed — ^they form no exception to the rule that living protoplasm cannot be digested. This brings forward the question of so much importance in the problems of immunity: Do living bacteria enter phagocytes, or are they first killed by extracellular agencies before they can be taken up? At the present time it seems to be positively established that leucocytes do take up bacteria which are still viable, and which may either grow inside the leucocytes or may be destroyed by intracellular processes. ^^ On the other hand, leucocytes do not take up extremely virulent bacteria, and hence the question as to the relative importance played by the leucocytes and by the body fluids is still undetermined. It is probable that phagocytosis by fixed tissue- cells is of much less importance in checking bacterial growth than is phagocytosis by leucocytes. Thus Ruediger's experiments showed that emulsions of organs, with the exception of bone-marrow, do not destroy streptococci which are readily destroyed by leucocytes. How- ever, the phagocytic activity of certain endothelial cells, especially in lymph sinuses and the Kupffer cells of the liver, is so great that these cells may equal or surpass the leucocytes in bactericidal power. Leu- cocytes do not seem to bind bacterial toxins. ^^ Indigestible substances may remain in cells, particularly in fixed tissue cells, for very long periods, if the substances are chemically in- ert. The leucocytes seem to transfer the indigestible particles which they have engulfed to other tissues, particularly to the lymph-glands; this is probably accomplished by phagocytosis of the laden leuco- cytes by the macrophages of the lymph sinuses, but how the insoluble particles are later transferred to the gland stroma or perilymphangial tissues, where they are chiefly found in such conditions as anthracosis, etc., is quite unknown. " Briscoe, Jour. Path, and Bact., 1907 (12), GO. " See Iluediger, Jour. Amcr. Med. Assoc, 1905 (44). 19S. "* Pettersson, Zcit. Ininninitat., 1911 (S), 498. Koozarenko, however, states that horse leucocvtos ucutrnli/.c diphtheria but not tetanus to.\in. (Ann. Inst. Pasteur, 191.'') (29), 190.) THEORIES OF CIIEMOTAXIS AND PHAGOCYTOSIS 259 Leucocytes contain substances which are stronfj;ly bactericidal, in- dependent of the action of the blood serum, and which have been called endolysins;^'' they are resistant to 65° or even higher, and seem to be bound rather firmly to the protoplasm of the leucocytes, for they resist extraction except by vigorous methods; they have a complex structure like the amboceptor-complement bacteriolysins of the serum, and are not specific (Weil).^^ They do not pass through porcelain filters readil}^ are precipitated by saturation with ammonium sul- phate, and resemble the enzymes in many respects."^ It is probable that the endolysins act upon bacteria that have been phagocyted, and perhaps also upon free bacteria when liberated in suppuration through disintegration of the leucocytes. Lymphocytes and macrophages seem to be devoid of this endolysin.'^" Phagocytosis of hving virulent bacteria may not always be an un- mixed benefit. Besides the obvious possibility of transporting the bacteria and spreading infection, we have also evidence that living bacteria may be protected through phagocytosis, against the action of bactericidal substances in the blood and tissues (Rous and Jones). '^^ THEORIES OF CHEMOTAXIS AND PHAGOCYTOSIS On the assumption that leucocytes obej- the same laws in their mo- tions as do the amebse, studies of the latter and of other forms of protozoa have furnished most of the ideas, hypotheses, and theories of the forces involved in leucocytic activities. The structural rela- tion of the leucocyte to the ameba is striking, although by no means complete; the relation of their activities is even closer. Each is a microscopic, independent, unicellular organism, moving freely in all directions by means of pseudopodia and protoplasmic streaming, taking other smaller bodies into its substance and digesting them, reacting similarly to like stimuh, and containing similarly a nucleus and many granules. The differentiation of the protoplasm of the ameba into a clear outer ectosarc and an inner granular endosarc is perhaps an important difference, but as far as the two forms of cells have been studied, the effect of this difference in structure does not seem to have been considered. That the unicellular protozoa, devoid of any central nervous system, and without any apparent co-ordinating mechanism, seem able to move about in a purposeful way, going toward food supplies and away from injurious agencies, toward or away from hght, heat, and chemicals, has long attracted the interest of physi- ologists, particularly as in these single-celled organisms we may look for the simplest conditions of existence and the most elementary hfe processes. It seems absurd to imagine that a paramoeciiwi goes toward " For general review see lOing, Zeit. Iramunitat., 1910 (7). 1. «8Arch. ft Hvg., 1911 (74), 289. " Manwaring, Jour. Exp. Med., 1912 (16), 250. " See Schneider, Arch. f. Hvg., 1909 (70), 40. ^1 Jour. Exper. Med., 1916 (23), 601. 260 INFLAMMATION a dilute acid because it "likes it," that an ameba rejects a piece of ^lass because it "does not taste good," as we explain similar mani- festations in higher forms; furthermore, it has been shown by Verworn "that minute enucleated fragments of protozoan cells react to stimuli ^ust as does the entire cell, and, therefore, it seems that the only possi- ble explanation of movements in protozoa must be a direct reaction of the stimulated part to the stimulus. The nature of the stimulus and the nature of the stimulated substance must determine the nature of the resulting reaction, and most of the observations so far made suggest that these reactions can be explained according to the known laws of the physics of fluids. An ameba, or a leucocyte, may be looked upon as a drop of a colloidal solution, surrounded by a delicate sur- face layer which is more or less readilj^ permeable to solvents and to substances in solution, and suspended in a fluid of quite different composition. Siirface Tension. — Such a drop of fluid suspended in another different fluid obeys well-known laws of physics. The particles of each fluid are all under the influence of a very considerable force, called the cohesion pressure, which tends to draw them together closely. Within the drop each particle is subjected to this force alike from all sides, so that the forces neutralize one another, and each particle is as free as if there were no cohesion pressure. But the particles on the surface are subjected to unequal pressure, for that of the fluid outside the drop is different from that inside, and so the pressure on the surface particles is equal to the difference of the cohesion pressure of the two fluids; this constitutes the surface tension. It is this tension that pulls in upon the surface continually, causing it to tend always to reduce the free surface to a minimum, which condition exists perfectly in the sphere. The amount of cohesion affinity is very different in different fluids, and therefore some have a high surface tension and some a low. When a substance dissolves in another the surface tension is a resultant of the surface tension of the two substances, and hence the surface tension of a liquid may be raised or lowered by dissolving various substances in it. Artificial Imitations of Ameboid Movement Imagine a drop of fluid suspended in water — let it be a drop of protoplasm, or oil, or mercury; the drop owes its tendency to assume a spherical shape to the surface tension, which is pulling the free surface toward the center and acting with the same force on all sides. The result is that the drop is surrounded by what amounts to an elastic, well-stretched membrane, similar to the condition of a thin rubber bag distended with fluid. If at any point in the surface the tension is lessened, while elsewhere it remains the same, of necessity the wall will bulge at this point, the contents will flow into the new space so offered, and the rest of the wall will contract; hence the drop moves toward the point of lowered surface tension. Conversely, if the tension is increased in one place, the wall at this point will con- tract with greater force than elsewhere, driving the contents toward the less resistant part of the surface, and the drop will move away from the j)oint of increased tension. The rcsemblaiice of these ciianges of form and tlic type of motion produced, to ameboid movement, is apparent, and nmch experimenting has been done to determine how IMITATIONS OF AMEBOID MOTION Jdl far tlic processes of motion as shown by anuilise and leucocytes can be reproduced by fluid drops under various conditions of experiment, and to ascertain if such ameboid movement of living cells can be entirely explained by the laws of surface tension. Gad/- in 1878, pointed out the resemblance to ameboid motion of the changes in shape observed in drops of rancid oils in weak alkaline solution. These changes in shape are due to the formation of soaps which lower the surface tension of the drop in places, so that the fluid flows toward these places and produces pseudopodium-like projections. G. Quincke'^^ later ascribed the contractions and other movements of amebse to alterations of the surface tension of the living substance in relation to that of the surrounding medium, believing the sub- stances responsible for the alterations to be albuminous soaps. Biitschli^^ found that drops of "foam structure" made by mixing rancid oil and potassium carbonate solution show "protoplasmic streaming" when placed in glycerol, and that they exhibit positive chemotaxis toward soap solution and other chemicals, the motion be- ing accompanied by current formation in the drops. The "pseudo- podia" formed by the drops also show currents rushing along their axes and returning by way of the surface. Heat leads to increased activity of motion. The motions were ascribed by Biitschli to the bursting of some of the superficial globules of the foam, which then spread over the surface of the drops, lowering its surface tension at the point of contact. He believed that ameboid motion, likewise, depended upon rupture of surface globules of protoplasm, for the "foam structure" of which he has been the leading advocate. Bernstein, ^^ basing his work on some observations of Paalzow, ob- served that a completely inorganic substance, a drop of quicksilver, could be made to imitate ameboid motion under the influence of chemical changes. If a crystal of potassium dichromate is placed near a drop of quicksilver in a nitric acid solution, as soon as the yellow color made by diffusion of the dichromate reaches the drop the quicksilver begins to show motion and advances toward the crystal. This movement is due to local oxidation of the surface mercury, which lowers the tension on that side of the drop, toward which the mercury then flows. If the crystal is removed, the drop follows, often flow- ing about it as if to take it in, but soon again withdrawing when the acid dissolves away the oxide formed on the surface, only to return again later. All these movements, which may be very life-like, are readily explained by changes in surface tension that take place under the influence of the bichromate and the acid, and are unquestionably referable to surface phenomena. '2 DuBois Reymond's Arch. f. Physiol., 1878, p. 181. " Wiedmann's Annalen, 1888 (35), 580. "* "Protoplasm," translation bv Minchin, London,11894. '» Pfliiger's Arch., 1900 (80), 628. 262 INFLAMMATION Artificial Amebae. — By far the most suggestive experiments on the simulation of ameboid activity by non-living substances are those of Hhumbler (1898) in his great work, "Physikalische Analyse von Lebenserscheinungen der Zelle."'* On the assumption that the living protoplasm was but a more or less tenacious fluid, following the simple physical laws of fluids, especially in relation to its sur- face tension, he devised a number of experiments to determine the correctness of these views. An ameba may be regarded as such a mass of viscid fluid, in a medium in which it is nearlj^ or quite insoluble; it is also constantly undergoing chemical changes within itself, and taking substances from or secreting them into the surrounding water. To reproduce partly these conditions a drop of clove oil is placed in a mixture of glycerol and alcohol; the alcohol and clove oil are miscible, the glycerol merely retarding the diffusion."^ Such a drop of oil will move about, changing its form and sending out pseudopodia much as an ameba does. These movements are undoubtedly due to changes in the surface tension brought about by the irregular mixing of the alcohol and the clove oil. The effect of chemotaxis upon an ameba can likewise be imitated with such an "artificial ameba." If some stronger alcohol is carefully introduced into the fluid near the drop, the surface tension on that side will be lowered, and the drop will flow in that direc- tion. The effect of chemical changes within the drop upon its motion may be demonstrated similarly by injecting a little alcohol into the substance of the drop near one edge — the drop will send out a pseudopodium on that side, and perhaps flow along in the direction of the pseudopodium. We can imagine that metabolic changes in the body of an ameba may account for many of its seemingly purpose- less movements by altering surface tension in some part of its circumference. Thermotaxis, the effect of heat in modifying or impelling ameboid motion, may- be equally well demonstrated in such an "artificial ameba," the drop being "posi- tively thermotactic," and flowing rapidly toward a heated point in the solution, because heat lowers the surface tension. Even as highly specialized a process as the taking of food may be closely simu- lated experimentally. Ameba? seem to possess the faculty of selecting substances that are suitable for their food, crawling over particles of sand, wood, etc., and rejecting them when they are pushed against or into the surface of the ameba, which, however, readily takes up bacteria, diatoms, alga^, etc., digests them, and later throws out the undigested particles. If there is any property of the ameba that suggests voluntary action, it seems to be exhibited in the choice of its food, although this is not so well developed a selective process as might be expected, for ameba; will take up many harmful objects, and they may be made to fill them- selves so full of useless substances that they cannot take up food. However, a drop of chloroform in water, which makes a good artificial ameba, if "fed" with various substances, will refuse some and take in others in a surprisingly life-like manner. Pieces of glass or of wood placed in contact with the drop, exert no influence; if pushed into the substance of the drop, they carry the surface ahead, and on being released they are thrown out with some force. If a piece of shellac, paraffin, styrax, or Canada balsam be brought in contact with the surface of the drop, however, the drop flows around it immediately, and takes it within its sub- stance, where it is soon dissolved. Even more strikingly like phagocytosis and intracellular digestion, however, is the result of a similar experiment with a piece of glass covered with shellac; the chloroform "amel)a" takes it up as readily as it does the shellac alone, but after all the coating is dissolved away the piece of glass is then cast out of the drop. The resemblance to the engulfing, digestion, and excreting of indigestil)le particles of bacteria, etc., by amel):r. is so striking that it seems impossible that there can be any fundamental differences in the two processes. It will al.so be noticed that the drop takes in only what it can dissolve and rejects what it cannot. One of the most remarkable actions of the ameba\ which seems almost cer- tainly the result of voluntary action, is this: Oftentimes in feeding, an ameba gets hold of a suitable nuiterial which is in the form of a long tlucad, much too long for the amel)a to surround. It then proceeds to coil up tlie (hreatl within its body, by stretching a slight distance along the thread, bending over, and forming a bend in the thread, and by repeating the i)rocess it crowds the tiiread into a '» Arch. f. Entwicklungsmechanik, 1898 (7), 103. "The details of these experiments are as given briellj' bv Jennings, Jour, of Applied Microscopy, 1902 (5), 1597. ARTIFICIAL AMEB/E 263 neat coil within its body, where it can be digested. The process is done so sys- tematically and with svich evident adoption of the moans at hand to the desired end, that it seems as if it must be an adaptation of tlie amolia to circumstances, the result of long experience or of heredity. That an artilicial ameba can per- form the same manouvers seems hardly credible, but it is readily done with almost no dilTerence in detail. If the chloroform drop is given a long fine thread of shel- lac, it proceeds to bend the thread in the middle, and to send pseudopodia out along the thread to pull it into the drop, coiling it up inside as the chloroform softens the substance of the thread, until it is all contained within the drop, pro- vided, of course, that it is not too long (a thread six times as long as the chloroform drop may be taken in completely)- The bending and coiling of the thread in this experiment is entirely in accord with the known laws and i)henomena of surface tension. Fully as striking an ameboid action as the coiling up of a thread too long to be taken in, is the building, by some of the protozoa clo.sely related to the ameba {Difflugia) of a shell which the animal seems to form by cementing together grains of sand, or diatom shells, or other suitable particles. The particles are united so closely and fitted together so well that they are almost i)erfectly free from crev- ices. Even this process is accurately imitated in Rhumbler's experiments. If a drop of oil is mixed with fine grains of quartz sand, and dropped into 70 per cent, alcohol, the grains are thrown out to the surface, where thej- adhere to the surface of the drop and to one another exactly as do the particles in a difflugia shell. So well fitted are the particles that the artificial shell may remain intact for months, and resemble the natural shell indistingui&hahly. Furthermore, the phenomenon of cell division can be imitated to some extent by oil droplets. Biitschli considered that the cleavage furrow of dividing cells represented an area of greater surface tension, and McClendon imitated cell division as follows: He suspended a drop of rancid oil and chloroform between water and salt solution, and allowed sodium hydrate to flow from pipettes against two opposite points in the droplet, whereon the surface tension was lowered and the drop bulged at these points, the band of higher surface tension constricting the drop between these two points. Burrows states that the changes seen in cells dividing beneath the microscope correspond well to these experimental observations.'^ Relation of the Above Experiments to the Phenomena Exhibited by Leucocytes in Inflammation The experiments cited indicate strongly, to saj^ the least, that amebse, and presumably leucocytes, react to stimuli of various kinds, chiefly through the effect of these stimuli upon surface tension. The stimuli may come from within the cell, being in this case the result of changes in composition brought about by metabolic processes; such chemical products alter the tension of the surface nearest their point of origin, causing what appears to be spontaneous motion. Stimuli acting from without may be chemical, thermal, electrical, or mechani- cal, but in any event they act as stimuli to motion through their eft'ect upon surface tension; if they decrease the surface tension the cell goes toward them; if they increase the tension, the cell moves away.''* The behavior of leucocytes in inflammation may be explained on these purely physical grounds very satisfactorily, as follows: At the point of cell injury or of infection, substances are produced that exert positive chemotaxis, as can be shown by experiments both '8 See Trans. Congress Amer. Phys., 1913 (9), 77. "' OH-ions decrease, H-ions increase the surface tension of leucocytes (Sch- wyzer, Biochem. Zeit., 1914 (60), 306, 447, 454), which may explain the fact that lactic and other acids exhibit negative chemotaxis. 264 INFLAMMATION outside and inside the body; these substances are chemotactic because they influence the surface tension of the leucocytes, and since with most if not all the products of cell disintegration the effect is to lower surface tension, the chemotactic effect is positive. As the chemotactic substances are produced, they diffuse through the tissues until they reach the walls of a capillary, through which thej'- begin to pass, pre- sumably most rapidly through the thinnest parts of the wall, the "stomata" and intercellular substance. The leucocytes passing along in the bore of the capillary will be touched by the chemotactic sub- stances most on the side from which the substances diffuse; the sur- face tension will be lowered on this side, causing the formation of pseudopodia and motion in this direction. When the leucocytes come in contact with the wall, their surfaces, because saturated with the chemotactic substances, will have a tension much the same as that of the cells of the capillary wall, which are likewise saturated with the same substances, and the two surfaces will tend to cling to one another; explaining the phenomenon of adhesion of leucocytes to the capillary wall, when, according to the usual description, "the leucocytes be- have as if either they or the capillary wall had become sticky."^" Sur- face tension of the leucocytes will be least nearest the points where the most chemotactic substances are entering the capillary, namely, the stomata; hence the pseudopodia will form in this direction and flow through the openings, the rest of the cytoplasm flowing after and dragging the nucleus along in an apparently passive manner. Since it is the cytoplasm that seems to be chiefly affected in these processes, the nucleus appearing to be rendered inert by its relatively dense and fixed structure, the leucocytes with most cytoplasm are most active in migration, while those with the least, the lymphocytes, are affected relatively little or not at all. Once through the vessel wall, the motion continues in the same manner, toward the side from which the chemotactic matter comes, just as the mercury drop flows toward the crystal of potassium dichro- mate, or the drop of oil flows toward the alcohol. If the leucocyte meets a substance that lowers its surface tension sufficiently, it will flow around the object and enclose it, just as the chloroform drop flows about the piece of shellac or balsam; this constitutes phago- cytosis. The motion of the leucocyte will continue in a forward di- rection until one of several possible things happens : (a) The leucocyte may reach a point where the chemotactic substances are so thoroughly diffused that the effects on its surface are the same on all sides; there will then be no tendency to move in any direction, {b) It may r(;a(;h a material that exerts a marked jiositivo iiithience upon it, causing much lowering of the surface tension, ])ul which is so large *" Kr('il)i(!li (An^h. f. Dcrin.atol., 1912 (114), SS.")) describes as chemical changes in the vessel walls during; the early stages of inflammation, a diffuse sudanophile change throughout the endothelial cells, in the form of fine, dust-like particles. Probably this change dei)ends simply on an aggregation of the intracellular lipoids. AMEBOID MOTION 2()r) that the cytoplasm flowing along its surface cannot surround it; other leucocytes will experience the same change, their cytoplasm will fuse together because of the equal lowering of their surface tension, and soon we got a mass of leucocytes with fused cytoplasm surround- ing the oliject, forming a "foreign body giant-cell." (c) The leuco- cyte may reach a place where the concentration of the chemicals is so great that chemical changes are produced in its cytoplasm. If these changes are of a coagulative nature, the surface of the cell will be stiffened so that it cannot migrate further; if of a solvent nature, the leucocyte is destroyed, (d) It may reach the margin of an area where the preceding leucocytes have become coagulated or otherwise rendered immobile, so that they block its path, while it is held fixed by the at- traction on this side, (c and d explain the formation of solid leu- cocytic walls about areas of inflammation, and the frequent absence of leucocytes within the central necrotic areas.) (e) The formation of chemotactic substances may cease because the substance causing the inflammation has been used up, or because the bacteria have been destroj^ed, or from any of the causes that terminate inflammation. Those leucocytes still advancing will reach a point where there is as much chemotactic substance behind as in front — they will then stop advancing. ^^ As the fluids exuded in the central portion continue to dilute the chemotactic substances and wash them out, there will soon be less chemotactic substance in the center of the inflamed area than there is farther out, hence the leucocytes will move away from the center toward the periphery, following the chemotactic substances back into the blood-vessel and the lymph-stream. These are the conditions that exist at the close of the inflammatory process, which results in the dispersion of the leucocytes. General leucocytosis can be explained equally well on the same grounds. Chemotactic substances from the area of inflammation enter the blood-stream, and so, in a very dilute form, pass through the bone- marrow. The chemotaxis in the blood will be greater than that of the marrow, and the leucocytes will move toward and into the blood. As long as the blood contains more chemotactic substances than the marrow, leucocytosis will increase, to stop when the amount in blood and marrow is alike or when there is less in the blood than in the marrow. Behavior of Tissue=cens and Formation of Giant=cells. The free cells of the tissues involved in inflammation can, of course, obey the same influences as the leucocytes, and apparently do so in so far as they are not checked by structural impediments to flowing motion; i .e., the more closely a cell is related to a single drop of fluid ^' The phagocytic action of leucocytes in vitro is decreased by substances that lower the surface tension, e. g. chloroform (Hamburger, K. Akad. Wetensch., 1911 (XIII (2)), 892). Ether-soluble substances from bacteria have no effect on phagocytosis (Miiller, Zeit. Immunitat., 1908 (1), 61). 266 INFLAMMATION protoplasm, the more closely does it resemble in the simplicity of its reactions the "artificial ameba." An illustration of the chemotaxis of epithelial cells is furnished by B. Fischer,^- who found that stained fats cause growth and migration of epithelial cells in the direction of the fat. Cells with much cytoplasm are best fitted to move freely, as a rule, and hence we see chiefly the large endothelial cells of the lymph sinuses and the serous cavities, and the large hyaline and granular cells of the blood acting as phagocytes, for phagocytosis is no different from ameboid motion which continues about a particle until it is sur- rounded; likewise we see the "epithelioid" and large endothelial cells with their abundant cytoplasm fusing together to form giant-cells. (Note that such giant-cells are formed particularly in conditions in which the epithelioid cell is more abundant than is the leucocyte, e. g., tuberculosis and other chronic inflammations. The cells that fuse about an infected catgut ligature are the leucocytes, for they are most abundant in such a place.) A good illustration, also, is the giant-cell formed by fusing of leucocytes about blastomyces in minute abscesses in the epithelium in blastomycetic dermatitis; the epithelial cells cannot flow or coalesce well because of their abundance of stiff keratin and their specialized cell-wall, and hence do not participate; the leucocytes are individually too small to surround the fungus cells, and hence they flow about them in the abscess exactly as they will do experimentally in a test-tube or in a guinea-pig's abdomen (Hektoen). The method of growing tissues in vitro permits of observation of the process of giant-cell formation, and establishes that, for foreign body giant-cells at least, they are formed by fusion of wandering cells (Lambert). ^^ The formation of giant-cells is, on this ground, but an amplification of ameboid movement and phagocytosis. The fusing of the individual cells is due to the lowering of their surface-tension by the materials diffusing from the body which is to be absorbed, until the surface of each cell becomes alike, when the surface tension at the point where each cell is in contact becomes zero and the cytoplasm runs together. Objections to the above Hypothesis. — Phj'sical explanations of ameboid movement seem to fit very perfectly the known facts concerning the actions of leucocytes. There arise but a few ditliculties in applying these laws to leucocytic action; one is the phagocytosis of chemically inert bodies, such as coal particles, tattooing materials, stone dust, etc. We know that anioba" al^so may take up such inert materials, altliough tliey generally refuse them, and it is believed that the particles exert some local injury to the cell wall that leads to an alteration in its tension. Ameba; seem also sometimes to excrete a sticky substance over their surfaces or over the foreign nuitter that is to be engulfed, which excretion seems to be the result of surface stimidation. Possibly leucocytes do the same. We nuist bear in mind, however, that the protoplasmic cells luvve much greater possibilities for action than the "artificial ameba," since within the jn'otoplasm countless cliemi- cal changes are going on wliicli nmst cause continual alteration in surfac(> tension; it is quite possible that mere mechanical action nuiy alter cheinica! action at the " Miinch. med. Woch., 190G (53), 2041. 83 Anatomical Record, 1912 (G), 91. SUPPURATION 207 point of contact, so that the injurinj^ particle niaj' become surrounded throu^^h local liquefaction of the protoplasm. With the aineba, unfortunately, the explanation of all its activities by purely physical' analogies is apparently not so successful. iUthouKh simple pseudopodia may be produced experimentally, and their formation explained readil}' on the surface tension basis, yet we find many forms of pseudopodia in the great family of ameba". Some of them are branching, some are fixed in extension, some have a stiff elastic axis. It would also l)e difficult to explain cilia as produced by changes in surface tension, yet we find in some protozoa tiiat pseudopodia may take on the persistence and action of cilia, and that cilia may seem to change into pseudo- podia. Jennings has made a most extended study of the relations of the "Be- liavior of Lower Organisms"*' to the physical theories of ameboid motion, and is unable to corroborate the claim that the processes that go on in "artificial ameba>" exactly reproduce tliose of living ameba-, or to accept the statement that living protoplasm behaves exactly as any similar drop of fluid would under the same conditions. He states that the currents set up in artificial ameba- by changes in surface tension are not the same as those in living ameba-, contrary to Rhumbler and to Biitschli. The movement of ameba, he maintains, is not due to the flowing of the contents of the cell in a central, axial current out into the pseudopodium and back on the sides, as occurs in the artificial ameba; but rather to a rolling for- ward of the upper surface over the anterior edge to the lower surface, where it becomes fixed to the surface on which the ameba is crawling. The part played by surface tension, he claims, is in the case of ameba- a very sul)ordinate one, and it is not sufficient to explain the movements of the living cell. However the discussion concerning the amebae may turn, it must be appreciated that there are some important differences between even the ameba and the leucocyte. The latter has by far the simpler organization, and approaches in structure, and presumably, therefore also in response to stimuli, more closely to the simple drop of colloid matter. It has no pulsating vacuoles, no specialized pseudopodia, never forms shells or coverings, and does not conjugate as do the amebse. The external surface of the leucocyte is much simpler, an important fact in connection with surface tension effects, for in the leucocyte the surface seems to be practically undifferentiated, naked protoplasm; whereas in amebae it is formed of a well-differentiated "ectosarc," which has marked motile powers, being able to contract sufficiently to cut an injured ameba completely in two. At the very least the surface tension explanation of leucocytic action agrees per- fectly with most of the observed actions of leucocytes, and it is the only reasonable theory offered. There seems to be no middle ground be- tween such a physical theory and a metaphysical theor}^ which would endow a single cell, without organs or nervous system, with the reasoning powers of highly developed animals, a position incompati- ble with the entire evidence of experience. SUPPURATION" For the formation of pus two conditions are necessary: (1) the ac- cumulation of leucocytes, and (2) necrosis and liquefaction of cells ** Publication No. 16, Carnegie Institute, Washington, 1904; also see American Naturalist, 1904 (38), G25. *^ Inflammatory Exudates, their formation and composition, are considered in Chapter xiv. 268 INFLAMMATION and tissue elements. Many leucocytes may be present in a tissue without suppuration; e. g., erysipelas. Necrosis of cells with their gradual liquefaction and absorption may also occur without suppura- tion; e. g., infarcts, aseptic liquefaction necrosis, etc. Hence for sup- puration to occur there must be produced substances with positive chemotaxis, to cause accumulation of leucocytes, for if a necrotic area is devoid of leucocytes, it does not suppurate; e. g., caseous tubercles. Secondly, necrosis must occur, for digestion and liquefaction of living cells and tissues does not take place. Only substances meeting these requirements — -i. e., causing positive chemotaxis and cell necrosis — will cause suppuration. Therefore, although bacterial infection is the usual cause of suppuration,*^ it may be produced by many other substances, among which the following are the best known: Bacterial proteins, even from non-pathogenic bacteria; oil of turpentine, mer- cury, croton oil, silver nitrate solutions (5 to 10 per cent.), and certain vegetable proteins (vegetable "caseins"). An excellent example of the importance of leucocytes for suppura- tive softening is the caseous tubercle, which is usually free from leucocytes and does not undergo suppuration. If for any cause leuco- cytes are attracted into the caseous area, softening and pus formation promptly occur. Hence Heile*'^ found that while pus from a "cold" tuberculous abscess will not digest fibrin and does not give the biuret reaction, both reactions appear after a leucocytosis has been brought about by injection of iodoform. It was formerly considered that the softening was due to the digestive action of the enzymes of the in- fecting bacteria, many of which were known to produce digestive enzymes dissolving protein culture-media; e. g., Staphylococcus pyo- genes. Although to some extent these enzymes may be a factor in causing the softening of the fixed tissues and of the killed leucocytes, their effect is probably insignificant as compared with the enzymes liberated by the leucocytes, as shown by the production of active experimental suppuration under aseptic conditions with turpentine, croton oil, etc.** Suppuration is, therefore, the result of three proc- esses: (1) Necrosis of cells; (2) local accumulation of leucocytes; (3) digestion of the necrotic cells, fibrin, and tissue elements by en- zymes which are derived from three sources, as follows: (a) the leucocytes; (6) the infecting bacteria (if such arc present); (c) the fixed tissue-cells. Possibly small quantities of enzymes are also intro- duced in the blood plasma, but these are probably very inconsiderable. *" Buchner considers that bacteria will not produce suppuration unless they are broken down so that their -pyogenic proteins are released; e. g., anthrax bacilli cause suppuration when acting locally, as in malignant pustule, but not when they are causing septicemia, because only in the former case are their pyogenic proteins liberated 8^ Zeit. klin. Med., 1904 (55), 508. ** Apparently suppuration may occur in herpes zoster vesicles in the absence of bacteria, according to the findings of Kreibich (Wien. klin. Woch., 1901 (14) 683). COMPOSITION OF PUS 269 Normal serum, and probably also normal cells, contain antibodies for the proteolytic enzymes of the leucocytes, and hence neutralization or destruction of these antibodies must be an important factor in de- termining the rate and amount of suppuration.*^ The influence of the antionzj^mos is well siiown by the rabh)it, with serum rich in antienzymes and leucocytes poor in protease, so that infections with pus cocci do not usually lead to the formation of liquid pus (Opie). In man we see a similar relation, in that exudates rich in serum do not suppurate because the enzymes are inhibited by the serum; but if the excess of serum is removed suppuration may then occur. With an excess of enzyme (i. e., leucocytes) the inhibiting effect may also be overcome, and suppuration then begins. Variations in the proportion of leucoprotease and serum antiprotease determine, therefore, the occurrence of suppuration, and the inflammatorj'' re- action is seen to be fundamentally the same as the humoral reactions of immunity, in that in each case the essential process is the provision of proteolytic enzymes to remove foreign or abnormal protein sub- stances. In inflammation the proteolytic enzymes are brought in the leucocytes, in humoral reactions the enzymes are present free in the plasma. The antiproteases may be of the nature of lipoids, probably with unsaturated fatty acids ( Jobling) . The proteolytic enzymes of the leucocytes and tissue-cells have been previously considered in connection with the subject of autolysis (Chap, iii), and it is necessary here only to call attention to the fact that these enzymes are of at least two varieties: (1) Proteolytic enzymes of the polymorphonuclear leucocytes, which act best in alka- line mechum (Opie^"); (2) autolytic enzymes of the tissue-cells, which act best in an acid medium or after a preliminary acidification (Hedin, et al.). The mononuclear leucocytes contain, like the tissue-cells, enzymes acting in an acid medium. The antienzymatic action of serum is favored by an alkaline reaction, but is altogether lost in an acid medium (Opie). Composition of Pus Because of its method of production, pus consists of the follow- ing substances: (1) The constituents of the exuded blood plasma; (2) the constituents of the leucocytes (and tissue-cells) that exist free in the pus; (3) the products of digestion of the proteins of the leuco- cytes and necrosed tissues. All analyses of pus that are recorded in the literature are in harmony with the above statements. In general the analyses consider pus as composed of two chief portions, the pus corpuscles and the pus serum. As is to be expected, the composition of pus-corpuscles is simply that of a large mass of leucocytes, which »9See Opie, Jour. Exper. Med., 1905 (7), IHG; 1907 (9). 207; Arch. Int. Med., 1910 (5), 541. 90 Jour. Exper. Med., 1906 (8), 410. 270 INFLAMMATION contain minute quantities of substances taken up from the pus serum by absorption and phagocytosis.^^ The old analyses of pus-corpuscles by Hoppe-Seyler^2 ^re given in the following table: Table I. Quantitative Composition of Pus-cells {in 1000 parts of the dried substance). I II Proteins 137.621 Nuclein 342.57 [ 685.85 673.69 Insoluble bodies 205. 66 J Lecithin \ i /i o oo 75 . 64 Fat |14d.Sd 75 QQ Cholesterol. 74.00 72.83 Cerebrin 51 . 99 \ i no ca Extractive bodies 44.33/ iu^.»4 Mineral Substances in 1000 Parts of the Dried Substance NaCl 4.35 Ca3(P04)o 2.05 Mg3(P04)2 1.13 FePOi 1.06 PO4 9.16 Na 0.68 K trace As abnormal constituents of the leucocytes contained in abscesses may be mentioned glycogen, fat (from phagocytosis and from "fatty degeneration" of the leucocytes), and "peptone" (Hofmeister).^^ Pus serum differs from blood-serum chiefly in the substances added to it through the proteolytic changes that occur in the pus, and also in that it has lost its antiproteolytic property, containing instead free leucoprotease. The fibrinogen that escapes from the vessels into sup- purating areas becomes so altered that pus will not coagulate, even upon addition of fibrin ferment (defibrinated blood). The reaction of the serum is usually slightly alkaline, becoming strongly alkaline if much ammonia is produced, which occurs especially if there is sec- ondary infection with the organisms of putrefacton. Sometimes, however, lipase derived either from bacteria or from the cells causes the splitting of sufficient amounts of fatty acids from the fats to make the reaction acid; lactic and other fatty acids are also sometimes formed. Presumably the nature of the infecting organism will modify the reaction, for some (e. g., staphylococcus) cause an acid formation in media, while others (e. g., pyocyaneus) cause an alkaline reaction. Pneumococcus pus is said to become markedly acid.^^ Hoppe-Seyler's analysis of pus serum gave the following results, which ^1 The electrical conductivity of whole pus is somewhat greater than that of blood, and pus plasma conducts much more than whole pus, because of the resistance of the leucocytes (Tangl and Bodon, Biochem. Zeit., 1917 (84), 183). *^ Med.-Chem. Untersuchungen. " Zeit. physiol. Chem., 1880 (4), 268. " Netter, Bougault and Salanier, Compt. Rend. Soc. Biol., 1917 (80), 97. COMPOSITION OF PUS 271 show no considerable deviation from the composition of blood plasma, except in an increased proportion of fatty matter and extractive substances. Table II Quantitative composition Plasma of pus scrum normal I II III Water 913.7 905.65 908.4 Solids 86.3 94.35 91.6 Proteins 63.23 77.21 77.6 Lecithin 1.50 0.561 Fat 0 . 26 0 . 29 [ 1.2 Cholesterol 0.53 0.87 1 Alcohol extractives 1.52 0.73\ Water extractives 11.53 6.92/ 4.0 Inorganic salts 7 . 73 7 . 77 8.1 Quantitatively the chief abnormal constituent of pus serum is the so-called "p^jin" of the older writers, which is nucleoprotein de- rived from the decomposing leucocytes, and hence increasing in amount progressively with the age of the pus;^^ it is characterized by its insolubility in acetic acid. The same substance is found more abundantly in the entire pus, on account of the presence of the cells, and when treated with 10 per cent. NaCl solution it forms a stringy mass which was formerly called "Rovida's hyalin substance." Glu- cothionic acid, derived from the leucocytes, is also present in pus.^^ In the pus serum are found all the other constituents of the. leuco- cytes, including particularly lecithin, cholesterol, fats (and soaps), cerebrin, "jecorin," and glycogen; and also the usual components of the blood-serum as well as some small quantities of pigment derived from decomposed red corpuscles. The products of autolysis are of particular interest, and they are found in varying amount, but usually less abundantly than might be expected, probably because of their solubility and consequent rapid absorption. Albumoses and peptones seem to be constantly present (Shattock).^^ The common occurrence of albumosuria during sup- puration presumably depends on the absorption of digestion products from the pus,^^ but true peptone has not been satisfactorilj- identified in the urine. Leucine and tyrosine have also frequentlj^ been found in pus,^^ but Taylor^ could find no workable traces of either monoam- 56 Strada, Biochem. Zeit., 1909 (16), 193. 5« Mandel and Levene, Biochem. Zeit., 1907 (4), 78. " Trans. London Path. Soc, 1892 (43), 225. °^ Literature on albumosuria, see Yarrow, Amer. Med., 1903 (5), 452; Elmer, ibid., 1906 (11), 169; Senator, International Climes, 1905 (IV), series 14, p. 85. See also "Albumosuria," Chap. xxi. " Miiller (Cent. inn. Med., 1907 (28), 297) recommends the tyrosine reaction with MiUon's reagent as a means of dififerentiating tuberculous from ordinary pus, the former not giving the reaction because of lack of leucocytic enzymes; but there is disagreement as to the constancy of this reaction in pus (Dold, Deut. mod. Woch., 1908 (34), 869). 1 Univ. of California Publications (Pathol.), 1904 (1), 46. 272 INFLAMMATION ino- or polyamino-acids in a liter of pus, which may depend on their having been either absorbed or transformed into ammonium com- pounds. Presumably this is in part the explanation of the large urea excretion in persons with extensive suppuration, as observed by Ameuille.- From the nucleoproteins 'purine bodies are formed and may be found in the pus. The relation of the purine bases to local leucocytosis is shown by Heile,^^ who found in cold tuberculous ab- scesses a proportion of purine bases equal to 0.5 per cent., in similar ab- scesses after injection of iodoform, 1.57, and in acute suppuration, 10.7. Spermin crystals are also occasionally found in old pus col- lections.^ Free fatty acids and volatile fatty acids, such as butyric, lactic,^ valerianic, and formic, have been found. Products of bac- terial activity, such as bacterial proteins and pigments (e. g., pyo- cyanin), may also be present. It is probable that in many instances these autolytic products are bactericidal, and thus help to terminate the infection. Direct tests have shown that the autolysate of fibrin is bactericidal for staphylococci and streptococci.^ See also discussion of "Autolysis of Exudates" (Chap. iii). All the numerous enzymes of the blood plasma, the leucocytes and the tissue-cells, are present in pus. Thus Achalme® found evidence of the presence of the following enzymes in pus: proteolytic en- zymes,^ lipase (splitting monobutyrin), diastase, rennin (coagulating milk), gelatinase, catalase, and oxidase, the last being very abundant. These seem to exist chiefly in the leucocytes, the pus serum being quite free from them. No evidence could be found of enzymes act- ing on amygdalin, saccharose, inulin, or lactose. Fibrin ferment is said to be absent from pus, which is quite surprising in view of the fact that this enzyme is generally considered as being derived chiefly from the leucocytes. Presumably the bacteriolytic "endolysins" of the leucocytes are also present in pus. There is little difference in the effect on metabolism produced by a sterile suppuration and one due to localized bacterial infection (Cooke and Whipple),^ one of the chemical features in each being a precipitous and sustained rise in the urinary N excretion. Presumably the reac- tion in both cases results from toxic products of protein cleavage, rather than from bacterial secretions in the case of septic inflanmiation. Probably only part of the excessive urinary N comes from the local injury, the greater part being derived from toxicogenic destruction of tissue proteins. i^Bull. Acad. M(:^d. Paris, 1917, (78), 8. 3 See Williams, Boston Med. and Surg. Jour., 1901 (145), 355. '' d-lactic acid i.s a constant constituont of i)us from the pleura (Ito, Jour. Biol. Chom., 1910 (2(5) 173). " Bilancioni, Arch, (li Farmacol., 1911 (11), 491. 8 Compt. Rend. Soc. Biol., 1899 (51), 568. ^ Concerning proteolytic enzymes of pus see Opie, Jour. Exper. Med., 1906 (8), 410. 8 Jour. Exp. Med., 1918, (28), 222. COMPOSITION OF SPUTUM 273 Sputum" The chomistiy of sputum may be properly considered in this con- nection. In reaction, sputum is ordinarily alkaline, but in case of marked bacterial decomposition in cavities the reaction may become acid. Its specific gravity varies from 1.008 to 1.026, usually varying directly with the number of leucocytes; the average specific gravity is about 1.013 . The greenish color frequently observed depends gen- erally upon blood-pigment (except in case of icterus), although in some instances the pigment is of bacterial origin. Renk'° has studied the proteins of sputum with special reference to the loss of protein to the body and its relation to cachexia. In three patients (consump- tives) studied, the daily amount of sputum of two averaged 145 grams for each; for the third it was 82 grams. This contained (aver- age) 5 to 6 per cent, of soHds; including mucin, 2-3 per cent.; protein, 0.1-0.5 per cent.; fat, 0.3-0.5 per cent.; ash, 0.8-0.9 percent. The daily loss of nitrogen was 0.75 gram, which eciuals about 6 per cent, of the total daily nitrogen output of persons under condition of starva- tion." Wanner^- found characteristic variations in the amount of protein in sputum from different conditions, as follows: in bronchitis the amount of protein is very small ; in bronchiectasis protein is pres- ent, but the amount of uncoagulable nitrogen (due to autolysis) is relatively large; in phthisis as well as in bronchiectasis the amount of protein does not exceed 1 per cent.; in pneumonia it may reach 3 per cent., but it is highest in pulmonary gangrene. Any protein content that causes more than a shght turbidity on boihng indicates an inflammation; e. g., in case of doubt between a chagnosis of pneumonia and infarct a high protein content speaks for the former. Rogers^^ stated that the sputum in every case of tuberculosis shows albumin,'^ but this has been questioned, especially as to chronic or quiescent cases. ^* Albumin, or better, coagulable protein is also present in the sputum of patients with pulmonary edema and pleurisy. According to Works^^ in active tuberculosis there is usually 0.2 per cent, or more of coagulable protein in the sputum. The mucin of sputum yields 9 Complete bibliography given by Ott, "Chem. Pathol, der Tuberc," Berlin, 1903; Falk, Ergebnisse Physiol., 1910 (9), 406; Plesch, Hanb. d. Biochem., 1908 (HI (1), 7. '"Zeit. f. Biol., 1875 (11), 102. 11 Plesch (Zeit. exp. Path. u. Ther., 1906, Bd. iii, July) found that 4.8 per cent, of all the absorbed calories were lost in the sputum in an advanced case of phthisis. Under similar conditions the amount of salts excreted by the sputum may equal or exceed that in the urine (Falk, loc. cit.).^ 12 Deut. Arch. klin. Med., 1903 (75), 347. " Presse Med., 1910 (18), 289; 1911 (19), 409; also Ganz and Hertz, iUd., 1911 (19), 41; Kaufmann, Beitr. lOin. d. Tuberk, 1913 (26), 269; Hempel-Jorgensen, ibid., p. 392. 1^ Review by Cocke, Amer. Jour. Med. Sci., 1914 (148), 724. 15 Fischberg and Felberbaum, Medical Record, Oct. 21, 1911; Acs-Nagy, Wien. klin. Woch., 1912 (25), 1904. 1' Jour. Amer. Med. Assoc, 1912 (59), 1537. 18 274 INFLAMMATION 33.6 per cent, of glucosamin when split with HCl, which gives an in- dex of the quantity of mucin; this is highest in chronic bronchitis and lowest in pneumonia and phthisis. Kossel found 0.1-0.33 gm. of nucleins in the sputum dailJ^ The following table by Bokay (taken from Ott) gives the propor- tion of the organic constituents of sputum in parts per thousand: Table III Bronchitis in typhoid Fibroid phthisis Phthisis, early in apex Phthisis, [ Phthisis, Phthisis, cavities \ advanced ! advanced Fatty acids as fat Free fatty acids . . Soaps Cholesterol Lecithin Nuclein Protein 0.224 trace traces traces traces traces 0.898 0.845 0.184 0.380 0.4 traces 0.102 2.040 0.462 0.521 0.430 1.617 1.543 2.468 3.468 0.370 0.537 0.172 4.430 0.307 0.516 1.160 1.165 0.260 3.455 9.725 0.902 3.973 0.141 1.245 0.489 5.115 On account of the digestion of the exudates by the leucocj'tes, sputum contains proteoses, peptones, and amino-acids, generally in proportion to the richness of the exudate in leucocytes; they are, therefore, most abundant in pneumonia. Simon^^ found considerable albumose in phthisical sputum, but no nucleohiston or free histon. In febrile cases of tuberculosis the amount of albumose may exceed the coagulable albumen, which rarely exceeds one per cent, of the moist weight. ^^ Staffregen, however, could find no true peptone in phthisical sputum, but Stadelmann^^ found that such sputum con- tained enzymes hydrolyzing fibrin, and attributed this largely to bacteria. Probably most of the enzymes present in sputum come from the leucocytes. In the early stage of pneumonia the sputum has no proteolytic action, presumably because inhibited by the large amount of serum present; but with resolution active proteolytic prop- erties appear (Bittorf).^'' In tuberculosis sputum the tryptic and antitryptic properties fluctuate, and lipase is absent (Eiselt).-^ Pneu- monic sputum before the crisis has but slight action on peptids, but acquires marked peptolytic activity thereafter.-'- Most sputa con- tain enzymes spHtting casein and polypeptids.-^ Sputum may contain indole, derived either from the putrefying proteins or excreted from the blood. ^^ 1' Arch. exp. Path. u. Pharm., 1903 (49), 449. '« Prorok, Miinch. med. Woch., 1909 (56), 2053. i» Zeit. klin. Med., 1889 (10). 128. " Deut. Arch. klin. Med., 1907 (91), 212. 2' Zeit. klin. Med., 1912 (75), 91. " Abderhalden, Zeit. phvsiol. Chem., 1912 (78), 344. " Maliwa, Deut. Arch. klin. Med., 1914 (115), 407. " Binda and Cassarini, Gaz. Med. Ital., 1913 (64), 461. I COMPUSITIOX OF SPUTUM 275 Tho amount of fats seems to (icpoiul directly ui)on the number of pus-corpuscles and the age of the pus (i. e., the amount of fatty de- generation). Jacobson found from 0.08 to l.G grams of fatty matter per day, containing on an average 14.70 per cent, of soaps, 15.79 per cent, of higher fatty acids, 0-10 per cent, of water-soluble fatty acids, 13.58 per cent, lecithin, and 10.49 per cent, cholesterol. As to the inorganic substances, Bamberger found two types of spu- tum, catarrhal and inflammatory. In the inflammatory there is a deficiency in alkali phosphate, SO3 constitutes more than 8 per cent. of the salts, and the ratio, -^Y) equals ..• In catarrhal sputum the alkali phosphates constitute 10-14 per cent., j^ ^ = ^q' ^"d the SO3 is from 0.6-1.2 per cent. Chlorine is about the same in both forms. These differences are, however, not as constant as Bamberger believes, according to several later investigations. The results of his analyses are shown in the following table: Table IV Chronic Acute phthisis phthisis Water I 94.55 | 93.38 Organic matter 4 . 67 6 . 88 Inorganic salts 0 . 78 0 . 74 One hundred parts of the salts contain: Chlorine 35.78 33.40 SO., 0.70 0.80 P2O6 13.05 14.15 K2O 24.07 19.99 NaoO 27.90 31.69^ Calcium phosphate 1-63 4 . 32'* Iron phosphate 0.09 0.14 Magnesium phosphate 1 . 20 .... Ca and Mg carbonate and sulphate 1 . 74 0 . 22 Silicic acid 0.9 0.3 -^ Including magnesium. CHAPTER XII THE CHEMISTRY OF GROWTH AND REPAIR PROLIFERATION AND REGENERATION The factors that incite cells to proliferate, as well as those that cause the cessation of proliferation after it has once started, are too entirely unknown to permit of speculation as to their exact nature. It seems probable, however, that they are, as Ziegler says, "identical with the stimuli which excite or increase functional and nutritive activity," and these are certainlj' in manj'' instances of chemical na- ture. Thus the application of various irritating substances in not too concentrated a form (e. g., painting the skin with iodin) may lead to proliferation without causing discernible degeneration of the cells. Mallory's^ observations on the phenomena of proliferation and phagocytosis show that the same bacterial products which destroj' the cells when concentrated, when sufficiently dilute cause prolifera- tion of similar cells. Carnot and Lavlievre^ have obtained evidence that actively growing kidney tissue, whether fetal or adult regener- ating kidney, contains something which is capable of stimulating growth of renal epithelium when injected into other animals. Numer- ous dyes are known to stimulate cell growth greatly, {e. g., the growth of epithelium into oils containing sudan III, etc.) and sometimes seem to lead by virtue of this fact to cancer growth (e. g., cancer of the bladder in dye workers). Chemical products from decomposition of vegetable matter have a particularly active stimulating effect, so that what seem to be true cancers have been experimentally produced by paint- ing the ears of rabbits with tar (Yamagiwa). Manj^ other instances of proliferation in response to chemical stimuli might be cited, but in nearly all cases it is extremely difficult to determine that the proliferation is not, after all, reparative in compensation for degenerative changes, and, therefore, possibly obeying some other biological law than that of a simple reaction to a chemical stimulus. Perhaps the most striking example we have of growth stimula- tion by chemical agencies is furnished by the proliferation and hyper- trophy which take place in the uterus^ and mammary gland during pregnane}'. The sanu; phenomena can be produced by injecting the lipoid fraction of extract of placenta and corpus lutcum (Frank).* 1 Jour. Exp. Med., 1900 (5), 15. -' Arch. M('d. E.xper., 1907 (19), 388. ■'See Leo Loeb, Jour. Amer. Med. Assoc, 1908 (50), 1897; 1915 (04), 726. * Jour. Anier. Med. Assoc., 1920 (74), 47. 27() I'h'OLIFl'Jh'ATlOX AM) UEGEN ERATION '111 Even transplanted bits of uterine tissue are stimulated to j^row Ijy these substances, thus excluding possible nervous control of growth.* Dried placenta fed to mothers also increases the rate of growth of the suckling infant (Hammett).*^ The nature of the growth-stimulating agency in placenta and corpus luteum is unknown but it strongly resists chemical agents. However, it may be pointed out that tethelin, described by Robertson^ as the growth-promoting substance of the hypophysis, is soluble in lipoid solvents. Acromegaly and gigantism, give evidence that even far more than normal growth ma}' be produced, presumably through the agency of an internal secretion of the hypo- phesis, but whether tethelin actually is the substance responsible is at present unknown. This substance is obtained from the anterior lobe, about 10 mg. for each gland, and contains 1.4 per cent, of phos- phorous. It is said to retard growth of animals before adolescence and to increase post-adolescent growth; also it has been reported that wound repair is stimulated by tethelin.* At this time, however, the status of tethelin is not fully determined. The influence of the other ductless glands on growth is discussed further in Chapter xxii. The studies by Whipple and his colleagues on the repair of the liver after extensive chloroform necrosis indicate that a mixed diet rich in carbohj'drate is more effective in facilitating this repair than meat or fat, and that thyroid extract does not stimulate repair.^ Also the healing of wounds is more rapid in meat-fed than in fat-fed dogs. (Clark) ^^ Attempts to find specific substances that will cause in- creased rate of wound healing have so far been unsuccessful." Re- generation of blood protein after hemorrhage is said to be most rapid on a protein rich diet.^- When an incomplete protein, ghadin, is the sole protein of the diet the hemoglobin is not regenerated. Although proper nutrition is necessary for cell proliferation, yet it does not seem that excessive nourishment can lead to excessive cell multiplication, or by itself cause cell proliferation to take place. Oxygen and certain inorganic salts are essential for cell division even in the lowest forms, and among such simple organisms as sea- urchins and certain other marine forms segmentation of the unfertilized ova may be incited bj^ changes in osmotic concentration, leading eventually to formation of perfect larvae (J. Loeb, et. al.).''-^ In lower animals very dilute solutions of alkalies stimulate the rate of cell growth and somewhat higher concentrations cause extremely irregular cell division; in mammals the feeding of alkalies causes great wasting "' Frank, Surg. Gyn. Obst., 1917 (25), 329. " Jour. Biol. Chem., 1918 (36), 569; Endocrinology, 1919 (3), 307. ' See Jour. Exp. Med., 1916 (23), 631. ^ Review by Barney, Jour. Lab. Clin. Med., 1918 (3), 480. => See Arch. Int. Med., 1919 (23), 689. '" Bull. Johns Hopkins Hosp., 1919 (30), 117. 'iSee DuNoiiy, Anaer. Jour. Physiol., 1919 (49), 121. '2 Kerr et al, Amer. Jour. Physiol., 1918 (47), 456. '^See J. Loeb, Studies in General Physiology, Chicago, 1905. 278 THE CHEMISTRY OF GROWTH AND REPAIR as if through cell stimulation.'^ The products of nuclein hydrolysis are said to stimulate cell growth. ^^ Potassium salts seem to be particularly important for proliferating cells, and Beebe and also Clowes and Frisbie^® have found that actively growing malignant tumors are rich in potassium and poor in calcium, whereas in slow- growing tumors the reverse is the case. Dennstedt and Rumpf^'^ also found that in hypertrophy of the heart the amount of potassium is increased, while in chronic degeneration of the myocardium the calcium and magnesium are usually increased. The proportion of nitrogen in the different parts of the heart is not changed during hypertrophy (Bence),^^ but the amount of NaCl is much increased in hypertrophy.^^ Chemical studies of proliferation are lacking, ^^ except in regard to the development of the embryo, etc.^^ New tissues differ from adult tissues in having a large proportion of water, and in having a larger proportion of the ''primary" cell constituents and a smaller propor- tion of the various secondary constituents, since these last are largely products of the activity of the adult cell. Of the primary constitu- ents, the proportion of the nucleoproteins is particularly high, and a number of interesting facts concerning the nucleoproteins in cell di- vision have been determined. Most important, perhaps, is the clas- sical observation of Miescher, who found that during the migration of salmon up stream to the spawning grounds, during which time no food is taken, the proteins of the muscular tissue become largely transformed into the protamin type of protein (characterized by con- taining large proportions of the polyamino-acids, such as argininc histidine, and lysine),^- which unite with nucleic acids to form the abundant nucleoprotein of the spermatozoa and ova."^ Whether such a transformation of proteins occurs in mammalian cells during cell multiplication cannot be stated, but certainly from some source an additional supply of nucleoprotein is derived. Developing sea urchin 1^ Moore et al, Biochem. Jour., 1906 (1), 294; 1912 (6;, 162. '* Calkins et al, Jour. Infect. Dis., 1912 (10), 421. 1^ See "Tumors," Chap. xix. 1^ Zeit. klin. Med., 1905 (58j, 84. 18 Zeit. klin. Med., 1908 (66), 441. '9 Rzentkowski, ibid., 1910 (70), 337. "<* The composition of qranulalion tissue has been determined by Hirsch (Amer. Jour. Med. Sci., 1920 (159), 356, who analyzed the "castration granulomas" of swine. These are large inflammatory tumors, probably resulting from subacute infection of the operation wound, and consist of dense fibrous tissue with fow cells and a scanty blood supi)ly, but sometimes more or less edematous. His figures are as follows: Water 81.9 per cent.; solids, 18.1; lipins, 2.3; proteins, 14.4. The protein contained sulphur, 0.34 per cent; phosphorus, 0.32, purine N, 0.08. (Other details arc given on p. 519.) ^' Literature on the chemistry of growth given bj' Aron, Handbuch d. Biochem., Ergiinzungsband, 1913. ^^ Concerning protamins, see rt'sum6 by Kossel, Biochem. Centr., 1906 (5), 1 and 33. " 8{!e also (ireene, Jour. Biol. Chcm., 1919 (39), 435. PROLIFERATION AND REGENERATION 279 eggs synthesize great quantities of nucleoprotein,^* even when in a solution free from phospliatcs, and hero tlic only available sourc(! for the phosphoric acid of the nucleins would seem to be the phospholipins of the egg (J. Locb). The nucleoproteins during karyokinesis undergo a chemical change in that they become of a more acid type (presum- ably through splitting off of part of the proteins from the nucleic acid), which results in the characteristic increase in afh.nity for basic dyes, and the increased negative charge which is easily demonstrated." This suggests the participation of an enzyme in the process of karyo- kinesis, just as there seems to be in the production of pycnosis in de- generating cells, but there seems to be no conclusive evidence on this point. Gies^^ could find no enzyme in spermatozoa that incites cell division in the ova of sea-urchins (Arbacia). The fertilization of eggs makes them more permeable to ions," wliich possibly determines many of the subsequent changes. In metaplasia we have what may be interpreted as a chemical alter- ation due to mechanical stimuli, e. g., the formation of keratin by cells that ordinarily do not do so; the deposition of calcium salts and oste- oid transformation of connective tissues in rider's bone, etc. That such is the case, however, cannot be positively stated from the evidence at hand. CHEMICAL BASIS OF GROWTH AND REPAIR=8 We do not know just what substances are necessary to maintain individual cells in normal condition, what are needed to stimulate them to multiplication, or what elements they require to permit them to multiply, but it has been learned that certain definite materials are required by the organism as a whole. It is not sufficient that a given number of calories with a certain quantity of proteins, carbohy- drates, fats and salts be supplied; it is essential that certain specific constituents be provided among these foodstuffs. This fact was first clearly pointed out by Gowland Hopkins in 1906, although in 1897 Eijkman had discovered that beriberi and experimental neuritis might result from a one-sided diet of polished rice, and in 1902 Roh- mann reported that purified food stuffs do not suffice to maintain and rear mice. The proteins must not only provide a sufficient amount of nitrogen, but they must also provide certain specific amino-acids, as has been especially demonstrated by the investigations of Willcock and Hop- ^* Not accepted by Masing, Zeit. physiol. Chem., 1910 (67), 161. "See Gallardo, Arch. Entwickl. Organ., 1909 (28), 125; Pentimalli, ibid., 1912 (34), 444. 26Amer. Jour. Phvsiol., 1901 (6), 54. " See AlcClendoii, Carnegie Inst. Publ., 1914. No. 183. ^^ See Mendel, "Nutrition and Growth," Harvey Society Lectures, 1914-15; Amer. Jour. Med. Sci., 1917 (153), 1; Lusk, "Science of Nutrition," Saunders, Phila., 1917. 280 THE CHEMISTRY OF GROWTH AXD REPAIR kins^^ and Osborne and Mendel.^" Apparently the presence of some of the simple straight-chain amino-acids can be dispensed with (e. g., glycine), and the animal will grow and thrive if other nutritive supplies are adequate, but certain, at least, of the more complex cj'clic amino- acids must be provided. Furthermore, the requirements for growth (quantitatively speaking at least), seem to be something more than the requirements for mere preservation of health and equilibrium, for it was found that animals could live and preserve nitrogen equili- brium when the protein of the diet furnished at most small quantities of lysine, but young animals were unable to grow with such a restricted supply of this amino-acid. If lysine was added to the defective pro- tein (gliadin from wheat) the animal would then be able to grow at a normal rate. Of particular importance is the fact that animals can be kept in a stunted condition on such a deficient diet until they have reached an age at which normally all growth would have long since ceased, and then when supplied with sufficient lysine they will begin to grow and continue until full size is reached. ^^ This last observation proves that growth is not conditioned by age, and that we do not stop growing because a certain age is reached; the capacity for growth may remain latent and capable of exhibiting itself when proper conditions are furnished. But no amount of any amino-acid will cause a fulh^ grown animal to grow any more, so it would seem that the capacity for growth becomes extinguished when it has been utilized to a certain fixed extent, and remains potent until it has been completelj'^ utilized. If the only protein furnished contains no tryptophane the animal cannot maintain itself and slowly loses weight until it dies, unless tryptophane is supplied. If zein from corn, which yields neither ly- sine nor tryptophane, is the sole protein, then the animal cannot grow unless both lysine and tryptophane are added to the diet. So too. pure casein is not adequate to maintain growth because of its low content in cystine, but if cystine is added the nutritive value is much increased. That the pure isolated amino-acids can meet the deficiencies when added to the imperfect protein ration, demonstrates that proteins serve for food as amino-acids, and not as larger complexes. VITAMINES OR FOOD HORMONES AND DEFICIENCY DISEASES '- Not only must the proteins present certain essential chemical compounds to the living and growing organism, but also an adequate 29 Jour. Physiol., 1906 (35), 88. ^° Series of papers in Jour. Biol. Cheni., 1912, et seg. 'Mour. Biol. Chem., 1915 (23), 439. '2 See "Report on the Present State of Knowledge Concerning Accessory Food Factors (Vitainines), Special Report No. 38 National Health Insurance Act, London, 1919; "The Newer Knowledge of Nutrition," E. V. McCoUum, New York, 1919; Blunt and Wang, Jour. Homo Economics, 1920 (12), 1; also Sympo- sium in Jour. Ainer. Med. Assoc, 1918 (71), 937. V IT AMINES 281 supply of the essential inorganic salts and certain other, as yet un- identified, substances are necessary to permit of maintenance, growth and repair. It has long been recognized clinically that certain diseases, notably scurvy, may result from the absence of some essential in the food supply. More recently other diseases have been proved to have a similar cause, and the study of one of these, beriberi, has led to a better appreciation of the nature of the food essentials concerned. This disease seems to result from the use of polished rice as the chief constituent of the diet, and can be checked by feeding unpolished rice, or rice polishings, or even extracts of rice polishings, as first demon- strated by Eijkman. A somewhat similar condition may be produced readily in birds by feecUng them only polished rice, the chief feature being a severe neuritis, which is relieved with remarkable rapidity by supplying the food deficiency. This experimental neuritis of fowls {polyneuritis gallinarum) has served as a valuable means of study of diseases of this class, and led to the demonstration that not only ex- tracts of rice polishings, but also many other food materials, contain the essential materials without which health cannot be maintained. One of the early investigators of this subject, Casimir Funk,^^ gave to "the hitherto unrecognized essential dietary factors" the name "vi- tamines," which, in spite of certain logical objections, has been widely adopted; but as Lusk states, the term "food hormones" would be preferable in our present state of knowledge. Although so essential for life the amount required is very small, for whole rice is said to con- tain not over 0.1 gm. per kilo, and perhaps much less, of the active substance. McCollum^* has summarized the evidence that two classes of substances are necessary for maintenance. These he designated, for convenience, "/ai soluble A'" and '^water soluble B." It is the former that is lacking in xerophthalmia, and the latter in poljmeuritis. It now seems certain that other diseases are the result of deficiency in other specific substances, ^^ particularly scurvy, which seems to result from lack of a "water-soluble C." It also is an open question whether under the water-soluble B are included two separate vitamines, one antineuritic, the other growth-promoting.^^" As yet the exact identity of the active agents in water or fat solu- tions has not been determined. The fat-soluble vitamines seem to be especially abundant in butter, egg yolk, and cod liver oil, which pre- sumably accounts for the commonly accepted values of these partic- ular fats. They cannot be replaced by any of the known components of fats, including phosphatids, lipochromes, cholesterol, etc.,^^ and '' See Ergeb. Physiol., 1913 (13 j, 125, for review of his work. 3" Jour. Biol. Chem., 1916 (24;, 491. 3» See Jour. Biol. Chem, 1918 (33), 55. 35»See Mitchell, Jour. Biol. Chem., 1919 (40), 399. 36 See Drummond, Biochem. Jour., 1919 (13), 81; Palmer, Science, 1919 (50), 501. 282 THE CHEMISTRY OF GROWTH AND REPAIR are scanty or absent in lard, olive oil, and most vegetable oils. Funk believed the water-soluble antineuritic agents to be pyrimidine deriva- tives. They are dialyzable (Drummond) and are adsorbed by Fuller's earth (Seidell). Williams and Seidell" have found that hydroxypurines have marked anti-neuritic effects, and they sug- gested that an isomer of adenine- is responsible for the anti-neuritic action of yeast extracts. Later Williams^^ found an active hydroxy- pyridene, and suggested that the curative properties of yeast and rice polishings may be due to an isomeric form of nicotinic acid. These observations await confirmation, and we still are in the dark con- cerning the character of antineuritic vitamines.^^ The nature of the ' ' fat-soluble A " is, if possible, even less known than that of "water soluble B." Drummond's investigations^" show that it is somewhat heat resistant, but it is destroyed at 100° for one hour, apparently not through oxidation or hydrolysis. It cannot be ex- tracted from oils by water or dilute acid, but is extracted to some ex- tent by cold alcohol. If the fats are hydrolyzed at room temperature the active factor disappears, and it cannot be identified with any of the recognized components of fats. Because of its thermolability and other properties, Drummond is driven to the conclusion that ''fat- soluble A" is not a clearly defined chemical substance, but rather it is a labile substance, perhaps possessing characteristics resembling those of an enzyme. ^"^^ Vitamines, especially those that are water-soluble, also favor the growth of bacteria, ^"^ and are essential for the growth of yeast, so that Williams^^" has found it possible to determine the amount of this vitamine present in a food stuff by the rate of growth of yeasts thereon. Typhoid bacilli are said to produce vitamines during their growth, ^'^'^ and if it is true, as has been stated, that neither plants nor animals seem able to synthesize them, it would seem that they must be of bacterial origin. Yeast is known to produce water- soluble vitamine in particular abundance, but not the fat- soluble vitamine. Other sources of water-soluble vitamines are numerous, especially green vegetables and whole cereals, but they are not so abundant in meat or milk.^°« Why the vitamines are essential and how they act is unknown. It^s suggestive that they are found especially in cells with an active metabolism, but whether as a result of this activity or because essential " Jour. Biol. Chem., 1916 (26), 431. '« Jour. Biol. Chem., 1917 (29), 495. '^ See review by Drummond, Biochem. Jour., 1917 (11), 255. "» liiochem. Jour., 1919 (13), 81. ■"•"See also Steenbock and Boutwell, Jour. Biol. Chem., 1920 (41), 163. "''See D. J. Davis, Jour. Infect. Dis., 1917(21), 392; Kligler, Jour. Exp. Med., 1919 (30), 31. ^'"^ Jour. Biol. Chem., 1919 (38) 465; also Baehmann, ibid., (39), 235. ■""' Pacini and llussell, Jour. Biol. Chem., 191S (34), 43. """See Osborne and Mendel, Jour. Biol. Chem., 1919 (39), 29; 1920 (41), 515. i DKFICIKXCV IJISKASES 283 for cell growth is undeterminctl." Voddor''- has suggested that Ihe aiiti-ncuritic vitaniinc is essential for growth and repair of the nervous tissue, and in its absence normal wear cannot be made good.'*' There is evidence that substances rich in the anti-neuritic vitamine stimulate growth in infants. ""^ Moore"*^ suggests that deficiency diseases may- be the result of lack of something needed to neutralize toxic sub- stances produced in metabolism or derived from outside sources, just as in poisoning with tri-nitro toluene and similar compounds there is no intoxication so long as the body can furnish sufficient glycuronic acid to neutralize the poisons. Dutcher"" finds some relation between catalase and vitamine content in experimental polyneuritis, and sug- gests that the vitamines stimulate oxidative processes which remove toxic substances. It is probable that more than one vitamine is neces- sary for maintaining normal conditions, and deficiency of one causes beriberi, of another scurvy, for some dietaries lead to one disease and some to the other. Quite possibly minor, or less well-defined im- pairment in health may often result? from quantitative deficiency in vitamine supplies. The main points concerning the most studied deficiency diseases may be summarized as follows : Beriberi^^ occurs in two forms — the dry polyneuritic type and the edematous or wet beriberi, and mixed forms. The dry type resembles the experimental polyneuritis of birds, mentioned previously, for in the birds edema does not accompany the polyneuritis that can be produced experimentally by feeding polished rice. There now seems to be no doubt that human beriberi is the result of the absence of certain es- sential elements in the diet, observed especially when the diet is polished rice, but possibly occurring with other deficient diets, for the necessary vitamine is of course present in many other foods than rice. Not only has a condition closely resembling human beriberi been pro- duced in animals, but also true beriberi has been experimentallj' produced in man by feeding on polished rice, as well as the repeated demonstration of both prevention and cure of the human disease by proper feeding or by administration of rice polishings or extracts " See Voegtlin and Myers, Amer. Jour. Physiol., 1919 (48), 504. " Jour. Amer. Med. Assoc, 1916 (67 j, 1494. ''^ McCarrison has observed hypertrophy of the adrenals in pigeons with experi- mental beriberi, although the other organs are atrophied. He considers a general nuclear starvation from lack of necessary nuclear nutritive materials to be the essential condition. (India Jour. Med. Res., 1919 (6), 275.) He found the- sex glands particularly atrophied, and Houlbert (Paris Medicate, 1919 (9;, 473, has found water-soluble vitamines essential for growth of sex glands. Emmett and Allen (Jour. Biol. Chem., Soc. Proc, 1920 (41), liii), obtained adrenal hypertrophy with thymus atrophy in rats fed diets deficient in water-soluble B, but not with diets deficient in A. Growth of tadpoles also seemed to be more accelerated by B than bv A. ^* Daniels et al, .Amer. Jour. Dis. Chil., 1919 (18), 546. ^^ British Publ. on Munitions, No. 11. «» Jour. Biol. Chem., 1918 (36), 6.34. *^ Full discussion and bibliography given by Vedder in his book "Beriberi," New York, 1913. See also Jour. Amer. Med. Assoc, 1916 (67), 1494. 284 THE CHEMISTRY OF GROWTH AND REPAIR thereof. From rice polishings has been obtained a crystaUine sub- stance, of which a dose of 20 to 30 milHgrams will cure a polyneuritic bird. As stated above, the pure active substance has not been isolated and its exact nature is undetermined. Vedder believes that it is something that is needed for the repair of nervous tissue, so that in its absence the nervous tissues degenerate. The paralysis, he believes, depends more on central than peripheral- nerve changes, since the degeneration of the nerves precedes the paralj^sis and maj^ persist long after the paralysis has disappeared. As rice polishings relieve the cardiac symptoms, which are important features of beri- beri, it is to be assumed that the vitamine is essential for the heart metabolism. Furthermore, heart muscle contains vitamine which will protect from polyneuritis birds fed on polished rice. This does not seem to be identical with the vitamine isolated by Funk, for while it relieves the cardiac symptoms and dispels the dropsy of wet beriberi, it does not cure the paralytic symptoms of dry beriberi, according to Vedder. This author "has a growing belief that dry and wet beriberi are separate and distinct diseases, which are, how- ever, generally associated." Rice polishings, he says, clear up beri- beri dropsy quickly, but do not affect the paralysis unless the polishings have been hydrolyzed. Walshe^^ calls attention to the fact that starved fowls live long enough to develop beriberi, yet nevertheless do not show it, so he thinks that there must not only be a deficiency factor, but also some positive factor, which may be the abundant carbohydrate of the rice diet. Possibly in the absence of the vitamine the carbohydrate metabolism is altered, with the production of toxic substances. Furthermore, in spite of the marked clinical results obtained with rice polishings, the disease is not always cleared up as readily as might be expected if only a lack of vitamines was concerned, and, therefore,, there still remains the possibility that some infectious factor may play at least a subsidiary part in human beri-beri (see Mitchell 3^''). Keratomalacia or Xerophthalmia,'^ a condition of opacity of the cornea, followed by ulceration and blindness, seems to be specifi- cally due to lack of the fat-soluble vitamine which is present in egg yolk, butter fats, green leaves, etc., but not in lard or in many vege- table oils. This disease can be produced readily in experimental animals by feeding diets free from proper fats, and is relieved by administration of small quantities of these fats. I have had the opportunity to observe numerous instances of xerophthalmia among the famine sufferers in Roumania, and to observe its prompt relief under cod liver oil feeding. It should not ha confused with simple " (iuart. Jour. Med., 1918 (11), 320. ^8 Bloch, Ugeskr. f. Laeger., 1918 (80), 815. DEFICIENCY DISEASES 285 eye infections wliicli arc likely to occur in poorly nourished laboratory animals/^ Nutritional Dropsy. ("War Dropsy" or Famine Edema). "^^ This condition, which was observcul extensively durinj^ the war, especially among Russian prisoners in Germany, has been seen wherever famine occurs, and is undoubtedly caused bj^ dietary deficiency. Apparently it is independent of scurvy. As it is often associated with xerophthalmia it has been thought to depend on al)sence of fat-soluble vitamines. It seems more probable, however, that it results from low caloric supply, although protein deficiency combined with exces- sive fluid and salt intake in the effort to maintain life with weak soups, are probably important factors.^' Most of the adult subjects have been receiving 800 to 1200 calories per day, containing })ut 30 to 50 gms. of protein. It is much more likely to appear in undernourished persons who are compelled to work than in equally starved persons at rest, since work, and also cold, increase the caloric deficiency. Numerous studies of metabolism and blood chemistry in persons exhibiting war dropsy have given concordant results, which show the extreme depletion of the body in all nutritive reserves.*^ The blood shows hypoglucemia, decrease in potassium, fatty acids, phospholipins and residual nitrogen, with anincrease in NaCl, and of acetone bodies and ammonia from starvation acidosis. There is also a decrease in the amount of protein in the blood even to one-half the normal amount, with hj^dremia and a less marked decrease in both red and white cells. The lack of reserve nitrogenous material both in blood and tissues is shown by the fact that when patients with war edema are fasted a few days the N excretion may fall to 2 to 3 gms. per day, while in absolute starvation of previously normal persons the N output usually is 10-12 gms. (Falta). Schittenhelm and Schlecht suggest that the edema is merely the result of injury to the capillary endothelium, in common with all the other tissues, whereby their permeability becomes increased. Famine edema seems to be closely related to the edema often observed in infants kept on a preponderatingly starchy diet, such as • barley water, for long periods. Here most striking degrees of dropsy are observed, which seem in all respects similar to famine dropsy. There is probablj^ also a close relation to the edema of pernicious anemia and cachexia. The relation of this form of edema to the " See Bulley, Biochem. Jour., 1919 (13). 103. "Full Review by Schittenhelm and Schlecht, Zeit. exp. Med., 1919 (9), 1; Maver, .lour. Amer. Med. Assoc, 1920 (74), 934. ^' See Guillermine and Guyot, Rev. Med. Suisse Rom., 1919 (39), 115; Falta, Wien. klin. Woch., 1917 (30),' 1637; Schittenhelm and Schlecht, Zeit. exp. Med., 1919 (9), 82. "See Schittenhelm and Schlecht, loc. cit.; Feigl, Biochem. Zeit., 1918 (85), 365; Jan&en, Miinch. Med. W'ocK, 1918 (65), 925; Biirger, Zeit. exp. Med., 1919 (8), 309. 286 THE CHEMISTRY OF GROWTH AND REPAIR edema of wet beriberi has not been determined, but it is highly prob- able that their origin has something in common. Both are dropsies due to diet deficiency, and it may well be that the deficiency is the same in each case. In both these conditions, as well as in the "Mehlnahrschaden" of starch-fed babies, there is the common ele- ment of relatively excessive carbohydrate supply, which may have something to do with the dropsy. The clinical evidence is against the view that nutritional edema depends on a lack of specific vitamines.^-" Experimental work supports the clinical evidence as to the etiology of nutritional edema. Miss Kohman^^ has found that rats fed diets composed chiefly of carrots often develop a severe edema, which is prevented by supplying protein, but not by butter fat or starch. Evidently neither fat-soluble nor water-soluble vitamines are respons- ible. It was found that on a dry diet of equal caloric and protein de- ficiency the rats are not so likely to develop edema. Experiments done in my laboratory by M. B. Maver"^" agree fully with those of Miss Kohman. Apparently low protein and high fluid intake are the most essential factors, although relatively high carbohydrate must also be considered. Scurvy would seem almost certainly to be a deficiency disease, but there has been much disagreement as to this point, especially among those who have studied experimental scurvy in animals. There is room for doubt that the expei'imental disease in animals is identical with human scurvy, at least there is reason to believe that more than one concHtion has been described as scurvy in experimental animals. Apparently guinea pigs, however, develop readily a disease which resembles scurvy very closely both anatomically and in its relation to dietary conditions. Hess, who has studied especially infantile scurvy, finds that orange juice given intravenously will relieve scurvy, and thus apparently disposes of all theories of gastrointestinal disorders as the responsible factor. Artificial "orange juice," (containing sugar, citric acid and inorganic salts in the proportions found in natural orange juice) is ineffective, so that apparently scurvy is the result of lack of some undetermined substance present in orange juice as well as in other fresh vegetable foods. As yet we have no evidence as to the character of this "vitamine," which has been designated as "water-soluble C," but it is probably quite distinct from either water- soluble B or fat-soluble A,^^ and Hess believes that our ordinary dietary probably does not contain any great excess of the antiscorbutic element, since scurvy so readily appears when the necessary vegetable foods arc reduced in amount. According to Chick and Hiune this vitaminc is present in living vegetable and animal tissues, in largest «2«See Burger, Zeit. Exp. Med., 1919 (8), 309. " Denton and Kohman, Jour. Biol. Chem., 1918 (36), 249; Kohman, Amer. Jour. Physiol., 1920 (51), 378. "Cohen and Mendel, Jour. Biol. Chem., 191S (35), 425. ,•-»' DEFICIENCY DISEASES 287 amounts in fresh fruits and grcrn vegetables, to a less extent in root vegetables and tubers. It is present in small amount in fresh moat and milk, and has not 3'et been detected in yeast, fats, cereals, pulses. The explorer, Stefansson,^^ has reported observations indicating the presence of antiscorbutic substances in raw meat, and their absence or deficiency in well-cooked meat and tinned foods. Evidently this antiscorbutic element is very unstable, since even drying vegetables at moderate temperatures, 6.5-70°, and cooking or salting meats, or heating with weak alkalies, destroys or greatly reduces their antiscor- butic value. ^^ Pasteurization of milk also reduces the preventive value of this food" which, in its raw state, contains in abundance all necessary factors for nutrition, but apparently little more of the antiscorbutic substance than is barely sufficient to maintain health. Pellagra^^ probablj- belongs among the deficiency diseases, despite numerous attempts to account for it as an infectious disease. The work of Goldberger^^ is especially valuable in affirmative evidence of the relation of dietary deficiency to pellagra. Here again we are entirely uninformed as to the nature of the deficiencj'. Goldberger®'' sums up his conclusions as follows: "The pellagra-producing dietary fault is the result of some one, or, more probably, of a combination of two or more of the following factors: (1) a physiologically defective protein supply; (2) a low or inadequate supply of fat-soluble vitamine; (3) a low or inadequate supply of water-soluble vitamine, and (4) a defective mineral supply. The somewhat lower plane of supply, both of energy and of protein, of the pellagrous households, though appa- rently not an essential factor, ma}', nevertheless, be contributory by favoring the occurrence of a deficiency in intake of some one or more of the essential dietary factors, particularlj^ with diets having only a narrow margin of safety. The pellagra-producing dietary fault may be corrected and the disease prevented by including in the diet an adequate supply of the animal protein foods, particularly milk, in- cluding butter and lean meat. " McCollum calls attention to the fact that a diet is not adequate unless it contains active metabolizing protoplasm, as found in green leaves, eggs, meat and milk; and pellagra-producing diets are largely composed of seed foods and pork fat. To make cereal grains diete- tically satisfactory there must be added inorganic elements, a protein, and substances containing "fat soluble A."^' " Jour. Amer. Med. Assoc, 1918 (71), 1715. " Givens and Cohen, Jour. Biol. Chem., 1918 (36), 127; Amer. Jour. Dis. Chil., 1919 (18), .30. *' See Hess, Amer. Jour. Dis. Chil., 1919 (17), 221. '* Concerning metabolism in pellagra see ]\Iyers and Fine, Amer. Jour. Med. Sci., 1913 (145) 705. Chemical changes in the central nervous system described by Koch and Voegtlin, Hygienic Lab. Bull. 103, 1916. "Public Health Rep.,' 1914 (29), 1683; 1915 (30), 3117, 3336. «» Jour. Amer. Med. Assoc, 1918 (71), 944. «> Jour. Biol. Chem., 1919 (38), 113. 288 THE CHEMISTRY OF GROWTH AND REPAIR Whatever the deficiency in diet may be, pellagra seems to develop most often in persons whose diet is preponderatingly maize seed products. Only in countries where maize is the chief dietary staple does pellagra occur with any great frequency, and in those countries where part of the population lives chiefly on maize, and other groups live on other foods, pellagra occurs chiefly or only in the former group. I have had the opportunity to observe much pellagra in Roumania during a period of protracted and serious food shortage, and this relation to maize was most striking and convincing. The peasants of this country have for their cliief food a thick mush of boiled, coarsely ground corn meal, called mamaliga, supplemented by such other foods as they can secure. Dwellers in the towns rely on bread from wheat flour as their chief carbohydrate supply, and have a much more abun- dant and varied list of accessory foods. Pellagra is prevalent in Rou- mania, but restricted to the maize-eating peasants, and in very definite relation to their inability to secure accessory food stuffs. While Roumanian physicians seem generally inclined to accept the theory that spoiled maize is responsible, my own observations would indicate that the chief difficulty is lack of accessory foods. The relation of maize to pellagra becomes particularly striking if we compare Rou- mania with a country where maize is not a staple food, such as Korea. Here for centuries a large part of the population has existed on the verge of starvation, the chief food being rice. Although here beriberi is common enough, especially among those who can afford the luxury of polished rice, pellagra is not observed, despite a much greater deficiency in total food supply, both as regards calories and acces- sories, than prevails in Roumania. Despite the abundant evidence of the relation of dietary deficiency there are those who interpret existing evidence as establishing or making probable that pellagra is nevertheless essentially an infectious disease. ^^ The compromise view that pellagra is an infectious disease which can only manifest itself among those suffering from dietary deficiency has also been supported, especially by INIcCollum.''^ Rickets. — Mellanby^^ holds that this disease results from a defi- ciency in fat-soluble vitamine, although admitting that the efficiency of malt extracts and lean meat in preventing experimental rickets is not in harmony with this hypothesis. As the total growth of rachitic puppies on a diet poor in fat-soluble A is about normal, he suggests that this agent is not necessary for growth, but merely for making growth normal. Without "A" the development of teeth is much interfered with. The recognized value of cod liver oil in rickets is in support of this view. Hess, McCollum and others do not accept the hypothesis that rickets is solely the result of lack of fat-soluble A. " See Jobling and Peterson, Jour. Infect. Dis., 191G (IS), 501. " Proc. Amer. Philos. Soc, 1919 (58), 41. «' Jour. Physiol., 1919 (.52), liii; Lancet, 1919 (19G), 407. DEFICIENCY DISEASES 281) The former has observed children who have hved a long time in per- fect health on a diet with a minimum of vitamine-containing fats.*° McCollum finds that rats develop a condition apparently identical with true rickets when kept on a diet deficient in any two of the three essentials, viz., protein, calcium, fat-soluble A. A diet lacking only one essential seems to be well borne. "Jour. Amer. Med. Assoc, 1920 (74), 217. CHAPTER XIII DISTURBANCES OF CIRCULATION AND DISEASES OF THE BLOOD THE COMPOSITION OF THE BLOOD The function of the blood being to maintain an equilibrium in the temperature, chemical composition and osmotic pressure between all parts of the body, it follows that it is never of exactly the same composition in any two places or at any two times. To the extent that every tissue is continually giving off something to the blood, we may consider that every organ is a factor in its formation, and as a result of this multiplex origin of the blood, the substances it may con- tain are beyond enumeration. There are probably but few chemical substances occurring in the tissue-cells that do not also occur in greater or less amount in the blood. In addition to these there are also the substances characteristic of the blood itself, besides a host of substances of unknown nature, apparently manufactured in re- sponse to the stimulation of substances entering the body from out- side; for we find that the blood of every adult individual contains substances that make him immune to a multitude of diseases that he has had in childhood, as well as substances that in later life protect him to a greater or less degree from infection by such organisms as the colon bacilli of his intestine, the pneumococci and streptococci in his throat, etc. We have learned of these defensive substances within very recent times, and also of the "antienzymes" that possi- bly protect the blood from the digestive enzymes of the body cells. What other substances of importance we may yet find in the blood is an open question. There are no apparent limits to the possibilities of the study of the blood, for it represents a little of every organ, and much that is characteristic of itself. In discussing briefly the substances that have been isolated from the normal blood, before considering the changes that occur in it during pathological conditions, we may roughly divide the blood into the formed elements and the plasma in which they are suspended. THE FORMED ELEMENTS.— By weight, tlie red corpiu«clos constitvito from 40 to 50 por cent, of tli(> l)l()0(l, tlie i)crccnt:i}i;c varyinji; uiiiUn- (lilTc-ront conditions, while llie total weif^ht of the leucocytes and platelets is insi<;iuticant. 'I'he henio- fr|ol)in constitutes troni Sti to 94 i)er cent, by \veit;;ht of tiie solids of the red cor- puscles, but the i)liysical and chemical relations that it bears lo the stroma of the corpuscles are as yet undetermined (see "Hemolysis"). t)f tlie remaining constit- 290 COMPOSITION OF BLOOD '291 uents of the corpuscles, from 5 to 12 ywr cent, consist of proteins, i)robal)ly chiefly globulins and nuclcoprolcins; 0.3 to 0.7 per cent, of lecithin; and about 0.2 to 0.3 per cent, of cholesterol (Iloppe-Seyler). The outer coat of the red corpuscles docs not seem to be equally permeable for all substances, and therefore we find the composition of the fluid p(jrtion of the cell quite different from that of the plasma about it. The salts of tlie corpuscles consist largely of potassium pho.s- phate, a little sodium chloride, some magnesium, but no calcium,' which is quite different from their proportion in the plasma. Probably many of the other con- stituents of the plasma, especially urea, ])enetrate the red corpuscles to a greater or less degree, but most of them, particularly the sugar, remain chiefly in the plasma. Hemoglobin, the most characteristic constituent of all the heterogeneous com- ponents of the blood, is a compound protein, and probably exists combined with some other constituent of the corpuscle, most probably the lecithin. It splits up readily into a protein, glohin, and an iron-containing substance, hemochromo- gcn, which readily takes up oxygen to form hematin. Only about 4 to 5 per cent, of the hemoglobin is hemochromogen, and iron constitutes but about 0.4 per cent. Hematin may be further split up into other substances, which will be considered in the discussion of "Hemorrhage." The leucocytes consist chiefly of nucleoproteins, with probably some globulin, and they also contain glycogen, phospholipins, and cholesterol. The blond-platelets are believed to be largely nucleoprotein, but little is known of their actual composi- tion; microchemical examination shows no evidence of either fat or glycogen.* BLOOD PLASMA differs from blood-serum in that the latter is formed from the former through the removal of the fibrinogen through its conversion into fibrin. Serum, therefore, contains no fibrinogen, but more fibrin ferment; otherwise it is practically the same as the plasma.^ It is well for us to appreciate that the blood is fundamentally a tissue, with its more solid structural elements lying in a pro- toplasm, the plasma, somewhat more dilute than the protoplasm of other tissues but in other respects much the same. Proteins. — Fibrinogen has the general properties of a globulin, with also a peculiar tendency to go into the insoluble form, fibrin. (This process will be discussed under "Thrombosis.") In the plasma are also other globulins,'* one soluble in water (pseudo-globidin), the other insoluble in water {euglobulin) . Serum-albumin, another protein of the plasma, probably consists of two or more varieties of albumin. There are also nucleoproteins {prothrombin) and non- coagulable proteias, which being poorly understood have been variously considered as glycoproteins, or mucoids, or albumoses. The serum proteins seem to be closely related to, or compounded with, the lipins of the plasma. Other Constituents. — The fat of the plasma varies much according to the time which has elapsed after the taking of food; in fasting animals it amounts to from 0.1 to 0.7 per cent. The sugar fluctuates less, being normally about 0.1 per cent., whUe the urea has been estimated at 0.03 per cent. Most of the sugar is dex- trose; but probablj' there is some levulose, possibly some pentose and other forms, and possibly also sugar combined with lecithin (jccorin) or other substances. Soaps, cholesterol, and phospholipins exist free in the plasma. There are also the numerous nonprotein nitrogenous substances that are excreted in the urine. Plasma differs strikingly from the corpuscles in that its inorganic substances are chiefly sodium and chlorine, while potassium and phosphoric acid are almost entirely absent. Another important fact is that when the plasma is combusted,, the acid radicals remaining do not suffice to balance the bases, indicating that much of the inorganic bases is joined with organic substances, probabh' as ion- ' The current statement that corpuscles are impermeable for calcium is refuted by Hamburger (Zeit. physikal. Chem., 1909 (69), 663). - Aynaud, Ann. Inst. Pasteur, 1911 (25), 56. ^ In the process of clotting certain changes occur, probablj^ physical, that may make the plasma more or less toxic (see Anaphyla.xis) and apparently alter its biological properties, since the reinjection of a person's own defibrinated serum may cause marked physiological and therapeutic effects {e. g., autoserotherapy in psoriasis). Especially noteworthv is the vasoconstrictor effect of defibrinated blood (see Hirose, Arch. Int. Med., 1918 (21), 604). * Literature given by Rowe, Arch. Int. Med., 1916 (18), 455. 292 DISTURBANCES OF CIRCULATION protein compounds. The alkali joined to the protein is non-diffusible, and con- stitutes about five-sixths of the total alkali. The concentration of the electrolytes of the blood has been determined by ascertaining the lowering of the freezing-point, which in human blood averages about 0.526°; this corresponds closely to the effect of a salt solution of 0.9 per •cent, strength. About three-fourths of the dissolved molecules of the blood-serum are electrolytes, and about three-fourths of these are molecules of NaCl, most of which are in the dissociated state.^ The calcium content is very constant, about 9 to 11 mg. per 100 cc. of plasma. Enzymes. — A large number of enzymes exist in the blood, the following being among those that have been detected: diastase, glucase, lipase, thrombin, rennin, and 'proteases. The proteases and perhaps the other enzymes are held in check to a large extent by " antiferments" that are also present (see "Enzymes")- In relation to the antiferments are the innumerable antibodies that exist normally in the serum for foreign proteins, foreign cells, and for bacteria and their toxins, as well as those resulting from reaction, etc. The proportions in which the constituents of the plasma normally occur have been determined by Hoppe-Seyler and by Hammarsten, as follows:* Table I No. 1 No. 2 Water 908.4 917.6 SoUds 91.6 82.4 Total proteins 77 . 6 69 . 5 Fibrin 10.1 6.5 Globulin 38.4 Seralbumin 24 . 6 Fat 1.2 Extractive substances 4.0 Soluble salts 6.4 12.9 Insoluble salts 1.7 No. 1 is an analysis by Hoppe-Seyler. No. 2 is the average of three analyses made by Hammarsten. Reaction. — If we titrate the blood plasma with an acid, we liberate much of the alkali from the proteins, dissociate all the Na2C03 present, as well as the NaHCOs and the sodium phosphate, and find in this way that the entire fresh blood contains neutralizable alkali corresponding tc a solution of Na^COs of about 0.443 per cent, strength (Strauss;. In other words the blood has a quantity of alkali in combination that can be drawn upon to neutralize acids to the extent indicated by the above figures. The real alkalinity of a fluid, however, is dependent upon the number of free OH ions in the solution; and Hober has determined by physico-chemical methods that the concentration of OH ions in blood is but little greater than in distilled water.'' Michaelis* has found the H+ concentration of the blood to be 0.45 X 10"', as contrasted with neutrality at 38° which is H4- = 1.5 X 10~'. The interchange between COo, phosphates and carbonates in the blood is such that it is impossible for any considerable quantities of free H or OH ions to exist, and the protoplasm is thus protected from an excess of either. The capacity of the blood to neutralize acids and alkalies is sometimes referred to as its "buffer value."^ According to Henderson'" not more than five parts of excess free H or OH ions can be present in ten billion parts of protoplasm. An alkalinity is impossible because this would cause an increased osmotic pressure which the kidneys would regulate; acidity is impossible because death would result ^ Concerning relation of conductivity to freezing-point see Wilson, Anier. Jour, of Physiol., 1906 (16), 438. '^ For complete analyses of the blood see Abderhalden, Zeit. physiol. Chem., 1S9.S (25), 106. ' For bibliography on Alkalinity of Blood, see Henderson, Ergcbnisse Physiol., 1909 (8), 254. 8 Deut. med. Woch., 1914 (40), 1170. " See Levy and Rowntree, Arch. Int. Med., 1916 (17), 525. »" Amer. Jour. Physiol., 1907 (18^. 250; 1908 (21;, 427. IlKMOh-h'/fACK 29;i from the inability of the blood to carry CO2. The blood and tissue jjrotcinf also can bind much of either H or OH ions," so that the preservation of neutrality is elaborately guarded. In the tissues, because of the profluc.tion of acids during metabolism, the Il-ion concentration is slightly higher than that i)f the blood, being estimated by Michaelis at exact neutrality, 1.5 X 10"'. Presumably one impor- tant purpose of the e.xact regulation of reaction is to provide proper conditions for enzyme action. The alkali of the blood exists in part as alkaline salts, carbonate and phosphate (the diffusible alkali), and partly combined with protein {non-diffusible alkali). As the corpuscles are richer in diffusible alkali than the plasma or serum, the num- ber of corpuscles modifies the alkalinity of the blood decidedly. Much importance is attached to the question of the alkalinity of the blood for two reasons: first, in certain conditions of disease the blood contains so much of organic acids that the alkali is partly saturated and the power of the blood to carry CO., is lessened, with serious results (see "Acid Intoxication," Chap, xx); and, second, the bacteri- cidal power of the blood is found to vary according to its alkalinity. In fact, metabolic activity seems generally to be favored V\v certain degrees of alkalinity; for example, J. Loeb^- found that sea-urchin eggs develop with much greater rapidity if a small amount of OH ions is free in the .-:ea-water. Brandenburg'^ states that the non-diffusible alkali varies according to the amount of protein in the blood; in pneumonia and acute nephritis he found it low. In cancer the titrable alkalinitj' is distinctly increased, and Moore and Walker'* hold that this is due to an increased alkalinity of the proteins of the blood. Awerbach'* claims that in severe high fevers the bactericidal effect of the blood alkalinity is increased (see also "Passive Congestion" for further discussion concerning the relation of alkalinity to bactericidal power). Viscosity of the Blood." — Normal blood is about five times (4.5 times, Austrian) more viscous than water, chiefly because of the corpuscles and the dissolved pro- teins. This viscosity does not vary directly with the specific gravity or the hemo- globin, but is closely related to the number of red corpuscles (Burton-Opitz) ; laking the corpuscles increases the viscosity considerably. Most salts increase the viscositj', but some, especially iodides, are said to reduce it. Carbon dioxide increases viscosity greatly, even when in amounts possible in the circulating blood. Anemia decreases the viscosity, approximately in proportion to the number of corpuscles; polj'cythemia is accompanied by a corresponding increase; leukemia, because of anemia, shows a decrease; in nephritis there maj' either be an increase or a decrease in the viscosity, not corresponding in any way to the blood pressure. Cardiac disease with edema shows low viscosity because of the anemia and hydre- mia, but if there is polycj'themia and no edema the viscosity may be high. .Jaun- dice causes an increase, diabetes gives variable results. Typhoid causes no charac- teristic change beyond that resulting from anemia, and in pneumonia the cyanosis and salt retention usually cause an increase (Austrian). Gullbring'" found the viscosity to vary directly with the per cent, of neutrophiles. As blood viscosity depends largely upon the corpuscles, it increases with reduction in the size of the lumen of the tube through which it passes, unlike a true solution; hence with narrow capillaries the viscosity is abnormally high until we reach the point where the corpuscles plug the capillary. HEMORRHAGE Hemorrhages result from an altered condition in the vessel-walls, which may be due either to trauma or to chemical injuries. Of the chemical agencies causing hemorrhages, bacterial products arc the 11 See Robertson, Jour. Biol. Chem., 1909 (6;, 313; 1910 (7), 351. 1= Arch. f. Entwicklungsmechanik, 1898 (7), 631. 13 Deut. med. Woch., 1902 (28), 78; Zeit. f. klin. Med., 1902 (-45), 157. 1^ Biochem. Jour., 1906 (1), 297; good discus.sion of blood reaction. 15 Med. Obosrenije, 1903, p. 596. 1^ Review of literature by Determann, Zeit. klin. Med., 1910 (70), 185; also Austrian, Johns Hopkins Hosp. Bull., 1911 (22), 9. See also Traube, Internat. Zeit. physik-chem. Biol., 1914 (1), 389. '" Beitr. klin. Tuberk., 1914 (30), 1. 294 DISTURBANCES OF CIRCULATION most important practically, but many poisons, such as phosphorus, formaldehyde, yhytotoxins (ricin, abrin, and crotin), and zootoxms (snake venoms) cause numerous and abundant hemorrhages. For- merly, the tendency was to ascribe hemorrhages from the above causes to mechanical injury of the vessels by thrombi, or by emboli of ag- glutinated corpuscles, but the work of Flexner^^ has shown that venoms cause hemorrhages by injuring the capillary walls, so that actual rents are produced by the intravascular pressure, and it seems highly probable that hemorrhages are produced by other chemical substances in a similar way. We may, therefore, refer such hemor- rhages to an endotheliotoxic action of the poison, or to a solvent effect upon the intercellular cement substance. In the case of ordinary chemical poisons the endotheliotoxic action is not specific, but with some of the toxins it seems to be quite so; for example, rattlesnake venom contains an endotheliotoxic substance ijiemorrhagin) , which seems to be a specific poison for endothelium, and which is the most dangerous constituent of the venom. If we immunize animals against tissues containing much endothelium (e. g., lymph-glands), their serum will be found to contain endotheliotoxins, so that when tliis serum is injected subcutaneously into a susceptible animal, large local hemor- rhages result; if injected into the peritoneal cavity, there results marked desquamation of the endothelial cells, which soon undergo de- generative changes (Ricketts).^^ It is quite probable that the bac- terial poisons that cause marked hemorrhagic manifestations likewise contain endotheliotoxins, although this matter does not seem to have been investigated. Even hemorrhage by diapedesis seems to be due to, or at least associated with, chemical changes in the capillary walls, for Arnold-*' found that when capillaries from which diapedesis had occurred were stained by silver nitrate, dark areas were found between the endothelial cells. As silver nitrate is a stain for chlorides, and dark- ens intercellular substance because it is rich in sodium chloride (Macallum), it is probable that there is an increase in the amount or a difference in the method of combination of the chlorides of the cement substance between the endothelial cells at the places where red corpuscles escape. M. H. Fischer-^ suggests that diapedesis results from a change in the endothelial cells, which under the inthi- ence of acids or other agents of metabolic origin become excessively hydrophihc, swell up, and become so softened that corpuscles may pass directly through the cell, just as a drop of mercury can pass through a sufficiently soft jelly without leaving a hole in the jelly. Ilemori'hage in cachetic conditions is often ascribed to changes in the vessel-walls due to malnutrition, but it is diffi.cult to imagine " Univ. of Penn. Med. Bull., 1902 (15), 355. " Trans. Chicago Path. Soc, 1902 (5), 181. 20 Viichow's Arch., 1875 (()2), 157. 2' "Nephritis," New York, 1912, p. 78. HEMOinaiAdE 295 capillary walls suffering from lack of nourisliiiicnt, even with the poorest of blood, and it seems more probable that the hemorrhages are due, even in cachexia, to chemical constituents of the blood that in- jure the endothelium. Hemorrhages that follow re-establishment of the circulation after complete occlusion, however, may be the result of asphyxial clianges in the capillary walls, presumably colloidal swell- ing of the cells. After severe hemorrhages the blood shows a decrease in specific gravity and viscosity, an increase in surface tension and electrical resistance, and either increase or decrease of the freezing-point de- pression, all these changes being transient if the individual is other- wise normal.'^ (See also Secondary Anemia.) There is a rapid absorption of fluid from the tissues and tissue spaces, resulting in a dilution of protein and formed elements, but not of salts. For the same reason the density of the blood decreases in direct relation to the proportion of the total blood that has been lost.-^ The alkali reserve of the blood is somewhat lowered by severe hemorrhage, ^^ but there is not a marked acidosis. The total nitrogen of the blood of course falls, but there is a tendency for sugar, urea and non-protein N to increase, and there is increased elimination of creatine in the urine, presumably from destruction of muscle tissue to replace the lost blood proteins. There is said to be a decreased permeability of vessels, resulting in reduced exudative processes. ^^ The proportion of the several blood proteins is variably altered after repeated hemorrhages; the sugar is little affected^^ but there may occur a marked rise in the content of immune bodies, especially specific agglutinins.^^ Rapid hemorrhages cause a decrease in the coagulation time because of a decrease in antithrombin and a slight increase in prothrombin, in spite of a decrease in fibrinogen.-^ If the blood is withdrawn re- peatedly in large amounts, centrifuged, and the washed corpuscles reinjected suspended in isotonic salt solution (plasmaphaeresis), life can be maintained even after 4 to 5 times the total volume of blood has been removed and washed. This is possible because of rapid reformation of the plasma, and the blood shows the changes characteris- tic of secondary anemias. ^^ Lipemia is often produced by severe or repeated hemorrhages, with a great increase in the phospholipins of the plasma and corpuscles. ^^^ Changes in the Extravasated Blood. — These begin soon after its escape. In most situations sufficient fibrin ferment is formed to " Oliva, Folia clinica, 1912 (3), 213. "Richet et al, Compt. Rend. Acad. Sci., 1918 (166), 587. 2), 29-4; Marcliand, in Krehl and Marchand's Handb. allg. Pathol., 1912, II (1), 307. Also later references in this text. ^^ See Ilurwitz and Lucas, Arch. Int. Med., 191G (17), 543; Minot et al, ibid., 191(5 (17), 101. '"See Lee and Robertson, Jour. Med. Res., 1916 (3;j), 323; Hess, Proc. See. Exp. Biol. Med., 1917 (14), 96. ^0 Ainer. .Jour. Dis. Children, 1914 (8), 386. " Whipple, Arch. Int. Med., 1913 (12), 037. HEMOPHILIA 299 although the left heart is frequently enhirf^ed, there is usually no in- creased blood pressure demonstrable; furthermore, conditions of high blood pressure, such as nephritis, do not cause hemophilia. The theory of ''hydremic plethora" is also without good foundation. The most natural place to look for the fundamental fault is in the blood, but speaking strongly against this is the occasional occurrence of "local" hemophilia; e. g., in this type of hemophilia wounds of the skin may behave as in normal individuals, whereas any injury of the mucous surfaces is followed by pronounced hemophilic bleeding ;^^ in other cases the hemophilic bleeding is limited to regions above the shoulders; in still another class the bleeding is always from one organ, e. g., the kidneys. Nevertheless, a great deal of investigation of the blood has been done, at first chiefl}' with negative results. There are no characteristic changes in the cellular elements of the blood, beyond the changes common to all secondary anemias, excepting possibly a decrease in the number of w^hite corpuscles with a relative increase in the number of lymphocytes as observed by Sahh; the platelet count is normal. No constant alterations in the salts of the blood have been found, calcium usually being normal ;^^ and the proportion of water, fibrinogen and the several other proteins, the alkalinity, and the osmotic pressure of the serum all seem to be normal. Metabo- Hsm is unchanged, except possiblj^ for calcium loss in some cases.** Since bleeding is normally stopped principally by coagulation, a de- ficiency in fibrin or its antecedents might be expected, but most studies on this point have shown a normal amount of fibrinogen in the blood of hemophilics, the frequent formation of large tumors of clotted blood at the bleeding points supporting the experimental evidence that the blood contains an abundance of fibrinogen. The "bleeding time" following punctures in the skin is not excessive. As to the rate of clotting, Sahli," who avoided a number of errors made in earlier investigations, found that in the intervals between the at- tacks of hemorrhage the rate of the coagulation of the blood is con- stantly much slower than normal. During an attack of bleecUng the coagulation time approaches the normal; indeed, it may be faster than normal; apparently this is due to a reaction on the part of the organism to the loss of blood. If blood is collected directly from the site of bleeding the coagulation time is very rapid, because of the ac- cumulation of fibrin ferment from the clot over which the escaping blood flow^s. Yet in spite of the normal coagulability of the blood and the rapid clotting after the blood escapes from the vessel, bleeding continues for long periods before it can be stopped. As he found no general change in the properties of the blood to account for the bleed- " Abderhalden, Ziegler's Beitr., 1904 (35), 213. " Ivlinger and Berg, Zeit. klin. Med., 1918 (85), 335, 406. " Kahn, Amer. Jour. Dis. Children, 1916 (11), 103; Laws and Cowie, ibid. 1917 (13), 236; Hess, Bull. Johns Hopkins Hosp., 1916 (26), 372. 300 DISTURBANCES OF CIRCULATION ing, and as local influences seem to be important in hemophilia, Sahli advanced the plausible hypothesis that the cause of hemophilia lies in hereditary deficiency of the fibrin-forming substances, thi'om- bokinase or zymoplastic substance (see "Thrombosis"), in the vessel- walls, so that when the vessels are injured there is no local production of fibrin such as occurs normally. Local hemophilia may be explained readily as a local deficiency in fibrinoplastic material. In general hemophilia even the leucocytes may exhibit the same defect, in which case clotting of the blood is diminished even outside the tissues. This hypothesis seems to be in excellent agreement with many of the facts now known, but there yet remains to be demonstrated such a lack ol" fibrin-forming elements in the vessel-walls and other tissues of a hemo- philic subject, and a single autopsy of a hemolytic subject gave, on the contrary, a very active thromboplastic extract from the vessels (Gressot).^^ The tissues of one case studied by Lowenburg and Rubenstone,*^ however, showed in the liver and thyroid a decreased capacity to accelerate coagulation. With the improved methods of study of the factors in coagulation of blood introduced by Howell, it has been found by him and cor- roborated by others^'^ that in hemopliilia there is constantly a defi- ciency in prothrombin, the other factors being practically normal in amount, and as in other hemorrhagic conditions there is no equal alteration in the prothrombin, they look upon this change as an essential characteristic of hemopliilia. Fonio, and jXIinot and Lee, however, find that the blood platelets of hemophihcs are remarkably ineffective in causing coagulation of either normal or hemophilic plasma, although normal platelets cause normal coagulation of hemo- philic plasma, and therefore conclude that there is some deficient activity on the part of the platelets in spite of their occurrence in nor- mal numbers in hemopliilia. The significance of the platelets is shown especially clearly by the observation of Ledingham and Bedson*^ that antiplatelet serum will produce a purpuric condition when in- jected into animals of the species furnishing the platelets, although no similar effect is produced by antileucocj'te or antierythrocyto serum, Hess^^ states that there may be an hereditary purjnira, sometimes occurring in the females of hemophilic families, difl'ering from hemophilia in a deficiency in the number of platelets, hemor- rhages following local congestions or puncture wounds and exhibiting an increase in the "bleeding time." "Zeit. klin. Med., 1912 (76), 194. Since corroborated bv Minot and Lee." "Jour. A.mer. Med. Assoc, 1918 (71), 1196. *' HowoU, .Vroh. Int. Med.; 1914 (VA), Tti; llurwitz and Lucas, ibid., 1916 (17), 543; Minot and Lee, ibi per cent. Other conditions (with anemia), average 0.079 per cent. 0.362 per cent. Health}' organs 0.084 per cent. 0.090 per cent. Iron is also found in the hemolymph glands, sometimes more abundantly than in the spleen (Warthin).*^ Extensive studies on the protein metabolism of pernicious anemia by Rosen- quist'" showed that there is a considerable destruction of tissue proteins, as indi- cated by nitrogen loss, but that at times nitrogen may be stored up for brief periods. At times there may also be an excessive elimination of purine nitrogen, indicating destruction of nuclear elements. Calorimetric studies show the metab- olism to be slightly above normal.^' In anemia due to Bothriocephalus quite similar changes were observed. Hunter^- describes the condition of the urine in pernicious anemia, particularly with reference to the elimination of much ''pathological urobilin,"^' which seems to be produced by intracellular destruction of hemoglobin. Iron may also appear in the urine in increased quantities." There is an increased elimination of oxy- proteic acid nitrogen, other urinar}' nitrogen constituents being normal (Kahn and Barsky)."^ No constant metabolic changes follow splenectomy in pernicious anemia.'^ Summary.^*^ — Putting together the above findings, we see tliat in pernicious anemia we have everj^ evidence that excessive hemolysis is taking place, and the fact that continued poisoning by toluylendia- mine^^ and other hemolytic poisons, such as that of Bothriocephalus ,^^ may give rise to a condition resembling pernicious anemia very closely, indicates strongly that hemolytic poisons are the cause of pernicious anemia. Histological studies show the same thing, and, as Warthin^^ says: "The hemolj^sis of pernicious anemia does not differ in kind from that occurring normally or in certain diseased conditions; the difference is one of degree only." The hemolysis seems to go on chiefly inside of phagocytic cells instead of in the blood, probably because the phagoc3'tes pick up the corpuscles as soon as they have been injured 8« Lancet, 1903 (i), 283; similar results obtained bv Evffel, Jour. Path, and Bact., 1910 (14), 411. 89 .\mer. Jour. Med. Sci., 1902 (124), 674. 90 Zeit. klin. Med., 1903 (49), 193 (literature). See also Minot, Bull. Johns Hopkias Hosp., 1914 (25), 338. 91 Mever and DuBois, Arch. Int. Med., 1916 (17), ^65; Grafe, Peut. Arch, klin. Med., 1915 (118), 148. See also Tompkias et al., Arch. Int. Med.. 1919 (23), 441. 92 British Med. Jour., 1890 (ii), 1 and 81. 93 See also Mott, Lancet, 1890 (1), 287; and Syllaba, Abst. in FoUa Hematol., 1904 (1), 283. 9^ Keunerknecht, Virchow's Arch., 1911 (205), 89. Not confirmed by Quecken- stedt. Zeit. klin. Med., 1913 (79\ 49; bibliography. 95 Denis, Arch. Int. Med., 1917 (20), 79. 9^ Review on etiology of pernicious anemia given bv Vogel, Jour. Amer. Med. Assoc, 1916 (66), 1012. 9" Syllaba," Hunter** (loc. cit.). 9* In horses a condition resembling pernicious anemia seems to be produced by a toxic product of the larva; of a fly. Oestrus cqui, which is found in the walls of the stomach of anemic horses (Seyderhelm, Arch. exp. Path. u. Pharm., 1914 (76), 149). 308 DISTURBANCES OF CIRCULATION by the hemolytic poisons. In some instances cholesterol administra- tion improves the anemia, which suggests that the poison attacks the lipoids of the corpuscles,^' as so many hemolytic agents do. Both- riocephalus anemia, which so closely resembles the "pernicious" form, seems to be caused by a hemolytic lipoid, ^ presumably a cholesterol ester of oleic acid, and there is a growing tendency to associate hemoly- tic lipins with the etiology of pernicious anemia.^ However, although in hemolytic anemias there is an increased amount of unsaturated lipins in the blood^ Medak* did not find the isolated lipoids to be particularly hemolytic.^ (See Hemolysis, Chapter ix.) The origin and the nature of the specific hypothetical poisons have been sought in vain. Some authors have referred them to infections of unknown nature, occurring perhaps in the mouth and gastrointestinal tract ( Hunter), ^^ or to hemolytic products of intestinal putrefaction,^ or to faulty metabohsm. For example, Iwao^ has found that tyramine (p-oxyphenyl-ethylamine) produces in guinea pigs a blood picture resembling pernicious anemia, and this amine ma 5^ be produced either in the intestines or during metabolism. Others, with perhaps the best of grounds, would ascribe pernicious anemia to a multiplicity of causes, which produce a protracted slight hemolysis that continues until the hematogenetic organs give out, their exhaustion being perhaps hastened by the influence of the toxic substances in the blood; hematogenesis then becomes insufficient to replace the lost corpuscles, and the picture of pernicious anemia is established.^ The relatively great proportion of the iron that is stored in the liver supports the view that the hemolysis takes place in portal territory, and many other facts point to the same conclusion, but it is not generally accepted that the spleen plays an essential role in causing pernicious anemia through excessive phagocytosis or production of hemolytic poisons.' »^ See Reicher, Bed. klin. Woch., 1908 (45), 1838. 1 Tallquist, Zeit. klin. Med., 1907 (61), 427; Arch. exp. Path. u. Pharm., 1907 (57), 367. ^ SeeLiidke and Fejes, Deut. Arch. klin. Med., 1913 (109), 433. 3 See King, Arch. Int. Med., 1914 (14), 145; Csonka, Jour. Biol. Cheni., 191S (33), 401. ■» Biochem. Zeit., 1914 (59), 419. « See McPhedran, Jour. Exp. Med., 1914 (18), 527. ^ See Kiilbs (Arch. exp. Path. u. Pharm., 1906 (55), 73), who found the in- testinal contents of patients with chronic intestinal disorders to contain hemo- lytic substances of undetermined character. IfcMiiolytic lijioids in the intestinal contents have been described by Bergor and Tsuchiga (Deut. Arcli. klin. Med., 1909 (96), 252) and Liidke and Fejes, loc. cit.; but this observation failed of con- firmation by Ewald (Deut. med. Woch., 1913 (39), 1293). Herter (Jour. Biol. Chem., 190i) (2), 1) suggested a relation between intestinal infection with B. acrocjencR capsulatiits, wliich produces hemolytic substances, and pernicious anemia. ^ Biochem. Zeit., 1914 (59), 436. '^ Sec also limiting, .Johns Hopkins llosp. Bull., 1905 (16), 222; Pai)penheim, Folia Serologica, 1910 (10), 217. «See Hirschlield, Zeit. klin. Med., 1919 (87), 165. DISEASES OF THE BLOOD 30'.) Leukemia In Icukoinia the chciiiical changes in the red corpuscles take a less prominent position, resenil)ling cither those of a secondary anemia or chlorosis, while the enormous numl)er of leucocytes is the prominent feature and causes marked alterations in the composition of the l)lood. Large quantities of nucleoprotcins and also of the intracellular enzymes are introduced into the blood by the excessive leucocytes. As the leucocytes are constantly breaking down, more or less of the products of their decomposition are present in the blood and appear in the urine. Because of the relatively slight metabolic activity of the lymphocytes tiio various chemical alterations are all less marked in Ijmiphatic than in myelogenous leukemia.^" There is a notable reduction in antibody production in leukemia, ^^ presumably because of the changes in the bone marrow; it is said that typhoid infection in leukemics may fail to result in agglutinin formation. Chemistry of the Blood.— Co n.sidering the quantitative alterations in the con- stituents of the blood, we find the specific gravity lowered, but not so much as itjwould be in a simple anemia with equally low hemoglobin, for the loss of hemo- globin is partly compensated bj" the increase in leucocytes and their products. Fibrinogen is usually increased in myelogenous leukemia.'- The serum shows but slight change in specific gravity, a slight decrease in proteins'^ being com- pensated by an increase in the NaCl. The freezing-point of the blood is lowered (Cohn^*), which is probably due to the increase in crystalloidal products of cel- lular decomposition. Erben" found that in lymphatic leukemia the serum contains less cholesterol than normal, although the fat content may be rather high. Calcium is frequently found increased, probably because of destruction of the bone tissue. In the red corpuscles the proportion of iron, protein and potassium is decreased as is also that of the cholesterol, that of the lecithin and water being somewhat increased. The total amount of potassium and iron in the blood is decreased, but the P2O5 in the ash is increased because of the large amount of nucleoprotein in the blood. A number of the earlier writers describe a decreased alkalescence which probably is due to the deficiency in the fLxed alkali of the pro- teins. There is an increased excretion of iron in the urine and feces.'"' The poor coagulation 0/ leukemic blood has long been known, but the reason for it has not yet been ascertained. Some investigators have reported a deficiency in fibrin, while others have found it increased. More recent reports, however, indicate that there is no marked change in either the amount of fibrinogen or of the fibrin-ferments. Erben'- found a normal amount of fibrin in the blood in lym- phatic leukemia; and in three cases of myelogenous and one of lymphatic leukemia, Pfeiffer'" found the amount of fibrinogen nearly normal. This is quite remarkable in view of the fact that in ordinary forms of leucocytosis both the amount of fibrin- ogen and the rapidity of clotting are increased. It is, therefore, extremely difficult 1° Stern and Eppenstein have observed that the striking proteolytic power of the leucocytes from the blood in myelogenous leukemia is not shown by the leucocytes in lymphatic leukemia (Sitz. d. Schles. Ges. f. vaterland. Kultur, June 29, 1906). " Rotky, Zent. inn. Med., 1914 (35), 953. 12 Erben, Zeit. klin. Med., 1908 (66), 278; full details on composition of the blood in leukemia. " Little change was found in the protein content of the serum by Heudorfer.. Zeit. klin. Med., 1913 (79), 103. "Mitteil. aus dem Grenzgeb. Med. u. Chir., 1906 (15), H. 1. 1* Zeit. klin. Med., 1900 (40), 282. '«Kennerknecht, \ irchow's Arch., 1911 (205), 89. " Cent. f. inn. Med., 1904 (25), 809. 310 DISTURBANCES OF CIRCULATION to understand the poor coagulability of leukemic blood, but study of the factors of coagulation by modern methods may clear this up, for in one case so[studied VSTiip- ple'^ found an increase in antithrombin. Decomposition Products. — Of particular interest is the finding" in the blood of decomposition products of the leucocytes, which are probably produced by autolysis of the leucocytes. (See Leucocytic Enzymes, Chapter iii.) Normal leucocytes are rich in autolytic enzymes, which under ordinary circumstances seem^ to' be held in check by the antienzymes of the blood. In leukemia this antienzyme action seems to be insufficient to prevent leucocytic autolysis, for even in freshly drawn blood proteoses for at least non-coagulable proteins) may be present.'' According to Erben, this is true only of myelogenous leukemia, the fresh blood in lymphatic leukemia not only being free from non-coagulable protein, but further- more this product of proteolysis does not soon develop when the blood is kept aseptically at incubator temperature. This is, of course, what one wouldi expect in^view of the well-known enzyme-richness of the polymorphonuclear leucocj'tes neutrophile cells seem to be the chief source of proteoses, since their granules soon and the scarcity of proteolytic enzymes in lymphocytes. Erben states that the disappear in blood that is undergoing autolysis, whereas the eosinophiles preserve their granules well, and true proteoses are not present in blood rich in mast cells (i. e., myeloma). The marrow, spleen and lymph glands are found strongly pro- teolytic (according to the plate method), in myelogenous leukemia, but in lym- phatic leukemia and pseudoleukemia, only the marrow shows a slight acti^Hty.-<' Schumm^i found in the blood in a case of myelogenous leukemia several varieties of proteoses, most abundant being the so-called deutero-albumose; in another he also found peptone, leucine, and tyrosine. In addition he demonstrated the auto- lytic nature of the changes that occur in leukemic blood after death (see also "Autolysis in Leukemia," Chap. iii). Most observers have failed to find alhumose in the urine in leukemia. Because of the involvement of the bone marrow, small amounts of Bence-Jones protein, as well as Morner's bodj"-, may be found in the urine. -^ Kolisch and Burian^^ not only found nucleoprotein constantly, and albu- mose frequently, but in one case of lymphatic leukemia they found histon in the urine, which undoubtedly came from nucleoprotein decomposition. The oxidase reaction is conspicuous in certain of the cells of myeloid leukemia, especially the large, non-granular cells of acute leukemia,-* but it is not known that these oxidases influence the chemistry of the disease. In spite of the richness of leucocytes in lipases the serum shows no increased lipolytic activity.-^ Protein Metabolism. — Stejskal and Erben-*^ studied the metabolism of a case of myelogenous and of a case of lymphatic leukemia, and found the nitrogen loss much greater in the myelogenous form, although food-absorption was better than in the lymphatic ; they consider that protein-destroying forces are at work in myelo- genous leukemia, similar to those of cancer cachexia, so that nitrogenous equi- librium cannot be attained. As the most characteristic products of decomposition of nucleoproteins are the purine bases, one would also expect to find them present in leukemia, and early writers mention the finding of purine bases and uric acid in the blood and spleen. The urinary findings in this respect have been very variable. Ebstein" observed the complication of leukemia with gout which he considered a coincidence, and also noted uric-acid concretions in the urinary passages in four cases. Numerous other authors have described increased uric-acid elimination, while some have observed increase in the purine bases, either with or without uric-acid increase. Magnus-Levy^^ observed a particularly large uric-acid output in acute leukemias, i« Arch. Int. Med., 1913 (12), 637. 19 For literature see Erben, Zeit. f. Ileilk. (Int. Mod. Abt.), 1903 (24), 70. 20 Jochmann and Zicglor, Mlinch. med. Woch., 190G (53), 2093. 2' Ilofmeister's Hcitr., 1903 (4), 442; Dcut. med. Woch., 1905 (31), 183. " Boggsand (iuthrie, Hull. .lohns IIoi)kins Hosp., 1913 (24), 3GS. 23 Zeit. kliii. Med., l.Si)G (29), 374 (literature on albuminuria in leukemia). " Dunn, Ouart. Jour. Med., 1913 (ti), 293. " Caro, Zeit. klin. Med., 1913 (78), 28(j. "Zeit. f. klin. Med., 1900 (39), 151. " For lit(M-ature see resume 1)V Walz in Cent. f. Patiiol., 1901 (12), 985. 2sVirchow's Arch., 189S (152), 107. LEUKEMIA 'A 1 1 but also fouiul that the relation be. ween the number of leii(;ocytc.s and the utif acid is extreincly variaI)Io. Soiuctiincs the iiitroKen loss is very great — even as much as 20 ^m. per day — and, corresponding with the destruction of nucleoj)roteins and the resulting uric-acid formation, phosphoric-acid excretion is often greatly increased — even up to 15 gm. ])er day. On the other hand, the results obtained by many other writers have been in every respect extremely varia- ble; some have found no increase in uric acid, some even report a decrease; likewise the P2O;, has been found even less than normal. For example, in a carefully studied case of lymphatic leukemia, Henderson and Edwards'-"-' found during six months no excessive excretion of uric acid or phosphoric acid. Zalesky and Erben found likewise no coasidcrable increase in the uric acid in lymphatic leukemia, but in myelogenous leukemia the uric acid was much increased; on the other hand, the amount of elimination of purine bases was reversed in the two forms, and creatin was decreased in both. Lipstein'° found no excessive elimination of amino-acids even in myelogenous leukemia. An increase in calcium is quite constantly ob- served, and attributed to the bone destruction-^ occurring in this disease. Undoubtedly these variations in results depend upon the known fluctuations in the course of the pathological processes of leukemia; the number of leucocytes, the size of the lymphatic organs, and the general condition of the patient all vary greatly from time to time, often with remarkable rapidity, and the excretion of products of metabolic activity must vary likewise. It can hardly be questioned that the enormous increase in the amount of lymphoid tissue in the body and blood must give rise to a greatly increased nuclein catabolism, with consequent appearance of its products (uric acid, purine bases, and phosphoric acid) in the urine. This seems to be well demonstrated by the increased elimination of uric acid and purine bases, together with a general increase in the nitrogen output that has been frequently observed following the therapeutic use of a:-rays in leukemia, which is attributed to the increased autolysis that x-rays are known to produce. Radium has a similar effect, increasing enormously the urinary total nitrogen, urea, ammonia, less markedly the uric acid, but especially the phosphates.^' According to Rosenstern^'- the x-rays affect chiefly the leucogenic tissues rather than the adult leucocytes. Lipstein also found an excessive elimination of amino-acids, in the urine of leukemic patients treated by .c-rays.^^ According to Curschmann and Gaupp,^* the blood of leukemic patients who have been exposed to x-rays contains a specific leucocytotoxin, which may be produced by a process of autoimmunization against the leucocytic substance set free by the disintegrated leucocytes. Capps and Smith^^ have obtained similar results. A'-rays seem not to alter the total metabolism appreciably.''^ Charcot's crystals (also called Charcot-Ley den and Charcot-Neumann crystals) represent a peculiar and striking product of nuclear destruction that has fre- quently^ been found associated with leukemia. These crystals were first observed by Robin3 7(1853) in leukemic tissues, but have been named after Charcot, who, with Robin, described their properties. They were described by Charcot as color- less, refractile, elongated octahedra; insoluble in alcohol, ether, and glycerol; soluble in hot water, acids, and alkalies; size variable, from 0.016 by 0.005 mm. up. These crystals have been found not only in the tissues and blood of cadavers, but 23 Amer. Jour, of Physiol., 190.3 (9), 417. 30 Hofmeister's Beitr., 1905 (7), 527. 31 Knudsonand Erdos, Boston Med. Surg. Jour., 1917 (176), 503. 32 Miinch. med. Woch., 1906 (53), 1063. " Literature on effects of x-rays in leukemia, see Arneth, Berl. klin. Woch., 1905 (42), 1204; Musser and Edsall, Univ. of Penn. Med. Bull., 1905 (18), 174; Rosenberger, Miinch. med. Woch., 1906 (53), 209; Williams, Biochem. Jour., 1906 (1), 249; Lessen and Moraw^tz, Deut. Arch. klin. Med., 1905 (83), 288; Koniger, Deut. Arch. klin. Med., 1906 (87), 31. 34 Miinch. med. Woch., 1905 (52), 2409. 36 Jour. Exp. Med., 1907 (9), 51; see also Klieneberger u. Zoeppritz, Miinch. med. Woch., 1906 (53), No. 18; Milchner u. Wolff, Berl. klin. Woch., 1906 (43), No. 23. 36 Arch. Int. Med., 1917 (19), 890. 3'^ Literature given by v. Leyden, Festschrift fiir Salkowski, Berlin, 1904, p. 1. 312 DISTURBANCES OF CIRCULATIOX also occasionally in the freshly drawn blood of leukemics. Poehl^s believes thorn to be the same as Bottcher's spermin crystals, and derived from decomposed nucleins. Schreiner considers that these spermin crystals are phosphoric acid salts of spermin (C2H5N), or, as Majert and Schmidt give it, C^HioXo, with the structure UN NH, thus being similar to, although not identiral with, piperazin. The entire question of the composition of spermin is still un- settled, ^^ however; and it is probable,' furthermore, that the crystals found in leukemia are not identical with the crystals observed in semen. Crystals that appear similar are also found in asthmatic sputum, empyema, and ascites fluid, bone-marrow, and tumors, and it has been suggested that they are derived from or related' to the oxyphile granules of the eosinophiles.*"^ This view implies an agreement with Gumprecht's opinion that the crystals seen in bone-marrow, asthmatic sputum, etc., are not spermin, but of protein nature. As can be seen, the nature and significance of Charcot's crystals are, at the pres- ent time, quite undetermined. Summary. — The chemical changes observed in leukemia depend upon the excessive quantity of leucocytes and lymphoid tissue, which undergo processes of disintegration at irregular intervals, with the result that the products of nucleoprotein destruction (uric acid, purine bases, and phosphoric acid) appear in the urine in increased quantities. As the large neutrophiles contain abundant autolytic enzymes, the prod- ucts of cell autolysis (proteoses, amino-acids, and products of nucleo- protein destruction) may appear at times in the urine and in the blood; because of the small amount of such enzymes in the lymphocytes, these changes are all much less marked in lymphatic leukemia. Charcot's crystals, which are perhaps derived from leucocytic nucleoproteins, may be found in the blood and tissues. The changes in the red cells are chiefly those of a secondary anemia, with occasionally some chlo- rotic features. The chemical findings of leukemia throw no light whatever upon the cause of the disease. Pseudoleukemia and Hodgkin's disease show only the evidences of a secondary anemia, without the chemical changes of either leukemia or pernicious anemia. There seems to have been little study of tlu' chemical processes of these diseases. Moraczewsld^i reports a study of metabohsm in one case, designated by him as pseudoleukemia and so quoted in subsequent literature, although the only leucocyte count mentioned in the original article was 171,000. This case showed some retention of nitrogen and calcium, with little change in the phosphorus and purine bases in the urine. HYPEREMIA Active Hyperemia This condition is associated with but few chemical changes. Cei- tain chemicals may cause active hyjxM-emia; some locally, as in the " Deut. med. Woch., 1895 (21), 475. ^* Literature, see HaTiimarsten, Amer. Transl., lUOl, p. -120. ■•» Literature, see Floderer, Wien. klin. Woch., 190:i (16), 276; Predtetschenskv, Zeit. klin. Med., 1<)0() (5«)), 29. *' Virchow's .Vrch., 1898 (151), 22. HYPEREMIA 313 case of irritants, such as alcohol, ether, ammonia, mustard, etc., which act neither by producing; a local vasodilator stimulus or Ijy paralyzing the vasoconstrictors. Other substances may produce active hypere- mia in special vascular areas, e. gr., cantharides causes active hyperemia in the kidneys, probably because of its elimination through these organs; pilocarpin causes active hyperemia in the salivary glands and skin, which is associated with increased function. In general, func- tional activity is associated with active hyperemia, and GaskelP^ has suggested that this is due to atonicity of the vascular muscle, the result of decreased alkalinity of the lymph flowing away from the active organ along the vessel-walls, it having been found that alkalies cause a tonic contraction and acids an atonic dilation of arterial muscle. '" Pathological active hyperemia is seldom of long enough duration to lead to any alterations in the tissues in which it occurs. The blood itself remains unchanged, except that the venous blood going from the part contains much less CO2 and more oxygen than usual, because more oxygen is brought to the tissues than can be used.'*^ Passive Hyperemia Passive hyperemia is almost equally unassociated with chemical changes, especially in its etiology, as it depends solely upon mechan- ical factors. Some chemical alterations result, however, from the changes in the stagnating blood, wliich may, if the obstruction to out- flow is severe, become of venous character in the capillaries of the con- gested area. Oxidation in the tissues is, therefore, impaired, and some fatty changes may result, e. g., in the center of congested liver lobules. Waste products accumulate, and possibly noxious products of meta- bolism are formed under lack of oxidation; either from these causes or solely from pressure and lack of nutrition there is a tendency to atrophy of the more specialized parenchymatous cells, and a proliferation of connective tissues. The atrophy of parenchyma is seen particularly in the liver, the increase of connective tissue in the spleen. ^^ In the Ividney neither atrophy nor stroma proliferations are pronounced, but the renal function is greatly impaired, since it depends upon the amount and quality of the blood brought to the kidney.""^ Whether connective-tissue proliferation in hyperemia depends upon overnutri- tion or upon irritation by waste-products, or is compensatory to par- enchymatous atrojihy, may be looked upon as still an open question. *^ Quoted by Lazarus-Barlow, "Manual of General Pathology," 1904, p. 126. "See discussion by WooUey, Jour. Amer. Med. Assoc, 1914 (63), 2279; and by Adler, Jour. Pharm., 1916 (8), 297. " Polycythemia (Vaquez-Osler disease) is accompanied by an increase in the total nitrogen of the blood, in proportion to the number of erythrocytes; but the nitrogen content of the individual erythrocyte is decreased, (v. Jaksch. Zent. inn. Med., 1912 (33), 397). *^ See Christian, Jour. Amer. Med. Assoc, 1905 (45), 1615. " See Rowntree and Geraghty, Arch. Int. Med., 1913 (11), 121; Nonnenbruch, Deut. Arch. klin. Med., 1913 (110), 162. 314 DISTURBANCES OF CIRCULATION Probably only the first two factors apply to the connective-tissue growth observed in the congested spleen, the clubbing of the fingers in congenital heart disease, or the thickening of the subcutaneous tissues in passive congestion of the lower extremities. Changes in the Blood. — Venous blood differs from arterial, not only in its increased load of CO2 and other waste products, but also in other ways. Venous blood generally clots less readily than arterial blood. *^ It contains more diffusible alkali because the CO2 combines with and tears away part of the bases that are held by the proteins, especially in the corpuscles, and so alkaline carbonates are formed and enter the plasma. Blood from the jugular vein on this account contains 20-25 per cent, more diffusible alkali than carotid blood (Hamburger).'*^ Since the bactericidal power of the blood has been shown to increase directly with the alkalinity, this property may be of importance in pathology. For example, the relative infrequency of infections in the right side of the heart may not depend solely upon lessened liability to endocardial damage, as generally considered, but is possibly due in part to the greater bactericidal power of venous blood. The same property probably explains the favorable results obtained in the treatment of local infections by artificially produced passive congestion. ^^ Too severe a stasis, with resultant edema, prob- ably favors local infection.^" V. Fodor^^ found that animals surviving infections show an in- creased blood alkalinity, whereas in those that died, the alkalinity was decreased; also, he found the resistance increased b}^ intravenous injections of alkalies. Other observers^- have noted a decrease in re- sistance after injecting acids into the blood. According to Calabrese, the alkalinity of the blood increases in immunization of animals against toxins, while Cantani found the injection of toxin followed by a decrease in alkalinity. Hamburger has shown that the bac- tericidal power of the blood may be increased in vitro bj^ shaking it with CO? as a result of the increased alkalinity, aided, perhaps, by some slight bactericidal power of the CO2 itself; he also found the blood more strongly bactericidal in venous congestion than normally, and the lymph from a congested part was also found more strongly bactericidal than normal lymph. Hamburger^^ has also found, how- ever, that chemotaxis is, if anything, slightly decreased under the in- fluence of CO2, as also is phagocytosis; large amounts of CO2 may ^'Vierordt (Arch. f. lleilk., 1S7S (19), 19:5) found coa{i;ul:ition faster in the l)lood in passive conj>;estion than in normal venous blood; but Ihisebrock (Zcit. f. Biol., 1882 (18), 41) found that if the stasis is protracted, the coagulation becomes delayed because of the excess of COa- •"* Virchow's Arch., 1S99 (150), 329; also. "Osmotischcr Druck und lonenlehre." " Hee Bier, "IIypcr;emie als lleilmittel," Leipsic, 1903. '0 Glasewald, C-ent. Chenz. Med. Chir., 1915 (IS), 507. f" Cent. f. Bakt., 1S90 (7), 753. •■^ Literature, see Ibunburf^er (loc. citJ'*), p. 281 '"•' Virchow's Arch., 1899 (150), 329. THROMBOSIS 315 reduce the phagocytic power for coal particles by 2.5-50 per cent. Hamburger's results as to the bactericidal power of human blood in venous stasis have been confirmed by Laqueur.'"* Schiller ascribes this not to increased alkalinity, but to disintegration of leucocytes with liberation of bactericidal substances. ^^ The blood in the veins and capillaries in passive congestion is gen- erally richer in corpuscles than normal, perhaps because of some loss of water,^*^ although this is not constant, applying particularly to more recent or more local processes; in long-continued stasis, as in congenital heart disease, the blood may be diluted." In the concen- trated blood of passive congestion the corpuscles may number six to eight minions per cubic millimeter, while the concentration of the solids of the serum may be at the same time reduced (Krehl). The viscosity of such blood is higher than that of normal blood. ^^ In acute stasis the proportion of serum proteins, especially the albumin, increases with the duration of the stasis; no changes occur in the non- protein constituents of the blood (Rowe).^^ THROMBOSIS The chemistry of thrombosis in most respects resolves itself into the chemistry of fibrin formation, a subject which is so extensively con- sidered in most treatises on physiological chemistry and physiology that it does not seem desirable to give here anything more than the essential principles involved in the clotting of the blood, as now under- stood, as an introduction to the consideration of the same process as it occurs under pathological conditions. In spite of innumerable in- vestigations, our knowledge of the actual participants and processes involved in the formation of fibrin is in a very unsatisfactory and fragmentary state. Some facts seem well established, however, and we have a general idea of the subject that may be applied with ad- vantage to the consideration of thrombosis. Fibrin Formation^o Several different substances seem to be concerned in the formation of fibrin, of which the first of importance is its antecedent, fibrinogen. Fibrinogen is a simple protein, related to the globulins, and differing chiefly in its ready coagula- bility, not only by fibrin ferment, but also by heat, salts, and other coagulating agencies. By itself, however, it shows no tendency to coagulate spontaneously. According to Goodpasture," fibrinogen is formed through the combined activity " Zeit. exp. Path. u. Therap., 1905 (1), 670. " Beitr. klin. Chir., 1913 (84), H. 1. 56 Grawitz, Deut. Arch. f. klin. Med., 1895 (54), 588. ^' See Krehl, " Pathologische Physiologic," 1904, p. 201. 58 Determann, Zeit. klin. Med., 1906 (59), H. 2-4. " Jour. Lab. Clin. Med., 1916 (1), 485. ^° For literature and full discussion see Hamniarsten's or Mathew's Physiolog- ical Chemistry; Morawitz, Ergebnisse der Physiol., Abt. 1, 1904 (4), 307, and Handbuch d. Biochem., 1908 II (2), 40; Leo Loeb, Biochem. Centr., 1907 (6), 829; Howell, Harvey Lectures, 1917. " Amer. Jour. Phvsiol., 1914 (33), 70. 316 DISTURBANCES OF CIRCULATION of the liver and intestines, although earlier writers have variously described its formation in the bone marrow, leucocytes, liver or intestines. The amount of fibrinogen present in the blood is actually quite small, the fibrin formed in nor- mal clotting being but 0.1 to 0.4 per cent, of the weight of the blood. Acted upon by the fibrin-ferment, it yields the characteristic insoluble protein fibrin, in crystalline form under favorable conditions, ^^ but we do not know definitely what changes the fibrinogen undergoes in this process. Fibrin resembles in its insolu- bility the proteins coagulated by heat, alcohol, etc., but when kept aseptically for some time, it becomes again dissolved; this process of fibrinolysis probably de- pends upon proteolytic enzymes, which fibrin, in common with other substances of similar physical nature, has the property of dragging out of solution and holding firmly. Undoubtedly entangled leucocytes are also an important factor in the fibrinolysis,^' which is greatly increased in phosphorus poisoning and when the liver is excluded from the circulation, a fact suggesting that the liver may form inhibiting substances. Theories of Fibrin Formation. — The great problem is the nature and the place and manner of origin of the fibrin-forming enzyme, generally called fibrin-jerment (also plasmase, thrombin and coagulin). The most fundamental theory of the origin and nature of fibrin-ferment is that of Alexander Schmidt, which may be briefly described as follows: The ferment, Schmidt believed, exists in the plasma in an inactive (prozyme or zymogen) form, which he called prothrombin Lpon disintegration of the leucocytes there is set free a substance, which, acting upon the prothrombin, converts it into the active thrombin; this activating agent Schmidt designated as the zymoplastic substance. With various modifications this stands to the present day as a basic theory. It having been shown that calcium facilitates the formation of fibrin, Pekel- haring advanced the idea that the prothrombin does not exist in the plasma, but is liberated from the leucocytes, and, combining with the calcium of the plasma, forms the thrombin. Morawitz considers three substances necessary for the formation of thrombin. (1) the prothrombin or throjnbogen, which he believes originates in the blood-plates; (2) the zymoplastic stibstance or thrombokinase, which is liberated from the leucocytes into the plasma; (3) calcium salts. Howell,''* however, explains coagulation as follows: Circulating blood normally contains all the necessary factors for fibrin formation, i. e., fibrinogen, prothrombin and calcium. But there is also present an inhibiting substance, antithrombin,^^ which prevents the calcium from activating the prothrombin into thrombin. In shed blood there appears a thromboplastin, derived from the platelets or the tissues, which neutralizes the antithrombin and thus permits thrombin to form. Rett- ger^'^ holds that the coagulation of the blood is not a true enzyme action at all, while Bordet and Delange*^ consider that thrombin is formed by the interaction of cytozyme from the platelets or tissue cells, and serozyme of the plasma. Mathews follows Woolridge and considers the clotting of the blood as essentially the crj^s- tallization of a phospholipin-protein compound, blood plasma, the stability of which compound is easily upset in many ways. The fibrin threads are essen- tially liquid crystals coming out of a saturated solution, the blood plasma, which is practically a dilute protoplasm. It will not serve our purpose, however, to go further into the hypotheses and disputes over these questions, which are detailed more fully in the literature previously cited, but it may be stated that mimerous American observers have found Howell's theory to fit well with both experimental and clinical observations on the variations in the coagulability of the blood. The question has been raised as to whether the leucocytes or platelets secrete their fibrin-forming constituent (be it thrombokinase or prothrombin is a matter ''^ See Howell, Amer. Jour. Physiol., 1914 (35), 143; Hekma, Internat. Zeit. physik. chem. Biol., 1915 (2), 279. "' Sec Morawitz, loc. cit.; also Rulot, Arch, internat. d. Physiol., 1904 (1), 152. 8* Amor. .Jour. Physiol, 1911 (29), 187. ** The antithrombin is formecl by the action of a phosphatid from the liver (heparin) upon a pro-antithrombin (Howell and Holt, Amer. Jour. Phvsiol., 191S (47), 328). «« Amer. Jour. Physiol., 1909 (24), 40(). «^ Ann. Inst. Pasteur., 1912 (20), ()57. See also Leo and Vincent, Arch. Int. Med., 1914 (13), 398. COAGULATIOX OF HIJX)!) 317 of minor importance to the pathologist) or Uberate it only after their disinte- gration. As far as pathological processes go, the latter seems to be the case, the disintegration ajjparently occurring whenever tiie leucocytes come in contact with a foreign body or with dead and injured tissues. The stroma of red corpuscles also contains thrombokinase."" Of the substances that may be isolated from tissues, ce-phalin is found especially active in producing thrombosis, and may be related to or identical with the thromboplastin.^" Tissue Coagulins. — Among the other i)oints that are of importance in i)atho- logical conditions is the fact that not only the leucocytes, but also tissue-cells, can liberate fibrin-forming substances {coaffidtiifi is the non-committal term ap- plied by Loeb). Howell considers that the effect of the tissue "'coagulins" is merely to neutralize the antithrombin of the blood, if such coagulins actually e.Kist; possibly there is thromboplastin in the tissues. These coagulating agents are present in tissue extracts and are liberated whenever the tissues are injured; muscle is rich in coagulin, as are also the liver and kidney, and, which is par- ticularly important, the blood-vessel wall (L. Loeb). Pieces of these tissues })laced in contact with fibrinogen solution will bring about prompt clotting. An- other important fact is that the coagulins, whether derived from leucocytes or from the tissues, have a certain degree of sjiecificity — that is, they act solely or most rapidly with fibrinogen of blood of the species from which they are obtained.'" In some instances this specificity is absolute, but more generally (particularly in the mammalia) it is only relative. Loeb also found that the amount of tissue coagulin was not decreased in organs altered by phosphorus poisoning, although during experimental autolysis the coagulins disappear. Wlien tissue coagulins and blood coagulins act together, the effect is greater than the sum of their inde- pendent actions, indicating the probability that they combine in some way to produce a particularly active coagulin. The blood coagulins are quite different from the tissue coagulins in many important respects, and the coagulins cannot be looked upon as a single substance of different origins. Blood-platelets. — It is still undetermined just what part the platelets play in coagulation. The well-known observation that in thrombosis the fibrin is often first formed about masses of platelets clinging to the wall of the vessel indicates that they participate in the process, and Bizzozero and others have maintained that the platelets and not the leucocytes are the source of the prothrombin. Numerous studies on the relation of the platelets to disease conditions have in- dicated a certain parallelism between their number and the tendency to coagula- tion observed in the various diseases (Welch). Howell believes the platelets to be the chief source of thromboplastin, which neutralizes the antithrombin of the blood and thus causes clotting. Wright and Minof^ find that a viscous meta- morphosis of the platelets is intimately associated with the early stages of coagu- lation. Bordet and Delange consider the platelets of more importance than the leucocytes in producing participants of the coagulating mechanism. The histo- logical evidence of the importance of the platelets in thrombus formation is con- clusive (see Zurhelle, Derewenko), and Cramer and Pringle"^ state that coagulation cannot occur without platelets. However, the blood of fishes, birds and reptiles clots although no platelets are found in these animals. Human thoracic lymph also is devoid of platelets yet it clots; but it may contain products of platelet disintegration to explain this clotting (Jordan)."'* Kemp'* concludes, from a thorough review of the subject, that the blood-platelets are usually normal or subnormal in number during acute infectious diseases, but increase rapidly if the disease terminates by crisis; iii pernicious anemia the number is always greatly diminished, although in secondary anemias they may sometimes be increased; in purpura hemorrhagica the number of plates is enormously diminished, which is «« Barratt, Jour. Path, and Bact., 1913 (17), 303. "Howell, Amer. Jour. Physiol., 1912 (31), 1; MacLean, ibid., 1916 (41), 250; 1917 (43), 586. '"Leo Loeb, Univ. of Penn. Med. Bull., 1904 (16), 382; Muraschew, Deut. Arch. klin. Med., 1904 (SO), 187. " Jour. Exp. Med., 1917 (26), 395. " Quart. Jour. Exper. Physiol, 1913 (6), 1. " Anat. Rec, 1918 (15), 37. "" Jour. AmQv. Med. Assoc, 1906 (46), 1022. 318 DISTURBANCES OF CIRCULATION perhaps related to the slowness of the clotting of the blood in this condition. Duke" states that when the platelet count falls below 10,000 per cubic mm. there is delayed coagulation and a tendency to purpura; with counts above 40,000 there is usually no hemorrhagic tendency."" If the platelet count is reduced arti- ficially (by benzene, diphtheria toxin) a similar purpuric tendency is observed. Poisons that in large doses reduce the platelet count, will increase it if in small doses. Calcium Salts. — The exact significance of calcium in fibrin formation is still unsettled. Blood from which the calcium has been precipitated will not coagu- late, and the addition of calcium salts will promptly cause it to do so. The vari- ous hypotheses advanced to explain the way in which calcium influences the clotting process are not in agreement. One hypothesis is that the calcium ions are necessary for the transformation of prothrombin into thrombin (Pekelharing, Hammarsten, Morawitz), the thrombin consisting of a compound of prothrombin, calcium salts, and thrombokinase. Howell considers that no kinase is necessary, the calcium activating the prothrombin whenever it is not inhibited by anti- thrombin. Modification of Coagulability. — If blood is drawn into a glass vessel well coated with oil or vaseline, through a cannula similarlj^ pro- tected, no coagulation will take place; but if any unoiled foreign sub- stance enters, even particles of dust, coagulation begins at once. The explanation is that the leucocj^tes do not liberate their coagulating substances until they have been injured by contact with some foreign body, and the experiment proves the importance of this action of the leucocytes, as well as explaining why the blood does not coagulate dur- ing life. The classical experiment of the ligation of a vein without injury to the endothelium, which permits the blood to remain stag- nant for a long period without clotting, depends upon the same fact, namely, that normal endothehum neither liberates coagulin itself nor injures the leucocytes so that they disintegrate. Loeb recalls the observation of Overton that lipoids are important constituents of the cell membranes, and suggests a similarity between the vessel lining and the oiled cannula, but analyses of aortic endothelium have shown a rather low lipin content (8.41-9.25 per cent.), although peritoneal endothehum has much more (13 to 15 per cent.).'''^ The suggestion that the vessel walls contain an anti-coaguUn could not be confirmed by Loeb. Since leucocytes are constantly undergoing disintegration in the blood and tissues under normal conditions, it might be asked why they do not cause clotting then and there. In explanation Loeb ad- vances his observation that the coagulins are destroyed during cell autolysis, and suggests that when leucocytes normally thsintcgrate, the coagulins are first destroyed by autolysis. It has also been shown that the cells and serum contain substances which inhibit or prevent coagulation, and it is possible that these play an important part under normal conditions in preventing coagulation by products of cell "Jour. Exp. Med., 1911 (14), 265; Arch. Int. Med., 1912 (10), 445; Jour. Amer. Med. Assoc, 1915 (65), 1600. ""Gram (Arch. Int. Med., 1920 (25), 325) states that platelet counts below 100,000 goncrallv accompany a tendency to bleed. He gives the normal figure as 200,000 to 500,000, but usually over 300,000. "Tait, (iuart. Jour. Exp. Physiol., 1915 (S), 391. COAGULATION OF BLOOD 319 disintegration, nnicli as other antienzymes are supposed to act in prevontins autodigcstion of living cellp. Coagulation of drawn blood may be retarded experimentally by re- moval of the calcium by precipitation as oxalate, fluoride, etc.; also by diminishing the oxygen and increasing the COi,, by addition of solu- tions of neutral salts in large amounts, by diluting greatly with water, or by keeping the blood cold. Bile salts retard coagulation markedly, by interfering with the conversion of fibrinogen into fibrin." Coagu- lation may be hastened by moderate heat, by whipping, exposure to air, by contact with much foreign matter, and by the addition of watery extracts from many different tissues and organs. Poisons that destroy the platelets reduce the coagulation (Duke). Of particular interest pathologically is the retardation of coagulation that follows injections of proteoses (the so-called "peptone" solutions) and also various other protein-containing solutions, such as organ extracts, bacterial toxins, snake venoms, eel serum, extract of leeches or of Uncinaria, impure nucleo-protein solutions, or solutions of various colloids. Most of these substances e. g., peptone, eel serum, cause reduction of coagulability when injected into animals, and are without effect on blood removed from the body. A few, however, prevent coagulation of drawn blood (snake venom, extract of leeches). When substances of the first class are injected in sufficient quantities, there occurs first a period of accelerated coagulation which may, particularly in the case of organ extracts, cause prompt death from intravascular clotting; if the animal survives, there follows a period of decrease or total inhibition of coagulability of the blood, both within the ves- sels and after removal from the body. The first period of increased coagulability undoubtedly depends upon the formation of a large amount of fibrin-ferment, but it has not yet been satisfactorily ex- plained how the inhibition of coagulation is produced. Apparently the fibrin-ferment formed at first is rapidly destroyed, but it is thought by some that it is converted into a substance that neutralizes the fibrin-ferment that may be formed later, or that a true anticoagulin is formed. It is also among the possibilities that all the available prothrombin or thrombokinase is used up during the first stage of acceleration. As before mentioned, the blood and tissues contain substances that inhibit coagulation, and it may be that these are secreted in excessive amounts, a view which is receiving much support from recent observations. According to Davis"" injection of tlirombin is followed quickly by an increase in the amount of antithrombin in the blood. It has been found that in animals deprived of the liver no coagulation-inhibiting substances are formed in the blood after in- jection of proteoses, hence Delezenne believes that the substances of this class act by causing a destruction of leucocytes, thus hberating a " Haessler and Stebbins, Jour. Exp. Med., 1919 (29), 445. "«Amer. Jour. Physiol., 1911 (29), IGO. 320 DISTURBANCES OF CIRCULATION . substance which increases coagulation and also another substance retarding coagulation; the first of these is destroyed by the liver, leaving the retarding substance to act unopposed.^* Leech extract (hirudin) prevents clotting by means of an antiferment action, combining with the thrombin. ^^ Snake venom, however, acts upon the thrombokinase (Morawitz). Coagulability of the Blood in Disease. — In disease the altera- tions in the coagulability of the blood depend upon much the same factors. In all conditions associated with suppuration and leucocyto- sis the amount of fibrinogen is increased. This is especially true of pneumonia.^° The fluidity of the blood in septicemia is probably dependent upon the appearance of the coagulation-inhibiting phase that follows the action of the products of cell destruction, including among them proteoses. In this connection should be mentioned the observation of Conradi,^^ who found that among the products of au- tolysis is a coagulation-inhibiting substance which is not destroyed by heat, diffuses readily, and in general behaves unlike the proteins. This or similar substances may well play a part in affecting coagulation in infectious diseases, and Whipple^^ has found a decreased coagula- bility in septicemia because of the presence of an excess of anti- thrombin. It may also be mentioned that animals soon acquire an immunity against proteoses, so that their inhibiting influence is no longer shown. This corresponds to the observation of Kanthack^^ that immune serum against venom neutralizes very effectively the anticoagulating principle of venom; an amount of antiserum alto- gether insufficient to neutralize the toxic properties of venom will neutralize its property of preventing clotting. The bacterial prod- ucts may also modify coagulation, and L. Loeb^"* has found that different organisms are unequally effective in this respect, Staphylo- coccus aureus being much more powerful in causing coagulation than any others tested ;^^ typhoid, diphtheria, tubercle, and xerosis bacilli and streptococci being without any apparent effect, while pyocyan- eus, prodigiosus, and colon bacilli occupy an intermediate position. Furthermore, after the organisms are killed bj- boiling, this effect is greatly reduced, showing that it does not depend merely upon the mechanical action of the bacteria, but probabi}' upon bacterial prod- ucts contained in the culture-media. ^* Tlic manner in which gelatin injections afl'ect the blooil coagulability is not yet understood (see Hoggs, Deut. Arch. klin. Med.. 1<)04 (79), 539); Moll (Wien. klin. Woch., 1903 (16), 1215) found an increase in fibrinogen. " Hirudin may contain antikinase (Mellanbv, Jour, of I'hvsiol., 1909 (38), 441). 8" Dochez, Jour. Exp. Med., 1912 (10), 093. " Hofmeister's Beitr., 1901 (1), 137. s'' Arch. Int. Med., 1912 (9), 305. '* Cited 1)V Lazarus-Barlow, p. 141. «< Jour. Aicd. Research, 1903 (10), 407. *^ Much (Biochem. Zeit., 1908 (14), 143) states that staphylococcus contains thrombokinase. COAGL'LATIOX OF HLOOI) 321 After phosphorus poisoning the blood may become non-coagula- ble, whicli Jacoby*^ ascribed to an absence of fibrinogen in the blood, because of a fibrinogen-destroying ferment in the liver. Doyon^^ has made a similar finding in chloroform necrosis of the liver, but he at- tributes especial importance to an excess of antithrombin liberated from the liver in these conditions. Whipple has also found a de- crease in fibrinogen with chloroform necrosis and cirrhosis of the liver. ^^ In other instances of decreased coagulability the fibrinogen is present, generally in normal amounts. After death the blood be- comes incoagulable because the fibrinogen is destroyed through a process similar to that of fibrinolysis;*^ this fibrinolysis may be com- plete as early as ten hours after death. The other proteins of the blood do not seem to be correspondingly attacked. Thrombokinase is also scanty in cadaver blood, but there seem to be no coagulation- inhibiting substances present. In anaphylactic shock the coagula- bility is reduced or abolished, associated with which is a leucopenia.^° Whipple^^ states that the antithrombin-prothrombin balance in the blood is in delicate equilibrium, but preserved by strong factors of safety. The prothrombin factor is rarely involved, most notably in melena neonatorum and aplastic anemia, and such conditions may be relieved by injecting normal blood, through the added prothrombin. The antithrombin factor is often excessive in hemorrhagic conditions, especially with hepatic injury, or it may be lowered and lead to throm- bosis from relatively slight injuries. Obviously the injection of nor- mal blood will harm rather than help patients with hemorrhage due to excessive antithrombin. Antithrombin is often found increased in diseases of the blood-forming organs, e. g., leukemia, possibly as a reaction to the products of disintegration of corpuscles; and hence hemorrhagic tendencies are noted in these diseases. In icterus the notable tendency to hemorrhage seems to depend upon the binding of the calcium of the blood. by the bile pigments,^^ and administration of calcium may bring the coagulation time back to normal with a cor- responding decrease in the hemorrhagic tendency. Pfeiffer^^ estimated the fibrin content of the- blood in disease, and found it increased in diseases with leucocytosis (pneumonia, rheuma- tism, erysipelas, scarlet fever), except leukemia, where it was normal; "Zeit. physiol. Chem., 1900 (30), 175; also Doyon et al, Compt. Rend. Soc. Biol., 1905 (58), 493. "Compt. Rend. Soc. Biol., 1905 (58), 704; Jour. phys. et path., 1912 (14), 229. «8 Bull. Johns Hopkins Hosp., 1913 (24), 207. 8' Morawitz, Hofmeister's Beitr., 1906 (8), 1. ^° The incoagulability of menstrual blood is ascribed to a lack of fibrin ferment by Bell (Jour. Path, and Bact., 1914 (18), 462) and to an excess of antithrombin by Dienst (Miinch. med. Woch., 1912 (51), 2799). "1 Arch. Int. Med., 1913 (12), 637. " Lee and Vincent, Arch. Int. Med., 1915 (16), 59. »3 Zeit. klin. Med., 1897 (33), 214; Cent. f. inn. Med., 1898 (19), 1. 21 322 DISTURBANCES OF CIRCULATION in diseases without leucocytosis (typhoid, malaria, nephritis), the fibrin was normal in amount. Stassano and Billon^^ have, further- more, shown that the amount of fibrin-ferment varies directlj^ with the number of leucocytes in the blood. Kollmann^^ found an increase in the fibrin of eclampsia, which Lewinski^^ could not substantiate. In experimental infections of anitnals Langstein and Mayer^" found a specific increase in pneumococcus sepsis, which undoubtedly bears an important relation both to the characteristic fibrinous nature of the alveolar exudate in pneumonia, and the striking amount of fibrin found in pneumococcus pleuritis, peritonitis, etc. Mathews^^ found an increase in the fibrin with all experimental suppurations. The coagulation time determined by different methods, in which different conditions for coagulation are presented, varies from 2 to 30 minutes; with most methods it is 5 to 8 minutes. ^^ In general, coagulability is not constantly if at all altered bj^ fever, cancer, dia- betes, slight secondary anemias, or many other diseases, and in nor- mal conditions it remains fairly constant. In infants the coagulation time is slightly shorter than in adults. The coagulation is hastened after considerable hemorrhages, in endocarditis, and perhaps in aneu- rism and thrombosis; and is commonly delayed in the acute exan- themata, in hemophilia, in purpura neonatorum, and occasionally in some other diseases.^ There is entire lack of agreement concerning the reputed acceleration of coagulation by oral administration of cal- cium salts, and retardation by citrates; and the supposed thrombo- plastic influence of gelatin cannot be shown consistently by direct observations. In jaundice, calcium salts probably have an effect, since here the cause of the deficient coagulation seems to be the fixation or precipitation of the blood calcium by the bile pigments. The bile salts also prevent the conversion of fibrinogen into fibrin.''^ It seems probable that the measurement of the time required for coagulation to take place in vitro does not exactly represent the tendency of the same blood to coagulate in the body of the person from whom it is obtained; for example, the injection of foreign serum has a notable effect in stop- ping hemorrhages, but the coagulation time of the recipient's blood is not correspondingly altered. Whipple's observation that with a low fibrinogen content the blood may coagulate in normal time, and yet the clots be too delicate to stop hemorrhage, explains at least part of the discrepancy; and of similar signifiance is the fact that with a very " Compt. Rend. See. Biol., 1903 (55), 511. »5Cent. f. Gyniik., 1897 (21), 341. »« Pfliiger's Arch., 1903 (100), (ill. " Hofnieister's Bcitr., 1903 (5), 09. »« Amer. Jour. Physiol., 1899 (3), 53. *" Full review and bibliography by Cohen, Arch. Int. Med., 1911 (8), 084 and 820. ' See Dochez (Jour. Exp. Med., 1912 (10), 093), who found some delay in coagulation in pneumonia. Corroborated.by Minot and Lee, Jour. Anier. Med. Assoc, 1917 (08), 645. THROMBOSIS 323 low platelet count the blood may coaguhite as rapidly as normal, but the clots do not shrink and become firm (Duke). Hence with a se- vere purpura hemorrhagica wc may have a normal clotting time. In other conditions with normal coagulability, hemorrhages may re- sult from excessive fibrinolysis which causes solution ol" the clot, espe- cially in hepatic diseases. ^ The Formation of Thrombi If we apply the facts brought out in the preceding discussion rela- tive to the factors in the coagulation of blood, to the manner and conditions under which thrombi are formed in the circulating blood, we find explanations for many of the features of thrombosis. Welch^ describes the steps in the formation of a thrombus after injury to the vessel-wall, as follows: First, there is an accumulation of blood- platelets adhering to the wall at the point of injury. Leucocytes, which may be present in small numbers at the beginning, rapidly in- crease in number, collecting at the margins of the platelet masses and between them. Not until the leucocytes have accumulated does the fibrin appear. As Welch remarks, these findings afford no conclusive evidence as to whether fibrin-ferment is formed from the leucocj'tes or from the platelets, but since the fibrin does not appear until after the leucocytes have accumulated, and also since small thrombi may consist solely of platelets without fibrin, it seems probable that the leucocytes must be looked upon as the chief source of the ferment. If the blood is made incoagulable by injection of hirudin, injury to the vessel-walls causes the formation of thrombi composed entirely of platelets (Schwalbe). Sometimes small clots may form without the apparent participation of either platelets or leucocytes. These purely fibrinous thrombi seem to start from injured endothelial cells, par- ticularly in inflammatory conditions, such as pneumonic lungs, and give the impression that the coagulin is derived from the endothelial cells. Zurhelle attributes by far the most important part to the platelets, an opinion shared by many, including Derewenko,'* who holds that the coagulation of' blood with entirelj^ occluded vessels is quite distinct from true thrombosis because of the lack of platelets in stagnant blood. ^ Clots formed in the absence of platelets do not shrink like proper thrombi (Duke). The process of clotting in the stoppage of hemorrhage offers some 2 See Goodpasture, Bull. Johns Hopkins Hosp., 1914 (25), 330. ^ Albutt's System, vol. 6, complete discussion of the general features of throm- bosis; also see Kiister, Ergeb. inn. Med., 1913 (12), 667; Zurhelle, Ziegler's Beitrage, 1910 (47), 539; Schwalbe, Ergebnisse Pathol., 1907 (XI (2) ) 901; Lubarsch, AUg. Pathol., Vol. 1, Wiesbaden. 1905. See Aschoff, Ziegler's Beitr., 1912 (52), 205, and Arch. Int. Med., 1913 (12), 503, concerning the mechanics of thrombus formation. * Ziegler's Beitr., 1910 (48), 123. ^ Not accepted by Schwalbe, loc. cit.^ 32-1 DISTURBANCES OF CIRCULATION differences from intravascular clotting, in that the coagulins of the tissue-cells also come into play. It is rather difficult to determine how much of the coagulation depends on these, and how much on the coagulins of the leucocytes, for the same conditions that favor libera- tion of tissue coagulins, i. e., much laceration and destruction of the tissue, also favor the disintegration of leucocytes by offering large ^reas of surface for contact. Loeb is of the opinion, however, that -of the two, the latter factor is the more important. It may be re- called that the joint action of tissue and blood coagulins is greater than the sum of their individual actions, which also must be an im- portant factor in causing clotting in bleeding wounds. As to the relative importance of stagnation and vessel injury in producing thrombosis, we know that total stasis in an uninjured vessel may not result in thrombosis, and, on the other hand, extensive in- jury or large calcified plaques in the intima of the aorta may also cause no thrombosis because of the rapidity of the blood flow; and, furthermore, clotting may occur even in intact vessels under the influ- ence of substances liberating fibrin-ferment in the blood; e. g., snake venoms, nucleoprotein injections, and possibly in disease. As the red corpuscles contain thromboplastic substances we may have thrombi formed when hemolytic agents are present in relatively stagnant blood, even without injury to the vessel-walls.^ Presumably the clot- ting does not occur when the stream is rapid, because any fibrin- ferment that may be liberated by injured leucocytes or endothelium is swept away before fibrin can become attached to the vessel-wall; or, according to Howell's hypothesis, because the current brings an excess of antithrombin to the point where the thromboplastin is being formed. Naturally, the combination of an injured vessel-wall, a slow current, and a high coagulability offer the most favorable conditions, and we owe to Welch the appreciation of the fact that in a large pro- rportion of all thrombi, even those caused by apparently purely me- rchanical agencies (e. g., cardiac incompetence), bacteria are present and probably determine the injury to the vessel-walls and the libera- tion of fibrin-ferment.^ We have previously referred to L. Loeb's . observations on the effect of bacteria in causing coagulation of the blood. Hyalin thrombi are frequently the cause of extensive degenerative lesions in the viscera, and although commonly formed of red corpuscles, they do not stain at all like normal corpuscles, presumably because a certain proportion of the hemoglobin has been altered or lost through hemolysis. Of particular interest is their reaction to Weigert's fibrin stain, by which they often, but not always, stain intensely, especially when hardened in Zenker's solution; a fact that has been the cause of much confusion in earlier studies. Flexner* first appreciated the nature of these thrombi as originating from agglutinated red corpuscles, although Hobs, 6 Dietrich Cent. f. Path. (Verhandl.), 1912 (23), 372. ^ Welch, Venous Thrombosis in Cardiac Disease, Trans. Assoc. Amer. Phys., 1900, vol. 15. 8 Jour. Med. Research, 1902 (8), 316. EMBOLISM 325 Ziegler, and others had earlier suggested that hyalin thrombi^were formed from red corpuscles. Boxnieyer' independently arrived at the same conclusion as Flexner, in studying hyalin thrombi as the cause of necrosis in the liver of animals infected with the hog-cholera bacillus. Flexner produced hyalin thrombi by injecting corpuscles agglutinated by ricin, or by injecting ricin itself, or hemolytic substances such as ether or foreign serum. As the thrombi become old, the cor- l)uscles lose their form and color and produce the typical hyalin appearance. iVarce'" proved conclusively the dependence of the thrombus formation upon agglutination, for he secured the same results, including the liver necrosis, by injecting specific agglutinating serums. He states that fibrin threads may oc- casionally be found at the periphery of the larger thrombi, but never in the smaller ones. It is extremely probable, from Flexner's observations, that in the thrombosis produced by injecting various toxic substances into the blood, the so-called "fibrin ferment thro7nbosis," the thrombi are merely agglutinative thrombi, devoid of fibrin; this is undoubtedly true for many of the thrombi observed after poisoning with the powerfully agglutinative snake venoms (see Chap. vi). Bacterial hemag- glutinins may also cause the formation of hyalin thrombi.'^ On the other hand, some, at least, of the hyalin capillary thrombi are undoubtedly composed of soft masses of fibrin which have not become fibrillar, although the successful staining by fibrin stain is not final proof of the fibrinous nature of a thrombus. The liver necrosis which follows ether injections in animals is caused by fibrinous thrombi which result from liberation of coagulins by the injured cells (L. Loeb ). Secondary Changes in Thrombi. — The changes that occur in thrombi after they have existed for some time are largely due either to ingrowth of new tissue or to calcification, the latter of which will be considered in a separate chapter The only other change of interest from the chemical standpoint is the central softening which may occur in any large thrombus, but is particularly often observed in the large globular thrombi'found in the heart. The center of the thrombus may be so completely softened that it resembles a sac of pus, the contents, according to Welch, consisting of necrotic fatty leucocytes, albuminous and fatty granules, blood-pigment and altered red corpuscles, and occasionally acicular cr^'stals of fatty acids. Undoubtedly this softening is related to the process of fibrinolysis previously described, and depends upon digestion of the fibrin by leucocytic enzymes,^^ but the fact that the central portion alone undergoes softening is of interest, suggesting that the antibodies for leucocytic proteases, which Opie^' found present in normal serum, prevent digestion at the surface of the clot. The same fact indicates that the tissue fibrinolysins^^ do not play an active part in softening clots. Embolism Emboli offer little of chemical interest, because of the purely me- chanical nature of their origin and of the effects they produce. ^^ An exception exists in the case of fat embolism, for the manner in which the fat is removed from the blood has invited considerable specula- tion.^^ Part of the fat is ehminated in the urine, ^^ but the problem of how it escapes from the glomerular capillaries is not satisfactorily explained; large emboh undoubtedly lead to rupture of the capillary 9 Jour. Med. Research, 1903 (9), 146. 1° Jour. Med. Research, 1904 (12), 329; ibid., 1906 (14), 541. 11 Pearce and Winne, Amer. Jour. Med. Sci., 1904 (128), 669. 12 Barker, Jour. Exp. Med., 1908 (19), 343. 13 Jour. Exper. Med., 1905 (7), 316. "Fee Fleisher and Loeb, Jour. Biol. Chem., 1915 (21), 477. 1* Fat embolism may follow poisoning with potassium chlorate (Winogradow, Virchow's Arch., 1907 (190), 92). i« Full discussion by Beneke, Ziegler's Beitr., 1897 (22), 343. 1' Discussed by Sakaguchi, (Biochem. Zeit., 1913 (48), 1) who finds a little fat in the normal urine. 326 DISTURBANCES OF CIRCULATION walls, and probably some fat also escapes through stomata or similar intercellular openings. Fat may also escape in the bile, and some is probably taken up by the tissue and endothelial cells by phagocytosis. Beneke found that the fat becomes partly emulsified by the mechanical action of the blood current, aided to a slight extent by saponification. The larger droplets after lodging -in the capillaries are surrounded by leucocytes, to which Beneke ascribes an active part in the removal of the fat as fine droplets by phagocytic action. We may well believe, however, that the lipase of the plasma is an important agent in disin- tegrating the emboli, although its action is limited because of the rel- atively small surface which the large drops offer for attack. After fat droplets have been taken into the cells, they presumably are util- ized in metabolism like normally acquired fat, as described previously. The amount of fat free in the blood in fat embolism may be sur- prisingly large. BisselP* found from 2 to 6.5 per cent, in the venous blood of several typical cases, although sometimes figures within normal limits (0.2 to 0.6 per cent.) were found. The higher quantities represent such a great amount of free fat in the blood, even without considering the quantity held in the capillaries, that it is scarcely possible for it all to have come from the fractured bones. Air embolism presents some features of interest from the chemi- cal standpoint, especially in those cases following sudden decrease in at- mospheric pressure in persons who have been exposed for some time to pressures considerably higher than that of the 'atmosphere (diver's palsy, caisson disease, etc.). This form of air embolism is due to the fact that fluids can dissolve much more gas at high pressures than at low pressures; consequently when the abnormall}^ great pressure to which divers, caisson workers, etc., are subjected is too suddenly re- duced to that of the atmosphere, the excessive gas that was absorbed during the period of high pressure by the blood and tissue fluids is released, and forms bubbles in the tissues and blood. The bubbles in the nervous tissues may cause paralyses of various sorts, or death; those in the blood may, if in suffi.cient amount, cause serious or fatal capillary obstruction. The bubbles consist chiefly of nitrogen, bo- cause the power of the blood to hold oxygen in combination is so great that not much of this gas becomes freed. ^^ The body fluids of normal persons contain about 675 c.c. of nitrogen, all told, but at 22 pounds pressure this is increased to 1350 c.c, while but about 50 c.c. of free oxygen would be present (Langlois). Carbon dioxitlc is so readily combined in the blood that none is free even at high pressure, al- though McWhorter-" reports that the gas collected from t he right side of the heart in a fatal case contained 20 per cent. COo antl 80 per cent. 18 Jour. Aincr. Med. Assoc, lOKi (G7), 1926. "• Tliis subject is fully discusscni by Leonard Hill in "Recent Advances in Physiolof^y and Biocheniistry;" London, lOOti. 2« Anicr. .lour. Mod. Sfi.,"l<)10 (LJ'.)), 'M'.i; Krdinan, //>((/., lOL} (lb")), fyiO. INFARCTION 327 nitrogen. Possibly some oxygen may also be released from solution during decompression.-^ At body temperature fats can dissolve five times as much nitrogen as serum or plasma,'^ which probably accounts for the severity of the changes in the nervous system with its rich lipoid content and delicate structure. Air embolism following obstet- rical operations or surgical operations about the neck and chest pre- sents chiefly mechanical features, ^^ and large quantities of air must be present to cause dangerous obstruction to circulation. ^^ Gas-bubbles may be produced in the blood soon after death by B. aerogenes cap- sulatus, but it is not probable that they are formed before death and cause air emboHsm. Infarction . The changes that occur in infarcted areas are of much interest in connection with the study of autolysis, for the absorption of the ne- crotic tissue of aseptic infarcts is purely a matter of autolj'sis. Ja- coby-^ found by ligating off a portion of a dog's liver, and keeping the dog alive for some time afterward, that in the infarcted tissues so produced leucine and tyrosine could be detected, just as they are found in liver tissue undergoing autolysis outside of the body. So, too, proteoses may appear in the urine when any considerable amount of tissue is cut off from its blood-supply. The processes of autolysis which occur in ordinary sterile infarcts are, however, extremely slow, and it is doubtful if enough of the products are ever in the blood or urine at any one time to be detected or to cause noticeable effects. For example, in an infarct of the kidney which was known to be al- most exactly fourteen weeks old, there still remained a layer of ne- crotic cortex one millimeter thick, quite unabsorbed during this time. If we examine such aseptic infarcts in various stages, we get the im- pression that the digestion is accomplished by leucocytes acting on the peripher}^ of the infarct, and not entering the dead area deeply, pre- sumably because of a lack of chemotactic substances in the dead cells. On the other hand, it seems probable that the tissue enzymes them- selves play an important part in the autolysis, for if we implant into animals pieces of tissue in which the enzymes have been destroyed by heating to boiling, it will be found that the cells and their nuclei re- main unaffected for many weeks; whereas if sterile unheated pieces of tissue in which the enzymes are still active are implanted, they lose their nuclear stain and begin to disintegrate relatively soon, without apparent participation by the leucocytes.-^ Ribbert-^ found that in 21 HUl and Greenwood, Proc. Royal Soc. (B), 1907 (79), 284. "Vernon, ibid., p. 366; Quincke, Arch. exp. Path. u. Pharm., 1910 (62), -164. -^ Review of literature by Wolff, Virchow's Archiv., 1903 (174), 454. -^ See Hare, Anier. Jour. Med. Sciences, 1902 (124), 843. " Zeit. phvsiol. Chem., 1900 (30), 149. 2« Wells, Jour. Med. Research, 1906 (15), 149. " Virchow's Arch., 1899 (155), 201. 328 DISTURBANCES OF CIRCULATION experimentally produced anemic infarcts in the kidneys of rabbits the nuclei retain their staining property well for nearly twenty-four hours, becoming then small and deeply stained, undergoing karj^orrhexis, and in large part disappearing from the convoluted tubules inside of forty-eight hours, although some nuclei may persist in the glomerules for three or more days. In human infarcts, Ribbert believes, the process goes on faster, for he has osberved here a loss of nuclei within twenty-four hours. These nuclear changes undoubtedly depend upon autolysis, but it is probable that the enzymes concerned reside in the cytoplasm rather than in the nucleus, for I have observed that cells of lymphoid type, with practically no cytoplasm, generally retain their nuclear stain much longer than cells with more cytoplasm; this is particularly noticeable in splenic infarcts, where the ]\Ialpighian corpuscles retain their staining affinities much longer than the pulp elements. Whether the destruction of the nuclei is accomplished by the ordinary intracellular proteases, or by special nucleoprotein- splitting enzymes (nuclease, ^^ etc.), remains to be determined. It is quite possible, however, that the first changes consist of a splitting of the nucleoproteins of the nucleus by the autolytic enzymes, liber- ating the nucleic acid, which gives the nuclei the characteristic intense staining with basic dyes (pycnosis) observed in areas of early anemic necrosis. The nucleic acid may then be further decomposed by the nuclease or similar enzymes. Taken all together, then, it would seem that the nuclear and cellular alterations that make up the character- istic picture of anemic necrosis are brought about by the intracellular enzymes — 'an autolytic process. The removal of the dead substance, however, seems to be accomplished rather by the invading leucocytes, through heterolysis. The relatively small part taken by the intracel- lular enzymes may possibly be due to the seeping through them of alka- line blood-plasma, for autolytic enzymes are not active in an alkaline medium; the leucocytic enzymes, however, act best in an alkaline medium. 2^ About the periphery of infarcts is usually observed more or less fat deposition (Fischler),^" particularly in the endothelial cells (Ribbert). This is not peculiar to infarcts, however, for Sata^' found a similar peripheral fatty metamorphosis common to all necrotic areas. The basis of this is possibly the persistence of the cell lipase, wliich syn- thesizes fatty acid and glycerol diffusing into the necrotic area with the plasma, unchecked by normal oxidative destruction of these substances. (See "Fatty Degeneration," Chap, xvi.) Hemorrhagic infarcts offer in addition to the changes conunon to anemic infarcts, the alterations occurring in' the blood-corpuscles. "» Sachs, Zoit physiol. Chcm., 1905' (4C), '337; Schittonhclni, ibid., 354.' ^^ More fully discussed by Wells,-" loc. cit., and under necrosis, Chap. xv. 30 Cent. f. Path., 1«)()'2 (13), 417. " Ziegler's lieitr., 1900 (28), 461. INFARCTION 329 Panski''^ found that after ligation of the splenic vein of dogs the red corpuscles begin to give up their hemoglobin in about three hours. After twelve hours fibrin formation begins in the tissues, the corpus- cles continue to give up hemoglobin and become cloudy in appearance. Later, iron-containing pigment is formed in the cells beneath the cap- sule, but in the deeper tissue even the iron normally present in the spleen tissue seems to disappear ;^^ this possibly depends upon the fact that pigment reacting for iron, hemosiderin, is formed only in living cells under the influence of oxygen, or it may be that acids formed during autolysis dissolve it. During autolysis in vitro, how- ever, Corper^* found no evidence of removal of iron from insoluble or coagulable compounds. The hemolysis is probably produced either by the action of autolytic products, which are notoriously hemolytic, or perhaps also by direct attack of tissue and blood proteases upon the corpuscles. Other retrogressive changes that may occur in infarcts, such as sep- tic softening and calcification, are not greatly different from the same processes occurring in other conditions, and will be considered with the discussion of these processes. '2 "Untersuchungen iiber den Pigmentgehalt der Stauungsmilz," Dorpat, 1890. " See also M. B. Schmidt, Cent. f. Path., 1908 (19), 416. ="• Jour. Exper. Med., 1912 (15), 429. CHAPTER XIV EDEMA' As the term edema indicates the excessive- accumulation of lymph (which may be either normal or modified in composition) in the cells, intercellular spaces, or serous cavities of the body, the problems of edema are inseparably connected with the consideration of the proc- esses of physiological formation and removal of lymph. For many years the study of these processes has been a favorite field of investi- gation by physiologists, and the great battle-place of the "vitalistic" and ''mechanistic" schools; and to this day the forces that determine the formation of lymph and its subsequent absorption have not been completely understood. By the application of the principles of phys- ical chemistry to the problem, however, great advances have recently been made, which seem to render our understanding of both Ijanph- formation and its pathological accumulation in the tissues nuicli clearer and more nearly accurate than they were before. We shall first consider, therefore the physiological formation of lymph, before taking up the subject of edema. Composition of Lymph. — Lymph consists of material derived from two chief sources. The greater part consists of fluid passing out of the capillaries into the tissue spaces; here it is modified by the addition of products of nietabolism derived from the tissue-cells, and by the subtraction of materials that the cells utilize in their metabolism. It is, therefore, essentially a modified blood plasma, and the modifications the plasma undergoes are so slight, that, under ordinary conditions, lymph shows on analysis no considerable differences from blood plasma, except a relative poverty in proteins, due chifly to the impermeability of the capillary walls for colloids. Its quantitative composition varies greatly^ depending upon the conditions under which it is collected, whether during activitja or rest, etc. The following tables of analyses have been collected by Hammarsten: 12 3 4 Water 939.9 934.8 957.6 955.4 Solids 60.1 65.2 42.4 44.6 Fibrin 0.5 0.6 0.4 2.2 Albumin 42.7 42.8 34.7 1 Fat, Cholesterol, Lecithin 3.8 9.2 [ 35.0 Extractive bodies 5.7 4.4 I Salts 7.3 S.2 7.2 7.5 1 and 2 are analyses of lymph from the thigh of a woman, 3 is from the contents of sac-like dilated vessels of the spermatic cord, 4 is lymph from the neck of a colt. ' A complete bibliography is given by Meltzor, .\merican Medicine, 1001 (S), 19 cl sctj.; also by Klemensiewicz, in Krelil ami Marchaml's ll;iiull)uch d. allg. Path., 1912, II (i), 311; Magnus, llandbuch d. Hiochem., 1908, 11 {2), 99; C.er- hartz, ihifl., j). 116. 330 FORMATION OF LYMPH 331 Chyle differs from lymph chiefly in the presence of large quantities of fat; during starvation the lymph and the chyle are of practically the same composition. Normal lymph contains much less fibrinogen than does the blood plasma, and hence coagulates slowly. Lipase and other enzymes have been found in the lymph, as in the plasma. The products of tissue metabolism added to the lymph by the cells may render it toxic (Asher and Barbera'). Under pathological condi- tions the lymph may be greatly altered, becoming poorer in solids under some conditions of edema, and becoming rich in proteins and blood-corpuscles under inflammatory conditions, until it partakes of the characteristics of an inflam- matory exudate (see analyses of transudates and exudates). An important fact to consider is, that of the entire water of the body but about one-tenth is in the blood. About two-thirds of the entire weight of the body is water, which is mostly in the cells and tissues, firmly bound by the colloids, only an unknown but smaller portion being as free movable fluid, and even here always associated with more or less colloid. A body weighing 60 kilos will, therefore, have 40 kilos of water, of which but about 4 kilos is blood. FORMATION OF LYMPH^ Filtration Theory. — The simplest possible conception of lymph formation is that it is merely the result of filtration of the liquid con- stituents of the blood through the capillary walls under the influence of the blood pressure. This "filtration theory" was supported origi- nally by Ludwig, and it was a prominent factor in the early appli- cations of mechanical principles to biological processes. In support of this theory were advanced the results of numerous experiments in which it was shown that increasing the blood pressure by means of ligating the veins, or by causing arterial dilata ion, resulted in an in- ■ crease of the lymph flowing out of the lymph-vessels of the part. Also, when the blood pressure is raised by epinephrine or by other means, a large proportion of the fluid leaves the blood vessels; con- versely, when the blood pressure is suddenly lowered by hemorrhage there is a rapid passage of fluid from the tissues into the blood. The experimental results were not always favorable to the theory, how- ever, particularly in the experiments in which blood pressure was raised by arterial dilatation; often the flow of lymph was Kttle in- creased, even when the arterial flow and pressure were greatly in- creased. iSlevertheless, the filtration theory held for many years, not only as an explanation of lymph formation, but also as an explanation of urinary secretion and of the secretion by other organs. It was only within a comparatively short time that it became clear that filtra- tion alone could not account for all the phenomena of secretion. For example, in many lower forms with undeveloped circulatory systems, and almost no blood pressure, secretion goes on vigorously; the pres- sure of glandular secretions may be much higher than the blood pressure within the capillaries; the activity of secretion is by no means 2 Zeit. f. Biol., 1898 (36), 154. ^ See review by Asher, Biochem. Centralblatt, 1905 (4), 1. 332 EDEMA in proportion to blood pressure, etc. If in glandular secretion, there- fore, fluids are removed from the blood and transferred into an ex- cretory duct through the action of some force other than that of the blood pressure, it is probable that lymph formation is equally inde- pendent of blood pressure. On this basis Heidenhain advanced his — Secretory theory of lymph formation, in which he suggested that lymph is the product of an active secretion by the endothelial cells of the capillaries, just as saliva is the product of the activity of the glandular cells. He showed that certain chemical substances may stimulate lymph flow, independent of blood pressure, just as pilocar- pine and other drugs may stimulate the secretion of saliva. These lymph-stimulating substances, which he named lymphagogues, fall into two distinct classes. One which includes such substances as peptone, leech extract, strawberry juice, extracts of crayfish, mussel or oysters, and numerous other tissue extracts, are characterized by causing the secretion of a lymph which is rich in proteins, even richer in proteins than the blood plasma; and, furthermore, there is no simultaneous increase in urinary secretion. Heidenhain considered that these sub- stances caused lymph secretion by stimulating the capillary endothe- hum in a specific manner; as they caused no appreciable rise in blood pressure the increased lymph secretion certainly could not be attrib- uted to filtration. This independence of the lymph flow of blood pressure is most conclusively shown by posttriortem lymph secretion; for example, Mendel and Hooker'* observed lymph flow for four hours after death, in a dog that had received an injection of peptone eight minutes before being killed.^ The second class of lymphagogues includes crystalloidal substances, such as sugar, urea, and salts. ^ Lymph secreted under the influence of these substances is poorer in protein than ordinary h'mph, and at the same time an increased urinary secretion is produced. With these crystalloidal lymphagogues the amount of effect is in inverse proportion to their molecular weight, which means that their effects depend upon the number of molecules in solution rather than upon their nature; in other words, the stimulation of Ij^mph by crystalloids is dependent upon the osmotic pressure of the crj'stalloids. Heiden- hain explained their action as follows: The crystalloids are secreted into the lymph-spaces by the action of the capillary endothelium, and there, owing to their raising osmotic pressure, cause a flowing of water out of the vessels. The difficulty here is to explain why the crystalloids while still in the vessels do not attract the fluids from * Amer. Jour, of Physiol., 1902 (7), 380. ' A fact not sufficiently taken into account is that blisters filled with serum, i. e., an inflaniinatory edema, may be produced in dead bodies bv liurns or scalds. (See Leers and Raysky, Virchow's, Arcli., 1909 (197), 324). * The action of many other substances has been investigated bv Vanagawa, Jour. Pharmacol., 191() (9), 75. FORMATION OF LYMI'II 333 the lymph-spaces into the blood, and so cause rather a lessened lymph secretion. While admitting that in pathological conditions {e. g., passive con- gestion) pressure and filtration 7najj play an important part, Heiden- hain considered that an active secretion by the endothelial colls is the chief factor in the normal formation of lymph. The means by which the cells perform this function was unknown; it was considered as an example of "vital activity," Heidenhain meaning by this term such chemical and physical forces of living cells as are unknown or not understood at the present time, rather than any metaphysical concep- tion of living matter, such as many vitalists assume. Other observers, corroborating Heidenhain's results for the most part, have modified, or amplified his theory. Asher and his collabo- rators, for example, ascribe the work done in causing lymph forma- tion to the cells of the various 'issues and organs, rather than to those of the capillar}'- wall. The increased flow of lymph from the salivary gland that occurs during its activity they consider due to the work of the gland cells, and its function the removal of products of metab- olism. The action of such a lymphagogue as peptone they ascribe to its stimulation of cellular activity, particularly in the liver, where it causes an increased formation of bile. Gies^ and Asher also ob- served that after an injection of crystalloidal lymphagogues, such as sugar, a prolonged flow of lymph occurred after the death of the animal, proving completely that such lymphagogic action is inde- pendent of blood pressure. Potocytosis. — In explanation of the process by which the cells, whether en- dothelial or tissue-cells, pass fluids through themselves from one place to another, Meltzer^ has made an interesting suggestion, as follows: Considering the prop- erty of endothelial cells to act as phagocytes, MacCallum* has shown that solid granules (e. g., coal pigment, carmin) are taken through the walls of the lymphat- ics by the phagocytic activity of their endothelial cells. Meltzer suggests that in a similar way the endothelial cells may transport through the vessel-walls not only solid particles, but also, by the same mechanism, substances in solution; and for this hypothetical process he suggests the name "potocytosis." There can be little question that cells do take up substances in solution, and sometimes this is done in an apparently selective manner; e. g., the taking up of bacterial toxins and vegetable poisons in the peritoneal cavity by the leucocytes. Presumably the mechanism of "potocytosis" is not different from that of phagocytosis, chemo- tactic forces determining the occurrence of the process. No experimental evi- dence has been advanced as yet for this very plausible hypothesis. Permeability of Capillaries. — In explanation of the variabihty in the amount and composition of the lymph. Starling^ has introduced the factor of altered permeability of the capillary walls, which pre- sumably depends upon the number and size of the pores. He found that normally the lymph coming from the lower extremities contains 7 Amer. Jour. Physiol., 1900 (3), p. xix; Zeit. f. Biol., 1900 (40), 207. « Johns Hopkins Hosp. Bull., 1903 (14), 1. 'Lancet, 1896 (i). May 9, et seq.; Schafer's Text-book of Physiology, vol. 1. 334 EDEMA only 2 per cent, to 3 per cent, of proteins, while lymph from the intes- tines contains 4 per cent, to 6 per cent., and lymph from the liver con- tains 6 per cent, to 8 per cent, of proteins; hence he considers that the liver capillaries are the most permeable, i. e., have the largest pores, so that more of the large colloid molecules can escape from them. The effect of lymphagogues of the first class (peptones, etc.) he attributes to their poisonous properties, and the consequent injury to, and altera- tions in, the capillary wall. The crystalloidal lymphagogues, he believes, act by first attracting fluids from the tissues into the blood with a resulting "hydremic plethora," which in turn leads to in- creased blood pressure and consequent filtration of a watery fluid out of the vessels. He considers, therefore, that the amount and quality of the lymph produced in any part are determined solely by two factors, the intracapillary blood pressure and the permeability of the capillary walls. In connection with this question of the permeability of the capil- lary walls, Meltzer suggests that the contractility and irritability of the endothelium may be a potent factor in determining the size of the pores in the capillary walls. When in a tonic condition, the endothe- lium is firmly contracted about the pores, keeping their size small; when the endothelial cells become relaxed by any cause, such as poi- sons, high blood pressure, poor nourishment, etc., the pores are en- larged, and increased escape of fluid results.^" It must be borne in mind, however, that most histologists do not now admit that capillary walls contain pores. M. H. Fischer holds that the endothelial cells undergo changes in consistency through changes in the affinity of the cell colloids for wa- ter; especially under the influence of acids the endothelium ma}' be- come much more fluid and of greater permeability. Adolf Oswald^" says that the normal capillary wall is somewhat permeable for the less viscous blood proteins (albumin and pseudoglobulin), and in in- flammation this permeability becomes increased so that the more vis- cous euglobulin and fibrinogen can pass through. Osmotic Pressure. — Still another possible factor in causing fluid to leave the vessels is osmotic pressure. Heidenhain refers to this cause the transudation produced by crystalloid lymphagogues, al- though in a rather unsatisfactory manner. As a result of the more recent studies of physical chemistry, and its application to biological processes, we have learned to appreciate the importance of osmotic pressure in cell activities (see Introductory Chapter), and in the question of lymph formation it occupies a particularly important place. We may consider it as follows: In the blood we have certain proportions of readily diffusible crystalloids and of non-diffusible "" Evidence of the contractility of capillary walls is discussed by Krogh, Jour. Physiol, 1919 (52), 457. "> Zeil. f. exp. Path., 1910 (8), 226. ^ FORMATION OF LYMPH 335 colloids. If no metabolic processes were going on in the tissues, wo should have the diffusible substances leaving the vessel-walls (leaving out, for the present, any question of secretory activity on the part of the endothelium) until an osmotic equilibrium is established in the tissues and in the blood. As a matter of fact, however, the blood pro- teins are not absolutely non-diffusible, but small (juantities do pass through the capillary walls, and so lymph under such a hypothetical condition would consist of a mixture of the same osmotic concentration as the blood plasma, with about the same proportion of crystalloids, but a smaller proportion of proteins; this, it will l)e noticed, is just about the composition of normal lymph. During life, however, the cells of the tissues are causing metabolic changes in these lymphatic constituents, and these changes consist chiefly in breaking down large molecules of proteins, carbohydrates, and fats into much smaller molecules. Now the osmotic pressure of a solution is dependent upon the yiumher of molecules and ions it contains, hence by breaking down these few large molecules with very little osmotic pressure into many small molecules, the osmotic pressure in these cells and tissues be- comes raised above that of the blood-vessels, and consequently water flows out of the vessels because of the increased pressure. We see here the probable explanation of the stimulating influence of metabolic products upon the formation of lymph, noted by Hamburger, Heiden- hain, and others. For suggesting and urging the importance of osmo- tic pressure in the formation of lymph we are indebted particularly to Heidenhain, v. Koranyi,^^ J. Loeb,^^ and Roth.'^ Loeb shows very clearly the relative greatness of the water-driving force of osmotic pressure as compared to that of blood-pressure, by his statement that the osmotic pressure of a physiological salt solution is about 4.9 atmos- pheres, which is twenty times as great as the blood pressure with which we have to do in ordinary physiological experiments. In varying osmotic conditions we may readily see an explanation for the increased lymph flow that occurs during tissue activity; namely, it is due to the in- creased formation of metabolic products. Many of the lymphagogues may act similarly by stimulating metabolic activity, with resulting in- crease in the formation of osmotic pressure-raising products of metab- olism in the organs; e. g., the increased lymph flow from the thoracic duct that follows stimulation of hepatic activity by injection of pep- tone (Heidenhain) or ammonium tartrate (Asher and Busch).'* As we shall see later in considering edema, osmotic pressure may play an important part in the pathological formation of lymph. It must be admitted, however, that there are many difficulties in the way of accepting unqualifiedly the original views as to the importance of " Zeit. f. klin. Med., 1897 (33), 1; 1898 (34), 1.- 12 Pfliiger's Arch., 1898 (71), 457. 13 Englemann's Arch., 1899, p. 416. " Zeit. f. Biol. 1900 (40), 333. 336 EDEMA osmotic pressure in lymph formation. ^^ For example, the lymph contains more chlorides and may have a much higher osmotic pressure than the serum of the same animal (Hamburger, Carlson, et al.).'' Variable Capacity of Colloids for Water. — Colloids of the type of the tissue proteins, i. e., hydrophil colloids, imbibe water with great avid- ity, until a certain proportion of water is present, the proportion varying under different conditions. The importance of this force in the production of edema and related processes was first pointed out by Martin H. Fischer, and has been developed extensively by him." The amount of water which a given hydrophil colloid, such, for exam- ple, as gelatin, or fibrin, will take up, is greatly modified by the reac- tion of the solution and by its content of electrolytes. Very small concentrations of acids or alkalies will greatly increase the amount of water absorbed, while salts reduce it, and the different acids and salts vary in their effects; thus hydrochloric acid causes a greater swelling of colloids than a corresponding strength of sulphuric acid, and calcium chloride depresses the swelhng more than potassium chloride. The effect of the salts is made up of their constituent ions. Non-electrolytes have relatively little effect. The forces developed by this affinity of colloids for water are enormous; thus, to prevent the taking up of water by starch requires a pressure of over 2500 at- mospheres, dried gelatin will take up 25 times its weight of water, and fibrin as much as forty times. Different colloids differ greatly in their affinity for water and in the way in which this afl&nity is modi- fied by electrolytes, and change in a colloid may greatly alter its capacity for swelling; thus, jS-gelatin, which can be formed from ordi- nary gelatin by the action of proteolytic enzymes, has greater capacity for swelling than the original gelatin. Gies especially lays stress on this factor, that is, the alterations of the hydrophilic tendencies of the tissue colloids by enzymes.'^ In the plant world we find striking examples of this character; thus, the succulence of some plants results 15 Gunzberg (Arch. ncerlandphysioL, 1918 (2), 364) states that the passage of water into the intercellular spaces is due to the electrical properties of the mem- brane separating the circulating fluid from the tissues. The element potassium and the ioas H and OH play an important part in this electrical osmosis which is able to drive the fluid in the opposite direction to osmotic pressure. Thus, a dialyzing sack containing Ringer solution minus K immersed in Ringer solution loses weight. Perfusion of frogs with Ringer solution minus K produces marked edema. 16 Amer. Jour. Physiol., 1907 (19), 3G0; 1908 (22), 91. 1' See Fischer's Monograph, "Oedema and Nephritis," New York, 1915; also numerous articles in the Zeit. f. Chem. u. Ind. d. Kolloide. An especially thor- ough discussion of this theory is contained in the Biochemical Bulletin, Vol. I., giving a bibliography of Fischer's work, together with articles on Gies' observa- tions on the modification of the hydrophilic tendency of proteins by enzyme action. 1* A definite and clear-cut example of the swelling of a tissue under the in- fluence of acid of metabolic origin is shown in the muscle cell in Zenker's waxy degeneration (Wells, Jour. Exper. Med., 1909 (11), 1). FORMATION OF DM I'll 337 from the conversion of i)ol3'-.sa('('li;iri(l(',s witli little li^'di'alion capacity into hyclrophilic pentosans and mucilages.'^ On the basis of the facts briefly suniniarized above, the proportion of water present in any cell or in anj^ fluid of the body which contains colloids, is assumed to be determined by certain factors, namely (1) the character of the colloids themselves; (2) the proportion and na- ture of acids or alkalies present in the fluids in and about the colloids; (3) the i^-oportion and nature of the salts. All these factors are changeable, and therefore the amount of water present in the cell or fluid varies accordingly. Thus, if a cell through its metabolism de- velops from such a non-electrolyte as sugar (which has no consider- able effect on the water content of the protoplasm), an organic acid, such as lactic acid, which has a large effect in increasing the affinity of the colloids for water, the cell will, presumably, take on more water, perhaps to a degree to cause intracellular edema. The acids diffusing from the cell into the intercellular spaces or into the lymph will cause equally well an increased affinity for water in the colloids here present, leading to intercellular edema. Conversely, neutralization of acids present in a colloidal solution, by alkaline salts brought by the blood, will decrease the affinity of the colloids for water which will escape from the colloids as they shrink. • This theory, which introduces a hitherto unappreciated factor into the considerations of lymph formation and edema, is of the utmost importance. It practically eliminates osmotic pressure, also the cell membranes so essential for the efficiency of this force, and in view of the difficulties that have arisen in trying to fit the cell membrane hypothesis and osmotic pressure to many facts of normal and patho- logical biology, an alternative hypothesis is welcome. As pointed out above, the forces involved in the swelling of colloids are so large as to be of great significance, and the amounts of electrolytes necessary to cause considerable variations in colloidal swelling are not more than can be present under normal and pathological conditions; conse- quently the possible influence of colloidal swelhng must be taken into account in all consideration of pathological processes. Whether or not it is capable of as universal application as Fischer mamtains, remains to be demonstrated, and there are,' indeed, some facts that do not seem to be in harmony with this theory. Summary. — We see from the above discussion that numerous the- ories have been advanced to explain the normal formation of lymph, and as their basis exist several different possible factors. Filtration, active secretion by the capillary endothelium, attraction by the tissue- cells, osmosis in response to formation of crystalloids outside the ves- sels, and changes in the affi.nity of colloids for water; all have been shown to be possible causes of lymph formation. It is highly probable that in a certain wa}' all are involved, particularly if we accept the 19 MacDougal and Spoehr, Plant World, 1918 (21), 245. 22 338 EDEMA view of the physical school that "secretion" and "attraction" by the cells are merely the outcome of physical forces; the causes of lymph formation then reduce themselves to absorption, filtration and diffusion. There has been, until recently, no question but that lymph does escape from the vessels through simple filtration, for the pressure inside the capillaries is presumably greater than outside, the capillary walls are not water-tight and they are not impermeable to the sub- stances dissolved in the plasma.^" Likewise osmotic exchanges surely go on between the vessels and the tissue-cells, and the conditions which determine the water content of our colloid solutions constantly vary. The question that remains is, do these few factors account for all of the lymph formation, and are they sufficient by themselves to explain the physiological regulation and the pathological variations in the lymph flow? They are purely physical or mechanical causes, and the "vitalist" school will claim that they are inadequate and that "vital activities" of the cells play the deciding role. But at present the evidence that is being accumulated seems to point more and more strongly to the conclusion that these "vital activities" are but the result of simple well-known physical forces acting under very complex conditions — complex because of the large number of different chemical compounds occurring together, and the varj^ing in- fluence of circulation, food supplies, cell structure, etc. ABSORPTION OF LYMPH By no means all the fluid that escapes from the vessels, nor all the products of cell metabolism, are carried away in the l3'mph — a con- siderable and perhaps the greater part of them is absorbed back into the capillaries directly. A classical proof of this is the experiment of Magendie, who observed that if poisons were injected into the leg of an animal, which had been separated from the body entirely except for the blood-vessels, that poisoning developed in the usual manner. In such experiments the lymph-vessels are severed and probably largely occluded; hence it does not solve the question as to whether substances are absorbed by the blood-vessels under normal condi- tions. Orlow found, however, that during absorption of fluid from the peritoneal cavity there is no perceptible increase in the lymph flow from the thoracic duct. Addition of sodium fluoride, a protoplasmic poison, was found to interfere with this absorption, for which and other reasons Heidenhain and Orlow considered that the absorj:)tion doixMuled upon the "vital activity" of the cells. More nearly reproducing normal conditions were the experiments of Starling and Tubby, who found *° Hill ("Recent Advances in Physiology and Biochemistry," 190G, p. 618) dis- putes the possibility of such a thing as filtration prcssiu'c, on the ground that the structures within the capsule of an organ must all be under tlie influence of the blood pressure alike; but with the presence of an outlet for the fluid, as in glands with ducts, filtration pressure surely can apply. ABSORPTION OF LYMPH 339 that mcthylonc-bluc or iiuligo-carniiii(> injected into the pleura or peritoneum appeared in the urine long before it colored the lymph in the thoracic duct.^^ Adler and Mcltzcr found evidence, however, that not all the absorption is accomplished by the blood-vessels, for obstruction of the thoracic duct retards absorption. That the ab- sorption is not dependent solely upon the circulation and blood pres- sure is shown by the fact that absorption from the peritoneal cavity occurs in dead bodies (Hamburger, Adler and Meltzer). The nature of the mechanism by which fluids are taken into the blood-vessels is still unknown. We can easily understand the en- trance of injected poisons and coloring-matters from the tissues into the blood, because they are more concentrated at the point of injection than in the blood, hence they may diffuse directly through the capil- lary wall. Likewise we can understand the diffusion of water from a hypotonic solution into the blood, but how a solution of the same concentration as that of the blood can enter the blood is difficult to ex- plain. Cohnstein and also Starling attribute this absorption to the proteins of the blood in the following manner: After a fluid is injected into the tissues or serous cavities there occurs a diffusion exchange between this fluid and the blood, until the concentration of the crystal- loids in each is equal; but the proteins of the blood cannot diffuse, and as they exert a positive although very slight osmotic pressure, this difference in osmotic pressure in favor of the blood causes diffu- sion of the extravascular fluid into the blood. Roth has also applied this idea in a rather comphcated manner to the absorption occurring in metabolic processes (see Meltzer), but it must be admitted that it is an unsatisfactory solution of the problem. Fischer would ascribe the passage of fluid to the relative affinity of the colloids of the blood and of the tissues for the fluid, and this would be towards the blood when- ever the blood colloids had, from whatever possible cause, a greater affinity for the fluid than the tissue colloids. Passage of the fluid from the tissues into the lymph stream was very easy to understand in the light of the older conception of the lym- phatic circulation, namely, that the lymph-vessels were merely con- tinuations of the interstitial spaces; we could then assume that as soon as the fluid left the blood-vessels it was practically within the lymphatic system, and was crowded along the lymphatic channels by the vis a tergo, aided by the valves of the lymph-vessels and the intra- thoracic vacuum. But it now seems, particularly through the studies of MacCallum,22 that the lymphatic vessels form a closed system, not in communication with the interstitial spaces. This being the case, we have to explain the passage of the lymph through the walls of the lymphatic vessels, and this is a problem which is not by any means a simple one, and which has yet to be investigated. It is significant 21 See Mendel, Amer. Jour. Physiol., 1899 (2), 342. " Johns Hopkins Hosp. Bull., 1903 (14), 105. 340 ■ EDEMA that the thoracic lymph has a higher osmotic pressure than the blood of the same animal (Luckhardt),^^ so that the lymph which enters the duct must do so against the osmotic pressure. THE PATHOGENESIS OF EDEMA With the facts and h3''potheses mentioned in the preceding para- graphs in mind, we may consider their bearing on the production of abnormally large accumulations of lymph in the tissues, that is, edema. We can imagine any one of the following factors as causing or helping to cause such a pathological accumulation: 1. Obstruction to outflow through the lymph- vessels. 2. Increased blood pressure. 3. Decreased extravascular pressure. 4. Increased permeability of the capillary walls. 5. Increased filterability of the blood plasma. 6. Osmotic pressure changes — either an extravascular increase or an intravascular decrease. 7. Changes in the affinity of the colloids for water. These may be taken up one by one, and considered in relation to their bearing upon the general problem of edema. 1. Obstruction to Outflow through the Lymph=vessels. — Be- cause of the very abundant anastomosis of the lymphatic vessels it is extremely difficult or impossible to cause any appreciable obstruction to the lymphatic circulation by occlusion of lymphatic trunks in the limbs or organs of the body, and in pathological conditions this possi- ble cause of edema is seldom actually observed. The chief instance of edema from lymphatic obstruction is observed after occlusion of the thoracic duct by tumors, tuberculous processes, animal parasites, or thrombosis; such occlusion is usually followed by rupture of the duct or its tributaries, with the production of chylous ascites or chylothorax, and chyluria. Filaria or their ova may occup.y so many of the lymph- atic channels of an extremity (leg) or part (scrotum) that the anasto- motic channels are thoroughly blocked, with a resulting local edema that in course of time is followed by the production of inflammatory connective tissue and elephantiasis.-^ Chronic lymphangitis or plug- ging of the lymph vessels by cancer cells may also result in lymphatic obstruction to such an extent that chronic edema results. It would seem, from Opie's experiments,-^ that the acute edemas may at times depend upon lymphatic obstruction, for he found that exi)erimental edema of the liver, produced by cantharidin, seems to be determined by inflammatory processes which occlude the sinuses of the l^-mph glands through which the hepatic l3'niph passes. =" Anicr. .Jour. Physiol., 1910 (25), 345. ''' Miiiison, Allbutt's System, 1897 (ii), 10S2. " Jour. Expcr. Med., 1912 (16), 831. PATIiaCESESIS OF EDEMA 341 Another way in which i'(hMna may \)v caused oi- infhienced by lymphatic obstruction is generally overlooked, but it is possibly of great importance; namely, from pressure upon the lymph channels by dilated vessels in hyperemia, or by cellular exudates and swollen tissues in inflammation. We see evidence of this in the rapid absorp- tion of exudates that frequentl}- follows the removal of but a part of the fluid in a chest cavity; apparently the decrease in pressure frees the paths of absorption and permits them to take up the remaining fluid. In inflammatory edema the lymphatic obstruction is probably not great, for Lassar found that the amount of lymph escaping from an edematous extremity is much greater than from a normal one; but in the case of strangulated hernias or other conditions in which edema results from circular constriction, obstruction of the h^mphatic vessels may be a factor of no mean importance. In general stasis the increased pressure in the veins of the neck may interfere with the passage of the fluid out of the thoracic duct into the blood. There is no difficulty in understanding edema from the above causes — it is simply a passive congestion of the lymphatic circulation, and no chemical factors are involved. The nature of the fluid found in such forms of edema will be discussed later. 2. Increased Blood Pressure. — This takes us back to the filtra- tion theory of lymph formation, and as it is generally conceded that more or less fluid escapes from the vessels by this mechanical process, ^^ the questions to be decided are: Can and does increased blood pres- sure, alone and wdthout other aiding factors, cause edema? If not, does it play an auxiliarj^ part in producing edema, and how important a part may this be? Many experiments have been performed with the object of answering these questions, with more or less conflicting re- sults. Cohnheim demonstrated that vasodilation (active hyperemia) alone will never bring on an edema; and many observers state that ligation of the femoral or other large veins will not cause edema in animals. However, when the vein is occluded, and the arteries are dilated by cutting their vasoconstrictor nerves, then edema may result (Ranvier, Cohnheim); but whenever venous outflow is impeded, we have other factors than simply increased pressure to consider, for the nourishment of the parts is decidedly impaired, and, as we shall see later, this may be of much greater importance than is the associated rise in blood pressure. To produce edema in the lungs by mechanical forces it is necessary to ligate the aorta and its branches, or the pul- monary veins (Welch), As such high pressures do not occur in any pathological concHtions, it is safe to assume that increased pressure alone is not capable of causing by itself the pulmonarj^ edema so fre- quently observed clinically. Welch, ^^ however, has supported the ^^ A rise of blood pressure leads to an increase in the hemoglobin of the blood, presumably because the fluid is forced out into the tissue spaces (Scott, Amer. Jour. Physiol., 1917 (44), 298). " Virchow's Arch., 1878 (72), 375; see also Meltzer {loc. cit.). 342 EDEMA hypothesis that a disproportion between the working power of the left ventricle and of the right ventricle may lead to pulmonary edema through pulmonary hyperemia. In the edema of passive congestion, increased blood pressure would seem to be an important factor, and there is no doubt that with an increased pressure of the degree ob- served in such conditions some increase in the hanph flow would result ; but from the evidence at hand it is improbable that the amount of lymph so secreted would ever be more than the lymph-vessels could carry away. Even the added obstruction to lymphatic flow due to pressure upon the lymph capillaries by congested blood-vessels, and the resistance to the lymph escaping from the thoracic duct offered by the increased pressure in the subclavian vein, would not satisfac- torily account for the edema of cardiac incompetence. Not to go into details here, it may be stated that the impression prevails that uncom- plicated rise in blood pressure is not sufficient by itself to produce edema. Some of the reasons for belittling this factor will be brought out in the subsequent discussion. 3. Decreased Extravascular Pressure. — This factor is particu- larly prominent in the so-called "ede^na ex vacuo," which occurs after the absorption of an area of tissue so located that the surrounding tissues cannot contract or fall in to fill the gap, e. g., brain softening, serous atrophy of fat. A still better example, however, is the edema that follows local decrease in atmospheric pressure in "cupping." In these instances the edema depends partly upon increased transu- dation, and partly on the retention of the fluid in the tissues, because it cannot well leave them against the atmospheric pressure. The idea advanced by Landerer that decreased elasticity of the tissues was a possible cause of edema has been attacked by Boninger.^^ During the early stages of edema the elasticity of the skin may be measurably decreased,^^ even when no edema is demonstrable by palpation, but this is not evidence that any loss of elasticity occupies a causative relation to the edema. The tissues can take up water until as much as six kilos has been added to the weight of the entire body before any edema can be detected by palpation (Widal). Edema ex vacuo is again an illustration of edema due to purely mechanical causes, but it is of little practical importance. 4. Increased Permeability of the Capillary Walls. — The im- portance of this factor in the production of edema was first demon- strated by Cohnheim and Lichtheim, who found that the jiroduction of an enormous increase in the amount of fluid in the blood (hydremic plethora) by injecting large quantities of salt solution, caused an edema of the viscera and serous cavities, but not any subcutaneous edema until tlic skin had been irritated l)y some means, such as hot water, iodiu, etc. By this irritation the capillary walls are injured, and an 2" Zeit. exp. Path. u. Ther., 1905 (1), 163. "Schade, Zeit,. exp. Patli., 1912 (11), 369. PATHOGENESIS OF EDEMA 343 excessive escape of the blood fluids follows. Magnus also showed that poisoning with arsenic, which injured the vessels, favored the experimental production of edema by transfusion. Starling, as noted before, observed that the permeability of the capillaries varies normally in different organs and tissues, which dotorminos quantita- tive and qualitative differences in the lymph normally flowing from various vascular areas. Heidenhain's "lymphagogues of the first class, " which are all poisonous substances, probably act by increasing the permeability of the capillaries, and in this way they produce local urticaria, which is often observed as a result of poisoning by these same lymphagogues, e. g., shellfish and strawberry poisoning. Just what changes are produced in the capillary walls that render them more permeable we do not know. Possibly in some instances it is a partial solution of the intercellular cement substances, possibly an enlargement of the stomata through loss of tonicity of the endothelium (Meltzer), sometimes it may be actual death of the endothelial cells, or, as Heidenhain and Cohnheim thought, it may be a stimulation of the endothelial cells to increased secretory activity. Fischer believes that a change in the hydrophilic tendency of the colloids, induced especially by acids formed in asphyxiated conditions of the cells, alter their structure and with that their permeability. Under pathological conditions increased permeability of the capil- lary walls is probably one of the chief factors in the production of certain forms of edema. We see evidence of it particularly in inflam- matory edema, with its protein-rich exudate. It cannot be doubted that in such conditions actual physical alterations take place in the capillaries, when we see that the slightly diffusible proteins escape from the vessels in the same proportions as they exist in the plasma; there can be here no question of heightened cell activity or increase in osmotic pressure, especially not when we note the indistinguishable transition of such an inflammatory exudate into one containing leu- cocytes and red corpuscles, which must pass through openings of some kind in the vessels. Edema due to inflammation and poisoning certainly depends to a large degree upon alterations in the vessel- walls. The question remaining is, do edemas that are not asso- ciated with distinct inflammatory or toxic influences depend also upon the vascular permeability? — does increased permeabiHty ever lead to the formation of protein-poor transudates? Cohnheim was inchned to attribute nearly all edema to this cause, for in passive congestion, or nephritis, or any of the common causes of edema, it is easy to find reason for the belief that poisons may be present in the blood; and as there was good evidence that the blood pressure alone could not account for the edema, it was natural to ascribe all these forms of edema to the action of toxic substances upon the capillary walls, lead- ing to increased permeability; or, what might amount to the same thing, increased secretory activity of the endothelium, as understood 344 , EDEMA "by Heidenhain. It is impossible at this time to eliminate as non- existent this secretory-activity doctrine, but, as we hope to show later, there exist other factors in all these non-inflammatory edemas that are suff'.cient to account for the edema without our having recourse to this hypothesis. For the present, therefore, we may consider altered capillary permeability as an essential factor in edemas characterized by protein-rich fluids (exudates), and state that the influence of al- tered permeability in the production of protein-poor fluids (trans- udates) is not proved, and is perhaps not of importance, although the evidence of recent studies on experimental nephritis seems to point more and more to the importance of vascular changes in acute nephritis, at least. ^° 5. Increased Filterability of the Blood Plasma. — This takes us back to Richard Bright's conception of renal drops}'. He im- agined that through the great loss of albumin in the urine the blood became so thinned and w^atery that it could filter through the vessel- walls, while normal plasma, he thought, was too thick and viscid to do so. The same idea was applied to the edemas of cachexia in cancer, etc., chlorosis, and all forms of edema associated with a decrease in the corpuscular or protein elements of the blood. With our present knowledge of diffusion of crystalloids and colloids we can readily ap- preciate that a decrease in the blood colloids, such as might occur in these diseases, could not facilitate the filtration of fluids through the capillary walls to any considerable degree. On the other hand, the amount of colloids in the blood will greatl}' modify' the amount of fluid held in the blood; e. g., acacia is used in intravenous injections because it holds in the blood vessels a large amount of fluid by virtue of its hydrophilic character. Stewart and Bartels considered that in renal dropsy the increased filterability of the plasma was not due so nuich to the loss in albumin as to retention of water, which caused an hydremic plethora. But this factor was soon eliminated, for it was found that complete anuria, produced by ligating both ureters, does not cause edema; and also that to produce an edema by increasing the water of the blood it was necessary to increase it many times an nmch as it can ever be increased by disease. Simply increasing the proportion of water by removing part of the blood and injecting a corresponding amount of salt solu- tion did not cause edema (Cohnheim and Lichtheim). We may, there- fore, look upon the hypothesis of increased filterability of the blood as chiefly of historic interest, and not important in the causation of edema. In the presence of other factors for the production of edema, however, the amount of fluid in the vessels is important; thus Pearce''^ found that in experimental uranium nephritis hydremia exerted a marked influence on the i)r()du(;ti()n of edcMiia. 30 Sec Sehmid and Schlayer, Deut. Arch. klin. Med., 1911 (10-4), 44. »' Arch. Int.. Med., l'»OS (3), 422. PATIKXyENESIS OF ICDKMA 345 (). Disparity of Osmotic Pressure in Favor of the Tissues and Lymph over the Blood. — On a proccdiiifi; pa^o wo luiv(! already con:sidercd the incans by which changes in osmotic pressure in the tis- sues are brought about, and how they may lead to an accumulation of fluid. The importance of osmotic pressure in causing pathological edema was suggested by J. Loeb''- in his studies on the physiological action of ions. He stated that edema occurred when the osmotic pressure was higher in the tissues than it was in the blood and lymph, and the cause was to be sought in conditions that lowered the osmotic pressure of the blood and lymph or raised that of the tissues. This condition he found in the accumulation of metabolic products: — in the case of muscle, tetanization of a frog's muscle for ten minutes raised the osmotic pressure over one atmosphere; separating a muscle from its blood-supply led to such an increase in osmotic pressure that it took up water from a 4.9 per cent. NaCl solution, which has a pres- sure of over thirty atmospheres. When we consider that in his studies on lung edema Welch was able by ligation of the aorta to raise the blood pressure less than ifo atmosphere, we begin to appreciate how much more powerful are the physico-chemical forces that are at work in the body than is the blood pressure, even of the aorta itself. Loeb found that whenever oxidation is impaired in a tissue its osmotic pressure rises, which he ascribed to the accumulation of in- completely oxidized metabolic products, particularly acids, and as a result the muscle takes up water and becomes edematous. On this basis we might explain the edema of venous stagnation as due to ac- cumulation of products of metabolism, partly because of impaired oxidation, partly, perhaps, because of their slow removal in the blood on account of the circulatory disturbance. The so-called "neurotic" edemas may possibly be explained by local increase in metabolic ac- tivity brought about by nervous stimuli, which causes increased forma- tion of substances raising osmotic pressure in the stimulated tissues. In renal edema the retention of water also seems to depend rather on osmotic pressure than on circulatory disturbances or alterations in the vessel-walls, for it has been shown that retention of chlorides, which the diseased kidneys do not eUminate normally, is an important cause of the dropsy in some cases. The chlorides accumulating in the tissues lead to an increased osmotic pressure, which causes the abstraction of water from the blood and its retention in the tissues. (The details of this subject will be considered later.) Conversely, Meltzer and Salant found that salt solution is absorbed from the peritoneal cavity more rapidly in nephrectomized rabbits than in normal rabbits because metabolic products accumulate in the blood and raise its osmotic pres- sure above normal; and it was observed by Fleisher and L. Loeb^^ that the rate of absorption of fluid from the peritoneal cavity is in- creased when the osmotic pressure of the blood is raised. 32 Pfluger's Arch., 1898 (71), 457. " Jour. Exper. Med., 1910 (12), 510. 346 EDEMA There are 3ome difficulties, however, in applying the influence of osmotic pressure as an explanation of all edemas. For example, in edema of the lungs, as Meltzer points out, what is the force that drives the fluid into the empty air-cells? Equally difficult to explain as the result of osmotic disturbance is the distribution of fluid that is seen in cardiac dropsy. The fluid does not accumulate in the tissues where metabolism is greatest, or where the most oxygen is used; but rather in the inactive subcutaneous tissues and in the serous cavities. Possi- bly the original transudation does occur in the muscles and sohd viscera, and the fluid is then mechanically forced out of them into the surrounding tissue-spaces, later settling according to the laws of gravity or according to the distensibility of the tissues. It is im- portant to take into consideration the fact that demonstrable edema does not manifest itself until a very large quantity of fluid has been retained by the body — as much as six kilos, according to Widal. Increased Hydration Capacity of the Tissue Colloids. — According to Fischer's theory this factor is of greater importance than any of the preceding, and of chief importance in increasing the amount of water present in the tissues are organic acids formed during metabolism. For example, the great power of asphyxiated muscle to take up water from a strong salt solution, which J. Loeb ascribed to the osmotic pressure of the acids formed in asphyxia, is attributed by Fischer to the influence of these acids upon the capacity of the colloids for water, and this explanation seems to be in better agreement with the facts, especially since Overton has shown that even if all the proteins, carbohydrates and fats in a muscle were split into the greatest possi- ble number of simple molecules and ions, the resulting osmotic pres- sure would not be sufficient to account for the amount of water taken up. Furthermore, when cells with demonstrable semi-permeability die, they at once lose their semi-permeability, and in consequence their osmotic pressure falls — but dead cells and tissues often exhibit great power of taking up water and becoming edematous.^'* It is an in- disputable fact that edema is especially associated with conditions of asphj^xiation, and the attempt to explain this by the increased osmotic pressure of the products of incomplete oxidation seem to harmonize with the facts far less successfully than the apphcation of the prin- ciple of colloidal swelling. A common error of the critics of this theory is that of assuming that free acid must be present to cause swelling. This is not at all true. An amount of acid far less than enough to saturate the acid-binding property of a protein or to be detected by indicators will greatly increase the amount of water which " The secreted fluid of postmortem thoracic lymph flow diff"ors from normal thoracic lymph in bein{; more cloudy, often bloody, contains more solids, has a higher molecular concentration with decreased electrical conductivity (.lappelli and d'Errico, Zeit. f. Biol., 1907 (50) 1), all of which findings are in agreement with the hypothesis that postmortem lymph flow depends upon changes in the cells, caused oy asphyxia and not dissimilar to the changes of acute nephritis. PATHOGENESIS OF EDEMA 347 this protein will combine. Presumably the colloidal carbohydrates and lipoids may also play a part in the water absorption of tisr^ues. Fischer's theory of edema, in his own words, is this: "A state of edema is induced whenever, in the presence of an adequate supply of water, the affinity of the colloids of the tissues for water is increased above that which we are pleased to call normal. The accumulation of acids within the tissues brought about either through their abnor- mal production, or through the inadequate removal of such as some consider normally produced in the tissues, is chiefly responsible for this increase in the affinity of the colloids for water, though the possi- bility of explaining at least some of the increased affinity for water through the production or accumulation of substances which affect the colloids in a way similar to acids, or through the conversion of colloids which have but little affinity for water into such as have a greater affinity, must also be borne in mind." In support of this theory he advances evidence which he interprets as indicating that: (i) "An abnormal production or accumulation of acids, or condi- tions predisposing thereto, exist in all states in which we encounter the development of an edema. (2) The development of an edema in tissues is antagonized by the same substances which decrease the affinity of the (h> drophilic) emulsion colloids for water (salts) and is unaffected by uhe presence of substances which do not do this (non- electrolytes). (3) Any chemical means by which we render possible the abnormal production or accumulation of acids in the tissues is accompanied by an edema." There are many features of lymph formation and edema with which this theory seems to harmonize well, and others with which it does not seem to agree so well, if at all, so that at this time it is a fair statement that the theorj' is under consideration, but the limita- tions of its applicability have not yet been agreed upon. It has met with much adverse criticism, some of which was poorly founded, but the fact cannot be disputed that the amount of water that colloids will hold varies greatly with changes in the colloids. We may not know absolute^, at present, whether the changes that take place in the colloids during life are great enough to alter their water content appreciably, but it is highly probable that they are. In many in- stances the principles of colloidal hydration offer the best explanation of observed conditions, and their application often elucidates matters more satisfactorily^ than any other working hj'pothesis. Certainly the}' cannot be disregarded in considering the factors that may come into play in producing edema. Summary. — We find that a number of factors may be considered as responsible for edema, some of them being prominent in one in- stance, some in another, but in few cases can we consider one factor alone as the sole cause. In most of the forms of edema, such as those due to renal disease and cardiac disease, it now seems probable that 348 EDEMA either osmotic pressure changes or changes in the affinity of the tissue colloids for water, play the most important part; whereas in inflamma- tory edema there can be no question that alteration in the capillary walls is the most essential factor. But the mechanical factor of blood pressure cannot be disregarded, although by itself seldom sufficient to cause edema; associated with other factors it is undoubtedly an im- portant agency, for there are few edemas that arc not associated with increased blood pressure. Hydremia and hydremic plethora may al- most be disregarded, except in so far as they may cause altered metab- olism in the tissues, injury to vessel-walls, over-saturation of the blood colloids, and decreased osmotic pressure within the vessels. Ljanph- atic obstruction is possibly a factor of some secondarj^ importance if we consider that distended vessels and tense tissues may occlude the lymph capillaries. Special Causes of Edema We may now consider which of the above factors are at work in bringing about edema under the conditions in which it is usually observed clinically. Before taking up the detailed consideration of edematous conditions, however, it may be well to call attention to the fact that our knowledge of edema, and especially its clinical recog- nition and study, has been handicapped by the lack of a suitable ob- jective method of detecting and measuring edema. We are in the same position in respect to edema that we were to blood pressure when the only measure was the clinician's forefinger. An attempt to remedy this defect has been made by Schade,^^ whose ''elastometer" reveals and measures degrees of edema not discernible bj^ the palpating finger. A study of edema with this instrument in the hands of Schwartz^^ has revealed many interesting facts, but as yet the appa- ratus is too complicated for general clinical use. "Cardiac" Edema. — Passive congestion introduces nearly all these aforementioned factors, for in addition to the increased blood pressure there is also an opportunity for changes in the capillary wall, either from stretching and thinning of the cells and cement substances, or from "loss of tone" in the endothelium surrounding the stomata (Meltzer), or from toxic injury by accumulated products of tissue metabolism. When the stasis is nearly complete, or if it is comp'ete for a time and then relieved, the endothelium may })e injured through lack of nourishment. As the edematous fluid in chronic passive con- gestion is usually of a watery type, poor in proteins, the edema is probably less dependent upon capillary permeability than upon other factors, except in the ease of acute stasis, when the fiuid partakes of the character of the exudates. Presumably the accunuilation of crystalloids within the tissues also plays a part in this form of edema, as the osmotic pressure is raised in tissues having deficient oxygen " Arch. Int. Med., 1910 (17), 390 and 4r-,9. NEPHRITIC EDEMA 349 supply. But Fischer liolds tluit tl)c retluction in oxidation acts chiefly by increased production of acids, which greatly increase the affinity of the tissue colloids for water and at the same time alter the colloidal state of the capillary endothelium so that the capillaries ijecome more permeable. Finally, there is probably more or less obstruction to lymphatic outflow because of the increased pressure on the lymphacic channels, and perhaps, also, in the case of cardiac incompetence, ob- struction to the discharge of lymph from the thoracic duct into the subchu'ian vein against the high intravenous pressure. Renal Edema. — We must recognize under this heading two dif- ferent types of edema. ■ In acute nephritis (e. g., in scarlatina) toxic materials appear to be the chief cause, and, as Senator contends, in- jure alike the capillaries of the renal glomerules and of the subcu- taneous tissues; in each case there results an increased permeability which is manifested by albuminuria as a result of the injury to the renal capillaries, and by edema as a result of the injury to the tissue capillaries. This sort of edema is allied to that produced by peptone and similar 13'mphagogues, and we might well imagine that the mech- anism consisted merely in an injur}- to the capillaries through which excessive fluid is driven by the blood pressure, were it not for such observations as those of ]\Iendel and Hooker,^*' who found that post- mortem flow is increased b}- these h'mphagogues also. We can hardly account for the force exhibited in postmortem lymph flow on any other ground than that it is furnished by osmotic pressure or colloidal absorption unless we wish to fall back upon "vital activity" of the surviving cells. Hence it is probable that even in the edemas of toxic conditions, such as acute nephritis, physico-chemical factors play a part, the responsible substances probabl}'' being abnormal or excessive metabolic products of the cells affected by the poisons. An interesting observation made by Bence" is that nephrectomized rabbits develop an edema even when they are given no water at all; this would seem to indicate an increased affinity of the tissues for water when the renal functions are deficient. Hydremia is always a favoring factor, however, and probably important in nephritic edema, ^* while nearly all students of acute experimental nephritis find evidence that the resulting edema depends very much upon the changes in the vessel- walls.39 In the more common edema of chronic nephritis we have to con- sider, among other factors, the blood pressure. That this is not an essential or even important cause, however, is shown by the fact that edema is usually much, less marked in interstitial nephritis with high blood pressure than it is in parenchymatous nephritis with a 3« Amer. Jour, of Physiol., 1902 (7), 380. " Zeit. f. klin. Med., 1909 (67), 69. 38 1 earce, Arch. Int. .Med., 1909 (3), 422. 39 See Schmidt and Schlaver, Deut. Arch. klin. Med., 1911 (104), 44; Pollak Wien. kUn. Woch., 1914 (27), 98. 350 EDEMA « much lower pressure. Toxic substances are, of course, also present in the blood, and may alter capillary permeabihty; these toxic substances may account for the lo-calized edemas and erythemas sometimes ob- served in nephritis. But probably most important is the action of the crystalloids which the kidney does not excrete, and which seem to be stored up in the tissues, where they cause transudation of water under the influence of their osmotic pressure. For example, Rzent- kowski"*" found that the average lowering of the freezing-point by the edematous fluid in nephritis was 0.583°, in cardiac dropsy it was 0.548°, and in tuberculous pleuritis 0.526°. This indicates ':hat the osmotic concentration of the fluid is highest in renal dropsy, and supports the belief that here and in cardiac dropsy osmotic pressure plays a more important part than it does in inflammatory exudation. ^^ Of the crystalloids that cause accumulation of fluid in the tissues, sodium chloride seems to be the most important. Retention of Chlorides in Edema. ^^ — From the investigations made by numer- ous clinicians, especially the French, it appears that — (1) in nephritis with edema a retention of sodium chloride frequently occurs; (2) that elimination of the chlorides is often increased during periods of improvement of the edema; (3) that a reduction of the amount of chlorides in the diet sometimes causes a great improvement in the edema, while administration of chlorides may make the edema much worse. There are, however, observations that also indicate that chloride retention does not account for many cases of renal drops}', for com- monly the above-mentioned conditions are not fulfilled." Nevertheless, it cannot be denied that chloride retention is sometimes an important causative factor in the edema of parenchymatous nephritis." If the retained chlorides obeyed the ordinary laws of diffusion, we should expect them to become distributed alike in the blood and tissues, so that they would merely cause an equal increase in the fluids of the blood and of the tissues; that is to say, there would be an hydremic plethora due to retention of water in the body by the accumulating chlorides. But, according to a number of observers, there is a specific retention in the tissues, which Strauss calls " historetention," and which explains the local edema. The way in which the historetention is produced is, however, not understood, and not all observers accept this hypothesis. If chlorides do bear a causative rela- tion to edema, the predilection of the subcutaneous tissues for edematous accumu- lations may be explained by the observation that when salt is given to an animal an undue proportion (28-77 per cent.) accumulates in the skin." In many conditions other than nephritis, there is also a chloride retention (e. g., pneu- monia, cardiac incompetence, sepsis, typhoid), and the edemas observed in these diseases may possibly depend upon chloride retention, as many French authors ^0 Berl. klin. Woch., 1904 (41), 227. ^1 Epstein (Amer. Jour. Med. Sci., 1917 (154), 638) calls attention to the de- crease of serum proteins (sometimes GO to 70 per cent.) and ascribes the edema to lowered osmotic pressure of the blood from loss of colloids. Low protein content of the blood might more probably favor edema by reducing the amount of fluid which the blood can hold as a hvdrophile colloid. '2 Literature, r6sum6 by Widal and Javal, Jour. Physiol, et Pathol, 1903 (5), 1107 and 1123; Rumpf, Miinch. med. Woch., 1905 (52), 393. Review in Albu and Neuberg's "Mineralstoffwochsel," Berlin, 1900, pp. 171-178; Georgopulus, Zeit. klin. Med., 1900 (60), 411; Christian, Boston Med. and Surg. Jour., 1908 (158), 416; Palmer, Arch. Int. Med., 1915 (15), 329. "See Blooker, Deut. Arch. klin. Med., 1909 (96), 80; Fischer, "(Edema and Nephritic." ** See Borchardt, Deut. med. Woch., 1912 (38), 1723. " Schade, Zeit. exp. Path. u. Ther. 1913 (14), 1. Also gives an interesting discussion of the relation of the skin to euenui. INFLAMMATORY EDEMA 351 supRest. Kunijif, indeed, often found more chlorides in edematous fluids of non- nephritic origin than in nephritic edema.'*' Fischer holds that the retention of chlorides in edema is secondary and not primary, for he found that tissues made to take up more water through acidification, also take up an increased amount of chlorides. Inflammatory Edema. — Although here the alterations in the cap- illary walls play an essential role, as shown by the protein-rich na- ture of the exudates, yet most of the other factors are added. In- creased blood pressure is prominent; lymph outflow is impeded by plugging of the lymphatic channels by clots and leucocytes, and by pressure on the outside; there is, undoubtedly, an excessive forma- tion of metabolic products in the tissues, to cause exosmosis, and the asphyxial conditions in inflamed tissues favor acid formation which ma}' cause in the colloids an increased affinity for water. According to Oswald'*^ the permeability of the vessels for proteins becomes spe- cifically altered in inflammation, so that not only the less viscous albumin and pseudoglo])uhn pass through their walls, but also the more viscous euglobulin and fibrinogen. To this class of edemas belong also the urticarias which follow the ingestion of various toxic sub- stances, many of which can be shown experimentally to be lympha- gogues. A good example is the urticaria which often follows the injection of antitoxic or other foreign serums, particularly their re- peated injection; in experimental animals such a serum may cause death very quickly by acute pulmonary edema. All these poisons probably produce urticarial edema by injury to the capillary walls in the subcutaneous tissues, and possibly changes in the hydrophilic properties of the tissue colloids are also produced by the poisons. In the action of vesicants especially, it may well be questioned if changes in the capillary walls and active hyperemia are not supplemented by local metabolic alterations. The edema which follows the sting of insects, which are known to secrete into the wound such acids as formic, seems to be a particularly good illustration of the production of edema by the influence of acids on the tissues (Fischer). Neuropathic Edema. — Until we understand better than we now do the manner in which nervous impulses modify metabolism, it will be difficult to estimate properly the importance of nervous impulses in the production of edema. That nervous control is a possible factor is well shown by manj^ experiments; for example, simple ligation of the femoral vein in animals does not cause edema, but if the sciatic nerve is cut the vasoconstrictors are paralyzed, and edema mav follow (Ranvier).^^ In this case the nervous influence is only indirect through its vasomotor effects. Similarly, stimulation of vasodilator *^ Breitmann (Zentr. inn. Med., 1913 (34), 633) describes under the name of "soda dropsy" a form of edema which results from excessive administration of sodium bicarbonate to correct acidosis in diabetes. " A. Oswald, Zeit. exp. Path., 1910 (8), 226. *^ Similarly, pulmonary edema follows experimental hydremia onlj- when the vagi are cut (F. Kraus, Zeit. exp. Path., 1913 (14), 402). 352 EDEMA fibers may cause edema. It is furthermore possible that nervous stimulation may lead to excessive metabolic activity, with an ac- cumulation of crystalloidal products and acids sufficient to cause edema when supplemented by active congestion and some resulting pressure upon the lymph-vessels. There are certainly many instances in which edema seems to depend upon nervous disturbance; for ex- ample, edema in the area of distribution of a neuralgic nerve; sudden joint effusions in tabetic arthropathy; and especially the typical "angioneurotic" edema. ^^ The only explanation that seems open is the one given above, namely, a combination of local h^-peremia and in- creased metabolic activity. Even the urticarias of apparently me- chanical origin (urticaria factitia), show evidence of a toxic action, in that there occurs a severe nuclear fragmentation (Gilchrist).^" Hereditary Edema. — In a number of families there has been ob- served a peculiar inherited tendency to the occurrence of acute attacks of local edema, which not infrequently have proved fatal when in- volving the glottis. ^^ There can be little question that these instances of hereditary edema depend upon a nervous affection of some kind, it being practically an angioneurotic edema; but how the edema is pro- duced, and what the nature of the nervous alteration may be, are as mysterious as are most other so-called "nervous inheritances." There also are cases of congenital edema, which may occur repeatedly in the fetuses of the same mother and cause habitual miscarriage;^^ and still another class of cases in which the children are born apparently healthy, but develop fatal dropsy w^hen a few weeks old.^'' Nothing is known as to the cause of this condition. Patein*^ has analyzed the fluid in a case of congenital ascites and found it somewhat more like an exudate than a transudate. Nutritional Edema ("War Dropsy" or Famine Edema). ^(Discussed under Deficiency Diseases, Chapter xii.) COMPOSITION OF EDEMATOUS FLUIDS^^ As is well known, the composition of edematous fluids varies greatly according to the cause of the edema and the place where it occui's. In general, non-inflammatory edemas (transudates) contain much less protein than do the inflammatory exudates, as is shown by the follow- ing taljle of analyses by Halliburton^'* and by Rornheim's-'" deter- minations of proteins in ascitic fluids. '"' Metabolism in angioneurotic edema is discussed bv ISIilliMand Pepper, Arch. Int. Med., 191G (18), 551. '-"Johns Hopkins Hosp. Bull., 1908 (19), 49. "Literature, see Fairbanks, Amer. Jour. Med. Sci., 1904 (127), 877; Hope and French, Quart. Jour. Med., 1908 (1), 312; Crowder, Arch. Int. Med., 1917 (20), 840. " W. Fischer, Berl. klin. Woch., 1912 (49), 2403. "Edgeworth, Lancet, 1911 (181), 211). "Jour. IMiariu. et Cliim., 1910 (102), 209. " Many data are given by Cierliarlz, ilamlbuoh dor Bioi'heniio, 190S, II (2), 137. '-0 Adaiiii, Allbutt's System, 189(5 (1), 97. " (Quoted by llamnuirsten, "Physiological C'homistrj-." COMPOSITION OF EFFUSIONS Table I 353 Parts per 100 ot fluid Sp. gr. Total protein Fibrin Serum- globulin Serum- albumin Acute pleurisy . . . Acute pleurisy. . . Acute pleurisy . . . Hydrothorax "1 Aver, of 3 cases / 1.023 1.020 1.020 5.123 3.4371 5.2018 0.016 0.0171 0.1088 3.002 1 . 2406 1.76 2.114 1.1895 3.330 1.014 1.7748 0.0086 0.6137 1.1557 Table II Ascitic fluid in Parts of protein to 1000 c.c. fluid Max. Min. Mean Cirrhosis of the liver 34.5 16.11 5.6 10.10 9 69-21 06 Bright 's disease 15.6-10.36 Tuberculous and idiopathic Carcinomatous peritonitis. . peritonitis. . . . 55.8 5 1 . 20 18.72 27.00 30.7-37.95 35 . 1-58 . 96 The specific gravity varies nearly in direct proportion to the amount of proteins, that of transudates usuall}^ being below 1.015, and exu- dates above 1.018, although there are many exceptions. Indeed, it is often very difficult to decide whether a given fluid is an exudate or a transudate. ^^ According to Rzentkowski,^^ the transudates at the moment they pass out of the vessels are simply solutions of crystal- loids in water and quite free from protein; the small amount of protein found in transudates he ascribes to protein pre-existing in the tissue- spaces. This idea is hardly acceptable in view of the known per- meability of the vessel-walls for proteins in normal conditions; more probably in cardiac and renal dropsies the quantity of protein escap- ing from the vessels is not greatly different from normal, but the 68Rivalta (Rif. Med., 1903; Biochem. Centr., 1904 (2), 529) has suggested the following test to distinguish exudates and transudates: Into a beaker con- taining 200 c.c. of water with 4 drops of glacial acetic acid, let fall a few drops of the fluid to be tested. If an exudate, a bluish-white line is left transiently behind the sinking drops, due to precipitation of the euglobulin and fibrinogen. This test, and also certain modifications (see Rivalta, Policlinico, 1910 (17), 676), seem to give quite reliable results. (See Ujihard, Berl. klin. Woch., i914 (51), 1112). With tuberculous effusions Rivalta's test is positive, but not Mo- relli's test, which consists in dropping the fluid into saturated HgClo solution, a yellowish ring of albuminate forming with non-tuberculous exudates, and a gran- ular precipitate with transudates. (See Zannini, Gaz. degli Osped., 1914 (4), 461). Memmi (Clin. Med. Ital., 1905, No. 3) suggests the larger content of lipase as a means of distinction of exudates. Tedeschi (Gaz. degli Osped., 1905 (26), 88) states that egg-albumen fed in large amounts appears in transudates and not in exudates, and can be detected by the biological precipitin test. Sugar is found more often in transudates (Sittig). " Virchow's Arch., 1905 (179), 405. 23 354 EDEMA excessive fluid escaping in these conditions carries with it no addi- tional proteins, and to this extent transudates in statu nascendi are protein-free. Transudates, even when produced by the same cause, vary in com- position in different parts of the bodj', presumably because of varia- tions in the permeability of the vessels in different vascular areas; just as pleural, pericardial, peritoneal, and meningeal fluids normally differ from one another. Thus C. S. Schmidt^" found the composition of the transudates in different pares of the bodj^ of a patient who died of nephritis to have the following composition: Table III Pleural Peritoneal Subarachnoid i Subcutaneous Water Solids. Organic matter. . Inorganic matter 963.95 36.05 28.50 7.55 978 . 91 21.09 11.32 9.77 988.70 11.30 3.60 7.70 As in this case, the general rule is that while che proportion of salts remains nearly constant, the proportion of protein in edematous fluids in different localities varies in decreasing order as follows: (1) pleura; (2) peritoneum; (3) cerebrospinal; (4) subcutaneous.®^ In the last-named location the specific gravity of edematous fluids may be as low as 1.005, and the proteins even less than 0.1 per cent. (Hoffmann"^). An increase in solids occurs after the eft'usion has existed for some time, presumably because of absorption of water and salts, leaving a slowly increasing proportion of proteins. Further- more, the composition of the patient's blood has considerable influ- ence on the composition of the effusion; this is particularly true in the case of ascites from portal obstruction, the contents of the blood coming from the intestine during digesoion modifying the composition of the ascitic fluid, *^^ Thus Miiller,®* in a case of portal vein throm- bosis, found in the ascitic fluid of a patient on an ordinary mixed diet, 0.179 per cent, nitrogen; on a protein-rich diet, 0.2494 per cent. N; on a protein-poor diet, 0.1764 per cent. N. In cachectic conditions the proportion of proteins is less than in stronger individuals, and, as in the blood plasma, the albumin decreases more rapidly than the globuhn as the cachexia advances (Umber). *^^ Physical Chemistry of Edema Fluids. ^ — The differences be- tween transudates and exudates depend almost solely on their protein contents, for the non-protein elements are almost identical with *" Hoppe-Seyler's Physiol. Chemie. "1 Javal (Jour. phys. et path., 1911 (13), 50S) places the fluids in this order: serum, peritoneal, pleural, sulK-utaiieous, cercl)rospinal. «2 Deut. Arch. kliu. Med., 1S,S9 (44), 313. «=• .See Deni.s and Miiiot, Arcli. Int. Med., 1C17 (20), 879. «^ Deut. Arch. klin. Med., 1903 (76), 563. «6 Zeit. klin. Med., 1903 (48), 364. PHYSICAL CHEMISTRY OF EFFUSIONS 355 the lymph and blood-serum, which naturally must be so since any original or temporary deviation in osmotic pressure must be rapidly cquahzed by diffusion. Thus Bodon^^ finds the concentra- tion of the electrolytes nearly constant in spite of considerable dif- ferences in composition of various edema fluids, indicating that the serosa permits passage of inorganic salts ahvaj^s in the same con- centration, while holding back the organic substances. Transudates contain an excess of NaCl over other electrolytes, while in exudates the proportion of electrolytes other than chlorides is increased over the findings in transudates." The surface tension of exudates is lower than that of transudates,"^ depending chiefly upon the globulin content. Rzentkowski"^ found some slight differences in molecular concentration as indicated bj^ the freezing-point; in tuberculous pleu- risy the average lowering was 0.523°, that of the serum being —0.56"; in cardiac dropsy the subcutaneous fluid gave —0.548°, and in renal dropsy —0.583°; tuberculous peritonitis, —0.523°; cirrhosis —0.536°; carcinomatous edema —0.547°. Of these figures, the most significant is the comparatively high molecular concentration of the fluid in nephritis, supporting the contention that the cause of renal edema is retention of crystalloids.^'' Tieken^^ has found the results in transu- dates, exudates, and other body fluids show^n in Table IV, Table IV Nature of fluid Sp. gr. Freezing- point of effusion, -°C. Freezing- point of blood, -°C. Disease Pleuritic effusion j 1,016 Pleuritic effusion ' 1,018 Pleuritic effusion 1,018 Pleuritic effusion 1,020 Pleuritic effusion 1,016 Pleuritic effusion 1,018 Pleuritic effusion 1,030 Pericardial eft'usion 1,018 Pericardial eft'usion 1,016 Pericardial effusion 1,012 Ascitic fluid 1,024 Ascitic fluid 1,020 Ascitic fluid 1,018 Ascitic fluid 1,013 Ascitic fluid 1,035 Hydrocele fluid 1,016 Cerebrospinal fluid 1,018 Cerebrospinal fluid 1,016 Cerebrospinal fluid 1,020 Cerebrospinal fluid. Cerebrospinal fluid. Cerebrospinal fluid. 1,014 1,017 -0.55 -0.55 -0.54 -0.55 -0.55 -0.64 -0.60 -0.55 -0.56 -0.56 -0.60 -0.57 -0.5S -0.62 -0.65 -0.56 -0.62 -0.64 -0.64 -0.56 -0.56 -0.56 I - 0.56 0.55 0.56 0.56 0.56 0.56 0.58 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.58 0.56 0.58 0.68 0.64 0.56 0.56 0.56 Pneumonia, lobar. Pneumonia, lobar. Tuberculosis. Tuberculosis. Tuberculosis. Valvular heart disease. Empyema; cyanosis. Pericarditis. Pericarditis. Hydropericardium. Cirrhosis of liver. Cirrhosis of liver. Tuberculous peritonitis. Organic heart disease. General peritonitis. Tuberculosis. Uremic coma. Uremic coma. Uremic coma. Tuberculous meningitis. Epidemic meningitis Epidemic meningitis. «« Pfluger's Arch., 1904 (104), 519; also see Galeotti, Lo Sperimentale, 1901 (55) 425. " Griiner, Biochem. Jour., 1907 (2), 383. «8Trevisan, Zeit. exp. Path., 1911 (10), 141. ^^Loc. dt.,^^ and also Berl. klin. Woch., 1904 (41), 227. ^^ Purulent exudates may show a high molecular concentration ( — 0.84° in one case), due to decomposition of the proteins into crvstalloids (Rzentkowski). " Amer. Medicine, 1905 (10), 822. 356 EDEMA The very high figures for effusions in nephritis and cardiac incom- petence indicate the concentration of crystalloids in these fluids, and support the behef that in the formation of both, osmotic pressure is an important factor." Edema fluids are usually alkaline except when bacterial changes lead to acid formation, but they are always able to neutrahze less acid than the blood of the same individual (Opie). Bodon^^ found, how- ever, that while they contain alkali that can be neutralized by titration against acids, yet they resemble the blood in being neutral as far as the presence of free OH ions is concerned. Protein Contents. — As indicated in the tables given previously, these vary greatly in quantity in various fluids ;^^ the quantitative relations of the different varieties of proteins have been less studied. Serum-albumins and globuhns constitute by far the largest part of the proteins, fibrinogen being scanty except in some inflammatory exudates, so that coagulation very seldom occurs spontaneously; pneumococcus exudates seem particularly rich in fibrinogen, which coagulates rapidly and firmly. The differences in the proportion of different serum proteins in transudates is attributed by A. Oswald^^ to the relative viscosity of these proteins which determines their ability to pass through the capillary walls. The viscosity of serum proteins varies in the following increasing order: albumin, pseudo- globuhn, euglobulin and fibrinogen; hence in transudates we may find only the first two, or perhaps only the albumin, while in exudates the two latter appear. Joachim''^ found in pleural transudates and exudates that the proportion of albumin, euglobulin, and pseudo- globulin is always lower in hydrothorax than in pleurisy. Of dif- ferent forms of ascites, the largest proportion of globuhn and the smallest of albumin occur in cirrhosis; while with carcinoma the pro- portions are reversed. In general the albumin is more abundant than the globulin, ^^ but, as Umber*^^ has found, the proportion of albumin sinks more rapidly in cachexia than does the globulin, corresponding to the similar changes in the blood proteins. The amount of protein lost in exudates is strikingly shown by one of Umber's cases of can- cerous ascites; during one year the fluid removed by paracentesis contained not less than three kilos of pure protein, the patient weigh- ing but 55.5 kilos. Several authors have found in inflammatory ascitic exudates a protein having physical and chemical properties much resembhng " Meyer and His (Deut. Arch. klin. Med., 1905 (85), 149) claim that the knv- ering of the freezing-point is less than that of the blood in exudates while form- ing, the same as the blood while stationary, and greater during absorption, which they consider indicates a "vital process" on the part of the cells. "See also v. Jaksch, Zeit. klin. Med., 1893 (23), 225; Kzentkowski (loc. cit.) " ■>* Zeit. exp Path., 1910 (8), 22G. ■'^ Pfiiiger's Arch., 1903 (93), 558. "i See Epstein, Jour. Exp. Med., 1914 (20), 334. COMPOSITION OF EFFUSIONS 357 mucin; it has been especially studied by Unibcr," who finds it (juite similar to the synovial mucin isolated in arthritis by Salkowski, and calls it serosaviucin. Non-Protein Organic Contents. — Proteoses,"' leucine, and tyrosine may be present in small quantities in exudates, being produced by autolysis" (Umber); and also mucoid substances (Hammarsten). Nucleoproteins may be present from leucocytic disintegration in exudates, as well as the products of their further splitting, such as purines and phosphates, daldi and Appiani'*" found uric acid constantly in amounts between 0.0055 g. and 0.0714 g., in all exudates, of which seven were tuberculous and two neoplastic. In three transudates amounts from O.OOt) to 0.011 g. were found. Allantoin is said to have been found in exudates (Moscatelli),*' but this is doubtful. All the other innumerable components of plasma may be found in edematous fluids; thus sugar^^ and urea (Carriere)**^ are often present, as well as other ex- tractives. The amount of urea varies quite as it does in the blood of the same individual, ^^ and it seems probable that all the crystalloid substances present in the blood pass freely into and from inflammatory exudates, so that an equi- librium between blood and exudates is approximated.*'' Sugar is said sometimes to be greater in amount in transudates than in the blood, but in exudates it is usually, if not alwaj's, lower than 0.1 per cent.*^ Glycogen is not present (Car- riere;.*^ By using more accurate methods than have been employed by most of the observers quoted above, Denis and Minot"" found urea, uric acid and creatinin to occur in exudates and transudates in the same concentrations as in the blood, but the sugar content of ascitic fluids is somewhat higher than that of the blood. Creatin, fats and cholesterol are much lower in transudates than in exudates in which they approach the concentration in the blood. In ascitic fluid the urea, uric acid and cholesterol are influenced by the diet. Lipins. — Lecithin is always present, partly bound to globulin and partly free (Christen). ^^ Cholesterol is present particularly in fluids that have been standing for a long time in the body, appearing often as visible crystals shining in the fluid; it probably originates from degenerating cells. Ruppert has described a case of pleural effusion with 1.129 per cent, of cholesterol when tapped the first time, 0.22 per cent, the second and 0.05 per cent, the third. Hedstrom reported finding in an old pleural effusion, 4.5 per cent, of chole.sterol; one year later there was but 0.09 per cent. Zunz*' has described a carefully studied case in which 14 aspira- tions were made; the cholesterol content was about 3 per cent, at first, but fell suddenly to 0.48 per cent, and then remained between 0.5 per cent, and 0.28 per cent. Lecithin varied from 0.1 to 0.04 per cent. As there did not seem to be enough cells present in the fluid to have yielded the obtained cholesterol through their disintegration, Zunz suggests that it may have been secreted by the walls "' Zeit. klin. Med., 1903 (48), 364; also Hoist, Upsalalakar. Forhand., 1904, p. 304. '8 Opie, Jour. Exp. Med., 1907 (9), 391. '3 Histidine and arginine were found in a carcinomatous exudate by Wiener (Biochem. Zeit., 1912 (41), 149). 8° Riforma Med., 1904, p. 1373; also Carriere, Compt. Rend. Soc. Biol., 1S99 (51), 467. 81 Zeit. physiol. Chem., 1899 (13), 202. 82 Sugar was found in only 8 of 23 fluids by Sittig (Biochem. Zeit., 1909 (21), 14) ; but is present in pulmonary edema fluid in proportion equal to or even greater than the blood (Ivleiner and Meltzer). 83 Javal and Adler, Compt. Rend. Soc. Biol., 1906 (61), 235; Rosenberg, Berl. klin. Woch., 1916 (53), 1314. 8^ Wells and Hedenburg, Jour. Infect. Dis., 1912 (11), 349; Scheel, Nord. Med. Laeg., 1916 (77), 610. 85 Hegler and Schumm, Med. Klinik, 1913 (9), 1810. 86 Compt. Rend. Soc. Biol., 1899 (51), 467. 87 Arch. Int. Med.. 1917 (20), 879. 88 Cent. f. inn. Med., 1905 (26), 329. 83 Travaux Ambulance de L'Ocean, La Panne, 1918, Tome II, Fasc. 1. 358 EDEMA of the cavity. Weems^° has described a similar case, with 1.39 per cent, in the first fluid drawn, but smaller amounts in fluids withdrawn later; this patient had a marked hypercholesterolemia. ArnelP^ found 0.41 i)er cent, of cholesterol in a tuberculous pleurisy. In most of these cases some fats have been present, Weems finding 0.33 per cent, and Ruppert 0.36 per cent. Toxicity. — Contrary to earlier ideas, transudates are not demonstrably toxic, even in nephritis (Baylac,^^ Boy-Teissier,.'^ Lafforcade'^), and therefore the toxic manifestations frequently observed after reduction of edema in nephritis, and ascribed to absorption of poisons in the transudates, are probably due to some other cause. In inflammatory exudates, of course, the causative agents as well as the products of cell destruction render the fluids poisonous. Enzymes and Immune Bodies. — All the enzymes of the plasma may appear in edematous fluids, being in all cases probably more abundant in exudates than in transudates. According to Carriere,^^ oxidases are inconstant, even in exu- dates. Lipase is said to be much more abundant in exudates than in transudates. ^^ (Concerning proteolytic enzymes see "Autolysis of Exudates," Chap, iii.) The various immune bodies, cytotoxins, hemolysins, bacteriolj'sins, agglutinins, etc., seem to pass freely into both transudates and exudates, and their presence is not characteristic of either, ^^ but as a rule the proportion is much higher in exudates.^* Peptid-splitting enzymes are usually found in such fluids, ^^ especially tuberculous exudates,' and these enzymes seem to be different from both erepsin and trypsin. Probably this type of enzyme is more often present than trypsin. Antitryptic activity is usually high, unless exhausted by the presence of much trypsin from cell-rich exudates. Purulent fluids are usually poor in opsonins;- in non-purulent fluids the opsonin content varies with the amount of proteins.' Turpentine exu- dates may sometimes be more strongly bactericidal than the serum of the same animal.'' Exudates usually contain about as much complement as the serum, but in suppuration the complement disappears; transudates contain little of either complement or hemolysins.^ Precipitin Reactions, etc. — Edematous fluids have been often used as a source of material in immunizing animals against human proteins. The precipitins thus formed are specific for human serum or for the proteins of the effusion, but can not be used to differentiate a transudate from an exudate, or a hj'drothorax fluid from an ascites fluid (Quadrone).^ Immune bodies, complement, agglutinins and antitoxins are present in effusions; e. g., the common use of blister fluid for the Widal test. Furthermore, according to Hamburger,^ edema fluid is distinctly more bactericidal than normal lymph. Varieties of Edematous Fluids^ On the preceding pages have been mentioned the chief differences in the characters of the effusions in the usual sites, ^ with their varia- 90 Amer. Jour. Med. Sci., 1918 (156), 20. " Hygiea, 1917 (79), 737. '••2 Compt. Rend. Soc. Biol., 1901 (53), 519. '■>^ Ibid., 1904 (56), 1119. 9^ Gaz. heb. Med. et Chir., Jan. 28, 1900. "^ Compt. Rend. Soc. Biol, 1899 (51), 561. 9" Zeri, II Policlinico, 1903 (10), No. 11; Memmi, Clin. med. Ital., 1905, Xo. 3; Galletta, Chn. med. Ital. 1911 (50), 143. '■" Granstrom, Inaug. Dissert., St. Petersburg, 1905. 98 Not corroborated by Ludke, Cent. f. Bakt., 1907 (44), 268. See also Delrez Bull. acad. Rov. Med. Belg., 1919 (29), 733. 99 Hall and WiUiamson, Jour. Path, and Bact., 1911 (15), 351. iSee H. Koch, Zcit. Kinderhoilk., 1914 (10), 1. 2 Opie, Jour. Expcr. Med.. 1907 (9), 515. 3 Bohmc, Dcut. Arch. klin'. McmI., 1909 (96), 195. Mlastaedt, Zcit. Inimunitat., 1912 (13), 421. "■ Aronstamm, Cent. f. Bakt., 1914 (74), 326. eCent. f. Bakt. (Ref.), 1905 (3()), 270. ' Virchovv's Arch., 1899 (156), 32S). " Chemistry of Pus and Sputum are discussed under Inflammation, Chapter xi. 9 Literature and r6sume on pleuritic exudates, see Ott, Chcm. Pathol, dor Tubcrc, 1903, {). 392. COMPOSITION OF EFFUSIONS 359 tions in protein contents, which variation agrees with StarUng's state- ment that the permeabiHty of the capillary wall for proteins differs normally in different localities. Some of the other effusion fluids not mentioned previously have particular properties of some interest. Subcutaneous Efifusions.^" — When of non-inflammatory origin these are vcrj' watery, having ordinarily a protein content of from 0.1 to 0.2 gm. per 100 c.c, there being more globulin in nephritic than in cardiac dropsy. The non-coagulable nitrogen and chloride content are not so high as in the blood of the same patients, but the ash is the same as that of the serum. The specific gravity may be as low as l.OOo, but the solids increase with the duration of the edema. Hydrocele and Spermatocele Fluids. — These have been studied particularly by Hammarsten, who found the average result of analyses of seventeen hj'drocele fluids and four spermatocele fluids as follows: Table V Hydrocele Spermatocele Water... 938.85 986.83 Solids 61.15 13.17 Fibrin 0.59 Globulin 13.25 0.59 Seralbumin 35.94 1.82 Ether-extractive bodies 4 . 02 ] Soluble salts 8.60 10.76 Insoluble salts 0 . 66 J Marchetti" found in ten specimens of hydrocele fluid rather higher results for the solids than did Hammarsten. He found 57.8 to 104.2 p. m. of solids, contain- ing organic sub.stances 48.8 to 95.02, and inorganic substances 8. 10 to 9.56; proteins, 33.5 to 90.19; ratio of globulin to albumin as 2.56 to 9.11. Among the pro- teins is found 1 to 4 p. m. that is not precipitated by heat. Corresponding with the analytic results, the specific gravity of hydrocele fluid is higher, 1.016 to 1.026 as against 1.006 to 1.010 for spermatocele fluid. Cholesterol is often abundant in hydrocele fluids, appearing to the naked eye as glistening scales. Patein'^ found sugar in most specimens of hydrocele. Apparently hydrocele fluid stands inter- mediate in properties between transudates and exudates.'^ Usually it contains but little of the immune bodies from the blood (Delrez).^* Meningeal Effusions. ^^ — Normal meningeal fluid differs from all other serous fluids in being clear and watery, in its low specific gravity (1.004 to 1.007), in containing but a trace of protein which is chiefly globulin (with a trace of proteose (?) ), and 0.05-0.13 per cent, of a reducing substance that is probably glucose, ^^ which is decreased in 1° See Epstein, Jour. Exper. Med., 1914 (20), 334. 11 Lo Sperimentale, 1902 (56), 297. 12 Jour, pharm. et chim., 1906 (23), 239; also Compt. Rend. Soc. Biol., 1906 (60), 303. 13 Vecchi, Gaz. Med. Ital., 1912 (63), 211; Epstein, Jour. Exp. Med., 1914 (20), 344. 1^ Resume by Blumenthal, Ergeb. der Physiol., 1902 (1), 285; Blatters and Lederer, Jour. Amer. Med. Assoc, 1913 (60), 811; Herrick and Dannenberg, ibid., 1919 (73), 1321; Levinson, Amer. Jour. Dis. ChUd., 1919 (18), 568; Becht, Amer. Jour. Physiol., 1920 (51), 1. 1* Schloss and Schroeder, Amer. Jour. Dis. Child., 1916 (11), 1; Hopkins, Amer. Jour. Med. Sci., 1915 (150), 847. 360 EDEMA acute suppurative meningeal inflammation (Jacob). ^^ There is nor- mally in the adult from 60 to 150 cc, and Frazier estimates that from 360 to 720 cc. is secreted daily. HalHburton gives the following analyses of pathological accumulations of such fluids : Table VI (Spina bifida) Case 1 Case 2 Case 3 Water 989.75 989.877 991.658 Solids 10.25 10.123 8.342 Proteias 0.842 1.602 0.199 Salts 1 Q .^^ / 0.631 3.C28 Extractives/ ^■^'^^ \7.89f 5.115 The percentage of solids in spina bifida is thus a little higher than in normal meningeal fluids. In hydrocephalus the percentage of solids is rather greater, as seen in Table VII. Table VII (Hydrocephalus; Case 1 Case 2 Case 3 Water 986.78 984.59 980.77 Solids 13.22 15.41 19.23 Proteins and extractives 3.74 6.49 11.35 Salts 9.48 8.92 7.88 Normal cei'ebrospinal fluid seems to be h3'pertonic to the serum of the same animaP^ and slightly more alkaline than the blood. ^^ In meningitis the alkalinity is often lowered. ^^ The alkaU reserve is nearly constant in systemic diseases, except diabetes (McClendon),'° and is practically the same as that of the blood. By gas chain meas- urements Levinson-i found the spinal fluid almost neutral (pH = 7.4-7.6); in epidemic meningitis it is 7.3-7.4. According to Fuchs and Rosenthal-- the average freezing-point of the cerebrospinal fluid is lowered about the same in all diseases (A = —0.52° to 0.54°) ex- cept in tuberculous meningitis, where it is much less (average —0.43°). The amount of potassium is about the same as in the blood, -^ and not increased in degenerative diseases of the central nervous system;-* after death the amount is much increased by post-mortem changes. Calcium is almost constant at 5 mg. per 100 cc, or about one-half as much as in the plasma." In diseases associated with destruction of brain tissue, such as general paralysis and epilepsy,-^ choline or some '^ Brit. Med. Jour. 1912 Oct. 26. »' Ravaut, Presse rnod., 1900 (8;, 128; Zanier, Cent. f. Fhysiol., 1896 (10), 353. •8 Hurwitz and Tranter, Arch. Int. Med., 1916 (17), 828. >» Levinson, Arch. Pediatrics, 1916 (33), 241. 2" Jour. Amer. Med. Assoc, 1918 (70), 977. 2> .jour. Infect. Dis., 1917 (21), 556. " Wion. mcd. Presse, 1904 ^45), 2081 and 2135. "Myers, Jour. Biol. Cheni., 1909 (6), 115, literature. "^ Rosenbloom and Andrews, Arch. Int. Med., 1914 (14;, 536. "Halverson and Borfreini, Jour. Biol. Chem., 1917 (29), 337. 2* Concerning si)inai fluid in cpilcp.sy see Larkin and Cornwall (Jour. Lab. Clin. Med., 1919 (4), 352. MENINGEAL EFFUSIONS 361 other base" may be found in the spinal fluid. (See "Choline, " Chap. iv.) ^ Under pathological conditions the amount of protein varies greatly and to some extent characteristically. Thus, in syphilis the euglobulin is so greatly increased that it is readily identified by various precip- tation methods,-^ while in more acute inflammations fibrinogen ap- pears.-* According to Mott^'' the fluid is especially rich in nuclein in progressive paral^'sis, and lipoids are increased in the fluid in do- generations of the central nervous system. Pathological fluids show also specific alterations in their colloidal property of preventing pre- cipitation of colloidal suspensions b}' electrolytes (the "Goldzahl" of Zsigmondy).'^ The surface tension is higher than that of the serum and is not characteristically altered in disease. ''^ The increased organic matter of pathological fluids raises the permanganate reduction index." In epidemic meningitis there is more positively charged protein while in tuberculous meningitis there is more negatively charged protein, which can be distinguished by suitable precipitants (Tashiro and Levinson).^^ Cholesterol can be found in all cases of mental disease, the amount not bearing any relation to the type of ps3'chosis (Weston) ;^' ordinar- ily' 0.2 to 0.7 mg. per 100 c.c. is found. The changes in PoOs content in disease are doubtful,^^ while the amount of reducing substances is said to be increased in disease. ^'^ In general the inflammatory fluids in the spinal canal resemble exudates elsewhere, but usually the con- centration of the different components is relatively low, except the chlorides. ^^ Normal cerebrospinal fluid contains no antiprotease (for leucoprotease), as does the fluid in many cases of chronic inflamma- tions; in acute inflammation proteases ma}- appear (Dochez^*). Pep- tid-splitting enzymes are especially abundant in meningitis. ■*" Anti- bodies pass from the serum into the cerebrospinal fluid only in minimal amounts or not at all, except when inflammatory exudation occurs, and even then the antibody concentration is usually low,'*^ and even 2" Kaufmann, Zeit. physiol. Chem., 1910 (66), 3-13; Laignel-Lavastine and Lasusse, Compt. Rend. Soc. Biol, 1910 (68), 803. 28 See Xoguchi, Jour. Exp. Med., 1909 (11), 604. =' See Mestrezat, Rev. d. Med., 1910, p. 189; Kaflfka, Deut. med. Woch., 1913 (39), 1874. 3° Lancet. July 9, 1910. " Lange, Zeit. Chemother., 1912 (1), 44; Spat, Zeit. Immunitat., 1915 (23),'426; Vogel, Arch. Int. Med., 1918 (22), 496. 32 Ivisch and Remertz, Miinch. med. Woch., 1914 (20), 1097. " See Hoffman and Schwartz, Arch. Int. Med., 1916 (17), 293. 3* Jour. Infect. Di.s., 1917 (21), 571. 35 Jour. Med. Res., 1915 (33), 119. 3® Apelt and Schumm, Arch. Psj-chiat. u. Xervenkr., ISOS (44), 845. " Jacob, Brit. Med. Jour., Oct. 26, 1912. 3s Javal, Jour. phv.s. et path, gen., 1911 (15), 508. 39 Jour. Exp. Med., 1909 (11), 718. "> Major and Xobel, Arch. Int. Med., 1914 (14), 383. *iLemaire and Debre, Jour, physiol. et path, g^n., 1911 (13), 233. 362 EDEMA simple chemicals enter the normal spinal fluid but very httle,^- ex- cept perhaps alcohol. ^''^ According to Rosenbloom'*'* there is no crea- tin or creatinine. It contains normally from 2 to 4 mg. of amino-N per 100 c.c, or about half that in the blood, without definite changes in syphilis. ^^ There is almost the same amount of urea as in the serum of the same person, i. e., 20" to 42 mg. per 100 c.c.*® In uremia the non-protein constituents of the spinal fluid increase with those of the blood, but to a less degree. Substances giving the ninhydrin test appear in meningitis,'*^ but Rosenberg states that even with the highest indicanemia"*^ no indican is found in the spinal fluid. Sugar is present in from 0.07 to 0.085 per cent, and is not modified signifi- cantly in mental diseases;*^ it is reduced in meningitis but increased in uremia.^" There is only a very small amount of diastase, not bearing any constant relation to the cell count. '^^ Xanthochromia. — In cases of retention of spinal fluid, usually in a lumbar cul-de-sac, it may assume a yellow color although free from blood pigment, containing much globulin and coagulating spontane- ously (Froin's syndrome) . The color is apparently due to concentra- tion of plasma held for some time in the spinal canal and may be from bilirubin. ^^ Most usually this condition accompanies tumor of the spinal cord.^^ Wound secretions obtained from large aseptic wounds, mostly amputation stumps, have been studied by Lieblein.'''' The reaction is generally alkaline, globulin and albumin abundant, but fibrinogen scanty, total nitrogen being less than that of the blood and decreasing from day to day; the proportion of albumin increases and globulin decreases as healing progresses. Occasionally albumoses were found, but only on the first day in aseptic wounds; if found later, they gener- ally were antecedent to suppuration (concerning suppuration see "Inflammation, " Chap. xi). Blister fluid is generally rich in solids and proteins (40-65 p. m.). In a burn blister Morner^^ found 50.31 p.m. proteins, among which were 11.59 p. ni. globulin and but 0.11 p. m. fibrin; also a substance reducing copper oxide, but not pyro- catechin. By refractometric determinations the amount of protein in blister fluids is in direct proportion to the amount in the blood. ^'' Antibodies of all " See Rotky, Zeit. klin. Med., 1912 (75), 494. " Schottmiillerand Schumm, Neurol. Zbl., 1912 (31), 1020. " Biochem. Bull., 1916 (5), 22. « Ellis, elal, .Jour. Amer. Med. Assoc, 1915 (64), 126. « Ellis and Cullen, Jour. Biol. Chem., 1915 (20) 511. ^^ Nobel, Miinch. med. Woch., 1915 (62), 1355, 1786. ^» Berl. kl. Woch., 1916 (53), 1314. ^' Weston Jour. Med. Res., 1916 (35), 199; Kraus and Corneille, Jour. Lab. Clin. Med., 1916 (1), 685. " Leopold and Bernhard, Amer. Jour. Dis. Chil., 1917 (13), 34. Discussion of chemistry of spinal fluid in children. " Jicsclike and Pincussohn, Deut. med. Wochs., 1917 (43), 8; Katakura, Kyoto Jour. Med. Sci., 1916 (13), 1. " Bauer and Spiegel, Deut. Arch. klin. Med., 1919 (129), 18. 6» Review by Sprunt and Walker, Bull. Johns Hop. IIosp., 1917 (28), 80; Elsberg and Rochfort, Jour. Amer. Med. Assoc, 1917 (68), 1S02. " Beit. klin. Chir., 1902 (35), 43. " Hammarsten, "Physiological Chemistry." 59 Engel and Orszag, Zeit. klin. Med., 1909 (67), 175. CHYLOUS EFFUSIONS 363 sorts seem to pass readily into blister fluids'' although the complement-fixation reaction is not so stroiiij as with the blood."* Hydrops of Gall Bladder. — The watery fltiid contains 99 per cent, water, a mucin-like substance, but no otiier proteins and no bile acids." Fetal Bronchiectasis. — The fluid resembles closely liquor amnii.'" Chylous Effusions.^' — Fat may be present in effusions in sufficient quantity to cause a milky appearance, either from escape of chyle from a ruptured or obstructed thoracic duct, or through fatty degen- eration of the cells in the effusion or the lining of the walls of the cavity. The former are designated as chylous, the others as chyU- form or adipose fluids, but it is not always easy to distinguish be- tween them. The composition of the fluids in true chj'lous exudates will var}' according to the food taken and the amount* of fat the food contains, and will resemble the composition of chyle, except to the extent that it is modified by the effusion or absorption going on in the cavity. They are characterized by strong bactericidal powers as evi- denced by lack of putrefaction after long standing. Analyses of human chyle are scanty. Panzer^^ found 90.29-94.53 per cent, water; 5.47-9.71 per cent, solids; 0.80-1.04 per cent, inorganic salts; 2.16 per cent, coagulable protein; 6.59 per cent, ether-soluble material; also diastatic enzyme, soaps, and occasionally traces of cholesterol, lecithin, and sugar. Carlier,^^ in a specimen from a child, obtained very similar results, except that the salts were much less abundant. The proteins and fats vary greatly with the diet; thus Sollmann^^ found variations in the proteins from 1.85 to 6.5 per cent. Edwards^'^ found that of 31 definitely estabUshed cases of chylous or chyliform ascites studied at autopsy, in 21 there was established the existence of a rupture in the thoracic duct or lacteals. Boston^^ in 1905 was able to collect 126 cases, including both chylous and chyliform ascites, and notes an associated eosinophilia in a case studied b}- him. Chylous ascites fluid often, but not always contains sugar," but it maj^ disappear after having once been present; the amount of fat is small, usually about 1 per cent., and the fluid is rich in sohds. If due to a ruptured thoracic duct, it may be possible to detect special fats taken in the food, e. g., butter-fats (Straus).®^ The reaction is " Eisenberg, Deut. med. Woch., 1909 (35), 613. 58 Buschke and Zimmermann, Med. Ivlinik, 1913 (9), 1082. 59Sjoquist, SvenskaLak. Handl., 1916 (42), 1291. «o Koeckert, Amer. .Jour. Dis. Chil., 1919 (17), 95. " Literature bv Gandin, Ergeb. inn. Med., 1913 (12), 218. " Zeit. phvsiol. Chem., 1900 (30), 113. " British xMed. Jour., 1902 (ii), 175. «< -\mer. Jour. Phvsiol., 1907 (17), 487; see also Hamill, Jour. Physiol., 1906 (35), 151. "Medicine, 1895 (1), 257; also see "Chem. u. morph. Eigenschaften fett- haltige Exsudaten," St. Mutermilch, Warschau, 1903; Comey and McKibben, Boston Med. and Surg. Jour., 1903 (148), 109. « Jour. .'^ler. Med. Assoc, 1905 (44), 513. «' For example, v. Tabora (Deut. med. Woch., 1904 (30), 1595) found as high as 0.864 per cent, of sugar in a typical case. " Arch. Physiol, et Pathol., 1886 (Ser. 3, vol. 8), 367. 364 EDEMA usually alkaline or neutral, and some specimens coagulate spontane- ously. Specific gravity varies from 1.007 to 1.040, the average being about 1.017. Perhaps the most important characteristic is the varia- tion produced by changes in diet.^^ Zdarek'^" found in a chyle-cyst 2.7 per cent, of fats, 7.2 per cent, of proteins, and 0.05 per cent, of sugar; feeding of fats increased their amount in the cyst and star- vation decreased it. Schumm''^ found in the solids of such a cyst 35.76 per cent, of fat, some of which was in the form of calcium soap. Chyloihorax fluid is, of course, quite similar to that of chylous ascites. Thus, Buchtala^^ found 91.34 per cent, of water, 8.66 per cent, solid, 4.86 per cent, protein, 2.5 per cent, fat, 0.26 per cent, cholesterol, and 0.94 per cent. ash. Similar figures were obtained by Salkowski^^ and others. Chyluria,''* which seems to depend upon an abnormal communication between the lymphatics of the receptaculum chyli and the kidney,^^ shows no particular chemical features beyond those of an admixture of a considerable amount (100 to 1000 c.c. per day) of chyle with the urine. Carter"' found the amount of fat in the urine to rise with increase of fat in the food. Pecker^^ observed a rise from a former average of 1.5 gm. fat per liter to 9.75 gm. after eating oils and milk. In some cases chyle escapes directly into the bladder or ureter from the lymphatics, in others the fat may be excreted directly from the blood, independent of lymphatic abnormality; in some cases the fluid entering the urine is true chyle and in others it is lymph. Ascites adiposus is characterized by the absence of sugar and by a higher percentage of fat, the maximum observed being 6.4 per cent. It is ascribed to fatty metamorphosis of cells, particularly in carcinomatous and tuberculous exu- dates; Edwards was able to show experimentally that a transudate may change from serous to cellular, and later come to contain fat. Pseudochylous effusions are also observed, not only in the abdominal and thoracic cavities, but even in the fluid of the edematous legs and scrotum ; these resemble chylous fluids in being turbid or milky, but are said to contain little or no fat. The turbidity is ascribed chiefly to lecithin, which is largely combined with the pseudoglobulin of the fluid (.Toachim).'^ Possibly in some cases the turbidity is partly or largely (Poljakoff)'* due to poorly dissolved proteins. Strauss*" has noted the occurrence of this form of ascites particularly in chronic parenchy- matous nephritis, but believes the turbidity has a local origin. Hammarsten has observed turbidity due to mucoid substances, as also have Gouraud and Corset.*' The pseudo-chylous effusions have a lower freezing point, a lower specific gravity, ^^ A sample of the composition of 1 liter of chylous ascitic fluid is shown by the analysis in the case studied by Comey and McKibben {loc. cit) : Specific gravity, 1.010; solids, 21 gm.; protein 9.75 gm.; urea, 1.28 gm.; fat, 1.45 gm.; inorganic matter, 8 gm.; peptone (?) and sugar, present; fibrinogen, mucin, nucleo-albumin, and uric acid absent. '« Zeit. f. Heilk., 1906(27),!. '1 Zeit. physiol. Chem., 1906 (49), 266. ^' Zeit. physiol. Chem., 1910 (67), 42. " Virchow's Arch., 1909 (198), 189; also Tuley and Graves, Jour. Amer. Med Assoc, 1916 (66), 1844; Patein, Jour, pharm. Chim., 1915 (11), 265. ''* Review of literature by Sancs and Kahn, Arch. Int. Med., 1916 (17), 181. " See Magnus-Levy, Zeit. klin. Med., 1908 (66), 482. ""^ Arch. Int. Med., 1916 (18), 541. "Jour, pharm. chim., 1917 (16), 139. See also Patein ibid., 1917 (16), 230 78Muncli. mod. Woch., 1903 (50), 1915; also Christen. Cent. f. inn. Med. 1905 (26), 329; Wallis and Scholberg, ()uart. Jour. Med., 191f (3), 301; 1911 (4), 153. '» Fortschr. d. Med., 1903 (21;, 1081; also Haushalter, Compt. Rend. Soc. Biol., 1910 (68), 550. 8« Note to Poljakoff's article;" also Biochem. Centr., 1903 (1), 437. 8' Coin))!. Rend. Soo. Biol., 1906 (60), 23. PNEUMOTHORAX 365 lower fat and greater lecithin content than typical chylous ascites. Gandin," however, questions the possibility of always differentiating the three types of turbid fluids as above indicated. Collecting all the recorded analyses in the literature he finds wide discrepancies, as indicated in the following table: (The maximum and minimum percentage figures are given for each component determined quantita- tively, with the average in parentheses.) Adipose Chylous (Chyliform) Pseudochylous Ether extract 0.065-9.2(1.65) 0.1-4.3(1.15) 0.007-1.86(0.25) Cholesterol + in 7, — in 2 + in 4 + in 3, — in 2 Lecithin -j- in 4, — in 1 + in 3 + in 20, — in 2 Sugar 4- in 46, — in 28 -f in 1, — in 4 + in 15, — in 14 Dry residue 3.1-10.6(6.2) 1.6-11.7(5.1) 1.2-7.6(2.9) Protein 0.9-7.7(3.5) 0.6-6.8(3.0) 0.1-4.2(1.4)' "Pepton" + in 6, -in 4 + in 1, - in 2 + in 1, - in 5 Ash 0.1-1.0 (0.59) 0.45-1.03 (0.65) 0.49-0.90 (0.73) It is quite evident that although the pseudochylous fluids usually contain little fat, they often contain more than the minimal content found in the other forms. Each type of fluid overlaps the others in one respect or another. Gandin states that to produce a turbid fluid but 0.01-0.1 per cent, of finely emubionized fat is necessary, and he believes that milky fluids always mean admixture of chyle, rejecting the terms pseudochylous and chyliform as unwarranted. He admits that fluids may contain droplets of fats not emulsionized, and hence not milky, which may be properly called adipose fluids. There are no characteristic chemical differences in the fats extracted from the different types of fluids. Chemistry of Pneumothorax In connection with the subject of exudates the above topic may appropriately be considered. The composition of the gases found in the pleural cavity in pneu- mothorax will necessarily vary greatly according to the cause. If the pleural cavity is in free communication with the exterior, the gas will be simply slightly modified air; for example, Ewald^^ found the following proportions in the gases in such a pneumothorax: CO2, 1.76 per cent.; O, 18.93 per cent.; and 79.31 per cent. N. Here the proportion of CO2 is even a little less than in ordinary expired air, which contains 3.3-3.5 per cent. When air .enters a closed pleural cavity and no effusion follows, it is slowly absorbed until a mixture of about 90 per cent. N, 4 per cent. O and 6 per cent. CO2 results; but if there is a serous effusion the oxygen disappears nearly or quite completely (Tobiesen).*^ In a seropneumothorax Ewald found 8.13 per cent, of CO2, 1.26 per cent, of O, and 90.61 per cent, of N, which is quite similar to the proportions of the gases in dry pneumothorax. Puru- lent pneumothorax generally shows more CO2 than the serous form, the average in the former being 15-20 per cent., in the latter 7.5-11.5 per cent. The average of the analyses in six cases of pyopneumothorax is given by Ewald as 18.13 per cent. CO2, 2.6 per cent. O, and 79.81 per cent. N. In open pyopneumothorax the gas approaches more closely the composition of air, but usually shows a slight excess of C()2; it is thus possible by a determination of the carbon dio.xide to determine quite accurately whether a given pneumothorax is in communication with the outside air. The transformation of a purulent into a putrid pneumothorax is accompanied by an increase of CO3, even as high as 40 per cent, having been found. The products of decomposition by the putrefactive saprophj'tes also are present, one analysis having shown 4.3 per cent, of hydrogen, 6.25 per cent, of methane, and traces of hydrogen sulphide. Infection of a pleural effusion by gas-producing organisms may also convert it into a pneumothorax, although this is not a common occurrence. The gases then 82 Complete literature and resume given by Clemens, in Ott's "Chem. Path der Tuberculose," Berlin, 1903, p. 406. " Beitr., z. lOin. d. Tuberk., 1911 (19), 451; 1911 (21), 109; Deut. Arch. klin. Med., 1914 (115), 399. 366 EDEMA present are the same as the organisms produce in similar culture-media, modified somewhat by absorption. The anaerobic gas-producing organisms have been found as the cause of such gaseous accumulations; it is questionable if the ordinary pathogenic organisms can cause a pneumothorax, since they are for the most part not capable of producing gas. The colon bacillus produces gas in sugar-containing media, but the amount of sugar in the pathological exudates is too small to'yield any considerable amount of gas; an exception is the pleural effusion in diabetes, and pneumothorax from infection of the pleural effusion in a diabetic by B. coli has been reported. Complete quantitative analyses of the gas in this form of pneu- mothorax seem not to have been made, but May found about 20 per cent, of CO2. The combustibility of the gas has i^frequently been noted, and is probably dueto hydrogen and methane. CHAPTER XV RETROGRESSIVE CHANGES (NECROSIS, GANGRENE, RIGOR MORTIS, PARENCHYMATOUS DEGENERATION) NECROSIS We recognize that a cell is alive through its reproducing, func- tioning, and its taking on and utilizing nutritive substances; yet at the same time we appreciate that a cell may do none of these things and still be alive. For example, a bacterial spore is quite inert physi- cally, and exhibits no chemical activity, yet it is by no means dead, since it still possesses the latent power to assume again an active exist- ence under suitable conditions. In pathological conditions we are accustomed to recognize the fact that a cell is dead by certain altera- tions in its structural appearance, particularly disintegrative changes in the nucleus; but this is exactly equivalent to recognizing that an animal is dead by the appearance of postmortem decomposition, for most of the characteristic histological changes of necrosis are merely postmortem changes in the cell. A cell may be dead and show ab- solutely none of these microscopic disintegrative changes, either because it has not been dead long enough for them to have taken place, or because the changes have been prevented by some means, just as we can prevent the appearance of postmortem decomposition by embalm- ing. For example, if we examine microscopically the mucous mem- brane of the stomach of a person who has died immediately after taking a large quantity of carbolic acid, although to the naked eye this mucous membrane is hard, white, and definitely necrotic, yet we find the histological picture presented by the cells almost absolutely un- changed from the normal. The cells are dead, but they have been so "fixed " that postmortem changes could not affect their structure. All cells examined by ordinary histological methods are, of course, dead — killed by the fixing agents outside of the body, in the same way that the carbolic acid fixes them within the body. It is evident, therefore, that it may be very difficult to determine always whether a cell is dead or not. Part of the difficulty, perhaps, hes in our failure to appreciate that not all parts of a cell die at the same time; i. e., the different chemical processes of the cell depend on its different intracell- ular enzymes, and these are not necessarily destroyed alike by the same agents. Even considerable respiratory activity may be ex- hibited by cells that have been killed.^" We recognize that after an animal is dead as a whole the various cells of its body do not die for some time as shown by the following i^See Haas, Bot. Gazette, 1919 (67), 347. 367 368 RETROGRESSIVE CHANGES examples: (1) We can cause the heart to beat for a considerable period after its removal from the body; (2) if we perfuse a mixture of glycocoll and benzoic acid through the kidney of a recentl}' killed animal, synthesis of these substances into hippuric acid will occur; and (3) the epithelium of the skin can be removed from the body of an animal long after death and transplanted successfully on another animal. So, too, in ordinary cell death (necrobiosis) not all the enzymes are destroyed together. When all are destroj'-ed at once, as by strong chemicals or by heat, the customary disintegrative changes do not take place. If, however, not all the enzymes are thrown out of function, then the others may be able to act, producing the disintegrative changes by which histologists ordinarily recognize cell death. These disintegrative changes are, for the most part, ap- parently brought about by the intracellular proteases, that is, through autolysis. This may be shown as follows:^ If we take two pieces of fresh normal tissues from an animal, and in one kill the enzymes by heating to 100° C, then implant both aseptically into the abdom- inal cavity of an animal of the same species, it will be found that the changes that follow in the two will be very unlike. In the un- heated tissue nuclear changes soon occur, so that they lose their ca- pacity for taking up basic stains, the cytoplasm becomes granular and fragmented, the tissue becomes friable so that it is difficult to secure good sections, and the changes are in general similar to those seen in areas of necrosis. The boiled tissue, on the other hand, retains its capacity for nuclear staining for months, except at the periphery, where it is slowly attacked by leucocytes and the enzymes of the blood plasma. Therefore it would seem that the characteristic changes of necrosis depend chiefly upon the intracellular enzymes, rather than upon the infiltrating plasma as Weigert^ and other early writers imagined. In areas of anemic necrosis (see "Infarcts") we have another case, in which the oxidizing enzymes are thrown out of function through lack of oxygen, while the other enzymes are, presumably, at first unaffected. From studies of infarcts it would seem that the intracellular proteases bring about the subsequent nuclear and cytoplasmic alterations, but that the eventual digestion of the area is accomplished by the invading leucocytes working slowly inward from the periphery. Apparently when the supply of materials from outside ceases, and when the oxidation processes of the cells no longer accomplish necessary steps of synthetic reactions or destroy products of protein catabolism, the proteases continue to split proteins without the balancing by the above-mentioned factors, with a resulting disintegration of the cells. Karyolysis and karyorrhexis are,"then, the result of an autolytic process, which is perhaps due to intracellular proteases that act spe- 1 Wells, Jour. Med. Research, 1906, (15),' 149. 2 (Jcnt. f. Path., 1891 (2), 785. NECROSIS 369 cifically on nucleoproteins, and w hich may l)e designated as nucleases.^ Nuclear staining by the usual methods depends upon an affinity of the acid nucleoproteins (in which the nucleic acid is not completely saturated by proteins) for basic dyes. Presumably in karyolysis the first step consists in a splitting of the nucleoprotein of the chromatin into nucleic acid and protein; this can be accomplished, according to Sachs, by the ordinary trypsin, and presumably, therefore, by the trypsin-like enzymes of the cell. Corresponding with this change we should expect the free nucleic acid to give an intense staining with basic stains, and this has frequently been described by those who have studied the cytological changes in anemic necrosis,^ and called pycnosis. As supporting this view still further may be quoted Arnheim's^ observation that in alkaline solutions the nucleus soon stains diffusely and weakly, and not at all after twelve to eighteen hours; this is to he explained by the fact that nucleic acid is both dissolved and neutralized by alkaline solutions. Acids developed in injured cells may, by combining with the basic elements of the nu- cleoproteins, render them still more acid and highly basophilic; thus, in muscles showing waxy degeneration from accumulation of lactic acid the muscle nuclei will be found pycnotio (see waxy degenera- tion). After the nucleic acid has been freed from the protein by the autolytic enzymes, it is still further decomposed by the "nu- clease" or similar intracellular enzymes that have the property of splitting nucleic acid into the purine bases that compose it — cor- responding with this change the hyperchromatio nucleus loses its affinity for stains, and karyolysis is complete. When extensive ne- crosis occurs there will result, therefore, an increased elimination of purines, as was found by Jackson and Pearce^ in animals with severe hepatic necrosis from hemotoxic serum. A careful analytical study of the changes taking place in the autolyzing spleen, for the purpose of correlating the chemical and microscopical changes, has been made by Corper,^ which corroborates the interpretation of necrosis advanced above. He found that during the stage when pycnosis is the chief feature there is no appreciable change in the nucleus; that is, the nucleic acid has not been split into free purines and the rest of its components; at this stage but little change has occurred in the lecithin, and a very slight amount of proteolysis is demonstrable. During the stage of karyorrhexis and karyolysis the most active disintegration is taking place, alDOut one-fourth of the nucleic acid becoming dis- integrated by the time all nuclear structures have disappeared; in the same period nearly half the lecithin (phosphatids) is hydrolyzed, while about one- fourth the coagulable protein has been hydrolyzed into non-coagulable compounds. After this stage the changes are very slow. It is somewhat surprising to find that when no vestige of nuclear substance remains in stainable form, there still remains three-fourths of the nucleic acid in an intact condition. Corper publishes ^ See Purine Metabolism, Chap, xxiii. ■* Schmaus and Albrecht, Virchow's Arch., 1895 (138), supp., p. 1; Ergeb. allg. Pathol., 1896 (3), 486 (literature). * Virchow's Arch., 1890 (120), 367. « Jour. Exper. Med., 1907 (9), 569. ' Jour. Exper. Med., 1912 (15), 429. 24 370 RETROGRESSIVE CHANGES a series of plates, together with the chemical details' thus establishing a standard whereby the histological changes can be interpreted in terms of the chemical changes which cause them. Autolysis of asepfcically preserved tissues outside the body is much more rapid than is the autolysis of infarcts and similar aseptic necrotic areas within the body. This may be due to either or both of two factors:^ First, autolysis is much slower in alkaline than in acid media; outside the body autolyzing tissues develop an acid reaction which favors their autolysis; within the body this is checked by the plasma. Second, the plasma contains inhibiting substances, which also may interfere with self -digestion in the body. In corroboration of the above may be recalled the fact that large necrotic areas show autolysis first in the center, where the alkaline, antagonistic body fluids presumably cause the least effect. Furthermore, it has been found by Wells^ that the histological changes of autolysis proceed much faster in tissues placed in serum that has been heated to destroy the antibodies than in unheated serum. Leucocytes, as Opie has shown, contain autolytic enzymes acting best in an alkaline medium, hence they perform their digestive function readily at the periphery of necrotic areas, and coagulated tissue proteins, when acted upon by body fluids, produce chemotactic substances which attract leucocytes to dead areas. ^° When a cell dies, certain physical changes occur that are probably of considerable importance. Bechhold says: "With the occurrence of death, protoplasm gelatinizes, Brownian movement of the smaller particles ceases, and the structure of the gel appears in the ultramicro- scope as a conglomeration of many reflecting platelets. It makes a substantial difference whether the protoplasm slowly dies or is suddenly killed by a fixative (alcohol, formalin, etc.). In the first instance there is a precipitation (flocculation), whereas, in the latter there is a stiffening; this difference may be readily recognized under the ultramicroscope." The permeability of the cell wall is almost immediately increased, so that all diffusible substances readily pass through, i. e., its semi- permeable character is lost. This we see particularly in plant cells, which lose their turgor with their semipermeability, and therefore the plant wilts. The cell structure is also disintegrated, and as a result coordination of the cell chemistry is at once destroyed. ^^ In- tracellular enzymes escape into the blood from areas of local death of cells, ^2 or as an agonal manifestation in general death. ^^ Various * Literature and more complete discussion under "Autolysis." 9 Jour. Med. Research, 190G (15), 149. 10 Burger and Dold, Zeit. Immunitiit., 1914 (21), 378. " See V. Prowazek, Biol. Centrbl., 1909 (29), 291. Pictet suggests that in dead proteins, aldehydes and amino radicals unite with one another to form cyclic compounds (Arch. sci. phys. nat., 1915 (40), 181). 12 Mandelbaum, Munch, med. Woch., 1914 (()1), 461. 13 Schultz, Miinch. med. Woch., 1913 (GO), 2512. NECROSIS 371 dyes which cannot penetrate hvinj; cells may stain dead or dying cells.'* These changes depend on alterations in permeability, and as permea- bility determines electrical resistance, Osterhout has used the resistance of plant cells as an indicator of vitality. He finds that normal cells have a rather constant resistance, which is reduced by anything that lowers the vitality' of the cell, and in direct proportion to the degree of injur}' or loss of vitality.'^ The temperature coefficient is also considerablj' lower in dead than in living tissue. ^^ When secondarj' disintegrative changes occur in the protoplasm, with the formation of many small molecules from the large molecules of the cell, both osmo- tic pressure and electrical conductivity increase rapidly. Changes in the permeability of cell protoplasm, however, may be of considerable degree without necessarily^ indicating serious injury of the cells (Oster- hout).^^ Death is accompanied by changes of the character of a monomolecular reaction, which is continually going on and w'hich is accelerated by the toxic agent. '^ Up to a certain point the reaction seems to be reversible. A principle of colloid chemistry, the alteration of colloids with time, has an interesting bearing on the question of aging and natural death of tissues.'*" It is characteristic of colloidal solutions (which, of course, is what cells are), that they continuously change in their properties, the change being generally in the direction of aggregation of the disperse colloidal particles, with a resulting ten- dency to precipitation or coagulation; the gels tend to decrease in elasticity and to become more turbid, associated with which are alterations in their perme- ability to crystalloids. A gelatin mass possesses its maximum elasticity three or four hours after it is first formed; and crystalloids penetrate fresh, quickly- formed gels at first more rapidly than later. As Bechhold says, we can imagine (1) a relation of such facts to the greater elasticity of young tissues; (2) to a pre- sumably greater permeability for crystalloids and hence more rapid metabolism; (3) to the decreasing water of the tissue with age (94 per cent, of water in the fetus of three months, 69-66 per cent, at birth, and 58 per cent, in adults); (4) to the demonstrated greater permeability of young nerve tissues for vital stains, etc. "In general we can say that the tissue colloids decrease in their water affinity (Quellbarkeit) both in animal organisms, which become poorer in water with age, and in plants, as shown by the hardening of older plant tissues." The bearing of these principles on the problem of senility and degeneration of elastic tissue, regeneration and many other subjects is obvious. Causes of Necrosis Anemia. — After the cutting off of blood-supply, cells soon undergo morphological changes that we recognize as indicating their death, and after a time they also become incapable of returning to their normal condition when the blood-supply is re-established, probably because of these structural changes. In just what way lack of nourishment " See Steckelmacher, Beitr. path. Anat., 1913 (57), 314. 15 See Science, 1914 (40), 488. 1^ Galeotti's earher observations with animal tissues (Zeit. f. Biol., 1903 (45), 65) do not harmonize with Osterhout'fe results, and Galeotti's idea that there is a special degree of ionization characteristic of living cells is not established. 1^ Botan. Gaz., 1915 (59), 242. 18 Osterhout, Jour. Biol. Chem., 1917 (31;, 585. 18" See H. Bechhold, "Die Kolloide in Biologic und Medizin," Dresden, 1912, „• 65. "372 RETROGRESSIVE CHANGES causes death has not been determined, but, as has been before suggested, it seems probable that it is because cataboUc processes are no longer balanced by anabolic processes, and with these latter oxi- dizing enzymes seem to be inseparab'y associated as far as our pres- ent knowledge shows us. That the loss of oxygen alone, with other materials presumably supplied to the cells in adequate amount, may cause necrosis, is shown by the presence of marked hepatic necrosis in animals kept a week in atmospheres extremely low in oxygen (5-9 per cent.)-^^ The nature of the chemical changes taking place in a cell when oxj^gen is deficient must be very different from the normal changes, and hence abnormal toxic substances may accumulate, e. g., excessive amounts of organic acids. Were it not that the proteolytic enzymes continue in action after nutrition is shut off, the cells might remain in a completely unaltered condition for an indefinite period, and capable of resuming their function when nourishment is again supplied, which is decidedly contrary to the facts. (The general features of anemic necrosis have been already discussed in the pre- ceding paragraphs, and also under the subject of infarction.) Thermic Alterations. — These have been studied particularly in connection with the cells of the lower organisms.^*' While some uni- cellular organisms can survive a temperature of 69°, most of them are killed at from 40°-45°. For the great majority of me+azoa the maximum temperature lies below 45°, and in the case of marine species below 40°.^^ The heating is accompanied by the appearance of granules in the cytoplasm, which become larger until the condi- tion of "heat rigor" sets in. Kiihne, in 1864, showed that in muscle cells, at least, there is contained a protein which becomes turbid through partial coagulation at 40°, and Halliburton-^ has found that in nearly all tissues are globulins coagulating at from 45°-50°; it is probable, therefore, that the granules formed in heated cells are produced through coagulation of these proteins. The importance of this coagulation in determining death is not yet fully established, but it would seem to be very great. Halliburton has observed that in both muscles and nerves to which heat is applied, contractions occur at various temperatures, corresponding exactly with the tem- peratures at which the several varieties of the proteins of the cells coagulate. Furthermore, Mott-^ has found that the temperature that is immediatel}'' fatal to mammals (47°) is exactly the same as the coagulating temperature of the lowest coagulating protein of 19 Martin, Loevenhart and Bunting, Jour. Exp. Med., 191S (27), 399. ^"Literature, see Davenport, "Experimental Morphology," New York, 1S97; Schmaus and Albrecht, Ergebnisse dcr Pathol., 1890 (."3, Abt. 1), 470. ^' Tlie adaptation of animal cells to high tcniperaturos i.s an interesting topic, especially in view of such results as those of Dallingor, who, by raising the tem- perature gradually during several years, caused flagollatn, with a normal maxinuuu of about 21° -23° to become capable of living at 70° (see Davenpotr.). " "Biochemistry of Muscle and Nerve," Phila., 1904. 2^ Quoted by Halliburton. NECROSIS 373 norve-cells. This fact is undoubtedly of great practical importance in causing death from fever, for although 47° C. (117° F.) is prob- ably never reached in man, yet application of much lower tempera- tures, even 42° (108° F.), for a few hours will cause coagulation of these proteins (all proteins coagulate at less than their ordinary coagulation point if the heating is continued for a long time). It would seem from the above observation that heat may cause cell death through coagulation of the proteins. Whether the cell death is in any way dependent upon destruction of the enzymes by heat has not l)een ascertained; but as most enzymes are not destroyed nmch be- low 60°-70°, it seems impro})able that they are greatly injured at the temperatures at which cells are killed. It is possible, however, that under the conditions in which enzymes exist in the cell they may be more susceptible to heat than under other conditions. Just how coagulation of cell globulins can determine the death of a cell is difficult to understand, unless the physical conditions of the cell are greatly altered thereby. Ordinarily we have in the cell an equi- librium between colloids in solution and colloids in the solid or gel state; if the colloids are rendered insoluble by heat, or by any other cause, so that this equilibrium is destroyed, serious alterations in the mechanism of all metabolism must result (Mathews). Other chem- ical reactions will also have their point of equihbrium altered by changes in temperature, and such alterations might well have disas- trous results. Different tissues show unequal susceptibility to heat. Werhov- sky^^ found the blood most affected by raising the temperature of living animals, next the liver, kidneys, and myocardium in order, the other tissues being little or not at all structurally injured. Ani- mals exposed to heat show a fall in the leucocyte count, followed by a rise in lymphocytes which persists; there is an extensive degeneration of cells in the spleen and lymph glands, followed by marked mitotic proliferation in the germinal centers. ^^ Cold^^" is well withstood by unicellular forms, and relatively poorly by more complex organisms, particularly by those with a highly de- veloped circulatory system; this is because individual cells are not greatly affected by freezing, whereas the circulatory channels are readily blocked by this cause. Bacterial cells are not killed by ex- posure for long periods to the temperature of liquid air^^ (—190°). Reduction of the temperature of plant cells to —13° may result in a granular transformation of the cytoplasm, often with rather seri- ous structural alterations. Cytoplasm seems to be more affected than the nucleus, for mitosis may occur slowly in plant cells at —8°, 2* Ziegler's Beitr., 18P5 (18), 72. " Murphy and Sturm, Jour. Exp. Med., 1919 (29),- 1. -^* Sy:^temic rffectsof cold reviewed by Foord, Jour. Infect. Dis., 1918(23), 159. 2 MacFadyrn, Lan et, 1900 (i), 849. 374 RETROGRESSIVE CHANGES and Uschinsky^^ noted that in animal tissues the nuclei were less af- fected b}^ cold than the cytoplasm. Blood seems little affected by freezing temperature, for du Cornu found that dog's blood kept on ice for five to ten days could be employed for transfusion without causing hemoglobinuria. Grawitz saw motion persist in human cili- ated epithelium kept for seven to nine days on ice. Cihated epi- thelium from the mouth of the frog may survive cooling to —90° and frog eggs are not killed by —60°. In many cells, however, the physical changes produced by freezing, and also by the subsequent thawing, are sufficient to render them incapable of further existence.'^* Cells devoid of or poor in water cannot be killed by freezing, hence it is probable that the currents set up about the crystals of ice in thawing, as well as the rapid contraction and expansion under the influence of the cold and the ice formation, are the cause of the effects of freezing, which, therefore, are not dependent upon chemical, but upon physical, alterations. In the case of warm-blooded animals, the gangrene following freez- ing depends not so much upon the freezing of the cells themselves as upon the formation of hyalin thrombi in the injured vessels (v. Reck-j linghausen, Hodara).^^ Kriege^'' found that if the freezing is transi- tory, the thrombi may again disappear; if over two hours in duration, they are persistent. Rischpler,^^ however, considers that cell death is due primarily to the effect of the cold upon the cells, and Lake^- found that for both isolated cells in culture and living tissues with intact blood supply, deai.b occurred at —6° C, this being the tempera- ture at which protoplasm freezes. On the other hand, Steckel- macher^^" found that freezing of liver tissue produced the same changes as ligation of the hepatic artery, i. e., increased permeabiHty of the cell wall followed by similar changes in the nucleus, suggesting that the changes produced by freezing depend on the vascular changes. Light.^'^ — ^Light may affect tissues seriously, apart from the effects of accompanying heat, although the experiments of Aron^^ indicate that insolation does not depend on the light raj^s, but solely on the heat. In the treatment of lupus by the Finsen method with concen- trated light rays, the action is largely a stimulating one, but associ- ated with or subsequent to a certain degree of cell injury. Ogneff^^ " Ziegler's Beitr., 1893 (12), 115. ^8 In plant cells it is the freezing and not the thawing that causes the harm (Maximow, Berichte Deut. Bot. Gesell., 1912 (30), 50-4). 29 Miinch. med. Woch., 1896 (43), 341. 30 Virchow's Arcli., 1889 (116), 64. " Ziegler's Beitr., 1900 (28), 541. 32 Lancet, Oct. 13, 1917. 32" Beitr. path. Anat., 1913 (57), 314. " Review by Bering, Ergeb. allg. Pathol., 1914, Abt. 1 (17), 790. See dis- cussion of the principles of the action of light on tissues by Bovie, Anier. Jour. Tropical Dis., 1915 (2), 506. •"• PliiU])i)ine Jour. Sci., B, 1911 (6), 101. a' Pfiiiger's Arch., 1896 (63), 209. NECROSIS 375 fouiul that inodcmte action of electric light, rich in violet and ultra- violet raj's, causes mitotic cell division; if the action is stronger, the cells undergo amitotic division and then become necrotic. Blue rays have but slight cytotoxic action, and rays further towards the red end of the spectrum are without demonstrable cffec '.. Light baths are said by Oerum"' to increase greatly the quantity of corpuscles and hemo- globin, while residence in the dark reduces these elements. The de- struction of bacteria by light is a well-known phenomenon, ^^ but it has been suggested that their destruction depends rather upon the action of substances produced in the culture-medium under the influ- ence of light than upon the effect of the hght upon the bacterial cells themselves. In view of the fact that enzymes and antibodies in solu- tion are quite readily weakened or destroyed by the action of light, it is possible that intracellular enzymes may be similarly destroyed by light, with resulting cell death. However, in the case of bacteria, at least, the effects of light seem to depend upon oxidation processes, for in the absence of oxygen, bacteria are not seriously injured by light, and D'Arc}^ and Hardy^* found that "active oxygen" is formed by the same portion of the spectrum that is most active in destroying bacteria. ^^ Light may also alter the solubility of cell proteins, espe- cially in the presence of various organic and inorganic substances that act as sensitizers, such as silica"! es, sugar, lactic acid or urea.^*^ In this may lie the cause of cataract, especially diabetic cataract. The general effect of light acting on organic substances present in plant and animal cells, is to produce from carbonyl-containing materi- als aldehyde or ketone compounds, whose reactivity and availability for important synthetic changes are conspicuous (Neuberg).*^ Whether oxidative processes are the cause of death in animal cells is not known, but we are familiar with many chemical reactions of vari- ous sorts that are initiated or checked by the action of light. ^^ Thus, bilirubin is oxidized into biliverdin, when acted upon by sunlight, even when not in contact with air; many vegetable oils are oxidized by sunlight, and it is probable that the oxidizing action of light upon organic compounds is of wide-spread occurrence. It is, therefore, quite possible that such oxidative changes may be the cause of necrosis produced by the action of light rays, especially as Bering'*^ has found that chemically active light rays have a direct action on oxidizing enzymes. 36 Pfluger's Arch., 1906 (114), 1. 3' Literature given by Wiesner, Arch. f«Hyg., 1907 (61), 1. 38 Jour, of Physiol., 1895 (17), 390. 33 See also Agulhon, who found that ultraviolet rays may attack enzymes to some extent in the absence of oxygen (Ann. Inst. Pasteur., 1912 (26), 38). "Schanz, Biochem. Zeit., 1915 (71), 406; Arch. Ophthal., 1918 (96), 172; Burge, Amer. Jour. Physiol., 1916 (39), 335; Neuberg and Schwarz, Berl. klin. Woch^ 1917 (54), 84. " Biochem. Jour., 1908 (13), 305. "See Davenport, "Experimental Morphology," 1897, p. 162. " Miinch. med. Woch., 1912 (59), 2795. 376 RETROGRESSIVE CHANGES It is yery probable that not all of the effects of exposure to the sun depend upon the heat rays, for there is evidence that the hght rays may also produce effects. This is definitely true in the case of indi- viduals or animals with certain pigments in their blood, notablj^ hematoporphyrin (q. v.). In them, not only may skin eruptions re- sult from relatively small exposure to light, but mice may be so sen- sitized that a few moments of exposure to light is fatal. ^* Artificial fluorescent substances, such as eosin, also sensitize tissues and proteins to light. ^^ Normal blood absorbs light rays in large amounts, as Finsen showed, and it is quite possible that changes in the chemistry of the blood result from the light rays. Exposure to the sun may cause a general leucocytosis with relative lymphocytosis.^^ According to HerteP^ the ultraviolet rays cause oxygen to spUt off the easily oxidizable compounds of protoplasm, and Bovie*^ found that they coagulate proteins; they also have a destructive effect on enzymes, ^^ serum complement'*^ and hormones. ^^ However, Burge,^^ found that exposure of living cells to ultraviolet radiation of sufficient intensity to kill the cells does not decrease to any appreciable extent the activity of the intracellular enzymes; the cell death he attributes to coagulation of protoplasm. Harris and Hoyt^^ advance evidence that the susceptibility of protoplasm to ultraviolet light is conditioned by selective absorption of the toxic rays by the aromatic amino-acids of the proteins. Toxins are reduced in activity by ultraviolet raj^s.^' X=rays^^ stimulate cell growth when applied in small amounts,^^ but larger amounts produce necrosis, which is peculiar in that an in- terval of several days, or even weeks, may elapse after the exposure before the necrosis manifests itself. Ellis^^ considers that the amount of necrosis is out of proportion to the changes in the vessels, which some have believed to be the cause of x-ray gangrene, and therefore that the cells must be directly injured," a view supported by Case- mir's^^ experiments with plant cells. The extensive studies of the " Hausmann, Biochem. Zeit., 1914 (67), 309. ** Full review on photodynamic action of light by Sellards, Jour. Med. Res. 1918 (38), 293. « Aschenheim, Zeit. Kinderheilk., 1913 (9), 87; Taylor, Jour. E.\p. Med., 1919 (29), 41. *' Zeit. Augenheilk., 1911 (26), 393. *« Science, 1913 (37), 24; see also Burge, Amer. Jour. Physiol., 1916 (39), 335. " Brooks, Jour. Med. Res., 1918 (38), 345. 60 Burge et al, Amer. Jour. Physiol., 1916 (40), 426. " Amer. Jour. Physiol., 1917 (43), 429. " Science, 1917 (46), 318; Univ. Caiif. Publ. (Pathol.), 1919 (2), 245. " Ilartoch et al, Zeit. Immunitat., 1914 (21), 643. ** Full review by Colwell and Russ, "Radium, X-Rays and the Living Cell," London, 1915. Also see Richards, Science, 1915 (42), 2S7. " See Schwarz, Munch, med. Woch., 1913 (GO), 2165. " Amer. Jour. Med. Sci., 1903 (125), 85. " Allen (Jour. Med. Research, 1903 (9), 462) states that protozoa ami vinegar eels are killed by long exposure to a;-rays, whereas plants are dooiiledly stiinulatod in their growth. " Med.-Naturw. Arch., 1910 (2), 423; rdsuni6 on a-rays. NECROSIS 377 Hertwigs show that the chromatin is chiefly affected, which presum- ably explains the fact that immature cells, and cells in active division, are more sensitive to a:-rays than adult cells, and that monstrosities develop from eggs exposed to radiant energy. As far as histological changes show, hard rays produce less but (juite the same changes as soft rays. That .r-rays have a marked effect on metabolism has been abundantly established.^'' According to Musser and Edsall,®" the ef- fect of :r-rays upon metabolism is unequalled by any other therapeu- tic agent, and is manifested by excessive elimination of the products of protein destruction, which arise particular!}' from the lymphatic structures. '^' These changes have been studied, therefore, particu- larly in connection with the treatment of leukemia {g. v.). In con- sequence of the injury to the blood-forming tissues, resistance to bac- teria is decreased (La wen). ^^ The renal epithehum seems also to suffer injury in some cases. ""^ Exposure of the entire body of animals, or large areas of hemato- poietic tissue in man, leads to profound changes. Chief of these are destruction of lymphoid cells, pigmentation of the spleen, destruction of bone marrow cells, primary rise in poh^morphonuclear cells followed by a fall to below normal, stead}' decline in lymphocyte count, and an increased resistance of the red cells to radiation. "^^ So marked may be the effect of x-rays on the marrow and spleen that antibody forma- tion is greatly depressed (Hektoen).^^ After heavy doses marked metaboHc changes occur which indicate a profound intoxication, there being vomiting and diarrhoea, high non-protein N in the blood and a great increase in the urinary N (Hall and Whipple).®^ These authors also observed necrosis in the intestinal epi '"helium. Presumably these reactions are similar to those observed following superficial burns, and depend on disintegration of tissue proteins with production of toxic substances. The long-continued action of a:-rays upon the skin has, in many cases, led to the formation of cancer, apparently because the pro- liferation stimulated by the rays progresses until it exceeds normal ^^ See Harvey (Jour. Path, and Bact., 1908 (12), 548), concerning the effects of x-rays. «o Univ. Penn. Med. Bull., 1905 (18), 174; also Edsall and Pemberton, Amer. Jour. Med. Sci., 1907 (133), 426. ^' A peculiar selective action for the generative cells is also shown by x-rays, which cause marked atrophy of the ovaries and testicles. In the latter it affects chieflv the germinative cells, sparing the cells of Leydig. (See --Ubere-Schonberg, Munch, med. Woch., 1903 (50), 1850; Frieben, ibid., 1903 (50), 2295; Specht, Arch. f. Gvn., 1906 (78), 458; Thaler, Deut. Zeit. f. Chir., 1905 (79), 576; Reif- ferscheid, Zeit. f. Gyn., 1910 (34), 593. 62 Mitt. Grenz. Med. u. Chir., 1908 (19), 141. " See Schulz and Hoffman, Deut. Zeit. f. Chir., 1905 (79), 350; Warthin, Amer. Jour. Med. Sci., 1907 (133), 736. 6^ Resume by Gudzent, Strahlentherapie, 1913 (2), 467. See also Taylor et al, Jour. Exp. Med., 1919 (29), 53. " Jour. Infect. Dis., 1918 (22), 28. « Amer. Jour. Med. Sci., 1919 (157;, 453. 378 RETROGRESSIVE CHANGES bounds." Likewise leukemia has been observed several times in roentgenologists, presumably produced in the same way.®^ As the metabolic changes produced by rc-rays indicate an extremely high rate of autolysis, one may ascribe the effects either to a stimulat- ing effect of .T-rays upon autolytic enzjanes, or as Neuberg^^ does, to an inhibitive action of a;-rays and radium rays upon the other intra- cellular enzymes without a corresponding deleterious effect upon the autolytic enzymes.^" This hypothesis agrees with the facts at hand, bui more details concerning the effects of these rays upon various enzymes are needed. The long latent period before the appearance of necrosis after exposure to x-rays is difficult to explain^ and agrees rather with the hypothesis of slow proliferative and obstructive changes in the blood-vessels. Radium, which shares with x-rays the power of causing tissue necrosis, does not have so marked an effect upon the blood, '^^ nor do the ultra-violet rays (Linser and Helber).''^ In general, radium has much the same effect on tissues as .r-rays,^^ but seems rather to stimu- late the action of most enzymes ;^^ autolysis, however, is not increased (Brown). '^^ Radium partially destroys the growth-promoting "vit- amines" of yeast, which may account for some of its effects on tumors (Sugiura and Benedict). Radium also causes severe skin lesions and a general lymphocytosis in those exposed to it for long periods.''^ Active deposit of radium emanation injected intravenously into animals is highly toxic, even small doses causing fatty degeneration in the liver associated with giant cell formation and hyperchromatic nuclei; larger doses cause multiple hemorrhages and death with severe enteritis. Lesions also occur in the kidneys, lungs, spleen and bone marrow. '^^" In proper amounts radium stimulates plant meta- bolism (Gager). Thorium-x also attacks specifically the leucocytes, ^^ so that by proper dosage an animal may be made practically leucocyte- s' See review by Wyss. Beitr. z. klin. Chir., 1906 (49), 185; Porter and Wol- bach. Jour. Med. Res., 1909 (21), 357. «8 See Jagic and Schwarz, Berl. klin. Woch., 1911 (48), 1220. «9 Zeit. f. Krebsforschung, 1904 (2), 171; also Meyer and Bering, Fortschr. Roentgenstrahlen, 1911 (17), 33; Richards, Amer. Jour. Physiol., 1914 (36), 400. '" Some authors have believed certain of the effects of .r-rays to be produced by choline liberated through the decomposition of lecithin. (See Benjamin and Reuss, Munch, med. Woch., 1906 (53), 1860.) '1 See Millet and Mueller, Jour. Cancer Res., 1918 (3), 127. ^2 Deut. Arch. kUn. Med., 1905 (83), 479. '3 Review by Guyot, Cent. allg. Path., 1909 (20), 243; also see Mills, Lancet 1910 (179), 462; Richards, Science, 1915 (42), 287. Full bibliography by Sugiura and Benedict, Jour. Biol. Chem., 1919 (.39), 421. 7^ Loewenthal, Berl. klin. Woch., 1910 (47), 287; Kionlca, Med. Klinik, 1911 (7), 68.5. Denied by Gudzent, Zeit. Strahlenther., 1914 (4), 666. '6 T. R. Brown, Arch. Int. Med., 1912 (10), 405. " See Ordway, .lour. Amer. Med. Assoc, 1916 (66), 1. '*"Bagg, Jour. Cancer Res., 1920 (5), 1. " See Plesch et al, Zeit. exp. Path., 1912 (12), No. 1; Schweizer, Miinch. med. Woch., 1916 (63), 341. NECROSIS 379 frco,^^ whi(;h has been used for experimental studies on the functions of the leucocytes. Electricity. — The effects of the electric current upon cells are de- scribed by Davenport as follows: A weak constant current causes a centripetal flowing of the protoplasm (in Actinosphaerium) ; if the current is increased or long continued, the cytoplasm of the pseudo- podia becomes varicose, and droplets are formed which soon burst, causing a collapse of the protoplasmic framework. Finally, the protoplasm on the anode side begins to disintegrate, and the loose particles move toward the positive electrode; eventually the cell structure may be entirely destroyed. A similar disintegration of the anode side of ameba has been observed by McClendon,^^ which he attributes to anions which cannot pass through the cell wall, and therefore accumulate on that side of the organism. If an alter- nating current is used, both anode and cathode sides of the cell are affected. In moving organisms electric currents determine direction of motion, even certain vertebrates (tadpoles, fish) being made to orient themselves according to the current. The nucleus seems to be more susceptible to harm by electric currents than the cytoplasm (Pfeffer),^" and there seems to be no oxidation-process involved in cell destruction by electricity (as is the case with light rays), for the effects are much the same in the absence of oxygen (Klemm). Schmaus and Albrecht state that the effect of electricity upon proto- plasm depends upon a loosening of the cohesion and a solution of the constituents of the cell (vacuolization), which last is, perhaps, due to direct chemical alterations. It may be suggested that the electric current causes a migration of ions toward one or the other pole of the cell, in this waj^ separating the movable inorganic ions of the ion-protein compounds of the cell from the immobile colloidal pro- teins, with consequent serious alterations in the chemistry of the cell. Zeit^^ found that continuous currents kill bacteria through the pro- duction of antiseptic substances in the culture-medium, but do not harm them directly. Jellinek^" has studied extensively the cause of death after severe electric shocks, and finds that there are produced intracerebral hemor- rhages and degeneration of the nerve-cells, which are sufficient to explain the death of the individual without having recourse to the more indefinite idea of "shock." Cunningham^^ considers fibrillary contraction of the heart as the cause of death. ^^ Spitzka and Ra- ^^ There is no increase in antitrypsin from this leucocyte destruction (Rosenow, Zeit. exp. Med., 1914 (3), 377). '^ Pfitiger's Arch., 1911 (140), 271. *'' Literature given by Davenport, "Experimental Morphology." 81 Jour. Amer. Med. Assoc, 1601 (37), 1432, literature. 82 Virchow's Arch., 1902 (170), 56; Lancet, 1903 (i), 357. " New York Med. Jour,, 1899 (70), 581. 8^ Full discussion by Jelliffe in Peterson and Haines' "Legal Medicine and Toxicology," 1903 (1), 245. 380 RETROGRESSIVE CHANGES dasch^^ find changes in the brains of electrocuted criminals, which indicate a sudden liberation of gas about the blood vessels, along which the current passes. The amperage seems to be far more important in determining the effect of a current than the voltage or wattage.^® Chemicals cause cell death whenever they are of such a nature as either to coagulate the cell proteins or to destroy its enzymes. The action of such substances as sulphuric acid, strong caustics, etc., hardly calls for explanation. Phenol (carbolic acid) may cause ne- crosis and gangrene even when in very dilute solution; this appears to be due more to the production of hyaline thrombi of agglutinated red corpuscles in the capillaries than to direct action upon the cells. In some unpublished experiments on the subject of "carbolic acid gangrene," I found this action of phenol very striking when dilute solutions were placed on the web of a frog's foot, under the micro- scope; as soon as the solution penetrated to a capillary, stasis with fusion of the corpuscles occurred in a very few seconds. Similar results have been obtained by Rosenberger.^^ Some poisons seem to cause necrosis without destroying the autolytic enzymes, in which case the cells are rapidly digested; at least, such an hypothesis seems -to explain best the changes seen in the liver in chloroform poisoning, acute yellow atrophy, eclampsia, etc.^* Not all poisons, b}^ any means, cause cell death — tetanus toxin, morphine, and other alkaloids cause death of the individual as a whole without usually causing pri- mary necrosis of any of the cells. Cell death does not necessarily de- pend upon destruction of all the cellular enzymes, as has been pointed out previously. Thus, bacteria may be killed by many chemicals which seem not to affect their autolytic enzymes seriously. Any con- siderable excess of either H or OH ions is incompatible with cell life, and it is possible that at times the production of acids within a cell may be sufficient to cause death;^^ e. g., in the kidney in acute nephritis (M. H. Fischer), or in the muscle in waxy degeneration ( Wells). ^° It is quite probable that many of the poisons act by interfering with the oxidative capacity of the cells; this seems almost certain in the case of chloroform necrosis, and even bacterial poisons (diphtheria and typhoid) were found by Pitini^^ to decrease the oxidizing power of the cells. The term, "protoplasmic poison," has been variously used and de- fined. Kunkel says that a protoplasmic poison "is a poison which, 86 Amer. Jour. Med. Sci., 1912 (144), 341. 89 Jellinek, Wien. klin. Woch., 1913 (26), 1793. 8' Verb. Phys. Med. Gesellsch. z. Wurzburg, 1900, vol. 34. 88 Wells, Jour. Amer. Med. Assoc, 1906 (46), 341. 8" The partial protection afforded by a rich carbohydrate diet against the necrogenic action of chloroform, phosphorus and renal poisons, as observed by Opie and Alford (Jour. Exp. Med., 1915 (21), 1), may depend on the antiketogenic effect of carbohydrates. »" Jour. Kxp. Med., 1909 (11), 1. »' Biochem. Zeit. 1910(253.257. NECROSIS 381 witliout producing directly evident alterations, harms or kills all living protoplasmic structures." HgCla is such a poison, whereas H2SO4, bromine, and similar substances that destroy all life through their strong chemical action are not included in this category. The j)rotoplasmic poisons presumably act by combining with one or more of the constituents of cell protoplasm; e. g., HgCL i)robably combines with the proteins, chloroform with the cell lipoids (physically?). By means of his special teehnic Barber^- is able to introduce minute quantities of poisons into living cells and observe their effect on the cytoplasm; HgCl2 is thus found to be most toxic, while AS2O3 is relatively inert. Mathews'-^^ has shown that the toxicity of ions depends on the ease with which they part with their electrical charges, and the toxicity of a salt is a function of the sum of the toxicity of the ions; hence the toxicity of a salt is in inverse proportion to its decomposition tension. Kunkel suggests that oxalic acid and fluorides are poisons because they combine the cell calcium, and barium salts may be poisonous because they precipitate the SO4 ions. We can readily imagine that the combining of even one of the essential con- stituents of the cell may so upset the normal chemical processes that the cell no longer takes up substances to repair its waste, and hence necrosis ensues. ^^ Physical agents may cause necrosis, usually in ways too obvious to require explanation. With most cells, large portions of the cyto- plasm can be destroyed without serious results, for so long as the nucleus is intact the cytoplasm can be reconstructed. The fact that necrosis frequently follows relatively slight injuries of the nucleus is perhaps best explained by considering that injury to the nuclear membrane modifies the permeability of the nucleus for substances in solution, which might readily affect its metabolic activities to a serious degree. It is possible, also, that solvents of lipoids, such as chloro- form, etc., produce much of their deleterious effects by modifying the permeability of the cell, if the semipermeability of cell mem- branes depends largely upon the lipoids they contain. ^^ Physical injury of even slight degree may bring on severe alterations in cells, however, and indeed may cause severe chemical alterations. We know that many chemical reactions can be brought about by slight mechanical disturbances, e. g., the explosion of fulminate, nitrogen iodide, etc., and it is quite possible that mechanical disturbances can, likewise, cause chemical changes in the protoplasm. JNIechanical injury of cells under the microscope results in an apparent increase 52 Jour. Infect. Dis., 1911 (9), 117. '3 Amer. Jour. Physiol., 1901 (10), 290; NichoU, Jour. Biol. Chem., 1909 (5), 453. ^* It is hardly profitable here to go further into the theories of the action of poisons, which are generally extensively considered in the treatises on toxicology and pharmacology (also by Davenport, loc. cit). 5^ See Pascucci, Hofmeister's Beitrage, 1905 (6), 552. 382 RETROGRESSIVE CHANGES in the acid reaction of the part involved (Chambers) ^^ and Hkewise traumatized nervous tissues develop an acid reaction (Moore). ^^ Many lower animals devoid of a nervous system respond to mechanical stimuli by chemical activity; e. g., the production of phosphorescence by marine organisms when agitated by an oar, etc. Possibly, the secretion of thrombokinase by the leucocytes, which occurs whenever they come in contact with a foreign body, is an example of a similar reaction to a mechanical stimulus. Even in urticaria factitia the sim- ple mechanical irritation which suffices to produce the wheals is fol- lowed very quickly by extensive nuclear fragmentation,^^ but it may be that unknown poisons are present in the hypersensitive skin and cause the karyorrhexis, and not the trauma alone. We have no good evidence that mere contact with a chemically inert foreign body unac- companied by cellular injury, can cause death of tissue-cells.^^ How- ever, Chambers ^^" states that simple trauma, even mere compression, of the eggs of asteria may cause them to coagulate into a solid mass. Extreme changes in osmotic pressure may lead to cell death, either by causing structural alteration in the cell (e. g., the bursting of plant- cells in water), or concentration of the electrolytes may become so great that the colloids are thrown out of solution, as in the ordinary salting-out processes of the laboratory. It is doubtful, however, if osmotic changes per se ever become so abnormal within the animal body (except in experimental conditions) as of themselves to cause cell necrosis. Varieties of Necrosis Coagulation Necrosis.' — This name is applied to necrotic areas that are firm, dry, usually pale yellowish in color, and observed prin- cipally in areas of total anemia or tuberculosis. The question has been long disputed as to whether a true coagulation occurs in such tissues or not. Necrosis produced by heat, carbolic acid, corrosive sublimate, etc., is naturally a coagulation necrosis, the cells of the affected area having undergone true coagulation; i. e., the conversion of their soluble colloids (sols) into the insoluble "pedous" modification. Whether the same change occurs in areas of anemic necrosis is not so well established. If the part contains a fair amount of plasma the liberation of the tissue coagulins from the dead cells will cause a con- version of the fibrinogen into fibrin — this can usually be demonstrated microscopically, but the presence of fibrin is not constant, and its quantity is usually insufficient to explain satisfactorily the condition »« Amer. Jour. Physiol., 1917 (43), 1. " Proc. Soc. Exp. Biol. Med., 1917 (15), 18. 08 Gilchrist, Bull. Johns Hopkins Hosp., 1908 (19), 49. 90 Meltzer (Zeit. f. Biol., 1894 (30), 4G4) has shown that bacteria may be killed by violent agitation, which causes disintegration of the cells. 09" Trans. Roy. Soc. Canada, 1918, p. 41. ^ Literature by Jores, Ergebnisse der Pathol., 1898 (5), IG. COAGULATION NECROSIS 383 of coagulation necrosis in infarcts, etc., as Weifrert maintained.^ Schmaus and Albrecht believe that a true coagulation of the cell proteins does occur in anemic infarcts, etc., for they found that the cells of kidneys with lijrated vessels contain at first granules soluble in water and salt solution; after forty-eight hours the granules cannot be dissolved in these solvents or in weak acetic acid, but are soluble in 2 per cent. KOH; after five to six days the granules are insoluble even in KOH. Beyond these experiments, we seem to have no proof of the occurrence of intracellular coagulation within areas of coagula- tion necrosis due to anemia; exact chemical studies on this point are much needed. Since tissue-cells contain coagulins for fibrinogen, it is possible that they also contain coagulins for cell-proteins, but this remains to be established. We do not know whether Chambers' observations on the spontaneous coagulation of tiaumatized asteria eggs^^o are applicable to other cells. Bacteria produce substances coagulating milk and fibrinogen. Bergey^ calls attention to the coagulation of serum by enzymes and acids produced by bacteria, and RuppeP found that the tubercle bacillus produces substances precipitating proteins; hence coagulation necrosis in bacterial infec- tions may be brought about in this way, and SchmolP has shown that the necrosis occurring in tubercles is associated with an almost complete coagulation of the cell-proteins. Necrosis associated with inflammatory exudation is, of course, ac- companied by coagulation of the fibrinogen of the exudate (e. g., diphtheria); this type of coagulation necrosis is chemically a simple fibrin-formation and readily understood. The peculiar hyaline de- generations of parenchymatous cells (e. g., Zenker's degeneration of muscles) are often included under this class, but it would seem more probable that the processes consist rather of the fusion of the structural elements of the cell into a homogeneous substance than a true coagulation. When necrosis is produced by chemical means more or less coagulation of some of the soluble proteins probably takes place; even in plant cells this coagulation of dead protoplasm is described.^ Liquefaction necrosis occurs particularly in the central nervous system, where the cell substance seems not to undergo the coagulative ^ Weigert believed that the dead area becomes permeated by plasma containing fibrinogen, which is coagulated in and between the cells. He put much weight on an increase in size of the necrotic area, which is by no means constant, as he intimated; necrotic areas are inelastic, and when death occurs thej- do not shrink with the fall of blood pressure as the surrounding tissues do, and hence they may appear to project from the surface of the dead organ when thej- did not do so during life. According to Moos (Virchow's Archiv., 1909 (195), 273) the plasma does not permeate infarcted areas to the extent that Weigert assumed. 3 Jour. Amer. Med. Assoc, 1907 (49), 680. * Zeit. phvsiol. Chem., 1898 (26), 218. 6 Deut. Arch. klin. Med., 1904 (81), 163. ^ Gaidukov. Zeit. chem. KoUoide, 1910 (6), 260; Lepeschkin, Ber. Deut. Bot. Gesell., 1912 (30), 528. 384 RETROGRESSIVE CHANGES changes described in the preceding paragraphs. Whether this is due to a lack of tissue-coaguhns or to a difference in cell composition cannot be said, but the large proportion of lipoids in brain tissue is probably an important factor. Probably "edema ex vacuo" is re- sponsible for much of the accumulation of fluid, due to the anatomical conditions that prevent a shrinking or collapse of the tissues to fill in the gap, and the lack of connective-tissue formation. Aseptic softening in general may be safely ascribed to digestion of proteins by cellular enzjnnes, either from the dead cells or from the leucoc3'tes. Suppuration is merely a form of liquef active necrosis, in which such digestion is particularly rapid because of the large number of leucocytes that are present. Necrosis of the gastric mucosa or of the pancreas is also followed by rapid liquefaction, through the action of the digestive enzymes of these tissues. When necrosis is accompanied by edema (as in superficial burns), the fluid enters the cells in large amounts, and in this way another form of liquefaction necrosis may be produced. Bacterial enzymes may be a factor in producing liquefaction of dead tissues, but with most pathogenic forms there is little proteolytic activity.^ Caseation. — This term is applied to a form of coagulation necrosis in which the dead tissue has an appearance quite similar to that of cheese. If we bear in mind the fact that cheese is a mixture of coagu- lated protein and finely divided fat, and that in caseation we have a coagulation of tissue proteins associated with the deposition of con- siderable quantities of fat, the reason for the gross resemblance of the product of this form of necrosis to cheese is apparent. SchmolP has analyzed caseous material, and found it almost entirel}" free from soluble proteins or proteoses. The protein material is almost solely coagulated protein, which in its elementary composition is related to the simple proteins or to fibrin, and not at all to the nucleoproteins. The extremely small amount of phosphorus present in the caseous material indicates that the products of disintegration of the cell nuclei must diffuse out early in the process. Caseation is, therefore, char- acterized by a coagulation of the proteins and a dissolving outTof Jhie nuclear components. Schmoll does not explain the cause of coagula- tion, however. It may be that it is the same as in the coagulation of anemic infarcts (since tuberculous areas are decidedly anemic), or possibly the tubercle bacillus produces substances coagulating pro- teins, as Ruppel states is the property of "tuborculosamin." Indeed; Auclair^ claims that the fatty substance that can be extracted from tubercle bacilli by chloroform is the cause of the caseation. Dead tubercle bacilli do not produce true caseation, however, according to Kelber;^" hence the substance causing the necrosis evidently docs not ^ See Bittrolff, Beitr. path. Anat., 1915 (60), 337. 8 Deut. Arch. Idin. Med., 1904 (81), 163. 9 Arch. m6d. exper., 1899, p. 363. 10 Quoted by Uurck and Oheindorfer, Ergebnisse der Pathol., 1899 (6), 288. CASEATION 385 diffuse readily from the bodies of the bacilli. Comparison of the chemical composition of bovine and human tuberculous lesions witli the corresponding normal tissues by Caldwell'^ gave the following results: The tubercle walls and the caseous material from lymph gland tubercles contain a lower percentage of water than does the normal tissue. In normal bovine liver tissue, the percentage of water present is less than that of the tubercle walls or of the caseous material from liver tubercles. The specimens of oasQous material from lymph gland and liver tubercles approach each other closely in their water content, the average being about 75% for the bovine material. The alcohol-ether-soluble substances from normal bovine lymph glands form about 24.4% of the dry weight, or about 4.4% of the moist weight. The walls of the lymph gland tubercles contain a distinctly larger amount of lipins than does the caseous material or the normal tissue. On the contrary, the walls of liver tubercles are poor in lipins as compared with the normal tissue, and they contain a smaller amount of fats than does the caseous material from these tubercles. When calculated on the basis of the dry weight, the caseous material from lymph gland tubercles contains a smaller percentage of lipins than does normal lymph gland tissue. When the ash is deducted, this difference disappears and the content of lipins becomes equal to or slightly greater than that of the normal tissue, but less than that of the tubercle walls. When calculated on an ash-free basis, the lipin content of the caseous material from liver tubercles is distinctly less than that of the normal tissue but greater than the lipin content of the tubercle walls. Cholesterol forms about 6.5% of the lipins from normal bovine lymph glands, or about 1.5% of the dry weight. The lipins from the walls of lymph gland and liver tubercles contain, in every case, 2-3 times as much cholesterol as do the lipins from the normal tissues. This is an actual increase also when calculated on the basis of the dry weight. The caseous material contains even a larger percentage of cholesterol than do the tubercle walls. Phospholipins constitute about 32% of the lipin fraction of normal bovine lymph glands, or about 7.9% of the dry weight; the corresponding values for normal liver are 41.2% of the fats, or 14% of the dry weight. The phospholipin content of the fats from the tubercle walls is slightly less than that of the normal tissues, while there is a very marked reduction in the phospholipin content of the lipins from caseous material of bovine origin. In the specimen of caseous material from human lymph glands, phospholipins formed 30.9% of the total lipins. The iodin numbers obtained from the fats of the tuber- culous specimens from lymph glands are higher than those from the normal tissues. This observation does not hold true for the liver specimens. In the latter, there is no difference noted between the iodin numbers obtained for the lipins from normal and tuberculous specimens, although the values are practically the same as those from the fats from the lymph gland tubercles. In the residues of caseous material left after extraction with alcohol and ether the nitrogen content remains relatively high; in fact, the reduction in nitrogen content is only slight when the calculations are made on ash-free residues. The percentage of nitrogen does not differ much from that obtained from the normal proteins of these tissues. In specimens of caseous material in which there are no macroscopic evidences of calcification other than the presence of sandlike particles, calcium sometimes forms as much as 15% of the residue left after extraction of the fats. In such residues, the phosphorus content may reach 9%. The amount of purine nitrogen in the walls of lymph gland tubercles is only slightly more than iiali that of normal lymph gland tissue, and the amount is apparently much less in the caseous material. In the residues from the walls of liver tubercles, purine nitrogen is present in only slightly higher percentage than in the normal liver. The results here obtained would seem to indicate that the pur- ines are even more abundant in the caseous residues of liver tubercles. The amount of material which enters the water solution during extraction is distinctly less from caseous material than from the residues of normal tissues. " Jour. Infect. Dis., 1919 (24), 81. Full review on composition of tuberculous tissues. 25 386 RETROGRESSIVE CHANGES The abundance of fat in caseous material on microscopic exami- nation is very striking. In addition to the figures obtained by Caldwell, Bossart^^ found from 13.7 per cent, to 19.4 per cent, of the dry sub- stance of caseous material soluble in alcohol and ether. In the scrap- ings from tuberculous bovine glands I have found 22.7-23.9 per cent, of the organic material soluble in alcohol and ether. ^^ Of this soluble material, Bossart found 25 to 33 per cent, of cholesterol, and Leber^'* found 38.31 per cent., which is a much higher phospholipin pro- portion than Bossart detected. Caldwell found cholesterol higher and phospholipins lower in caseous than in normal tissues. The total amount of lipins, however, constituted a smaller percentage of the dry weight than in the normal tissues from which the caseous material originated. Presumably these fatty materials are derived chiefly from the disintegrated cells; this is probably true of the phos- pholipin and cholesterol, but the fact that in histological preparations most of the fat is found about the periphery of the caseous area,^^ sup- ports the belief that it has wandered in from the outside.^® A certain proportion of the fat is possibly derived from the bodies of the tubercle bacilli, which usually contain about 40 per cent, of fatty matter; but it has not been determined whether the fat from this origin forms an appreciable part of the fatty matter of caseous material. Caseous areas persist for extremely long periods of time without undergoing absorption, which indicates that the autolytic enzymes are destroyed early in the process, presumably by the toxins of the tubercle bacillus; corresponding to this Schmoll found autolj'sis very slight indeed in caseous areas, and even when the caseous material breaks down to form a "cold abscess" the fluid differs from true pus in containing less free amino-acids, e. g., tyrosine is missing." Caldwell also obtained lower figures for extractives in caseous than in normal tissues. Because of a lack of chemotactic substances no leucoc3'tes enter to remove the dead material, in consequence of which caseous material gives no evidence of containing proteases, according to the Miiller-Jochmann plate method. That the failure of absoprtion is not due to a modification of the proteins into an indigestible form is shown by the rapid softening of caseous areas when, through mixed infection, chemotactic substances are once developed and leucocytes enter. Jobling and Petersen^^ suggest that in caseation the autolysis is inhibited by the soaps of fatty acids, which are abundant in caseous areas and have a marked antitryptic effect. ^* Quoted by Schmoll, loc. cit.^ " Wells, Jour. Med. Research, 1906 (14), 491. ^* Quoted by Schmoll.^ 1^ Sata, Ziegler's Beitr., 1900 (28), 461. ^^ Fischler and Gross (Ziegler's Beitr., 1905 (7th suppl.), 344) could find no fatty acids in caseous areas bv histological methods, i' See Muller, Cent. inn. Med., 1907 (2S), 297. IS Jour. Exp. Med., 1914 (19), 239; Zcit. Immunitat., 1914 (23), 71. PANCREATIC FAT NECROSIS 387 Fat Necrosis" ThrouKh iisaj.'c (his Icrm Ikis come to indicate a specific form of necrosis of fat tissue, which is characterized l)y a focal, circumscribed arran Carnegie Inst. Publ., 1915, No. 203. " Cesa-Bianchi, Frankf. Zeit. Path., 1909 (3), 723. " Morgulis, Howe and Hawk, Biol. Bull., 1915 (28), 397. " Biol. Bull., 1918 (34), 149. ''* Hlemons, Bull. Johns Hopkins Hosp., 1914 (25), 195. "^ Jour. Biol. Choin 1910 (25), 201. " Review of general features by Landsteiner, Ziegler's Beitr., 1903 (33), 237. CLOUDY SWELLING 397 gests that the change consists in a coagulation of the cell proteins, which idea is supported by the similarity of the niicTOscopic changes observed in the cells and the earliest microscopic changes observed in cells after heating gently to about their maximum thermal point. On the other hand, the granules in cloudy swelling are generally de- scribed as being soluble in dilute acetic acid and dilute KOH, which indicates that they are not the result of ordinary heat coagulation. If we bear in mind, however, that cloudy swelling probably does not represent one single change, it may be possible to arrive at some understanding of the chemical changes that occur in the process. Albrecht''^ considers, with good reason, that we may have a granular appearance of cells which is simply an exaggeration of the normal granular structure, and, although it may be observed in tissues mod- erately affected by toxins, or in starvation, or in transitory anemia, the change is still to be looked upon as little more than physiological in response to stimuli and overwork. Such a "cloudy swelling" may also occur in cells in the beginning of autolysis, or simply under the influence of salt solution. If the injury is greater, however, as in profound sepsis, or extreme local anemia, the granules becomes coarser, less soluble in acetic acid and KOH, and droplets resembling "myelin" make their appearance. If the injury is still more severe, true coagu- lation of the granules occurs, and they become insoluble, the fatty droplets become more prominent, and the cell reaches a condition that may with propriety be termed necrosis or fatty degeneration, or both. There is no very sharp line separating necrosis and cloudy swelling, especially if we consider only the changes in the cytoplasm. In the earliest stages the granules are perhaps due, in some cases, to simple aggregation of the colloids, without the development of a true coagulation, and so the granules are still soluble. Possibly bacterial, toxins may also cause soluble precipitates, but this does not appear to have been established. Halliburton has shown that temperatures that may be reached in high fevers can cause turbidity in solutions of cell proteins, and hence heat precipitation may be partly responsi- ble for the turbidity of cells in cloudy swelling, but it is doubtful if the granules thus formed would be soluble in acetic acid. A careful discussion of the character and characteristics of this process is given by Bell,"^ wdio concludes that the term cloudy swelling is sound only as a gross description, since microscopically the cells may be found to show albuminous granules, or fatty metamorphosis or simple edema. When present, the granules are of unknown nature — they are not identical with Altmann's granules, although Aschoff and Ernst^^ both consider that many of them are derived from the mitochondria. An enormous number of granules may be present in the renal cells " Verb. Deut. Path. Gesell., 1903 (6), 63. '8 Jour. Amer. Med. Assoc, 1913 (61), 455. " Verb. Deut. Path. Gesellsch., 1914 (17), 43 and 103. 398 RETROGRESSIVE CHANGES without demonstrable impairment of function.^" They may disap- pear dm-ing acute infections, and they bear no constant relation to fatty changes. We may speak with more assurance concerning the swelling of the cell, and attribute it to an edema of the cell contents, it having been shown that in cloudy swelhng the water content of the organs is in- creased.^^ This might be produced by a rise in osmotic pressure due to abnormally rapid splitting of proteins with incomplete oxidation of the substances formed, which results in formation of many crj^stalloid molecules with high total osmotic pressure, from a smaller number of colloid molecules with almost no osmotic pressure. It has frequently been shown that the cell-walls do not lose their semipermeable character until the death of the cell occurs; hence in cloudy swell- ing water diffuses in much more rapidly than the crystalloids can diffuse out,^^ causing a hydropic swelling. This hypothesis is sup- ported by the observations of Cesaris Demel,^^ who found that by modifying the osmotic conditions of the cells, particularly epithelial cells, he could closely reproduce many of the characteristic features of parenchymatous degeneration. It is possible, also, that too high concentration of crystalloids within the cells may be a factor in the precipitation of the cell colloids. In view of the fact that in the earliest stages of autolysis, histologic and microscopic changes closely resembling those of cloudy swelhng are pronounced, and that organs the seat of cloudy swelling notoriously undergo autolysis with extreme rapidity after death, ^"^ we may also consider that this process is possibly in part responsible for the change of ordinary intra vitam cloudy swelling. The appearance of fine granules of lipoid substance^"" (myelin or "protagon") in cells during autolysis and during cloudy swelling is in support of this idea, and chemical analysis of organs showing cloudy swelling gives definite evidence of autolytic decomposition of the pro- teins and an increase in the water content. ^^ Presumably this increase in water is the cause of the lowered specific gravitj'- of organs exhibiting parenchymatous degeneration.^^ Landsteiner, through his studies of cloudy swelling in human material, also came to the conclusion that autolysis is an important element in its production. Martin H. Fischer^*^ applies the principles of colloidal chemistry to the problem and concludes that the changes of cloudy swelling 80 Shannon, .lour. Lab. Clin. Med., 191G (1), 541. 81 Schwenkenbecher and Ingaki, Arch. exp. Path. u. Phann., 1906 (55), 203. 82 See introductory chapter concerning osmosis; also discussion of edema. 83 Lo Sperimentale, 1905; Cent. f. Path., 1905 (IG), 613. 84 See Medigreceanu, Jour. Exp. Med., 1914 (19), 309. s^Orglcr, Virchow's Arcli., 1904 (176), 413; Hess and Saxl, ibid., 1910 (202), 149. 86 Verh. Deut. Path. Gesell, 1903 (6), 76. 87 See Olsho, Arch. Int. Med., 1908 (2), 171. 88 "Oedema and Nephritis," New York, 1915, p. 455; also Zeit. Chcm. u. Indust. Colloide, 1911 (8), 159. CLOUDY SWELLING 399 may be ascribed to acids developed in the cell. It is of significance that Chambers^'-* has found that even slight mechanical injury of iso- lated cells under the microscope produces a demonstrable acidity in the protoplasm. Electro-negative proteins in the cell are precipitated by weak concentrations of acids, forming the granules in the cells, which can be dissolved again by a stronger concentration of acid as in the characteristic clearing of granular cells by acetic acid. The swelling is explainable by the increased affinity for water of other cell proteins under the influence of acids. This theory is supported by good experimental evidence and has much in its favor, the chief question being whether the blood cannot, under ordinary conditions of circula- tion, furnish sufficient neutralizing salts to prevent acidification in the cells to cause cloudy swelling. s^Amer. Jour. Physiol., 1917 (43), 1. CHAPTER XVI RETROGRESSIVE CHANGES (Continued) Fatty, Amyloid, Hyaline, Colloid, and Glycogenic Infiltration and Degeneration FATTY METAMORPHOSIS In 1847, in the first number of his Archiv, Virchow divided the forms of fatty changes that may occur in pathological conditions into two groups — "infiltration" and "degeneration" — -a division that has since become classical. By infiltration he indicated the ex- cessive accumulation of fat in the cells in the form of large droplets, without destruction of the nucleus or irreparable damage to the cells, and by the use of the term infiltration he implied his belief that the fat entered the cell from without. When the fat remained in the form of fine droplets and the cell became much disintegrated, Virchow considered that the fat was derived from the breaking down of the cell proteins, and hence the process was considered to be a fatty degenera- tion of the protoplasm. Since that time scarcely any other subject in pathology has been more warmly discussed than that of the origin of the fat in fatty degeneration, and an appalling amount of literature has accumulated concerning the questions involved. It will be im- possible to give more than the essential facts that have been developed, referring the reader for the full details of the discussion and evidence to the numerous compilations of literature, particularly those of Rosen- feld,' and to the original articles cited in the text. I "Fat Formation," Ergebnisse der Physiol., Abt. 1, 1902 (1), 651; ibid., 1903 (2), 50. Also see discussion in the Verh. Deut. Path. Gesell., 1904 (6), 37-108, and the review by Leathes in his "Problems in Animal Metabolism," 1906, pp 71-121, and "The Fats " Monographs on Biochemistry, London, 1910; von Fiirth, "Chemistry of Metabolism," Amer. Transl., New York. 1916. Concerning theor- ies of role of lipase in fat metabolism see Chap. iii. Other reviews of literature on pathological fat formation by Christian, Johns Hopkins Hosp. Bull, 1905 (16), 1; Lohlein, Virchow's Arch., 1905 (180). 1; Pratt, Johns Hopkins Hosp. Bull. 1904 (15;, 301 (particular reference to heart); Wohlgemuth, Handbuch d. Bio- chem., 1909, III (1), 150; Magnus-Levy and Meyer, ibid., 1910, IV (1), 445; Dietrich, Ergebnisse der Pathol., 1909, XIII (2), 283. Concerning Obesity see V. Bergmann, Handbuch d. Biochem., 1910, IV (2), 208. Later references of im- portance cited in the text. 400 FATTY METAMORPHOSIS 401 Physiological Formation of Fat Concerning the normal formation of fat we may summarize the evidence as follows : (1) A large proportion of the fat of the body comes from the fat taken in the food, as also does the fat of the milk. This can be shown, as Rosenfeld particu- larly demonstrated, by starving an animal until it is as free from fat as possible, then feeding with a large amount of some fat that is of a type different from that normally found in the animal; the new fat that it then laid up in the fat depots of the animal will partake of the characters of the fat given in the food. In case the animal is lactating, the milk-fat will also resemble the fat of the food. As a matter of fact, the body fat is not of constant composition, even in the same individual; it varies greatly with age, having much less olein in infancy than in later years, varying somewhat in composition in the different fat depots in the same body, and apparently being more or less modified by diet. (2) Fat may also be formed from carbohydrates. According to Rosenfeld, this fat differs from the fat formed on mixed diet in having less olein in proportion to the palmitin and stearin, and it is deposited particularly in the subcutaneous and mesenteric tissues rather than in the liver. Man does not seem to form fat readily from carbohydrates, but rather burns them to protect his proteins; on the other hand, swine and geese readily form fat from carbohydrates. As the fatty acid radicals of ordinary fat (C18H36O2, Ci6H.-;202, C1SH34O2), are much larger than the carbotiydrate radicals, a process of synthesis must be involved in the formation of fat from carbohydrates.^ (3; Proteins are a possible source of fat, but it has not been established that they are either a common or an important source of fat in either physiological or pathological conditions, or, indeed, that they really ever do form fat. Upon this statement rests our present tendency to refute the long-cherished conception of fatty degeneration as a true degeneration of cell proteins into fat, as suggested by Virchow. This view was supported by the earlier work of ^"oit and his school, who believed that they had demonstrated that animals could form fat from protein food, and their work was for a long time accepted as correct. Later Pfliiger and his pupils pointed out what seem to have been essential errors in these investigations, and, after much discussion and experimentation, the majority of physiologists now support the view advanced in the sentence opening this paragraph. Since proteins contain carbohydrate groups, and since fats can be formed from carbohydrates, the possibility of the formation of fats from the proteins in this indirect way cannot be denied. It is also possible that the nitrogen-containing groups may be split out of the amino-acids of the protein molecule, and that the non-nitrogenous residues can then be built up into fatty acid molecules as large as the molecules of stearic, palmitic, and oleic acids; but we have no proof that either of these processes occurs in the normal cell or in the cell that is undergoing degeneration. Atkinson and Lusk' have obtained evidence of some fat formation from meat fed to a dog, but this was only slight and obtained with difficulty. Pathological Fat Accumulation For a long time fatty degeneration was looked upon as one of the chief evidences that fat was formed directly from protein, for the cell protoplasm seemed, morphologically, to be changed directly into fat in this process. Additional support was also claimed from the supposed increase in fat in the ripening of cheese;^ from the forma- tion of abundant fat by maggots living in fat-poor blood or fibrin; and by the apparent conversion of proteins into fatty acids and soaps 2 This, Magnus-Levy suggests, may be accomplished through lactic acid which is formed from sugar, and then, after reduction to an aldehyde, several of these molecules are combined into the higher fatty acid. See Leathes, loc. cit., p. 82. 3 Proc. Natl. Acad. Sci., 1919 (5), 2-16. * Even the increase of fat in ripening cheese is doubtful (Nierenstein, Proc. Royal Soc, B., 1911 (83), 301; Kondo, Biochem. Zeit., 1914 (59), 113). 26 402 RETROGRESSIVE CHANGES in the postmortem change, adipocere. But it has now been well es- tabhshed that there is no true conversion of protein into fat in the fatty- degeneration produced experimentally by poisoning with phosphorus, etc.,^ and the other supposed instances of fat-formation above cited' have been discredited by various methods which it will not serve our purpose to discuss here, beyond mentioning that one of the chief sources of error Hes in the fact that many fungi and bacteria^ can form fat from protein. It having been rendered probable that fat was not formed by dis- integration of the protein of the degenerating cells, it remained to determine what the source of the fat observed in the cells under patho- logical conditions might be, and this part of the problem has been largely cleared up by Rosenfeld. This investigator proceeded as fol- lows: Animals were starved until they were extremely poor in fat, then fed upon easily identified foreign fats, such as mutton tallow (which has a liigh melting-point and can combine with httle iodin) or linseed oil (which has a low melting-point and can combine with much iodin). The animals under these conditions laid up in their fat depots, including the liver as well as the subcutaneous tissues, large quantities of these foreign fats. By starving again for a few days the foreign fat was removed from the liver, leaving still a large amount in the other storehouses, and the animals were then poisoned with phosphorus or other poisons that cause a typical fatty degener- ation of the liver and other viscera. When the fat was extracted from the fatty hver of these animals, it was found that the new fat that had appeared in the liver during the process was not normal dog fat (which it should have been if formed by degeneration of the cell pro- teins), but was, in part, of the same type as the foreign fat which the animals had deposited in their subcutaneous tissues and other fat storehouses. Furthermore, it was found that animals starved to an extremely low fat content do not develop the typical fatty liver of phosphorus-poisoning, a fact which Lebedeff had already noted in a ease of phosphorus-poisoning in an emaciated patient. Of similar significance is the fact that in fatty human livers the iodin number, normally high, falls as the amount of fat increases until it is approxi- mately that of adipose connective tissue.'^ Therefore, it seems evident that the fat accumulating iii the liver during fatty degeneration is not derived, as Virchow thought, through a tJ-ansformation of cell proteins into fat, but rather is an infiltrated fat brought in the blood fro^n the fat deposits of the body to the disintegrating organ. This work has since been corro- borated and extended by many observers, and its correctness can now ^See Taylor, Jour. Exp. Med., 1899 (4), 399; Shibata, Biochem. Zeit., 1911 (37), 345. " See Beebe and Buxton, Amer. Jour, of Physiol., 1905 (12j, 466; Slosse, Arch. Internat. Physiol., 1904 (Ij, 348. ^ Leathes, Lancet, Feb. 27, 1909; Hartley and Mavrogordato, Jour. Path, and Bact., 1908 (12;, 371; Jackson and Pearce, Jour. Exp. Med., 1907 (9), 578. FATTY METAMORPHOSIS 403 hartlly be questioned.'* "Fatty degeneration/' tliercforc, at least in some cases, differs from "fatty infiltration" chiefly in the fact that in the former the process is associated with serious injury to the cell, caused by the action of toxins or loss of nutrition, while in the latter the cell is not seriously injured and is capable of returning to its nor- mal condition whenever the fat is removed. '■* Fatty "Degeneration" without Infiltration. — By showing that new fat in fatty livers is infiltrated fat, Rosonfeld did not entirely clear up the subject, for, in the course of his analyses of organs that were micro- or macro-scopically the seat of fatty degeneration, he found that there is not always any correspondence between the amount of fat that seems to be present, as determined by microscopic methods, and the amount that chemical analysis shows to be present. This is particularly true of the kidney. Thus, the amount of fat and lipoids, or lipins, present in normal kidneys (dog) was found to vary between 18.5 per cent, and 29.12 per cent, of the dry weight, the aver- age being 21.8 per cent.; whereas, after producing a typical "fatty degeneration" by means of phosphorus and other poisons, the lipin content was still found to be between 16.9 per cent, and 22.6 per cent.^" In all instances the amount of lipins in kidneys showing typical fatty degeneration under the microscope was found equal to or less than the normal amount — it was never increased. The same conditions were found to obtain in human kidneys that showed fatty metamor- phosis. Microscopic examination of specimens stained with the speci- fic fat stains,'' therefore, gives no indication of the amount of fat 8 Schwalbe (Verh. der Deut. Path. Gesell., 1903 (6), 71) claims that in a sim- ilar way iodin compounds of fat can be demonstrated to be transported into the fatty organs. His analyses were merely qualitative and by quantitative deter- minations 1 was unable to corroborate his conclusions (Zeit. f. physiol. Chem., 1905 (45), 412). ^ A striking proof of the lack of injury associated with fatty infiltration is shown by the fatty infiltration frequently seen in the liver, especially of alcoholics, in which it may be difficult to find, microscopicallj', anj' cell cytoplasm because of the fat, the tissue looking like fatty areolar tissue; and j^et there may be no clinical evidence whatever that the liver fimotion has been impaired by the process. 1" Concerning the normal intracellular fats see introductory- chapter. " Fat-staining involves several principles of interest in this connection. (See reviews by BuUard, Jour. Med. Res., 1912 (27), 55 and Escher, Corrlil. Schweizer Aertze, 1919 (49), 1609.) Osmic acid (OsO^), the longest used for this purpose, is reduced to OsOi by oleic acid, imparting a black or dark-brown color to the fat; but it does not stain staurated fatty acids, such as palmitic or stearic acid. Thus, Christian found in pneumonic exudates fat that stained by other methods but not by osmic acid, apparently because it contained no oleic acid (Jour. Med. Research, 1903 (10), 109). Sudan 111 and scarlet II {fat inmccau) are two sj-nthetic dyes which stain fat in a purely physical way, entering and remaining in the fat-droplets because they are much more soluble in fat than they are in water or alcohol. (Fulh^ discussed by Michaelis (who introduced scarlet R) in Virchow's Arch., 1901 (164), 263; and by Mann, "Physiological Histology," p. 306.) These stains have the advantage of staining all sorts of fats and not staining other substances that may reduce osmic acid. Fatty acids and soaps may be stained with copper acetate, which forms a green copper salt, and thus be distinguished from fats (Benda, Vir- chow's Arch., 1900 (161), 194). J. Lorrain Smith (Jour. Path, and Bact., 1907 (12), 1) has introduced as a fat dye, Nile blue sulphate, which forms a blue salt with free fatty acids, while neutral fats are stained red bj' the oxazone base. 404 RETROGRESSIVE CHANGES contained in a degenerated kidney. A pathologic kidney containing 16 per cent, of lipins ( 18 per cent, is about the average amount in normal human kidneys) may show extreme "fatty degeneration" under the microscope, whereas another kidney may contain as much as 23 per cent, of hpins, yet not show any fat whatever by staining methods. The explanation of this remarkable discrepancy is as follows: Every tissue and organ seems to contain a greater or less amount of lipins, varying from 5 per cent, to 20 per cent, of the total dry weight of the organ in the case of most of the important tissues, yet this is usually held in such a form that it cannot be stained by any stains available for the purpose. Thus in the kidneys, as before remarked, we may have as much as 23 per cent, of lipins present and yet be unable to stain any of it by ordinary methods. The greater part of this seems to be essential to the cell, for it cannot be removed by the most extreme starvation; e. g., the liver of the most emaciated dogs may contain 10 per cent, to 20 per cent, of fatty substances. Furthermore, the same resistance is shown by part of the fat to extraction with ether. A certain proportion of the fat can be extracted readily in twenty- four hours or less by ether, but after this time no more can be made to leave the tissues. Apparently the rest of the fat is held in a com- bination that is insoluble in ether, and a large proportion of this fixed material is not simple fat, but lecithin, cholesterol, and compounds of these lipoids. It has also been demonstrated that fatty acids can combine with amino-acids to form compounds (lipo-peptids) very similar in their properties to these "masked" fats.^^ By digesting the tissue for a short time by pepsin, however, the fixed lipins become freed, so that they can then be readily dissolved out in ether. We see, therefore, that much of the fat of normal cells is so firml}- combined that it cannot be dissolved in ether, and under normal conditions all, or nearly all, of it cannot be stained. (This applies particularly to the parenchymatous organs; the fat of the areolar tissue is all readily extracted — Taylor.) By the use of Ciaccio's method for microscopic demonstration of intracellular lipoids, BelP'' has been able to demon- strate in those cells that are fat-free by ordinary methods sufficient lipoidal material to account for the normal "invisible fat," which is probably identical with the "liposomes." But when pathological changes in the cells result in decomposition of the cell protein through autolysis, or produce physical changes in the colloids that hold the lipins emulsionized, part of this normally invisible fat is set free, and, becoming visible, "phanerosis," ^'^ produces the so-called "fatty degen- eration." This explains the observations of Rosenfeld, cited above, that kidneys may show much fat to the naked eye and microscopically, when they actually contain even less than normal amounts of fat. Tay- 12 Bondi, Biochem. Zcit., 1909 (17), 543. "Internat. Monats. Anat. u. Physiol., 1911 (28), 297; Jour. Med. Res., 1911 (24), 539. '* Klein perer, Deut. med. Woch., 1909 (35), 89. FATTY METAMORPHOSIS 405 lor^^ advanced this explanation, and supported it experimentally by showing that during fatty degeneration this protected fat actually is liberated, some two-thirds becoming ether-soluble in an experiment performed with phosphorus-poisoned frogs. Mansfeld""' also found that in animals poisoned with phosphorus, the proportion of fat which is present in a form free from protein union in both blood and viscera, is increased, while the firmly bound fat is decreased. As further support may be mentioned the fact that organs undergoing experi- mental autolysis show microscopically an apparently typical fatty degeneration, although analyses show that no actual increase in fat occurs. ^^ Relation of Anatomical to Chemical Changes. — From the facts brought out in these various experiments we must consider that the anatomically estabhshed condition of "fatty degeneration" represents either or both of two conditions: (1) It may result from an increase in the normal quantity of fat in an organ undergoing paren- chymatous degeneration, through an infiltration of fat from the out- side; this is particularly true of the fatty degeneration of the hver, presumably because the hver normally receives the relatively saturated body fats to work them over into the more labile desaturated fats; (2) there may be no increase in the total amount of fat, but the invisible fat becomes visible through autolj^sis or hydration changes in the cell proteins. Thus, Bainbridge and Leathes'^ found that after ligation of the hepatic artery there is a marked fatty degeneration of the hver, without an increase in the amount of fat according to analysis. (3) Finally, of course, both factors may occur together. Of these various forms, in only the first would the chemist consider the organ "fatty, " although from a morphological standpoint the second form is entitled to rank as a true "fatty degeneration," and the form that will occur seems not to depend upon the cause of the cell injurj^, but rather upon the organ under consideration. In a studj'- of the relation of the mor- phological to the chemical changes Rosenfeld^^ arrived at the following results : Normal human hearts contain, on an average, 15.4 per cent, of lip- ins; the hearts showing fatty degeneration contain 20.7 per cent, on an average. -° The pancreas, which normally contains 15.8-17.4 per '* Jour. Med. Research, 1903 (9), 59. " Pfluger's Arch., 1909 (129), 63. '' Dietrich, Arb. path. Inst. Tubingen, 1906 (5), H. 3; Hess and Saxl, Virchow's Arch., 1910 (202), 149; Ohta, Biochem. Zeit., 1910 (29), 1; Shibata, ibid., 1911 (31), 321. The significance of the increase of lipins observed in perfused kidneys by Gross and Vorpahl is made doubtful by the article of Underhill and Hendrix. Jour. Biol. Chem., 1915 (22), 471. " Bioehem. Jour., 1906 (2), 25. " Berl. klin. Woch., 1904 (41),587. ^° The amount of phospho-lipins in the heart is usually nearly constant, but alimentary fat may accumulate in the myocardium under certain conditions. See Wegelin, Berl. klin. Woch., 1913 (50), 2125; Bullard, Amer. Jour. Anat., 1916 (19), 1. 406 RETROGRESSIVE CHANGES cent., also contains an increased amount when showing fatt}' degenera- tion. The liver, however, takes on by far the greatest amount of fat after "steatogenetic" poisons, ^^ and the microscopic picture usually gives a very good approximation of the amount of lipins it contains. ^- Apparently in these organs any excessive fat above the normal is observable microscopically, although the normal lipin content is not, and only in these three organs could Rosenfeld find an actual increase in fat after poisoning with phosphorus, etc. It would seem, on the other hand, that there is not often a real increase in the fat content of the "fatty" kidney.^^ Normal spleen contains 14.2 per cent, of lipins, and lung 17.3 per cent., but in both, "fatty degeneration" results in a lowering of this quantity. Degenerations in the nervous tissue, which Virchow considered the best evidence of the conversion of protoplasm into fat, also show a marked decrease in lipins, and voluntary muscle shows no increase in the normal quantity after poisoning. In general, these experiments support the contention of Taylor concerning the disclosure of the invisible fat through autolysis. -■* An explanation of many of the discrepancies lies in the newer studies on The Relation of the Lipoids to Fatty Metamorphosis.-^ — Until within a few years the significance of the intracellular lipoids in fatty degeneration and related processes was not appreciated, beyond the ^1 In fatty livers in phosphorus-poisoning the amount of fat may reach 75 per cent, of the dry weight. Accompanying the fat increase are increase in water and a relative or absolute decrease in proteins, probably due to cell autolysis. In acute yellow atrophy a similar decrease in protein occurs, but without an in- crease in fat. (See v. Starck, Deut. Arch. klin. Med., 1884 (35), 481.) 22 See Helly (Beitr. path. Anat., 1914 (60), 1) who examined 100 human livers which showed all variations in microscopic fat content, and chemically from 7.36 to 74.43 per cent, of lipins (dry weight). He found that there was usually a good correspondence between microscopic appearance and analytic re- sults, altho some marked and unexplained discrepancies were observed. Gener- ally the fat content was from 10 to 30 per cent, of the dry weight, with 19 to 21 per cent, the most common figures. When there is much fat present in the liver the fat content of the bile is increased (Le Count and Long, Jour. Exp. Med., 1914 (19), 234). ^^ This is contradicted by Landsteiner and Mucha (Cent. f. Path., 1904 (15), 752), and by Lohlein (Virchow's Arch., 1905 (180), 1) and Rosenthal (Deut. Arch. klin. Med., 1903 (78), 94), but is supported by Orgler {ibid., 1904 (176), 413), and Dietrich, Verb. Deut. Path. Gesell., 1907 (11), 10. See also the later studies by Rosenfeld on the effects of various steatogenetic poisons on different organs, in Arch. f. Exp. Path. u. Pharm., 1906 (55), 179 and 344. It is probable that the truth lies between the opposing views, namely, the kidney may under some conditions take up fat from tlie blood, but it does so to a much less extent than the liver, and it may sometimes show marked fatty change anatomically without corresponding increase chemically. 2'' Pieces of tissue implanted into animals may show a perii)lieral fatty meta- morphosis or infiltration, yet show upon analysis a decreased fat content (Dietrich, Verb. Deut. Path. Gesellsch., 1905 (9), 212).' ^f* Literature by Leathes, "The Fats," London, 1910; Bang, Ergebnisse der Physiol., 1909 (8), 463, also, "Chemie u. Biochem. d. Lijjoide," Bergniann, Wies- baden, 1911; Kawanuu'a, Virchow's Arch., 1912 (207), 4()9, also "Die Cholester- inesterverfcttung," Fischer, Jena, 1911; Asclioff, Zieglor's Beitr., 1909(47),!, also Festschr. f. IJnna, 1911 p. 23; SchuUz, l<;rgel)nisse d. Pathol., 1909 (XIII;), 253; Ilanes, Bull. Johns Hopkins Hosp., 1912 (23), 77; Anitschkow and Chala- tow. Cent. f. Pathol., 1913 (24), 1. FATTY METAMORPHOSIS 407 fact that in most organs showing fatty changes the quantity of choles- terol and lecithin is not greatly changed. In 1902 Kaiserling and Orgler described under the non-committal name of "myelin" certain intracellular droplets that may be found in the cells of the normal adrenal cortex, and in amyloid kidney's, pneumonic exudates, tumor cells, retrogressive thymus tissue, corpus luteum, and bronchial secre- tions, and which differ from fat in being doubly refractile (anisotropic) when viewed through Nicoll prisms, and in staining but shghtly gray with osmic acid, although taking up other fat stains well. As explained in Chapter i, the myelins are probably mixtures of Upins, cholesterol-esters being prominent, and in many conditions in which fat-like vacuoles are prominent in cells, leading to the diagnosis of fatty degeneration, these substances are responsible, presumably having been liberated from combination with the cell proteins in some cases, in others actually being increased in the cell. This condition, which Aschoff refers to as a cholesterol-ester fatty metamorphosis, is especially seen in the parenchyma cells derived from the urogenital anlage — ^that is, the adrenal cortex, kidnej", testicle and corpus lu- teum. Aschoff states that doubly refractile droplets can be formed by lecithin and phosphatids generally, oleates, cholesterol esters, cho- lesterol v^hen dissolved in phosphatids or fats or fatty acids, as well as by cholesterol esters dissolved in fats. Of these the most im- portant quantitatively is the cholesterol ester group, -*^ and the anal- yses of Windaus have shown that in pathological processes the increase is much greater in the cholesterol esters than in the free cholesterol. Cholesterol compounds stain different!}^ from neutral fats, being more yellow than red with sudan III, and grajash rather than black with osmic acid. Pathologically the anisotropic droplets are also found especially in the above-named tissues, but also in tissues the site of chronic inflammation, including the mucosa of the gall bladder where they may be of importance in the formation of cholesterol concre- tions. They are also found in the alveolar epithelium in pulmonary inflammation, in atheromatous patches in arteries, in many tumors, in most cells," including even the adipose tissues themselves,-^ and occasionally in varied pathological tissues.'^ Perhaps the most con- spicuous deposits are in the epithelium of the "large white kidnej's," and in xanthomas. In Gaucher's disease there is also a remarkable lipoid accumulation in the foamy phagocytic cells. ^'^ According to Munk" true lipoid degeneration always means a serious injury to the cell, but there seem to be many exceptions to this. Indeed, according 26 See also Verse, Ziegler's Beitr., 1911 (52), 1. " Ciaccio, Cent. f. Path., 1913 (24), 50. 28 Cramer, Jour. Physiol., 1917 (51), xi. 29 Pathological decrease in lipoids may also be observed, especially in the ad- renal cortex, usually under the influence of toxic agents; e. g., Hirsch found a marked decrease in delirium tremens (Jour. Amer. Med. Assoc, 1914 (63), 21S6). 30 See Wahl and Richardson, Arch. Int. Med., 1916 (17), 238. 31 Virchow's Arch., 1908 (194), 527. 408 RETROGRESSIVE CHANGES to Anitschkow and Chalatow {loc. cit). the feeding of foods rich in cholesterol may cause the appearance in the liver of great quantities of anisotropic droplets, lipoid deposits in the aorta, enlargement of the adrenal cortex, and the presence in practically all tissues of semifluid, doubly refracting crystalline structures (cholesterol steatosis).^- In cells undergoing autolysis the fat-like "myehn" droplets which appear, differ from the above in not being anisotropic, but are un- doubtedly closely related to them in composition. These "myelin" droplets are also found in cells showing cloudy swelling, presumably representing cell lipoids set free through changes in the cell proteins. They are. characterized by staining with osmic acid but not by sudan III, which shows them not to be simple fats nor yet lipoids, but they are undoubtedly precursors of true fatty degeneration;^^ they prob- ably consist chiefly of lecithin, with more or less free fatt}' acids and relatively little cholesterol (Aschoff). It is possible to distinguish the lipoids of cells, whether normal or pathological, from the neutral fats by means of Ciaccio's method.^* This consists in a preliminary treatment with bichromate, which ren- ders the lipoids insoluble; the tissues can then be hardened and im- bedded by the usual methods which remove the unchromated fats, leaving the lipoids stainable by sudan III. By tliis method BelP° has been able to stain the lipoids in the normal kidney and other tis- sues, in sufficient amount to account for all the so-called "masked fat," which thus seems to be, as also indicated by chemical evidence, largely lipoidal. Jastrowitz^^ has studied the relation of lipoids to fats in the fatty changes produced by various means, and finds that in severe fatty changes with much transported fats there may be little change in the lipoids; with blood poisons which cause little increase in total fats, the lipoid content of both blood and organs may be high; usually the phosphatid content is unchanged or slightly increased, but it may be decreased. The proportion of cholesterol to neutral fats is usually within normal limits in tissues showing fatty changes.^' The mito- chondria seem to be compounds of phospholipins with proteins, and these agglutinate and form fathke droplets in phosphorus poison- ing.38 presumably they play an important role in fatty metamor- phosis. Cells in tissue cultures, however, may take up fat droplets from the surrounding medium {i. e., fatty infiltration), without any association with or changes in the mito-chondria.^' "See also Anitschkow, Deut. med. Woch., 1913 (39), 741; Wesselkin, Vir- chow's Arch., 1913 (212), 225; Rubinstein, Compt. Rend. Soc. Biol., 1917 (80), 191. 33 Hess and Saxl, Virchow's Arch., 1910 (202), 149. 3^ Cent. f. Path., 1909 (20), 771; Arch. f. Zellf., 1910 (5), 235. 36 Jour. Med. Res., 1911 (24), 539. 38 Zeit. exp. Path. u. Ther., 1914 (15), 116. 3' Czyhlarz and Fuchs, Biochem. Zeit., 1914 (63), 131. 38 Scott, Amer. .lour. Anat., 1916 (20), 237. 39 M. R. Lewis, Science, 1918 (48), 398. FATTY METAMORPHOSIS 409 Summary. — We must conclude, tiierofore, that fatty degeneration of an organ means, in the case of the liver, myocardium, and pan- creas, an infiltration of neutral fat from outside into cells which have been degenerated by the action of poisons or other injurious influ- ences, plus a certain amount of apf)arent increase in fat because of the setting free of previously invisible fats and lipoids normally present in the affected cells. In the kidney, spleen, and muscles an increase of fat seldom occurs from these causes, but the cells may show a marked fatty metamorphosis through the setting free of the invisible intracellular fat and lipoids by autolytic or physico-chemical changes. In the adrenal, kidney, and often in other tissues, the fatty material present in the cells is characterized by being doubly refractile, and then consists chiefly of cholesterol esters, together with greater or less quantities of phosphatids, fatty acids, soaps and neutral fats. Pathogenesis of Fatty Metamorphosis Nevertheless, the old anatomical distinction of infiltration and de- generation still remains, provided we do not hold to the original idea that the term degeneration implies that the cell protein has been con- verted into fat; for we must recognize that under some conditions the cells may take up great quantities of fat without suffering any appre- ciable degenerative changes, whereas in other instances the appear- ance of fat is associated with marked and complete disintegration of both nucleus and cytoplasm. Furthermore, we have yet to explain why, under some conditions, the fat is removed from the fat depots to be stored up in the liver or other organs. By applying the commonlj' accepted ideas concerning fat metabolism, a satisfactory explanation seems to be possible. Fat is always utilized and trans- ported in the form of its two constituents, fatty acid (or soaps) and glycerol, which are diffusible and soluble. It enters and leaves the cells in this condition, being split or combined, as may be necessary to produce equihbrium, by the action of lipase, which is present within the cells and in the blood and lymph. Under normal conditions there is little free visible fat in the cells of the parenchymatous organs, because it is largely used up through oxidation of the glycerol and fatty acids by the action of the intracellular oxidases, ^yhe^e there is abundant lipase and but little oxidative activity, as is the case in the areolar fat tissue, fat accumulates in large amounts. When, for any reason, the oxidative power of the parenchymatous organs is re- duced, fat accumulates in them as it does in the fat depots normally, and we have an excess of fat in the parenchymatous cells; thus, in pubnonary tuberculosis, severe or protracted anemias, etc., a great accumulation of fat occurs, particularly in the Kver, where normally active oxidative processes continually balance the action of the abun- dant Hpase of the liver-cells. The liver being normally concerned in the preparation of fat for metabolism, it is also perfectly possible to 410 RETROGRESSIVE CHANGES have an accumulation of fat in the normal liver merely as a result of increased function, and hence fatty changes may be purely physio- logical in this organ. ■^'^ ' If the fat accumulates in cells that are structurally normal or nearly so, the fat-droplets fuse together under the pressure of the cytoplasm, and we get the picture of a typical fatty infiltration; in- deed, the only tissues in which we get this typical infiltration are^the liver and the fatty areolar tissue, in both of which the process is pre- sumably physiological in character even if not always physiological in degree. If the cells are much disintegrated through the action of the poison, — e. g., phosphorus, bacterial toxins, etc., — ^the accumulat- ing fat-droplets are not crowded into one large droplet, but He free in the granular debris of the disintegrating cell, constituting the typi- cal appearance of fatty degeneration. Fatty degeneration is usually brought about by poisons, while abnormal fatty infiltration depends usually upon decreased oxidation, due to lack of either oxygen or hemoglobin in the blood. If the anemia is extreme, however, the cells degenerate, and then we find a true fatty degeneration caused by lack of oxygen. "^^ Thus, in an anemic infarct fat accumulates about the periphery of the dead area,^^ probably because fatty acids and glycerol diffuse in slowly from the surrounding parts where circulation still goes on, and are built up into fat by the cell lipase, for in anemic areas the intracellular oxidases cannot destroy these substances as they nor- mally do, because of lack of oxygen. The accumulation of fat in dead areas depends, therefore, on the fact that the constituents of fat can diffuse into the dead tissue, whereas the oxygen, being held in the cor- puscles, cannot enter the anemic area.^^ It is also possible that where fat is set free by autolysis of dead tissue, or when for any cause free fat or lipoid material is present in the vicinity of living cells, it may be phagocyted or in some way infiltrate the cells, causing a fatty meta- morphosis by absorption (Dietrich). It is to be supposed that poisons also cause fatty degeneration in a similar way — -by interfering with oxidation. We have much evi- dence that in phosphorus, chloroform, and other poisoning associated with fatty degeneration of the liver, oxidation is impaired."*^ If we imagine for a moment, a cell in which oxidation is checked by any means, we shall have in this cell the lipase and the proteol3^tic enzymes not balanced, as they normally are by the action of the oxi- dases, and hence the processes of cell autolysis and of the accumula- " See Coope and Mottram, Jour, of Physiol., 1914 (49), 23; Helly, Beitr. path. Anat., 1914 (60), 1. ^' Mohr (Zcit. exp. Path., 190G (2), 434), denies that oxidation is decreased in anemia; and in a man with but about half the normal lung area the metabolism was not found iiltored to any extent by Carpenter and Benedict, Amer. Jour. Physiol., 1909 (23), 412. « Fisciiler, Cent. f. Path., 1902 (13), 417. " See Griesser, Ziegler's Beitr., 1911 (51), 115. " See Welsch, Arch. int. de pharm. et therap., 1905 (14), 211. FATTY METAMORPHOSIS 411 tion of fat by the lipase will go on uncontrolled. The result will be a disintegrated cell containing many fat-droplets, i. e., fatty degen- eration.^^ In cloudy swelling there also appear droplets stained with osmic acid but not by sudan 111, which Hess and Sa.xl"' have shown to result from intravitam cell autolysis, and to be a precursor of true fatty degeneration. Work with cells in tissue cultures indicates that fatty changes of all types may occur independently of the circulation. Lamberf^ states that the amount of fat in the culture cells is roughly propor- tional to the amount in the culture medium, and cells rich in fat may move actively and undergo normal mitosis. Lewis, however, observed fatty changes in cells growing in fat-free media, and made the espe- ciall}'' interesting observation that cells grown in 2.5-3 per cent, alco- hol will show a rich fat accumulation. Also, an accumulation of fats and lipoids in cells grown in the presence of such steatogenetic poi- sons as phosphorus and Oleu7n pulegii has been observed by others, "'^ which indicates that free cells behave the same under the influence of such poisons as the cells of the fixed tissues. The process of unmasking the masked fats is explained by M. H. Fischer^'' on a physical basis, as follows: The fats of the cells are distributed as an emulsion in a hydration compound of water with hydrophilic colloids, notably proteins and soaps. Such an emulsion breaks down whenever the hydrophilic colloid is either dehydrated or diluted beyond certain ranges. As the usual conditions that cause fatty degeneration, such as poisoning with phosphorus, arsenic, etc., or local circulatory disturbances with local acidosis, all tend to de- hydrate some of the cell colloids and to dilute others, it would seem probable that the appearance of the fat droplets in the cells is the result of such changes in the colloids that previously held them in an emulsion too fine to exhibit readily visible fat particles. The relation of cloudy swelling to fatty degeneration is readily explained on this ^^ Interference with oxidation does not necessarily irnply destruction of the oxidases. As yet we know practically nothing concerning the oxidases of the cells in disease, and the above hypothesis has yet to be demonstrated. Duccheschi and Aluiagia (Arch. Ital. Biol., 1903 (39), 29) found the normal amount of lipase in phosphorus-livers, but also observed no decrease in ability to oxidize salicylic aldehyde, which, however, does not prove a normal power to oxidize fats. Gierke's observation (Ziegler's Beitr., 1905 (37), 502) that glycogen and fat accumulate under identical conditions might be cited as indicating decreased oxidative power. Wells (Jour. Exper. Med., 1910 (12), 607) found that the power of liver tissue to oxidize purines was not decreased by the maximum degree of fatty degeneration, but Waldvogel (Deut. Arch. klin. Med., 1907 (89), 342) found that obese persons can burn fatty acids which arise in metabolism less readily than normal; and Quinan (.Jour. Med. Res., 1915 (32), 73) found the ester-splitting lipolytic enzymes of the liver much reduced in the liver of chloroform necrosis, but the relation of these esterases to true lipases is not known. ^« Virchow's Arch., 1910 (202), 149. *' Trans. Assoc. Amer. Phys., 1913 (9), 93; Jour. Exp. Med., 1914 (19), 398. " Krontowski and Poteff, Beitr. path. Anat., 1914 (58), 407. ^^ Fischer and Hooker, Science, 1910 (43), 468; Fischer, Fats and Fatty Degen- eration, Wiley, New York, 1917. 412 RETROGRESSIVE CHANGES basis, as follows: When a local acid intoxication of a cell occurs, some of the proteins will swell and others will precipitate, resulting respectively in the swelling and cloudiness of the cells characteristic of cloudy swelhng; but at the same time the emulsifying capacity of these proteins will be impaired, permitting the coalescence of the fat droplets and the resulting picture will be that of fatty degeneration. Summary .^ — Fatty metamorphosis involves changes of two kinds. First, infiltration of fat, which occurs when the oxidative power of the cells is decreased, so that fat is not destroyed, but is accumulated from the blood under the influence of the hpase of the cells; if there is not any serious injury to the cells, the histological changes consist in the accumulation of one or a few large droplets of fat in each cell, constituting the condition known anatomically as ''fatty infiltration." This occurs, pathologically, chiefly in the Hver. If at the same time the cytoplasm is disintegrated through autolylic changes, the fat- droplets do not fuse, but remain as small, more or less discrete, fat granules among the granules of cell debris, constituting the micro- scopic picture of "fatty degeneration"; this condition occurs particu- larly in the heart and hver. Second, each cell contains a large amount of fat and hpoids (5-25 per cent, of its dry weight), which is so combined that it cannot be detected microscopically; this may be hberated during the autolytic processes and colloidal changes of cell disintegration and become visible, constituting a macroscopical and microscopical degeneration, but without any actual increase in fat — this condition occurs particu- larly in the kidney and nervous system. Third, a combination of both of the above processes — infiltration of fat and liberation of masked intracellular fat — may occur simultaneously in an organ. *^ Fourth, in certain cells, especially in the kidney, adrenal, ovary and some tumors, there may be a great increase in the lipoids of the cell, ''lipoidal degeneration," and especially of cholesterol esters and free cholesterol, part of which is infiltrated and part set free from com- bination in the cytoplasm. PROCESSES RELATED TO FATTY METAMORPHOSIS ADIPOCERE This apparent transformation of the substance of dead bodies into a wax-like material was for a long time looked upon as evidence of a transformation of protein into fat, but in the light of more recent in- vestigations this view can hardly be held. Adipocere is the product of a process that occurs particularly in bodies buried in very wet ^' The above conception of the processes involved in fatty metamorphosis is more fully discussed by the writer in other publications (Jour. Amer. Med. .\ssoc., 1902 (38), 220; ibid., 1906 (46;, 341). Ribbert (Deut. med. Woch. 1903 (29), 793) has also advanced a similar explanation for the morphological differences between fatty "degeneration" and "infiltration," i. e., that the degenerative changes are independent of fatty accumulation. ADIPOCERE 413 places or lying in water, and results in an apparent replacement of the muscles and other soft parts (but not the glandular organs) by a mass consisting of a mixture of fatty acids in crystaUine and amor- phous form, and soaps, particularly ammonium, magnesium, and calcium salts of palmitic and stearic acid (the oleic acid largely disap- pearing during the process). Analysis of samples of adipocere by Ruttan^- gave the following figures: Composition of Human and Pig's Adipocere I II III Pigs Human Human (mature) hard soft 94.1 82.9 75.8 0.8436 0.8397 0 8410 1.436 1.437 1.439 60-63° 52-54° 50-51° 201.7 207.3 203.8 207.0 211.0 212.2 6.04 9.65 12.52 34.75 11.8 271.0 266.0 264 0 70.82 71.78 59.2 5.24 8.87 11.6 14.80 8.24 7.8 1.21 0.91 0.90 0.16 0.15 0.83 0.87 0.69 0.75 4.41 6.76 12.3 0.665 1.93 4.14 0.035 0.054 0.574 i.99 2 25 Ether soluble, per cent Specific gravity at 100° C. . . Refractive index 65° C Melting- point Acid value Saponification value Iodine value Acetyl value Mean molecular weight Saturated fat acids, per cent Unsaturated fat acids Hydroxy fat acids Stearin and palmitin Olein Unsaponified matter Calcium soaps Protein Ammonia Ash Ammonium and other soluble soaps were absent. The hydroxy- stearic acids, w^hich are so characteristic of adipocere, are formed from the oleic acid of the original triolein. Cholesterol has also been found in adipocere. ^^ The resulting material is absolutely resistant to putrefaction, and hence remains intact for many years. This replacement of the soft parts is, however, only apparent, for the total weight of a body in this condition is much Hghter than that of the original body; indeed, one is always surprised at the hght weight on Hfting such a specimen. Adipocere occurs almost exclusively in fat bodies, and it seems probable that all the soaps and fatty acids found are formed from the original fats of the corpse.^^ These gradually flow into the places left by the disintegrating muscle, etc., a process that occurs readily in cadavers, according to Zillner;"^ or the infiltration may be accomplished through "Jour. Biol. Chem., 1917 (29), 319; Trans. Rov. Soc. Can., 1916 (10), 169. *' Van Itallie and Steenhauer, Pharm. Weekblad, 1917 (54), 121. " Fatty changes in the viscera may favor their transformation into adipocere (Muller, Vierteljahrs. gericht. Med., 1915 (50), 251). " Vierteljahrsch. f. gericht. Med., 1885 (42), 1. 414 RETROGRESSIVE CHANGES diffusion of the ammonium soaps formed during the decomposition. As the subcutaneous fat is hardened by the formation of soaps, and the bones remain to hold the parts in position, the general form of the body is preserved, creating the impression that its entire substance has been converted into adipocere, when the total mass may actually weigh but twenty pounds or so, and, according to Zillner's estimate, not more than one-tenth of the muscle substance is replaced by adi- pocere. This false impression is probably responsible for much of the mistaken idea concerning the conversion of tissue proteins into fatty acids. Thus, Schmidt^'' found that in early Egyptian mummies 60 per cent, of the weight of the lungs and 30 per cent, of the spleen con- sisted of fatty acids, and fell into the usual error of considering this conclusive evidence of transformation of proteins into fat. Numerous attempts have been made to prove that muscle could be thus converted into fatty acids and soaps, but although success has been claimed by a few, the results are not entirely convincing. ^^ Bacteria can convert proteins into fats, beyond a doubt, and they may do so to some slight extent in adipocere formation, but probably this factor is not important. In the light of our present conception of fat metabohsm it is prob- able that the process of adipocere formation occurs as follows: The fatty acids of the fat tissue are combined by the ammonia formed during putrefaction, removing these fatty acids from the normal balance of fat and fatty acids in the fat tissue; as a result, the lipase of the fat tissue continues to split the fat, and more fatty acids are produced, which likewise go to form soaps. This continues until practically all the neutral fat has been decomposed, the glycerol dif- fusing rapidly away. The soluble soaps, which the bacteria do not attack, diffuse into the softened muscle tissue, which they gradually replace in part. In the meantime, from the more soluble ammonium soaps, calcium and magnesium soaps are being slowly formed, accord- ing to the usual rule of double decomposition (that the least soluble salt will be formed under such conditions) ; or else, if an acid reaction develops, free fatty acids are precipitated. The oleic acid seems to be converted into the higher fatty acids (Salkowski).-""* It is also possible that the saponification is due to the gradual action of the alkaline fluids produced in decomposition of the tissues, or to the alkalinity of the water in which the body lies. Possibly bacteria may be re- sponsible for this decomposition of the fats rather than the body lipase, for Eijkman^^ has observed that certain bacteria growing in fat-containing agar produce calcium, ammonium, and sodium soa[is, simulating adipocere. ^'^ " Zeit. allg. Physiol., 1907 (7), 3G9. " See Rosenfeld, Ergcb. dor. Physiol., Abt. 1, 1902 (1), 659. " Festschr. f. Virchow, 1S91, p. 23; corroliorated by Schiitze. "Cent. f. Bakt., 1901 (29), 847. «»See also Ccvidalli, \'it>rteljahrschr. froiifhtl. Mod., 190G (32), 219; and Schiitze, Arch. Ilyg., 1912 (70), 110. LIPEMIA 415 Zillner** gives the following scheme of the changes that take place in a cadavrr undergoing aflipocorc formation: (1) Migration of fluid contents of the body (imbibition of blood and transudation) — one to four weeks. (2) Decomposition of superficial epidermis, then of corium — ^first two months. (3) Decomposition of muscle and gland parenchyma, until only the inorganic part of the bones and the con- nective and elastic tissues remain — three to twelve months. (4) Migration of neutral fat, crystallization and partial saponification of the higher fatty acids in the panniculus; transformation of the blood pigment into crystalline form — four to twelve or more months. ^^ LiPEMIA Normally the blood contains a considerable amount of fats and lipoids, varying somewhat, but not greatly, with the diet. The older literature gave figures varying widely, but analyses by more modern methods"- give figures for the ether-soluble constituents of the normal plasma (before breakfast) ranging ordinarily from 0.57 to 0.82 per cent., of which cholesterol and phosphatid"'' are about equal (0.2 to 0.3 per cent.) with very httle neutral fat (0.1 to 0.2 per cent.). In various diseases, exclusive of diabetes, the total lipin content was found by Bloor to be about normal, but the proportion of the different lipins varied somewhat. After taking fat-rich food, however, there may be a considerable excess of the food fats in the serum, and it is, there- fore, extremely difficult to say just when the amount of fat in the blood is large enough to be considered as a lipemia, especially since after every fatty meal there is enough fat in the blood to make it turbid.'"'^" B. Fisher"* states that we may speak of a pathological lipemia when we have a distinctly cloudy blood or serum, which is clarified by shaking with ether through the dissolving out of fat which can then be sepa- rated from the ether. We may, however, sometimes find turbid plasma with normal hpin content and clear plasma with hyperlipemia (Gray). Earlier writers described, incorrectly, lipemia in many con- ditions, but recent writers mention it chiefly as occurring in alcoholism"^ and diabetes. By far the greatest amounts of fat are observed in the latter condition, and diabetic lipemia is always accompanied by an acidosis, although acidosis often occurs without lipemia. Experi- ^* Sclerema neonatorum is caused by hardening of the subcutaneous fat, perhaps because of a low proportion of oleic acid. (Beyer, Verh. Deut. Path. Gesell., 1908 (12), .305.) C. S. Smith, however, found normal oleic acid but a high figure for free fatty acids. Others have described high melting points for the fat, believing the condition to be merely an exaggeration of the normally high proportion of palmitic and stearic acids of infant fat tissues (Smith, Jour. Cut. Dis., 1918 (36), 436). «2 Bloor, Jour. Biol. Chem., 1916 (25), 577. ^^ Concerning blood lecithin see Feigl, Biochem. Zeit., 1918 (90), 361. ®^" Neisser and Brauning, Zeit. exp. Bath. u. Ther., 1907 (4), 747. " Virchow's Arch., 1903 (172), 30. R6sum6 and complete literature. ^^ Also occurs in 'experimental alcoholism (Feigl, Biochem. Zeit., 1918 (92), 282; Bang, ibid., (90), 383). 416 RETROGRESSIVE CHANGES mental pancreatic diabetes may be accompanied by lipemia.^^ In- creases in the blood lipoids, not usually of sufficient magnitude to cause a distinct lipemia, may be found in nephritis (Bloor), cirrhosis, tabes and paralysis (Feigl).^^ Neisser and Derlin^^ found 19.7 per cent, of fat in the blood of a patient with diabetic coma (after death 24.4 per cent, was found) whose urine contained 0.8 per cent, of fat, and through analysis of this and other material came to the conclusion that the fat comes directly from the chyle; i. e., that it is food fat, not body fat. Fischer found an average of 18.129 per cent, in his case, including at least 0.478 per cent, of cholesterol, with no lipuria and very small amounts of fatty acids; of the fat, about 67.5 per cent, was olein. Ringer®^ has found 14.4 per cent, of lipins, including 2.14 per cent, of cholesterol. As high as 27 per cent, of fat has been found in the blood.''" In many cases the increase is chiefly in the lipoids, liyoidemia,"^^ and in acidosis there is said to be an especial increase in cholesterol i^Adler).''^ Study of a large number of diabetic bloods by Gray'^^ gave the following results: Normal lipin values are seldom found, the most marked increases being in the total glycerides, next the total fatty acids, then the cholesterol, and least the phospholipins. Increase of both cholesterol and glycerides seems to be pathognomonic of chronic diabetic lipemia, as in alimentary hpemia the increase is in the fatty acids. The greater the duration of the diabetes the lower the lipins, and high figures give a bad prognosis, being usually associated with acidosis. Hyperglycemia and hyperlipemia do not run parallel.^* In general, the amount of blood lipins increases with the severity of the disease, ^^ the averages in a large series of analyses being as follows: normal, 0.59 per cent.; mild diabetes, 0.83; moderate, 0.91; severe, 1.41. The changes concern chiefly the plasma. Coexistent nephritis does not modify the blood hpin figures. When the lipemia is accompanied by icterus the fats may clear up and a clear serum is present, despite a high fat content. ^"^ It is an important question whether, with high quantities of fat in the blood, fat embolism may result, for it is possible that at least some of the cases of diabetic coma are due to such fat embolism in the cerebral vessels. Ebstein^^ considers this a possible, but not a common, occurrence, because the droplets are too small to cause oc- " Seo, Arch. exp. Path. u. Pharm., 1909 (Gl), 1. " Biochem. Zeit., 1918 (88), 53; (90), 1. «» Zeit. klin. Med., 1904 (51), 428. "o Proc. Soc. Exp. Biol. Med., 1917 (15), 40. ^» Frugoni and Marchetti, Bed. klin. Woch., 1908 (45), 1844. ^1 See Weil, Munch, med. Woch., 1912 (59), 2096. " Berl. klin. Woch., 1910 (47), 1323. " Boston Med. Surg. Jour., 1917 (178), 16. '* Corroborated by Bang, Biochem. Zeit., 1919 (94), 359. " See Jour. Amer. Med. Assoc, 1917 (69), 375. '8 Feigl and Querner, Zeit. exp. Med., 1919 (9), 153. " Virchow's Arch., 1899 (155), 571. LIPEMIA 417 elusion of the vessels unless they conibine to form large droplets. Fischer doubts if the droplets ever fuse together enough to cause em- bolism, supporting his contention both by experiments and clinical records, but cases have been reported as fat emboUsm from diabetic lipemia.'^^ The cause of lipemia has not yet been satisfactorily determined. In alcohohsm it is commonly ascribed to a failure to burn fat, because of the presence of the more readily oxidized alcohol, and the common coexistence of diabetes and hpemia suggests for both a common cause; i. €., lack of oxidation of fat and sugar. In corroboration may be cited the occurrence of lipemia in other conditions associated with defective oxidation; i. e., pneumonia, anemia,^'-* phosphorus-poisoning. As we are still unfamiliar with the essential factors and steps in the oxidation of fat, it would be mere speculation to attempt to explain further the reason for the failure of destruction of the fat. The origin of the fat in hpemia is likewise undetermined. Ebstein considers that it arises partly from the food, partly from fatty degeneration of the cells of the blood, the vessel-walls, and the viscera. Neisser and Derlin consider it as merely food fat coming from the chyle and accumulated in the blood. Fischer beheves that it is largely derived from the fat depots, and that because of loss of the Hpolytic power of the blood it cannot be rendered diffusible, and hence it cannot enter the tissues where it is normally consumed. Sakai^" also found a low lipase content in the blood and suggests that fat entering the blood is unable to leave it because of defective hpolysis. Klemperer and Um- ber hold that it comes from disintegration of tissue cells, but are unable to determine the cells concerned. Ervin^^ attributes diabetic hpemia to the glycogen deficiency of the cells, assuming that glycogen acts as a protective colloid which holds the intracellular fats in emulsion. Bloor's studies^- support strongly the view that the fats come from the food, for he found hpemia only in diabetics receiving fat in their food, and under fasting an existing hpemia disappears. Choles- terol increases parallel with the fat, while lecithin is relatively little increased. Verse^^ says that a lasting lipemia can be produced by feeding rabbits mixed cholesterol and oil, but not with either of these alone. In severe diabetes without hpemia the hpins are all much increased in the plasma, but with the relative proportions about as in normal individuals, although with a tendency for the fats to accumu- late in excess. The facts that fat oxidation depends upon carbohy- drate oxidation, and also that in diabetics excessive fat feeding is "8 Hedren, Svenska Liik. Handl., 1916 (42), 933. " See Boggs and Morris (Jour. Exper. Med., 1909 (11), 553), who produced lipemia bv repeatedly bleeding rabbits. 8" Biochem. Zeit., 1914 (62), 387. " Jour. Lab. Clin. Med., 1919 (5), 146. " Jour. Biol. Chem., 1916 (26), 417; 1917 (31), 575. 83 Miinch med. Woch., 1916 (63), 1074. 27 418 BETROGRESSIVE CHANGES usual, are probably significant in the causation of diabetic lipemia. (See also cholesterolemia.) Pathological Occurrence of Fatty Acids Fatty acids occasionally occur free in pathological processes. The best example of this is fat necrosis (q. v.)., where crystals of fatty acids appear in the necrotic fat-cells, arising through splitting of fat. and later becoming combined with calcium from the blood. Similar crystals, consisting of a mixture of palmitic and stearic acids, fre- quently called margarin or inargaric acid crystals, may be found in decomposed pus, in sputum from bronchiectatic cavities and from gangrene of the lungs, in gangrenous tissue, and in atheromatous areas. According to Schwartz and Kayser^'* the free fatty acids, at least in pulmonary gangrene, arise from Hpolysis by bacterial action rather than by the hpase of the tissues. Eichhorst found crystals of fatty acids in the neighborhood of acute patches of sclerosis in the central nervous system in multiple sclerosis, and McCarthy*^ found them in a spinal cord undergoing secondary degeneration from compression. Whipple^^ describes a case with deposits of fatty acids and neutral fat in the wall of the intestine and the mesenteric glands, while soaps and fatty acids are said to be present in excess in chronic appendicitis.^^ Soaps and fatty acids, especially oleic acid and oleates, are highly toxic, and their profound hemolytic power has been thought of im- portance in pathological conditions, especially bothriocephalus anemia. ^^ (See Hemolysins, Chap, ix.) The fatal dose of sodium oleate for rabbits is 0.15 gm. per kilo (Leathes). The salts of higher fatty acids above capric are hemolytic, while those from caproic down are not, nonoic acid salts being the turning point (Shimazono).^^ The toxicity of soaps may be related to their marked power to inhibit proteolytic enzymes.^" The fatty acids may be stained green by copper acetate, according to Benda's method, and if then treated with hematoxylin, they turn black. ^^ With Nile blue sulphate they stain blue, forming a blue salt, while the neutral fats are stained red by the oxazone base (J. L. Smith). Fischler and Gross^- state that fatty acids are present in atheromatous areas and about the margin of anemic infarcts, but are not recognizable by this method in such fatty degenerations as pneumonic exudates, caseation, etc. Klotz^^ considers that calcium «< Zcit. klin. Med., 1905 (56), 111. "Univ. of Penn. Med. Bull., 1903 (IG), 141. «« liull. Johns Hopkins Hosp., 1907 (IS), 382. "Anthony, Jour. Med. Res., 1911 (20), 359. '* Faust, Suppl. Bd., Schiuiedebcrp's Arch., 1908, p. 171. 89 Z. Immunitat., Ilef., 1911 (4), 650. "".lobling and Petersen, Jour. Exp. IMod., 1914 (19), 251. «' Fischler, Cent. f. Path., 1904 (15), 913. 82 Ziegler's Beitr., 1905 (7th suppl.), 343. " Jour. E.\p. Med., 1905 (7), 633. rATIIOLoaiCAL OCCURRENCE OF CHOLESTEROL 419 soaps are forinod as the fiist step in pathological calcification, accord- ing to microchciuical evidence; but a chemical investigation of the same question did not give the writer positive results.'-" In fatty cells, especially in the liver, crystals are often found and interpreted as fatty acids, which are really crystals of neutral fats."^ Pathological Occurrence of Cholesterol" Cholesterol in crj^stals is found under somewhat the same conditions as the fatty acids, and although cholesterol is not a fat, but an alco- hol, its ph3^sical properties are so similar that it may be considered in this place. (See "Gall-stones," Chap, xvii, for further discussion.) The characteristic large flat plates of cholesterol may be found in any tissue in which cells are undergoing slow destruction, and where absorp- tion is Y>oor. Therefore, they are found frequently in atheromatous patches in the blood-vessels, encapsulated caseous areas, old infarcts and hematomas, inspissated pus-collections, dermoid cysts, hydrocele fluids, etc.; especially large amounts occur in the cholesteatomatous tumors of the ear and cranial cavity. ^^ In degenerative conditions of the central nervous system^^ choles- terol may be present in the spinal fluid (Pighini''^), and in an old pleural effusion as much as 3 to 4 per cent, of cholesterol has been found^ (See Pleural Effusions, Chap, xiv.) Windaus- found that normal aortas contain about 0.15 per cent, cholesterol, while in two atheromatous aortas he found 1.8 per cent, and 1.4 per cent., the increase being more in the cholesterol esters than in the free cholesterol. Amyloid kidneys, however, show an increase only in the cholesterol esters, and not at all in the free cholesterol. (See Relation of Lipoids to Fatty Metamorphosis, p. 406.) Ameseder^ found that 28.56 per cent, of the ether extract of atheromatous aortas was cholesterol. Tlie claim of Chauffard that arcus senilis, xanthelasma, and other ocular condi- tions depend on cholesterol deposition is not substantiated by Mawas'* but Verse^^ observed corneal opacity in rabbits fed cholesterol and oil. In liquids the crystals form glistening scales; in fresh tissues they may be recognized by their solubility in ether, cholorform, hot alcohol, etc., and by their color reactions. In histological specimens prepared by the usual methods the cholesterol is dissolved out, but the resulting clear-cut clefts are quite characteristic. In fresh specimens in which 9* Wells, Jour. Med. Research, 1906 (14), 491. 95 Smith and White, Jour. Path, and Bact., 1907 (12), 126. 9^ Concerning the chemistry of cholesterol see introductory chapter. 9^ See Bostroem, Cent. f. Path., 1897 (8j, 1. '^ Southard has described cholesterol concretions up to 2 cm. diameter in the brain and cord. (Jour. Amer. Med. Assoc, 1905 (45), 1731.) 99 Riforma Med., 1909 (25), 67. 1 Ruppert, Miinch. med. Woch., 1908 (55), 510; Zunz, Hedstrom, and others (see Chap. XIV). 2 Zeit. physiol. Chem., 1910 (67), 174. ^Zeit. physiol. Chem., 1911 (70), 458. * Monatsbl. f. Augenheilk., 1912 (13), 604. 420 RETROGRESSIVE CHANGES cholesterol crystals are present, on treatment with five parts concen- trated sulphuric acid and one of water, the edges of the crystals be- come carmine red, then violet. Concentrated sulphuric acid plus a trace of iodin colors the crystals in sequence, violet, blue, green, and red. Hirschsohn^ recommends a reaction with a 90 per cent, solution of trichloracetic acid in HCl, which gives red, then violet, then blue. The results of microchemical examination are said not to agree at all quantitatively with analytic results.*^ Since all cells contain cholesterol,^ it is perhaps accumulated as one of the least soluble products of their disintegration. The origin of the normal cell cholesterol is unknown, but that which is liberated by normal disintegration of cells seems to be retained and worked over.^ It is not destroyed during autolysis.^ Cholesterol is generally con- sidered, but without convincing proof, to be a product of protein de- composition; if this is true, then the cholesterol found in disintegrating tissues may be formed from the cell proteins during their decomposi- tion.^'' Apparently cholesterol crystals may be slowly removed, the chief factor probably being the giant-cells that are often found surrounding them,^> and the large "foamy" endothelial cells that take up especially the uncrystallized cholesterol. In general they behave as inert foreign bodies. Xanthomatous masses of various kinds all seem to be composed of deposits of cholesterol esters which lead to proliferative and phagocytic reactions in the fixed tissues.'^ Cholesterolemia.^' — Normal blood contains 0.16 to 0.17 per cent. (Gorham and Myers) of cholesterol, of which about 55 per cent, is in the corpuscles, but in pathological conditions the amount in the plasma varies greatly (Bacmeister and Henes)." Cholesterol-rich diet causes a slight increase,'^ but a more marked increase is said to be obtained in pregnancy," nephritis, early arteriosclerosis, obesity, diabetes, and obstructive but not in hemolytic jaundice." According to some observations, in nephritis the amount of cholesterol bears no relation to the albuminuria, and in uremia it may be low; acute febrile diseases usuallj' show a lowered cholesterol, which is unchanged in tuberculosis. Stapp^* describes 6 Pharm. Centralhalle, 1902 (43), 357. « Thavscn, Cent. allg. Pathol, 1015 (2()\ 433. 7 See Doree, Biochem. Jour., 1909 (4), 72. 8 Ellis and Gardner, Proc. Royal Soc, London, 1912 (84), 461. 3 Corper, Jour. Biol. Chem., 1912 (11), 37; Shibata, Biochem. Zeit., 1911 (31), 321. '" Of historical interest is Austin Flint's idea that cholesterol in the blood is an important factor in intoxications, especially in icterus (Amer. Jour. Med. Sci., 1862 (44), 29). All recent evidence is to the effect that cholesterol is not toxic. " See LeCount, Jour. Med. Research, 1902 (7), 160; Corper, Jour. Exp. Med., 1915 (21), 179; Stewart, Jour. Path, and Bact., 1915 (19), 305. 12 Literature given by Rosenbloom, Arch. Int. Med., 1913 (12), 395. > 3 Bibliography by Dewey, Arch. Int. Med., 1916 (17), 757; Gorham and Myers, Arch. Int. Med., 1917 (20), 599; Pacini, Med. Record, 1919 (94), 441. " Deut. med. Woch., 1913 (39), 544. " See Luden, Jour. Biol. Chem., 1916 (27), 257. 1^ The blood of the fetus corresponds closely to that of the mother in respect to free cholesterol but contains no cholesterol esters. (Slemons and Curtis, Amer. Jour. Obst., 1917 (75), 569.) " Rothschild and Felson, Arch. Int. Med., 1919 (24), 520. "Deut. Arch. klin. Med., 191S (127), 439; corroborated by Epstein, Amer. Jour. Med. Sci., 1917 (154), 638. AMYLOID 421 marked cholesteroleinia as accompanying severe parenchymatous nephritis, but not cluonic interstitial typos. K'ollert and FinKor'"" state that hypercholestero- lemia up to ().2S per cent, may be found, but only when the kidney is excreting lipoids, and believe that the cholesterol is at least partly responsible for albuminuric retinitis. Bloor,^" however, found no change in the blood cholesterol in nephritis. The blood content has been reported as low in febrile cutaneous di.seases, but high in afebrile cutaneous diseases associated with eosinophilia.^" However, DiMiis^' states, after examination of a largo nundier of cases, that hypercholestero- loinia was found only in diabetes, and that low cholesterol values are found in cachexia or prostration, but are not characteristic of anj' particular disease. In Japan low values have been observed in beriberi, high in homiplegia.^'* Numerous investigators have described hypercholesterolemia in patients with gall-stones (7. V.) and attribute a causal relation thereto. The importance of the cholesterol of the blood in hemolysis and protection therefrom has been discussed under that subject; in anemia there is usually hypocholesterolemia. Experimental hypercholesterolenua in animals leads to a deposition of choles- terol in various organs, especially the aorta,*' kidneys and liver, accompanied by degeneration in the parenchymatous structures, and excretion of cholesterol in the urine and bile; gall-stones may be formed (Dewey). Sometimes lipoid-filled endothelial cells become so abundant in the spleen as to resemble Gaucher's dis- ease (Anichkov, McMeans**). Excessive cholesterol in the blood reduces phago- cytic activity and antibody formation in experimental animals.*^ Robertson be- lieves cholesterol to have an accelerative action on cancer growth, related to its hydroxyl radical, '^^ but in cancer patients there seems to be no cholesterolemia (Denis). ^^^ The ratio of free cholesterol to cholesterol esters in normal human blood is nearly constant, the esters being about 33.5 per cent, in the blood and 58 per cent. in the plasma; in preganncy and during fat absorption the proportion of choles- terol esters is high, in cancer and nephritis it is low.^^ The blood of the fetus contains free cholesterol but no cholesterol esters. AMYLOID28 Virchow, in 1853, made the first study of the nature of the substance characteristic of ''lardaceous" degeneration, and considered it to be a sort of animal cellulose, because it often became blue if treated with iodin followed by sulphuric acid. To this resemblance in staining reaction we owe the unfortunate, misleading, but generally used, name amyloid."^ It was but a few years (1859) before Friedreich and »»" Mtinch. med. Woch., 1918 (65), 816. 1^ Jour. Biol. Chem., 1917 (31), 575; also Kahn, Arch. Int. Med., 1920 (25), 112. 2" Fischl, Wien. klin. Woch., 1914 (27), 982. " Jour. Biol. Chem., 1917 (29), 93. 22 Bull. Naval Med. Assoc, Japan, Feb., 1919. " See Adler, Trans. Assoc. Amer. Phys., 1917 (32), 255. 24 Jour. Med. Res., 1916 (33), 481. The material in the cells in Gaucher's disease is perhaps a protein-phosphatid compound (Mandelbaum and Downey, Fol. Hematol., 1916 (20), 139. 2^ Dewey and Nuzum, Jour. Infect. Dis., 1914 (15), 472. " Jour. "Cancer Res., 1918 (3), 75. ^^^ Luden reports an increase (Jour. Lab. Clin. Med., 1916 (1), 662. " Bloor and Knudson, Jour. Biol. Chem., 1917 (29), 7; 1917 (32), 337. -8 General literature to 1893, see Wichmann, Ziegler's Beitr., 1893 (13), 487; also Lubarseh, Ergeb. allg. Path., 1897 (4), 449; discussion in the Verh. Deut. Path. Gesellsch., 1904 (7), 2-51; Davidsohn, Virchow's Arch., 1908 (192), 226, and Ergebnisse allg. Path., 1908 (12), 424. -^ In view of the fact that this substance is chemically related to chondrin, and that it also closely resembles this substance physically, it has seemed to the writer that the name '"'chondroid" would be much more appropriate than any of the many more or less misleading and inappropriate titles that are at present in use. The very multiplicity of these terms, however, prohibits any attempt to introduce still another. A particularly unfortunate source of confusion exists in the use of the name amyloid for a vegetable substance, formed by the action of acids upon cellulose. 422 RETROGRESSIVE CHANGES Kekule showed that the substance in question was of protein nature; their methods were very crude, but the main fact was soon better substantiated by Klihne and Rudneff (1865). Krawkow,^'' however, in 1897 gave us the first good idea of the composition of amyloid sub- stance through his ampKfication of Oddi's^^ observation that amyloid organs contain chondroitin-sulphuric acid, finding that amyloid is a compound of protein with this acid, similar to nucleoprotein, which is a compound of nucleic acid and protein. This work has received general acceptance, although a later paper by Hanssen^^ reports a study of amyloid material isolated in pure condition from sago spleens by mechanical means, which contained no chondroitin-sulphuric acid, although the amyloid organs taken in toto do contain ah excess of sul- phur as sulphate. This important contradiction to prevailing ideas has not, so far as I can find, been subjected to investigation by others, with the exception of a casual remark by Mayeda^^ that a prepara- tion of amyloid which he had made did not yield sulphuric acid. Chondroitin-sulphuric acid, which has been studied especially by Morner and by Schmiedeberg,^^ has the formula C18H27NSO17, according to tlie latter, and yields on cleavage chondroitin and sulphuric acid, as follows. Ci8H27NSOn+H20 = C18H27NO14+H2SO4 Kondo,^^ however, gives it an empirical formula of C15H27NSO16, there being ap- parently two equivalents of the base for each SO4 group. Levene and La Forge'^ have demonstrated that chondroitin-sulphuric acid consists of sulphuric acid, acetic acid, chondrosamine which is an isomer of glucosamine, and glucuronic acid. It unites with histones and forms a precipitate.^^ Chondroitin is a gummy substance which in turn may be split into acetic acid and a reducing substance, chondrosin. Chondroitin-sulphuric acid is the characteristic component of car- tilage, but it is also found in the walls of the aorta and other elastic structures (Krawkow). It has also been found in a uterine fibroma and in bone tissue by Krawkow, but could not be found in the parenchymatous organs, normal and pathological, or in chitinous structures. Morner has also found it in a chondroma. Chemistry of Amyloid. — Krawkow separated amyloid from nu- cleoprotein, to which it is most closely related, by dissolving both sub- stances from the minced amyloid organs with ammonia, precipitating with acid, and then taking up the amyloid with Ba (OH) 2 solution, in which the nucleoprotein does not dissolve. Amyloid thus isolated is a nearly white powder, which is easily soluble in alkalies, but slightly in_^acids, and is very resistant to pepsin digestion. The elementary composition was found by Krawkow to be approximately as follows : C = 49-50%; H = 6.65-7%; N = 13.8-14%; S = 2.65-2.9%; P in traces only. ^» Arch. exp. Path. u. Pharm., 1897 (40), 196. 31 Ibid., 1894 (33), 377. 32 Hiochem. Zcit., 1908 (13), 185. " Zeit. physiol. Chem., 1909 (58), 475. 3* Morner," Skand. Arch. Physiol., 1889 (1), 210; Zcit. physiol. Chem., 1895 (20), 357, and 1897 (23), 311; Sciimiodeberg, Arch. exp. Path. u. Pharm., 1891 (28), 358. Hoc also Levene and La Forge, Jour. Biol. Chom., 1913 (15), 09 and 155; 1914 (18), 123. •••f' Biocliem. Zeit., 1910 (2G), 116. 3« Pons, Arch, internat. physiol., 1909 (8), 393. AMYLOID 423 Quite similar analytic results have been obtained by Neuberg,'^ who corroborated Krawkow's finding of a body of apparently similar composition in tlie normal aorta. Neuberg has studied especially the protein constituent of the amyloid compound, and found it character- ized by a high proportion of diamino-nitrogen,"^ as compared with most proteins, as shown in the following table giving the percentage of the total N contained in each of the three forms, amid-nitrogen (ammonia), monamino-acids, and diamino-acids: Tahle I Monamino- acid nitrogen Diamino- acid nitrogen Amid nitrogen Liver amyloid . . . . Spleen amyloid . . . Aorta "amyloid". Gelatin Casein 43.2 30.6 54.9 62.5 76.0 51.2 57.0 36.0 35.8 11.1 4.9 11.2 8.8 1.6 13.4 The variations in the composition of the different amyloids, as shown in the above table, indicate that the protein group may vary in different organs in different cases, and also indicate that the "amyloid- like" substance of normal vessels is not the same as the pathological substance. Corresponding variations were found in the apportion- ment of the sulphur between that which is in the form of oxidized sul- phur and the unoxidized sulphur. The proportion of the different amino-acids in the protein constituent of amyloid is strikingl}' like that of thjmius histon, and entirely dissimilar to the apparently closely related elastin, as shown by Table II. Table II Cleavage products (in percentages) Amyloid Elastin Thymus histon Glycocoll 0.8 ! 25.8 Leucine 22.2 j 45.0 Glutaminic acid 3.8 0.7 Tyrosine 4.0 i 0.3 a-Proline 3.1 1.7 Arginine - 13.9 0.3 Lj'sine 11.6 I .... 0.5 11.8 3.7 5.2 1.5 14.5 7.7 "Verh. Deut. Path. GeselL, 1904 (7), 19. ^^ Corroborated by Jackson and Pearce (Jour. Exp. Med., 1907 (9), 520), but not by Mayeda (Zeit. physiol., Cliem., 1909 (58), 469), who found histidine, which Neuberg had missed, and a lower arginine and lysine content than histon re- quires. 424 RETROGRESSIVE CHANGES This carries out the resemblance of amyloid to nucleoproteins, and, likewise, Neuberg found amj-loid very slowly digested bj^ pepsin, and much better by trypsin, although less rapidly than simple protein; it is also destroyed by autolytic enzymes, for amyloid tissues readily undergo autolysis. ^^ Neuberg considers, from the above results, that amyloid is probably a transformation-product of the tissue protein, similar to the transformation of simple proteins into protamins that occurs in the testicle of spawning salmon as they go up the streams, as shown by Miescher's classical studies. Raubitschek'*° found that isolated amyloid, when used for immune reactions, behaved like a specific protein, different from the normal proteins of the animal from whence it came and apparently biologically the same in different spe- cies. (This observation awaits confirmation.) Krawkow considers that amyloid differs from normal chondroitin- sulphuric acid compounds, such as cartilage, in that in the latter the acid radical is in a loose combination with the protein, while in amy- loid the combination is a very firm one, perhaps in the nature of an ester. The occurrence of the typical amyloid reaction in what ap- pears otherwise to be normal cartilage, occasionally observed in senile tissues, may be due to the transformation of loosely bound into firmly bound chondroitin-sulphuric acid. In any event, amyloid is not essen- tially a pathological product, but rather a slightly modified normal constituent of the body. However, in view of the contradictory results of Hanssen and Mayeda, as yet uncontroverted, the chemical nature of amyloid must be considered as undetermined. An important con- sideration is that amyloid deposition occurs under similar conditions in all sorts of animals, including birds; it is very often found in the livers of antitoxin horses, and mice are especially prone to a severe amyloidosis after relatively slight and brief infectious processes. ^^ Staining Properties. — The classical reaction for amyloid is its staining a reddish brown when treated with iodin (best as Lugol's solution) in the fresh state. Such stained specimens, if afterward treated with dilute sulphuric acid, usually become blue or greenish, but may merely turn a deeper brown. Occasionally old compact amyloid may stain bluish or green with iodin alone. The iodin reaction disappears in specimens that have been kept for some time in preserving fluids, or in tissues that have become alkaline, and is generally less persistent than the metachromatic staining by methyl-violet or methyl-green, which color the amyloid red. Oc- casionally an otherwise typical amyloid will fail to react to iodin, but will stain well witii methyl-violet. All these variations may occur in different specimens from the same body, and the blue iodin-sulphuric acid reaction is usually given well only bj' splenic amyloid. These variations probably dejjcnd upon the age and stage of development of the amyloid, or upon secondary alterations, and are perhaps related to Neuberg's observations on the difference in -composition of amyloid of different origins. Krawkow studied these reactions with pure, isolated amyloid, and found evidence that the iodin reaction depends upon the physical properties of the ^* Concerning the absorption of amyloid see Dantchokow, Virchow's Archiv., 1907 (187), 1. "Verh. Dcut. Path. Gesell., 1910 (14), 273. " See Finzi, Lo Speriment., 1911 (05), 483; Davidsohn, Virchow's Arch., 1908 (192), 22(). AMYLOID 425 amyloid, while the inethyl-violct stain is a chemical reaction, and hence the iodin reaction is much the more readily altered or lost. As Dickinson^- says, amyloid stains with iodin simph' as if it absorbed the iodin more than does the surrounding tissue. Krawkow believed that the methyl-violet reaction is due to the dye forming a compound with the chondroit in-sulphuric acid, for he found that these substances unite with one another to form a rose-red precipitate. Hanssen, however, holds that the dj'es react with the protein, the iodin with some other, unknown labile substance. Schmidt found that implanted pieces of amyloid lost their iodin reac- tion as they underwent digestion, while the methyl-violet reaction was still very distinct. ■•* It is evident, therefore, that iodin is not by itself a specific stain for amyloid, especially as glycogen gives a similar reaction,** while true amyloid may not react. Leupold** summarizes his investigations as follows: Amyloid is a complex of different substances which are differentiated by micro-chemical reactions. The protein ground substance of the amyloid is refractory to the typical amyloid reactions. The group which is responsible for the methyl-violet reaction is intimately combined with this protein substance and is separated from it only by the action of alkali. The groups which give respectively the iodin and the iodin- sulphuric acid reactions are closely related to each other. Nevertheless the iodin- sulphuric acid reaction is a completely independent one and is not a modification of the iodin reaction. The occurrence of different colors in the iodin-sulphuric acid reaction depends upon different degrees of oxidation. Amyloid is an emulsion colloid in the gel state. After oxidation with potassium permanganate it is soluble in ammonia, NaOH and Ba (OH) 2. Conjugated sulphuric acid plays an important part in the production of amyloid in the organism. The.existence of large amounts of conjugated sulphuric acid produces amyloid which gives the iodin reaction. The methjd-violet reaction also depends on the presence of con- jugated sulphuric acid; however, for its production there must probably occur a reduction in the amyloid protein. The group which gives the iodin-sulphuric acid reaction occurs through decomposition and perhaps does not depend upon the sulphuric acid. The Origin of Amyloid This question has not been at all cleared up as yet by the advances made in our knowledge of the chemistry of amyloid substance. The fact that chondroitin-sulphuric acid is a characteristic constituent suggests that this bodj^ niay be liberated in considerable amount dur- ing the destructive processes to which amyloidosis is usually sec- ondary; this idea is further supported by the fact that amyloidosis occurs particularly after chronic suppuration in bone and lungs, both of which tissues, according to Krawkow, contain chondroitin-sulphuric acid. This idea was not substantiated, however, by the experiments made by Oddi and by Kettner,*^ who fed and injected into animals large quantities of the sodium salt of chondroitin-sulphuric acid with- out producing amyloid changes. Unpublished experiments of the writer with the same material, as well as with ground-up cartilage and with mucin, were equally unsuccessful. Likewise mice injected by " Allbutt's System, vol. 3, p. 225. "Litten (Verh. Deut. Path. GeselL, 1904 (7), 47) states that thionin and kresyl-violet are the most specific stains for amyloid, which they color blue; whereas methyl-violet stains red not onlv amyloid but also mucin, mast cell granules, and the ground substance of cartilage. V. Gieson's stain usually colors amyloid pale yellow, and hyalin red. ** See Wichmann, Ziegler's Beitr., 1893 (13), 487. " Beitr. path. Anat., 1918 (64), 347. « Arch. exp. Path. u. Pharm., 1902 (47), 178. 426 RETROGRESSIVE CHANGES Strada^^ with the nucleoprotein of pus, the so-called pyin, or with chrondroitin-sulphuric acid, did not develop amyloidosis. Oestreich^* injected cancer patients with chondroitin-sulphuric acid for thera- peutic purposes, but no amyloidosis resulted. As it is possible to cause amyloidosis experimentally in animals, especially chickens and rabbits, by causing protracted suppuration or chronic intoxication with bacterial filtrates, these negative results speak strongly against the idea of a transportation of chondroitin-sulphuric acid, but do not determine it finally. They may also, with propriety, be used in sup- port of the statement of Hanssen that amyloid does not contain chon- droitin-sulphuric acid. Leupold"*^ advances the following hypothesis: In chronic suppuration a soluble protein circulates in the blood which stimulates the formation of "defensive ferments." This protein substance, under certain conditions, is deposited in organs where large amounts of sulphuric acid occur. For the development of amyloid there are necessary three factors: A preformed protein, an increased amount of conjugated sulphuric acid, and an inefficiency of the amyloid-filled organ to eliminate the increased amount of conju- gated sulphuric acid. There is usually much difficulty in producing amyloid experiment- ally, for in only a certain proportion of cases are the experiments posi- tive (in but about one-third of Davidsohn's'*'* 100 trials, and many other experimenters have been much less successful). ^° Davidsohn, faihng always to get amyloid experimentally after the spleen had been removed, suggests that this organ (in which amyloid is usually earliest and most abundantly observed) produces an enzyme, which causes a precipitation of amyloid in the tissues from a soluble precursor brought in the blood from the site of cell destruction. Schmidt^^ considers it probable that some enzymatic action causes a precipitation or coagula- tion of the substance in the tissue-spaces or lymph-vessels. Amyloid is never deposited in the cells themselves, ^^ and it seems to be now generally considered that the amyloid material is infiltrated in the form of a soluble modification or precursor and that it is not manu- factured in the organ where it is found. It is an interesting fact that a practically identical substance is formed in all tissues and in al species of animals, even when the cause is quite difi"erent. Whether the precursors are brought to the organ in solution, or in leucocytes, is unknown — probably the former. Pollitzer^^ states that in various infections, especially coccus infections, chondroitinsulphuric acid is excreted in the urine; if this is correct it has an undoubted bearing on *' Hiochein. Zeit., 1909 (16j, 195. "Zeit. Krebsforsch., 1911 (11), 44. " Verb. Dent. Path. Gesell., 1904 (7), 39. »" See Tarclietti, Deut. Arch. klin. Mod., 1903 (75), 526. Hirosc, Bull. Johns Hop. Hosp., 1918 (29), 40. *' Verh. Deut. Path. Gesoll., 1904 (7), 2. " See Kbert, Virchow's Arch., 1914 (210), 77. " Deut. mod. Woeh., 1912 (3S), 1538. HYALINE DEGENERATION 427 the genesis of amyloidosis. The presence of glyeothionic acid in pus'* is of similar significance. The hypothesis that amyloid is formed from disintegrating red corpuscles is probal)!}^ incorrect. Amyloidosis is produced by the most varied species of bacteria and by their toxins, although the staphylococcus is usually most effective in experimental work.^^ Neither is suppuration absolutely essential, for injection of toxins alone (c. g., in preparing diplit heria antitoxin'^*"'), without suppura- tion, may produce amyloidosis, as also frequently does syphilis without suppuration and, less often, many other non-suppurative conditions (e. g., tumors). Wago^^ reports finding a widespread "amyloid-like" degeneration in rabbits immunized with sterile pancreatic extracts. Local amyloid accumulations are of some interest in considering the genesis of the usual generalized form. They occur particularly as small tumors in the larynx, bronchi, nasal septum, and eyelids; as all these tissues are normally rich in chon- droitin-sulphuric acid, it seems probable that the amjdoid arises from a local overproduction of chondroitin-sulphuric acid, which becomes bound with proteins in si(u. This makes it seem more probable that, in spite of the lack of positive experimental evidence, general amyloidosis is due to liberation of excessive quanti- ties of chondroitin-sulphuric acid in the sites of tissue destruction. Another form of local amyloid is seen particularly in the regional lymph-glands of suppurating areas; e. g., the lumbar glands in vertebral caries, the axillary glands in shoulder-joint suppuration. This local amyloidosis is undoubtedly due simply to the fact that these glands receive first, and in largest amounts, the cau.se, what- ever it may be, of the amyloid production.*^ Less readily explained are cases of extensive amyloidosis limited to the heart.*' Corpora amylacea will be found discussed under "Concretions" (Chap. xvii). HYALINE DEGENERATION*" Much confusion concerning this condition may be avoided if we ap- preciate that the term hyaline indicates a certain physical condition, which may be exhibited by many substances of widely different na- ture and origin. There is 7io one chemical compound, ^'hyalin," which, accumulating in cells or tissues, produces a hyaline appear- ance. The limits of the application of the term "hyaline degenera- tion," even to histological findings, is not agreed upon, but in general it is used to apply to clear, homogeneous, pathological substances that possess a decided affinity for acid stains, such as eosin. Some- what similar substances, usually of epithelial origin, which do not " Mandel and Levene, Biochem. Zeit., 1907 (4), 78. ** In a series of experiments directed to ascertain, if possible, which constituent of pus might be the cause of amyloid formation, 1 was unable to secure amyloid by protracted intoxication of rabbits by Witte's "peptone," which consists chiefly of proteoses (Trans. Chicago Path. Soc, 1903 (5), 240). ** See Lewis, Jour. Med. Research, 1906 (15), -449. "Arch. Int. Med., 1919 (23), 251. *^ Quite unexplained is the cause of the rarely observed localization of amyloid in the wall of the urinary bladder. See Lucksch (Verb. Deut. path. Gesell., 1904 (7), 34). ConcretioiLs giving the amvloid reactions have been found in the pelvis of the kidney. (Schmidt, Cent. f. Pathol., 1912 (23), 865. Mivauchi, ibid., 1915 (26), 289.) *9 See Hecht. Virchow's Arch., 1910 (202), 168. *" General literature, seeLubarsch, Ergeb. allg. Path., 1897 (4), 449. 428 RETROGRESSIVE CHANGES take either acid or basic stains strongly, are usually called "colloid." We may properly consider that pathological hyalin can be divided into two chief classes according to its origin: (1) connective-tissue hyahn; (2) epithelial hyahn. Connective=tissue hyalin is characterized, like amyloid, by be- ing deposited in or among the fibrillar substance of connective tissues, and not within the cells themselves, but there are undoubtedly several different sorts of chemical substances responsible for various forms of connective-tissue hyalin. One form is closely associated with amyloid, being found in organs showing amyloid degeneration, or in other tis- sues in the same body. In experimentally produced amyloidosis in animals it has been shown that such a hyahne substance may appear before the amyloid, which eventually replaces it; hence, it has been suggested that hyalin is a precursor of amyloid. ^^ Such hyalin differs from true amyloid only in its failure to give the characteristic stain- ing reaction of amyloid; in all other respects, e. g., cause, location, termination, it is the same. As it has been shown (see preceding sec- tion) that the staining properties of amyloid are very inconstant, it is probable that the above-described variety of hyahn is merely an in- completely developed, or occasionally a retrogressively altered amy- loid. However, it is probably not necessary, as some authors have thought, that amyloid should always pass through this hyaline stage in its formation. Quite different, without doubt, is the form of hyalin observed in scar tissue. This variety develops almost constantly in any scar-tissue after the blood-supply has been reduced to a minimum through con- traction, and is seen characteristically in the corpora fibrosa of the ovary, fibroid glomerules in chronic nephritis, thickened pleural, peri- cardial, and episplenitis scars, etc. Such hyahne substance occurs independent of the usual causes of amyloid, affects only abnormal fibrous tissue, never changes into amyloid, and is prone to undergo calcification — it surely has no close chemical relation to the form of hyalin that does become amyloid. Presumably, it is similar in na- ture to the collagen of normal fibrous tissue intercellular substance, which has undergone physical rather than chemical changes into a homogeneous hyaline substance. For its physiological prototype it has the thick "collagenous" fibers of the subcutaneous connective tis- sue. Probably of quite different origin is the hyalin that develops from elastic tissue, as seen best in the thick-walled, partly obliterated arteries of the senile spleen; and less characteristically in the early stages of arteriosclerosis, since here the preceding form of connective- tissue hyalin may also occur. Although arterial elastic tissue is related chemically to amyloid, these hj^alinc vessels do not develop the usual amyloid reaction, but retain more or less of the specific, clastic " SeeLubarsch, Cent. f. Pathol., 1910 (21), 97. COLLOID DFXiENERATION 429 tissue stains. Presumably (his I'oiiii ol" hyalin is an increased and physically altered elastin.**- Epithelial hyalin occurs within the cells, and includes substances of presumably widely diverse chemical nature, from the keratin of squamous epithelium to the small intracellular hyahne granules of carcinoma and other degenerating cells (Russell's fuchsin bodies)/'^ Fuchsin bodies are found also in plasma cells and, less often, in other cells, including granulation tissue; the fuchsin bodies of this class are beheved by Brown^^ to be derived from red corpuscles, a view also held by Saltykow, but not accepted by all pathologists/'^ Extracellu- lar substances of hyaline character, but of unknown composition, may also be produced by epithelium, e. g., hyaline casts in the renal tubules. The composition of none of these forms of hyalin is known, except that by using microchemical methods Unna®® has found evidence that keratohyalin consists of two elements, one of acid character, appar- enth' derived from the chromatin, and a basic substance resembling the globulins. Many other pathological materials of widely differing nature may, under certain conditions, assume a hyaline appearance; e. g., fibrinous exudates and thrombi, degenerated muscle-fibers (Zenker's or "waxy" degeneration), tumor-cells, (cylindroma), etc. In all of these the chemical nature of the parent substance or substances is probably much less altered than its physical appearance, but whether the change is related to the process of protein coagulation or not is unknown. Occasionally hyalin, both in epithelium and connective-tissue, takes on a crystalline structure (Freifeld).^^ COLLOID DEGENERATION This term, also, has a very indefinite meaning, and is applied to many different conditions by various authors. Thus, v. Reckhng- hausen includes under this name amyloid, epithelial hyaline, and mu- coid degeneration. Marchand includes hyaline connective-tissue degeneration, and, also, as do most other writers, the mucoid degenera- tion of carcinoma. Ziegler rightly protests against the inclusion of mucin under this heading, but includes the corpora amylacea. On ac- count of the chscovery by Baumann of the specific chemical nature of thyroid colloid it becomes particularly unfortunate that the term ''colloid" has such a wide and uncertain apphcation. It would seem that the safest view to take is that the word colloid is merely morpho- " See Schmidt, Verb. Deut. path. Gesell., 1904 (7), 2. *^ Literature, see Hektoen, Progressive Med., 1899 (ii), 241. 6* Jour. Exp. Med., 1910 (12), 533. «5 See discussion, Verh. Deut. path. Gesell., 1908 (12), 265; Miinter, Virchow's Arch., 1909 (198), 105. « Berl. klin. Woch., 1914 (51), 598. "Ziegler's Beitr., 1912 (55), 168; also Goodpasture, Jour. Med. Res., 1917 (35), 259. 430 RETROGRESSIVE CHANGES logically and macroscopically descriptive of certain products of cell activity or disintegration, which have nothing in common except the fact that they form a thick, glue-like or gelatinous, often yellowish or brownish substance. There is 7io one definite substance colloid, accord- ing to the usual usage of the word in pathological literature, but many different protein substances may assume the appearance to which the name "colloid" is given. Looking at the matter in this way, we must recognize as the usual "colloid" substances, the following chemical bodies: Thyroid colloid, the physiological prototype of the group. This consi.sts of a compound of globulin with an iodin-containing substance, thj^roiodin, the com- pound protein being called by Oswald iodothyreoglobulin. It occurs pathologi- cally only in cystic and similar changes in the thyroid or accessory thyroids. Being a specific product of the thyroid with definite physiological properties, it manifestly has only a morphological relation to the other forms of colloid found in degenerating tumors, etc. In cysts of the thyroid, and less often in tumors, there is occasionally found a more dense "colloid" material of deeper color, the "caoutchouc colloid" of the Germans; this seems to result largely from transformation of red corpuscles in hemorrhagic cysts (Wiget).** (The nature of thyroid colloid is discussed more fully under " Diseases of the Thyroid," Chap, xxii.) Mucin, when secreted in closed cavities, as in tumors, where it becomes thick- ened by partial absorption of the water, may take on a "colloid" appearance while retaining its chemical and tinctorial characteristics. This is particularly observed in the "colloid" carcinomas which arise especially from the mucous membrane of the alimentary tract. This substance is, of course, quite specific both in its chem- ical nature and its origin from specialized epithelial cells, and the process should properly be considered as a "mucoid degeneration." Pseudomucin, which differs from mucin in not being precipitated by acetic acid, is a common component of ovarian c.ysts, and when somewhat concentrated by absorption of water, forms "typical colloid." Because it is alkaline, this form of colloid tends to stain rather with the acid dyes (eosin, acid fuchsin, etc.), while true mucin stains with basic dyes. Several varieties of pseudomucin have been described by Pfannenstiel, and their properties will be considered more fully in the section on "Ovarian Tumors" (Chap. xix). The clear, glassy, yellowish sub- stance contained in small cavities of ovarian tumors, which is usually called "colloid," consists of nearly pure pseudo-mucin. All these substances yield a reducing substance on boiling with acids, which is a nitrogen-containing body, glucosmnin.^^ Simple proteins (e. g., serum-globulin, serum-albumin, nucleo-albumin, etc.) may, when in solution in closed cavities, become concentrated through absorption of water until they produce the phj^sical appearance of "colloid." Probably the colloid contents of dilated renal tubules, cavities in various mesoblastic tumors, etc., are produced in this way. MUCOID DEGENERATION Mucin, in its typical form, is a compound protein, consisting of a protein radical and a conjugated sulphuric acid which contains a nitrogenous sugar. Hence, when boiled with acids, mucin yields a substance reducing Fchling's solution. Mucin is acid in reaction, probably because of the presence of the sulphuric acid and, therefore, is characterized microchemically by staining with basic dyes. It is readily dissolved in very weak alkahne solutions, is precipitated by 68 Virchow's Arch.. 1906 (185), 416; von Sinner, ibid., 1915 (219), 279. «9Zangerle, Munch, med. Woch., 1900 (47), 414. MUCOID DEGENERATION 431 acetic acid, and its physical properties when in solution are quite characteristic. The term mucin, however, probably covers a number of related but distinct bodies. Some, such as the pseudotnucins, are readily distinguished by not being precipitated by acetic acid, and by being alkaline in reaction; others yield reducing substances with- out previous decomposition with acids (paramucin); while even among the "true" mucins certain differences in solubility exist. ^^ The studies of Levene^^ indicate that the non-protein radicals of mucins are of two sorts: One, chondroitin-sulphuric acid, contains the nitrogenous hexose, chondrosamine, and is found in cartilage, tendons, aorta and sclera; the other, mucoitin-sulphuric acid, has as its carbohydrate chitosamine, and is found in gastric and umbilical cord mucin, vitreous humor, cornea and ovarian cysts. In the mammalian body we find mucin occurring in two chief lo- calities: (1) as a product of secretion of epithehal cells; (2) in the interstices of connective tissue, especially of tendons. ^^ (The resem- blance of synovial fluid to mucin is more physical than chemical.) There is also evidence that mucin or a related body constitutes the cement substance between all the body-cells. Corresponding to these two chief sources of mucin we find mucoid degeneration occurring as distinct processes in mucous membranes (or tissues derived therefrom) and in connective tissue. Epithelial Mucin. — As epithehal mucin represents a distinct product of specialized cells, it is questionable if the ordinary applica- tion of the term degeneration in the sense of the conversion of cell- protoplasm into mucin, is correct. Certainly the mucin formation of catarrhal inflammation is merely an excess of a normal secretion, and the degenerative changes that may be present in the epithelial cells are produced by the cause of the inflammation, and are not dependent upon mucin formation. Even in the extreme example of mucoid degeneration seen in carcinomas derived from mucous mem- branes (the so-called "colloid cancers"), the epithelial degeneration is not necessarily to be interpreted as a conversion of cell-cytoplasm into mucin, but is largely due to the pressure of secreted mucin upon the cells within the confined spaces of the tumor. The mucin in these forms of mucoid degeneration is chemically the same as the normal mucin coming from the same source, but mixed with larger or smaller quantities of other proteins derived from cell degeneration or from '" For special consideration see Cutter and Gies, Amer. Jour. Phvsiol., 1901 (6), 155. '"■ Jour. Biol. Chem., 1918 (36), 105. '^ Schade (Zeit. exp. Path., 1913 (14), 23) says that the long controversy concerning the intercellular substance of mammalian connective tissue is settled by the work of Lier (Ledermarkt-Collegium, Frankfurt, 1909, p. 321), who found it to be a mucin similar to that of tendon or umbilical cord. Its behavior in edema supports this observation. That there are some chemical similarities in the protein moiety of epithelial and tendon mucin is indicated by their immuno- logical inter-reactions (Elliott, Jour. Infect. Dis., 1914 (15), 501). 432 RETROGRESSIVE CHANGES vascular exudates, and we do not yet know certainly the chemical character of the secretion of normal mucous membranes." (The stringy, mucin-like substance seen in some purulent exudates is prob- ably composed largely of nucleoproteins and nucleo-albumins derived from the degenerating leucocytes^ and is not true mucin.) Connective=tissue Mucin. — Excessive formation of connective- tissue mucin is observed most characteristically in myxedema {q. v.), but may also occur in connective tissues that are poorly nourished or otherwise slightly injured; it is seen particularly in the connective tissues surrounding the epithelial elements in adenomas and carcino- mas. In the walls of large blood vessels there is a mucoid connective tissue, rich in mucin, which may be increased in arterio-sclerosis (Bjorling).^^ Connective-tissue tumors (^myxosarcoma, myxofibroma, or myxoma) may also show a great quantity of mucinous intercellular substance, but many of the so-called myxomas are in reality merely edematous j&bromas or polypoid tumors, in which the resemblance to true myxoma is largely structural rather than chemical. This form of mucoid degeneration seems to be merely a reversion to the fetal type of connective tissue, which is characterized, as in the umbilical cord, by an excessive accumulation of a mucin-containing fluid intercellular substance, and a paucity of collagenous fibrillar structure. Appar- ently, when connective tissue reverts to an embryonal type, either from intrinsic causes (tumor formation), or when the nourishment is insufficient, or possibly when the normal stimulus to cell growth is absent (myxedema), the mucoid characteristics of fetal tissue reappear. The presence of mucin in the tissues seems to cause no reaction, and its absorption causes no harm. Rabbits that I injected with large quantities of pure tendon mucin almost daily for two to four months, showed absolutely no deleterious effects, either locally or con- stitutionally. Some of the French authors^^ claim that mucin pos- sesses a slight bactericidal power. On the other hand, Rettger^^ and others have found an apparently typical mucin produced by certain varieties of bacteria. GLYCOGEN IN PATHOLOGICAL PROCESSES" It seems probable that all, or nearly all, cells contain larger or smaller quantities of glycogen, but it may be insufficient in amount to be detected either microscopically or chemicall3^ Glj^cogen seems to be formed within the cells from the sugar of the blood, through a process of dehydration and polymerization, and to be reconverted whenever necessary into sugar, by a reverse process of hydrolysis. It ^^ See Lopcz-Suarez, Biochcm. Zeit., 1913 (56), 167. ^* Virchow's Archiv., 1911 (205), 71. "ArloiiiK. Coinpt. Rend. Soo. Biol., 1902 (54), 306, and 1901 (53), 1117. ^« Jour. Med. Research, 1903 (10), 101. " Bibliography by Gierke, Ziegler's Beitr., 1905 (37), 502, and Ergebnisso Pathol., 1907, XI (2), 871; Klestadt, ibid., 1911, XYU), 349. GLYCOaENIC IXFIl/rh'ATION 433 is quite possible that both of these processes represent merely the reversible action of an intracellular enzyme, but this has not been estal)li8lio(l. We do know, however, that soon after death the intra- cellular glycogen is rapidly converted into dextrose.^" Properties of Glycogen. — Glycogen is frequently called an "animal starch," having the same f^eiieral composition as the starciies (('bIIk.Oo)^, and apparently, like the starches, it represents a relatively insolul)le resting stage of sugar in the course of metabolism. It is readily soluble in water, forming an opalescent, colloidal solution, and, therefore, has no efTect on osmotic pressure, and it is not difTusible."^ Because of its solubility and the rapidity with which postmortem change to dextrose occurs, specimens that are to be examined microscopically for glycogen must be hardened while very fresh in strong alcohol, in which glycogen is insoluble.*" One of the most characteristic reactions is the port-wine color given by glycogen when treated with iodin; this reaction may be applied micro- scojncally, solution of the glycogen being avoided by having the iodin dissolved in a solution of gum arable or in glycerol. Salivary ptyalin rapidly converts gly- cogen into glucose, and this reaction may also be used microscopically to prove that suspected granules are glycogen. However, failure to find glycogen micro- chemically does not alwaj's mean its absence from a tissue.*' Physiological Occurrence According to Gierke, the normal glycogen of cells resembles fat in that part of it disappears during starvation, while the rest cannot be removed in this way and probably is something more than a reserve food-stuff. In distribution glycogen somewhat resembles fat, being abundant in the liver*'^ and muscles, but Gierke considers that the microscopic evidence of the quantity of glycogen present in the cell agrees better with the results of actual chemical analysis than is the case with fat. Ilusk,*^ however, finds only a general agreement, with marked exceptions. Neither iodin nor Best's carmin stain are absolutely specific for glycogen, but Gierke believes that we may safely consider a substance as glj^cogen when it is homogeneous, rather easily soluble in water and more so in saliva, gives the usual iodin reaction, and stains bright red with Best's carmin solution.*^ With these controls, the microscopic findings were found to agree closely with the results of direct chemical analysis, and glycogen was found microscopically visible in muscle, liver, lung, heart, uterus, and skin (but not in the brain,*^ where it may be demon- strated chemically in minute quantities). Glycogen is commonly said to be especially abundant in fetal tissues, but it is not present in all fetal cells, *^ nor is it always most abundant in the most rapidly growing tissues. Although both fat and glj'^cogen are quite abundant in fetal muscle and liver tissues, the liver of early embryos does not contain either.*' Invertebrates and the lower vertebrates have more than the higher forms. In mammalian adults the liver and muscle contain the most glycogen, cartilage '* Literature concerning physiology of glvcogen bv Pflliger, Pfliiger's Arch., 1903 (96), 398; and Cremer, Ergeb. der Physiol., 1902 (1. Abt. 1), 803. '^ See Gatin-Gruzewska, Pfluger's Arch.. 1904 (103), 282. *° According to Helman (Cent. f. inn. Med., 1902 (23), 1017), glycogen may be found in specimens preserved in alcohol as long as fifteen years. *i Bleibtreu and Kato, Pfluger's Arch., 1909 (127), 118. *^ In the livers of two executed criminals Garnier (Comp. Rend. Soc. Biol., 1906 (60), 125) found respectively 4 per cent, and 2.79 per cent, of glycogen. »3 Univ. of California Publ., Pathol., 1912 (2), 83. ** Concerning staining methods see Ivlestadt, loc. cil.'''' ** Mav be present in fetal nervous tissues. (Gage, Jour. Comp. Xeurol., 1917 (27), 451). «8 See Glinke, Biol. Zeit., Moskau, 1911 (2), 1. " Adamof! (Zeit. f. Biol., 1905 (46), 288) contests the idea that the amount of glycogen is in direct relation to growth energy; see also Mendel and Leaven- worth (Amer. Jour. Physiol., 1907 (20), 117), who found no particular abundance in the tissues of the fetal pig. 28 434 RETROGRESSIVE CHANGES standing next, and it is also present in squamous epithelium (particularly the mid- dle layers), especially that of the vagina (Wiegmann), but not in slightly stratified (cornea), transitional, or cylindrical epithelium. Normal human kidneys do not seem to show glycogen, but it may be present in the kidneys of mice, rabbits, and cats. There is considerable in the heart muscle.*^ The amount in different skele- tal muscles varies,^^ usually being especially abundant in the diaphragm. Gly- cogen is most abundant in the uterus at the time of child-birth, and is abundant in the placenta; but it is also present in the uterus and tubes independent of preg- nancy.^" After pancreas extirpation, Fichera^^ observed a disappearance of all visible glycogen, except a little in the cartilage and stratified epitheliuni ; hence he considers the glycogen-content as a function of cell nourishment. Fat and gly- cogen often occur together, although one may be present without the other (Gierke). Presumably the failure to find glycogen in certain cells depends rather on a failure of technic than on a total absence of glycogen. There has been some diversity of opinion as to whether glycogen occurs as granules in the living cell, or whether the granules are formed from a homogeneous substance by hardening fluids. In view of the clear-cut, definite spaces it may leave in cells when dissolved out, glycogen probably occurs as granules, especially when present in abnormally large quantities. Ervin^^ believes that glycogen, like fat, may exist within the cells so finely divided that it cannot be stained bj'' glycogen stains. The studies of Arnold have shown that in many cells the glycogen takes on a definite structure in close relation to the plasmosomes. It has been suggested that the intra-epithelial hyaline bodies (Russell's fuchsin bodies) are glycogen, which idea is probably not correct. Habershon and others have suggested that eosinophile granules are either glycogen or related to it. The presence of glycogen in the cells seems to cause no injury to the cytoplasm, and if it again disappears, the cells become quite normal. ^^ Even the nuclei may contain granules of glycogen without evident permanent injury. Pathological Occurrence According to the results obtained by Fichera and Gierke, it seems probable that glycogen accumulation is produced under the same conditions as are fatty changes, i. e., when oxidation is locall}^ or generally impaired. Fat and glycogen are, therefore, often found together in the margins of infarcts and of tubercles, in passive con- gestion of the liver, and in heart muscle with fatty changes due to severe anemia. The glycogen, being more labile, seems to disappear early when the cells become necrotic, and hence glycogen is not pres- ent in older necrotic areas where the fat still persists. (This proba- bly accounts for the frequently repeated statement that glycogen and fat do not occur together.) Ervin^- believes that glycogen is impor- tant in holding intracellular fats emulsionized, and hence in its ab- sence in diabetes the fats become visible as fatty degeneration — hence the inverse ratio of glycogen and fat. Whether the glycogen can be transformed into fat, perhaps forming an intermediary stage in a trans- formation of protein into fat, has not been determined, but there 88 Berblinger, Ziegler's Beitr., 1912 (53), 155. 8M.ipska-Mlodowski, Beitr. path. Anat., 1917 (64), 18. 90 McAllister, Jour. Obs. Gyn. Brit. Emp., 1913 (34), 91. 91 Ziegler's Beitr., 1904 (30), 273, literature. 92 Jour. Lab. Chn. Med., 1919 (5), 14(). ''Yet Teissier (Compt. Rend. Soc. Biol., 1900 (52), 790) believes the amount normally present in the liver is strongly bactericidal, and in a later publication (ibid., 1902 (54), 1098) considers that it is toxic to liver-cells. ^^Vndelstadt (Cent. f. Bact., Abt. 1, 1903 (34), 831) found that under certain conditions gly- cogen impedes hoinoly.sis by normal serum. • GLYCOGENIC INFILTRATION 435 seems to be little doubt that it is infiltrated from outside the cell, and not formed directly from degenerated protein. It seems to be deposited only in cells that are still living, although it can become split up in dead cells. All cells, but especially muscle-cells and leucocytes, seem able to lay up glycogen in visible amounts under cer- tain conditions. In inflamed areas glycogen is found in both ti-ssue- cells and- leucocytes, but not in cells showing nuclear degeneration (Best, Gierke). In pneumonia the leucocytes of the exudate, and to a less extent the alveolar epithelium, contain glycogen as well as fat. In tubercles glycogen is found in the cells which contain ba- cilli, and it is generally present in the epithelioid cells, rarely in giant cells, not at all in lymphoid cells or in the necrotic elements (De- vaux). Liver glycogen is altered most in poisoning, being reduced by phosphorus, arsenic, chloroform, HgCU, and many other poisons; the amount is reduced when death from any cause is slow, or when putrefaction has occurred, but it is increased in carbon monoxide poisoning (Alassari).^^ In rabbits, at least, it is deposited in the liver first about the central vein, and in fasting animals it disappears first from the periphery. ^^ It seems to have a marked protective effect in phosphorus poisoning. ^^ Glycogen in Tumors. — Glycogen has been observed frequently in tumors. Brault believed the quantity an index of rate of growth, on the principle that glycogen appears most abundantly in embryonal tissues, and therefore in tumors the amount of glycogen should agree with the degree to which the cells have gone back to the embryonic tj'pe. Lubarsch considered that only tissues normally containing glj^cogen give rise to glycogen-containing tumors. Gierke could cor- roborate neither of these ideas, and considers that glycogen appears in tumors under exactly the same conditions in which it appears in other tissues; i. e., when cell nutrition and oxidation are impaired. Ap- parently, however, both the embryonic origin and local retrogressive changes determine the deposition of glycogen in tumors. Glycogen is particularly abundant in squamous epithelium of epitheliomas that have gone on to hornification;^'^'' in testicular tumors, hyperneph- romas, parathyroid tumors (Langhans),^^ endotheliomas, chondromas, and mj^omas, and it also occurs in the connective tissues surrounding tumors. Of 1544 tumors of all sorts examined by Lubarsch, ^^ 447 (or 29 per cent.) contained glycogen microscopically; fibromas, oste- omas, gliomas, hemangiomas were always free from glycogen; and lipomas and lymphangiomas nearly always. Adenomas are almost " Gaz. degli Ospedali, 1906 (27), 537. " Ishimori, Biochem. Zeit., 1913 (48), 332. 9« See Simonds, Arch. Int. Med., 1919 (23), 362. ■^^^ In mouse tumors Haaland found glycogen only in squamous cell carcinoma, and in the connective tissue surrounding other tumors (Jour. Path, and Bact., 1908 (12), 439). ^' Virchow's Arch., 1907 (189), 138. "ss Virchow's Arch., 1906 (183), 188. 436 RETROGRESSIVE CHANGES equally free from glycogen (two positive in 260 specimens), while it was constant in teratomas, rhabdomyomas, hypernephromas, and chorio- epitheliomas. Fifty and seven-tenths per cent, of the sarcomas and 43.6 per cent, of the carcinomas showed glycogen, most abundant in squamous-cell epitheliomas; columnar-celled carcinomas contain gly- cogen much less often, and it is always absent in "colloid cancers." Animal parasites, in common with other invertebrates, usually show abundant quantities of glycogen. ^^ It has been found in pro- tozoa, as well as in all varieties of intestinal worms. According to Barfurth, nematodes in glycogen-free animals may contain glj'cogen. The glycogen is found chiefly in the connective tissues of the intestinal parasites, but in some of the nematodes it occurs chiefly in the sexual organs and muscle-cells. The walls of the hydatid cysts contain much glycogen, which is, perhaps, related to the usual presence of sugar in their contents. If Habershon's contention is correct, that eosinophile granules are related to glycogen, we may have here an explanation of the occurrence of eosinophilia in infection with animal parasites. (See also "Animal Parasites," Chap, v.) Glycogen in Leucocytes. — The occurrence of glycogen in the blood has aroused much interest, particularly in relation to its diag- nostic value. Many leucocytes contain granules that stain with iodin, and although it is possible that these are not all granules of glycogen, yet, for the most part, they probably represent this substance in excessive quantities. The granules are observed chiefly in the poly- morphonuclear neutrophiles, but seldom in large and small mononu- clear cells and eosinophiles.' Occasional granules are also found free (or perhaps contained in blood-platelets) in all blood, whether normal or pathological. 2 Hirschberg^ states that normal animals of all species have leucocytes giving an iodin reaction for glycogen if proper technic is used, but which is not obtained by the ordinary iodin-gum solution method unless the glycogen is rendered abnormally insolu- ble by toxic injury; this is an explanation for the relationship of iodophilia and infections. According to Wolff-Eisner the leucocytes in myeloid leukemia contain no glycogen granules. It does not seem to be settled whether the glycogen is taken on by the leucocytes at the place of pathological lesion, or in the bone-marrow under the in- fluence of circulating poisons, or both. Habershon states that from 1 to 16 per cent, of all leucocytes normally contain glycogen granules, ^^ Elaborate treatise on occurrence of glycogen in lower animals by Barfurth, Arch, mikros. Anat., 1885 (25), 269; also liusch, Arch, internat. physiol., 1905 (3), 49; Brault and Loeper, Jour. Phys. et Path. Gen., 1904 (6), 295 and 720. 1 See Bond, Brit. Med. Jour., Feb. 3, 1917. =^ Literature— Locke and Cabot, Jour. Med. Research, 1902 (7), 25; Locke, Boston Med. and Surg. Jour., 1902 (147), 289; Reich. Bcitr. klin. Chir., 1904 (42), 277; Kiittner, Arch. klin. Chir., 1904 (73), 438; Gulland, Brit. Med. Jour., 1904 (i), 880; Habershon, Jour. Path, and Bact., 1900 (11), 95; Wolff, Zeit. klin. Med. 1904 (51), 407. 3 Virchow's Arch., 1908 (194), 367. GLYCOGENIC INFILTRATION 437 and Wolff believes that the glycogen seen in leucocytes represents normal glycogen made insoluble through injury. This may explain why the leucocytes in an infected area may give iodin reactions when the leucocytes in the circulating blood do not. Locke gives the occurrence of this abnormal iodin staining of the leucocytes (termed iodophilia) as follows: "Septic conditions of all kinds, including septicemia, abscesses, and local sepsis (except in the earliest stages), appendicitis accompanied by abscess formation or per- itonitis, general peritonitis, empyema, pneumonia, pyonephrosis, sal- pingitis with severe inflammation or abscess formation, tonsillitis, gonorrheal arthritis, and hernia or acute intestinal obstruction where the bowel has become gangrenous, have invariably given a positive iodophilia, and by its absence all these cases can be ruled out in diag- nosis. In other words, no septic condition of any severity can be present without a positive reaction. Furthermore, the disappearance of the glycogen granules in the leucocytes in from twenty-four to forty-eight hours following crisis with frank resolution in pneumonia, and the thorough drainage of pus in septic cases, is of considerable importance." Clinical experience, however, seems not to have ac- corded any constant significance to iodophilia.* In exudates glycogen is found in the leucocytes as long as they retain their vitality, but disappears soon after retrogressive changes begin; hence it is not usually present in old sterile pus. Loeper^ made quantitative estimates of the glycogen in exudates, finding from 0.59-0.62 gram per Liter in cellular pneumococcus pleural effusion, 0.25 gm. in cellular tuberculous effusion, but only traces in serous tuberculous effusion and in an old tuberculous pyothorax. A pneu- monic lung contained 0.85 gm. of glj^cogen per kilo, and traces were found in pneumonic sputum and in the contents of tuberculous cavi- ties. It is very abundant in tuberculous sputum, as much as 2 to 3 per cent, in advanced stages, but absent in bronchial catarrh; in pneu- monia 0.05 per cent, was found, in putrid bronchitis 0.25 per cent. (Pozzilli). When glycogen solution (1 per cent.) is injected into the peritoneal cavity, the endothelial cells and invading leucocytes be- come loaded with glycogen granules. Glycogenic Infiltration in Diabetes. — Although in diabetes the chief normal storehouses of glycogen, the hver and muscles, are either poor in or free from glycogen, yet in other tissues in diabetes the most marked accumulations of glycogen are found, the granules frequently fusing in the cells into droplets larger than the nucleus. When dissolved out in ordinary microscopic preparations, the clear round space left is exactly hke the space left by a fat-droplet, except that the margins show a tendency to take the basic stain for some unknown reason. In even the most extreme cases, however, the nucleus * See Bernicot, Jour. Path, and Bact., 190G (11),' 304. » Arch. M6d. Exp., 1902 (14), 576. 438 RETROGRESSIVE CHANGES is well preserved although it, too, may contain large masses of glycogen, in which case there is no glycogen in the cytoplasm.'' Gl3'cogen is found particularly in the epithelium of Henle's tubules," in heart muscle, and in the leucocytes. Fiitterer describes masses of glj'cogen in the cerebral capillaries, resembling an embolic process; it is also present in the tissues of the eye.^ Experimental diabetes (pancreas extirpation, piqure) produces a marked glycogenic infiltration.^ We cannot yet change van Noorden's statement: "We lack the biological explanation as to why certain cells retain the capacity to store glj'co- gen and even exert it more actively than before, whilst the proper organs for the storage of glycogen have lost it." * Askanazy and Hubschmann, Cent. f. Path., 1907 (18), 041. ' See Fahr, Cent. f. Path., 1911 (22), 945. * Shimagawora, Klin. Monatsbl. Augenheilk., 1911 (12), G82. 9 Huber and MacLeod, Amer. Jour. Physioh, 1917 (42), 019. CHAPTER XVII CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS CALCIFICATION' Pathological calcifioation occurs in two forms: one is a precipita- tion of calcium in secretions and excretions of the body; the other is the deposition of calcium salts in the tissues themselves. The former, which includes not only concretions in general, but probably also the deposition of calcium salts in the cells and tubules of the kidney,- both in disease and in experimental calcification after cer- tain poisonings, is readily enough explained in most instances by rec- ognizable alterations in the composition of the secretions, which lead to simple chemical precipitations. With this form we shall deal in the subsequent consideration of concretions, but, in referring to calci- fication, shall indicate only depositions from the blood directly into the tissues.^ Relation of Calcification to Ossification. — In normal ossification we have to deal with the accumulation of lime salts within the stroma or cells of a tissue that has usually undergone certain preparatory changes in the way of formation of a more or less homogeneous ground substance, but has not suffered a total loss of vitality, although vitality is possibly decreased. Pathological calcification is similar, in so far as we have to deal with deposition of quite the same salts in tis- sues that have suffered either total or partial loss of vitality, and which very frequently indeed are hyaline. What appear to be essential differences are these : (1) In calcification the lime salts always remain in clumps and masses, often fusing to greater or less degree, but never with the diffuse even permeation of tissue seen in ossification. (2) All the cells within a calcified area, if not dead at the beginning of the process, eventually disappear for the most part, and we have sooner or later a perfectly inert mass, practically a foreign body, instead of a specialized tissue as in ossification. (3) Ossification is accomplished only in varieties of connective tissue, but calcification may involve any sort of cell or tissue provided it is degenerated sufficiently. Furthermore, anj^ area of calcification is likely to be replaced by bone, no matter what tissue may be involved; ap- parently the presence of calcium salt deposits in any part of the body can stimu- late the connective tissues to form bone,* but in the absence of calcium salts even the cells which are normally osteogenic will not form bone. 1 Literature and resum6: Pfaundler, Jahrb. f. Kinderheilk., 1904 (60), 123; Wells, Jour. Med. Research, 1906 (14), 491, and Arch. Int. Med., 1911 (7), 721; Hofmeister, Ergebnisse Phvsiol., 1910 (9), 429; Schultze, Ergebnisse Pathol., 1910, XIV (2), 706. 2 See Wells, Holmes and Henry, Jour. Med. Research, 1911 (25), 373. ^ Normally the calcium content of the blood is quite constant, about 9-11 mg. per 100 c.c. serum, and the quantit}' is not modified by most diseases, except nephritis in which the serum calcium is reduced; also in eclampsia, tetany and jaundice. (Halverson, Mohlcr and Bergcim, Jour. Biol. Chem., 1917 (32), 171.) * See Nicholson (Jour. Path. Bact., 1917 (21), 287) concerning heterologous ossification. 439 440 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS Composition of the Deposits in Calcification.^ — The composi- tion of the inorganic salts in calcified areas in the body seems to be practically the same, if not identical, whether the salts are laid down under normal conditions (ossification) or under pathological condi- tions. With the blood continually passing between the bones and the calcified areas, the composition of the two must inevitably become similar or identical. This may be shown by a table giving the pro- portion of inorganic salts found by analysis of normal bone, and the proportion found in calcified materials.^ Mg3(P04)2 CaCOs Cai(P04)j Pathological Calcification Bovine tuberculosis Bovine tuberculosis Bovine tuberculosis Bovine tuberculosis (softened gland) Human tuberculosis Calcified nodule in thyroid . . . Thrombus, human Normal Ossification Human bone (Zalesky) Human bone (Carnot) Human bone (Carnot) Ox bone (Zalesky) Ox bone (Carnot) 0.S4 12.8 0.9 13.1 1.2 11.7 1.5 7.6 1.2 10.1 0.85 13.4 1.1 11.9 1.04 ±12.8 1.57 10.1 1.75 9.2 1.02 1.53 11.9 85.9 85.4 86.4 90.6 87.8 85.4 86.5 83.8 87.4 87.8 86.1 85.7 Iron may be present in pathological calcification, and, according to Gierke,^ in the fetus the entire skeleton contains iron as far as it has calcified, most at the points of active ossification. This statement has been questioned by Hiick and others, who believe that most of the iron demonstrable in normal ossification is the result of an arti- fact, for calcium deposits seem to have a great afiinity for iron. Be- cause of this, pathological calcium deposits take up iron from old hemorrhages in the vicinity, and so in many areas where there have ^ MacCordick (Lancet, Oct. 18, 1913) has advanced the interesting hypothesis that calcific deposits during life exist mostlj^ as soft masses, like unset mortar. Only when sufficient accumulation of CO2 occurs, as after death, or in the center of large areas of low vitality, such as fibroids, do the deposits become hardened; e.g., in a gangrenous leg the calcified vessels are stiff and brittle, while higher up in the living tissues they are soft and pliable. This would explain why we do not more often observe fractures of calcified arteries. As yet this hypotliesis has not received the critical tests its importance deserves. If true it will explain the cases of extensive calcification of the pericardium in which the heart is so encased that function would seem impossible if the deposit were rigid during life. (Sec Trans. Chicago, Pathol. Society, 1911 (8), 109, for consideration of pericardial calcification.) However, Klotz (Jour. Med. Res., 1916 (34). 495) has questioned the correctness of MacCordick's views on the basis of the occa- sional occurrence of fractures of calcified arteries, but without experimental evi- dence contradicting MacCordick. « Wells, loc. cit. ' Virchow's Arch., 1902 (167), 318. PATHOLOGICAL CALCIFICATION 441 been hemorrhages, especially in the vicinity of elastic tissue, th(!re occur actual "calcium-iron" incrustations." S. Ehrlich^ states that elastic fibers in the vicinity of hemorrhages take up the iron-contain- ing derivative of the blood-pigment, and this acts as a mordant for subsequent calcium deposition. Analysis of similar deposits in a syphilitic spleen by Gettler^" showed the presence of large amounts of silicates as well as calcium and iron. Potassium was much less than in normal spleen tissue. The presence of iron in normal ossification is supported by Sumita'^ and Eliasscheff.*^ In the so-called iron-lime lung Gigon** found but a trace of calcium and much sodium and potassium. Structure of Calcified Areas. — As before mentioned, in calcifi- cation there is not the same uniform infiltration of the ground sub- stance with lime salts that occurs in bone, yet the calcified area is possessed of a ground substance of organic material which does not dissolve in weak acids that remove the salts. There is no definite ratio between the lime salts and this albuminoid matrix, however. At first the salts occur in granules, which may become fused to a greater or less degree. It has been thought by some that the deposition occurs in the form of "calcospherites." These are small calcareous bodies, usually of concentric structure, which were first described by Harting. They appear to occur widely distributed in normal tissues, both animal and plant, and seem to be the result of the formation of insoluble calcium salts in the presence of colloidal substances, just as urinary and other concretions are formed about an organic nucleus. If calcium chloride and soluble carbonates are allowed to combine very slowly to form calcium car- bonate in a solution of egg-albumen, these or indistinguishable bodies are formed, which on being dissolved are found to possess an organic stroma that exhibits a marked affinity for any pigmentary substance that may be present. Apparently, when the proper concentration exists, the salts in crj'stallizing hold between the crystals the albuminous substances by which they are surrounded. Dastre and Morat believe that the substratum is lecithin, which others have found occupying a similar place in prostatic concretions. Calcospherites have been found in tumors, in cystic cavities, and in bodies with beginning decomposition. It may be men- tioned in passing that Littlejohn'^ observed the abundant formation of calcium phosphate crystals in bodies that had been immersed for some time in sea water. Oliver has found calcospherites in the tissues of a cancer of the breast. Pettit'^ found calcospherites in a sarcoma of the maxilla, presenting insensible transi- tions into the substance of the osseous tissue, and he suggests the possibility that the calcospherite formation may be related to the formation of bone. It seems, however, that they are probably more closely related to the formation of the shells of invertebrates, which are largely composed of carbonates in crystalline structure with an organic ground substance between them, and very little phos- phate indeed. « See Gigon, Ziegler's Beitr., 1912 (55), 46; Sprunt, Jour. Exp. Med., 1911 (14), 59; Klotz, Johns Hop. Hosp. Bull.; 1916 (27), 363. " Cent. f. Pathol., 1906 (17), 177. '» Symmers, Gettler, Johnson, Surg., Gyn. Obst., 1919, (28), 58. " Virchow's Arch., 1910 (200), 220. 12 Ziegler's Beitr., 1911 (50), 143. >3 Edinburgh Med. Jour., 1903 (13), 127. 1* Arch. d. Anat. Micros., 1897 (1), 107. 442 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS Occurrence of Pathological Calcification As far as we know, calcification seldom occurs in normal tissue, except in the formation of bone. Often the infiltrated tissue is com- pletely dead, as in infarcts, organic foreign bodies, caseous areas, and particularly in old inspissated collections of pus. It may be said that any area of dead tissue that is not infected, and that is so large or so situated that it cannot be absorbed, will probably become infil- trated with lime salts. Most frequently calcified, next to totally necrotic tissues, are masses of scar-tissue that have become hyaline sub- sequent to the shutting off of circulation in the scar by contraction of the tissue about the vessels. Elastic tissue also seems prone to an early calcification, and it is not uncommon to see the elastic laminae of small arteries calcified in an apparently selective manner. A pe- culiar form of calcification is that frequently found in ganglion-cells of the brain which have become degenerated or necrotic, particularly in the vicinity of old hemorrhages; the cells become infiltrated with lime salts until a complete cast of the cell, with dendrites and axis- cylinder well impregnated, is formed. The calcification of renal epi- thelium obtained experimentally by temporary ligation of the renal vessels or by the administration of certain poisons, is more closely related to the formation of ordinary urinary concretions than to tissue calcification, the calcium being present as the phosphate only.^^ Cal- cification of epithelial cells does occur, however, and seems to be pre- ceded by hyaline changes, in which hyaline substance the calcium is later deposited, as in epithelial pearls, for example. Metastatic Calcification. — What is perhaps the only exception to the rule that some form of tissue degeneration is required before cal- cification occurs, is the "metastatic calcification" of Virchow.^^ In conditions with much destruction of bone, as osteomalacia, caries, osteosarcoma, etc., deposits of lime salts have been found distributed diffusely in various organs, particularly in the lungs and stomach. As much as 13.38 per cent, of the dry weight of the lung and 12.15 per cent, of the kidney have been found as CaO in such a case.*^ As there is no evidence that these organs have been the site of any dif- fuse tissue necrobiosis before the calcification occurred, it seems prob- able that the deposits have been made in practically or quite normal organs, because of oversaturation of the tissue fluids by calcium salts. The fact that the lung and stomach, and also to a less degree the kid- ney, are picked out, suggests that the calcification is related to the fact that in these same organs we have the excretion of acids into their cavities, which leaves the fluids in the substance of the organs correspondingly alkaline, and an increase in the alkalinity of the iMour. Med. Iles^earch, 1911 (25), 373. "Virchow's Arch., 1855 (8), 103; review l)V Kockol, Dent. Arch. klin. Med., 1899 (U4), 332. Bil)lioKraphy and review by Wells, Arch. int. Med., 1915 (15), 574. " Virchow's Arch., 1909 (197), 112. PATHOLOGICAL CALCIFICATION 443 fluids makes the calcium salts decidedly less soluble. In the stomach the calcium deposits arc limited to the interglandular tissue about the upixn- portion of the shmds of the fundus, exactly corresponding to the parietal cells which are supposed to secrete the acid. Pre- sumabl}^, under normal conditions, the amount of calcium in the blood is too slight to be thrown down in this way, but when oversat- urated because of the calcium absorption in the skeleton, precipita- tion occurs in the parts of the bodj^ where the alkalinity of the blood or tissue fluids is greatest, or the CO2 concentration least. There also occurs a true metastatic calcification in the large arteries, pulmonary veins, and beneath the endocardium of the left side of the heart; that is, always in the places where the blood contains the least CO?. This fact supports the hypothesis that the CO? is an important factor in the solution of calcium salts in the blood, and that when there is an over- saturation with calcium it is deposited where the CO2 is least abun- dant. When the amount of calcium in the blood is increased by injecting or feeding calcium salts, depositions of calcium salts may take place in injured tissues, ^^ or even in normal tissues, as in Tanaka's experiments.^^ Extensive calcification may take place in the lungs without any evident bone disintegration, nor yet nephritis which has been thought at times to lead to enough calcium retention to account for metastatic calcification (Harbitz).^° A few cases of extensive subcutaneous calcification of unknown etiology have been described, but their relation to metastatic calcification is doubtful, as they seem to be localized deposits. ^^ Some have attempted to include the calcification of the vessels and other tissues in old age in the metastatic calcifications, ascribing the origin of the salts to the senile absorption of bone, but senile calcifica- tion is probably dependent rather upon the extensive hyaline degenera- tion of the connective tissues that occurs in the senile scleroses,-^ a change which seems to be more physical than chemical.-^ Chemistry of the Process of Calcification In analyzing the etiological factors in the production of pathologi- cal calcification for the purpose of determining the chemical changes that occur in the process, we have the following facts upon which to base the consideration: (1) The calcium salts must come from the blood, where they are 18 See Thayer and Hazen, Jour. Exp. Med., 1907 (9), 1. i^Biochem. Zeit., 1911 (35), 113; (38), 285; see also Katase, Beitr. path. Anat., 1914 (57), 516. 2« Norsk Mag. Laeg., 1917 (78), 1129. 21 See Mosbacher, Deut. Arch. klin. Med., 1918 (128), 107. ^^ Under the name of "calcium gout," M. B. Schmidt has described a case with generalized deposition of calcium in other tissues than those usually affected in metastatic calcification (Deut. med. Woch., 1913 (39), 59). ^^ See analyses of elastin from calcified and normal aortas by Ameseder, Zeit. physiol. Choin., 1913 (85), 324. 444 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS held in solution or in suspension by the proteins, either as the car- bonate and phosphate themselves, or as calcium-ion-protein com- pounds, or perhaps both. This suspension or solution is an unstable condition, possibly only because of the extremely small proportion of calcium in the plasma (about 1 : 10,000), and, therefore, capable of being overthrown by increased alkalinity of the blood, changes in the proteins or CO2 content, or changes in the quantity or composition of the calcium salts. It is probable, from the work of Barille, that the calcium of the blood exists as a soluble complex double salt, tri- basic calcium-carbon-phosphate (P208Ca2H2: 2C02(,C03H)2Ca), this compound being possible because of an excess of CO2. (2) Retrogressive changes in the tissues are a sine qua non except in metastatic calcification. Hyaline degeneration, the chemical nature of wliich is not understood, is a very favorable condition, as also is necrosis when absorption is deficient. (3) In the areas that are to become calcified the circulation is very feeble, the blood plasma seeping through the tissue as through any dead foreign substance of similar structure, without the presence of red corpuscles to permit of oxidative changes. We may, therefore, imagine that the deposition of calcium salts in such areas of tissue degeneration depends upon one or more of the following conditions: (1) Increased alkalinity or decreased CO? in the degenerating tissues, causing precipitation of the inorganic salts in the fluids seep- ing slowly through them. (2) Utihzation of the protein of the fluids by the starved tissues so completely, because of its slow passage through them, that the calcium cannot be held longer in solution. (3) The formation within the degenerated area of a substance or substances having a special affinity for calcium. (4) Production of a physical condition favoring the local absorp- tion of salts, the least soluble salts accumulating in excess. The first of these conditions seems to come into play especially in metastatic calcification, already discussed. We have no evidence that in degenerating tissues, much less in normal ossification, there is an alkahne reaction developed; but rather the contrary, an acid reaction is more usual. But, as explained below, decrease in the CO2 content in calcifying tissues, especially when combined with other changes, may be of importance. Lichtwitz24 especially has laid emphasis on the possible part played by changes in the proteins in inducing calcification. He advances the idea that precipitation of the colloids in the degenerated area, as in caseation, decreases the amount of crystalloids which can be held in solution, wherefore the least soluble salts, those of calcium, arc precipitated; by laws of osmotic pressure more calcium in solu- " Deut.-med. Woch., 1910 (36), 704. PATHOLOGICAL CALCIFICATION 445 tion will then enter to establish equilibrium, be precipitated, and make way for more calcium, until the amount of deposit prevents further osmotic diffusion. Although suggestive in regard to patho- logical calcification, and probably of importance in the formation of concretions, this conception is difficult toapi)ly to normal ossification; also in pathological calcification one would expect precipitation of calcium to occur in the outermost surface of the degenerated area, soon leading to a shell of inorganic material which would limit the deposition. The possibihty of the formation of calcium-binding substances within the degenerated area has always seemed the most attractive, and has received the most attention by investigators. Of the special substances that might be present in such areas that would have a high affinity for calcium, phosphoric acid usually receives first con- sideration, since it is as phosphate that most of the calcium is bound, and also since the possible sources of phosphoric acid in decomposed nucleoproteins and lecithin are so obvious. Less considered in the past, fatty acids offer another possibility, especially in view of the fatty degeneration that so frequently precedes calcification. Proteins might also be formed that would combine calcium, especially dcutero- albumose, which Croftan^^ states has a high degree of affinity for calcium, and which would be present in areas undergoing autolysis. Formation of Calcium Soaps. — In favor of the possibility that the calcium is first bound as soaps are the following facts: Calcifica- tion occurs chiefly in places where fatty degeneration has occurred, such as tubercles, atheromatous vessels, etc. In fat necrosis fatty acids are formed, which soon combine wuth calcium to form calcium soaps. Virchow observed calcification in the form of soaps in a lipoma, and Jaeckle^® found that a calcifying Lipoma contained 29.5 per cent, of its calcium in the form of calcium soaps. Klotz-^ ob- tained staining reactions in calcifying tissues that suggested the pres- ence of soaps, which he also extracted by solvents, and he strongly urges, as the first step in the formation of pathological calcified masses, that the calcium is first laid down as soaps, afterward under- going a transformation into the less soluble phosphate and carbonate. Fischler and Gross-*^ also obtained microchemical reactions for soaps in the margins of infarcts and in atheromatous areas, but not in caseous areas; they therefore consider that calciimi-soap formation is an important step in the process of pathological calcification, but that it is not essential. The value and the interpretation of the his- tological evidence of the participation of calcium soaps is, however, open to question. =5 Jour, of Tuberculosis, 1903 (5), 220. 2«Zeit. phvsiol. Chem., 1902 (36), 53. "Jour. Exper. Med., 1905 (7), 663; 1906 (8), 322. 28Ziegler's Beitr., 1905 (7th suppl.), 339. 446 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS On the other hand, Wells, ^^ studying large quantities of material chemically, found at most doubtful traces of calcium soaps in calci- fying matter, even in the earliest stages, and also very small amounts of other soaps or fatty acids, and, therefore, questions the occurrence of calcium soaps as an essential step in calcification, although not doubting that under certain conditions {e. g., calcifying lipomas, fat necrosis) this may occur. In calcification at all stages the propor- tion of calcium carbonate and phosphate was found quite constant, and exactly the same as in normal bone; namely, in the proportion expressed by the formula 3(Ca3(P04)2:CaC03, which Hoppe-Seyler advanced to express the composition of the salts of bone. Hence it seems probable that there are no essential differences between the processes of ossification and pathological calcification,^" and there seems to be as yet no reason for assuming that in the former calcium soaps constitute an essential step in the process. Phosphoric Acid in Calcification. — It has generally been as- sumed that in normal ossification the calcium is combined by phos- phoric acid, which probably is derived from the cartilage cells, possibly through autolysis of the nucleoproteins or some similar process. ^^ Grandis and Mainini,^^ by using microchemical methods, thought that they found evidence that the phosphorus of ossifying cartilage is converted from an organic combination into an inorganic form (P2O5), which then takes up calcium from the blood. The methods used have been questioned, and Pacchioni,^^ from his studies, was inclined to the opinion that the calcium entered the cartilage already combined as phosphate. Wells implanted into the abdominal cavity of rabbits various tissues that had been killed and sterilized by boiling, and found that tissues rich in nucleoproteins showed no tendency to take up calcium in greater amounts than did tissues poor in nucleoproteins, which result speaks against the idea that phosphoric acid derived from nucleic acid combines the calcium. On the other hand, im- planted dead cartilage soon became thoroughly impregnated with calcium salts, which seemed to be deposited in the same proportion as to carbonate and phosphate as in bone. Physical Absorption of Calcium Salts. — As there could be no question of "vital activity" on the part of this boiled cartilage, it seems most probable that there exists in cartilage a specific absorp- tion affinity for calcium salts, similar to the absorption affinity that Hofmeister^^ observed exhibited by other organic colloids (gelatin 29 See review in Arch. Int. Med., 1911 (7), 721. ^^ Dyes that stain the bones when fed to living animals (madder) also stain pathological calcific deposits (Macklin, Anat. Kecord, 1917 (ll), 387). ^' Hanes, who observed that the i)hosi)hatids disaj)pcar from the liver of the developing chick, suggests this as a source of the phosphoric acid required for ossification (.Jour. Exper. Med., 1912 (16), 512). ^2 Arch, per la sci. Med. Torino, 1900 (24), 67. " Jahrb. f. Kinderheilk., 1902 (56), 327. " Arch, exper. Path. u. Pharm.. 1891 (28), 210. I OSTEn}rALAnA 447 disks) toward various crystalline suhstaiiccs in solution. It is of sig- nificance that the substances in which calcium is deposited are, in most instances, of similar physical character, being homogeneous and often hyaline, although of the most varied chemical composition; in other words, they agree much more in physical than in chemical struc- ture. Also we find that hyaline tissues with an affinity for calcium often exhibit a similar affinity for other substances, such as pigment and iron.^^ Hofmeister advances the hypothesis that when the cartilage or other matrix becomes saturated with calcium salts, any decrease in COo content of the solution will lead to a precipitation of calcium salts, thus restoring to the cartilage its power of absorbing more calcium salts whenever the fluid comes to it with a higher degree of saturation with calcium salts and CO2. This hypothesis is in har- mony with Barille's observation that when the C02 is reduced the complex carbon-phosphate of calcium precipitates a mixture of car- bonate and phosphate in the same proportions as found in bones and calcific deposits generally. The fact that this ratio (10 to 15 per cent. CaCOs and 85 to 90 per cent. Ca?(P03)4), is found in all stages of calcification, is entirely in favor of the above hypothesis, and opposed to the idea that any special chemical precipitant formed in the calcifying area is responsible for the deposition of calcium. Taken all in all, the evidence seems in favor of the view that normal ossifica- tion and pathological calcification (except metastatic calcification and the calcification of fat necrosis and other areas of necrotic fat tissue) depend more upon physico-chemical factors and variations in CO2 concentration than upon the presence of chemical precipitants in the tissues. This view is supported by the observation of IMacklin^'' that calcifying and ossifying tissues become stained alike with madder fed during their formation, through the deposition of stained calcium salts from the blood. Osteomalacia" In this condition the quantity of inorganic salts in the bone is greatly decreased, while, at the same time, their place is taken in part by new-formed osteoid tissue; as a result, the proportion of the weight of the bone formed by inorganic salts is reduced to as lew as 20 to 40 per cent., instead of being from 56 to 60 per cent., as in normal bone. This has suggested that the cause of the disease may be a solution of the lime salts by some acid, but Levy^^ found that in osteo- malacia the proportion of calcium carbonate and phosphate in the " See Sprunt, Jour. Exp. Med., 1911 (14), 59. '« Jour. Med. Res., 1917 (36), 493. " See also review in Albu and Neuberg's "Mineralstoffwechsel," Berlin. 1906, pp. 124^127; bibliography by Zesas, Cent. Grenz. Med. u. Chir., 1907 (10), 801; full discussion by McCrudden, Arch. Int. Med., 1910 (5), 596; 1912 (9), 273. 38 Zeit. physiol. Chem., 1894 (19), 239. 448 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS bones remains constant, as also does the proportion of calcium and phosphoric acid; if the decalcification occurred through solution by lactic or other acids, he argued, the carbonate should be decomposed first, ^^ whei'eas the lime salts seem to be taken out as molecules of calcium carbonate-phosphate; i. e., in the same proportion as they exist in the bone. On the other hand, it has been found in Pawlow's laboratory that dogs kept for long periods after a pancreatic fistula has been established, develop a condition resembling osteomalacia,^" which would seem most reasonably explained as due to the constant loss of alkali in the pancreatic juice. Furthermore, investigation of Levy's objection to the acid solution theory has led to the observa- tion that when mixtures of calcium carbonate and phosphate are in colloids they are dissolved at equal rates. *^ Histologically, absorption seems to depend largely upon a direct eating out of bone tissue, both organic and inorganic substance, by osteoclasts (Cohnheim), followed by a formation of an uncalcified osteoid tissue. (Senile osteoporosis differs chiefly in that no new osteoid tissue is formed.) According to Dibbelt^^ when osteomalacia is experimentally induced in pregnant dogs and then recovery is allowed to take place, the decalcified bone substance present in the active stage does not become calcified, but is absorbed and replaced by new bone. Studies of metabolism in osteomalacia have shown a loss of calcium by the body, especially in the urine, as shown by the following table given by Goldthwait et al.:^^ Limbeck Neumann Goldthwait CaO in urine (gm.) CaO in feces 1.773 3.834 3.859 1.800 Total excreted 5.607 2.965 11.65 11.26 5.66 Total in food 4.56 Loss of CaO 2.9G5 0.39 1.10 McCrudden also found a considerable retention of nitrogen and sulphur, which may be retained in the new-formed osteoid tissue; magnesium^^ is also retained, probably being substituted for calcium in the bones. It is known that when magnesium and strontium are ^' Goto reports that in experimental HCl acidosis the bones lost 20 per cent, of their CaCOa without appreciable loss of phosphate (Jour. Biol. Chenu, 1918 (36), .355). "> Babkin, Zeit. Stoflfwechsel, 1910 (11), 561; Looser, Vcrh. Deut. Patli. Gcsell., 1907 (11) 291. " Kranz and Liesegang, Deut. Monat. Zahnhcilk., 1914, p. 628. ^2 Arbeit. Path. Inst. Tubingen, 1911 (7), 559. *' Goldthwait, Painter, Osgood and McCrudden, Amer. Jour. PhvsioL, 1905 (14), 389. ** Corroborated by Cappezzuoli, Biochem. Zoit., 1909 (16), 355. RICKETS 449 given to growing animals they will partially replace the calcium in the bones,'*'' while it is said by Etienne''*^ that excessive feeding of calcium itself leads in time to decalcification of the bones. Zuntz**^ found the respiratory metabolism in osteomalacia within normal limits, but tending to be low; protein metabolism shows nothing striking, but there is a high excretion of phosphoric acid through the feces. Castration of women with osteomalacia has been frequently, but not always, followed by improvement or recovery,^** and Neumann, and also Goldthwait, have found that in these cases the calcium loss is replaced by a marked calcium retention after the operation. What the relation of the ovaries to calcium metabolism or to osteomalacia may be has not yet been ascertained. Scharfe^'-* and Bucura^" both state that there are no characteristic or constant structural altera- tions in the ovaries in osteomalacia McCrudden^' found that the improvement in calcium metabolism observed after castration may be but temporary, and therefore believes that the primary cause of the disease does not lie in the ovaries. He is of the opinion that re- peated drains on the calcium of the bones, incited most often by preg- nancy, occasionally by tumors, sometimes by unknown causes, result in an excessive reaction to the stimuli, so that eventually the losses become too great to be made up; that is, osteomalacia is an exaggera- tion of a normal process resulting either from excessive stimulation of that process, or a failure to recover when the stimulus ceases. The beneficial effects of castration are probably ascribable chiefly or solely to the prevention of pregnancy. Osteitis deformans seems to be a localized osteomalacia. The relation of the adrenals to osteomalacia advocated by Bossi,^^ is of questionable significance, and there is no definite evidence as to any relation of exophthalmic goiter^^ or the para- thyroids,^* although hyperplasia of the parathyroids has been des- cribed.^^ RlCKETS^s As wdth osteomalacia, chemical studies of the bones in rickets have thrown httle light upon the etiology or pathogenesis of this condition. *5 See Lehnerdt, Zeit. exp. Med., 1913 (1), 175. •»« Jour. Physiol, et, Path., 1912 (14), 108. " Arch. f. Gyn., 1913 (99), 145. ^» Bibliography by Schnell, Zeit. Geb. u. Gyn., 1913 (75), 178. *3 Cent. f. Gyn., 1900 (24), 1216. =" Zeit. f. Heilk., 1907 (28), 209. " Amer. Jour, of Physiol., 1906 (17), 211. " Zent. f. Gyn., 1907 (31), 69 and 172. " Tolot and Sarvonat, Rev. d. Med., 1906 (26), 445. " Erdheim, Cent. med. Wiss., 1908 (46), 163. ^5 Bauer, Frankfurter Zeit. Pathol., 1911 (7), 231. " Complete literature and full discussion by Pfaundler, Jahr. f. Kinderheilk., 1904 (60), 123; also see Albu and Neuberg, " Mineralstoffwechsel," Berlin, 1906. pp. 119-124; symposium in the Verhandl. Deut. Path. Gesellsch., 1909 (13), 1. Metabolism studies by Meyer, Jahrb. Kinderheilk., 1913 (77), 28. 29 450 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS As the following table (taken from Vierordt^^) shows, there is a marked deficiency in the proportion of inorganic salts in the bones in rickets. The proportion of the different salts seems to be quite the same as in normal bone. Normal bone of a two months old child Rachitic bones Tibia Ulna Femur Tibia Humerus Ribs Vertebrae Inorganic matter 65.32 34.68 64.07 35.93 20.60 79.40 33.64 66.36 18.88 81.12 37.19 62.91 32.29 67.71 57.54 1.03 6.02 0.73 33.86 0.82 56.35 1.00 6.07 1.65 34.92- 1.01 14.78 0.80 3.00 1.02 72.20 7.20 26.94 1 ,. .„ Magnesium phosphate 0.81/ 4.88 1.08 60.14 1 6.22/ 2.66 0.62 81.22 Collagen (or ossein) Fat More modern analyses^** show a relative increase in water and magnesium, with a persistence of the normal ratio of calcium phosphate and carbonate. ^^ Cattaneo®° finds the increase in magnesium to vary in different parts of the skeleton, being greatest in the ribs. Aschen- heim states that the blood of cliildren with rickets shows greater variations from the usual CaO content (8-10 mg. per 100 c.c.) than are found in normal children,^' which is not corroborated by others."' As an essential difference from osteomalacia is the fact that in rickets there is a failure on the part of the osteoid tissues to calcify, whereas in osteomalacia absorption of calcified tissue takes place with subsequent substitution by osteoid tissue. Furthermore, in rickets the deficiency in calcium is said to be present only in the bones, "^ whereas in osteomalacia the soft tissues are also poor in lime salts. According to SchmorP^ the first structural abnormality in rickets is a failure to lay on calcium by small islands of cartilage in the zone of preparatory calcification. None of the various hypotheses as yet advanced to explain this defective ossification has satisfactorily accounted for all the observed facts. That a deficiency of calcium in the food is the cause of rickets is a most natural assumption, but it has not been proved that this is the case. Young animals fed on calcium-poor foods show, naturally enough, defective development of the bone,*''* but this differs essentially from ^^ Nothnagel's System, vol. 7, part ii, p. 21. " Gassmann, Zcit. physiol. Chem., 1910 (70), IGl. ^' The bones and muscles in Barlow's disease show quite the same deficiencj'' in calcium as in rickets (Bahrdt and Edelstein, Zeit. Ivinderheilk.,il913 (9), 415). 60 La Pediatria, VII, 497. 8' Jahrb. Kinderheilk., 1914 (79), 446. "2 There is a decrease in the calcium of the muscles according to Aschenheini and Kaumheimer (Monatschr. f. Kinderlieil., 1911 (10), 435) " Verhandl. Deut. Path. Gesell., 1905 (9), 248. 6^ See Weiser, Biochcm. Zeit., 1914 (66), 95. RICKETS 451 rickets in that the bone formed is defective chicfl}' in amount rather than in quality (Stoltzner), Furthermore, such "pscudo-rachitic bone" possesses a marked affinity for calcium salts, and takes them up as soon as the}' are supplied (Pfaundler). As the blood in rickets contains nearly normal amounts of calcium^^ it seems quite certain that calcium starvacion is not the fundamental trouble. In view of the fact that rickets is not solely a disease of bone tissue, but that all the various important viscera, as well as the muscles and tendons, show pathological changes, it seems most reasonable that rickets should be looked upon as a constitutional disease, in which the bone changes are prominent chiefly because the disease occurs at a time when the bone tissue is most actively forming and when the other organs are relatively quite completely developed. Stdltzner,^'' finding evidence that rickets does not depend upon either lack of calcium in the food or deficient absorption of calcium, and that the blood in rickets is of nor- mal alkalinity, looks upon the failure of calcification as depending upon an abnormality in the calcified bone tissue itself." He finds evidence of a preliminary alteration in normal osteoid tissue which prepares it to take the salts out of the blood, and Pfaundler-^^ supports this view, suggesting that this preparatory change in the osteoid tissue may depend upon autolysis, which is perhaps deficient in rickets. ^^ On the other hand, after extensive experimental work, Dibbelt^^ comes to the conclusion that rickets results from excessive elimination of calcium into the intestine, presumably because of the presence of precipitating substances in the intestinal contents, such as P-Oo from casein. Agreeing with Dibbelt that the excessive elimination of cal- cium is chiefly through the feces, Schabad^" aftei' equally extensive investigations, believes .that calcium starvation in children, from defec- tive absorption, may cause at least a pseudo-rickets, indistinguishable clinically or chemically from true rickets. But the fact that children with rickets show nearly normal figures for blood calcium does not agree with these calcium starvation hypotheses. As with osteomalacia, attempts have been made to associate with the etiology of rickets defects in the ductless glands, especially the adrenals,'" thymus, ^^ and parathyroids,'- but as yet without convinc- ing evidence. ^^ There has also been an attempt to include rickets "Howland and Marriott, ' Trans. Assoc. Amer. Phys., 1917 (32), 307. «« Jahrb. f. Kinderheilk., 1899 (50), 268. ^'' How metallic i)hosphorus causes growing bones to laj' on increased calcium is an unsolved problem, but a striking fact. (See Phemister, Jour. Amer. Med. Assoc, 1918 (70), 1737.) «8 See also Nathan, Med. News, 1904 (84), 391. *^ Articles in the Arbeiten a. d. Path. Inst. Tubingen, Vols. 6 and 7; also ^'erh. Deut. Path. GeselL, 1910 (-14), 294; Miinch. nied. Woch., 1910 (57), 2121. '0 Arch. f. Kinderheilk., 1909 (52), 47; 1910 (53), 381; 1911 (54), 83; Fortschr. Med., 1910 (28), 1057. '"■ Stoeltzner, Verh. Deut. Path. Ges., 1909 (13), 20. ^2 Erdheim et al, Frankfurter Zeit. Path., 1911 (7), 178. '^ Concerning the chemical changes of osteogenesis imperfecta (congenital fra- gility of bones), see Schabad, Zeit. Kinderheilk., 1914 (11), 230. 452 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS among the diseases that depend upon specific deficiencies in the diet, especially the fat-soluble "vitamines" which cod-liver oil supplies abundantly. So far, however, this hypothesis is not positively estab- Jished.^^ (See also Rickets, under ''Deficiency Diseases" Chapter xii.), CONCRETIONS All pathological concretions appear to be laid down according to a definite law. There must first be a nucleus of some substance differ- ent from the substance that is to be deposited, and which is most frequently a mass of desquamated cells, but may consist of clumped bacteria, masses of mucus, precipitated proteins, or a foreign body of almost any sort. Upon this nucleus substances crystallize out of solution, much as cane-sugar crystallizes on a string to form rock candy, but with the important exception that among the crystals is usually deposited more or less mucin or other organic substance, which forms a framework in which the crystals lie, and which remains, if the crystals are dissolved out, as a more or less perfect skeleton of the concretion. In no case would the concretion form were it not that the solution is overcharged with some substance, but not infrequently it is the presence of the nucleus that leads to the precipitation of the substance; i. e., the nucleus may play either a primary or a secondary role. With few exceptions, the dissolved substance is deposited in crystallme form, although the crystalline structure may in time partly disappear through condensation or through filling of the interstices with some other material. Even so structureless a substance as amyloid may, when forming concretions, appear in a crystalline form (Ophiils). The structure of a concretion depends upon two factors: The crystals tend to be deposited at right angles to the surface, and thus give a radiating structure; but the rate of deposition is usually irregular, and during the periods of quiescence the surface tends to become covered with mucin or other organic substances, hence we also get a concentric, laminated structure. Frequently both of these lines of formation are easily discerned, but either one or the other may be- come obscured. Concretions consist, therefore, of mixtures of colloids and crystal- loids deposited from solutions of the same character, and hence the application of the principles of colloidal chemistry throws much light on the conditions of their formation."'''^ Colloidal solutions hold in solution greater quantities of crystalloids than simple solutions, for the reason that at the surface of each colloidal particle there is a zone in which the crystalloids are more concentrated than elsewhere, thus permitting more crystalloids to be dissolved in the solvent ""^ See Paton, Findlay and Watson, Brit. Med. Jour. Dec. 7, 1918; Mellanbv* Lancet, 1919 (196), 407. ^6 See Schade, Munch, med. Woch., 1909 (5G), 3; 1911 (58), 723; Zeit. exp. Path., 1910 (8), 92; also Lichtwitz, Ergeb. inn. Med., 1914 (13), 1; also his mono- graph "Ueber die Bildung der Ilarn- und Gallensteine," Springer, Berlin, 1914. BILIARY CALCULI 453 between the colloidal particles. On the other hand, the concentra- tion of the crystalloids on the surface of the colhjidal particles causes the colloids to serve as the starting point of precii)itation whenever the crystalloids are in excess. When the crystalloid goes out of solu- tion, therefore, it will form crystals or precipitates which are most intimately associated with the colloids, as we see when uric acid crystallizes out of urine, taking with it the colloidal pigments by which it is absorbed. Or, if the colloids are precipitated, the solvent power of the solution is reduced, and the crystalloids will deposit in intimate relation to the colloids. As Schade pointed out, if a colloid precipi- tates in an irreversible form {e. g., fibrin), the concretion will be per- manent, as with ordinary concretions, but if the colloid precipitate is reversible the mass may be dissolved again, as with the precipitate of urates in the tubules of the infant's kidney. Biliary Calculi'^ As may be judged from the above statements, concretions are never composed of one substance in a pure form, but usually consist of a mixture of the constituents of the fluid in which they are developed. This is particularly true of gall-stones, which contain in greater or less quantities several or all of the constituents of the bile. While cholesterol forms the greater part of nearly all biliary concretions, and is present in greater or less amounts in all, calcium salis of the bile- pigments are always present; usually inorganic salts of calcium (car- bonate and phosphate) are also present, as well as small amounts of fats, soaps, lecithin, mucus, and other products," and occasionally traces of copper, '^^ iron, and manganese. '^^ The quantity of bile salts, the chief constituent of the bile, is usually extremely minute, apparently only so much as may percolate into the crevices of the con- cretion. However many stones there may be in a gall-bladder, they usually are all of approximately the same composition and structure. In gall-stones from the domestic animals the proportion of inor- ganic salts is usually much higher than it is in man. Naunyn has classified gall-stones according to their composition, as follows: 1. "Pure" Cholesterol Stones. — The purity is only relative, since even the purest alwaj'S contain some pigment as well as a stroma and a nucleus; but the amount of cholesterol may reach 98 per cent., and is usually over 90 per cent. Crystalline structure is usually well marked, while stratification is slight. The color varies from nearl}^ pure white to yellow, or even brown on the surface. 2. Laminated Cholesterol Stones. — These consist of about 75-90 per cent, of cholesterol, and differ from the preceding form in containing more pigment, which "8 Bibliography by Bacmeister, Ergeb. inn. Med., 1913 (11), 1. ^^ Fischer and Rose found about 0.1 gm. carotin in 1280 grns. gall stones from cattle. (Zeit. physiol. Chem., 1913 (88), 331.) '8 See Mizokuchi, Cent. f. Pathol., 1912 (23), 337. "^Gall-stones have been found enclosing droplets of mercury. (Xaunyn, Frerichs.) 454 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS is deposited in layers alternating with the white layers of cholesterol. The pig- ment here, as in all other gall-stones, consists always of the calcium salts of the pigments — not of pure bilirubin and biliverdin themselves. Considerable cal- cium carbonate is also usually present, particularly in the green layers of biliverdin- calcium. 3. Common Gall-bladder Stones. — The composition of this form is but little different from the above, the chief difference being in the structure. They pre- sent externally a firmer crust, usually distinctly laminated; in the center is a softer pigmented nucleus which frequently shows a central cavity containing fluid. Such calculi are not distinctly crystalline in structure, and are small, seldom larger than a cherry. 4. Mixed Bilirubin-calcium Calculi. — These generally occur singly, but some- times in groups of three or four, and are of large size. Although the chief con- stituent is bilirubin-calcium, there is always much cholesterol, often over 25 per cent. Copper and traces of iron may also be present. Their structure is lamin- ated, with sometimes a crystalline cholesterol nucleus. 5. "Pure" Bilirubin-calcium Calculi. — In addition to the chief constituent, bihver din-calcium, hilifuscin and hilihumin^^ are practically always present. Bili- humin is at times the chief ingredient, and may form over half of the substance; hilicyanin is rarely present. There is always some cholesterol, but sometimes only traces. These calculi are small, from the size of a grain of sand to that of a pea, and they occur in two distinct forms. One form is of wax-like consistence; the other is harder, steel-gray or black in color, with a metallic luster. Pure bili- rubin and biliverdin, not combined with calcium, are practically never present in concretions. 6. Rarer Forms. — (a) Amorphous and incompletely crystalline cholesterol gravel. Cholesterol externally giving a pearly luster; pigment in the center. (b) Calcareous Stones. — Consist chiefly of a mixture of calcium carbonate and bilirubin-calcium. Calcium carbonate may occur either as a superficial crust, or as small masses within an ordinary calculus; calcium sulphate and phosphate occur rarely in traces. Stones consisting mainly of calcium carbonate are ex- tremely rare in man, but more frequent in cattle and other herbivora, in which all forms of concretions contain much calcium, either combined with pigment or as carbonate and phosphate. A calcium oxalate gall-stone has also been described. ^^ (c) Concretions with included bodies, and conglomerate stones. (d) Casts of Bile-ducts. — Occur particularly in cattle, and consist chiefi}^ of bilirubin-calcium. Rarely and imperfectly formed in man. Aschoff and Bacmeister differ somewhat from Naunyn as to the composition of gall-stones, which they classify as follows: 1. Pure cholesterol stones. 2. Stratified cholesterol-calcium stones. 3. Cholesterol-pigment-calcium stones. 4. Composite stones, composed of cholesterol and a mantle of cho- lesterol and calcium. 5. Bilirubin-calcium stones, usually found in the bile passages of the liver. 6. The very rare calcium carbonate stones. Formation of QaII=stones. — Until quite recently our views con- cerning the chemistry and pathology of the formation of gall-stones *" Biliverdin differs from bilirubin in containing one more atom of oxygen in the molecule, and it is easily formed from bilirubin — even exposure to air will slowly bring about the oxidation. Bilifusciit is a still more oxidized derivative — so much so that it does not give Cmeliii's reaction (with HNO3+HNO2) for bile- l)igments. Bilihumin rei)resents the most oxidized of these products, is brown in color, and is the chief constituent of the residue left after treating gall-stones with ether, alcohol, and chloroform to dissolve out the cholesterol. ** Montlaur, Bull. sci. i)harmacol., Vol. IS, p. 19. BILIARY CALCULI 455 were dominated by the observations and conclusions of .Naunyn*^ and his pupils. Former observers, having learned that bile normally contains cholesterol (Hammarstcn found from 0.06-0. IG per cent, in human bile), sought the cause of gall-stones in either an increased elimination of cholesterol by the liver, or a decrease in the power of the bile to hold the cholesterol in solution. Thus Frerichs, finding that the presence of large amounts of bile salts and an alkaline re- action favored the solution of cholesterol, imagined that a diminu- tion of either bile salts or alkalinity led to the precipitation of the cholesterol. Naunyn and his pupils, however, not finding that the amount of cholesterol present in the bile depends upon the amount taken in the food or the amount present in the blood, and that it did not vary in disease, except when gall-stones were present, con- cluded that the cholesterol of the bile is neither a product of general metabolism nor a specific secretion-product of the liver. Finding that pus and the secretions from inflamed mucous membranes (bron- chitis) contained as much cholesterol as did normal bile, and often more, they concluded that the chief source of cholesterol in gall-stone formation was from the degenerating and desquamated epithelial cells of the gall-bladder and bile tracts. This idea was supported by the large amount of cholesterol found in the contents of gall-bladders shut off from the common duct, and by the formaton of gall-stones in such isolated gall-bladders. Some further evidence has since been brought forward in favor of this same view,^^ but others, finding no abundance of cholesterol in the wall of the gall-bladder have not accepted this origin.'*'* On the basis of Naunyn's hypothesis the ordinary steps in the for- mation of a cholesterol concretion are as follows: Some injury to the mucous membrane of the bile tracts is the starting-point; this injury is usually produced by infection, the colon and typhoid bacilli being the most common organisms in this process. ^^ Through the degenera- tion of the epithelial cells an excess of cholesterol is formed, while at *2 An English translation of this classic work, by A. E. Garrod, has been pub- lished by the Sydenham Society, 1896, vol. 158. *^ Thus Wakeman (quoted by Herter, Trans. Congress Amer. Physicians, 1903 (6), 158) was able to cause an increase in the cholesterol of the bile in the gall- bladder of dogs by injecting into it HgClo, phenol, or ricin. At first the choles- terol seems to be contained largely in the degenerating desquamated cells. Also the interesting case of a cholesterol calculus in a pyosalpinx, described by Thies (Arb. Path. Inst. Tubingen, 1908 (6), 422), shows the possibility of an inflamma- tory origin for such concretions, and independent of bile. ■«^ Aschoff, Miinch. med. Woch., 1906 (53), 1847 and 1913 (60), 1753; Aschoff and Bacmeister, "Cholelithiasis," Gustav Fischer, Jena, 1909; Laroche and Flandin, Compt. Rend. Soc. Biol., 1912 (72), 660. 8^ See Cushing (Johns Hopkins Hosp. Bull., 1899 (10), 166), who produced gall-stones experimentally by injecting typhoid bacilli into the circulation after injuring the gall-bladder. Literature on the relation of bacteria to gall-stones given by Funke, Proc. Path. Soc, Philadelphia, 1908 (11), 17; also see Rosenow who finds that streptococci are often responsible (Jour. Infect. Dis., 1916 (19). 527). Grieg notes the frequent occurrence of gall-stones in rabbits immunized with cholera vibrios (India Jour. Med. Res., 1916 (3), 397). , 456 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS the same time the desquamated cells and clumped bacteria offer suitable nuclei upon which the cholesterol begins to crystallize out. Apparently after the calculi have reached a certain size they cause sufficient mechanical injury to keep up the cell degeneration and chol- esterol formation, even after the infection has subsided. A certain amount of infection and inflammation is a favoring condition, however, for Harley and Barratt^*' found that fragments of cholesterol calculi introduced aseptically into the gall-bladders of dogs were slowly dis- solved and disappeared, but this was prevented by infecting the gall- bladder with B. coll. According to Naunyn's investigations, it is not an alteration in the composition of the bile, as formed in the liver, which causes the precipitation of cholesterol, but rather the presence of the nidus, and the production of large quantities of cholesterol in immediate proximity to this nidus, that determines the formation of a concretion. In case the bile stagnates in the gall-bladder, the choles- terol that is being constantlj^ formed by the normal disintegration of surface epithelium accumulates, until, even without infection, there forms a sediment of soft yellowish and brownish masses, consisting chiely of cholesterol and bilirubin-calcium. From this material calculi may eventually form, and by their irritation lead to further formation of cholesterol and increased growth." But bacteriological studies indicate that generally an infectious influence is present in cholelithiasis, and bacilli may be found alive in gall-stones for remark- ably long periods. Recent applications of colloidal chemistry add much to our under- standing of gall-stone formation. Thus, Lichtwitz points out that the colloids of normal bile, all of which are electro-negative, may be precipitated by positive serum colloids coming from the blood when the gall-bladder is inflamed; hence we get a precipitate of cholesterol, bilirubin and proteins. When the colloids are thus thrown down the solvent power of the bile for the alkali earths it contains is decreased, and so calcium or magnesium are added to the mixture. Cholesterol is in solution in the bile as an emulsion colloid, and when stagnation of the bile leads to absorption or disintegration of the chelates and fats which keep it in solution, the droplets become confluent, and then crystallization takes place (Schade) with formation of spheroliths, and eventually a crystalline cholesterol calculus. If even the slightest pressure is brought to bear on the myelin-like masses before they crystaHize, however, they will be pressed into scales, and the common laminated structure results; hence crystalline calculi are single, while multiple gall-stones are laminated, with perhaps partial crj^stallization between the lamellse. Also when the gall-stones result from inflamma- tion, and there is much serum colloid present, the stones are lamellated 88 Jour, of Physiol., 1903 (29), 341; see also Ilansemann, Virch. Arch., 1913 (212), 139. " Concerning the structure of gall-stones see Ribbert, Virchow's' Arch., 1915 (220), 20. BILIARY CALCULI 457 because these colloids deposit in that form {e. g., corpora amylacea and other protein concretions). These considerations explain the for- mation of gall-stones in the gall-bladder from either inflammation, or stagnation without inflammation. Aschoff and Bacmeister,**** however, hold that the usual series of events in the formation of gall-stones is first the formation of a pure cholesterol stone without inflammatory cause, because of actual in- creased excretion of cholesterol by the liver, because of cholesterolemia; or because of resorption of solvent substances from stagnating bile: these primary cholesterol stones then cause inflammation and occlu- sion, leading to the formation of the common mixed stones. Bac- meister ascribes more importance to calcium than do most o<"her investigators, in which he is supported by Rosenbloom,*^ while Kuru^" states that fibrin is usually present. Boscnbloom reports a small series in which concretions composed chiefly of calcium were found in all cases with a history of infection, while in cases without infection the stones we;*e cholesterol. More recent studies of the cholesterol content of the blood and bile also have reacted against the concept that all the cholesterol of gall- stones comes from the wall of the bile tract through inflammatory changes. It has been found that patients with gall-stones often show a hypercholesterolemia;^^ that pregnancy, which seems to be a predis- posing cause of cholelithiasis, is accompanied by hypercholesterolemia; that in races subject to cholelithiasis there is more cholesterol in the diet and in the blood than in those races that seldom have gall-stones (DeLangen) ; that with hypercholesterolemia there is an increased out- put of cholesterol in the bile, and that experimental hypercholester- olemia may lead to the formation of gall-stones without evident infection of the bile tracts (Dewey^^). As far as the existing evidence permits one to draw conclusions, it would seem probable that both local and systemic conditions are of iniyortance iii gall-stone formation. Apparently, gall-stones may form from cholesterol derived from the inflamed bile tract walls, independent of the amount of cholesterol present in the bile; but presumably they may derive part if not all the cholesterol from the bile in some cases. In either event, a hj^percholesterolemia will favor their formation, and hence any given condition of injury to the gall bladder will more often give rise to concretions in persons with a high cholesterol content in the blood. ^^ Changes in the bile itself may be produced by disease 88 Ziegler's Beitr., 1908 (44), 528. 89 Jour. Amer. Med. Assoc, 1917 (69), 1765. »" Virchow's Arch., 1912 (210), 433. »i Henes, Surji., Gyn. and Obst., 1916 (23), 91. »2 Arch. Int. Med., 1916 (17), 757; see also Aoyama, Deut. Zeit. Chir., 1914 (132), 234. 9^ This relation of hypercholesterolemia and infection to cholelithiasis is sup- ported by the extensive observations of Rothschild and Wilensky (Anier. Jour. Med. Sci., 1918 (156), 239, 404, 564; Arch. Int. Med., 1919 (24), 520), who find some types of cases accompanied by cholesterol increase, which is missing in many cases of inflammatory cholelithiasis. (See also Reimann and Magown, Surg., Gynec. and Obst., 1918 (26), 282; Fasiani, Arch. Sci. Med., 1918 (41), 144. 458 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS of the liver that will alter bile compositon in such a way that its capac- ity to sustain cholesterol in solution or suspension will be lowered, ^"^ and this factor also cannot be dismissed as without importance; tran- sient thickening of the bile, such as may occur in any febrile disease, may also very possibly initiate precipitation and stone formation. More and more this last factor is receiving consideration, together with hypercholesterolemia, as of importance in producing cholelithiasis. Rovsing,^^ quoting Boysen's analysis of 200 autopsy cases of choleli- thiasis, which showed that all recent deposits and the centers of older concretions consisted of calcium-pigment, especially emphasizes this transitory concentration of bile. It was formerly supposed that the calcium-pigment concretions were produced by the presence of excessive calcium in the bile, derived particularly from lime-laden drinking-water, but it has been demon- strated that increase of calcium in the food does not cause an increase in the amount in the bile. Furthermore, on concentrating bile, which contains both bilirubin and calcium, the free bilirubin separates out and not the calcium compound of bihrubin; and also, Naunyn found that the bile salts prevent precipitation of calcium-bilirubin, even when calcium salts are added in considerable amounts. Appar- ently it is the presence of positively charged protein substances that leads to the precipitation of this electro-negative substance from bile, and hence the formation of pigment calculi is also favored or initiated by inflammation of the bile tracts, particularly as most of the calcium salts seem to come from the mucous membrane ;^'^ later, as we have seen, these pigment concretions often become covered with cholesterol derived from the injured epithelium, and the common mixed calculi are then formed. In view of the fact that much of the pigment in these calculi is composed of the oxidation products of bilirubin, especially bilihumin, it is possible that oxidation processes in the stagnating bile are important causes of the precipitation; Naunyn suggests that bacteria may be the cause of the oxidation. Pigment calculi are par- ticularly important as the starting-point of the larger mixed calculi. Aufrecht,^^ indeed, holds that gall-stone formation usually begins with particles of pigment that are expelled from the liver cells as such, and ordinarily are discharged into the intestine; if they make their way back into the gall-bladder they form the nuclei of concretions. It is possible, Naunyn believes, for the pigment to be later gradually replaced by cholesterol. "^ See D'Amato, Biochcm. Zoit., 191.'5 (G9), 353. '•^ Hospitalstidende, 1915 (5S), 249. "" This commonly-held view is denied by Liehtwitz and Book (Dent. med. Woch., 1915 (41), 1215), who fonnd the calcium content of bile from fistnlns to be from 65-84 mg. per liter, and in bladder l)ile to vary from 85 to 325 mg., but not according to the presence or absence of inflammation. »' Deut. Arch. klin. Med., 1919 (128), 242. URINARY CONCRETIONS 459 Urinary Calculi'* These differ from the bile concretions in two important respects: First, there is no evidence that any considerable part of their con- stituents may come from the walls of the cavities that contain them ; they are usuall}^ deposited on account of an over-saturation of the urine, or on account of a change in composition of the urine, which renders them insoluble. Second, the composition of urinary calculi is usually less mixed than that of biliary calculi, although seldom, if ever, is it pure. Thus, Finstcrer found but six concretions composed of onlj' one substance, in a collection of 114 calculi. As with the bile, the chief constituent of the urine (urea) is so soluble that it never forms concretions, but only the less soluble minor constituents are thrown down. For the formation of calculi, however, it is not sufficient to have merely an excess of a substance in the urine, for we may have deposition of urates, phosphates, or uric acid in simple crystalline form without the formation of calculi. A nucleus of some sort is present as well as a binding suhsiance,^^ which is often mucus derived from the walls of the passages, although the center of the concretion most often consists of uric acid or urates. Although the amount of colloidal material in urine is relatively small, yet it undoubtedly plays an important part in maintaining in solution the less soluble crystalloids, which are especially the urates and calcium oxalate. Normal urine contains no colloids which form irre- versible gels, and hence ordinary deposits can be readily dissolved, but in inflammatory conditions there appears fibrinogen which readily forms the irreversible fibrin, and conditions thus become favorable for the formation of concretions of any crystalloid with which the urine may be saturated or over-saturated at the time (Schade). Possibly other colloids may play a similar role. Aschoff and Kleinschmidt' hold that most urinary calculi begin as primary calculi, formed inde- pendent of inflammation from excess of the main constituent (uric acid, oxalates, xanthine, but chiefly ammonium urate) ; this calculus forms the crystalline nucleus of the laminated secondary deposits of other substances, chiefly uric, acid, oxalates and phosphates, all being deposited without inflammation. The inflammatory formations con- sist chieflj'^ of ammonio-magnesium phosphate and ammonium urate, usually deposited on a foreign body or a primarj^ calculus. The ex- tensive study of the microscopic structure of urinary calculi by Shattock,- shows also that a nucleus of cells or other organic material is, at least in uric acid calculi, extremely rare, the center being almost always a primary crystalline deposit from a supersaturated solution. "* General Bibliography given by Finsterer, Deut. Zeit. klin. Chir., 1906 (80), 41 -4; and Lichtwitz.'= ^' Hippocrates appreciated the existence and importance of the mucoid binding substance in urinary concretions (Schepelmann, Berl. klin. Woch., 1911 (48), 525). ^"Die Harnsteine," Berlin, Julius Springer, 1911. 2Proc. Roy. Soc. Med., Path. Sec, 1911 (4), 110. 460 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS Calculi formed because of changes in the urinary composition in- dependent of evident infection are often called "primary," in con- tradistinction to those arising from changes in composition brought about by infection and ammoniacal decomposition. Because of the injury produced by a primary calculus, infection frequently results, and then the primary calculus may become the nucleus of a secondary calculus; indeed, on account of the change of reaction, the crystalloids of the primary calculus may be dissolved out, and their place taken by the secondary deposit {metamor'phosed calculi). In structure urin- ary calculi usually show both radiating and concentric lines of forma- tion, and when the chief constituents are dissolved away, an organic framework remains. They are generally classified according to their prominent component, as follows: Uric=Acid Calculi. — Uric acid is but shghtly soluble, only one part dissolving in 39,480 of pure water at 18°, and it is even less soluble in the presence of acids.^ The presence of sodium diphosphate in the solution makes it much more soluble, and various organic bodies also favor its solution, among them being the urinarj^ pigments. As can be seen, the maintenance of uric acid in solution is by a small margin, even in normal conditions; hence the mere cooling of the urine fre- quently suffices to cause an abundant deposition of uric acid combined with pigment, as the familiar "brick-dust" deposit. The formation of uric-acid calculi is, therefore, not only a question of the amount of uric acid in the urine, but depends even more upon the amount of the substances that hold it in solution, and as both these factors are sub- ject to wide variations under both physiological and pathological con- ditions, uric acid and urates are common in urinary concretions. The older literature indicates that the most common calculus is of this nature, but a number of recent analyses indicate that the im- portance of uric acid and urates has been overestimated. On the con- trary, this material rarely forms a considerable part of the calculi, but is usually present in greater or less amount in most or all urinary calculi (Kahn).^ It is probable, however, that uric acid is important as furnishing the primary nucleus of calculi of preponderatingly cal- careous or mixed composition. Apparently there are marked differ- ences in the prevailing composition of calculi in different countries; in China, for example, Pfister'^ found eleven of twelve calculi composed of uric acid. Uric acid is eliminated combined chiefly with sodium, potassium, and ammonium; according to some authors, as abiurate, according to others, as a quadriurate. If the urine is excessively acid, it con- ^ Concerning solubility of uric acid in urine see Haskins, Jour. Biol. Chem., 1916 (26), 205. * Arch. Int. Med., '1913 (11), 92; review of literature. Rosenbloom, (Jour. Amer. Med. Assoc, 1915 (65), 161) found but two uric acid stones of twenty-six analyzed. 6 Zeit. Urol., 1913 (7), 915. URINARY CONCRETIONS 461 tains much acid j^hospliates, wliicli withdraw part of the bases from the uric acid, and this, when free, crystalhzcs out if in excess. Hence the formation of uric-acid concretions is favored by high acidity of the urine, by concentration of the urine, or by an increased ehmina- tion of the uric acid. The last may result from excessive nuclein- rich food, or from excessive catabolism of the tissue nucleoproteins (e. g., leucocytosis from inflammatory diseases or leukemia), which conditions are also usually associated with an increased urinary acid- ity. The chemistry of uric acid is discussed more fully in the chap- ter on Gout, Chap, xxiii.) Uric-acid calculi are formed chiefly in the pelvis of the kidney, but many pass into the bladder. They are quite hard, and yellow or reddish-yellow in color, because of the presence of urochrome and urobilin, the former of which seems to be chemically combined and the latter but physically, since it can be washed out with water. Uroerythrin or uromelanin (a decomposition product of urochrome) may also be present. Not infrequently calcium oxalate is present, .sometimes in considerable quantities. Other urinary constituents may be present in small amounts. In case the calculus enters the urinary bladder it may set up irritation leading to infection; the urine then becoming alkaline, calcium and ammonio-magnesium phosphate will be deposited upon the surface, and the uric acid will be more or less dissolved out and replaced by the phosphates (metamorphosis). Urate calculi occur chiefly in new-born or young infants, and rarely in adults. In the young they are related to, and may originate in, the deposits of urates in the pyramids of the kidney (the so-called urate or uric-acid "infarcts"), which have been supposed to result from the decomposition of the nucleoproteins of the nucleated fetal red corpuscles. (See Uric Acid, Chap, xxiii.) The concretions are composed chiefly of either ammonium or sodium urate, but potassium and even calcium and magnesium urate may be admixed. Their genesis in the young probably depends upon injury to epithelium by the excessive urates of the "infarcts," which affords a suitable nucleus for their start; their growth depends chiefly upon the concentration of the infant's urine. In adults they may arise secondary to an am- moniacal decomposition of the urine. Urate concretions are not com- mon; they are generally rather soft, and often much colored b}' pigments. Calcium oxalate calculi are, according to recent observers,* the most common urinary concretions.^ Often they show admixtures of urates or uric acid, which latter frequently constitutes the nucleus, and when urinary infection occurs they may in turn serve as the nucleus to phosphatic deposits. On account of the hardness and roughness of tliese stones they frequently cause bleeding, which may result in their ^ Concerning their structure see Fowler, Johns Hopkins Hospital Reports, 1908 (13), 507. 462 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS being very dark in color and containing blood-pigment. They are usually first formed in the pelvis of the kidney, and arise chiefly in persons excreting excessive quantities of oxalic acid. Normally but about 0.02-0.05 gram of oxalic acid is eliminated daily in the urine, apparently all as calcium oxalate, which is kept in solution by the acid phosphates. The amount may be increased by certain foods rich in oxalates, particularly rhubarb, grapes, spinach, etc.; also prob- ably by gastric fermentation.'^ Oxalic acid may possibly be formed from uric acid, and perhaps also from the carbohydrate group of proteins,^ and it is possible that abnormally large amounts arise from these sources under pathological conditions. During bacterial de- composition of the urine oxalic acid may be formed from uric acid (Austin). 9 Phosphate calculi are formed as a result of decomposition of the urine, with formation of ammonia from the urea.^" In the ammoniacal solution thus formed the magnesium is precipitated as NH4MgP04, the calcium as Ca3(P04)2, and calcium oxalate and ammonium urate are also thrown down, so that the concretions consist of a mixture of these substances, the magnesium salt being the most abundant. In none does one substance occur in a pure state. Pigments of various kinds, and more or less mucus or other organic constituents of the framework are also present. Phosphate calculi are the typical "sec- ondary" concretions, and they are formed usually in the bladder as a consequence of cystitis, but may be formed in the renal pelvis or in the urethra. In some cases the salts are precipitated in such large quantities that they form great masses of a sediment which does not aggregate into concretions. Occasionally stones consisting princi- pally of Ca3(P04)2 or CaHP04 are formed, but these are rarities. As the calcium taken in the food is chiefly eliminated in the feces, the amount in the urine does not vary directly with the amount in the food, and the formation of phosphatic concretions is always a matter or urinary reaction and not of diet.^' .4s these stones fuse to a black, enamel-like mass under the blow-pipe, they have been called "fusible calculi." Calcium carbonate calculi are formed frequently in herbivora, but they are very rare in the urinary passages of man, although occurring elsewhere intliebody ' Baldwin, Jour. Exp. Med., 1900 (5), 27. * See Austin, Boston Med. and Surg. Journal, 1901 (145), ISl. Contradicted by Wegrzynowski, Zeit. physiol. Chem., 1913 (83), 112. 9 Jour. Med. Research, 1906 (15), 314. "> Under the name "struvit stone," Pommer (Vcrh. deut. Path. Gesell., 1905 (9), 28) describes a urinary calculus composed of very pure ammonio-magnesium phosphate, forming the hard, rhoml)ic crystals known to mineralogists as "struvit." This is an example of a phosphate stone formed independent of ammoniacal decom- position, a rare occurrence. •'Osborne (Jour. Amer. Med. Assoc, 1917 (69), 32) has observed numerous cases of formation of phosphate calculi in tlie urinary bladder of rats kept on diets deficient in fat soluble vitamines. The reason for this association of diet and concretions is not known; possibly the dietary deficiency causes lessened resist- ance to urinary infection. URINARY CONCRETIONS 463 not infrequently. Occasionally these arc soft and chalky, hut if well crystallized, they are the liartlost of concretions. Cystine calculi'- are rare but very interesting formations. Cystine S-CH(N1I...)-C()()1I I is important as the sulphur-containing portion of the protein S-CII(NH2)-C()0II molecule. Under normal conditions all the cystine taken in food is completely oxi- dized and none (or uncertain traces) appears in the urine. In certain individuals the urine contains considerable quantities of cystine constantly {cyslinuria, see Chap xxi), and occasionally in these cases soft concretions of nearly pure cystine are formed in the urinary passages. Cystine calculi may reach the size of a hen's egg, are crystalhne in structure, and in the urine of such patients the characteristic hexagonal crystals may usually be found. The cystine of calculi is identical with that from proteins and may be associated with tyrosine.'^ Xanthine Calculi — Xanthine is the most abundant of the purine bases normally present in urine, but the total amount is extremely small. Like uric acid, it fluc- tuates in amount according to the amount of destruction of nucleoproteins, either of the food or of the tissues. Concretions consisting chiefly of xanthine, which is often mixed with uric acid, are extremely rare, but a few isolated specimens having been described. Rosenbloom could collect but six cases in the literature, adding one himself.'* Indigo calculi, derived from the indican of the urine through oxidation, have also been described a few times. Urostealith calculi, composed of fatty matter, have been occasionally observed. Although some of the concretions described under this head have really repre- sented foreign bodies introduced through the urethra (e. g., Kruckenberg's concre- tion of paraffin from a bougie), yet true fat concretions do occur. The origin of the fat in these stealiths is unknown; possibly it comes from degenerated epithelium. Horbaczewski'5 analyzed such a specimen which had the following percentage composition: Water 2.5 Inorganic matter 0.8 Organic matter (chiefly protein) 11.7 Fatty acids 51.5 Neutral fat 33 . 5 Cholesterol traces The fatty acids consisted of stearic, palmitic, and probably myristic acid. Cholesterol calculi have been found in the urinary bladder in a few instances, the cause being unknown. Horbaczewski^* describes one weighing 25.4 grams, found in a patient who had previously had cystine calculi; it contained 95.87 per cent, of cholesterol and but 0.55 per cent, of inorganic material. Gall-stones have been known to enter the urinary bladder through a fistula between the gall-bladder and urinary bladder. ^^ Fibrin "calculi," formed from blood-clots, often more or less impregnated with urinary salts, have occasionally been observed. Other proteins may also form simi- lar calculi.'" General Properties of Urinary Concretions.'^ — The hardness depends partly upon the chemical composition of the calculus, but more upon the rate and condition of formation (Rowlands, Kahn). 1- Literature concerning cystine, see Friedmann, Ergeb. der Physiol., 1902 (i), 15; Marriott and Wolf, Am. Jour. Med. Sci., 1906 (131), 197. 13 Abderhalden, Zeit. physiol. Chem., 1907 (51), 391; 1919 (10-1), 129. » N. Y. Med. Jour., Jan. 16, 1915. 's Zeit. physiol. Chem., 1894 (18), 335. 16 See Finsterer, Deut. Zeit. klin. Chir., 1906 (80), 426. 1^ See Morawitz and Adrian, Mitt. Grenz. Med. u. Chir., 1907 (17), 579. 1* Systems for procedure in determining the nature of urinary calculi are given by Hammarsten (Text-book of Phj'siol. Chem.) and by Smith (Reference Hand-book of Med. Sci.). 464 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS Under comparable conditions it is said that those composed of amorph- ous phosphates are the softest; next come those with some admixture of crystalhne phosphates. Urate concretions are harder than these, but are still softer than uric acid and crystalline phosphate calculi. Oxalates are usually the hardest, except for the rare crystallized calcium carbonate stones. Cystine and amorphous concretions can be scratched with the finger-nail, while even the hardest varieties of calculi can be scratched with a wire nail. Genersich^^ gives the following degrees of hardness for different calculi: Cholesterol, 1,5- 1.6; ammonium urate, 2.5; soft phosphate (Mg), 2.6; hard phos- phate (Ca), 2.75; uric-acid stones (also salivary and prostatic calculi, atheromatous patches, and phleboliths), 2.9; calcium oxalate (also rhinoliths and lung stones), 3.3-3.5; calcium carbonate stones of herbivora, 4.5. But the hardness or gross appearance of a urinary calculus give little or no indication of its chemical composition. The rate of growth also varies according to composition, but is, of course, much modified by other factors. Oxalate and urate stones grow most slowly, phosphate stones most rapidly. A urate stone has been known to increase by about two ounces during seven and one half years, while a catheter fragment or other foreign body may be- come covered with a crust several millimeters thick in a few weeks. ^"^ Spontaneous disintegration of urinary concretions is limited almost solely to calculi composed entirely or largely of uric acid. Out of 121 cases collected by Englisch.^i in all but 7 this was the case, these being composed of calcium and magnesium phosphate (5), or calcium phosphate or carbonate (1 each). The disintegration is brought about through solution of the binding substance and me- chanical shattering of the stone into fragments. This occurs but rarely, Bastos- estimating that perhaps one calculus in ten thousand undergoes disintegration. Corpora Amylacea^^ In the case of these widely-spread concentric bodies we find the name misleading, for the bodies are not a form of animal starch, as was suggested by their laminated structure and iodiu reaction, nor are they so closely related to amyloid material as the name implies. Different authors disagree decidedly concerning the staining reactions of these bodies, but it may be said that the reactions are extremely inconstant. Sometimes the corpora are stained bluish or green with iodin, sometimes brown, often little at all; occasionally they react partly with methyl- violet, but more often they do not; sometimes por- tions of one body react one way, while the remainder behaves differ- ently. Seldom if ever do the ordinary concretions of the prostate " Virchow's Arch., 1893 (131), 185. 20 ZuckerkandLNothnagel's System, vol. 19, pt. 2, p. 229. 21 Arch. klin. Chir., 1905 (76), 961 (elaborate revie^v). " Folia Urol., 1913 (8), 81. " General literature, Posner, Zeit. klin. Med., 1889 (16), 144; Lubarsch, Ergeb. allg. Pathol., 1894 (lo) 180; Ophiils, Jour. Exp. Med., 1900 (5;, 111; Nunokawa, Virchow's Arch., 1909 (196), 221; Brutt, ibid., 1912 (207), 412. CORPORA A.^rVLACEA 465 give all the amyloid reactions characteristically, but the corpora amylacea of the lungs are much more likely to do so (Stumpf).^' It seems improbable that these bodies, which occur in the prostate of every adult, can be the same as the amyloid, which is seldom observed except as the result of serious processes of tissue destruction. Accord- ing to their structure they obey the usual laws of the formation of concretions, having a central nucleus and a structural framework of different composition from the chief substance. It seems most prob- able that they should bo interpreted as simple concretions of protein nature, which form under certain conditions when a nucleus of some sort (usually pigment, degenerated cells, or inorganic crystals) exists in a stagnating, protein-rich fluid. At limcs the resulting concretion may be of such a physical nature that it absorbs iodin readil}^ (just as they often show a marked absorption-affinity for pigments), and occasionally it may react metachromatically with methyl-violet, pos- sibly because of the presence of chondroitin-sulphuric acid derived from the mucin of the cavities where the concretions form, but per- haps for some other unknown reasons. Occasionally pure amyloid may form in the tissues typically concentric (or even crystalline) bodies, as in Ophiil's case, but this is the exception. It seems prob- able that corpora amylacea are usually protein concretions,^'^ and neither amyloid nor animal starch. Those formed in the central nervous system may be of myelin or neuroglia origin.-^ The small amount of material available prevents an accurate analysis of the corpora amylacea; it is known that they are very in- soluble in water, acids, alkalies, etc., behaving like coagulated protein in this respect. Even hot concentrated nitric acid will not dissolve them, according to Posner. This author considers lecithin and cho- lesterol to be important constituents, and by Ciaccio's staining method lipoids can be found in prostatic corpora amylacea. ^^ How- ever, it is said by Bjorling-^ that the ordinary hyaline and granular corpora do not contain fats or lipoids, but that a certain class of "lipoid" prostatic concretions contain many granules of this nature. The corpora amylacea of the lateral ventricles seem to consist chiefly of calcium salts deposited in a concentric arrangement through the medium of an organic basis. Posner considers that the presence of lecithin in prostatic corpora prevents their calcification, although this change occasionally does occur. Other Less Common Concretions Pancreatic Calculi.-^ — The cause of the formation of stones in the pancreatic duct is not definitely known, but apparently infection is the most important factor, " Virchow's Arch., 1910 (202), 134. " Ramsden's observations (Proc. Royal Soc, 1903 (72), 156) on the precipi- tation of proteins by the action of surface contact may have some bearing on the formation of such protein concretions. 2« See Lafora, Virchow's Arch., 1911 (205), 295. " Posner, Zeit. f. Urologie, 1911 (5), 722. 28 /bid, 1912 (6) 30. 2^ Literature by Scheunert and Bergholz, Zeit. phj'siol. Chem., 1907 (52), 338. 30 466 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS since simple experimental stasis will not cause their formation.'" The calculi consist usually of a mixture of calcium phosphate and carbonate, associated with more or less organic matter, including frequently cholesterol, but all the usual products of proteolysis may be present because of the presence of trypsin. Oc- casionally the calculi consist chiefly of calcium carbonate, which may be almost pure.'^ Shattock^2 ^^g observed a pancreatic concretion composed of calcium oxalate. Sodium phosphate and chloride, magnesium phosphate, and proteins have also been found in these concretions. Taylor^' describes a pancreatic con- cretion containing, according to the analyst, chiefly silicate (!), a finding difficult to understand or accept. Baldoni'^ found, on analysis of a stone weighing 3.1 grams, the following percentage composition: Water 3 . 44 Ash 12.67 Proteins 3 . 49 Free fatty acids 13 . 39 Neutral fatty acids 12 . 40 Cholesterol 7 . 69 Pigments and soap 40 . 91 Undetermined 6.01 Usually, however, pancreas stones consist chiefly of inorganic substances. Johnson and WoUaston report analyses of two stones, one containing 72.30 per cent, calcium phosphate and but 8.80 per cent, organic matter; the other 91.65 per cent, calcium carbonate, 4.15 per cent, magnesium carbonate, and but 3 per cent, organic matter. Legrand'^ found only 0.7 per cent, organic matter in another concretion which contained 93.1 per cent, calcium carbonate. Pancreatic juice being strongly alkaline, can hold but a small quantity of calcium salts in solution (normally but 0.22 part per thousand — C. Schmidt); presumably the little nor- mally present is held in the form of a colloidal suspension by the proteins. Possi- bly when stasis occurs, digestion of the proteins leads to the precipitation of the calcium salts, or, more probably, the excessive calcium is largely derived from the exudate from the inflamed ducts, as seems to be the case with the calcium of biUary calculi. Salivary Calculi.'^ — These have a similar composition, in the main, to the con- cretions of the pancreatic duct, except that they generally contain more organic matter, resembling in this respect the "tartar" of the teeth." Bessanez found in one 81.3 per cent, of calcium carbonate and 4.1 per cent, of calcium phosphate, whereas in another the carbonate was but 2 per cent, and the phosphate 75 per cent. Potties has described a calculus with a central portion composed chiefly of uric acid and a peripheral portion containing 69 per cent, of calcium phosphate and 20.1 per cent, of calcium carbonate. Harlay^* found in one specimen 15.9 per cent, organic matter, 75.3 per cent, calcium phosphate, 6.1 per cent, calcium carbonate. Roberg believes that bacteria alone do not usually cause salivary calculi to form, but that a foreign body entering the duct is the chief factor. In- creased alkalinity may also favor precipitation of calcium from the saliva. In Roberg's case of sialolithiasis the saliva was of normal composition. Intestinal Concretions.— These always have a nucleus of some indigestible foreign substance, most often hair, but sometimes cellulose structures or solid indigestible particles, including gall-stones, fruit-stones, bone, etc. The bulk of the concretions is usually made up chiefly of ammonio-magnesium phosphate, with some calcium phosphate, carbonate, and sulphate, protein matter, and occasionally '« See Lazarus, Zeit. klin. Med., 1904 (51), 530. Literature. " Rosenthal, Arch. f. Verdauungskr., 1914 (20), 619. " Brit. Med. Jour., 1896 (i), 1034. " Lancet, Dec. 18, 1909. " Schmidt's Jahrb., 1900 (268), 210. « Jour. Pharm. et Chim., 1901 (14), 21. '"Literature, see Roberg, Annals of Surgery, 1904'(39), t)69. " Particles of gold have been found in a salivary calculus by Maurin (Repert pharm., 1919 (30), 257), presumably derived from fillings. '8 Jour. Pharm. et Chim., 1903 (18), 11. INTESTINAL CONCRETIONS 467 calcium and magnesium soaps. Two intestinal concretions analyzed by Schuberg" had the following percentage composition when dried : Ammonio-magnesiuni phosphate 57. 1 63.9 Calcium i)hosphate If).? 23.8 Calcium carbonate 4.0 Calcium sulphate 3.0 0.7 Alcohol-ether extract 1.9 0.8 Other organic substances 21 .5 6.0 In countries where oatmeal is largely eaten, intestinal concretions are not infre- quent; they contain calcium and magnesium phosphate, about 70 per cent.; oatmeal bran, 15-lS per cent.; soaps and fats, about 10 per cent. (Hammarsten). Occasionally concretions consisting largely of fats and soaps are found, and after taking large doses of olive oil masses of solidified oil may be pa.ssed that are readily mistaken for softened gall-stones, for the removal of which the oil is usually given. The "fecal stones" found in appendices often show the structure of calculi, and, unlike other enteroliths, consist less of ammonio-magnesium phosphate than of calcium salts i^" soaps may be important constituents.''^ Bezoar stones are intestinal concretions probably coming from Capra aegngrus and Antelope dorcas. One variety consists chiefly of lithofellic acid, C2nH.-)fi04, which is related to cholalic acid, and gives an aromatic odor when heated. The other variety ("false bezoars") does not give the aromatic odor, and consists chiefly of ellagic acid, Ci4H60s, a derivative of gallic acid, and, therefore, probably derived from the tannin of the food of the antelopes. Intestinal "sand" occurs as (1) "false sand," consisting of particles of indi- gestible food, such as the sclerenchymatous particles in the flesh of pears and bananas;" and (2) true sand, consisting largely of inorganic material, and formed, according to Duckworth and Garrod,^-^ in the upper part of the large intestine. Analyses of specimens by Garrod showed the following composition: Water 12.4 f calcium oxide 54 . 98 Organic material 26.29 I phosphorus pentoxide 42.35 Inorganic material. . . 61 31 containing | carbon dioxide 2.20 [ traces of Mg, Fe, etc 0 . 47 Analyses by other observers have given similar results, the absence of the large proportion of magnesium found in larger concretions being striking. The color is usually brown, due chieflj^ to urobilin, unaltered bile-pigments being scanty. Preputial concretions sometimes form beneath a prepuce that cannot be retracted, through deposition of urinary salts on and in the accumulated smegma. ^^ The composition is, therefore, very mixed, and consists of an organic base contain- ing much cholesterol, fats, and soaps, incrusted with inorganic substances, of which ammonio-magnesium phosphate and calcium phosphate are usually the most abundant. Prostatic concretions originate in the corpora amylacea through growth ac- cretion of inorganic salts, until they may reach considerable size. Stern^' gives the following results of analysis of such a prostatic stone: Water 8.0 Organic matter. . . 15.8 Lime ' 37 . 64 Magnesia 2 . 38 Soda 1 . 76 Potash 0.5 Phosphoric acid 33 . 77 Iron trace '9 Virchow's Arch., 1882 (90), 73. *° Harlay, Jour, pharm. et chim., 1910 (2), 433. ^' Williams, Biochem. Jour., 1907 (2), 395. " Myer and Cook, Amer. Jour. Med. Sci., 1909 (137), 383. •"■^ Lancet, 1902 (i), 653. Full resume and literature. " See Zeller, Arch. klin. Chir., 1890 (41), 240. « Amer. Jour. Med. Sci., 1903 (126), 281. 468 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS Lung stones.*^ — These may be formed in the bronchi, through accretion about an inorganic nucleus, similar to the formation of calculi in other epithelial-lined passages; or they may consist of calcified areas of lung tissue or peribronchial glands, which have been sequestrated through suppuration and have entered the bronchi. In the latter case, the calculi present the usual composition of patho- logical calcified areas. That the expectorated stones frequently represent calcified tubercles is shown by Stern" and by Biirgi." who demonstrated tubercle bacilli in decalcified lung stones. The following percentage figures are taken from Ott:*' Calcium phosphate 52 . 0 72 . 8 Magnesium phosphate 1.0 Magnesium carbonate 2.0 Calcium carbonate 13 0 6.0 Fat and cholesterol 24 . 0 7.0 Other organic substances 4.0 10.0 Rhinoliths'^ are formed about nasal secretions, blood-clots, and most frequently about foreign bodies. They therefore contain much organic substance in addition to the inorganic salts deposited upon them Berlioz" gives the following table from the analysis of four specimens ; Weight of specimens, grams 1 3.75 2 1.34 3 0.63 4 0.95 Water Organic matter Calcium phosphate Magnesium phosphate Calcium carbonate Traces of iron 5.80 16.60 62.02 5.08 10.50 Doubtful. 5.10 18.20 60.61 6.28 9.81 Distinct. 4.00 16.00 61.40 3.93 14.67 Doubtful. 6.90 18.10 47.63 6.68 20.69 Distinct. Tonsillar concretions consist chiefly of carbonate and phosphate of calcium deposited upon the inspissated secretions and desquamated cells of the tonsillar crypts.^" According to some authors, leptothrix threads frequently form the nucleus of the concretions. Cutaneous concretions are occasionally observed, located chiefly in the sub- cutaneous tissue, often occurring multiple. The origin is possibly in dilated sebaceous glands with retained secretions. Unna considers that calcium soaps are formed as a first step, but an analysis of such material bj' Harley^^ showed 87.2 per cent, of ash, 12.8 per cent, organic matter, 0.9 per cent, of fat; calcium phos- phate constituted 65.2 per cent., and calcium carbonate 16.4 per cent. Gascard^^ found in similar material 23.4 per cent, organic matter, and of the inorganic matter, 91.1 per cent, was calcium phosphate, and 8.9 per cent, calcium carbonate. Gouty deposits observed in the subcutaneous tissues, as well as along the tendons, articular cartilages, etc., consist usually of nearly pure biurate of sodium and potassium. Ebstein and Sprague^^ found the composition of such material to be as follows : Uric acid 59 . 70 Tissue organic matter 27 . 88 Sodium oxide 9 . 30 Potassium oxide 2 . 95 Calcium oxide 0.17 MgO, Fe, P2O6, S traces "Literature. Poulalion, Thesis, Paris, 1891; Stern, Deut. mcd. Woch., 1904, (30), 1414; Burgi, Deut. med. Woch., 1906 (32), 798; Gerhartz and Strigel, Beitr. z. klin. Tubarc, 1908 (10), 33. *'' "Chem. Path, der Tuberc," 1903, p. 92. ■»8 Literature, Scheppegroll, Jour. Amer. Mcd. Assoc, 1896 (20), 874; Gcrber, Deut. med Woch., 1892 (18), 1165. *^ Jour. Pluirm. et Chiin., 1891 (23), 447. ^"McCarthy, Brit.Med. Jour., Oct. 28, 1911. ''' Jour. Pliarm. et Chim., 1903 (18), 9. »2 Ibid., 1900 (12), 262. " Virchow's Arch., 1891 (125), 207. PULMONARY INCRUSTATIONS 400 After a time, however, calcium salts may be deposited, and Dunin" has observed deposits resembling gouty tophi that were merely calcium salts. Pneumonokoniosis In a number of cases of the different forms of this con(htion quan- titative anal3'-ses have been made, which may be briefly discussed as follows: Not only docs the lung of every adult contain considerable amounts of coal-pigment stored up in the connective tissues (and also in the peribronchial glands), but also, which is perhaps less generally appreciated, considerable quantities of sihcates are also present (chal- icosis) from inhaled dust. Woskressensky^^ found silicates in all of 54 lungs examined, except two from infants. The lungs of individ- uals whose occupations do not expose them especially to dust inhala- tion contain increasing amounts of silicates in direct proportion to age; the silicates constitute then from 3.5 to 10 per cent, of the total ash of the lungs. There is always a larger proportion of silicates in the peribronchial glands than in the lungs, constituting from 6 to 36 per cent, of the ash, corresponding with Arnold's observation that in gold-beaters the glands contain more metal than the lungs. In stone-workers Schmidt found a higher proportion of Si02 in the lungs than in the glands. In normal adults the amount of coal-pigment is greater than the amount of silicates; in children the reverse is the case. ThoreP® reports that the lungs of a worker in soapstone contained 3.25 per cent, of ash, including 2.43 per cent, of soapstone. In siderosis iron has been found in the lungs in proportions varying from 0.5 per cent, to 7.9 per cent, of the dry weight, the last amount having been found by Langguth" in the lungs of an iron miner, which contained also 11.92 per cent, of Si02. An analysis of a lung from a knife-grinder is reported b}'' Hoden- pyl^ss which gave the following results: Total weight of dried and powdered lung, 48.1009 grams; total solids, 44.7986; ether-soluble substance, 14.6017. Composition of the ether-soluble substance: free fatty acids, 7.498; neutral fats, 4.044; cholesterol, 3.037. Proteins, 15.4759; charcoal (total carbon less protein carbon), 7.198; ash, 4.2903. The composition of the ash (in grams) was as follows: K2O, 0.2167; NaaO, 0.3523; CaO, 0.0965; Fe20„ 0.0879; AI2O3, 1.4628; SO3, 0.0704; P2O5, 0.9565; Si02, 1.2043. The amount of emery, rep- resented by the oxides of aluminum and silicon made up more than one-half of the ash, and the iron constituted about one-fourth. The man had worked at the trade of knife-grinder for about fifteen years. " Mitt. Grenzgeb. Med. u. Chir., 1905 (14), 451; also Kahn, Arch. Int. Med.^ 1913 (11), 92, and M. B. Schmidt, Deut. med. Woch., 1913 (39), 59. " C^.ot. f. Path.. 1898 (9), 296. 58 Ziegler's Beitr., 1896 (20), 85. " Deut. Arch. klin. Med., 1895 (55), 255. '« Medical Record, 1899 (56), 942. 470 CALCIFICATION, CONCRETIONS, AND INCRUSTATIONS McCrae^^ has analyzed the lungs of six gold mine workers, in South Africa, finding from 9 to 21.7 grams of ash per lung, of which 29 to 48 per cent, was silica; aluminum was also high, and an increased PoOs content was ascribed to the accompanying fibrosis. Klotz^° found from 1.2 to 5.3 grams of free carbon in each lung, of dwellers of Pittsburg, as contrasted with 0.145 and 0.405 grams found in the lungs of residents of Ann Arbor. Hirsch^^ analyzed four average Chicago lungs, finding in grams per lung: I II III IV Carbon 2.72 0.71 1.20 0.19 Silica 0.18 0.28 0.69 0.04 Calcium Oxide 0.45 0.12 0.02 0.05 ^5 "The Ash of Silicotic Lungs," John McCrae, Johannesburg, 1914. ^0 Aineh Jour. Publ. Health, 1914 (4), 887. General review on anthracosis. " Jour. Amer. Med. Assoc, 1916 (66), 950 CHAPTER XVIII PATHOLOGICAL PIGMENTATION' MELANIN^ Melanin occurs normally as the coloring-matter of hair, of the choroid of the eye, of the skin, in the pigment matter of many lower animals, and most strikingly as a defensive substance in the "ink'" ejected by squids to render themselves invisible in the water. Path- ologically melanin occurs chiefly as the result of an excessive pro- duction of this pigment by cells normally forming it, as in freckles, melanotic tumors, and Addison's disease (probably). Cells that do not normally form melanin probably do not acquire this power in pathological conditions. Pathological failure to form melanin is also observed, as in skin formed in the healing of wounds and after syphili- tic lesions; or in albinis77i, in w'hich the failure to form melanin may be attributed to hereditary influences.^ Occasionally in domestic animals, especially in calves, a congenital melanosis is observed in- volving many parts of the body."* A melanin or some similar pig- ment may be found in nerve cells (e. g., substantia nigra), and DoUey* beheves it to be a result of nuclear metabolism under conditions of depression. The function of melanin is evidently that of protection from light rays, and Young*' has found that isolated melanin from hu- man skin absorbs violet and ultra-violet rays. Probably this protec- tion is responsible, at least in part, for the relative infrequency of skin cancers in the colored races.'' Melanin seems always to be produced through metabolic acti\'ity of specialized cells. The idea, which was formerly advanced, that it is derived from hemoglobin as a product of disintegration, seems to have failed entirely of substantiation. In malaria we frequently find a diffuse pigmentation of the skin of such a nature as to suggest ' Literature by Oberndorfer, Ergebnisse Pathol., 1908 (12), 460, and Hueck, Ziegler's Beitr., 1912 (54), 68. 2 Literature and resume given by v. Ftirth, Cent. f. Pathol., 1904 (15), 617; Handb. d. Biochem., 1, 742. ' Gortner holds that dominant whites are due to the presence of antioxidase, while regressive whites have neither the power to form pigments nor to inhibit their formation (Amer. Naturalist, 1910 (44), 497). * See Caspar, Ergebnisse allg. Path., 1896 (III2), 772. ' Science, 1919 (50), 190. « Biochem. Jour., 1914 (8), 460. ' However, Hanawa found white areas in skin less affected by chemical irri- tants and infections than dark areas. (Dermatol. Zeit., 1913 (20), 761.) This is not in agreement with most observers who have found pigmented skin more re- sistant. (See Hanzlik and Tarr, Jour. Pharm.. 1919 (14), 221.) 471 i 472 PATHOLOGICAL PIGMENTATION strongly a melanin formation, and this has been cited as an example of the production of melanin from hemoglobin. Carbone has proved, however, that this malarial pigment is derived from hematin. The amount of iron contained in melanin has been much investigated, as bearing upon the question as to whether the melanin is derived from hemoglobin or not, and the results obtained by the best methods indi- cate that the amount of iron present is usually extremely small, and often it is entirely absent; furthermore, the presence of iron is no proof that the pigment is derived from hemoglobin, since other iron- protein compounds undoubtedly exist,^ — especially nucleoproteins, and chemical examination shows that melanin does not contain hemo- pyrrole groups.^ Composition of Melanin. — The elementary composition of different specimens of melanin examined by various observers has been found to vary greatly. This probably depends on three factors : First, it is extremely difficult to obtain melanin in a pure condition; second, the process of purification requires the action of strong acids and alkalies, which undoubtedly modify the composition of the melanin; thirdlj^, melanin is probably not a single substance of definite composition, but includes several related biit different bodies. The values found varj' for carbon from 48.95 to 60.02 per cent. ; for hydrogen from 3.05 to 7.57 per cent. ; for nitrogen, 8.1 to 13.77 per cent. Hofmeister gives, as a characteristic of melanins, that their elementary molecular composition is always nearly in the proportions N : H : C = 1 : 5 : 5. Gortner's^ studies have led him to accept the general principle that melanin is formed through the action of an oxidase on an o.xidizable chromogen, but that in keratinous structures there exist at least two types of melanins, one, a "nielano- protein," soluble in dilute acids and existing dissolved in the keratins; the other, insoluble in dilute acids, exists as pigment granules and is of unknown nature. Piettre'"* believes that melanin from sarcoma of the horse consists of a protein united to a pigment. Those whose studies of melanin formation have been made with the microscope, state that the nucleus is active in the process, ^^ and some find the melanin so closely related to the lipoids that they consider it a lipochrome.^^ A particularly prominent constituent of some melanins is sulphur,which has been found in as high proportions as 10 per cent, in melanin from sarcomas, and even 12 per cent, in sepia from the squid; in melanin from hair the sulphur is usually about 2-4 per cent.; but in choroid melanin, and in some other forms, sulphur seems to be absent. The proportions of sulphur obtained from the same specimen purified by different methods show wide variations, and hence v. Fiirth considers that neither the sulphur nor the iron are indispensable constituents of the melanin. Probably the melanin molecule contains atom-complexes that have a tendency to bind certain sulphur and iron compounds (e. g., cystine or hematin derivatives). There is much reason to believe that the melanin is derived from certain groups of the protein molecule that seem readily to form colored comjiounds. The aro- matic compounds of the protein molecule, such as tyrosine, phenylalanine, and tryptophane, readily condense with elimination of water and absorption of oxygen, to produce dark-colored substances. When proteins are heated in strong hydro- chloric acid, we obtain a dark-brown material, which closely resembles the melanins both in elementary composition and in general properties, so that it is referred to 8 Spiegler, Hofmeister's Beitr., 1907 (10), 253. »Biochem. Bulletin, 1911 (1), 207; r6sum6. *" Compt. Rend. Acad. Sci., 1911 (153), 782; also see Reprint from 1st Internat. Cong. Compar. Pathol., Paris, 1912. 1' Htaffel, Verh. Deut. Path. Ges., 1907 (11), 136; Schultz, Jour. Med. Res., 1912 (26), 65. 12 Dyson, Jour. Path, and Bact., 1911 (15), 298; Kreibich, Wien. klin. Woch., 1911 (24), 117. MELANIN 473 as "artificial melanin" or "melanoid substance." These substances, like the natural inelanins, when decomposed by fusing with caustic potash, yiehl skatolc, indole, and pyrrole derivatives, whicli are undoubtetily derived from the tyrosine and tryptophane of the protein molecule. Therefore, it seems probable that both the melanoid substances and the true melanins are formed from the chromojfen groups of the protein molecule through processes of condensation, elimination of water, and the taking up of oxygen.'* In the sepia sacs of the cuttle-fish, in meal-worms which form a melanin-like pigment, and in plants that produce the black Japanese lacquer, have been found oxidizinq enzymes that have the property of producing black pigment by their action upon tyrosine and other aromatic compounds. Neuberg" found that ex- tracts of a melanosarcoma of the adrenal could produce pigment from epinephrine and /3-oxyphenylethylamine, but not from tyrosine. The ink sacs of the squid contain an enzyme forming a pigment from epinephrine, apparently through oxi- dation and condensation. These enzymes may, therefore, possibly be responsible for the production of melanin in animal tissues, by causing oxidative changes in the chromogen groups of the protein molecule that are liberated by autolysis (see "Tyrosinase"), v. Ftirth urges strongly the view that both normal and patho- logical melanin formation depend upon the action of the tyrosinase or allied en- zymes in conjunction with autolytic enzymes; the latter split free the chromogen groups of the protein molecule, which are then oxidized by the tyrosinase, undergo condensation, and take up sulphur- and iron-holding groups and also other organic compounds, the entire complex forming the melanin. Bruno Bloch'^ has found that the occurrence of melanin in the skin corresponds to the location of cells with the capacity of oxidizing 3.4-dioxyphenylalanine, which is closely related in structure to epinephrine, and which he believes may be the usual antecedent of melanin. He has found this oxidizing property exhibited by the dark patches in variegated animals, but not by the white areas; the pig- mented ocular structures do not oxidize this substance. Properties of Melanin. — When isolated in a pure condition, melanin is a dark-brown substance of amorphous structure, no mat- ter how black the material from which it is derived may be.**' It is quite insoluble in all ordinary reagents except alkalies, in which some melanins dissolve easily, and some with difficult3\ Strong boiling hydrochloric acid scarcely affects non-protein melanins. By the action of sunhght or oxichzing agents on melanin-containing sections the pigment can be bleached out. The chief decomposition-products formed on fusing with alkalies are indole, skatole, and "melanic acid"; no cystine, leucine, tyrosine, or other amino-acids can be iso- lated. Most authors, therefore, consider the melanins as heterocyclic compounds standing in some relation to the indole nucleus. If melanin is injected subcutaneously into animals u'abbits and guinea-pigs), there appears in the urine a substance which turns dark brown after the urine has stood for some time (Kobert, Helman). The pigment is apparently reduced, particularly by the liver, to a colorless melanogen, which is eliminated in the urine. The same process occurs when melanin is produced in excess and enters the '3 See Herzmark and von Furth, Biochem. Zeit., 1913 (49), 130. 1* Zeit. f. Krebsforsch., 1909 (8), 195. 15 Bloch and Ryhiner, Zeit. exp. Med., 1917 (5), 179; Zeit. physiol. Chem., 1917 (100), 226. '^ Spiegler (Hofmeister's Beitr., 1903 (4), 40) claims to have isolated from white wool a white chromogen, closely related to melanin chemically, but Gortner (Amer. Naturalist, 1910 (44), 497) believes this to be a decomposition product of keratin, unrelated to melanin. 474 PATHOLOGICAL PIGMENTATION blood, as in the case of melanosarcoma, a colorless melanogen being formed which is excreted in the urine, constituting "melanuria." Occasionally the urine is dark when first passed, because of the pres- ence of melanin, but usually it must be subjected to oxidizing agen- cies (^bromine water, nitric acid, hypochlorites, etc.), or exposed to air to bring out the brown color. Helman^^ says that true melano- gen may be considered to be present in urine: \1) If the careful ad- dition of ferric chloride causes the development of a black precipi- tate. (2) If this precipitate dissolves in sodium carbonate, forming a black solution. (3) If from this solution mineral acids precipitate a black or brownish-black powder. All three reactions must be obtained, for substances other than melanin may give the first two. Especially to be distinguished are alkaptonuria, chronic intoxication with phenols, and some cases of extreme indicanuria.^* In support of the view that tryptophane is the mother substance of melanin is the fact that feeding tryptophane to melanurics increases the melanin excretion (Eppinger). The coloring power of melanin is very great, for urine containing but 0.1 per cent, of melanin has the color of dark beer (Hensen and Nolke), and the entire skin of a negro contains only about 1 gram of melanin (Abel and Davis). '^ Excessive quantities of melanin may be in part deposited in the lymph-glands and skin, causing diffuse pigmentation; it may be deposited in the endothelium lining the blood-vessels. Koberfc injected melanin into albino rabbits, but did not succeed in getting any deposition in the choroid or skin. Helman found some evidence of toxicity when large doses of melanin dissolved in sodium carbonate are injected into animals, but this is possibly due to the alkali rather than to the melanin. Melanotic Tumors.^" — Tumor melanin does not differ from mel- anin produced by normal cells in any essential respect. Usually it con- tains much sulphur, even as much as 10 per cent., yet Helman in eight specimens found but four that contained both sulphur and iron, in three only sulphur, in one only iron and no sulphur; therefore, tumor melanins show the same variations in composition as do normal mel- anins. Iron is frequently found microscopicallj'- in the pigment in melanosarcoma, but this is chiefly due to admixture of blood-pigment coming from extravasations of blood. The peculiar fact that melano- sarcoma is very common in white or gray horses, but very seldom " Cent. f. inn. Med., 1902 (23), 1017; Arch, internat. Pharmakodynam., 1903 (12), 271. ^8 Melanuria fully discussed by Feigl and Querner, Deut. Arch. klin. Med., 1917(123), 107. " Jour. Exp. Med.. 189G (1), 3G1. 2" Under the title Acanthosis Nigi'icans (see PoUitzer, Jour. Amer. Med. Assoc, 1909 (53), 1369) is included a group of cases of widespread cutaneous pigmentation with papillary hypertrophy, commonly associated with cancer, most often ab- dominal. While ascribed to action of the sympathetic nervous system injured by the cancer, this explanation is far from satisfactory, and the possibility of metabolic pigmentary disturbance must be considered. MELANIN 475 occurs in dark-coated horses, has not been explained. The frequent occurrence of mehmuria and mehincniia in patients witli iiiolanosar- coma is not duo to any peculiar property of sarcoma nichmin, but to the enormous quantity of melanin that is prochKuul by the tumor and set free in the degenerating; portions. Thus, while Abel and Davis'^ estimate that there is only about 1 gram of melanin in the entire skin of a negro, Nencki and Bordez have obtained from a sarc-omatous liver 300 grams of melanin, and estimate that the entire body con- tained 500 grams. Helman'^ states that the melanin may con- stitute 7.3 per cent, by weight of the fresh substance of some melanosarcomas. According to Lubarsch and to Helman, melanotic tumors rarely contain glycogen. As mentioned above, Neubcrg found that a melanotic sarcoma of the adrenal produced pigment from epinephrin and from /3-oxy- phcnylethylamine, but he failed to get positive results with melano- sarcomas of the eye and from the horse, but Alsberg-' succeeded in finding in melanosarcoma from the liver an enzyme oxidizing pyro- catechin and Jager- found that horse melanosarcoma extracts will oxidize epinephrin to a pigment. The "dopa reaction" of Bloch,^^ which depends on the presence of specific oxidizing enzymes in the cells, may be exhibited by the connective tissues quite generally throughout the body in some cases of melanosarcoma."" Eppinger-^ found that the urine of a patient with melanosarcoma gave intense reactions for indole and tryptophane, and that when tryptophane was fed to a patient there was a great increase in the melanuria. He therefore concludes that the power of the body to destroy the pyrrole ring is reduced, and instead it undergoes reduc- tion, methylation and union with sulphuric acid, to form an ethereal sulphate of methylpyrrolidine-hydroxy-carbonic acid (CH3-C5H9N2O4). Abderhalden-'' also found a relation to tryptophane, for in the urine of a melanuric was present a substance rich in tryptophane; and Primavera" found the urine in a case of melanosarcoma containing free tyrosine, fluctuating in amount with the pigment. Addison's disease is associated with the deposition of a pigment in the skin that is generally considered to be a melanin, differing from that produced normally in the skin only in quantity and not in origin or composition. ^"^ No satisfactory explanation of the relation of the adrenal to this pigmentation seems yet to have been made, al- though it is natural to assume that when the function of the adrenal is destroyed, substances accumulate in the blood that have a stimu- li Jour. Med. Res., 1907 (16), 117. 22 Virchow's Arch., 1909 (198), 62. •"" Matsunaga, Frankf. Zeit. Path., 1919 (22), 69. " Biochem. Zeit., 1910 (28 j, 181. 2^ Zeit. physiol. Chem., 1912 (78), 159. " Giorn. Int. Scienze Med., 1908 (29), 978. -« Concerning histogenesis of the pigment see Pforringer, Cent. f. Path., 1900 (11), 1. 476 PATHOLOGICAL PIGMENTATION lating effect on the pigment-forming cells. Abnormal protein catab- olism, with excessive accumulation of the chromogenic constituents of the protein molecule, has been suggested, as also have alterations in the influence of the sympathetic nervous system upon the chromo- phore cells, for nerve lesions (e. g., neurofibroma) often are accom- panied by pathological pigmentation of the skin.^^ It is significant that the active constituent of the adrenal medulla, the epinephrin, is an aromatic derivative closely related to tyrosine, since the production of pigment by the action of oxidizing enzymes upon such substances is well known. Furthermore, Neuberg has described a melanotic adrenal tumor which produced pigment by oxidizing epinephrine. On this basis the pigmentation of Addison's disease would seem to be the result of an abnormal accumulation or distribution of aromatic compounds, because of their failure to be converted into epinephrine. In support of this hypothesis is the observation of Meirowsky that the human skin contains an enzyme capable of oxidizing epinephrine to a pigment, and that pieces of skin kept warm will develop a postmortem pigmentation, and this is supported by Konigstein-^ who found that the pigmentation was greater in animals deprived of their adrenals or given injections of epinephrine. Bloch^^ believes that the pigmentation results from the precursor of epinephrine, 3.4-dioxyphenylalanine, which is oxidized in the epidermal cells to melanin. As exact chemical studies of the pigment in Addison's disease have not been made, however, we have no positive proof that it is a mel- anin, hence any speculation as to the cause of its formation is prema- ture. Carbone-^ claims to have isolated from the urine in Addison's disease a pigment that contains much sulphur, and which he considers similar to or identical with the melanogen of melanuria. A similar observation is reported by Eiselt.^'' v. Kahlden,^^ however, has ob- served crystals resembling hematoidin in the pigmented tissues. Ochronosis'^ is a condition characterized bj^ a black pigmentation of the cartilages, first described by Virchow in 1866. In 1904 Osler'^ reported two cases, and found but seven others in the literature to that time. Virchow suspected that the condition was due to a permeation of cartilage by hematin derivatives, but Hansemann, finding a case associated with melanuria, considered that the pigment is probably of metabohc origin. Hecker and Wolf studied the urine of a similar case, and concluded that the pigment must be melanin. Albrecht,^* " See r6s\im6 by Schmidt, Ergeh. der Pathol., 1896 (Bd. 3, Abt. 1), 551. " Wien. klin. Woch., 1910 (23), 616. 29 Giorno R. Acad. med. di Torino, 1896. '" Zeit. klin. Med., 1910 (69), 393; full discussion on the pigment of Addison's dlSGA-SG »iVirchow's Arch., 1888 (114), 65. 32 See Adler, Zeit. f. Krebsforsch., 1911 (11), 1; Poulsen, Ziegler's Beitr., 1910 (48), 346. "Lancet, 1904 (i), 10 (literature). »< Zeit. f. Heilk., Path. Abt., 1902 (23), 366. I OCHRONOSIS 471 however, suggested a relation of ochronosis to alkaptonuria, having found honiogentisic acid in the urine of a case reported by him (see "Alkaptonuria")- Osier's two patients were brothers with alkap- tonuria, the evidence of ochronosis consisting of discoloration of the cartilages of the ears. Langstein'"' has examined a specimen of urine preserved from Hansemann's case, and found no evidence of alkaptonuria.^*^ Pick" summarizes the results of his study of his case and of the literature, as follows: Ochronosis is a definite form of melanotic pig- mentation, the pigment of ochronosis being in most of the cases very closely related to melanin. The pigment, or its chromogcn, circulating freely in the blood, is imbibed not only by cartilage, but also by loose connective tissue, voluntary and involuntary muscle-cells, and epi- thelial cells, without any decrease in vitality of these cells being observable; however, degenerated tissues show the greatest amount of pigmentation. The diffuse pigment can become granular after a time; it is iron-free, but under certain circumstances may contain fat. This melanin arises from the aromatic nucleus of the protein molecule (tyrosine, phenylalanine), and the related hydroxylized products, under the influence of tyrosinase. In some cases the constant absorption of mi- nute quantities of phenol from surgical dressings seems to have been the cause of the condition. Besides this formation of pigment from such "exogenous" aromatic substances, however, it is probable that in alkaptonuria the "endogenous" aromatic substance (homogentisic acid) present may be converted into pigment by the tyrosinase. In many of the cases of ochronosis the pigment or a precursor may be excreted in the urine, which then undergoes spontaneous darkening when exposed to the air. The kidneys may also become pigmented and granular masses of pigment may be present in the renal tubules. Poulsen^*^ states that of the 32 known cases of ochronosis (in 1911) in 17 there was alkaptonuria, in 8 carbohc acid dressings had been used for long periods, and in the remaining 7 cases the cause was not determined. These facts are conclusive evidence of the origin of ochronotic pigment from aromatic radicals, whether these radicals are converted into true melanin or not. The localization of the pig- ment is explained by the demonstration by Gross and Allard,^^ that cartilage has a greater affinity than other tissues for homogentisic acid. Ochronosis can be produced experimentally with homogentisic acid, and often is associated with an arthritis. •*" There are, however, numerous cases of alkaptonuria without ochronosis. The ochronosis described in lower animals is not the same as human ochronosis, affect- 35 Hofmeister's Beitr., 1903, (4)145. 3« Also see Langstein, Berl. klin. Woch., 1906 (43), 597. " Berl. klin. Wochenschr., 1906 (43), 478. 38 Mlinch. med. Woch., 1912 (59), 364. 33 Arch. exp. Path. u. Pharm., 1908 (59), 384. " Gross, Deut. Arch. klin. Med., 1919 (128;, 249. 478 PATHOLOGICAL PIGMENTATION ing the bones rather than the cartilages (Poulsen),-'^ and being more properly designated by the name osteohemachromatosis (^Schmey).''^ Malarial pigmentation, according to Ewing/' may have any one of the follow- ing origins: (1) Pigment elaborated by the intracellular parasite. (2) Hematoidin de- rived from the remnants of infected red cells. (3) Hematoidin or altered hemo- globin deposited in granular or crj'stalline form from red cells dissolved in the plasma. (4) Bilirubin or urobilin granules or crystals. Of these, the pigment formed by the parasites has been considered by many as a true melanin, but this cannot be considered as established, especially as Ewing finds it to have the same relation to solvents as do the blood-pigments. Carbone and Brown^^ consider the malarial pigment to originate from hematin, with which it agrees in solubility, spectroscopic properties, and in containing iron. Pigmentation of the Colon.'*-^ — Sometimes the mucosa of the entire colon is found deeply pigmented, with a material of unknown character, but resembling in many respects a melanin. The cause of the condition is unknown. Abder- halden" has found pigments that seemed to be derived from tryptophane, while Niklas^^ attributes the coloration to tyrosinase activity. Pigmentation of the oral mucosa, with a pigment resembling melanin, has been described especiallj^ in pernicious anemia. It does not seem to be related to the adrenal.** LiPOCHROMES In normal plant and animal tissues occur pigments that are either fats or compounds of fat, or substances highl}- soluble in fats. In animals they occur normally in the corpus luteum, in the epithelium of the seminal vesicles, testicles, and epididymis; in ganglion-cells, especially in the sympathetic nervous tissue; in the Kupffer cells of the hver and in fat tissue. Pathologically, such pigments are found particularly in the muscle-cells in brown atrophy of the heart, and less abundantly in the epithelium of atrophied livers and kidneys (Lubarsch^^ and Sehrt^")- ^^1 are characterized b}^ staining b}' such fat stains as sudan III and scarlet R, and usually, but not constantly, by osmic acid; they are dissolved by the usual fat solvents. It is questionable if all pigments that stain for fat should be considered as true lipochromes, however, for their other reactions are variable; and Borst would distinguish these pathological pigments from the true lipochromes by calling them lipofuscins, including under this term the brown "waste pigments, " which Hueck believes to be formed from disintegrated lipoids or fatty acids. Many pigmentary substances *i See Ingier, Ziegler's Beitr., 1911 (51), 199. " Frankfurter Zeit. Pathol., 1913 (12), 218; also Teutschlaender, Virchow's Arch., 1914 (217), 393. ••3 Jour. Exp. Med., 1902 (6), 119. *Uour. Exper. Med., 1911 (13), 290. " Full review bv McFarland, Jour. Amcr. Med. Assoc, 1917 (69), 1946. <» Zcit. phy-iol. Chem., 1913 (85), 92. "' Mimch. med. Woch., 1914 (61), 1332. See also Hattori, Mitt. mod. Ge- scUsch., Tokio, 191() (30), No. 6. " See Weber, (>uart. Jour. Med., 1919 (12), 404. " Cent. f. Pathol., 1902 (13), 881. *» Virchow's Arch., 1904 (177), 248. See also Mayer ct al, Jour, physiol. et path. g6n., 1914 (16), 581. LIPOCIIROMES 479 are probably soluble in fats, and in this way t lie lipofusfins are formed.*' In the renal epithelium is found a pigment resembling the lipofuscins, increasing with age and not related to the urinary pigments." Typical plant lipochromes, as also the pigments of Staphylococcus pyogenes aureus and citreus, are colored blue by concentrated sulphuric acid with formation of small blue crystals of lipocyanin. With iodin- potassium-iodide solution they are colored green. Lipochrome of frog-fat stains blue with this solution (Neumann);^' lipochrome of the corpus luteum (called lutein) occasionally gives a faint blue with sulphuric acid or Lugol's solution (Sehrt); but the fat-holding pig- ments of the other tissues mentioned above do not give either of these reactions. Possibly these last are not true lipochromes, therefore, but rather pigments chemically or physically combined with fat. Cotte''^ believes that the true lipochromes of plants and animals have a cholesterol base, but the presence of glycerol in plant and bacterial lipochromes can be demonstrated by the acrolein test — possibly, therefore, both cholesterol and neutral fats are present. Melanins and pigments derived from hemoglobin do not stain with sudan III and are not soluble in ether, etc., and hence can be readily distinguished from the fatty pigments. It has been shown by Escher*^ that the pigment of the corpus luteum is identical with the carotin of carrots. Apparently carotin and xan- thophyll (a crystalline pigment from green plants) ^"^ are the chief pigments of milk fats, egg yolk, and probably of body fats.^^ In the body lipins these pigments accumulate throughout life because of their great solubility in lipins, which explains the high color of the fats of old persons. Carotin seems to be almost or quite devoid of toxicity, ^^ and in persons eating carrots in large C|uantities there may be enough pigment present in the blood {carotinemia) to produce skin pigmenta- tion resembling jaundice.^" The work of Palmer indicates that carotin and xanthophyll are much more widely distributed than was formerly appreciated. Ani- mals with colored fats owe the color to these plant pigments, which are also present in the blood of these same animals, but not in the blood of animals with colorless fats (swine, rabbits, dogs, sheep, goats), and the so-called lipofuscin of the ganglion cells has been shown to be ^^ Ciaccio (Biochem. Zeit., 1915 (69), 313) agrees with Hueck, and finds it possible to distinguish between pigments from phosphatids, which stain poorly with Sudan III, and those from free fatty acids which stain deeply with this dye. " Schrever, Frankf. Zeit. Pathol., 1914 (15), 333. " Virchow's Arch., 1902 (170), 363. " Compt. Rend. Soc. Biol., 1903 (55), 812. " Zeit. physiol. Chem., 1913 (83), 198. " Concerning plant pigments see review by West and Horowitz,Biochem. Bullet., 1915 (4), 151 and 161. ^^ See articles by Palmer and Eckles, Jour. Biol. Chem., 1914, Vol. 17 et seg. 58 Wells and Hedenburg, Jour. Biol. Chem., 1916 (27), 213. " Hess and Myers, Jour. Amer. Med. Assoc, 1919 (73), 1743; see also ibid., 1920 (74), 32. 480 PATHOLOGICAL PIGMENTATION carotin. '^^ Palmer found that carotin is the pigment of milk fat, body fat and corpus luteum of the cow, while xanthophyll with some carotin colors the egg yo\k, body fat and blood serum of the fowl. Chickens deprived of these pigments from the time of hatching have no pigment in their fats or egg yolks although the fowls are healthy and their colorless eggs are fertile. ^^ This work makes doubtful the existence of other fat-soluble intracellular pigments in man, such as lipofuscin, and Dolley states that even the typical lipofuscin of brown atrophy of the heart is sometimes insoluble in all reagents that dissolve fats. Xanthosis diabetica^^ also seems to depend on an excess of lipochromes in the blood, probably partly endogenous from mobilization of tissue fats and chiefly exogenous from the abundance of green vegetables in the diet. Accompanying hypercholesterolemia is usually present. CMoroma." — The pigment that causes the peculiar green color characteristic of these malignant growths, was considered by Chiari, Huber and others as a fatty substance related to or identical with the lipochromes. It commonly fades on exposure to air, and also when in the usual preservative fluids, to which it does not impart its color. The color may be brought back after formaldehyde preservation by H2O2 or by weak alkalies (Burgess)." Ottenberg" has suggested that the green color may be due to eosinophiles which abound in chloromas, since in fresh preparations eosinophile granules have a faint greenish tinge. It contains no iron, is soluble in absolute alcohol and in ether, and is usually, but not always (v. Recklinghausen), stained black with osmic acid.*^ Treadgold states that as the green color is not present from the beginning it would seem that cellular de- generation must play a part. Possibly a degeneration of the granules of the myelocytes and myeloblasts, aided by the products of hemoglobin disintegration, is responsible.*^^ Chromophile cells may be considered in this connection. Kohn^' has described certaincellswithadecidedaffinity for chromic ocid and its salts, found abundantly in the sympathetic nervous system, in the carotid gland, and in the medulla of the adrenal. They are also present in tumors derived from these organs. Extracts from such organs have a marked effect in raising blood pressure, and, according to Wiesel,'^* they are greatly involved in Addison's disease. The nature of the chromophile substance is unknown, but it can be fixed only by chromic acid or chromates; cells hardened by other means show merelj' spaces in the places occupied by this substance. It is generally believed to be the same as the epi- nephrine, but it does not always seem to parallel in amount the quantity of epine- phrine as determined chemically. Ogata''" states that the chrome reaction depends on the reduction of chromic acid to chromium dioxide by epinephrine. «» Dolley and Guthrie, Jour. Med. Res., 1919 (40), 295. Marinesco, however, says that the pigment of nerve cells resembles that produced during autolysis in ganglia (C. R. Soc. Biol., 1913 (72), 838). 6^ Jour. Biol. Chem., 1919 (39), 299. 62 Burger and Reinhart, Ziet. exp. Med., 1918 (7), 119. "Literature by Dock, Amer. Jour. Med. Sci., 1893 (106), 152; and Dock and Warthin, Med. News, 1904 (85), 971; Burgess, Jour. Med. Res., 1912 (27), 133. «* Amer. Jour. Med. Sci., 1909 (138), 505. *' The pigment of xanthelasma multiplex seems to be a fatty substance (Pocns- gen). Virchow's Arch., 1883 (91), 354. «6 Quart. .Jour. Med., 1908 (1), 239; Weber, Proc. Roy. Soc. Med., Clin. Med. Sec, 191G (9), 7. " Prag. med. Woch., 1902 (27), 325. «« Zeit. f. Heilk., Path. Alit., 1903 (24), 257. «» Jour. Exp. Med., 1917 (25), 807. HEMOGLOBIN 481 Blood Pigments"" Red corpuscles behave much as do other non-nucleated fragments of cells, undergoing disintegration rapidly and constantly when under normal conditions, as well as when subjected to various harmful in- jfluences (see "Hemoh^sis"), or when outside of the vessels in extrava- sations of blood. The processes and products of their disintegration arc, therefore, much the same whether occurring under normal or pathological conditions. The hemoglobin molecule is large and com- plex, and from it are derived many substances of the nature of pig- ments; indeed, hemoglobin itself may appear free as a pigment. Hemoglobin is a compound protein, consisting of a protein group {globin) and a coloring-matter {hematin or hemochromogen) .''^ The protein globin is of a basic nature, and seems allied to the histons; the hematin is, therefore, presumably acid. Hemoglobin ordinarily does not crystallize readily, especially the hemoglobin of man, and it is doubtful if it ever does so in the living tissues, although possibly this may occur in the center of large hematomas. In bodies that have undergone postmortem decomposition, and occasionally in specimens kept for microscopic purposes, irregular orange-yellow crystalline masses of hemoglobin may be found. This occurs particularly if the blood has been acted upon by hemolytic agents or has undergone putrefactive changes, and then is hardened in alcohol. The crystals are either oxyhemoglobin, or more often an isomeric or polymeric modification, parahemoglohin (Nencki). Hemoglobin also enters cells unchanged, imparting a diffuse yellowish color, and apparently it is non-toxic. ^^ If present in the blood in large enough amounts it is excreted un- changed in the urine, but at least one-sixtieth of the total number of red corpuscles must be in solution at one time to produce hemoglo- binuria; in man at least 17 c.c. of laked corpuscles must be injected to accomplish this." Addis^'* has developed the following conception of the metabolism of hemoglobin. Free hemoglobin, liberated especially by the phago- cytes of the spleen, is taken up by the other phagocytes, notably the Kupffer cells of the liver, which pass it on to the liver cells. The pigment moiety, hematin is separated from the globin, and converted through removal of its iron into bilirubin. The bilirubin excreted into the intestine is there reduced to urobilinogen, which is in part reab- sorbed and polymerized into urobilin, which in turn is possibly poly- merized into a larger complex. In the liver this urobilin complex has restored to its pyrrol nuclei the original side chains, and then is used ^"Literature by Schmidt, Ergebnisse der Pathol., 1894 (!■>), 101; and 1896 (nil), 542; Schulz, Ergebnisse der Physiol., 1902 (Ii), 505. '^ Halliburton and Rosenheim recommend that the name "hemochromogen" be dropped in favor of "reduced hematin'.' (Biochem. Jour., 1919 (13), 195). 72 Barratt and Yorke, Brit. Med. Jour., Jan. 31, 1914. " Sellards andMinot, Jour. Med. Res., 1916 (34), 469. '*Arch. int. Med., 1915 (15), 412. 31 482 PATHOLOGICAL PIGMENTATION to form new hemoglobin molecules. This hj^pothesis is merely ten- tative, but it affords a useful "working hypothesis" for the considera- tion of many phases of pigment metabolism. In the decomposition of hemoglobin the first step is the splitting of the globin (which does not form pigments) from the hematin, from which many pigments may be derived. Hematin. — The formula given for this substance by Nencki is C32H32N4Fe04 while Hoppe-Seyler proposed the formula C34H34 N4Fe05, although it is not certain that the hematin of all animals is the same. It is found frequently as an amorphous, dark-brown or bluish-black substance, in large, old extravasations of blood, but sel- dom in small hemorrhages. As a pathological pigment hematin is by no means so frequently found as its derivatives. Schumm^^ ob- served a patient with chromium poisoning w^ho showed for several days abundant hematin free in the blood. He has also found it in malaria, pernicious anemia, congenital hematoporphj^ria, and gener- ally with acute toxic hemolysis, including patients infected with B. eniphysematosus, when the hematin may be accompanied by me+he- moglobin without a corresponding urinary excretion of these pigments. Feigl found hematinemia in many cases of poisoning with the war gases. "^ Brown" found that solutions of hematin cause chills and fever, and suggests that his pigment may be at least partially responsible for the symptoms of malaria. '^^ Hematin has been beheved to spht up gradually into an iron-free pigment {hematoidin) and an iron-contain- ing pigment {hemosiderin). This change may be represented by the following equation, according to Nencki and Sieber:^^ C32H32N404Fe + 2H2O = 2C,6H,8N203 + Fe. (hematin) (hematoidin) However, finding that the pigment in the malarial spleen is hematin, Brown^" suggests that hematin cannot well be an intermediary prod- uct in hemoglobin disintegration, since this malarial pigment persists a very long time in the tissues without change. He has made other observations that led him to conclude that hematin is not an inter- mediary substance between hemoglobin and hemosiderin, but that when once formed it is destroyed very slowly, by oxidation rather than hydrolysis. Injected into rabbits it produces vascular lesions in the kidneys**' and in large doses causes a marked fall in blood pressure. *- Hematoidin may be found in old, large extravasations, as orange- " Zeit. physiol. Chem., 1912 (80), 1; 1913 (87), 171; 1916 (97), 32. '« Biochem. Zeit., 1919 (93), 119. " Jour. Exper. Med., 1912 (15), 580; 1913 (18), 96. '8 Disputed by Butterfield and Benedict, Proc. Soc. Exp. Biol., 1914 (11),''80. "Arch. exp. Path. u. Pharm., 1888 (24), 440; Brugsch and Vo.slumoto, Zeit. exp. Path., 1911 (8), 639. "ojour. Exper. Med., 1911 (13), 290; 1911 (14), 612. «' Arch. Int. Med., 1913 (12), 315. " Brown and Loevenhart, Jour. Exp. Med., 1913 (IS), 107. BLOOD PIGMENTS 483 colored or red rhombic plates, first described by Virchow. Some- times, however, hematoidin occurs in the form of yellowish granular masses, and it may be associated with lipoids; it is also found irt crystalline form in icterus (Dunzelt).^' It seems to be nearly or quite identical with the bile-pigment, bilirubin, and it is probably the source of this substance under normal conditions. When formed in excessive amounts, either through increased destruction of corpuscles in the vessels or in extravasations, the amount of bile-pigment is in- creased (see "Icterus"). Possibly some of the hematoidin becomes transformed (Urectly into urobilin, and is then eliminated in the urine. Hemosiderin'*'' is relatively insoluble, and, therefore, is more slowly removed when formed in hemorrhages, and more abundantly deposited in the tissues when formed after excessive hemolysis. In acute hemolytic anemia a third of the total iron of the blood may be deposited in the liver, spleen and kidneys within 24 hours. ^'^ In infarcts hemosiderin soon disappears (Schmidt), ^"^ presumably because dissolved by the acids formed during autolysis. According to Neu- mann, hemosiderin is produced only under the influence of living cells and in the presence of oxygen, while hematoidin arises independent of cellular activity;^^ but Brown^^ has found that hemosiderin can be formed during autolysis of the liver, especially when air is present, and therefore probably by an oxidizing enzyme. He suggests that in hemosiderin the pigment is still hematoidin, and that the formation of^hemosiderin takes place in the nuclei, the hemosiderin being made directly from hemoglobin without the intervention of hematin. It may also be formed from the iron-containing protein of the cells during autolysis, independent of hemoglobin. ^^ Milner^" considers that, under similar conditions, an iron-containing pigment is also formed, which differs from hemosiderin in having the iron so combined that it cannot react with the usual reagents; this pigment may later change into hemosiderin. Up to the present time we do not know the chemi- cal nature of hemosiderin, nor its exact fate in the body, but it is probably utilized in the manufacture of new hemoglobin, for it is known that the iron liberated when hematin is broken up in the body under experimental conditions is deposited and not eliminated (Mor- ishima).^^ Unstained hemosiderin generally appears in the form of brown *^ Cent. f. Path. 1909 (20) 966. 8-« See Neumann! Virchow's Arch., 1888 (111), 25; 1900 (161), 422; 1904 (177), 401; also Arnold, ibid., 1900 (161), 284; Leupold, Beitr. path. Anat., 1914 (59), 501. s^Muir and Dunn, .Jour. Path, and Bact., 1915 (19), 417. 86 Verh. Deut. Path. Gesell., 1908 (12), 271. 8^ The accumulation of iron in the liver which follows poisoning with hemolytic agents, is not prevented or diminished bv preliminary removal of the spleen (Meinertz, Zeit. exp. Path. u. Ther., 1906 (2), 602). S8 Jour. Exper. Med., 1910 (12), 623. 83 Sprunt et al, Jour. Exp. Med., 1912 (16), 607. 90 Virchow's Arch., 1903 (174), 475. »i Arch. exp. Path. u. Pharm., 1898 (41), 291. 484 PATHOLOGICAL PIGMENTATION^ or yellowish-brown granules, and not as crystals. After a time it is taken up and deposited to a large extent in the liver, spleen, bone- jnarrow, and kidney, either as hemosiderin or possibly as some other iron compound of similar nature. From these sites it seems to be later taken up to be utilized in the manufacture of new red corpuscles. Whenever there is hemosiderin deposition in the kidney, granules of, the pigment may be found in the urine, free or in cells (Rous).^^ All told the average human body contains about 3.2 grams of iron, of which 2.4 to 2.7 grams is in the blood. According to ]\Ieyer^^ iron is present in the body in three forms: 1. Not demonstrable by reagents because so firmly bound (hemoglobin). 2. Loosely bound iron, colored by (NHJaS acting for a long time (ferratin). 3. Salt- like compounds with proteins, and inorganic compounds, reacting at once with reagents. Ferratin is the iron compound in the liver, con- taining 6 per cent. iron. He believes that probably hemosiderin is not a definite substance, but merely indicates compounds of the third class. Iron pigments may be transformed from one class to another, e. g., in corpus luteum scars, whose age can be estimated, class three may be replaced by class two. We may have in the sputum and lungs " Herzf ehlerzellen " that either do or do not stain with ferrocyanide. In morbus maculosus, Kunkel found the pigment of the internal organs to be pure iron oxide. Hueck also holds that hemosiderin is an in- organic iron compound, loosely bound to proteins and fats, and that it never forms an iron-free pigment, as has been stated. He believes that there is very little iron in the tissues in a firm union like hemo- globin, and that by proper technic some iron can be stained in every organ which contains iron chemically demonstrable. Ischida''* be- lieves that an iron-containing pigment may be formed in striated muscles from the iron normally there, without requiring a hematoge- nous origin. Hematoporphyrin.^^ — -There are several closely related pigments de- rived from hematin that are appropriately grouped under the desig- nation of porphyrins, for they are not all identical with the pigments prepared artificially from hematin by Nenclci and called by him hema- topoi'phyrin and mesoporphyrin , the former apparently representing a reduction, the latter an oxidation product. "'' The porphj-rins found in the urine and feces are different from each other and from those prepared by Nencki." Physiologically, these pigments are of great interest, because of the close chemical relation they have been found " Jour. Exp. Med., 1918 (28), 645. " Ergel). der Physiol., 190.5 (5), 698; literature. »^ Virchow's Arch., 1912 (210), 67. *' Literature and full review by Giinther, Deut. Arch. klin. I\Icd., 1912 (105), 89; and by Jcsionek, Ergeb. inn. Med., 1913 (II), 525. »• Fischer and Meyer-Betz, Zeit. physiol. Chem., 1912 (82), 96. •7 H. Fischer, Munch, med. Woch., 1916 (63), 377; Zeit. physiol. Chem., 1916 (97), 109 and 148; Schuinin, ibid., 1915 (96), 183. BLOOD PIGMENTS 485 to bear to chlorophyll,^'^ with wliicli luMiioslobin is so closely rolatod functionally. It is also interesting to consider that whereas carnivora obtain much hemoglobin in their food, herbivora obtain much chlo- rophyll. Pathologically, porphjTin is of interest as a urinary pigment, being found normally in the urine in traces, but present in considerable quantities in many diseases, ^^ such as rheumatism, tuberculosis, various liver diseases, and, most strikingly, after the administration of sulphonal, veronal or trional. A congenital form of hematoporphy- ria occurs, in which the blood contains free hematin and a porphyrin (Schumm),^ about 0.3-0.4 gm. being usually excreted daily in the urine; in the blood it is accompanied by hematin and bilirubin. When in abun- dance it may color the urine a rich Burgundy red, and it is sometimes accompanied by a precursor, uro-fuscin. It is present in the bones of animals showing hemochromatosis and in the bones of persons- exhibiting the congenital form of "hematoporphyria," described by Giinther, which is accompanied by severe skin lesions that are ascribed to the action of light upon the skin sensitized by the hematoporphyrin. Hausmann'' and others have studied extensively the photosensitizing action exhibited by hematoporphyrin and other porphyrins, and find evidence suggesting a relationship between hematoporphyria and " hydroa aestiva," and other conditions in which the skin is abnormally sensitive to light. An acute form of porphyrinuria has been described, usually in women, and sometimes associated with ascending motor paralysis.''" Afterinjection of 0.2 gm. hematoporphyrin into his own veins Meyer- Betz'* found himself so sensitized to light that exposure to the sun caused severe skin reactions during a period of weeks, and exposure to the Finsen light produced severe ulceration; but little hemato- porphyrin escaped in the urine. Many other products of blood destruction tested on animals were without sensitizing effects. IVIeth- ylation of the pyrrol groups only partially removes the activity of hematoporphyrin. Porphyrin obtained from urine and feces by Fischer also sensitized mice to light. Sufficient doses of hematoporphyrin may sensitize mice so that they become narcotized and die in a few minutes after exposure to intense light, a true "light stroke." Pseudomelanosis. — When loosely bound iron is present in the tissues, and in the same tissues sulphides are produced tlirough bac- terial action, a discoloration with sulphide of iron will result, which is called pseudomelanosis, because the pigment resembles true melanin in its blackness. This is most frequenth'- observed as a postmortem '^ For literature see Abderhalden, "Lehrbuch der physiol. Chemie," 1906. 95 See Garrod, .Jour, of Physiol., 1892 (13), 598. iZeit. physiol. Chem., 1916 (98), 123; 1919 (105), 158. ' Hegler et al, Deut. med. Woch., 1913 (39), 842. = Biochein. Zeit., 1910 (30), 276; 1914 (67). 309. 3<»L6ttier, Corr.-bl. f. Schweizer Aerzte, 1919 (49), 1871. * Deut. Arch. klin. Med., 1913 (112), 476. 486 PATHOLOGICAL PIGMENTATION phenomenon in and about the abdominal cavity, and in the ordinary postmortem discoloration both the liberation of the iron from its firm organic combination, and the production of hydrogen sulphide, are the work of bacteria. Pseudomelanosis may occur intra vitam, particularly in the margins of infected areas, and it may also be ob- served in the intestines, liver and spleen, and about the peritoneum, in bodies examined immediately after death, before any evident post- mortem decomposition has set in. This seems to depend upon the previous intra vitam formation of hemosiderin, which is then combined by sulphur liberated from tissue proteins through bacterial action.^ Methemoglobin. — If hydrogen sulphide acts upon hemoglobin that has not been decomposed, a greenish compound of sidphur- methemoglobin is formed (Harnack^), which is the cause of the greenish color seen in the abdominal walls and along the vessels of cadavers. This union of hemoglobin and hydrogen sulphide occurs only when oxygen is present (oxyhemoglobin). The sulphur-hemoglobin com- pound is readily decomposed by weak acids, even by CO2, with the formation of methemoglobin, which in turn readily becomes decomposed to form hematin. During life sulphemoglobin may form in the cir- culating blood, the sulphur presumably coming from intestinal putre- faction, and hence the condition is called "enterogenous cyanosis," which term also covers methemoglobinemia produced by nitrites formed in the intestines. '^ The latter condition is also present in poisoning by phenacetin,^ aniline and acetanilid, and related pigments appear in the blood in poisoning with chlorates and nitrobenzol. Pneu- mococci and Streptococcus viridans, as well as some other bacteria, may produce methemoglobin.^ In infections with B. em-physem.atosus , Schumm found this pigment free in the blood. Van den Bergh^° has found sulphemoglobinemia in puerperal sepsis, and probably these pigments could be found in other conditions if sought. Hemof uscin is the name given by von Recklinghausen to the brownish pigment found in involuntary muscle-fibers, particularly in the wall of the intestine. It does not react for iron, and is insoluble in alcohol, ether, chloroform, or acids; therefore it is not a lipochrome. It is bleached by H2O2, and is often found associated with hemosiderin which is not bleached. Von Recklinghausen, and also Goebel,^' ascribe this pigment to an alteration of hemoglobin which enters the cells in dissolved form, but Rosenfeld,^- who has subnutted the mater- ' Ernst, Virchow's Arch., 1898 (152), 418. Literature. « Zeit. physiol. Chem., 1899 (2G), 558. ' West and Clarke, Lancet, Feb. 2, 1907; Davis, ibid., Oct. 26, 1912; Gibson Quart. Jour.Med., 1907 (1), 29; Long and Spriggs, /6/(/., 1918 (IH, 102; .Tamieson, ibid., 1919 (12), 81. 8 See Heuhner, Arch. exp. Path., 1913 (72), 241. 9 Cole, Jour. Kxp. Med., 1914 (20), 303; Blake ibid., 1916 (24), 315; Schumm, Zeit. physiol. (Miem., 1913 (87), 171. "• Nederl. Tijd. Geneesk., 1918 (1), 1774. " Virchow's Arch., 1894 (136), 482. " Arch. exj). Path. u. Pharin., 1900 (45), 46. HEMOCIIROMA TOSI S 487 iai to analysis after isolation, found 3.70 per cent, of sulphur, from which he considers that it is related to the nielanins or melanoid sub- stances. The substance is readily dissolved by alkalies, and con- tains no iron. According to Taranoukhinc,''' the pigment in the myo- cardium in brown atrophy of the heart is also derived from proteins, and is neither a lipochrome nor a hemoglobin derivative. Other observers, however, consider this pigment a lipochrome or a lipofuscin. It is probable that the name hemofuscin has been given to several different pigments, which resemble one another only in that they do not con- tain iron. Strater^"* says that the name hemofuscin cannot be used for the pigment of the involuntary muscles, as he finds evidence that it does not arise from hemoglobin and is probably a waste pigment; but hemofuscin is found in epithelial and connective tissue cells. Hemochromatosis."'' — -This name was given by von Reckling- hausen to a condition in which the organs and tissues throughout the body are abundantly infiltrated with two pigments; one, iron-con- taining, identical with hemosiderin; the other seems to be the same as the hemofuscin described above. It is usually distinguished from general hemosiderosis in which only the iron pigment is deposited, ^^ although there are numerous observers who believe that all the pigment in hemochromatosis contains iron, but in some of the pigment the iron is firmly bound and difficult of demonstration. The hemosiderin is found chiefly in the parenchyma cells of the glandular organs, espec- ially the liver and pancreas, which organs usually show marked inter- stitial proliferation. The hemofuscin is found in the smooth muscle fibers of the gastro-intestinal tract, blood vessels, and genito-urinary tract. Under the heading of local hemochromatosis, von Reckling- hausen grouped such conditions as brown atrophy of the heart and pigmentation of the intestinal wall, which probably are quite distinct from the generalized hemochromatosis, since the local form occurs as a physiological process in old age. Hess and Zurhelle found 38.7 gm. of iron in the liver in one case (the normal amount is 0.3 gm.), and Bernouille'^^ found 18.3 gm. or 2.95 per cent, of the dry weight in the liver, 2.65 per cent, in the pancreas, and the same in the spleen. Anschiitz found 14.69 per cent, in the lymph glands, 7.62 per cent, in the liver, and 5 per cent, in the pancreas of a case. Muir and Dunn^^ obtained the following percentage figures: Liver, 6.43; pancreas, 2.49; spleen, 0.825; retroperitoneal glands, 11.64; kidneys, 0.406; 13 RousskyArch. Patol., 1900 (10), 441. ,. 1^ Virchow's Arch., 1914 (218), 1. 15 Literature given by Sprunt, Arch. Int. Med., 1911 (81, 75; Potter and Milne, Amer. Jour. Med. Sci., 1911 (143), 46; Roth, Deut. Arch. klin. Med., 1915 (117), 224; McCreery, Canada Med. Assoc. Jour., 1917 (7), 481; Howard and Stevens, Arch. Int. Med., 1917 (20), 896. '^ In lower animals occurs a form of hemochromatosis affecting especially the bones, and sometimes mistaken for ochronosis. (See Teutschlacnder, \'irchow's Arch., 1914 (217), 393.) 1' Corr.-Bl. Schweiz. Aertze, 1911 (40), 610. 18 Jour. Path, and Bact., 1914 (19), 226. 488 PATHOLOGICAL PIGMENTATION adrenals, 0.121; heart, 0.714; skin, 0.188; small intestine, 0.14. (Other analytical results are given by Howard and Stevens.) Opie's conclusions concerning this disease are as follows: (1) There is a distinct morbid entity, hemochromatosis, characterized by widespread deposition of an iron-containing pigment in certain cells, and an associated formation of iron-free pigments in a variety of locahties in which pigment is found in moderate amount under physiological conditions. (2) With the pigment accumulation there occur degeneration and death of the containing cells and consequent interstitial inflammation, notably of the liver and pancreas, which become the seat of inflammatory changes accompanied by hyper- trophy of the organ. (3) When chronic interstitial pancreatitis has reached a certain grade of intensity, diabetes ensues and is the terminal event in the disease. Diabetes occurs in the majority of the cases of generalized hemochromatosis (50 of 63 collected by Sprunt) and was called by Hanot, "bronzed diabetes," because of the coloration of the skin. It has been suggested that the pigmentation is due to decomposition of the blood-corpuscles in the diabetic blood, but the pigmentation and sclerotic changes precede the diabetes, which is secondary to the atrophic and sclerotic changes in the pancreas. It seems probable that both the pigment formation and the tissue changes depend upon some intoxication, the origin and nature of the toxic agent being entirely unknown. In many cases it has seemed possible that alcohol might have been the inciting cause. There is no evidence of abnormal boold destruction which might account for the pigmentation, and ]\leltzer and Parker have suggested that the difficulty lies in the inability of the tissues to get rid of the iron set free in normal catabolism. INIetab- olism studies have indicated that there is some retention of food iron which may be interpreted as supporting but not proving this hypothe- sis.^^ Rous and Oliver, -° finding that protracted hemolysis of foreign corpuscles in rabbits produces a typical mild hemochromatosis, sug- gest that the liver cirrhosis is primary and renders this organ unable to deal adequately with the blood pigments, which therefore accumu- late in the organs and cause diffuse fibrosis. ICTERUS^' Pigmentation of the tissues of tlie body in jaundice depends upon the presence in them of bile-pigments, which usually have been formed in the liver and reabsorbed either into the lymph or blood (or botii). However, a pigment that seems to be chemically identical with bili- rubin (hematoidin) may be formed from hemoglobin liberated on the " See Howard and Stevens {loc. cit.) and McCluro, Arch. Int. Med., 1918 (22), 610. 2" Trans. Assoc. Anier. Phys., 1918 (3.3;, 132. 2' Literature l)y Stadelinann, "Der Icterus," Stuttgart, 1891; Minkowski, Ergebnisse der Pathol., 1S9.') (2), 079. ICTERUS 489 breaking up of red corpuscles, and possibly this may be produced in sufficient amounts outside of the liver to give rise to general icterus. Certainly the local greenish-yellow pigmentation occurring in the vicinity of extravasations of blood, due to hematoidin formation, may be looked upon as a "local jaundice," and in icterus hematoidin^^ crystals may be found in the tissues.-^ Bile-pigments. — Bilirubin is of a reddish-yellow color, and it Is the chief pig- ment of human bile. Its formula is CsiHssN-iOe or CssHjeN^Oe, and its relation to hematin, from which it is formed, is shown by the following formula, which ex- presses the manner in which blood pigment may be converted into bilirubin by the liver under normal conditions, and into hematoidin (its isomer) in the tissues and fluids of the body in pathological conditions: CaiHa^N.OjFe + 2H2O = C3.H38N4O6 + FeO. (hematin) (hematoidin or bilirubin) Bilirubin is not soluble in water, but dissolves in the alkaline body fluids as a soluble compound, "bilirubin alkali." It is very slightly soluble in ether, ben- zene, carbon disulphide, amyl-alcohol, fatty oils, and glycerol, but is more soluble in alcohol and in chloroform. Biliverdin, 0^411:1^^40$, as its formula indicates, is an oxidation product of bilirubin. Bilirubin in alkaline solutions will oxidize into biliverdin merely on exposure to the air, and the change from yellow to green of icteric specimens when placed in oxidizing solutions (e. g., dichromate hardening fluids) is due to the formation of the green biliverdin. Biliverdin is the chief pigment of the bile of carnivora, but it is also present in varying amounts in human bile. The various other biliary pigments, namely, bilifuscin, biliprasin, choleprasin,-* bilihumin, and bilicyanin, are probably not normal constituents of bUe, but are oxidation products of bilirubin, and are found chiefly in gall-stones (q. v.). A pigment similar to urobilin maj'- be present in normal bile. The total amount of pigments present in bile is probably not far from one gram per liter; rather under than above this amount. Etiology of Icterus. — ^Although hematoidin, which is isomeric if not identical with bilirubin, may be formed outside of the liver when red corpuscles are broken up in hemorrhagic extravasations, and possibly also when they are broken up within the vessels by hemolytic agents, yet it was formerly held that a true general icterus does not occur without the liver being implicated. This view rested on evi- dence of various sorts. First, the classical experiments of Minkowski and Naunyn,-^ which demonstrated that in geese the production of hemolysis by means of arseniuretted hydrogen leads to icterus, but if the livers of the geese have been previously removed, no icterus follows the poisoning. Second, the repeated demonstration that in icterus produced by septic conditions, poisoning, etc., which was formerly looked upon as Si "hematogenous" icterus, the urine contains bile salts as well as pigment, indicating an absorption of bile from the liver. Third, the finding of histological evidence that in so-called hematogen- " See Guillain and Troisier, Semaine Med., 1909 (29), 133; Widal and Joltrain, Arch. med. expcr., 1909 (21), 641. •" Dunzelt, Cent. f. Path., 1909 (20), 966. 24 SeeKiister, Zeit. physiol. Chem., 1906 (47), 294. " Arch. f. exp. Pathol, u. Pharm., 18S6 (21), 1. 490 PATHOLOGICAL PIGMENTATION ous icterus there occur occlusions or lesions of some sort in the bile capillaries, which can account for the reabsorption of the bile into the general circulation. ^^^ Therefore, it was beheved that the pigments that produce the general discoloration of icterus are, at least for the most part, manufactured by the liver, whatever the cause of the re- absorption of the bile from the Hver into the blood may be. That hemolytic agents cause icterus was explained by the fact that on account of the large amounts of free hemoglobin brought to the liver, excessive amounts of bile-pigments are formed, which render the bile so viscid that it blocks up the fine bile capillaries; on account of the low pressure at which bile is secreted, a slight obstruction of this kind is sufficient to stop entirely the outflow of bile, which then enters the capillaries of the liver and also, to a less extent, the lymphatics.-^ It is also possible that the hemolytic poisons injure the liver-cells so much that the minute intra- and intercellular bile capillaries become disorganized, and permit of escape of bile into the lymph-spaces and its absorption into the blood-vessels. ^^ Swelling of the degenerated liver-cells may also be an important factor in the occlusion of the bile capillaries; swelHng of the lining cells of the bile capillaries may also coexist, and fibrin may occlude them in toxic or infectious icterus. However, Whipple and Hooper-^ have obtained experimental evi- dence that after intravenous injection of hemoglobin into dogs with the liver excluded from the circulation, bile pigments appear in the urine and icterus is manifested in the fat tissues, from which observa- tions it is concluded that the Hver may not be the only place in which bile pigment can be formed from hemoglobin.''" Several authors have found bilirubin produced in hemorrhagic effusions located where the liver could have had no influence. ^^ We also recognize tj'pes of hemo- lytic icterus in which the liver does not seem to be concerned, and with bile pigments present in the blood and urine unaccompanied by bile salts (dissociated icterus), so that the old dictum of the essential implication of the liver in icterus seems to be incorrect.-'-' Joanno- vics^'' gives, as a result of a comparative study of icterus from bile obstruction and icterus from hemolysis, the following chief differ- 2« See Eppingcr, Ziegler's Beitr., 1903 (33), 123; Gerhardt, Miiuch. incd. Woch., 1905 (52), 889. Lang (Zeit. exp. Path. ii. Ther., July, 190G (3), 473) has demon- strated the presence of fibrinogen in the bile in i)hosph()rus-poisoning, which perhaps accounts for the "bile thrombi" ol)servod l)y Kppinger in toxic icterus. " See Mendel and Underbill, Amer. Jour. Phj^sioL, 1905 (14), 252; Whipple and King, Jour. Exp. Med., 1911 (13), 115. 28 Sterling, Arch. exp. Path., 1911 ((U), 468; Fiessinger, Jour. Physiol, et Pathol., 1910 (12), 958. Oertel (Arch. Int. Med., 191S (21), 73) suggests (hat intracellular preciintation of bile pigment within liver cells altered by poisons may prevent its excretion into the bile canaliculi. -'■Mour. Exi)er. Med., 1913 (17), 593 and 012. '" Attempts to i)roduce bile pigments from hemoglobin by bacterial action have been unsuccessful. (()uadri, Fol. Clin. Chim., 1914, No. 10.) =" Hooper and \Vhipi)le, Jour. Kx]). Med., 191() (23), 137. '■>- See Leiu'liiic, Ikntr. i)atli. Anat., 1917 ((14), 55. '•■' Zeit. f. llcilk.. Path. Abt., 1904 (25), 25. ICTERUS 491 ences: Icterus due to hemolysis appears sooner than icterus from bile-duct occlusion, and reaches a much higher degree; the obstruction in hemolytic icterus, when present, is intra-acinous; in stasis it is chiefly inter-acinous; in hemolytic icterus there is a large splenic tumor due to accumulation of degenerated red cells in the spleen, where they become disintegrated preliminary to the formation of bile- pigment. If the spleen is removed, hemolytic agents may not cause icterus, because the corpuscles are not then prepared for pigment formation. ^^ In obstructive icterus from gall stones there is a choles- terolemia in proportion to the amount of icterus, which is not usually true of icterus from other causes." Toxicity of Bile. — In any event, we must appreciate that in icterus not only are abnormally large quantities of bile-pigment present in the blood, but also usually the other less conspicuous constituents of the bile. Whole bile of rabbits is fatal to rabbits in doses of 0.2.5 to 0.5 cc. per kilo, by intraperitoneal injection, and about half as much intravenously (Bunting and Brown^"). Death is the result of changes in the myocardium, where necrosis is produced; and severe degenera- tive changes are also found in the kidneys and liver; when the bile is injected into the peritoneum, pancreatitis and fat necrosis result. The relative toxicity of the bile-pigments and the bile salts is not as yet uniformly agreed upon. Bile-pigments. — Bouchard" and others have claimed that the bile- pigments are far more toxic than the bile salts, which is contradicted by Rywosch and others. Bihrubin is normally present in the blood, and is probably responsible for the yellow color of the plasma. ^^ It is always present in excess in icterus of whatever degree. ^^ A series of analyses by Gilbert^" and others gave the following results: Normal blood-serum contains 0.027-0.08 gram bilirubin per liter; in obstructive icterus they found 0.7 to 1.0 gram of bihrubin per hter, in biliary cir- rhosis 0.33 gram per liter, in icterus neonatorum 0.2 to 0.5 gram; in pneumonia 0.068 gram was found. These figures, however, are far in excess of those described by later investigators. Bauer and Spiegel"*^ give figures of about 1 part in 100,000 to 200,000, or 0.01 to 0.005 gm. per hter. In icterus the highest figure given by these authors was 0.07 gm. There is a marked variation between different normal individuals, but for the same person the figures are nearly constant. The threshhold value for the blood seems to be about 1 part in 50,000; ** The etiology of icterus neonatorum (when not obstructive) has not been ascertained, but a natural tendency towards icterus is said to exist in the new- born, their blood containing much more bile pigment then than later. (Hirsch, Zeit. Kinderheilk., 1913 (9), 196; \lppa, Miinch. med. Woch., 1913 (39), 2161.) " Rothschild and Felsen, Arch. Int. Med., 1919 (24), 520. 3« Jour. Exper. Med., 1911 (1-4), -145. " Literature and discussion by Stadelmann, Zeit. f. Biol., 1896 (34), 57. 3s Blankenhorn, Arch. Int. Med., 1917 (19), 344; 1918 (21), 282. '9 Feigl and C^uerner, Zeit. exp. Med., 1919 (9), 153. "» Compt. Rend. Soc. Biol., 1905 and 1906. ^1 Deut. Arch. klin. Med., 1919 (129), 17. 492 PATHOLOGICAL PIGMENTATION when this proportion is exceeded the pigment begins to be deposited in the skin and excreted by the kidneys (v. d. Bergh), However, Blankenhorn believes that at times the bihrubin in the plasma is so bound that the kidneys do not excrete it, and yet it may be able to diffuse into the skin.'*^ AVith reduced renal function the amount of pigment in the blood may be increased without hepatic chsease. King and Stewart'*^ state that the amount of pigment in a lethal dose of whole bile will cause death, but the bile salts present in the same quantity of bile will not cause recognizable effects; uncombined pigment is more toxic than its calcium or magnesium salts. Bile from which the pigment is removed has very little toxicity. They suggest that calcium is increased in the blood in icterus as a protection against the toxic effects of the pigments. The combining of the calcium with bile pigment, however, renders it unavailable for fibrin formation, and this seems to be an important factor in the hemophilic tendency of icterus,'** and Pettibone records a marked decrease in blood calcium in protracted jaundice.*^ The decrease in available calcium may also be responsible for the bradycardia and some of the mental and nervous symptoms. Bile salts are said to be toxic, generally producing depression of the central nervous system, with resulting coma and paralysis; they are also decidedly toxic to cells of all sorts, causing hemolysis and marked destruction of tissue-cells. Small quantities of bile salts stimulate the central end of the vagus, and large amounts influence the heart itself; hence in icterus we observe a slowing, and often an irregularity, of the pulse, and the blood pressure is lowered. Al- though there has been much dispute as to whether the chief effects of icterus upon the heart depend upon action of the bile salts upon the vagus, or upon the intracardiac ganglia, or upon the muscle itself,"'*' yet Weintraud demonstrated that in some cases of icterus administra- tion of atropin, which paralyzes the vagus, stops the bradycardia, indicating the importance of the effects of the bile salts upon the vagus in causing this feature of cholemia. According to Meltzer and Salant,*''' bile also contains a tetanic element which disajipears from stagnating bile; the bile salts contain this tetanizing agent in less amount than does the whole bile. But King'** and others ascribe most of the effects of bile on the heart to the bile pigments, perhaps through abstraction of the calcium. Taurin given in 10 gm. and even larger " Corroborated by Meulengracht (Ugesk. f. Laeger., 1919 (SI), 1785) who states that when bilirubin reaches a certain concentration it passes into the tissues, but not into the urine until a higher concentration is readied. " Jour. Exi)er. Med., 1909 (11), 07:i *" See Lee and Vincent, Arch. Int. Med., 1915 (16), 59. ^i" Jour. Lab. Clin. Med., 191S (3), 275. " See Minkowski, Ergel). der Pathol., 1895 (2), 709. " Jour. Exp. Med., 1906 (8), 128; review and literature concerning toxicity of bile. " Sec King, Bigelow and Pcarce, Jour. V]\\\n\ Med., 1912 (14), 159. ICTERUS 493 doses by mouth, subcutaneously and intravenously to man produced no noticeable effects (Schmidt).'*^ Since the bile salts cause hemolysis, and since in even "hematogen- ous" jaundice they may enter the blood, it can readily be seen that in this way an increased formation of bile-pigment may be incited which leads to further obstruction to the outflow of bile from the liver, and a "vicious circle" may thus be established. The necroses observed in the liver in icterus, "icteric necrosis," are generally ascribed to the cj'totoxic effects of the bile salts, although it is difficult always to exclude infection extending along the bile-ducts to the liver tissue. Possibly the power of bile salts to dissolve lipoids may be responsible for the cytotoxic effects^" as well as for the hemolysis. The itching and irritation of the skin in icterus may be due to the effect of the bile-salts deposited in it, for pruritus is said to be absent in the pig- mentary jaundice of congenital hemolytic icterus. There is also an increase in the cholesterol in the blood, which may be related to the "xanthomas" that form in chronic icterus. ^^ Unfortunately we have no accurate method for quantitative determination of the amount of bile salts in the blood. A remarkable tendency to spontaneous hemorrhages, frequently ob- served in icterus, probably depends upon injury to the capillary endothelium by the bile salts, ^- while the protracted, often uncontrol- lable, hemorrhage that may occur from operation wounds in icteric pa- tients, is related to the slowed coagulation of the blood observed in icterus. The bile salts themselves may delay coagulation by interfer- ing with the conversion of fibrinogen into fibrin. ^^ The cytotoxic effect of the bile salts is also shown by the albuminuria of icteric per- sons, which frequently results from the renal lesions the bile produces. Although bile itself is toxic to many bacteria, especially the pneumo- coc.cus,'^'* 3^et in icterus the bactericidal power of the blood is lowered, and infections are prone to develop and to be severe; moreover, the growth of several species of bacteria is favored by bile.^" Croftan^'' summarizes the physiological effects of bile acids as fol- lows: (1) A powerful cytolytic action, affecting both blood-cor- puscles and tissue-cells. (2) A distinct cholagogue action. (3) In small doses (1-500) they aid coagulation. (4) In large doses (1-250 and over) they retard coagulation. (5) Slow the heart action. ^^ (6) In small doses they act as vasodilators; in large doses, as vaso con- " Schmidt et al, Jour. Biol. Chem., 1918 (33), 501. 50 Neufeld and Handel, Arb. kaiserl. Ges.-Amte, 1908 (28), 572. 51 Chauffard, Presse Med., 1913 (21), 81; Chvostek, Zeit. klin. Med., 1911 (73), 479; Pinkus and Pick, Deut. med. Woch., 1908 (34), 1427. " See Morawitz. Arch. exp. Path., 1907 (56), 115. " IIaes>ler and Stebbins, Jour. Exp. Med., 1919 (29), 445. *'* See Neufeld and Haendel, loc. cit. « See Meyerstein, Cent. f. Bakt., 1907 (44), 434. 56 New York Med. Jour., 1906 (83), 810; see also Faust, "Die tierische Gifte," Braunschweig, 1906, p. 29. " See Berti, Gaz. degli Osped., 1916 (37), 1233. 494 PATHOLOGICAL PIGMENTATION strictors. (7) Reduce motor and sensory irritability. (8) Act on the higher cerebral centers, causing coma, stupor, and death. Sel- lards^^ found that injection of bile salts into guinea pigs causes ulcer- ation and hemorrhage in the stomach. It is difficult to decide how much of the profound intoxication that is sometimes present in icterus ("cholemia" and "icterus gravis") to ascribe to the reabsorbed bile, for frequently there is an accompany- ing infection, and even if there is no infection the impairment of liver function by the obstruction of bile outflow must also be reckoned with. The liver is not only the great destroyer of toxic substances absorbed from the alimentary canal, but it is also an important seat of nitrogenous metabolism, interference with which may lead to ac- cumulation of many toxic nitrogenous substances in the blood. ^^ The long duration of severe icterus in some cases of occlusion of the bile- ducts, with relatively shght evidences of intoxication, would seem to indicate, however, that on the whole the bile is not so much respon- sible for the intoxication observed in icterus as are the associated conditions. On the other hand, in not a few instances it has been observed that escape of large quantities of bile into the peritoneal cavity may be followed by symptoms similar to those of icterus gravis; in these cases only the bile can be held responsible for the intoxica- tion.«" Dissociated Jaundice*^' is the existence of cither bile salts or bile pigment sep- arately in the blood. This may be produced either by the bile salts being ex- creted by the kidney, leaving only the less diffusible pigment in the blood, or by separate escape of bile salts from the liver into the blood. Also in true hem- olytic icterus we may have bile pigments present in the blood without bile salts. Congenital Hemolytic Icterus.'^- — This term describes a condition characterized by a chronic, non-obstructive jaundice, without evident intoxication. A similar condition is also observed developing in adults, without familial tendencies. The congenital form usually shows familial character, but isolated congenital cases do occur. It is the result of active hemolysis, apparently taking place chiefly in the spleen, and leading to an icterus without evident participation of the liver. The cause of the hemolysis is entirely unknown, although there is a marked fragility of the erythrocytes evidenced by reduction of their resistance to hypotonic solutions, and it results in a moderate anemia, with excretion of much uroliilin in both stools and urine; the blood contains biliru])in which is not excreted in the urine. The jaundice is usuall3^ unaccompanied by evidence of cholemia, icteric pruritus or hemophilia. The spleen is greatly enlarged and improvement has generally fol- lowed splenectomy but the exact relation of the spleen to the disease is not known." The frequent occurrence of gall stones in this condition may be the result of hyper- cholesterolemia from hemolysis. The metabolism of a case*^' showed loss of nitrogen, calcium, magnesivnn and iron, and a much increased uric acid excretion. These conditions may improve after operation.^^ "^ Arch. Int. Med., 1909 (4), 502. '« See Bickel, E.xper. Untersuch. iiber der Pathol, der Cholaemie, Wiesbaden 1900. «»See Ehrhardt, Arch. klin. Chir., 1901 (64), 314. 0' Hoover and Hlankenhorn, Arch. Int. Med., 1916 HS), 289. " See Richards and Johnson, Jour. Amer. Med. .Vssoc, 1913 (61), 1586. •3 See series of articles by Pearce el al., in Jour. Exp. Med., on Relation of Spleen to Blood Destruction. 6< McKelvv and Rosenbloom, Arch. Int. Med., 1915 (15), 227. « Goldschmidt, Pepper and Pearce, Arch. Int. Med., 1915 (16), 437. UROBILIN 495 The Pigmentation in Icterus. — Living tissues have but a slight tendency to take up bilc-pignicnts, much of the tissue-staining ob- served at autopsy being due to postmortem imbibition from the blood and lymph. Quincke*^® found that after subcutaneous injection of bihrubin only the connective tissue, both cells and intercellular fibrils, becomes diffusely colored; later, it fades out of the cells, leaving only the fibrils stained. Muscle-cells, fat-cells, and vessel-walls take up the pigment only after their death. If the jaundice continues for a long time, the subcutaneous deposits of bilirubin may undergo a slow oxida- tion, the color changing to an olive or to a dirty grayish green. The pigment in the connective tissues is at first in solution, but may be de- posited in a granular form after a considerable amount has accumu- lated. Bile pigments and bile salts may both be present in consider- able amounts in the blood and not pass through the kidneys, and also they may fail to pass into the tissues; hence we may have cholemia without icterus or choluria, because of the firmness with which the pig- ments are bound in the plasma (Hoover^^). The question whether in icterus the skin may be colored b}' other pigments than bilirubin, especially by its reduction product, urobilin, seems to have been decided negatively. Bile-pigment is probably not absorbed as such from the intestine in sufficient quantity to cause icterus. Such bile-pigment as enters the blood from the liver is ex- creted through the kidneys chiefly, but also in the sweat. Ordinarily, other secretions (milk, tears, saliva, sputum) are not colored in jaun- dice, but if the secretions are mixed with inflammatory exudations, they may then be colored (e. g., pneumonic sputum). When the bile- pigment is resorbed from the skin, it may be in part transformed into urobilin, which appears in the urine in increased amounts during the period of recovery from jaundice. Part of the bile-pigment is prob- ably eliminated by the liver after the cause of obstruction has been removed from the bile-passages. Urobilin" This pigment is probably formed chiefly, if not solely, from bile pig- ments by the action of reducing bacteria in the intestine. It is ex- creted in the urine only as its chromogen, urobilinogen, but in the feces both urobilin and urobihnogen may be found; when exposed to air the chromogen oxidizes quickly to urobilin. Addis^* states that bilirubin is reduced to urobilinogen in the bowel and is then largely absorbed, to be at once oxidized and polymerized into urobilin, two molecules of urobilinogen uniting under the influence of oxj^gen to form one of urobilin. In the liver the urobilin is largely worked over «« Vichow's Arch., 1884 (95), 125. «^ Bibliography and review by Meyer-Betz, Ergeb. inn. Med., 1913 (12), 734; Wilbur and Addis, Arch. Int. Med., 1914 (13), 235. 68 Arch. Int. Med., 1915 (15), 412. 496 PATHOLOGICAL PIGMENTATION to form new hemoglobin, and hence the functional capacity of the liver is indicated by the completeness with which it utilizes the uro- bilin, except in cases of excessive formation of urobilinogen as a re- sult of hemolysis. The amount of urobilinogen in the urine will be found increased, therefore, in hemolytic icterus, and decreased in ob- structive icterus. Exceptionally, urobilinogen may be formed from blood disintegrated in bloody effusions without evident participation of the liver, e. g., urobilinogenuria with hemorrhagic ascites, hemolytic poisons, etc. With a normal liver urobilinogenuria is found only when there is excessive hemolysis, otherwise urobilinogenuria occurs only with an injury to the liver parenchyma (Hildebrant). In general, the amount in the urine is an index of the amount of blood destruction.^^ There seems to be little if any retention by imperfectly functioning kidneys (Blankenhorn) and it can often be found in the urine when not demonstrable in the blood. Occlusion of the bile ducts stops an ex- isting urobilinogenuria by preventing the formation of urobilinogen in the intestine. Normally there is a very small amount of urobilin- ogen and related substances in the urine, which disappears when there is no bile in the intestine. Fromholdt^^ considers that increased ■ bacterial reduction in the intestines may by itself account for uro- bilinogenuria. The amount of urobilin and urobilinogen excreted in the feces, seems to vary directly with the amount of hemolysis, ^^ and the same is true for the duodenal contents.'^' The evidence of abnor- mal hemolysis is said to occur first in the stools, then in the duo- denal contents, and lastly in the urine; the presence of even small amounts of urobilinogen in the urine being evidence of a probable per- nicious anemia in the absence of signs of biliary and hepatic disease.^-" Digestive Disturbances in Obstructive Icterus." — In case the icterus depends upon the occlusion of the main bile-passages by stones, tumors, etc., the situation is complicated by the effects of the absence of this natural secretion in the in- testinal canal. Carbohydrate and protein digestion seem to be but little affected, especially the former, but the proportion of the ingested fat that appears in the feces increases from the normal 7-11 per cent, to GO- See Bayon, Lancet, 1912 (ii), 1579. '» See Leuenberger, Beitr. klin. (^hir., 1912 (SO), 20S. lo^Zeit. angew. Chem., 1919 (32), 333. ^' "Researches into Induced Cell Reproduction and Cniioer," London.] PROTEINS OF TUMORS 499 designating them as "auxetics." These seem to be present in the anthracene fractions of tar,'- wliich may explain the frequency of can- cer in workers in tar, soot and parafhn. Japanese investigators report that protracted irritation of rabbits' ears with tar leads to strikingly infiltrative proliferation of the epithelium, with metastasis.^'' The influence of various salts on cell gi-owtli has also been applied to cancer pathology, and while we have abundant evidence that chemical sub- stances may either stimulate or check cell growth, as well as regulate it, our biological chemistry has not yet given us any very substantial facts on these problems.^'* Nevertheless, numerous observations have been made concerning the chemistry of tumors, which, although they do not as yet throw any important light on the fundamental problems of tumor pathology, are of much interest. These may be briefly summarized as follows: A. CHEMISTRY OF TUMORS IN GENERAL (1) Proteins. — Earlier studies showed that tumor growths con- tain the same sorts of proteins as do normal tissues, apparently in about the same proportions, and in spite of certain contradictor}- re- ports this statement seems to be correct. In all probability the nucleoproteins of tumors share the specific characteristics of the nucleoproteins of the tissues from which they arise — at least this is the case with the nucleoproteins of lymphosar- coma, according to Bang.'^ This seems to have been confirmed by Beebe,'^ who found nucleo-histon only in lymph-gland tissue, but the distinction between thymus and lymph-gland nucleohiston is probably not so easily made as Bang intimates. Because of their richly cellular structure cancers may contain more nucleoprotein than the tissues from which they arise. ^^" However, Wells and Long'^ found the proportion of purine nitrogen in tumors of several classes to be much lower than might be expected from the nuclear content as shown by the microscope; also, Satta^^ found unexpectedly low phosphorus figures and Yoshimoto^^ found no parallelism between the number of nuclei and the nuclein content. The purines present in tumor tissues are quite the same in nature and proportion as in normal tissues (Wells and Long), as also are the nucleoproteins. Bergell and D6rpinghaus-° have studied the nature of the proteins '2 Norris, Biochem. Jour., 1914 (8), 253. 1' Yamagiwa, Mitt. Med. Gesellsch., Tokio, 1916 (30), 1. " A theory of cell division in cancer as a result of electric forces is given by Jessup et al, Biochem. Jour., 1909 (4), 191. 1* Hofmeister's Beitr.. 1903 (4), 368. i^Amer. Jour. Physiol, 1905 (13), 341. i«»Petrey, Zeit. physiol. Chem., 1899 (27), 398. 1" Zeit. f. Krebsfrsch., 1913 (12), 598. 13 Arch. Ital. Biol., 1908 (49), 380. '5 Biochem. Zeit., 1909 (22), 299. " Deut. med. Woch., 1905 (31), 1426. 500 THE CHEMISTRY OF TUMORS in tumors by determining the proportion of the various amino-acids that compose them. Because of the amount of material necessary for the ester method, they were obhged to use a mixture of various primary and secondary cancers and one sarcoma. The protein of this tumor-mixture was characterized by the very high proportion of alanine, glutaminic acid, phenylalanine and aspartic acid, there being from 5 to 10 per cent, of each. Leucine was very low, 5-10 per cent., as against 20 per cent., or higher, found in most normal tissues. Gly- cine and tyrosine were present in small quantities, and serine was prob- ably also present. Neuberg^^ found in cancer protein 1.3 per cent, of tyrosine, 17 per cent, of leucine, scarcely 1 per cent, of glutaminic acid, and 4.92 per cent, of glja^inc. In five human tumors of different sorts, Kocher-- found high figures for diamino-nitrogen; his averages were: arginine, 12.42; histidine, 4,86; lysine, 11.23; total, 28.47 per cent, of the protein nitrogen. Drummond's'^ careful studies in this field have shown that the diamino-acid content of tumors is generally slightly higher than in corresponding normal tissues, prob- ably varying directly with the amount of nuclear material, there being nothing found to indicate that the hexone bases are in any way respon- sible for increased growth. Strange, and as yet unexplained, varia- tions in tryptophane content in various tumors were found by Fasal,-"* some having a very high try{)tophane figure, while in others none could be found. Centanni" found that trjqitophanc and tyrosine inhibit, while skatolo and indole stimulate carcinoma growth. Certain authors have believed that the cancer cell has a specific chemistry,-*^ but most of these analyses, including that of Abderhalden and Mcdigreceanu,-^ seem to indicate that cancer proteins have much the same composition as normal proteins. Cramer and Pringle-^ find that there is less nitrogen in mouse cancers than in equal amounts of other mouse tissue, the decrease being in the coagulable nitrogen, incoagulable nitrogen being relatively increased; a given amount of nitrogen produces more cancerous than normal tissue. The water content of rapidly growing tissues, whether normal or cancerous, was found to be high. This corresponds with the analysis of Robin,-' who found the water content high and nitrogen low in carcinomas of the liver, suli)hur being especially low, and C^hisholnr*" has found the proportion of nitrogen in several human tumors lower than in the soma- "^^ .\rb. a. d. Path. Inst, zu Berlin, 1906, p. 593. " Jour. Biol. Cheni., 1915 (22). 295. "Biochem. Jour., 1910 (10), 473. ^^ Biocheni, Zeit., 1913 (55), 88. " Tumori, 1913 (2), 406. -» Blumenthal, Zeit. Krebsforsch., 1907 (5), 183. -' Zeit. phvsiol. Cliem., 1910 (09), 00. =« I'roc. Koval Soc, B., 1910 (S2), 315; Jour. Phvsiol., 1910 (50), 322. ■'"Cent. Phvs. Path. StotYwechs., 1911 (0). 577. Bull. .Vcad. M6J., 1919 (81), 799; Coinpt. Rend. Acad. Sci., 1919 (108), 1071. "« Jour. Pathol, and Bact., 1913 (17), 000. CARBOHYDRATES OF TUMORS 501 tic tissue. However, the lack of any marked specific individuality of cancer proteins when tested by immunological reactions, indicates a very close chemical agreement with normal tissue proteins. ' On account of the amount of autolysis going on in tumors the products of protein splitting are usually present. Beebe''' found in a number of tumors leucine, tyrosine, tryptophane, proteoses (biuret reaction), and in one glycine. Drummond" has found leucine, tyrosine and creatinine commonly present in water extracts of malig- nant tumors. Because of the deficient circulation in the tumors, the amino-acids accumulate in the cancer tissues in sufficient amounts to be detected, and may be found even when no macroscopic evidences of degeneration are present. Possibly on account of this poor absorp- tion no proteoses, peptones, or amino-acids could be found in the urine of cancer patients by Wolff ;''^ but Ury and LilienthaP-* found a posi- tive reaction for albumose in the urine in about two-thirds of all car- cinoma cases examined by them; however, it may be absent even in advanced stages. Lactic acid is also present in tumors, according to Fulci^^ and Saiki,^^ the latter finding 0.48 gm. of lactic acid per 100 gms. cancer of the stomach. According to Clowes" cancer tissues are much more permeable to ions than are normal tissues. (2) Other Organic Constituents. — These, in general, resemble the organic constituents of the tissue from which the tumor arises, for a structural resemblance to the parent tissue always exists, and as structural features depend largely on the proportion of the chemical components, a structural similarity fairh' implies a chemical similarity. For example, adrenal and renal tissue contain much lecithin and choles- terol, and hypernephromas show a similar composition; the fat of a lipoma is, in its qualitative features, almost identical with the normal fat of the same individual; tumor melanin shows no characteristic chemical distinction from normal melanin, etc. Glycogen has been particularly studied in tumors, especially be- cause of the erroneous idea advanced by Brault that the quantity of glycogen is in direct proportion to the mahgnancy. From a sum- mary of all the evidence, it seems that two chief factors determine the presence and amount of glycogen in tumors. One is the embryonic origin of the tumors; thus tumors of cartilage, striated muscle, or of squamous epithelium, which tissues normally contain much glycogen, are hkewise provided with an abundance of this material. Second, the occurrence of areas of impaired cell-nutrition favors the accumu- lation of glycogen in the degenerating tumor-cells just as it leads to 31 Amer. Jour. Physiol, 1904 (11), 139. « Biochem. Jour., 1917 (11), 246. 35 Zeit. f. Krebsforschung, 1905 (3), 95. 3* Arch. f. Yerdauungskr., 1905 (11), 72. " Gaz. internaz. dimed., 1910, No. 24. 36 Arch. m^d. exp^r., 1911 (23), 376. 37 Proc. Soc. Exp. Biol. Med., 1918 (15), 107. 502 THE CHEMISTRY OF TUMORS a similar accumulation in all other tissues (Gierke). ^^" The most ex- tensive consideration of this topic is reported by Lubarsch,^^ who found glycogen microscopically in 447 (or 29 per cent.) of 1544 tumors examined. It was present in but 3 out of 184 fibromas, osteomas, gliomas, hemangiomas, lipomas, and lymphangiomas, and in but 2 out of 260 adenomas from various parts of the body. It occurred in all teratomas, rhabdomyomas, hypernephromas, and syncytiomas. In 138 sarcomas glycogen was present in 70 (50.7 per cent.); of 415 carcinomas it was found in 181 (43.6 per cent.). In the squamous epithelial cancers 70 per cent, contained glycogen, while the mucoid or colloid cancers were always free from glycogen. The glycogen undoubtedly enters the cells from without, probably entering as sugar, and being converted into glycogen by intracellular enzymes. We have no rehable studies of the actual quantity of glycogen in various tumors, although Meillere^^ states that the microscopic and chemical examina- tion of tumors give corresponding comparative results, which Gierke states is generally true with glycogen estimations. Pentoses. — Neuberg'*'^ reports finding, as a product of autolysis of a carcinoma of the liver, a pentose which was not produced by autolysis of either normal liver tissue or the primary growth in the stomach. Beebe*^ found that in carcinoma of the mammary gland the percentage of pentose (xylose) is somewhat higher than the amount in normal mammary glands f about 0.23 per cent.). Carcinoma in the hver did not show any constant excess of pentose above that of normal liver tissue (about 0.38 per cent.). A primary carcinoma of the liver showed quite the same pentose and phosphorus content as normal liver tissue. In general, no constant relation of pentose to origin, mahgnancy, or degeneration of tumors was observed. Purines and Purine Enzymes. — ^The purines of both benign and malignant tumors have been studied by Wells and Long, ''- who found them the same as those in normal tissues, and in much the same rela- tive proportions. The proportion of the total nitrogen of tumors which is constituted by the purine nitrogen is less than would be ex- pected from the histological evidence of the amount of nuclear material contained in the tumors. Tumors also seem to contain much the same purine enzymes as the normal tissues. Thus, guanase seems uni- versally present in tumors derived from human tissues, and adenase is missing, although autolyzing tumors can disintegrate their nucleic acid (nuclease) and change the adenine radicals of the nucleic acid into hypoxanthine, presumably by way of adenosine and inosine ( Am- berg and Jones). Secondary tumors growing in the human liver do "« Ziegler's Beitr., 1905 (.37), 502. '8 Virchow's Arch., 190G (183), 188. 3» Coinpt. Rend. Hoc. Biol., 1900 (52), 324. " Berl. klin. Woch., 1904 (41), 1081; 1905 (42), 118. "' Arncr. Jour. Physiol, 1905 (14), 231. *■' Zeit. f. Krebsforsch., 1913 (12), 598. LIPINS OF TUMORS 503 not acquire the enzyme, xanthine-oxidase, which is a characteristic enzyme of this organ. Tlie hver tissue between the cancer nodules seems to oxidize purines less actively then normal liver tissue. Long-*' has also found similar conditions in tumors from sheep, pigs and cattle, observing that primary carcinoma of the liver does not contain xan- thine oxidase, a point of interest in view of the fact that in the develop- ment of mammals the xanthine oxidase does not appear until late. Water extracts from various tumors have been found to contain small amounts of free purines, chiefly adenine, guanine and hypoxanthine (Drummond).''- Lipins. — Tumor cells seem to contain much the same fats and lip- oids as normal cells, and, as far as known, in much the same proportions as characterize the cells from which the tumors arose. Thus Wells^* found that hypernephromas show the same high proportions of lecithin and cholesterol as he found in normal adrenal, and as are found in the renal cortex. Other malignant tumors have much less lipoids and fats( see Hypernephromas). A secondary carcinoma of liver cells, metastatic in the skull, was found by Prym^" to show the same sort of fatty infiltration that is characteristic of fatty liver cells. On ac- count of the poor blood supply of many tumors, fatty changes are usual, occurring under the same conditions and showing the same microscopic features as fatty degeneration in other tissues,''*' being more common in malignant than in benign tumors; especially abund- ant in squamous cell carcinomas, and scanty in sarcomas. Crystals of cholesterol or cholesterol compounds are described in tumors by White. '*^ Dewey** found the chief lipoid in jaw tumors to be choles- terol, with more or less free fatty acids and soaps, according to mi- crochemical determinations. Even lipoma fat shows no difference from normal fat,"*^ and the depot fat of tumor patients is quite the same as in patients with other diseases associated with equal wasting, *° in whom some increase in unsaponifiable material (cholesterol) is usual. Murray^^ says that the lipoids of degenerating uterine fibroids are strongly hemolytic, which may account for the so-called ''red degen- eration" of these tumors. Freund and Kaminer^- suggest that the fatty acids of tissues are of importance in determining whether a tissue is a suitable soil for secondary growth, these substances being deficient in tissues where growths develop. Mitochondria, which seem to be closely related to the intracellular lipins, show no constant differences " Jour. Exper. Med., 1913 (18), 512. *"■ Jour. Med. Res., 1908 (17), 461. « Frankf. Zeit. Path., 1912 (10), 170. « See Haga, Berl. klin. Woch., 1912 (49), 342; Joannovics, Wien. klin. Woch. 1912 (25), 37. *" Jour. Path, and Bact., 1908 (13), 3. *8 Jour. Cancer Res., 1919 (4), 263. " See Wells, Arch. Int. Med., 1912 (10), 297. *» Wacker, Zeit. physiol. Chem., 1912 (78), 349; 1912 (80), 383. °i Jour. Obst. Gyn. Brit. Emp., 1910 (17), 534. " Wien. klin. Woch., 1912 (25), 1698. 504 THE CHEMISTRY OF TUMORS in cancer, benign tumors and normal cells, except that sometimes in cancer they fix stains less firmly (^Goodpasture). ^^ There has been some effort to correlate the cholesterol and lecithin contents of blood and tissues with the rate of cancer growth; apparent- ly lecithin inhibits growth and cholesterol stimulates.^'' However, Bullock and Cramer" found much more cholesterol in a slowly growing mouse carcinoma than in a rapidly growing one, somewhat more phos- phatid in the latter, much more phosphatid in a sarcoma than in the carcinoma, and cerebrosides only in the latter; in necrotic portions of tumors they found an increase in simple fats. These figures are based on too few observations to be interpreted as yet. (3) Inorganic Constituents. — These have been studied by Clowes and Frisbie^*^ under exceptionably favorable conditions, in that the age of the tumor could be accurately estimated, in the inoculable carcinoma of mice. They found that rapidly growing tumors contain a high percentage of potassium and little or no calcium, whereas in old, slowly growing, relatively necrobiotic tumors the relation is reversed, the potassium decreasing greatly while the calcium increases. Mag- nesium is present only in traces, while the proportion of sodium fluc- tuates much less, but is usually greater than either the potassium or calcium, although in very old tumors the latter may become excessive. The most rapid growth, however, seems to occur in tumors in which both calcium and potassium are present in the ratio of K 2 3 Ca = l°^'2 Beebe^^ analyzed a number of human tumors with the following results: Phosphorus was found in proportion to the amount of nu- clear material, varying from 0.139 per cent, (uterine fibroid) to 1.06 per cent, (sarcoma). Iron varied from 0.013 per cent, to 0.064 per cent., probably depending on the amount of blood and nucleoproteins. Calcium is most abundant in old degenerated tumors, and potassium in rapidly growing tumors. These results, supported by Clowes and Frisbie's findings, indicate the importance of potassium for cell growth. Injection of potassium salts into mice increases their susceptibility to inoculation (Clowes),^^ while calcium decreases cancer growth (^Goldzieher).^^ Exposure of isolated cancer cells to calcium salts reduces their growth capacity when inoculated, apparently through reducing their water content; both effects arc counteracted by sodium " Jour. Med. Res., 1918 (38), 213. " See Robertson and Burnett, Jour. Exp. Med., 1913 (17), 3-14; 1016 (23), 631; Sweet et al, Jour. Biol. Chem., 1915 (21,, 309; Luden, Jour. Lab. Clin. Med., vols. 3 and 4. ^f- Proc. Royal Soc. London (B), 1914 (87), 230. •■sAmer. Jour. Phvsiol., 190.5 (14), 173. " Amor. Jour. Pliysiol., 1904 (12), 167. " British Med. Jour., Doc. 1, 1906. 6" Vorhandl. Dout. Path. Gesellsch., 1912 (15), 283. I ENZYMES OF TUMORS 505 (Cramer).*" A greater proportion of potassium was found in primary than in secondary growths by Mottram;''' sodium was the same in each; there is more potassium in squamous cell carcinoma than in round cell sarcoma. Robin*^'- states that in cancerous livers the cancer tissues contains more inorganic matter than the normal liver tissue about it. Cattley*' found the microchemic distribution of potassium the same in cancer as in normal cells, and the same seems to be true of manganese.*^ Schwalbe''^ found that cancer-cells contain iron in a condition de- monstrable by the Berlin-blue reaction, and occurring independent of hemorrhages. Tracy^® found that tumors reacted microscopically for iron, either free or in the form of an albimiinate, only in areas where hemorrhages had occurred. Nuclear or organic iron could be detected in the nuclei, occurring in a network arrangement. In other words, iron occurs in tumors, both quantitatively and qualitatively, exactly as in normal cells of the same type. The same writer" found in tumors by microchemical reactions, that phosphorus in the form of nucleo- proteins likewise shows no essential differences from its distribution in normal tissues. In this connection may be mentioned the observations of Hem- meter,*** who found that the cells of carcinoma of the mammary gland will shrink when placed in physiological salt solution or in the serum of the patient, whereas normal cells swell when placed in cancer-juice. This suggests that the osmotic pressure, and, by inference, the amount of inorganic constituents, is lower than in normal tissues. Crystal- loids, such as KI, diffuse readily into cancer tissue.*^ (4) Enzymes. — The rapid and extensive autolysis that occurs in tumors, as shown both morphologically and by the presence of the products of protein cleavage in them, indicates that tumor cells resemble all other cells in possessing intracellular proteolytic enzj^mes. Because of autolysis, puncture fluids in cancer of serous surfaces show an increased amount of incoagulable nitrogen (J\Iorris),^° and they may show free amino-acids ( Wiener), ^^ while there is a slight increase in the incoagulable nitrogen of the blood (Takemura).^- There is considerable but not undisputed evidence that cancer tis- s" Biochem. Jour. 1918 (12). 210. " Arch. Middlesex Hospital, 1910 (19), 40. «2 Compt. Rend. Acad. Sci., 1913 (156), 334. ^^ Lancet 1907 (172) 13. «^ Medigreceanu, Pro'c. Royal Soc, B, 1912 (86), 174. " Cent. f. Path., 1901 (12), 874. 66 Jour. Med. Research, 1905 (14), 1. " Martha Tracy, Jour. Med. Research, 1906 (14), 447. 68 Amer. Jour. Med. Sci., 1903 (125), 680. 69 Van den Velden, Biochem. Zeit., 1908 (9), 54; see also Wells and Heden- burg. Jour. Infect. Dis., 1912 (11), 349. '"Arch. Int. Med., 1911 (8), 457. '1 Biochem. Zeit., 1912 (41), 149. " Ibid., 1910 (25), 505. 506 THE CHEMISTRY OF TUMORS sue autolyzes somewhat more rapidly than corresponding normal tis- sues," and, according to Neuberg, Blumenthal and others,^'* that cancer extracts digest other tissues than themselves (heterolysis), a property not exhibited by extracts of normal tissues. Miiller and others would ascribe this heterolysis to the leucocytes present in the tumors. Nu- cleases have been found in tumors as in other tissues," and in general the enzymes which deamidize adenine and guanine (adenase and gua- nase) are usually present if the original tissue possessed these enzymes but no instance of the presence of xanthine oxidase or uricolytic en- zyme has been obtained (Wells and Long, loc cit^'^). Hamburger finds that the enzymes of cancer tissue upon which the glycyl-tryptophane and other enzyme tests for cancer are based, are ereptases, resembling in all their properties the ereptases of normal tissues, and not present in particularly large amount. However, Ab- derhalden^^ has found evidence that certain peptids may be split in a different way by cancer than by normal tissues, supporting those who hold that cancer enzymes are different from normal tissue enzymes. Autolysis of tumors is said to be augmented by x-ray, and especially by radium (Neuberg), and tumor tissue is readily digested by trypsin. The presence of ereptases in carcinomatous gastric juice has been especially studied because of its diagnostic possibilities, and the care- ful investigation of Jacques and Woodyatt''^ seems to show conclu- sively that such an enzyme is rarely present in gastric juice except when derived from a cancer present in the wall of the stomach, pro- vided peptolytic bacteria are excluded by filtration. Deaminizing enzymes may also be found in gastric cancer secretions.''^ In the blood of cancer patients there is usually an increased antitryptic activity, ascribable to the reaction against enzymes absorbed from the cancer; it is less pronounced with sarcoma.''^ The body tissues of patients dying with cancer show a low ereptic activity, but the same occurs in persons dying from other wasting diseases (Col well). ^^ This is also true of other tissue enzymes; — at least purine oxidizing enzymes are deficient in the liver tissue between secondary cancers (Wells and Long''-) and the catalase is also reduced in liver tumors (Blumenthal and Brahn)^^ and in the blood of tumor mice (Rosenthal);'^- in human blood the catalase may vary either side of normal.*^* Brahn'^'* reports '« See Yoshimoto, Biochem. Zeit., 1909 (22), 299; Daels and Delenz6, Bull. Acad. Med. Belg., 1913 (26), 833. '^ Bibliography by Hamburger, Jour. Amer. Med. Assoc, 1912 (59), 847. '5 Goodman, Jour. Exp. Med., 1912 (15), 477. '« Zeit. Krebsforsch., 1910 (9), 266. " Arch. Int. Med., 1912 (10), 560. '» Halpern, Mitt. Grenz. Med. Chir., 1915 (28), 709. '» Citronblatt, Med. Klin., 1912 (8), 138. 80 Arch. Middlesex Hosp., 1909 (15), 96. *' Zeit. f. Krebsforsch., 1910 (8), 436. See also Weidenfeld, Wien. klin. Woc-h., 1918 (31), 324. 82 Deut. med. Woch., 1912 (38), 2270. 8' Rohdenburg, N. Y. Med. Jour., 1913 (97), 824. »*Zeit. Krebsforsch., 1917 (16), 112. INTERNAL SECRETION OF TUMORS 507 that the hvcr tissue between sccoiuhiry cancer nodules, and also the liver in cases of cancer in the portal area, shows diminished catalase, lipase and lecithinase function, with increased autolysis, but these changes are not observed in the livers when the cancer is in other parts of the body. However, chol'ne has been found in necrotic sarcomas of rats,^^ whicli would seem to indicate the presence of enzj'ines dis- integrating lecithin. As mentioned elsewhere (See Melanin), me- lanotic tumors may contain enzymes oxidizing tyrosine, epinephrine, pyrocatechin, or other related aromatic substances, with the forma- tion of pigmentary substances. (See also, Autolysis in Tumors, chap, iii.) Other enzymes are also present in tumor cells. Buxton^*^ exam- ined a large number of tumors for their enzymes by the plate {auxan- ographic) method, and found considerable variations in different growths. All contained amylase (splitting starch) and lipase (split- ting butyrin). Most, but not all, tumors coagulated milk and liquefied casein, and also liquefied gelatin (rennin, proteases). Peroxidase was nearly always, and catalase always, present. Digestion of fibrin, co- agulated serum, and coagulated egg albumen could not be observed. Practically all tumors split glycogen. Tyrosinase could not be demon- strated. The fact that early embryonic tissues were found poor in enzymes*^ speaks against the common assumption that tumors repre- sent strictly an embryonic formation, but Long^^ found that xanthine- oxidase, which in normal development does not appear until late in fetal life, was absent from primary carcinomas of sheep livers, al- though normal adult sheep liver tissue is rich in this enzyme. MacFadyen and Harden^^ studied the juices obtained by grinding up tumor cells made brittle by liquid air, and found by direct methods (chiefly in breast cancers) invertase, maltase, amylase, proteases acting in both acid and alkaline solutions, catalase, oxidase, with per- haps traces of lipase and peroxidase, but no lactase. Tumors arising from the gastric mucosa, according to Waring,^" contain both pepsin and rennin; those from the pancreas, both pri- mary and secondary growths, contain trypsin, steapsin, amylase, and rennin. (5) Internal Secretion. — If tumors are derived from an organ with an important internal secretion, the tumor cells in many cases produce the same internal secretion, which seems to have the same functional properties as the normally produced secretion. Thus a metastatic growth from a thyroid tumor has been said to functionate in place of the resected gland; Gierke^^ found in about 20 grams of 85 Ellinger, Munch, med. Woch., 1914 (61), 2336. 8« Jour. Med. Research, 1903 (9), 356. " Ibid., 1905 (13), 543. 88 Jour. Exper. Med., 1913 (18), 512. 8' Lancet, 1903 (ii), 224. ^o Jour. Anat. and Physiol., 1894 (28), 142. " Hofmeister's Beitr., 1902 (3), 286. 508 THE CHEMISTRY OF TUMORS material from metastatic thyroid tissue in the vertebral column about 5 mg. of iodin, which was a trifle larger proportion than was present in the thyroid itself. Carlson and WoelfeP^ found much iodin in the metastases of a thyroid carcinoma of a dog, while in another dog whose cancerous thyroid contained no iodin the secondary tumors were also devoid of this element. I have also analyzed metastases from a carcinoma of the thyroid which contained no demonstrable iodin, despite the presence of colloid. Marine and Johnson^^ found that in two cases of cancer of the thyroid in man, and one in the dog, the cancer tissue showed no ability to retain iodin given by mouth, in contrast to normal thyroid and simple adenomas. Meyer-Hiirlimann and Oswald^'* have described a remarkable case of cystic carcinoma of the thyroid, from which in six weeks 2840 c.c. of secretion was obtained by puncture. It contained 0.077 mg. iodin per 10 c.c. (the patient having previously been given Kl) as compared with normal thyroid 'which contains 0.4 to 4 mg. per 10 gm. It contained both globuhn and albumin, the former corresponding to true thyroglobulin, even to in- creasing vagus irritability experimentally. The "adenomatous" nodules of the thyroid often show evidence of active secretion, Goetsch^^ having found their cells rich in mitochondria, while Graham'^*^ found that they take up iodin and metabolize it so that the adenomatous tissue produces the typical thyroid effect on the development of tad- poles. Adrenal cancers do not usually cause Addison's disease, per- haps because they functionate in the place of the destroyed gland (Lubarsch). In the characteristic production of cachexia, often apparently out of all proportion to the amount of tumor tissue, there would seem to be evidence that a peculiar and abnormal product of metabohsm is formed by cancer-cells, and extracts from cancers have been found toxic for protozoa." As yet, however, it has been impossible to demon- strate any characteristic toxic substance in cancers.^** Girard- Mangin^^ claims that malignant tumors contain colloidal poisonous substances in proportion to their softness, extracts causing paralysis and fall of blood pressure; but others have failed to substantiate this.^ Because of the constant disintegration of the tumor tissues, products of autolysis are formed, and undoubtedly enter the circulation in small quantities; possibly they are a factor in the systemic mani- festations of malignant growths, analogous to the action of cleavage " Amer. Jour. Physiol, 1910 (26), 32. "3 Arch. Int. Med., 1913 (11), 288. »* Korr.-Bl. Schweizer Aerzte, 1913 (43), 1468. »» Bull. Johns Hopkins Hosp., 1916 (27), 129. " Jour. Exp. Med., 1916 (24), 345. "Woodruff and Underhill, Jour. Biol. Chcin., 1913 (15), 401; Calkins. Jour. Cancer Res., 1916 (1), 205 and 399. "* See Blumenthal, Fcstschr. f. Salkowski, Berlin, 1904; Ilanseinann, Zrit. Krebsforsch., 1906 (4), 565. »» Presse M6d., 1906, p. 17)9; Compt. Rend. Soc. Biol., 1909 (67), 117. 1 See Bruschettini and Barlocco, Cent. f. Bakt., 1907 (43), 664. I HEMOLYSIS IN CANCER 509 products of foreign proteins which may produce "protein fever" and other toxic effects. It has often been observed that when ex- tensive necrosis is produced in experimental tumors in rats and mice the animals may show profound toxemia, presumably because of absorption of autolytic products. Since all normal tissue-cells produce substances through their me- tabolism that enter the circulation, it is quite certain that tumor-cells do likewise, and it is highly probable that the presence of abnormal quantities of such products, even if they are of quite normal compo- sition, may cause disturbances in the body. As yet, however, no such substances, either normal or abnormal, have been isolated, nor has their presence been demonstrated. Numerous isolated observations of ptomains or similar substances in the urine of cancer patients may be found in the literature,- but their importance is extremely question- able. A large proportion of cases of malignant tumors exhibit renal injury (Kast and Killian)^ but whether from products of the tumor o? from bacterial infection has not been determined. Hemolytic Substances. — -A number of observers have described the finding of hemolytic substances in cancer extracts. Bard^ observed that in hemorrhagic carcinomatous exudates in serous cavities the blood is rapidly hemolyzed, which is not the case in exudates from other causes, but this was not corroborated by Weil.^ Kullmann^ found that extracts of carcinomas contain hemolytic substances acting energetically both in the body and in vitro; these are soluble in alcohol and in water, are not complex in composition, are not specific for hu- man corpuscles, but are toxic for all varieties of corpuscles. Micheli and Donati*"' likewise found hemolytic substances in 8 of 15 tumors, of which 5 acted on all varieties of corpuscles, and 3 acted on only certain varieties; they regard the hemolytic substances as the products of autolj^sis in the tumors. WeiF also found the hemolytic property of tumor extracts to vary with the amount of necrosis, and to depend on dialyzable hemolytic substances distinct from the hemolysins of normal tissues. It is well known that among the products of autolysis of normal tissues are hemolytic substances. Whether the severe anemia frequently present in carcinoma is due, either largely or in part, to these products of autolj^sis is unknown, but it is very probable that they have some effect. Hemolysis in Cancer.^ — The blood serum of cancer patients has often a hemolytic action on the corpuscles of normal persons (Crile), but this property is quite inconstant, being present in 67 per cent, of a series of 472 cancer cases collected by Krida, while 15 per cent, of 2 See KuUmann, Zeit. klin. Med., 1904 (53), 293. 3 Proc. See. Exp. Biol. Med., 1919 (16), 141. * La Semaine M6d., 1901 (21), 201. 5 Jour. Med. Res., 1910 (23), 86. 6 Riforma Med., 1903 (19), 1037. ^ Jour. Med. Res., 1907 (16), 287. 510 THE CHEMISTRY OF TUMORS cases of other diseases and 2.6 per cent, of normal persons showed hemolytic activity of the serum. ^ Elsberg found that normal corpus- cles injected subcutaneously into cancer patients are hemolyzed, but Gorham and Lisser found this reaction positive in but 60 per cent, of their cases, the subcutaneous hemolysis not corresponding at all to the hemolytic activity of the patient's serum in the test tube. The stomach contents in cancer of the stomach, when ulcerated, are hemo- lytic (Grafe and Rohmer).^ The red corpuscles of cancer patients are said to have usually a greater resistance to hemolysis by cobra venom than normal corpuscles, but this is not characteristic, there being simi- lar alterations in other diseases.^*' The reputed power of the serum in cancer to protect corpuscles from hemolysis by oleic and lactic acid could not be demonstrated by Sweek and Fleisher.^^ An extensive review of the literature and methods led Cohnreich^^ to the conclusion that resistance of erj^throcytes to hypotonic solutions and to poisons vary independently of one another. He has devised an improved method for testing resistance to hypotonic solutions, which seems to vary directly with the amount of stroma and PO4 content, and finds that determinations of maximum and minimum resistance are of little value, as *"hese concern only a small part of the corpuscles; he therefore determines the "plurimum" resistance, involving most of the corpuscles. The most significant results were obtained in cancer of the alimentary tract, in which an increased resistance was always demonstrable. Farmachidis'^ finds the cobra venom resistance more specific for cancer than do most other investigators. (6) Metabolism in Cancer. — There are numerous observations indicating that cancer cachexia is in no way different from the cachexia of other conditions. The behavior of the nitrogen metabolism seems to be quite the same as in tuberculosis and other wasting diseases. There is the same excessive elimination of aromatic substances (phenol, ^* indican) and oxyacids (Lewin,^^ BlumenthaP^), which Lewin con- siders to arise from the abnormal metabolism of proteins, and not from putrefactive decomposition in the tumor or in the intestines. In rats with sarcoma, increased excretion of uric acid and creatin has been ob- served. ^^ There is also the same excessive elimination of mineral salts that is observed in pulmonary tuberculosis, and termed "demin- * Literature bv Gorham and Lisser, Amer. Jour. Med. Sci., 1912 (144), 103. 9 Deut. Arch.'klin. Med., 1908 (94), 239. •" Kraus, Ranzi and H. Ehrlich, Sitz. Ber. Akad. Wien., 1910 (119), 3; see also Grunbaum, Jour. Path, and Bact., 1912 (17), 82. " Jour. Med. Res., 1913 (27), 383. " FoHa Hematol., 1913 (16), 307, full bibliography. " Gaz. degli Osped., 1915 (36), 689. ^* Somewhat higher than average figures for phenol in the blood were found in sarcoma cases by Theis and Benedict (.lour. Biol. Chcm., 1918 (36), 99). 1^ Deut. med. Woch., 1905 (31), 218. i« Festsclir. f. Salkowski, Berlin, 1904. " Ordway, Jour. Med. Res., 1913 (23), 301. METABOLISM IN CANCER 511 eralization" b}' Robin,''* but no alteration in the excretion of chlo- rides.'^ As in other cachexias, the creatin content of the muscles is decreased.'" FraenkeP' finds evidence that there may be some diffi- culty in tryptoj)hano inc^tabolisni in tumors and in tumor patients, especially marked with melanotic tumors. Extensive respiratory studies by Wallersteiner-^ showed enormous variations in the amount of heat production in different cases, in about 10 per cent, of which figures as hip;h as those of severe fevers or exophthalmic goiter were obtained repeatedly; most of the cases showed high normal figures. Nitrogen loss did not ordinarily occur if the calorimetric findings were considered in the calculations; nitrogen equilibrium was maintained if sufficient nourishment was obtained and utilized. In general, metabolism in cancer resembles that of fever, and warrants the assumption of a toxic stimulation of tissue destruction. It is entirely possible that the pro- ducts of cancer protein destruction are responsible for this toxicogenic metabolic abnormality, since Vaughan has demonstrated that the effects of bacteria and foreign proteins are quite the same in their pyretic and toxic action. Salkowski demonstrated that the amount of colloidal nitrogenous material, precipitated from the urine by strong alcohol, is increased in cancer. Numerous observers have corroborated this, but find that a similar conchtion obtains in other cachectic diseases, although in cancer the amount of colloidal nitrogen seldom is as low as normal unless the tumor is removed. ^^ Much of this colloidal nitrogen seems to be in the form of "oxy-proteic acid" (Salomen and Saxl),-'* which is a mixture of incompletely oxidized polypeptids, containing much unoxidized sul- phur." The proportion of neutral sulphur in the total sulphur in the urine seems to be increased in cancer (Weiss), but not so constantly or characteristically as to be of great diagnostic value." Much clinical investigation has been made of these urinary changes, which has gen- erally substantiated the fact that there usually is more increase in colloidal nitrogen and ethereal sulphate in the urine of cancer than in other diseases, but that in no sense are these changes specific for cancer, and the fundamental metabolic disturbances responsible have ^^ Quoted by Lewin, he. cit.^^ Clowes et al. (5th Ann. Rep., X. Y. State Dept. of Health, 1903-4) report observing a slight chloride retention in cancer patients, and review the literature of metabolism in cancer. '8 Robin, Compt. Rend. Acad. Sci., 1913 (156), 1262. 20 Chisholm, Biochem. Jour., 1912 (6), 243. =' Wien. klin. Woch., 1912 (25), 1041. 22 Deut. Arch. klin. Med., 1914 (116), 145. 23 See Mancini, Deut. Arch. klin. Med., 1911 (103), 288; Semenow, Foha Urol., 1912 (7), 215; de Bloeme et al, Biochem. Zeit., 1914 (65), 345. 2* Wien. klin. Woch., 1911 (24), 449. 25 Killian reports finding in the blood two to three times the normal amount of nonprotein sulphur while the total sulphates remain normal. ("Cancer: Its Nature, Causes, Diagnosis and Treatment." By R. H. Greene, New York, 1918.) 2« Stadtmijller and Rosenbloom, Arch. Int. Med., 1913 (12), 276; Interstate Med. Jour., 1916 (23), No. 2; bibliographv. Kahn, Jour. Cancer Res., 1917 (2), 379. 512 THE CHEMISTRY OF TUMORS not been ascertained.^^ They seem more indicative of the excessive catabohsm of cachexia than of cancer tissue itself. SaxP* has as- cribed part of the increased sulphur elimination to abnormal excre- tion of sulphocyanid, and as small doses of sulphocyanides lead to increased oxyproteic acid in the urine he suggests that in cancer there is a specific disturbance in sulphocyanid metabolism, an hypothesis that awaits confirmation. Of similar status is the excessive excretion of glycuronic acid described by Roger. ^^ Israel, and also Engelmann, have reported the occurrence of a marked increase in the lowering of the freezing-point of the blood in carcinoma (as low as —0.60° to —0.63°, the normal being —0.56°), which they attributed to the presence of excessive products of protein decomposition in the blood. Engel,^" however, found no such in- creased lowering of the freezing-point in his cases, and questions the significance of the results of Israel and Engelmann. There may be a dietary increase in the blood sugar in cancer, ^^ which rises more rapidly and remains high longer than normal.^- The total protein of the blood is low, with some increase in the proportion of globulin as is usual in cachexia. ^^ According to Moore and Wilson^'' the acid-neutralizing power of the blood ("alkalinity") is increased in cancer; this is prob- ably related to if not the cause of the decreased HCl content of the gastric juice, which occurs whether the cancer is in the stomach or not. As this alkalinity is not associated with an increase in the inorganic bases of the blood, it may be that the proteins have an increased basicity. Although numerous other observers describe a decreased alkalinity as in other cachectic conditions, ^^ Menten,^^ making direct H-ion measurements, found an increase in alkalinity in the serum of nearly all cases of carcinoma and sarcoma. The blood in cancer contains less calcium than normal which results in a tendency to osteoporosis" and to deposition of calcium in the kidney epithelium;^'* there is an increase in the potassium of both the blood and tissues. ^^ Blood analyses in 189 cases of cancer, by Theis and Stone, ^^ gave usu- ally low figures for non-protein and urea nitrogen, but with amino-N " See Goodridge and Kahn, Biochem. Bull., 1915 (4), 118; Damask, Wien. klin. Woch., 1915 (28), 499; Sassa, Biochem. Zeit., 1914 (64), 195. 28 Biochem. Zeit., 1913 (55), 224. =»Bull. Hoc. Med. Hop., Paris, 1915 (31), 499. »" Berl. klin. Woch., 1904 (41), 828. 51 Williams and Humphreys, Arch. Int. Med., 1919 (23), 537. '- Rohdenburg, Bernard and Krehbiel, Jour. Amcr. Med. Assoc, 1919 (72), 1528. '' Loebner, Deut. Arch. klin. Med., 1918 (127), 397. 5' Biochem. Jour., 1906 (1), 297: Watson, Jour. Path, and Bact., 1909 (13), 429; Sturrock, Brit. Med. Jour., 1913 (2), 780. »^ See Traube, Int. Zeit. Physik.-Chem. Biol., 1914 (1), 389. 3" Jour. Cancer Res., 1917 (2), 179. " Goldzieher, Verb. Deut. Path. Ges., 1912 (15), 283. "8 M. B. Schmidt., Verb. Deut. Path. Ges., 1913 (16), 329. " Mottram, .\rch. Middlesex Hosp., 1910 (19), 40. *" Jour. Cancer Res., 1919 (4), 349. DIET AND TUMOR GROWTH 513 slightly above normal; uric acid and sugar were within normal limits. Cholesterol, fatty acids and total fats are generally increased in the blood in malignancy.'"" (7) Diet and Tumor Growth. — In general, any condition that decreases the nutrition of tlic body as a whole, or of the tissue in which a tumor is located, decreases the rate of growth of the tumor, in which respect neoplasms exhibit quite the opposite behavior to infectious processes. Thus, the older the individual the more slowly the tumor usually grows; ligation of the lingual artery retards the growth of cancer of the tongue; repeated pregnancy and lactation delay the progress of cancer in mice,"*^ suggesting that tumor cells have a greater avidity for nutritive elements in the blood than have ordinary somatic cells, but less than the cells of the fetus or of the active mammary gland. Numerous attempts have been made to determine the relation of tumor growth to specific dietary deficiencies. Sweet, Corson- White and Saxon"*- found that rats kept upon a diet deficient in specific amino- acids (lysine), so that body growth did not occur although nutrition was maintained, show a slower growth of implanted tumors than ani- mals on an adequate diet. Rous"*^ obtained similar results with some transplanted tumors, but not with all, nor with spontaneous tumors. Van Alstyne and Beebe*'* found that rats living on casein and lard showed much less growth of inoculated tumors than when lactose was added to the diet. Robertson and Burnett"*^ observed that the addi- tion of cholesterol to the diet increases the rate of growth and the de- velopment of metastases in inoculated rat tumors, which has been corroborated by others. This accelerative action depends on the hy- droxyl radical, although other hydroxy-benzol derivatives do not have this effect."*" The growth-promoting principle of the hypophysis, tetheUn, is also said to stimulate cancer growth. Funk"*^ found greater growth of inoculated sarcoma in fowls given normal diets than in those fed pohshed rice. Benedict and Rahe'*^ supplied vitamines by adding to an otherwise inadequate diet, just enough yeast to keep the rats in fair condition, and found that inoculated tumors grew, although extremely slowly, even when the animal itself could not grow. Evidently tumor cells cannot manufacture substances essential for growth i. e. vitamines. Corson-White*^ states that, generally, vitamine-rich diets favor tumor growth, especially if there is also an abundance of cholesterol. Fraenkel,^° however, observed no stimulat- ing effect from rice polishings or yeast extracts. "» Ee Niord et al., Arch. Int. Med., 1920 (25), 32. "^ See Maud Slve, Jour. Cancer Res., 1919 (4), 2.5. '- Jour. Biol. Chem., 191.3 (15), LSI ; 1915 (21), 311. « Jour. Exp. Med., 1914 (20), 433. " Jour. Med. Res., 1913 (24), 217. ** Jour. Exp. Med., 1913 (17), 344. ^« Jour. Cancer Res., 1918 (3), 75. *' Zeit. phvsiol. Chem., 1913 (88), 352. ^8 Jour. Cancer Res., 1917 (2). 1.59. " Penn. Med. Jour., 1919 (22), 348. *" Wien. klin. Woch., 1916 (29), 483. 33 514 THE CHEMISTRY OF TUMORS (8) Immunity Reactions in Cancer. — The fact that a certain degree of specific immunity can be developed against normal tissue cells (see Cytotoxins, Chap, x), has encouraged study of the possibility of se- curing immune antibodies which might be specific for cancer, and has led to much research on this subject, ^^ with results as yet of httle value. There is no doubt that the bodj^ has distinct powers to inhibit to a greater or less degree the growth of tumors, and to destroy many of the cells which escape from cancers into the lymph and blood, ^2 while in experimental animals inoculated tumors are in most instances unable to grow, and they may, after growth has once begun, recede or even disappear. Furthermore, animals may be made immune to tumors to which they would otherwise be susceptible . Many schemes of immunization of patients by injection of extracts or autolysates made from their own tumors, or similar tumors of others, have been tried, ^^ but in the hands of competent and critical observers the results seem to have been practically negative. ^^ It is not always kept in mind that inoculated cancers in rats and mice represent an artificial condition behaving very differently from spontaneous tumors. There is no lack of evidence that cancers do produce, in greater or less amounts, various antibodies of some degree of specificit}' for can- cer, which must be interpreted as evidence that cancer proteins are in some respects different from the normal proteins of the host; however, the amount and specificity of these antiboches seem to be low," and, in many observations, they have failed to be demonstrated. In- deed, Coca in his review states unqualifiedlj^, "The usual biological tests of complement deviation and specific precipitation fail to show the hypothetical antibodies, though a distinct cytotoxic influence can be demonstrated in the plasma of animals of foreign species that have been actively immunized against a tumor." His own experiments failed to demonstrate specific complement-fixation antibodies in patients injected with extracts of their own tumors. Lewin^'^ also fails to find conclusive evidence of the demonstration of specific anti- bodies in cancer, yet accepts the immunity which is produced by in- jections of virulent cancer material as an active immunity dependent upon cancer antibodies. It may, however, depend on a stimulation of the local cellular reactions that inhibits cancer growth. ^^ Pfeiffer^'* claims to find specific anaphylactic antibodies in the blood of cancer patients, but this has not been confirmed by several other observers. ^'■' " Literature by Coca, Zeit. Immunitat., 1912 (13), 525; Kraus et al., Wien klin. Woch., 1911 (24), 1003. " Reviewed by Wells, Jour. Amer. Med. Assoc, 1909 (52), 1731. " Review by Fichera, Jour. Cancer Res., 1918 (3), 303. "See Blumenthal, Zeit. Krebsforsch., 1914 (14), 491; Bauer, Latzel and Wessely, Zeit. klin. Med., 1915 (81), 420. " See Morgenroth and Bieling, Biochem. Zeit., 1915 (68), 85. ^« Folia Serologica, 1911 (7), 1013; literature. " Tyzzer, Jour. Cancer. Res., 1910 (1), 125. " Wien klin. Woch., 1909 (22), 989; Zeit. Immunitiit., 1910 (4), 455. " See Weil, Jour. Exp. Med., Oct., 1913, (18), 390. IMMUNITY REACTIONS IN CANCER 515 V. Dungern^" has described positive complement fixation reactions, partially specific for cancer and benign tumors, by using alcoholic extracts of the tumors or acetone extracts of human erythrocytes as antigen, but he interprets these reactions as not due to specific anti- bodies, but to abnormal products of metabolism.''^ The complement content of the blood is said to be slightly increased in cancer (Engel),^- but there is nothing characteristic about this. Ascoli and Izar*^ have applied the mciostagmin test {q. v.) and state that this gives very positive results in determining the existence of cancer, their work having been corroborated by many but not by all of those who have repeated it.^^ Burmeister'"'^ could obtain no reliable results with the epiphanin reaction. Freund and Kaminer'"' have found that the serum of cancer pa- tients is unable to dissolve cancer cells, as normal serum does, and even protects them against the lytic power of normal serum. The lysis is ascribed to a non-nitrogenous fatty acid, while the protective agent of cancer serum is said to be a "nucleo-globulin" which is in- creased in the serum in cancer. They also find that cancer extracts give a specific turbidity or precipitation with cancer serum, which is attributed to a carbohydrate content of the extract. According to Kraus and v. Graff" the serum of full term, pregnant women, and normal umbilical cord serum, behave like serum from cancer patients. In support of Freund and Kaminer's observation is the experiment of Neuberg^^ who found that cancer cells plus normal serum underwent digestion more rapidly than cancer cells plus cancer serum, as measured by the incoagulable nitrogen. A critical test of many recommended methods of serum diagnosis of cancer by Hal- pern^^ gave disappointing results. With the von Dungern technic he obtained 80 per cent, of positive results, with the meiostagmin reaction 85 per cent., but with the Abderhalden method but 30 per cent. The other methods he finds of little value. The testimony concerning the specificity of the Abderhalden reaction (q. v.) in cancer is so conflicting that it seems unprofitable to discuss it, the results varying from such «» Miinch. med. Woch., 1912 (59), 65,1093 and 285-4; also Rosenberg, Deut.med. Woch., 1912 (38), 1225. ^^ Farmachidis (Riforma med., 1918 (34), 382) states that onh^ with maUgnant disease occurs the activation by cobra venom of the hemolytic action of the serum in the complement fixation reaction. «- Deut. med. Woch., 1910 (36), 986. Not corroborated by Ordway and Kel- lert, Jour. Med. Research, 1913 (28), 287. " Munch, med. Woch., 1910 (57), 2129; Biochem. Zeit., 1910 (29), 13. " See Rosenberg, Deut. med. Woch., 1913 (39), 926; Wissung, Berl. klin. Woch., 1915 (52), 998. Roffo, Revista Inst. Bact., Buenos Aires, 1917 (1), 53. " See Burmeister, Jour. Inf. Dis., 1913 (12), 459; Bruggemann, Mitt. Grenz. :Med. u. Chir., 1913 (25), 877. «« Biochem. Zeit., 1912 (46), 470; Wicn. klin. Woch., 1911 (24), 1759; 1913 (26), 2108. " Wien. klin. Woch., 1911 (24), 191. «s Biochem. Zeit., 1910 (26), 344. " Mitt. Grenz. Med. Chir., 1913 (27). 370. See also Mioni, Tum6ri, 1914 (3), 697. 516 THE CHEMISTRY OF TUMORS as those cited by Halpern above, to 100 per cent, correct reactions de- scribed by others. '''' Coca'^^ obtained entirely unsatisfactory results with both the von Dungern complement fixation test and the Freund- Kaminer reaction. Many observations have been made on the antitryptic activity of the blood in cancer (see Chap, iii) which has usually shown an increase (in all but about 10 per cent, of the cases) ; but many other conditions, especially cachexia, may cause positive reactions. Cancer serum is said to have a lessened power to activate pancreatic lipase'^ when the disease is progressive, but on improvement or recovery this effect is increased. B. CHEMISTRY OF CERTAIN SPECIFIC TUMORS In the literature are to be found a few studies of chemical features of some forms of tumors, which may be briefly discussed to advantage. (1) Benign Tumors (a) Fibromas and Myomas. — The few specimens studied show but a small amount of nucleoprotein, as might be expected from the small amount of their nuclear material. Because of the tendency of fibromas to undergo retrogressive changes, the amount of calcium is likely to be large. No studies as to the special features of their col- lagen, as compared with normal connective-tissue collagen, seem to have been made. Lubarsch" found no glycogen (microscopically) in any of 66 fibromas he examined. Wells and Long^^* found that in uterine fibro-myomas but one per cent, of the total nitrogen is purine nitrogen, distributed as guanine, 44 per cent.; adenine, 31 per cent.; hypoxanthine, 25 per cent. The relatively large proportion of pre- formed hypoxanthine corresponds with the abundance of this purine free in normal unstriated muscle. Fibromyomas are able to deami- dize their guanine and adenine to xanthine and hypoxanthine, and con- tain guanase but not adenase. Extracts from uterine fibromyomas show practically the same composition as extracts of normal uterus.''* A uterine fibroid analyzed by Beebe^*^ contained 14.56 per cent, of nitrogen, 0.981 per cent, of sulphur, 0.139 per cent, of phosphorus, 0.013 per cent, of iron, 0.12 per cent, of calcium oxide, 0.44 per cent, of potassium, and 1.115 per cent, of sodium. The proportions of ni- trogen and sulphur are high as comi)ared with most tumors; the phosphorus, iron, and potassium are low, corresponding to the small amount of nucleoprotein and the slow rate of growth. If degeneration '" See de Crinis and Mahncrt, Ferine ntfrsch., 1918 (2), 103. ^1 Jour. Cancer Research, 1917 (2), 61. " Shaw-Mackenzie, Lancet, Nov. 8, 1919. " Virchow's Arch., 1906 (183), 188. 7* Zeit. Krebsforsch., 1913 (12), 59S. T"* Winiwarter, Arch. f. Gyniik., 1913 (100), 530. " Amer. Jour. Physiol., 1904 (12), 167. BENIGN TUMORS 517 is marked, the amount of calcium is greatly increased. Krawkow^^ found a trace of chondroitin-sulphuric acifl in a uterine fibroid. Lu- barsch found glycogen occasionally in richly cellular uterine leio- mj'omas, and in the vicinity of dcgonorating areas; however, 76 out of 85 showed no glycogen. Pfannensticl''* analyzed the alkaline fluid of a cystic fibromyoma, which coagulated spontaneously; it contained sugar, but no mucin or pseudomucin. The cysts were dilated lymph- spaces, and the fluid corresponded to lymph in composition. A similar result was obtained by Oerum,^'-* who found in the fluid scrum- albumin, serum-globuhn, and 0.358 per cent, of fibrin; the total pro- teins constituted 6.3056 per cent. Sollmann^" found in the "colloid" of a cystic degenerated fibromyoma both pseudomucin and paramucin (see "Ovarian Cysts"), which differed somewhat from the same sub- stances found in ovarian tumors. The, common occurrence of marked cardiac weakness in patients with uterine fibroids has led to the suggestion that in the fibroids some toxic product is formed which acts on the heart, or that both the fibroid and the heart defect might result from a common cause. The experi- mental evidence concerning the relationship is not convincing, and there is much ground for the belief that the heart suffers solely from the anemia common in these cases. ^"^ There is said to be a hemolytic poi- son, a lipoid according to Murray, ^^ formed in the degenerating fibroids which causes local hemolysis and "red degeneration," and there are cases of acute aseptic degeneration of fibromyomas which seem to have caused systemic intoxication. (6) Myxomas. — From a myxoma of the back Oswald*' obtained a mucin with the following elementary composition: C, 50.82; H, 7.27; N, 12.24; S, 1.19; P, 0.25 per cent. This differs from other mammalian mucins in the presence of phosphorus, but Oswald does not consider this a contamination. It also contained 12 per cent, of carbohydrate, apparently glucosamine. (c) Chondromas, like normal cartilage, always contain much glycogen (Lubarsch). Morner^^ found chondroitin-sulphm'ic acid in several chondromas that he examined, although Schmiedeberg had failed to do so. '^ Arch. exp. Path. u. Pharm., 1898 (40), 195. "Arch. f. Gyn., 1890 (38), 468. " Maly's Jahresber., 1884 (14), 462. 8« Amer. Gynecol., 1903 (2), 232. 81 See Jaschke, Mitt. Grenz. Med. u. Chir., 1912 (15), 249; ;McGUnn, Surg. Gyn., Obst., 1914 (18), 180. 82 Jour. Obs. and Gyn., 1910 (17), 534. " Zeit. physiol. Chem., 1914 (92), 144. s^Zeit. phvsiol. Chem., 1895 (20), 357. 518 THE CHEMISTRY OF TUMORS (d) Lipomas^^ have been studied by Schulz^*' and by Jaeckle.^^ The former found in a retroperitoneal hpoma 75.75 per cent, of fat, 2.25 per cent, of connective tissue, and 22 per cent, of water. Of the fat, 7.31 per cent, was in the form of the free fatty acids and 92.7 per cent, as neutral fats. The fatty acids of the fat consisted of 65.57 per cent, oleic acid; 29.84 per cent, stearic acid; 4.59 per cent, pal- mitic acid. Cholesterol was only qualitatively demonstrable. In the connective tissue was found chondroitin-sulphuric acid. Lubarsch found glycogen in lipomas only when they were degenerated. Jaeckle observed the formation of calcium soaps in a calcifying lipoma, the calcium being distributed as follows: calcium soaps, 29.5 per cent.; calcium carbonate, 28.61 per cent.; calcium phosphate, 41.89 per cent. The fats of lipomas he found practically identical with those of the subcutaneous tissues, except sometimes for a deficiency in lecitliin, as shown by the following figures: Composition op Fats in — Subcutane- ous tissue Lipoma I Lipoma II. Lipoma III Refraction, at 40° 50.6 197.3 0.25 63.7 74.1 70.9 0.39 0.196 18.5 6.2 0.084 0.32 50.1 197.7 0.33 59.0 68.6 65.7 0.31 0.155 24.9 5.1 50.9 197.7 0.35 64.0 74.4 71.2 0.48 0.24 '0.'34 50.5 Saponification number Reichert-Meisser number. . . . lodin number 195.9 0.35 64.1 Olein 74.5 Oleic acid 71.3 Acid number 0.67 Free acid 0.34 Palmitic acid 18.5 Stearic acid 5.9 Lecithin Cholesterol 0.015 Lipomas are able to hydrolyze fats and esters, their lipase behaving in all respects like the lipase of normal areolar tissue.**** Lipoma fat is hydrolyzed by lipase as readily as is normal human fat. No rea- son for the reputed unavailability of lipoma fat for the metabolism of the host could be found. It is doubtful if the fat of benign lipomas *^ In xanthoma tuberosum multiplex, which shows local deposits composed largely of cholesterol esters and contains also pigment with the properties of a lipochrome, the presence of hyper-cholesterolemia is disputed. (Roscnbloom, Arch. Int. Med., 1913 (12) 395; Schmidt, Dermatol. Zeit., 1914 (21), i:37). Edsall found the composition of the fat in the fatty tumors of adiposis dolorosa (Dercum's disease) but little different from that of normal fat. (Cjuotod by Dercum and McCarthy, Amer. Jour. Med. Sci., 1902 (124), 994). Martelli (Tumori, 1918 (6), 1) on histological grounds states that in 2-3 per cent, of the cells the fats are mi.xed with fatty acids, while cholesterol, phospholijiins and chromolijjins are very scanty; he attributes the condition to disturbed lipogenesis from enducrin-sympathetic disfunction. 8" Pfluger's Arch., 1893 (55), 231. 8^ Zeit. physiol. Chem., 1902 (36), 53. 88 Wells, Arch. Int. Med., 1912 (10), 297; full review. OVARIAN CYSTS 519 is entirely unavailable for metabolism, at least in all cases, but in malignant fatty tumors this seems to be true. Hirsch*^ and Wells have studied such a tumor in which, despite most complete exhaus- tion of the fat from the normal fat depots, about two pounds of fat and four and one-half pounds of protein were stored up in the growing tumor. This tumor was an edematous retroperitoneal lipo-sarco?na, weighing G9 pounds (the heaviest solid tumor on record) and its chemical composition is given below as compared with the composition of granulation tissue (castration granuloma of swine) . Fibro lipo sarcoma. Granuloma, per cent. per cent. Alcohol ether residue 6 . 53 15 . 84 Alcohol ether extract 2.94 2.28 Total soUds 9.47 18.12 Water 90.53 81.88 Alcohol ether residue (per cent, of solids) Total protein 66.91 91.16 Protein sulphur 0 . 65 0 . 33 Protein phosphorus 0 . 44 0.31 Total Purine Nitrogen 0. 13 0. 07 Lipins contained, per cent. Total nitrogen 0.17 0 . 09 Total sulphur 0 . 13 0 . 05 Total phosphorus 0.19 0 . 09 (e) Ovarian cyst contents have been studied more than almost any other tumor products, because in their gelatinous or slimy sub- stance are contained numerous interesting forms of proteins, many of which are combined with carbohydrates and related to the true mucins. These substances are frequently referred to under the names of pseudomucin, paralbumin, metalbumin, and ovarian "colloid," and belong to the class of "mucoids."^° In view of the fact that the flu- ids in the Graafian follicles of the ovary do not contain these particu- lar forms of protein, their presence in cysts derived from adventitious structures (Pfliiger's epithelial tubes) suggests a specific form of meta- bolism on the part of the epithelium of these structures. Serous cysts, formed by dilation of Graafian follicles, usually are small in size, and the contents resemble those of the normal follicles (Oerum),^^ consisting 'of a serous fluid with a specific gravity usually from 1.005 to 1.014 (occasionally 1.020 or more), and containing 1.0-4.0 per cent, of solids. Occasionally in these cysts the contents become solidified through absorption of the water, and a gelatinous or glue-like "colloid" content results. Mucoids are never present (Pfan- nenstiel).^- 89 Amer. Jour. Med. Sci., 1920 (159), 356. 9" Concerning mucoids see Mann's "Chemistry of the Proteins," 1906, pp. 541- 551. " See Malv's Jahresbericht, 1884 (14), 459. 92 Arch. f.^Gynsek., 1890 (38), 407 (literature). 520 THE CHEMISTRY OF TUMORS Proliferating cystomas contain the peculiar characteristic mucoid proteins mentioned above. Usually the contents are fluid, but of a pecuhar slimy, stringy character, due to the mucoid substance, and often opalescent or slightly turbid. The specific gravity is generally high — 1.015-1.030. The reaction is usually slightly alkaline to lit- mus, and neutral or slightly acid to phenolphthalein. If hemorrhage has occurred into them, the fluid is discolored, and may contain blood- pigments in crystalhne and amorphous forms. Small cysts often show a condensation of the proteins into a semisolid "colloid" ma- terial, but sometimes their contents resemble those of a serous cj^st. Often masses of proteins fall out of solution, forming yellowish floc- culi or large deposits half filling the cysts. As with all stagnant fluids of this type, cholesterol crystals are frequently found. The char- acteristic proteins are members of the class of pseudomucins, wliich are constantly present (Oerum). Intraligamentary papillary cysts contain a j^ellow, yellowish-green, or brownish-green liquid, which contains little or no pseudomucin; the specific gravity is usually high (1.0.32-1.036) and the fluid con- tains 9 to 10 per cent, of solids. The principal constituents are the simple proteins of blood serum (Hammarsten). According to the same author, the rare tubo-ovarian cysts contain a watery serous fluid with no pseudomucin. Chemistry of the Mucoids of Ovarian Cysts. — Pseudomucin has the following elementary composition: C, 49.75; H, 6.98; N, 10.28; S, 1.25; O, 31.74 per cent. (Hammarsten). In common with the true mucins it yields a sugar-like reducing body, which has been investigated by numerous chemists (Miiller, Panzer, Zan- gerle, Leathes, Neuberg, and Heymann*'). Panzer considers that this reducing substance is in the form of a sulphuric-acid compound, similar to, but not identical with, chondroitin-sulphuric acid. Hammarsten, however, did not find this substance constantly present. Leathes determined for the carbohydrate group the composition C12H23NO10, named it " paramucosin," and considers it a reduced chondrosin (which is the carbohydrate group of chondroitin-sulphuric acid). Neuberg and Heymann established, however, that the reducing body must come from chitosamin (CeHuNOs), and do not consider paramucosin a constant con- stituent of ovarian mucoids. The amount of reducing substance varies greatly in the mucoids found in different cysts; in some the mucoid yields but about 3 to 5 per cent., in others as much as 30 to 35 per cent., of reducing substance. Pseudomucin dissolves readily in weak alkalies, and differs from true mucin in that it is not precipitated by acetic acid, and from the simple proteins in that its solutions are not coagulated by boiling. With water a slinn', stringy, semi- solution is formed, resembling in appearance the material found in ovarian cysts. Leathes distinguishes two forms of ovarian mucoids: One, paramucin, occurs as a firm, jelly-like substance, which is converted by peptic digestion into easily soluble pseudomucin. Ovarian "coZ/oz'd" probably consists of a thickened pseudo- mucin, often mixed with other proteins. Pfannenstiel*- considers the "colloid" material as representing a modified pseudomucin, strongly alkaline and relatively insoluble, which he calls " pseudo-mucin /i." He also describes a very soluble mucoid found only in certain ovarian cysts, naming it " pseudo-jriucin y." The reason why these variations in the pseudomucins exist is not understood; they cannot be explained as due to variations in tiie cell type in the ej'st wall, although pseudomucin is probably the result of true secretion. The smallest cavities of ovarian cystadenomas contain nearly pure pseudomucin, which presents •3 Hofmeister's Beitr., 1902 (2), 201 (literature) OVARIAN CYSTS 521 a clear, glassy structure; the larger the cysts become, and the more turbid and thinner the fluid is, the more .'im])Ie are the proteins it contains. True mucin is never present in ovarian cysts. Pseudomucin occurs only in the glandular proliferating cystomas and the papillary proliferating cystadenomas, in the former appearing constantly and abundantly, in the latter not constantly and never abundantly (Pfanncnstiel). Paralhximin (Scherer) is a mi.\ture of pseudomucin with vari;ii)le amounts of simple proteins. Mctnlbumin (Schercr) is the same body that is called pseudomucin by Ilamiuarsten. Pnramucin (MitjukotT)"* is a mucoid differing from mucin and pseudomucin in reducing Fehling's solution directly, without having the carbohydrate group first split off by boiling with an acid. Hydrolysis of paramucin by PregP'' showed an absence of glycine, but traces of diamino-acids, and the presence of leucine, alanine, proline, aspartic and glutamic acids, tryptophane and tyrosine. Substances similar to pseudomucin have been occasionally found in cancerous ascitic fluid and in cystic fibromyomas (HoUmann); and they are abundant as constituents of the contents of the peritoneum in the condition known as "pseudo- myxoma peritonei,"^^ when the material is in realit.y the product of cells implanted on the peritoneal surface through the bursting of an ovarian cyst (or a cyst of the vermiform appendix (Frankel)).^^ The physically similar substance found in pathological synovial membranes by Hammarsten differs in yielding no reducing substance. Parovarian cysts arising from the Wolffian body present an entirely different content, which is a clear, watery fluid, with specific gravitj' usually under 1.010; the solids amount to but 1 or 2 per cent., and consist chiefly of salts (the ash being often over 80 per cent.), mostly sulphates and chlorides. They are usually (or always) free from pseudomucin, mucin, or other sugar- containing substances, and other proteins occur only in small amounts, unless the cyst is inflamed. Apparently mucoids do not form in cysts lined by ciliated epithelium (Pfannenstiel). Santi^* has studied the physical chemistry of ovarian cysts, and finds the freez- ing point very near that of blood, having no relation to density, viscosity or nitrogen content; the specific electrical conductivity is higher than that of blood serum. The physicochemical properties are less dependent upon chlorides, and more on other substances (Gruner).^^ (/) Dermoid cysts of the ovary contain, as their chief and most characteristic constituent, a yellow fat, which melts at 34°-39° and soUdifies at 20°-25°. Ludwig and Zeynek^ have examined over sixty such tumors, and found that the fatty material constantly contains two chief constituents: one, crystaHizing out readily, they believed to be cetyl alcohol, (CHs — (CHo) u — CH2OH) ; the other, remaining as an oily fluid, seems to be closely related to cholesterol, although not consisting of one substance alone. Small quantities of arachidic acid (Cvq, H40O2), as well as stearic, 'palmitic and myristic acid (C14H28O2), existing as glycerides, are also "pres- ent. Ameseder,^ however, found evidence that the supposed cetyl alcohol is really eikosyl alcohol (C20H42O). These substances are se- creted by the glands of the cutaneous structures of the cyst, and re- s'" Arch. f. Gynsek., 1895 (49), 278. 35 Zeit. physiol. Chem., 1908 (58), 229. 9« Literature by Peters, Monatschr. f. Geb. u. Gyn., 1899 (10), 749; Weber, St. Petersb. med. Woch., 1901 (26), 331. " Miinch. med. Woch., 1901 (48), 965. s^ Folia clin. chimica et microscop., 1910 (2), 73. 39 Biochem. Jour., 1907 (2), 383. iZeit. phvsiol. Chem., 1897 (23), 40. 2 76id., 1907 (52), 121. 522 THE CHEMISTRY OF TUMORS semble in composition sebaceous material, which is characterized by containing a large proportion of cholesterol partly combined with fatty acids. Dermoids sometimes contain masses of fattj' concretions which seem not to depend on chemical changes but on the presence of forma- tive nuclei and framework of desquamated epithelium; they consist of a mixture of neutral fats and cholesterol esters, with some free cho- lesterol.^ Cholesteatomas, in addition to their abundant cholesterol content, contain keratin.'' {g) "Butter" Cysts." — In the mammary gland retention cysts form, filled with products of alteration of the milk, including butyric acid and lactose (Klotz),® and these are called "butter cysts" or milk cysts. Analysis of the contents of such a cyst by Smita^ gave the following results, as compared with human milk: Cyst contents Human milk Fat 72.97 3.90 Casein 4.37 0.63 Albumin 1.91 1.31 Milk-sugar 0.88 6.04 Ash 0.36 0.49 Water 20.81 87.09 Fats consisted of — Cyst Cows' milk Stearin and palmitin 37.0 50.0 Olein 53.0 42.2 Butyrin 9.0 7.8 Occurring independent of lactation usually, but not always, are the ''soap cysts," which contain chiefly calcium and magnesium soaps, but also neutral fats, free fatty acids, and traces of cholesterol (Freund).8 (2) Malignant Tumors The chief general features of the composition of these growths have been considered in the discussion of the chemistry of tumors in general. A malignant tumor differs from a similar benign tumor chiefly in having usually a larger proportion of the primary cell con- stituents, and a smaller proportion of the secondary constituents and intercellular substances, since these are largely the product of the functional activity of the cells, which, in malignant tumors, do not often develop sufficiently to functionate extensively. Hence malig- nant tumors usually show a rather high proportion of the characteristic constituents of nucleoprotoins; /. e., phospiiorus and iron. If rapidly 'Lippert, Frankf. Zeit. Path., 1913 (14), 477. ^Risel, Verh. Deut. Path. Gesell, 1909 (13), 322. '■An "oil cyst" behind the ear has been fully analyzed by Kreis (Schweiz. Apoth. Ztg., 1918 (56), 81) and found to contain, in addition to much neutral fat and cholesterol, consideral)le amounts of high unsaturated hydrocarbons. * Arch. klin. Chir., 1880 (25), 49. ' Wien. klin. Woch., 1890 (3), 551; see also Zdarek, Zeit. phvsiol. Chem., 1908 (57) 461. « Virchow's Arch., 1899 (156), 151. MALIGNANT TUMORS 523 growinp;, thoy contain much potassium; if un(lcrp;oinp; much retrogres- sion, httlo potassium and a hirger amount of calcium (Beebe, Clowes and Frisbie). On account of the extensive disintegration, the products of autolj^sis are usually mu(^h more abundant than in b(mign tumors. The composition varies greatly with the origin, although to a less extent than with the benign tumors. In Fraenkel's laboratory^ it was found that cancers are often defective in tryptophane, and from a squamous cell carcinoma of the skin little or none of this amino-acid could be obtained, although normal squamous epithelium is rich in trypto[:)hane. Fasal,^'^ however, found usually a high tryptophane jBgure in cutaneous epithelioma, but very irregular results in other tumors. As Bang and Beebe have shown, the tumors arising from lymphatic tissues show the chemical characteristics of these structures, and contain histon nucleinate. Tumors from squamous epithelium develop keratin in direct proportion to the amount of maturity the cells reach. Even the most complex and specific products of metabolic activity may be developed by malignant tumors (e. g., thyroiodin, epinephrine, bile), and in a form and condition capable of performing function. As Buxton and others have shown, malignant tumors pro- duce a great variety of intracellular enzymes. The idea that glycogen is present in tumors in proportion to their malignancy has been dis- proved by Lubarsch, Gierke, and others; among the malignant tumors glycogen is found particularly in chorioepitheliomas, hypernephromas, and squamous cell carcinomas. Of particular importance is the ob- servation of Beebe, that the composition of metastatic growths is modified by the organ in which they are growing, so that they tend to resemble the organ serving as their host; which, however, does not hold for certain of their enzymes (Wells and Long). In a case of primary carcinoma of the liver, Wolter'^ found the tumor tissue richer in nuclein phosphorus and poorer in phosphatids than the adjacent liver tissue; cholesterol was 0.25 per cent, of the fresh weight, fatty acids 1.67 per cent, and water 82.33 per cent., the water of the normal tissue being 79.34 per cent. As to the special varieties of malignant growths, there is little as yet determined concerning their chemistry beyond what has been stated previously. The variations in compositon of tumors are largely the direct result either of their resemblance to some normal tissue or of degenerative changes that they have undergone. " Colloid" carcinoma may be mentioned specially, in view of the confusion caused by the lax use of the term "colloid" {q. v.). The fluid contents of colloid cancers of the gastro-intestinal tract are usually chiefly epithelial mucus, containing mucin mixed with a greater or less quantity of proteins from degenerated cells and serous effusion. This mucin is acid in reaction, is precipitated by acetic acid, 9 Wien. klin. Woch., 1912 (25), 1041. '» Biochem. Zeit., 1913 (55), 88. 11 Biochem. Zeit., 1913 (55), 2G0. ■ , 524 THE CHEMISTRY OF TUMORS and has an affinity for basic dyes.'^ The colloid cancers of the mam- mary gland, in which the "colloid degeneration" involves the stroma, probably contain a connective-tissue mucin analogous to that of the umbilical cord, as also do the myxosarcomas, if we may judge by their origin and staining reactions, but no exact chemical study of these substances can be found. Colloid cancers of the ovary, arising usually from the same structures as the ovarian cysts, contain pseudo- mucin or alhed bodies (see ''Ovarian Cysts"). Colloid tumors of thyroid tissue often contain the typical colloid of normal thyroid tissue, even when metastatic in other organs; in the tumor-colloid may be a relatively normal proportion of iodin (Gierke ^■^). Hypernephromas possess several interesting chemical features. For example Gatti^'* brought forward the fact that such a tumor analyzed by him contained 3.4735 per cent, of lecithin, which agreed very well with the amount of lecithin in normal adrenals. Beebe^^ found in the watery extract of a hypernephroma the following sub- stances: tryptophane, proteoses, glycogen, leucine, and tyrosine, in- dicating the occurrence of autolysis. About 29 per cent, of fat was present, which was all extractable without pepsin digestion, and the fat contained about 18 per cent, of its weight as cholesterol. Lecithin was also present, but not quantitatively determined. A study of the fats and lipoids of hypernephromas and other tumors gave the results shown in the following table i^*^ "3 a 0) u c3 a u o Hypernephromas °S3 ll fa Carcinoma of breast Sarcoma, second- ary, inlUvei 1 2 3 4 36.3 7.6 11.9 20.6 11.8 18.4 33.0 28.0 4.6 6.4 16,9 6.0 8.3 22.7 33.0 6.7 10.0 20.4 9.0 13.4 27.5 38.4 8.7 14.0 22.9 8.3 13.4 21.4 85.0 0.5 3.3 0.7 2.0 13.3 2.4 8.6 2.2 2.4 26.1 1.7 1.9 20.0 21.4 14.5 Cholesterol, per cent, total 'dry weight Cholesterol, per cent, dry, fat-free 0.9 1.6 1.2 1.9 Cholesterol, per cent, ether-soluble substance 4.3 11.0 Lecithin, per cent, total dry weight . . . Lecithin, per cent, dry, fat-free sub- 0.7 j 6.2 0.9 7.3 Lecithin, per cent, ether-soluble sub- stance 3.0 39.8 Hypernephroma No. 1. — Typical specimen, with the usual amount erf hemorrhage and necrosis; cells much vacuolated. Hypernephroma No. 2. — Similar^tolNo. 1. Hypernephroma No. 3. — Primary growth resembled more a papilloma than an ordinary hyper- nephroma in most places; no vacuolization of cells, little necrosis, and no hemorrhage. Hypernephroma No. 4, — -Tumor resembling a lipoma, witli a stroma in places resembling a fibrosarcoma in structure. In only a few areasj^were cells present resembling adrenal tissue most of the tissue resembling fatty areolar tissue. i-Thc fluid of a colloid cancer of the peritoiioviin examined by llnwk contained a protein resembling serosa mucin, containing 11.5 \ivv cent, of X and O.S per cent, of 8. (McCrae and Coplin, Amcr. Jour. Med. Sci., 1910 (151), -175.) 13 Hofmeistcr's Beitr., 1902 (3), 28(5. '* Virchow's Arch., 1897 (150), 417. "Amer. Jour. Physiol., 1904 (11), 139. "Wells, Jour. Med. Res., 1908 (12), 4G1; see also Steinke, Frankfurt. Zeit. Path., 1910 (5), 1G7. I MYELOPATHIC I'liDTI-:! WRI A 525 It will be at once observed that the two typical hypcrnoi)hromas, Xos. 1 and 2, show a marked resemblance to the normal adrenal in the proportion of fat and lipoids. (The lower figure for lecithin in No. 1 probably is due to the fact that this specimen had been preserved lon<;er than the others.) This was what was to be expected from the microscojjic resemblance of these tumors to adrenal tissue, and corroborates the results of Gatti's and Beebe's observations on iso- lated cases. More surjirisinf"; is the fact that equally comparable results were obtained in the hypernephroma (No. 3), which contained only cells free from vacuolization and not at all resembling adrenal cells. From this it may be concluded that in these tumors of adrenal origin the amount of fats and lipoids present cannot be estimated from the degree of cytoplasmic vacuolization of the cells, or the extent of necrosis; the fatty materials are an integral part of the cells, present in them as an essential constituent and not as the result of degeneration. The results of analysis of two carcinomas and a sarcoma indicate that the hypernephromas are peculiar in their close resemblance to adrenal tissue in respect to fat and to lipoid content. The amount of all these constituents in these three tumors is far below that found in the hypernephromas, although in the carcinoma of the breast the amount of simple fats is relatively large, as might be expected in view of the function of the cells from which it arose. It is interesting to note that a carcinoma of the gall-bladder shows a rather high proportion of its fatty material as cholesterol, for this observation may bear a relation to the well-known tendency of the ei)ithclium of the gall-bladder to form cholesterol. The large proportion of lecithin in the sarcoma of the liver may possibly be due to the in- fluence of the soil upon which the tumor was growing, but we need more informa- tion concerning the lipoid content of other malignant tumors arising in different sites. Renal hypernephromas reproducing the adrenal cortex in struc- ture do not contain epinephrine,^^ but tumors of the adrenal arising in the medulla may do so.^* Microscopically, hypernephromas con- tain much glycogen. The special tests for hypernephroma tissue recommended by Croftan seem not to be specific. ^^ Melanotic tumors produce melanin, which seems not to differ at all from the melanin found in normal pigmented structures. Hel- man-° found as high as 7.3 per cent, by weight of melanin in melano- sarcomas. (See also Melanin, p. 474, and Enzymes in Tumors, p. 505. Concerning Chloromas-^ see p. 480.) Multiple Myelomas and Myelopathic "Albumosuria" Multiple mijelom.as are of particular chemical interest, because of the appearance in the urine in such cases of the peculiar protein first described as an albumose by Bence-Jones,^^ and now, because of lack of grounds for its definite classification, generally known as the '^ Bence-Jones body" or " Bence-Jones protein." Because of the ex- tensive bone destruction there is also an excessive excretion of cal- cium,-^ and sometimes metastatic calcification may occur. ^^ This 1^ Greer and Wells, Arch. Int. Med., 1909 (4), 291; Brooks, Jour. Exp. Med., 1911 (14), 550; Ciaccio, Deut. Zeit. f. Chir., 1910 (104), 277. 18 Wegehn, Verb. Deut. Path. Ges., 1911 (15), 255. 19 Koerber, Yirch. Arch., 1908 (192), 356. 20 Arch, internat. Pharmacodyn., 1903 (12), 271. 2' Metabolism in chloroma does not differ from leukemia (Sakaguchi, INlitt. Med. Fak., Tokio, 1914 (13), 198). 22 For Hterature, see Rosenbloom, Biochem. Bulletin, 1911 (1), 161; Vance, Amer. Jour. Med. Sci., 1916 (152), 693. 23 Blatherwick, Amer. Jour. Med. Sci., 1916 (151), 432. 24 Tschistowitsch and Kolessnikoff, Virchow's Archiv.. 1909 (197), 112. 526 THE CHEMISTRY OF TUMORS variety of tumor differs from the standard types of malignant tumors in thac it involves the marrow of many bones simultaneously, in a very diffuse manner, without usually giving evidence of a true metastasis. In many respects it resembles the leukemias, pseudoleukemia, and chloroma, and it is extremely uncertain as to where in the classifica- tion of tumors and of the diseases of the blood-forming organs this disease should be placed. Histologically, the tumors show evidence of being derived from the specific cells of the marrow, either from the plasma cells (Wright) or from the neutrophile myelocytes-^ or their predecessors (Muir). Cases of myeloma without the proteinuria have been described, and also a few instances of the presence of apparently typical Bence-.Iones protein in the urine without myelomas, but with bone carcinomas, leukemia or chloroma.-'' Properties of the "Bence=Jones Protein." — Not to go into details, which are given in the hterature cited, the important facts con- cerning the " Bence-Jones protein," and its appearance in the urine {"myelopathic albumosuria," Bradshaw), are as follows: It is a protein, the exact nature of which has not been determined; at first considered an albumose because of its peculiar reactions to heat, its nature has since been contested, but the weight of evidence seems to be in favor of the contention of Simon that it is most closely related to the water-soluble globuHn of the blood. In certain cases it partly precipitates spontaneously from the urine," and it may crystallize in the renal tubules. ^^ Its most characteristic properties are the following: The coagulation temperature is low, varying from 49°-60° in various cases, and being considerably modified by the amount of salts and urea present in the solution. Probably the protein forms a molecular compound with the salts which is more stable at 100° than at lower temperatures (Hopkins and Savory). In many cases the coagulum is redissolved on heating, and reappears on cool- ing, but this characteristic feature is not always present, and often disappears in cases where at first it is present. A precipitate is formed by strong (25 per cent.) nitric acid, which disappears on heating and reappears on cooling. Strong hydrochloric acid causes a dense precipitate, which is quite typical (Bradshaw). No precipitate is produced by acetic acid, even in excess, and the addition of acetic acid to a hot coagulated specimen causes prompt solution of the coagulum. UnUke albumoses, this substance does not dialyze; the salt-free solution left in the dialyzing bag does not precipitate. A purplish-violet color is usually given with the biuret reaction, but it may be more reddish in color, especially if little copper is present. " Forman and Warren (Jour. Cancer Res., 1917 (2), 79) found the cells to con- tain granules giving the indol-phenol blue reaction and hence belonging to the myeloid group. ^'^ Glynn has described a glycoprotein resembling Morner's body, in the urine during myeloma (Liverpool Med. Chir. Jour., 1914, p. 82). A crystallizable pro- tein, resembling the Bence-Jones body, has been found in the urine of a woman with gastric cancer without any bone involvement (Schumni and Kimmcrle, Zeit. physiol. Chem., 1914 (92), 1). " Rosenbloom, Arch. Int. Med., 1912 (9), 255. " Loehlein, Cent. allg. Path., 1913 (24), 953. BENCE-JONES PROTEIN 527 Sulphur is readily split off by alkalies, reacting with lead acetate to produce lead sulphide (Boston). After standinp; in alcohol, by which the protein is precipitated, it loses its solu- bility (differing in this respect from albumose). As to the exact nature of this protein, httle can be said at the pre- sent time. Since protoproteoses, dcuteroprotcosos, and peptone are split off on digestion with pepsin, the molecule is evidently larger than that of any of the albumoses. The well-purified substance is free from phosphorus, and hence contains no nucleins; but it contains con- siderable sulphur (between 1 and 2 per cent.), which is readily split off. Like casein, it contains no hetero-group (lack of heteroproteoses on digestion), but differs in containing a carbohydrate group (in small amount) and in the absence of phosphorus. On hydrolysis Magnus- Levj^'-^ obtained glutaminic acid, tyrosine, and leucine, but no glycine. He found the nitrogen distributed as follows: amid-nitrogen, 9.9 per cent.; humin-nitrogen, 9.8 per cent.; diamino-nitrogen, 6.4 per cent. — which last was composed of : histidine, 0.9 per cent.; arginine, 2.4 per cent.; lysine 3.0 per cent. The extensive analytic studies of Hopkins and Savory^° show that the amino-acid grouping is that of a typical protein, with a liigh proportion of aromatic radicals, similar proteins not being found in the tumors or muscles of a typical case. In fact, the amino-acid content, as given below, indicates that Bence- Jones protein is as distinct from other proteins in chemical composition as in its physico-chemical properties. The amino-acids, in round numbers, were isolated in the following percentage proportions of the entire protein: Valine-leucine fraction, 14; glutamic acid, 8; aspartic acid, 2; proline, 2.7; phenylalanine, 4.8; t3T0sine, 4.2; tryptophane, 0.8; cystine, 0.6; arginine, 6; histidine, 0.8; lysine, 3.7; sulphur, 1.2. An important point in this work is the agreement in composition of the proteins from two different cases, being identical within the hmits of the analytic methods, showing that the protein is of constant and characteristic properties. Occurrence of "Myelopathic Albumosuria." — Not all eases of multiple myeloma show the presence of Bence-Jones protein in the urine, however, and it is present occasionally in other conditions. Multiple bone involvement by other tumors does not often cause "albumosuria."" There is no evidence that it occurs in the normal body, even in the bone-marrow, or that it is produced as a step in the splitting of any form of proteins. A few cases of supposed osteomala- cia have been reported, with the Bence-Jones bodj^in the urine, but on more careful investigation these seem to have been unrecognized mye- '9 Zeit. physiol. Chem., 1900 (30), 200. '« Jour, of Physiol., 1911 (42), 189. '' A case of this kind has, however, been described by Oerum (Ugeskrift f. Lager., 1904, No. 24), in which the bone tumors were multiple metastases of a gastric carcinoma. See also Boggs and Guthrie, Amer. Jour. Med. Sci., 1912 (144), 803. 528 THE CHEMISTRY OF TUMORS lomas (e. g., the cases of Bence-Jones and of Jochmann and Schumm). Similarly the case reported by Askanazy as leukemia with Bence- Jones protein in the urine, on reexamination was found to be multiple myeloma. However, at least eight eases of true chronic leukemia with Bence-Jones proteinuria have been reported.^- Coriat^^ describes a substance found in a pleuritic fluid which gave the reactions of the Bence-Jones body, and he believes that it may have been formed from serum globulin through the digestive action of the leucocytes or bac- teria, Zuelzer reports finding the same body in the urine of a dog poisoned with pyridin.^'* It is a striking fact that the kidneys elim- inate such great quantities of this protein without being permeable to the very similar normal blood proteins, and usually without show- ing evidence of structural changes. Also that when injected into animals it does not escape freely in the urine as it does in man. It may be found in the blood and exudates of patients with myeloma, ^^ as much as 7.8 per cent, having been found in the blood by Jacobson.^^ Miller and Baetjer" report finding a protein corresponding closely to Bence-Jones protein in the urine of three apparently normal persons and in two cases of hypertensive nephritis without evidence of bone disease, thus opening the question as to whether, after all, this protein is invariably associated with bone disease. Simon^^ has observed that the protein may be accompanied by dialyzable substances giving the ninhydrin reaction, probably amino-acids or peptids. Origin of the Protein. — As to the place of formation of this pe- culiar protein, there is much diversity of opinion. Magnus-Levy advanced against the idea that it is formed by the tumor cells, the following arguments: In the urine of myeloma patients are excreted great quantities of the protein,— as much as 30 to 70 grams per day, — whereas the total amount of protein in all the tumor tissue in the body seldom exceeds, or, indeed, equals this quantity. It seems im- probable that so little tumor tissue can form so much urinarj^ protein, and Magnus-Levy suggests that it must come from the food proteins as a result of altered protein metabolism. Against this view, however, are the following facts: (1) The Bence-Jones body has been found (but not constantly) in the myeloma tissue, but not in other organs or tissues; (2) the quantity in the urine is not dependent upon diet; (3) it is associated almost exclusively with this form of tumor. Simon considers it probable that the protein is formed from serum-globulin, perhaps by an enzymatic action of the tumor cells, and once formed, it is rapidly eliminated by the kidneys, as are all foreign proteins. 32 Boggs and Guthrie, Bull. .Johns Hopkins Hosp., 1913 (24), 368. " Amer. Jour. Med. Sci., 1903 (12G), 631. '* Wolgemuth (Arb. a. d. Path. Inst, zu Berlin, Festschrift, 190(5, p. 027) states that normal human bone marrow may contain true albumosos. "Taylor el al, Jour. Biol. Chem., 1917 (29), 425. "^ Jour. Urol., 1917 (1), 167. " Jour. AiiuT. Med. Assoc, 1918 (70), 137. ".Jour. Anicr. Med. Assoc, 1918 (70), 224. BENCE-JONES PROTEIN 529 Normal bone marrow does not contain this protein (Nerking^'*). Roscnbloom'"' has found evidence that Bence-Joncs protein may pos- sibly be derived from the osseo-albunioid of the bones. Weber and Ledin