ve +e ; nya a pisos ree, if : vie viene ie were ve toe + mas srs Ad vv." Andean ra da ttt 3 Tire vive ete ny. vier - aa vr, § or MeRet es ¥ Sieaty ele rt? peak aes y wre . \Veree CV viws eo pheeerwewengen eds (re Pere rwee seein tte nit einai AS (ia ie Sean We a ont ' be Vay pi Tr eh te y ‘5 ’ CAS eG LE j > { vt Rat Oh reey fi } Bitty Va eee . Aad ol 1 ” ; \ * Ne A ol yy j ft t : ; f fit, % a ’ 3 1 i iy * 7 ? 9 ; DFR ek it Wh Vo hae } i bs] 5 ‘ a* i } t ie ¥ P ‘ ‘ i f i f ,! : : x > - ‘ 1 THE JOURNAL OF IMMUNOLOGY VOLUME III BALTIMORE, MD. 1918 906196 CONTENTS NuMBER 1, JANUARY, 1918 A Rapid Method for the Production of Precipitin Antigen from Bacteria: An Attempt to Apply it to the Determination of the Type of Pneumo- coccus in Sputum. Charles Krumwiede, Jr., and W. Carey Noble... ... On Von Dungern’s Indigo Test for Syphilis. B. Fujimoto................. Extracts of Antibodies Obtained from Specific Precipitates of Typhoid- Antityphoid Serum'Complex. Israel Weinstein........................ The Specificity of Intracutaneous Absorption. G. H. Smith and M. W. (COOLS Ri ES ge ae RRR ci UGE tine Bir neh aA ULE OD LR CRA Sg OO ep ere RRR Specific Reactions of the Body Fluids in Pneumococcic Infections. G. R. ere yasmal © OS Pla chien tase soci ee ciel ae hae nc eis coe aco eee oe ae NuMBER 2, Marcu, 1918 Studies on the Antitrypsin of Serum. B. Fujimoto................5.....2% The Constancy of the Protein Quotient During Intensive Digestion and Prolonged Starvation. Samuel Hanson: /.0.79 ss ee ie ones eee The Immunologic Properties of Uveal Pigment. Alan C. Woods........... The Examination of the Blood Preliminary to the Operation of Blood Trans- 1 AUN) PY Wat) Osea a O Tel Raa ie A EO A Experiments upon the Passive Transfer of Antibodies from the Blood to the Cerebrospinal Fluid. John A. Kolmer and Shigeki Sekiguchi.......... The Isolation, Purification and Concentration of Immune Bodies. A Study CisimmunesHemoliysinss.> Mis Kosaka pects sae hia et A New Method of Estimating the Antitryptic Index of Blood Serum. T. Brailsford Robertson and Samuel Hanson.....................0ese-00- The Non-Influence of Injections of Trypsin upon the Protein Quotient in Blcodsserum.. Samuel iansona era eee eee eae Effects of Intravenous Injections of a Colloid (Gelatin) upon Rabbit Sera. LEWES “\ NGG) GN amen we OO aa eae a ea RA US Be Sar Par re Oe er NuMBER 3, May, 1918 The Influence of Active Normal Serum (Complement) upon Meningococci. I. The Opsonic Activity.of Fresh Normal Serum Alone and in Combination with Antimeningitis Serum for Meningococci. John A. Kolmer, Ikuzo oyamavanG LOLs) VMatsunamt: 2 occ cpp ils eit oom Weld sca onl aeieiae The Influence of Active Normal Serum (Complement) upon Meningococci. II. The Bactericidal and Protective Value of Fresh Normal Serum Alone and in Combination with Antimeningitis Serum for Meningococci. ois Viatsunami and John, iKolmerien...c sjecia atic an ueeeceeee lil 101 109 131 139 147 157 1V CONTENTS The Relation of the Meningococcidal Activity of the Blood to Resistance to Virulent Meningococci. Toitsu Matsunami and John A. Kolmer....... 201 Experiments on the Production of Antipoliomyelitic Serum in Rabbits. dean “he H. Dsenmis sos!) nr eee sooo oe oc Rein ee tote ohare ee 213 The Study of Problems of Immunity by the Tissue Culture Method. II. The Tissue Culture as a Means for Quantitatively Estimating Toxin and Antitoxin and Determining the Distribution of Antitoxin in Passively Immunized Animals. Montrose T. Burrows and Yoshio Suzuki........ 219 The Study of Problems of Immunity by the Tissue Culture Method. I. A Study of the Cells and Blood Plasma of Animals which are Naturally Resistant and Others which are Susceptible to Diphtheria and Tetanus Toxins. Yoshio 'Suzuki.l/.c a)... 25 {SUS Ah ee eee 233 NumBErR 4, Juuy, 1918 A Study of the Immunizing Properties of Bacterial Vaccines Prepared after Various Methods. M. W. Perry and John A. Kolmer....... aay, The Bactericidal Action of Whole Blood with a New Toctnrane, ae he Determination. George D. Heist, Solomon Solis-Cohen and Myer Solis=Cohens. ss .2a-d 5 Sh aes vse ea ee oars oe Eee 261 Complement Fixation with Protein Substances. Reuben L. Kahnand Archi- bald MeNeil..... eS eee ee Le ee PETS Pes co Soke yoo Nae ase ae< 277 A Note on the Relation Between Proteolysins and Hemolysins. Archibald MeNeiliand Reuben Il, Kahne. fac y:.<.1-e eee ee eee ao eeeeee 295 The Influence of Arsphenamine and Mercurie Chlorid, upon Complement and Antibody Production. Ikuso Toyama and John A. Kolmer........ 301 Proceedings of the American Association of Immunologists. Fifth Annual Meeting, held at the New Medical Laboratories and in the Hygiene Lab- oratory of the University of Pennsylvania, Philadelphia............... 317 NuMBER 5, SEPTEMBER, 1918 A Contribution to the Study of the Complement Fixation Reaction in Tuber- culosis: MC A> Wilson. 26. .840555,. ./..< agen ee ee 345 A Contribution to the Study of the Complement Fixation Reaction for Tuberculosis. Hassow) von. Wedel... ).aeeccee ee eee oe ee 351 The Réle of Immunity in the Conduct of the Present War. John A. Kolmer 371 On the Mode of Action in Vitro and the Preparation of Hemolytic Antibodies. A. 1K, Balls:and. John i: Korne.. 2280 oe ee eee ee ee 377 A Note on Bleeding Guinea-pigs and on Preserving Sheep’s Erythrocytes. Sok J: Wenner 5556 oid sod she 2 pnaa's ko oleae & ee I ee 391 Studies in Pneumonia. VIII. A Skin Reaction to Pneumotoxin. Charles Weiss‘and John A.. Kolmer:. ; ....0 20) saseme ee ee eee 397 A Method of Preparing Bacterial Antigens. James C. Small............... 413 A Study of Saponin Hemolysis. Tanemoto Furuhata...................... 423 CONTENTS Vi NuMBER 6, NoveMBER, 1918 Active Immunity in Experimental Poliomyelitis. H. L. Abramson and Herman Gerber........... ty Ao Oe eed cd aa MO 2? ot RE aN COURTS 435 Experimental Pollinosis. Preliminary Report. Henry L. Ulrich........... 453 Prompt Macroscopic Agglutination in the Diagnosis of Glanders. Olga R. AON ALIE LETS c sols sb POS CORR IO ae Oa aT te tT tir an en aA meen, 463 ae MAJOR RICHARD WEIL, M.O.R.C. , Richard Weil was born in New York City in 1876, next to the youngest child in a family of seven children. His early educa- tion was directly under the supervision of his mother, who was an exceptionably able and clear-headed woman. At the age of twelve he entered the sehool of Dr. J. Sachs where he prepared for college, entering Columbia University, class of 1896. Al- ready at this early age he showed a remarkably brilliant mind, which reasoned clearly and delved deeply into all subjects that presented themselves to him. At Columbia his intellectual ver- satility showed itself in different lines of study—prizes in English and Latin, honors in Greek, original research in science bearing testimony to his assiduous study and careful and thorough work. In recognition of his exceptional ability he was elected to the Phi Beta Kappa at the close of his junior year and he was in- vited by three departments, the Latin, the English and the bio- logical to become a member of their teaching staffs. Before graduating from Columbia, his mind had been opened to the great possibilities presented by the study of medicine, and this led him to decide upon medicine as a career despite all the other attractive fields that lay open before him. Before entering the College of Physicians and Surgeons, class of 1900, he spent a summer at Wood’s Hole at the biological station. In the Medical School he was an excellent worker and found time in off hours to do experimental work in the physiological laboratory. This work led to the degree of Master of Arts. On graduating from the medical college he became an interne at the German Hospital (October, 1900, to October, 1902), where he re- ceived an excellent practical training in medicine and surgery. After the completion of this period of practical training he went to Europe for one and a half years study of medicine and its allied sciences. His clinical work was done chiefly in Vienna under Nothnagel, Neusser and Naunyn, and all three developed 1 ll EDWIN BEER AND CARY EGGLESTON boundless enthusiasm in their eager pupil. Under Marchand and vy. Recklingshausen his scientific longings were further whetted and he obtained a glimpse of the wide reaches of scien- tific medicine. While at Strassburg he was detailed to work in a typhoid epidemic and acquitted himself most creditably. In 1904 he returned to New York to enter upon a career as physician-investigator, and in all the years that followed he lived up to this professional ideal. Upon taking up practice in 1904, he began his scientific in- vestigations in the realm of medicine at Cornell Medical Col- lege and in the following year was appointed Assistant in Experi- mental Pathology. Four years later he became Instructor in Cornell in the Department of Experimental Therapeutics, which position he held until 1911 when he became Assistant Professor in the same department. During this period he was active in investigative work, which followed along the lines of experimental pathology and in 1915 he was given the Assistant Professorship in the Department of Experimental Pathology. The following year this department was merged with that of Experimental Medicine and Dr. Weil was given the chair which he held at the time of his death. In 1904 he was also made Adjunct Pathologist to the German Hospital, which position he held until 1910. From 1908 until 1913 he was an Adjunct on the Visiting Staff to the Mt. Sinai Hospital, thereafter being a member of the Assistant Attending Staff. In the same year that he became an Associate Attend- ing at Mt. Sinai Hospital he was given the positions of Assist- ant Director of Cancer Research and Attending Physician in the newly reorganized General Memorial Hospital. Upon his appointment to the chair of experimental medicine in Cornell University Medical College he resigned the position of Assistant Director to the Memorial Hospital but continued there as At- tending Physician. During these thirteen years Dr. Weil not only performed his official duties in the positions Just men- tioned, but found time to conduct a fairly active private prac- tice and to delve deeply into scientific research. Among the problems to the solution of which he bent his energies the two MAJOR RICHARD WEIL ill to which he devoted the major portion of his time were those concerned with hemolysis and with anaphylaxis. To the former of these problems he contributed much of value, but the work by which he is known in this field is probably that concerning the variable resistance of human red blood cells to the hemolytic action of cobra venom. In the field of anaphylaxis he stood almost alone as an expo- nent of the cellular theory of its mechanism. Although his work in support of this theory was vigorously criticized by many of those who supported the humoral mechanism, and although he met opposition to his views at almost every turn, he lived to see some of the leading European investigators won over to his side. His untimely death found two new studies on one of the phases of this problem in course of publication. These studies gave promise of clearing up one of the most hotly dis- puted points regarding the mechanism of anaphylactic reactions in the higher mammals. All of the investigative work that came from his pen was noted for its remarkable clarity of presentation and logic. He was never content to approach any of the problems from a single point of view, but always attempted to see all sides in their true perspective and to attack the problems that pre- sented from as many different points as possible. While the workers in the same field in this country have not yet accepted the cellular mechanism of anaphylaxis, as have some of those abroad, the evidence which Dr. Weil adduced in its support is of such a nature that we feel safe in predicting its ultimate acceptance. It is neither necessary nor fitting that we should dwell fur- ther upon his contributions to medical science in these and other fields for his published works are known to all who are interested in the subjects with which they deal. Dr. Weil was a member of many medical organizations in- cluding the New York Academy of Medicine, the American Medical Association, the American Society for the Control of Cancer, the American Society of Clinical Investigation, the As- sociation of American Physicians and others of a more restricted 1V EDWIN BEER AND CARY EGGLESTON and narrower scientific nature. He occupied important offices in a number of scientific societies, having been vice-president of the American Association for Cancer Research and president of the Society for Serology and Hematology, and of the Ameri- ean Association of Immunologists. He also found time to per- form the duties of an Associate Editor of the JouRNAL oF Im- MUNOLOGY and of the American Review of Tuberculosis as well as having been the Editor of the Journal of Cancer Research from its foundation to the time of his death. The range of Dr. Weil’s activities was truly remarkable but it serves better than anything else to emphasize the striking features of his character. He was a man of broad interests and clear vision, and an indefatigable worker in every field that he entered. But it was not alone in medicine that he stood out above the average for he was keenly interested in the progress of the times; he was a lover of art, of history and of literature. He could talk well on almost any subject of an intellectual nature and was always ready to sharpen his wits in a friendly argument. He had a keen sense of humor and a ready wit com- bined with a manner of easy cordiality. It was not one but rather the combination of all of these attributes which won him the lasting friendship of all who came to know him. Finally, he was imbued with a profound spirit of patriotism which led him to tender his services to his. country at the outbreak of hostilities. Starting as a captain in the Medical Officers Reserve Corps, it was but a few months before he was made a major and detailed to Camp Wheeler as Chief of the Medical Service. There his devotion to the great task of com- bating pneumonia rapidly undermined his health and he him- self fell a victim to the disease which he was endeavoring to conquer. EpwIn BEER, Cary EGGLESTON. MAJOR RICHARD WEIL Vv BIBLIOGRAPHY 1899 Development of the ossicula auditus in the opossum. Ann. N. Y. Acad. Sci., 1899, 12, 103. An anomaly in the internal course of the trochlear nerve.. Jour. of Comp. Neurology, 1899, 9, 1. 1900 On the evidence of the Golgi methods for the theory of neuron retraction. (With R. Frank.) Arch. of Neurology and Psychopathology, 1900, 3, 265. 1904 Typhoid epidemiology. N. Y. Medical News, 1904, February 20 and March 5. 1907 Concerning a distinct type of hypernephroma of the kidney which simulates various cystic conditions of that organ. Ann. of Surgery, September, 1907. Hemolytic properties of organ and tumor extracts. Jour. of Med. Research, 1907, 11, 287. 1908 The hemolytic reactions in cases of human cancer. Jour. of Medical Research, 1908, 19, 281. The hemolytic reactions of the blood in dogs affected with transplantable lymphosarcoma. Archives of Internal Medicine, 1908, 1, 21. 1909 On the resistance of human erythrocytes to cobra venom. Jour. of Infectious Dis., 1909, 6, 688. On the specific acquired resistance of red blood cells. Proc. of the Soc. for Exp. Biol. and Med., 1909, 6, 49. Variation in resistance of human erythrocytes. Proc. of the Soc. for Exp. Biol. and Med., 1909, 7, 2. A method of testing the interaction of ferments and antiferments. (With S. Feldstein.) Proc. of the Soc. for Exp. Biol. and Med., 1909, 7, 61- Avoidance of hemolysis in transfusion. (With M. Rebling.) Amer. Jour. of Surgery, 1909, 23, 268. Serum reactions incancer. Jour. of Amer. Med. Assoc., 1909, 52, 407. 1910 The antitryptic activity of human blood serum. Amer. Jour. of the Med. Sci- ences, 1910, 139, 714. Experimental study of the antitryptic activity of human serum. Archives of Internal Med., 1910, 5, 109. Properties of ascitic fluids, especially in cases of cancer. Jour. Med. Research, 1910, 23, 85. 1911 On tumor-immunity in rats. Proc. of the Suc. for Exp. Biol. and Med., 1911, 9) 32. vl EDWIN BEER AND CARY EGGLESTON 1912 Bemerkungen zur Arbeit von P. Kuschakoff. Zeitschr. f. Immunitatsforsch., 1912, 13, 216. An experimental study of antianaphylaxis. (With A. F. Coca.) Proc. of the Soe. for Exp. Biol. and Med., 1912, 9, 102. 1913 The nature of antianaphylaxis. (With A. F. Coca.) Zeitschr. f. Immuni- taitsforsch., 1913, 17, 141. On a new factor in passive anaphylaxis. Proc. of the Soc. for Exp. Biol. and Med., 1913, 10, 110. The nature of anaphylaxis. Jour. of Med. Research, 1913, 27, 497. A study of the blood in rats recovered from implanted sarcoma. Jour. of Exp. Med., 1913, 18, 390. The effects of colloidal copper with an analysis of the therapeutic criteria in human cancer. Jour. of the Amer. Med. Assoc., 1913, 61, 1034. The intravascular implantation of rat tumors. Jour. of Med. Research, 1913, 28, 497. On antisensitization. Zeitschr. f. Immunititsforsch., 1913, 20, 199. The relation of anaphylaxis to the problem of human disease. Louisville Monthly Jour. of Med. and Surg., 1913, 20, 193. Studies in anaphylaxis: Ito IV. Jour. of Med. Research, 1913, 28, 243. V. Desensitization. Jour. of Med. Research, 1913, 29, 233. The cellular theory of anaphylaxis. Fourth International Congress, London, August, 1913. Section of Bacteriology and Immunity. Anaphylaxis in immune animals. Proc. of the Soc. for Exp. Biol. and Med., 1913, 10. Studies in anaphylaxis, VI: A study of the cellular theory by the graphic method. Jour. of Med. Research, 1913, 30, 87. 1914 Studies in anaphylaxis: (Jour. Med. Research, 1914, 30, 299-364): VII. The relation between antibody content and lethal dose in anaphylaxis. VIII. The function of circulating antibody and the avidity of cellular antibody. IX. The relations between partial desensitization and the minimal lethal dose in anaphylaxis. X. The presence of intracellular antigen as a factor in immunity. XI. The share of intracellular antigen in immunity and in desensitization. Theoretical considerations. Studies in Anaphylaxis: XII. Experiments in antisensitization. Zeitschr. f. Immunititsforsch., 1914, 23. 1s. XIII. The activation of antibody by the cell. Jour. Med. Research, 1914, 32, 107-120. The mechanism of anaphylactic shock. Proc. N. Y. Path. Soc., 1914, 14, 8. 1916 Sodium citrate in the transfusion of blood. Journ. of A. M. A., 1915, 64, 425. Chemotherapy andtumors. Journ. of A. M. A., 1915, 64. MAJOR RICHARD WEIL Vil The treatment of parotid tumors by radium. Journ. of A. M. A., 1915, 65. Anaphylaxis to formed or cellular elements. (With B. Denzer.) Proc. of the Soc. for Exp. Biol. and Med., 1915, 12, 7. Anaphylatoxin and the mechanism of anaphylaxis. Proc. of the Soc. for Exp. Biol. and Med., 1915, 18, 2. Equilibrium in the precipitation reaction. Proc. of the Soc. for Exp. Biol. and Med., 1915, 18, 2. : Equilibrium in the dissociation of precipitates. Proc. of the Soc. for Exp. Biol. and Med., 1915, 18, 2. Equilibrium in the combination and the dissociation of precipitates. Proc. N. Y- Path. Soc., 1915, 15, 7. 1916 Note on a skin reaction in pneumonia. Jour. Exper. Med., 1916, 23, 1. Immunological studies in pneumonia. Jour. Exper. Med., 1916, 23, 1. Studies in anaphylaxis: (Jour. of Immunology, 1916, 1, 1): XIV. On the relation between precipitin and sensitizin. XV. Equilibrium in precipitation reactions, equilibrium in combination. XVI. Equilibrium in precipitation reactions, dissociation. XVII. On the coexistence of antigen and antibody in the body. Chemotherapeutic experiments on rat tumors. Jour. Cancer Research, 1916, Li. The analysis of serum sickness. Proc. N. Y. Pathological Soc., N. S., 16, 3 and 4. Characteristics of the precipitation reaction. Proe. of the Soe. for Exp. Biol. and Med., 1916, 13, 8. Studies in Anaphylaxis: XVIII. The mechanism of delayed shock. Jour. of Immunology, 1916, 2, 95. XIX. Simultaneous injection of antigen and antiserum. The anaphyla- toxin theory of anaphylaxis. Jour. of Immunology, 1916, rl 1917 The excretion of congo red by the stomach. (With R. L. Cecil.) Proc. of the Soc. for Exp. Biol. and Med., 1917, 14, 72. Further studies in serum sickness. Proce. of the Soc. for Exp. Biol. and Med., 1917, 14, 60. Anaphylaxis in the dog. Proc. of the Soc. for Exp. Biol. and Med., 1917, 14, 6. The immune reaction to tuberculous infection. Jour. of A. M. A., 1917, 68, 972. The relation between antigen and antibody in the living animal. Jour. of Im- munology, 1917, 2, 399. The vasomotor depression in canine anaphylaxis. Jour. of Immunology, 1917, 2, 429. Studies in Anaphylaxis: XX. The reciprocal relations of antigen and antibody within the cell. Jour. of Immunology, 1917, 2, 470. XXI. Anaphylaxis in dogs; a study of the liver in shock and-in peptone poisoning. Jour. of Immunology, 1917, 2, 525. XXII. Anaphylactic reaction of the isolated dog’sliver. (With C. Eggleston.) Jour. of Immunology, 1917, 2, 571. A RAPID METHOD FOR THE PRODUCTION OF PRE- CIPITIN ANTIGEN FROM BACTERIA: AN ATTEMPT TO APPLY IT TO THE DETERMINATION OF THE TYPE OF PNEUMOCOCCUS IN SPUTUM CHARLES KRUMWIEDE, JR. anp W. CAREY NOBLE From Bureau of Laboratories, Department of Health, New York City Received for publication October 9, 1917 Various methods have been devised for obtaining a precipitin antigen from bacteria. Many of these require prolonged incuba- tion of broth cultures or more or less complicated methods of extraction where growth on solid media is used. In some at- tempts to use the precipitin reaction to determine the presence of antigen in sputa and feces, we developed the following simple and rapid method of making a bacterial precipitin antigen which seems applicable to most,and probably to all bacteria with the exception of the acid fast types. To a heavy suspension of bacteria in distilled water, add suffi- cient alkaline hypochlorite solution! to give a final concentration of 5 per cent; boil over the free flame or heat in a water bath for several minutes. If the bacteria have not dissolved, add more of the alkaline hypochlorite solution and reheat. Repeat this process till the bacteria are completely dissolved. This is shown by the appearance of translucency. Add several drops of an alcoholic solution of phenolphthalein. Then add ;-hydrochloric acid till the color is just discharged To this neutralized solu- tion, add several volumes of 95 per cent alcohol. A copious precipitate which increases on standing, should result. The precipitate is collected by centrifuging and decanting the super- natant fluid. Add normal saline solution to the sediment to give approximately a one to twenty solution by volume and 1 For convenience we have used the soluticn marketed under the trade name of “antiformin.”’ 1 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 1 2 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE extract at the temperature of boiling water or boil over the free flame for from three to five minutes. Then centrifuge to clear the solution of the insoluble debris. The supernatant fluid is the finished antigen. It can be prepared in a half hour or less if necessary. This antigen can be reheated for sterilization at 100°C., apparently indefinitely. Certain points must be considered in carrying out the various steps of the process if a concentrated and economical antigen is to be obtained Although broth cultures concentrated by sedi- mentation in the centrifuge may be used with success, it is pref- erable wherever possible, to use distilled water suspensions of organisms grown over large areas of agar. The initial suspension of organisms must be relatively heavy, in fact, as little fluid as is needed should be used in making the suspension. Furthermore, it must not be diluted unduly in adding the alkaline hypochlorite solution or in the process of neutralization. The alkaline hypochlorite solution should be added as concentrated as is feasible in measuring the desired amounts. If the neutralized extract is very concentrated several volumes of alcohol will give a prompt, copious and almost com- plete precipitation or so nearly complete that it would not be economical to try to precipitate the remainder by further addi- tion of alcohol. If, however, the extract is dilute, more alcohol is needed proportionately to obtain a precipitate and this separ- ates more slowly. Even in only moderately dilute extracts the addition of more than several volumes of aleoho] may only cloud the solution and flocculation will be very slight and much delayed. The finished antigen, even after centrifuging, may be slightly opalescent. ‘This is due to the presence of fine particles of bacte- rial debris, which are difficult to throw down. As the concen- trated antigen may be diluted from ten to forty or more times and as the diluted antigen is very clear this opalescence is not a factor. If for some special purpose it is necessary to use the antigen in its concentrated form, the presence of this bacterial debris must be considered as it is agglutinable and will therefore, bind antibody and constitute part of the precipitate. In one instance we collected the insoluble debris after boiling typhoid PRODUCTION OF PRECIPITIN ANTIGEN 3 bacilli in a 30 per cent solution of ‘“antiformin” for one-half hour; we washed the debris till no precipitable substance was present in the washings and then suspended the debris in normal saline. With dilutions of serum that gave precipitation with a finished antigen, the bacterial debris was promptly agglutinated. The control suspension without serum flocculated to some extent and settled out. Even the severe treatment given did not destroy the agglutinability of the bacterial “rests.” The presence of the fine debris can be avoided to some extent by not breaking up the sediment too vigorously in making the final extract. If for some special reason the debris is objectionable it can be re- moved by filtration. The most striking point in the method outlined above is the extreme resistance of the precipitable substances to the prolonged heating in alkaline hypochlorite solutions, which allows the rapid solution of dense suspensions with subsequent concentra- tion of antigen. This is really the only original step in the method as extraction with cold alkaline hypochlorite solution has been used before. The other steps are all well known. That which is new in the method is the combination of these steps so as to allow the rapid preparation of antigen freed from extraneous substances added in the process of preparation. We have used the method to prepare antigens from pneumococci, using the sediment from centrifugalized broth cultures, with agar cultures of B. typhosus, with types of B. paratyphosus, with B. diphtherice and with B. mallei and have obtained such satisfactory results from all that we see no reason why the method is not applicable to all bacteria soluble in an alkaline hypochlorite solution. Two questions arise in relation to this method. In the process how much precipitable substance is destroyed? We attempted to determine this by preparing antigens with different concentrations of alkaline hypochlorite solutions and employing different periods of heating. By preparing these so that the concentration of the end product was as comparable as possible we could see no evidence that there was any appreciable destruction of antigen. The other point is how far specificity is injured if at all, by the severe treatment. If the factors entering into the immune 4 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE reaction are considered; that is, standardization of antigen de- termined by dilution and the addition of varied dilutions of immune serum, with the subsequent employment of graded dilutions of the serum against the standard of antigen thus determined, as well as due consideration of the time factor; viz., the rapidity of the reaction, there is no evidence of loss of speci- ficity. If any of these factors are not considered, especially if the antigen is too concentrated, cross reactions become numer- ous and marked, especially with closely allied organisms. With types of bacteria that can be easily separated because TABLE 1 Precipitin reactions—pneumococcus sera SERUM TYPES AND DILUTIONS PNEU MOCOCCUS ANTIGEN | TIME Type I Type II Type IJI | 1-1 1-4 1-9 ie ee 1-9 1-1 er 1-9 | hours L|++*] + | 2] = pS) =) = pa Typed... - of ee Pees pineal me = * ae a } | | | 1} —-|- |} = |++}44+] 4] -— |] —] =| Type II.... | | Oo J 2) £ ) - | = 444/444) 44+) - | - | - | | | Type lil...4{ 2 [- = | =") cle ee. ee ee -}-|]-]-]-}- }41} +] - | j Antigen used = 0.2 cc. Serum solution = 0.2 ce. *++4+ = profuse precipitate. ++,+]|,+, +,+ = decreasing amounts of precipitate. of their distinctly different agglutinability even a relatively con- centrated antigen gives little evidence of cross reaction; that is, the antigen so prepared loses none of its specificity. The fol- lowing table illustrates this, only a slight cross being noted after long incubation. The antigens were made from broth cultures treated with antiformin and prepared according to the method already given above. It is not the place to go into the question of the relative speci- ficity of precipitins and of agglutinins. Because we encountered more marked cross reactions with the precipitin reactions with _ail be ite te as PRODUCTION OF PRECIPITIN ANTIGEN oO very closely allied bacteria than was evident using the same serum for agglutination reactions, we feared that we were de- stroying some specificity in these cases To exclude this possi- bility we prepared antigens according to the accepted method, that is, prolonged growth in broth and subsequent removal of bacteria by filtration and compared the results obtained from the use of such antigens with the results obtained with antigens pre- pared according to our method. With due regard to the factors mentioned above, that is, appropriate dilutions of antigens and of sera and consideration of the time factor, both types of anti- gens gave comparable results. The broth antigens being very much more dilute, did not introduce the factor of concentration of the antigens somarkedly. The following tables are representa- tive of the results obtained. No final method of standardizing the antigens has been evolved. A relatively satisfactory method, however, is to determine the volume of sediment obtained by alcoholic precipitation, using graduated centrifuge tubes, and making the final suspension by adding up to twenty volumes of saline. With comparable antigens in this way, and having obtained the appropriate dilu- tion of one antigen against its homologous serum other antigens can be similarly diluted in testing cross reactions of this serum. This was not carried further as the practical application of the precipitin reaction for differentiation of bacteria is relatively little used. The main interest of the method, in our mind is its great value in preparing material for teaching and for demon- stration purposes and for courses in immunology. It should also be of value in experimental procedures where a readily obtainable supply of concentrated antigen is needed. We have attempted to apply the method for diagnostic pur- poses for the extraction of antigen from feces and from sputa. The object in attempting to extract antigen from feces was to determine whether we could detect typhoid antigen in stools and in this way determine the presence of typhoid or allied bacilli without the necessity of a prolonged bacteriological examination. Thus far such attempts have been without success. Even extracts of feces in which 95 per cent of the bacterial flora are ee Ce 6 CHARLES KRUMWIEDH, JR., AND W. CAREY NOBLE typhoid bacilli fail to yield a precipitable extract. Evidently we are extracting other substances from the feces which appear ; TABLE 2 Comparison of broth and of antiformin antigens f ANTIFORMIN ANTIFORMIN ANTIGEN ANTIGEN BROTH ANTIGEN DILUTED 1:10 SERUM (oNDEGEEY) (ANTIGEN A) ees > DILUTION ORGANISM (ANTI- = 5 = = Pm = & | | TYPHOID) | 3 Bye 5 set ae x OMOEA GOS aaa 96/4 /| 22a ee Le amir hopantaeneleaeen (dae (arcade lor B. typhosus......... 126. esi etal ctrl gt ated tate let elton a thes Ea fle Ne eto ae it) a lfee aey = = ( bebecal Co 222 Ws ihe SE EE TS Ca eels B. paratyph ‘‘A’’.... 1:6 = = = = + 4] (Sea 1:12 ee ic ee (a be td all ae | Slee f Pea = == Ss == = — | = | = B. paratyph “‘B’’.... 126 — _ + — |SLC}+=]—|]-|= (e125 =) = 4) Sn ee | —|= 1:1 ++ 44+) + | +i [+1] - ae B. pullorum. .........: 1:6 a 1 ited ee + |+1)—-—|+|= 1:12 + {+] = | + HIl—-|= 72 1a spice mann beedueietr th lear ae | ag B. sanguinarium. ... 600 [FR EL EEL) CR eae toa2 “ie pric ler a= Ae el ek fit a | = SLCFP=" SLC S| = B. abortus equi...... . 136 = |) i= 1 SLEs) re = a Oiler 8 Pe a ee es 8a | Fag | - tae Oe ec esa Mee a a ery Breas esis 32. | es | a eer eS) eS ee ee | 1: 12) sulci Sale tle Sun eet | - i - S1.C = slightly cloudy. C = cloudy. + = slight flocculation. + = dis- tinct flocculation. + ; = flocculationmore marked. + + = heavy precipitate. Antigen 0.2 cc. and serum dilution 0.2 ec. used. * Overnight. + Symbols in these columns not comparable with those of very dilute antigen in table 3. The symbols in each table are in comparison with most voluminous precipitate obtained in series represented by a table. TABLE 3 Comparison of broth and of antiformin antigens BROTH ANTIGEN (1:5) ANTIFORMIN ANTIGEN eae (1:100) (ANTIGEN B.) ORGANISM EASE SOR! r (ANTI- TYPHOID) halt eae Tea bort Walt oo Toa hae (| 1-10 +l} ] ++] tl ++] ++ LO oilecteslbe actA Metect= ht) be) toe Bett DROSUSS. nase ener ee 1-50 + +]; |4+ a EA gall tke) 1-75 C. aE = —- + ata [| 1-100 — + aE = CQ. ae (| 1-10 — |SLOG|SLC} — Saat 1-20 Sailers 9) agli panes Cay se Ba paralypie sche 6. <8... 1-50 _ — = ae as “ a eS, |e || 1-100 - = ix = a = 1-10 | C C e = =| ten, omy |e | ot ee ADS oth a Peemmatypn. | B??..........- 1-50 = as. im BY st es 1-75 = _ = = his = 1-100 = — = 2 aS =~ | Pete tel eta at a ete Metco steaks 20 | +i | +1 | ++] 4! net 1835 FOU UTILS ees Aids SOE 1-50 = + at. + | ata | 1-75 SLC] = + — =i 1-100 — C + — ae 1-10 anil) |Waeseelarae) Sear = a 1-20 Ses eleltsese cei (se ae B. sanguinarium*........... 1-50 + }/4+]-] +1 tie | cet 1-75 C C aa = at 1-100 = — te = a (} 1-10 C C C a = + 1-20 ~ _ = = ke se B. Abortus equinus.......... 1-50 = = a fea = stir 1-75 _ — = = es a3 1-100 — _ = pe vs) = (i, 1-10 C C aa = = eet 1-20 - — =— as es ad ESRACOUPE Rares ere Rate 2 oe 1-50 -- = = = 2 zs 1-75 — - — = = es 1-100 — — = zs a Ze S1.C = slightly cloudy. C = cloudy. + = slight flocculation. + = dis- tinct flocculation. + | = more marked flocculation. +-+ = heavy precipitate. Antigen — 0.2 cc. and serum dilution 0.2 cc. used. * These two paratyphoid types of fowl origin were selected because of their very close agglutinative relationship one to the other and to B. typhosus. The extreme cross precipitation with these types in this series is no indication of loss of specificity. Compare with similar agglutinative results of Smith and Ten Broeck. (Journ. Med. Res., 1915, xxxi, 549.) 7 Overnight. 8 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE in the final antigens and interfere with the reaction. Further work is in progress to determine whether we can eliminate the inhibiting factors and to determine their character. The application of this method in preparing antigens from sputa was attempted with the hope of evolving a rapid method of determining the type of pneumococcus present. Modi- fications have been found necessary and although the method at present used is far from satisfactory and has only given positive results in a relatively small percentage of the specimens examined, it is given in the hope that it will serve as a basis to which further improvements can be added. We have been handicapped by lack of satisfactory material and further work will not be possible till the next pneumonia season.” The success of the method depends on several factors. One difficulty is that not infrequently the sputum submitted is saliv- ary or pharyngeal and not a specimen raised by coughing. Natur- ally a positive result can only be obtained if the sputum is com- paratively rich in pneumococci. The technical difficulties are first, the necessity of removing the albuminous and mucous material which otherwise passes over to the final extract, making it thick or gelatinous; secondly, the fact that the alcohol will not give a sufficiently prompt and complete precipitation where little antigen is present. The method has been modified in the attempt to overcome the first difficulty mentioned. To the sputum add sufficient antiformin to give a concentra- tion of from 3 to 5 per cent. If the sputum is thin, add the concentrated reagent; if it is very thick, dilute by adding one-half its volume of the appropriately diluted antiformin; that is, no more dilution should be made than is necessary for final digestion. The mixture is then heated to 100°C. and shaken till it becomes fluid. Add a few drops of phenolphthalein to the hot antiformin- sputum and neutralize with acid. Then add a sufficient excess of acid to cause coagulation of the coagulable substances in the fluid, centrifugalize and collect the supernatant fluid; neutralize and add several volumes of alcohol, making a sufficient mixture to fill a 15 ee. test tube. If a precipitate does not separate, let 2 Three successive Type I sputa just received have been positive. PRODUCTION OF PRECIPITIN ANTIGEN 8) the tubes stand for a while; then if a precipitate is formed, cen- trifugalize and decant the supernatant fluid; add 1 cc. of saline and heat to extract the sediment and centrifuge to clear. (If too much saline is added the extract may be too dilute and beyond the reaction stage.) The supernatant fluid is then added to the tubes containing the ‘‘type’’ sera and the tubes are incubated in the water bath. Each tube contains 0.2 ec. of serum and 0.2 ee. of antigenic fluid The rapidity of the reaction varies with the antigenic content of this fluid With a fluid that is rich in anti- gen, there may be seen an immediate clouding followed by a prompt separation of the precipitate, if serum and antigen are mixed in the tube, or by a heavy ring of precipitate if the antigen is carefully “layered” on the serum. With a fluid less rich in antigen, the reaction is slower; with little antigen, it may be absent. Should the finished antigen be thick an attempt may be made to use it after dilution, but this is usually unsuccessful. As is evident the method as far as evolved is not wholly satis- factory, but even in its present state, sputa rich in antigen can be examined and the type diagnosed in from one-half to one hour. Any increase in the number of positive reactions and in the rapidity of the reaction will depend on overcoming the technical difficulties already mentioned. Thus far no false reactions have been encountered unless the incubation was unduly pro- longed. These may be found to occur occasionally because of the presence of much extraneous material in the finished antigen. One fact has impressed us very strongly in carrying out the above work; viz., the very marked difference between bacterial and serum precipitation in regard to the dilution limits of the antigen. A serum antigen—as is well known—may be diluted as much as 1 to 50,000 times and still give a precipitate with a highly potent homologous immune serum. A bacterial antigen must be much more concentrated. The alcoholic precipitate as described when dissolved in twenty times its volume of saline cannot be diluted more than eighty or one hundred times at most. It may be that this is due to the fact that only a small fraction of the bacterial protein is precipitable. Another differ- ence is the extraordinary resistance of the bacterial antigen to 10 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE the action of heat and of chemical agents. Acidification and - boiling are without effect. In fact we have cleared antigens by adding abumin and coagulating it by heat and by acidification and neutralizing the filtrate. SUMMARY A simple method of preparing a precipitin antigen from bac- teria is presented, which allows of the preparation of such anti- gens with a rapidity and in concentration hitherto impossible. An attempt to modify the method to extract antigen from pneu- monic sputum to determine rapidly the type of pneumococcus present has been partially successful. It is presented as a possible basis for further improvements. ON VON DUNGERN’S INDIGO TEST FOR SYPHILIS B. FUJIMOTO From the Forensic-Medical Institute of the Imperial University at Tokio Received for publication, October 10, 1917 Von Dungern’s indigo test for syphilis, which has been de- scribed in the issue of the Miinchener Medizinische Wochen- schrift for September, 1915 (no. 36), was recently examined by Edward P. Flood (1) and it was proved by him that its results are not constant and therefore not practically applicable to the diagnosis of syphilis. A further study of this reaction was undertaken in our labo- ratory to determine more precisely its value. The reagents, which were used in our test, were prepared as follows. Though Flood stated that the preparation of the reagents was not easy and that he was obliged to introduce a modification of von Dungern’s procedure in order to get a clear solution of the in- digo, I was able to prepare them easily in the same manner as did von Dungern, namely: One gram of indigo was carefully triturated and 4 cc. of con- centrated sulphuric acid were added to it. After the mixture had stood for forty-eight hours (the indigo being then com- pletely dissolved), distilled water was added up to 100 cc. A gray sediment, which was caused by unpurity of the material, was thrown down. The solution thus obtained was finally clear and indigo blue in color. The Fehling solution no. 2 was prepared as usual; 173 grams of Rochelle salt were dissolved in 200 ce. of hot distilled water and 73 cc. of sodium hydroxide (specific weight 1.5) were added to this. Distilled water was finally added up to 500 ce. To 1.5 ce. of the indigo-sulphurie acid there were added 10ce. of distilled water and to 4 cc. of this solution there was added 1 cc. of Fehling no. 2. The final solution was greenish yellow 11 12 B. FUJIMOTO and free from any precipitate. This reagent was always used soon after the final mixture was prepared. At first, the test was carried out exactly as described by von Dungern in order to determine whether this test is practically applicable or not. The sera to be tested were obtained from the dermatological clinic and from the medical clinic of our university. The results of these tests with syphilitic and non- syphilitic sera are shown in table 1. In our tests they were con- trolled with the Wassermann reaction. TABLE 1 Dungern’s indigo test SEMI- NO 7 9 : c - REAGENT, 0.2 cc. INACTIVATED SERUM, 0.3 cc. EGAGULATION)| 7 iceant on liconaaniaras, Wassermann ( +) sera..................+--. 1 0 22 Wassermann ( — ) sera.............-...-+-- 1 2 | Almost all of the sera did not coagulate in our tests. Now comparing these results of the indigo test with the normal and the syphilitic sera, we see that the test is not specific, and that there is no difference between them at all. Such a non- specific reaction is evidently unavailable for the clinical diagnosis of syphilis. In order to determine at what point coagulation of sera occurs, the reagent was used in various amounts (0.2, 0.1, 0.05 ce.). The results are as follows. TABLE 2 Coagulation point of sera, to which the reagent was added in various amounts SEMI- NO ACTIVATED SERUM, 0.3 cc REAGENT | COAGULATION | ¢o aGULATION | COAGULATION 0.2 1 0 Peps Wassermann (+) sera......... 0.1 20 3 0 0.05 23 0 0 0.2 1 2 ol Wassermann (—) sera......... { 0.1 29 5 0 0.05 34 0 0 VON DUNGERN’S INDIGO TEST FOR SYPHILIS 13 Almost all sera both Wassermann positive and negative coagu- lated with 0.1 and 0.05 ce. of this reagent. But with 0.2 ee. most of them do not coagulate; that is to say, their heat coagu- lation is prevented by 0.2 cc. of the reagent almost in all cases. When 06.3 ce. of the sera was mixed with 0.2 ce. of distilled water or of physiological salt solution without any of the re- agent the heat coagulation occurred soon after fifteen or twenty seconds. In order to solve the question why the heat coagulation of the sera is prevented by the reagent, I have proceeded to the following tests. The reagent contains indigo, sodium sulphate, sodium and potassium tartrate and sodium hydroxide. Which of these sub- stances plays a part in preventing the heat coagulation of sera; is it the indigo or the salts or the alkali? I. INDIGO The indigo sulphuric acid was neutralized with barium hy- droxide, and we obtained thus a neutral solution of indigo free from sulphuric acid. This neutral solution was concentrated until it possessed the same depth of color as the original indigo sulphurie acid. Indigo solution (0.2 ec.) was added to the syphilitic and non- syphilitic sera in order to see whether the indigo alone has the power to prevent the heat coagulation of sera. The results are shown in table 3. TABLE 3 Tests with pure indigo solution free from acid L g . E a z 9 : b SEMI- NO PURE INDIGO SOLUTION (FREE FROM ACID), 0.2 cc. COAGULATION COAGULATION | COAGULATION Wassermann (++) sera0.3 cc.:............:. 6 0 0 Wassermann (—) sera 0.3 ce................. 4 0 0 Coagulation occurred always rapidly, in fifteen or twenty sec- onds, as if only distilled water had been added, which proved that indigo does not play any part in preventing the heat coagu- lation of sera. 14 B. FUJIMOTO II. COAGULATION TESTS WITH SALTS Comparing the very small quantity of neutral sodium sul- phate with that of the Rochelle salt in the reagent, we can neg- lect the former, because it is neutral salt and does not prevent the heat coagulation; on the contrary, it can even accelerate the coagulation in a small degree. : It was proved in test tubes that the Rochelle salt can acceler- ate coagulation even in the presence of the concentrated sodium hydroxide. Thus it is evident that neither indigo nor salts pre- vent the heat coagulation of sera. Ill. TESTS WITH ALKALI Finally it remains to test the action of sodium hydroxide. The alkalinity of the reagent was titrated. Flood states that 1 ce. of it requires 1.21 ce. of acid for its neutralization, but I have found not only by titrating but also by mathematical cal- culation from the molecular weights of the ingredients of the reagent that 1 cc. of it requires 3.3 cc. to 3.4 cc. of 4 acid for its neutralization. Acidity of indigo sulphuric acid in the reagent: at N x 4/100 x 1.5°/ (10 + 1.5) x 4/441) =—0.154 Ni Alkalinity of Fehling no. 2 in the reagent: (sodium hydroxide specific weight 1.5 = 17 N) 17 N x 73 / 500 x 1/ (441) = 0.496 N Therefore, the alkalinity of the reagent, must be 0.496 N (alkalinity) — 0.154 N (acidity) = 0.342 N (alkalinity) To determine at just what point of alkalinity the sera coagu- late, several Fehling solutions of different alkalinity were so pre- pared as its alkalinity in the test tubes may be each 0.12 N, 0.10 N, 0.08 N, 0.06 N, 0.04 N, 0.02 N. These solutions of different alkalinity were added (a) to the indigo and (b) to the reagent without indigo (viz., to 4 ec. of concentrated sulphuric / VON DUNGERN’S INDIGO TEST FOR SYPHILIS 15 acid there were added 96 ce. of distilled water). The results are shown in table 4. TABLE 4 Coagulation tests with solutions of different alkalinities ALKALINITY N SERA, 0.3 cc. REAGENT, 0.2 cc. 0.12 0.10 0.08 0.06 0.04 0.02 Syphilitic sera me kart Mommas, |e] ft ee ee eee ee ee | ee | Oe Non syphilitic sera Sed 2) (eh eigen aa Esa Mtg Mi eT ae meee er) (ein oa (mn ee eee els (—) = non coagulation. (+) = coagulation. (a) Reagent with indigo, (b) Reagent without indigo. As we see in the above table, each serum has its coagulation point in its 60 per cent concentration (serum 0.3 cc., reagent 0.2 cc.), and we can not find any difference, as to this coagulation point, between, syphilitic and non-syphilitic sera. TABLE 5 ALKALINITY IN TEST TUBES SERUM, 0.3 cc. 0.12 N 0.10 N 0.08 N 0.06 N { No. 1 = = umes fe Wassermann (+) sera INO25.,. + =f Bie +h Nowoes. = == ao + { No. 4 = 2 = si Wassermann (—) sera Nosot aan. er _ — ze 25 | No. 6 _ at se ae (+) = coagulation. (—) = non coagulation. 16 B. FUJIMOTO Now suspecting that not only the alkalinity but also variation in the concentration of the serum might serve to cause varia- tion in the heat coagulation, we have tested the sera in its 75 per cent concentration instead of 60 per cent, viz., to 0.3 ce. of serum was added 0.1 cc. of the reagent. The alkalinity of the reagent used here was so prepared that the alkalinity in the test tubes was 0.12 N, 0.10 N, 0.08 N, 0.06 N respectively. The results are shown in table 5. Comparing these results with those shown in table 4, we see that the more diluted the sera, the more easily their heat coagu- lation can be prevented by alkali. SUMMARY 1. Von Dungern’s indigo test for syphilis is practically un- available for the serum diagnosis of syphilis, because we can not find any difference as to this test between syphilitic and non- syphilitic sera. 2. Among the substances that the reagent contains neither indigo nor salts play any part in the inhibition of the coagula- tion, as von Dungern formerly thought. 3. The coagulation of sera in this reaction is prevented only by the alkali, as Flood proved. 4. The degree of alkalinity that can prevent the coagulation of serum, varies according to the individual, not according to the presence or absence of syphilis. 5. On the other hand, the alkalinity that prevents the coagu- lation of sera varies according to the concentration of the sera; that is, the more diluted the serum, the more easily may its heat coagulation be prevented by alkali. I wish to express my thanks to Prof. Dr. K. Katayama and Prof. Dr. S. Mita for suggesting the subject and for their kind advice. REFERENCE (1) FLoop: Journal of Immunology, 1916, 2, 69. EXTRACTS OF ANTIBODIES OBTAINED FROM SPECIFIC PRECIPITATES OF TYPHOID- ANTITYPHOID SERUM COMPLEX ISRAEL WEINSTEIN From the Laboratory of Bacteriology and Hygiene, New York University, New York City Received for publication October 11, 1917 An investigation was undertaken in this laboratory with a view to precipitating the antibodies from antityphoid sera by means of a specific antigen and subsequently dissociating this antigen- antibody complex, obtaining the antibodies in a solution as free from foreign protein as possible. As is well known, a number of antityphoid sera have been pre- pared and occasionally physicians have reported good results with them, but their use has never been general. In order to be successful with a bactericidal serum it seems necessary to inject large amounts. For example, Cole (1), in the treatment of pneumonia, by using very large quantities of antipneumococcus serum (from 190 to 700 ec.) has had success where others have failed. Owing to the danger of ‘serum sickness’ following the injection of large amounts of horse serum, physicians have hesi- tated to treat their patients with it. Thus, the importance of eliminating as far as possible all extraneous protein and at the same time the harmful effects that it produces, can readily be seen. It was with this aim in view that the experiments described below were begun. | Within the last fifteen years great interest has been shown by in- vestigators in regard to the fate of the antibodies in the precipitin reaction. Gay (2) was the first to demonstrate that complement was fixed by precipitates. He later (3) amplified his earlier studies and concluded that ‘“alexin fixation by a mixture of serum and antiserum is produced by an antigen-antibody complex distinct from precipitin- ily THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 1 18 ISRAEL WEINSTEIN ogen-precipitin but usually brought down by the precipitate in its - formation in such a way as to give the appearance that fixation is produced by the precipitate itself.” Muir and Martin (4), Toyosumi (5) and Zinsser (6) likewise showed that precipitates had complement- fixing properties. In the case of agglutinins, Von Eisler and Tsuru (7) showed that normal hemagglutinins were removed from the serum through precipitation while Landsteiner and Prasek (8) proved that both hemagglutinins and bacterial agglutinins were brought down with the precipitates. Gay and Chickering (9) found that the protective bodies in an antipneumococcus serum were carried down with the precipitate. Other investigators have shown that it is possible to break up the antigen-antibody complexes. Matsui (10) succeeded in extracting bactericidal bodies from cholera vibrios after the former had united with the latter. Muir (11) removed the hemolytic amboceptor from red blood corpuscles, while Landsteiner (12) separated agglutinins from their antigens. Chickering (13) was able to dissociate the pre- cipitates formed by the union of pneumococcus antigen and antiserum and get the antibodies into solution. He found that his extracts con- tained agglutinins and precipitins besides protecting susceptible animals as efficiently as the original serum. EXPERIMENTS AND RESULTS Through the kindness of Professor Park an antityphoid serum having high agglutinating and bactericidal titers was obtained. The serum was taken from a horse that had been immunized with increasing intravenous injections of dead and then living typhoid bacilli at definite intervals for a number of months. In order to determine whether antibodies beside the precipitins were carried down with the precipitate, the original serum, the supernatant serum and the extract obtained from the precipitate were, in each experiment, tested for their agglutinin content. This antibody was chosen in preference to the others because of the ease with which the tests could be made. However, the extract that appeared to be most promising was also tested for bactericidal, complement-fixing and protective bodies. The macroscopic agglutinin test was used. 1 cc. of a twenty-four hour typhoid broth culture was added to an equal amount of the PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 19 serum or extract. The tubes were incubated at 37.5°C. for one hour and then left over-night in the ice-box. Readings were made the following morning. In the preliminary experiments extracts were prepared according to the method that Chickering had found to be best. The precipitates, which had been washed three times with normal salt solution were emulsified in normal saline to which had been added a few drops of a 1 per cent sodium carbonate solution, and were then heated for one hour at 42°C. with occasional shaking. The method which Chickering used in preparing his antigen, namely, shaking in saline pneumococci that had been previously killed with acetone and dried in vacuo, was found to be entirely unsatisfactory in the case of the typhoid bacillus. This was undoubtedly due to the greater resistance of the typhoid bacilli to autolysis. In attempting to get a suitable antigen, the fol- lowing procedures were used: (1) saline suspension of bacteria killed by heating at 54°C. for 1 hour; (2) extracts of bacteria killed with heat or acetone, dried in vacuo over sulphuric acid and shaken in saline or distilled water for lengths of time varying from ten minutes to twenty hours; (3) extracts of bacteria dried, ground in a mortar and then shaken in saline, (4) extracts of bacteria shaken in distilled water from seven to sixteen hours and then allowed to autolyze at temperatures ranging from 37.5°C. to 54°C. for an additional sixteen hours ; (6) extracts of bacteria allowed to remain at 37.5°C. in saline or distilled water for ten days. Whenever distilled water was used the extract was rendered isotonic before adding it to the serum. None of the above antigens gave good results. Fair results were obtained with an extract of bacteria that had been killed with acetone, dried and ground eight hours a day for six days. This antigen removed approximately one-half of the agglutinins of the serum. The best results were obtained with an extract made by digesting the bacilli in a 2.5 per cent solution of antiformin. The bacterial growth on an agar slant in a quart Blake bottle was suspended in 3 cc. distilled water and was added to an equivalent amount of a 5 per cent antiformin solution. This mixture was allowed to remain at room temperature for one hour, after which it was 20 ISRAEL WEINSTEIN centrifugalized at high speed for three-quarters of an hour. The supernatant liquid was rendered neutral by the addition of nor- mal hydrochloric acid. Equal amounts of serum and antigen solution were shaken together and put in the incubator. At the end of two hours a dense precipitate was observed. After re- maining in the ice-box overnight the mixture was centrifugalized. The supernatant liquid was withdrawn, the precipitate was washed three times and then was extracted according to the method described above in alkaline saline solution equivalent to the amount of the serum used. The serum, supernatant and extract were then tested for their agglutinin content. The titer of the serum was 1: 20,000; that of the supernatant was 1: 4000, while that of the extract was 1: 1000. It is thus seen that approximately three-fourths of the agglutinins were removed from the serum, while one-sixteenth of that amount was recovered in the extract. By treating various dilutions of the antigen with normal horse serum as well as with immune serum, it was seen that the antigen acted specifically. At the suggestion of Prof. Charles Krumwiede the antigen was purified in the following manner. First, the antigen solution was precipitated in three volumes of absolute alcohol. After centrifugalization, the precipitate was dried and then taken up in an amount of normal saline solution equivalent to the original volume of the antigen. After being shaken at 37°C. for a few minutes the precipitate went into solution. This purified antigen was found to be just as potent in carrying down agglutin- ins from the immune serum as the untreated antigen. It had the advantage of being rid of the high salt content present in the untreated antigen due to the large amount of alkali in the anti- formin and its subsequent neutralization with hydrochloric acid. DETERMINATION OF THE AMOUNT OF ANTIGEN NECESSARY TO BRING ABOUT THE MAXIMUM PRECIPITATION OF ANTIBODIES In order to determine the optimum amount of antigen for precipitating the antibodies from the serum, various amounts PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 21 of antigen solution and serum were mixed together. The amount of serum remained constant, while the amount of antigen varied from eight times the volume of the serum to one-eighth of its volume. Various amounts of diluted antigen were also used in order to determine whether the process took place better in weak than in concentrated solution. The tubes containing the mixtures were incubated for two hours and then left in the ice-box overnight. Abundant. precipitates appeared in all of the tubes with the exception of those that contained 0.5 ce. and 0.25 ce. of antigen. After centrifugation, the supernatant liquids were pipetted off and titrated for their agglutinin content. In the last column of table 1 the actual titers of the supernatants are given, allowances having been made for the dilution of the serum by the antigen. TABLE 1 Comparative agglutinating power of whole serum and sera treated with various quantities of antigen. Titer of whole serum equals 1: 20,000 ACTUAL DILUTIONS Sie | cenilion| eran OF Pr SE Sa SNE | SERUM USED GEN USED s 8 s s S Ss 8 S 3 3 S S S SOLUTION cc. cc. 1 8.0 +/+/+/+}+|—|/—|—|—|-—|-/—|-| 1: 8000 2 6.0 +I+]+/+]+]— —|—|—|—|—| 1: 6400 3 4.0 +/+]+)/+|+]+]—|—|—|—|-—|—|—] 1: 6000 4 2.0 +/+/+/+|+]+]+/+!-!—|/—I-I-| 1: 8000 5 1.0 2 (1:2) |-+/-+}+/+/+/+/+/+]—!-!-/-l-| 1: 8000 6 0.5 2 (1:4) +/+ /+|+}-+/+/+]+]+/+]—|-|-] 1: 16,000 7 0.25 2 (1:8) | +/+|+)+/+/+]+]+]+]+/+/-|—| 1: 20,000 8 2.0 4 (1:2) |+/+/+/+/+/+]-/-/-]-|-]-]-] 1: 6000 9 2.0 16 (1:8) = |+)+/+/+/-|—|—/|-|-|-|-|-]-| 1: 9000 10 2.0 20 (1:10) |+)+|+/—|—|—|-|—|-|—|-|-]— 1: 8800 The results of this experiment show that little is to be gained by the use of large quantities of antigen. An amount of antigen solution, prepared as described above, equal to that of the serum seems to give as good results as can be obtained. If too little antigen is used there is, as might be expected, little precipitation. The diluted antigens seem to work a little better than the un- diluted ones; e.g., 2 cc. of a 1: 2 dilution reduces the agglutinating Ze ISRAEL WEINSTEIN titer of the serum to the same degree that 2 cc. of undiluted antigen does. DETERMINATION OF THE BEST METHOD OF EXTRACTING THE ANTIBODIES FROM THE PRECIPITATE Effects of alkali The effect of weakly alkaline and strongly alkaline solutions in dissociating the precipitates was tested. Twelve tubes, each containing 2 ec. of antigen and 2 cc. of serum were incubated for two hours and then put in the ice-box overnight. The pre- cipitates were washed. Those in the first seven tubes were suspended in 2 ce. saline with small amounts of a 1 per cent solution of sodium carbonate. The precipitates in the remain- ing tubes were suspended in 2 cc. of a solution of sodium carbon- ate or sodium hydroxide as indicated in table 2. All of the TABLE 2 Comparative agglutinating power of extracts treated with various amounts of alkali. Titer of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 8000 DILUTIONS TUBE ALKALINE SOLUTION AMOUNT pe TIME Se S = |8|8|2 /2/8|8 ce. 46 hour | 1 1.0 per cent Na,CO; 0 42 1 aelsaeeae i= 2 1.0 per cent NasCO; 0.01 42 | ee ee ee 3 1.0 per cent Na,CO; 0.02 42 1 Stas estea| lca at | al el 4 1.0 per cent Na,CO; 0.03 42 1 cote eitea| atte al el 5 1.0 per cent NazCO; 0.04 42 1 +{+/+)4+)—|—)— 6 1.0 per cent Naz:CO; 0.05 42 | a a et feed =| | 7 1.0 per cent Nas,CO; 0.1 42 1 See ase) | 8 1.0 per cent Na,CO; 2.0 42 1 +/+}+/+]—|-—|/— 9 0.5 per cent NaOH 250) 42 1 +/+)+)+ 5 a 10 1.0 per cent NaOH [2228 42 1 +/+/+/+)/4+/—|/— ib 2.5 percent NaOH | 2.0 42 1 j+/+/-|-—|-!-!- 12 5.0 per cent NaOH 2.0 42 | LN | ase fed Ge rd a > tubes were put in a 42°C. water-bath and shaken gently for one hour. The precipitate in tube 12 (5 per cent NaOH) went into solution at once; the precipitate in tube 11 (2.5 per cent PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 23 NaOH) and half of that in tube 10 (1 per cent NaOH) were dissolved at the end of an hour. The small amounts of alkali in the extracts evidently exert no beneficial effect, for the control tube (table 2) shows as high a titer as the others. Larger amounts of alkali are helpful. The results also clearly indicate that a strongly alkaline solution de- stroys the antibodies. Effect of time and temperature Ten tubes, each containing 2 ce. of antigen and 2 ce. of serum, were incubated for two hours and then put in the ice-box over- TABLE 3 Comparative agglutinating power of extracts of precipitates treated with weakly alkaline solutions for various lengths of time at different temperatures. Titer of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 6000 DILUTIONS a = = = = <— TEMPE RA- TUBE ALKALINE SOLUTION AMOUNT eee TIME a : 3 3 3 s ce. OF hours 1 1 per cent Na.CO; 0.05 42 1 oe ee 2S ee ee 2 1 per cent NasCO; 0.05 42 2 eee |e eh 3 1 per cent Na.CO; 0.05 | 49 3 2) tS a a P| 4 | 1 per cent Na,CO; 0.05 42 4 0 |+)+/+/+})-}— 5 | 1 per cent Na,CO; 0.05 ADO Esp ay feta eee eres | a 6 | 1 per cent Na.CO; 0.05 ADs Vt C6) sic Rl al el 7 | 1 per cent Na.CO; | 0.05 54 1 aE | pe : a per cent Na.CO; 0.05 54 2 Se a ee 9 1 per cent Na.CoO., 0.05 OM 24 I+/+)/+ ai alr: 10 | 1 per cent Na,CO. | §60.05 20 240 |+/+)—/-|-|- night. The precipitates were washed three times with normal saline solution and then each was suspended in 2 ce. of saline with 0.5 ce. of a 1 per cent sodium carbonate solution. The tubes were left at various temperatures for varying lengths of time. It can readily be seen from table 3 that practically nothing is gained by incubating the mixture for more than one hour at 42°C. At the end of that time a certain amount of dissociation of the precipitate has taken place. A longer time or higher tem- perature evidently adds nothing. 24 ISRAEL WEINSTEIN Effect of washing the precipitate The next point considered was whether any of the agglutinins were lost in washing the precipitate. Accordingly, the precipi- tates which had formed in four tubes, through the mixture of 2 cc. of serum and 2 cc. of antigen in each, were treated as fol- lows: No. 1 was suspended in 2 ce. of a 1 per cent sodium carbon- ate solution: Nos. 2, 3 and 4 were washed once, twice and three times respectively (2 cc. of normal saline solution being used TABLE 4 Comparative agglutinating power of extracts of washed and unwashed precipitates. Titer of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 4000 DILUTIONS TUBE PREPARATION OF SOLUTIION 2/8/8/8/=|2/8/8/8 1 | Precipitate treated with1 per cent sodium carbo- nate solution +/+)+/+/+/+]+]-|-— 2 | Precipitate, washed once, treated with 1 per cent sodium carbonate solution fff jj | — | — | — 3 | First washing +)+/+/+)4+/—|—-j-|— 4 | Precipitate, washed twice, treated with 1 per cent sodium carbonate solution aac pl bus mailed 5 | Second washing +)/+/—|—|—|-—|-|-|— 6 | Precipitate, washed three times, treated with 1 per cent sodium carbonate solution +/+)+/+)+)—|}—|—|— 7 | Third washing +)/—)}—|-—}|—|-—|-—|-|-— for each washing) and each was then suspended in 2 cc. of 1 per cent sodium carbonate. The extracts and the washings were tested for their agglutinin content. It will be seen from table 4 that there is a reduction in the agglutinin content of an extract of a washed precipitate. We believe this to be due to the fact that a small amount of serum, which remains on the unwashed precipitate, is carried over into the extract. The washings remove all of this serum. In other experiments it was found that when the precipitate was tightly packed in the centrifuge tube and the supernatant serum care- fully removed with a capillary pipette, the titer of the first wash- ing was not above 1:800. It may further be seen from table PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 25 _4 that the titers of the second and third washings are in no way comparable to that of the first. We may conclude that the amount of agglutinins lost through the dissociation of the precipi- tate by washings, when the process is carried on within a few minutes and at room temperature, is negligible. Tests for bactericidal bodies An extract was prepared in the usual way by treating the precipitate with an amount of slightly alkaline saline solution equal to the volume of the original serum. The whole serum, the supernatant and the extract were tested in vitro for their bactericidal power. Method. The three solutions were inactivated by heating at 56°C. for one hour and were then diluted with normal saline. One cubic centi- meter of each dilution was mixed with 1 ce. of a 1: 10 dilution of com- plement and 0.1 cc. of a 1: 2000 dilution of a twenty-four-hour typhoid broth culture. The complement was obtained from the pooled sera of three normal guinea-pigs. The tubes were incubated at 37°C. for two hours, at the end of which time the contents of each tube were plated in agar. The plates were incubated at 37°C. for twenty-four hours and then the colonies were counted. Controls were set up as follows: (1) culture with saline solution, in order to determine the number of bacilli used in each test; (2) culture with complement, in order to see whether the complement of itself had any bactericidal action; and (3) culture with each of the inactivated solutions, in order to see whether all of their native complement had been destroyed. By consulting table 5 it is seen that the whole serum shows a definite bactericidal action in a dilution of 1:40,000, the supernatant in a dilution of 1: 30,000 and the extract in a dilu- tion of 1: 20,000. We may roughly say, therefore, that about half of the bactericidal bodies present in the whole serum are recoverable in the extract. . Tests for complement-fixing bodies The whole serum, the supernatant and the extract were next tested for their complement-fixing power. The antigen pre- NUMBER OF COL- ONIES 3 26 ISRAEL WEINSTEIN TABLE 5 Comparative bactericidal power of whole serum, supernatant and extract | gs 7 | BR Sa Bo ous a B SOLUTION beageieed D Z oR ase ae ae & je} ce ce. @ontrols-eosae ee 0.1 (cout fe Complement controls. . < 0.1 1 Serum control......:.... | “Ot Whole serum 1: 10,000 = Ord Whole serum 1: 10,000 1 0.1 Whole serum 1: 20,000 1 0.1 Whole serum 1: 30,000 1 0.1 Whole serum 1: 40,000 1 | O20 Whole serum 1: 60,000 1 |} On Whole serum 1: 80,000 1 0.1 Whole serum 1: 100,000 1 Supernatant serum con- CLO Sere aye eee | 0.1 Supernatant serum | 1: 5,000 | 0.1 Supernatant serum | 1:5,000 | 1 0.1 | Supernatant serum | 1: 10,000 1 0.1 | Supernatantserum | 1: 20,000 1 0.1 Supernatant serum | 1: 30,000 1 | 0.1 | Supernatant serum | 1: 40,000 1 | 0.1) Supernatant serum | 1: 60,000 1 | 0.1) Supernatant serum | 1: 80,000 1 | 0.1 | Supernatant serum | 1: 100,000 Reig Extract control........ 0.1 | Extract 1 100 0.1 Extract 1: 100 1 | 0.1 | Extract 17; E000) |. wad 0.1} Extract 1: 5,000 il 0.1 Extract 1: 10,000 1 0.1 | Extract 1: 20,000 1 | On! Extract 1: 30,000 1 | 0.1 | Extract 1: 40,000 1 0.1 | Extract 1: 60,000 1 0.1] Extract 1: 80,000 1 0.1 Extract 1: 100,000 1 LO oe ee ge ee eee ae PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 27 pared with antiformin proved to be anticomplementary and therefore could not be used in making the tests. A suspension of ground bacilli was used in its place. Method. In preparing the antigen,! typhoid bacilli that had been erown on salt-free veal agar were suspended in normal saline solution and left in flowing steam in the Arnold sterilizer for one and one-half hours. The mixture was then centrifugalized and the supernatant liquid discarded. The bacteria were washed once with 5 volumes of absolute alcohol, twice with 5 volumes of ether and then allowed to dry at room temperature for forty-eight hours. The dried bacilli were ground in a mortar for one hour. 5 cc. of this bacterial powder were taken up in 55 cc. of normal saline solution. This suspension was used as antigen. 0.1 cc. of 1:50 dilution of the antigen, which constituted 2 units, was used in each test. For complement, the sera from ten normal guinea-pigs were pooled and 0.1 cc. of a 1:10 dilution of this serum was used in each test. This represented 1# units of complement. Dilutions of the serum, supernatant and extract were made and de- creasing doses of each dilution from 0.1 cc. to 0.01 cc. were put into tubes. To these, 0.1 cc. of antigen and the same amount of comple- ment were added. The volumes of the mixtures were equalized by the addition of normal saline solution. In each rack there was one tube which contained 0.2 cc. of antigen and 0.1 cc. of complement, with no serum. This constituted the antigen control. There were also two other tubes which contained 0.04 and 0.02 cc. serum respectively and 0.1 cc. of complement with no antigen. These served as the serum controls. After incubating the tubes for thirty minutes in a 37°C. water-bath, the hemolytic system was added. This consisted of 0.1 ce. of a 5 per cent suspension of washed sheep corpuscles plus two units of hemolysin. The unit of hemolysin was found to be 0.04 ce. of a 1: 6000 dilution. “The tubes were then put in the water-bath for an additional ten minutes, at the end of which time the results were read. In titrating the various solutions, the whole serum and supernatant were used in dilutions of 1: 10 and 1: 100; the extract, undiluted and in dilutions of 1:10 and 1:100. In the case of the supernatant serum, the serum controls were found to be anticomplementary to a certain extent, due undoubtedly to the fact that it contained some of the 1 This antigen was prepared in accordance with directions given by Miss M.A. Wilson of the Research Laboratories of the New York City Health Department. 28 ISRAEL WEINSTEIN “antiformin” antigen in it. This must be taken into consideration in interpreting the results. It would seem from table 6 that the extract had about one- tenth the complement-fixing power of the serum. Animal experimentation Historical note. No true typhoid infection can be produced in labo- ratory animals, with the possible exception of the anthropoid apes. The inoculation of small amounts of stock culture has little, if any, effect. The use of Jarge quantities, especially of a culture whose viru- TABLE 6 Comparative complement-fixing power of whole serum, supernatant serum and extract DILU- SOLUTION TION CONTROLS RESULTS OF TESTS 1:10 | Hemolysis 0.001 cc. Complete fixa- Whole serum. . tion 1: 100| Hemolysis 0.0008 cc. Weak fixation 1:10 | 0.04 Strong fixation 0.005 ec. Strong fixation 0.02 Weak fixation Supernatant Se en ear 1: 100} 0.04) 0.002 cc. Weak fixation Trace 0.02 f : Un- | Hemolysis 0.01 ce. Complete fixation dilut- Extract: = 25:2: ed 1:10 | Hemolysis 0.002 cc. Weak fixation 1: 100| Hemolysis 0.001 ec. Weak fixation lence has been increased by successive passages through animals, is followed by death. That death is due to intoxication was proved by Sirotinin (14), who showed that the animals succumbed to doses of dead as well as living bacilli. Investigators have reported definite lesions in animals that have died following an injection of typhoid culture. There is usually a congestion of the abdominal organs, especi- ally of the spleen, liver, kidney and intestinal lymph nodes. But Beumer and Peiper (15) have shown that these lesions are not specific. They were able to produce the same lesions through the injection of non-pathogenic soil and water bacteria. Metchnikoff and Besredka - Sil Wa. = PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 29 (16) have shown that a true typhoid infection can be produced in the _ higher monkeys. In the case of fifteen chimpanzees and one gibbon, that were treated with mixtures of pure typhoid culture and the fecal material of typhoid patients, they had only one negative result. Chan- temesse and Widal (17) increased the virulence of their culture by in- jecting it into guinea-pigs simultaneously with large doses of killed streptococci. They passed the culture from animal to animal, using decreasing doses of streptococci, until the virus acquired fixed characteristics. Protective power of the extract The object of the first experiment was to determine whether the virulence of the stock typhoid culture could be increased by following the method of Chantemesse and Widal (17), ice., by the simultaneous injection of typhoid and killed streptococ- cus cultures. Accordingly three guinea-pigs, were inoculated as follows: No. 1. 4 cc. of typhoid culture subcutaneously and 8 cc. of streptococcus vaccine intraperitoneally. No. 2. 4 ec. of typhoid culture subcutaneously. No. 3. 8 cc. of streptococcus vaccine intraperitoneally. No. 1 was dead within less than eighteen hours. No: 2 died after forty-eight hours. No. 3 survived. Cultures taken from the peritoneum and heart of no. 1 were positive. 1 cc. of the peritoneal exudate was mixed with 5 cc. of broth and incubated for five hours. Guinea-pig 4 was given 3 cc. of this mixture subcutaneously and at the same time 7 ce. of streptococcus vaccine intraperitoneally. The animal died within fourteen hours. Cultures from the heart and peritoneum were found to be positive. The culture obtained from the heart was inocu- lated into broth. Guinea-pig 5, which received 2 cc. of this broth culture intraperitoneally, became very sick within six hours after the injection. A general paralysis set in which was soon followed by death. The bacilli were obtained in pure cultures from the peritoneal exudate and heart blood. It was this virulent culture which was used in the later experiments. Eight guinea-pigs were each given intraperitoneally 2 cc. of culture. Three of them received in addition 1, 2 and 3 ce. 30 ISRAEL WEINSTEIN respectively of serum, while three others received equivalent amounts of extract. The two control animals died within twenty-four hours after the injection. 1 cc. of the extract, as well as 1 ec. of the serum, protected against a fatal dose of culture. In the case of mice, the minimal lethal dose was found to be 0.2 cc. injected intraperitoneally. 0.2 cc. of the extract, as well as 0.2 cc. of whole serum, was found to protect mice against the minimal lethal dose of culture. Kjeldahl determinations? Determinations of the nitrogen content of the whole serum and of the extract were made. The whole serum was found to contain 1.064 grams of nitrogen per 100 cc. The content of the extract varied from 0.018 to 0.028 grams per 100 cc. In other words, the nitrogen content of the serum was reduced to from 3: to = Of its original amount. DISCUSSION OF RESULTS In attempting to obtain an antigen that would precipitate antibodies from an antityphoid serum, it was soon found that - methods that gave good results with micrococci were of no use whatever with typhoid bacilli. Micrococeci autolyze rapidly in distilled water and even in normal saline. Typhoid bacilli, on the other hand, disintegrate very slowly. The best method of preparing a typhoid antigen appeared to be the digestion of the bacteria with 23} per cent antiformin and the subsequent neutralization of the solution with hydrochloric acid. Such an antigen not only produced an abundant precipitate when mixed with antityphoid serum, but also carried down with the precipi- tate antibodies other than the precipitins. In order to free the antigen of its high salt content and of any free chlorine that might have been present, it was precipitated with three volumes * These determinations were kindly carried out by R. L. Kahn of the Monti- fiore Home and Hospital Laboratory, New York City. PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 31 of absolute alcohol, the precipitate being subsequently redis- solved in normal saline solution. An equal amount of antigen solution and immune serum gave the best results. Larger amounts of antigen did not bring down greater quantities of antibodies from the serum. This is in keeping with the results of Dean (18) who found that the maxi- mum precipitate was obtained by combining equivalent pro- portions of antigen and serum. The mixture of antigen and serum produced the precipitation of large quantities of agglutinins, bacteriolysins, complement- fixing and protective bodies. Landsteiner and Prasek (8) found that bacterial agglutinins were carried down with the precipitate. Chickering (13) found the same thing in the case of the precipi- tates formed from mixtures of pneumococcus antigen and serum. That complement-fixing bodies are brought down with the pre- cipitate has been shown by the work of Gay (3) and Zinsser (6). The removal of protective bodies from an immune serum was demonstrated by Gay and Chickering (9) and later by Chickering alone. It was possible to dissociate the antigen-antibody complex to a certain degree and to obtain a portion of the antibodies in solution (about 5 per cent of the agglutinins, 50 per cent of the bacteriolysins and 10 per cent of the complement-fixing bodies that were present in the original serum.) Chickering had a similar experience. He found that while it was possible to remove most of the antibodies with a single extraction, a residue alwaysremained. This residue could not be removed by repeated extractions. Slightly alkaline solutions were found to be best for extraction. Strongly alkaline solutions would dissolve the precipitate and would also destroy the antibodies. A temperature of 42°C. was found to be necessary. Room temperature was entirely un- satisfactory. Muir (11) observed the same thing while attempt- ing to dissociate hemolysin from red blood corpuscles. Because a comparatively small proportion of the agglutinins are recoverable, it does not necessarily mean that the same thing is true of the other antibodies. Chickering, for example, 32 ISRAEL WEINSTEIN found that his extracts contained practically all of the pro- tective bodies that were present in the original serum but only a small part of the agglutinins. The work of a number of investigators shows that animal experimentation in the case of typhoid fever is a rather uncertain procedure, inasmuch as no true infection can be produced in ‘laboratory animals, with the possible exception of the anthropoid apes. The death of animals inoculated with typhoid is most probably due to intoxication. If large amounts of the culture are given, especially if the virulence of the culture has been previously raised by passage through animals, there is an invasion of the blood stream and probably an increase in the number of bacteria. The method of obtaining a virulent culture, described by Chantemesse and Widal (17), namely, the simultaneous inoculation of typhoid bacilli and streptococcus vaccine, was found to be effective. The enormous reduction in the protein content of the serum that is possible by extracting the antibodies, compensates for the antibodies that are lost. Such a solution as the extract, because of its low protein content, could be injected into typhoid fever patients without running the risk of having the serious reactions on the part of the patient that often follow the injec- tion of the whole serum. CONCLUSIONS 1. An extract of typhoid bacilli, formed by digesting the bacteria in 2.5 per cent antiformin and neutralizing the solution with normal hydrochloric acid, produces a voluminous precipi- tate when mixed with antityphoid serum. 2. An equal amount of such an antigen solution and serum produces the maximum precipitate. 3. Such precipitates contain not only precipitins, but also agglutinins, complement-fixing, bactericidal and protective bodies. 4. About 5 per cent of the agglutinins, 50 per cent of the bac- tericidal bodies and 10 per cent of the complement-fixing bodies present in the original serum can be extracted from the precipi- tate in a slightly alkaline solution at 42°C. a a PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 33 5. The extract contains from 3 to 39 the amount. of nitrogen that is present in the whole serum. The nitrogen content of the extracts having been reduced on the average to ;, the amount present in the whole serum, the agglutinins were concentrated approximately twice, the bactericidal bodies twenty times? and the complement-fixing bodies four times. 6. By passing stock cultures of typhoid bacilli through guinea- pigs, whose resistance is lowered by the simultaneous injection of streptococcus vaccine, it is possible so to increase the virulence of the culture that small doses will kill mice and guinea-pigs in a short time. 7. 1 ce. of extract will protect guinea-pigs against 2 cc. of typhoid bouillon culture, a fatal dose. 0.2 cc. of extract will protect mice against 0.2 ec. of culture, which is likewise a fatal dose. REFERENCES (1) Coxtz, R.: Jour. Am. Med. Assn., 1913, 61, 663. (2) Gay, F. P.: Centralbl. f. Bakteriol., 1905, 39, 603; Ann. de l’Inst. Pasteur, 1905, 19, 593. (3) Gay, F. P.: Univ. Calif. Pub. Path., 1911, 2, 23. (4) Muir, R. anp Martin, W. B. M.: Jour. Hyg., 1906, 6, 265. (5) Torosum1, H.: Centralbl. f. Bakteriol., 1908-1909, 48, 325. (6) Zinsser, H.: Jour. Exper. Med., 1912, 15, 529. (7) Von Etstmr, M. anp Tsuru, J.: Ztschr. f. Immunitiitsforsch., 1910, 6, 608. (8) LANDSTEINER, K. anp PraseEk, E.: Ztschr. f. Immunitiitsforsch., 1911, 10, 68. (9) Gay, F. P. anp CuHIcKERING, H. T.: Jour. Exper. Med., 1915, 21, 389. (10) Marsut, I.: Ztschr. f. Immunititsforsch., 1914-1915, 23, 233. (11) Murr, R.: “‘Studies on Immunity,’’ London, 1909, p. 12. (12) Lanpsrriner, K.: Miinchen. med. Wehnschr., 1902, 49, 1905. (13) Curckerine, H. T.: Jour. Exper. Med.; 1915, 22, 248. (14) Srrotinin, W.: Ztschr. f. Hyg., 1886, 1, 465. (15) Brumer anp Perper: Centralbl. f. klin. Med., 1886, 7, 633. (16) Mretcuntixorr, E. anp BrsrepDKA, A.: Ann. de l’Inst. Pasteur, 1911, 25, 193. (17) CHANTEMESSE AND Wrpat: Ann. de l’Inst. Pasteur, 1892, 6, 755. (18) Dean, H. R.: Proc. Roy. Soc. Med., Path. Sect., 1912, 5, part 3, p. 62. * We do not believe that a very good method for testing bactericidal bodies has yet been devised. The plate method, used in our experiments, is not suited for accurate work. THE JOURNAL OF IMMUNOLOGY, VOL. II, NO. 1 THE SPECIFICITY OF INTRACUTANEOUS ABSORPTION G. H. SMITH anv M. W. COOK From the Research Laboratories of the H. K. Mulford Company, Glenolden, Pennsylvania Received for publication October 17, 1917 In previously reported experiments (1) we have shown that the absorption of antigen from the cutaneous tissues of specifi- cally immunized animals proceeds at a rate markedly in excess of that occurring in normal animals. Horse serum, injected in- tracutaneously into guinea-pigs immunized against horse serum, disappeared from the site of injection much earlier than did horse serum similarly injected into normal pigs. The question was not determined, however, as to whether the heightened reactivity of the tissues, attendant upon the immune state, pertained to the specific antigen only or whether it indicated an activation of a general mechanism of elimination extending to non-specific antigens. The experiments here reported deal with this point. In the present work guinea-pigs immunized to one antigen were tested by the intracutaneous injection of the specific and also of a heterologous antigen to demonstrate to what extent absorption depends upon a specific factor. The technic employed was much the same as that already described. In several particulars, however, it was amplified so that a better check on the results might be obtained. Guinea- pigs immunized to normal horse serum were used with normal guinea-pigs as controls. As antigens for intracutaneous in- jection we employed both horse and goat serum, the former derived from a horse immunized to B. dysenteriae (Y-Hiss), and the latter serum secured from a goat immunized to the gonococcus. With such antigens check titrations could be made. The antidysentery horse serum could be detected by the precipitin 35 36 G. H. SMITH AND M. W. COOK f test with rabbit antihorse serum and by the demonstration of the dysentery agglutinins. Similarly, by the use of a rabbit anti- goat serum and the agglutination test with the gonococcus the presence of goat serum could be demonstrated. Not only were these tests made upon the tissue extracts prepared from the places of injection but titrations were also made by both tests upon the sera of the injected animals to determine the amounts of antigen absorbed into the circulation. Thus a measure was secured of both the rate of disappearance of antigen from the skin and the rate of appearance of antigen in the circulation. The table, which follows, gives the results obtained by the use of these titration procedures in several preliminary tests upon immunized and normal guinea-pigs, all of which received for the intracutaneous injections horse serum, the antigen speci- fic for the immunized guinea-pigs. It is evident from the data given that precipitin and agglutination tests are satisfactory means for measuring the transference of antigen. In addition, the data show the marked difference between normal animals and immunized animals. TABLE 1 The absorption of specific antigen IMMUNIZED PIGS NORMAL PIGS HOURS Tissue | Serum Tissue | Serum a. Precipitin tests 6 1: 25,600 1: 1,600 1: 25,600 ee LOO 12 1: 6,400 1: 6,400 1: 12,800 1: 1,600 24 1: 2,400 1: 9,600 1: 6,400 1: 3,200 48 1 400 1: 4,800 1: 2,400 1: 600 b. Agglutination tests 6 1: 12,800 1: 1,600 1: 12,800 Ls 150 12 1: 9,600 1: 12,800 1: 12,800 1: 2,400 24 1: 2,400 1: 12,800 1: 6,400 1: 8,000 48 1: 15200 1: 12,800 1: 1,200 1: 1,600 Examination of the figures given above shows that the agglu- tination titrations run parallel with the precipitin tests. Not only is this true of the tests conducted upon the tissue extracts, a & SPECIFICITY OF INTRACUTANEOUS ABSORPTION Si but also of those in which the serum of the injected animals was used. In addition, precipitin and agglutination tests were made upon the serum of both normal and immunized guinea-pigs, which had received no intracutaneous injections of antigen, in order to demonstrate normal agglutinins, if present, and to detect any horse serum remaining in the circulation from the immunizing treatment. Normal agglutinins were not present in the serum in amounts capable of having an appreciable bear- ing upon the titrations to be made upon the tissues and sera of the injected animals, nor was there a significant residue of horse serum remaining from the immunizing injections, since a positive reaction with a potent rabbit antihorse serum was never obtained in dilutions greater than 1: 200. Tissue extracts from both the normal and the immunized pigs, in the dilutions tested—1 : 50— failed to show either normal agglutinins for B. dysenteriae or precipitable horse serum. TABLE 2 The transference of precipitable horse serum. Precipitin tests HOURS | DISAPPEARANCE OF HORSE APPEARANCE OF HORSE CHANGE IN PRECIPITIN SERUM FROM THE TISSUES SERUM IN THE SERUM VALUE, GUINEA-PIG SERUM 6 1: 12,800 1: 1,600 1: 3,200 12 1: 6,400 1: 6,400 1: 1,600 24 1: 1,600 1: 12,800 1: 600 48 1: 1,600 1: 6,400 1: 400 Tests upon the sera of the immunized pigs showed that a certain amount of precipitin had been elaborated in the course of the immunizations. In this connection it is interesting to note that the precipitin of the serum decreased as the absorption of antigen proceeded. This point is illustrated in the following table which gives the results secured in one series of titrations made upon pigs immunized to horse serum. That the greater part of the intracutaneously injected serum passes into the circulation and is not immediately taken up by the tissues, cutaneous tissue being the index, is shown by the following. At the time of removal of the portion of skin into 38 G. H. SMITH AND M. W. COOK which the injection was made a section of uninjected skin from the other side of the animal was removed and extracted. Titra- tions conducted upon these extracts in parallel with extracts from the injected tissue gave values as indicated in table 3. TABLE 3 The distribution of absorbed antigen in immunized pigs a HORSE SERUM IN THE HORSE SERUM IN THE HORSE SERUM IN THE ELORIEE INJECTED TISSUE UNINJECTED TISSUE CIRCULATION a. Precipitin tests 6 1:12,300 | 1: 200 1: 1,600 12 1: 6,400 | 1: 400 1: 6,400 24 1: 2,400 1: 400 1: 12,800 48 1: 1,000 1: 400 1: 6,400 b. Agglutination tests 6 1: 12,800 1: 400 | 1: 1,600 12 1: 12,800 1: 200 1: 6,400 24 1: 3,200 1: 1,600 1: 12,800 48 it 800 1: 1,600 1: 12,800 The facts made evident in tables 1, 2, and 3 indicate clearly that the process of immunization has developed a mechanism whereby the specific antigen is removed from the cutaneous tissues to reappear in the serum at a rate far in excess of that occurring in normal animals. This conclusion is based solely upon the ability of normal and specifically immunized guinea- pigs to absorb horse serum. Before continuing the work ‘by introducing a second, heterol- ogous, antigen it was necessary to determine whether or not there exists in normal animals a selective action for a particular antigen. Accordingly an experiment was conducted in which normal pigs were injected intracutaneously with the antigens previously mentioned, that is, with antidysentery horse serum on one side of the body and with antigonococcus goat serum on the other. In making the intracutaneous injections every pre- caution was taken to insure the retention of equal amounts of each antigen by the tissues and an effort was made to introduce the needle at the same depth into the skin. The general ap- SPECIFICITY OF INTRACUTANEOUS ABSORPTION 39 pearance of the elevation, which remained after injection and the removal of the needle, indicated that the procedure was earried out with considerable uniformity. In the table which follows data are given showing the reactions secured with normal pigs. TABLE 4 The absorption of horse serum and goat serum in normal animals a. Precipitin tests HOURS HORSE SERUM IN THE SKIN GOAT SERUM IN THE SKIN 6 1: 12,800 1: 12,800 12 1: 6,400 . 1: 6,400 24 1: 2,400 Lid;200 48 1: 2,400 1: 2,400 HORSE SERUM IN THE CIRCULATION GOAT SERUM IN THE CIRCULATION 6 is 150 ta 200 12 1: 1,600 1: 3,200 24 1: 3,200 1: 3,200 48 1: 600 I<. 4 000 b. Agglutination tests DYSENTERY AGGLUTININS IN THE SKIN | GONOCOCCUS AGGLUTININS IN [HE SKIN 6 1: 12,800 1: 12,800 12 1: 12,800 1: 12,800 24 1: 6,400 1: 4,800 48 200 1: 1,600 DYSENTERY AGGLUTININS IN THE GONOCOCCUS AGGLUTININS IN THE CIRCULATION CIRCULATION 6 1 100 1 150 12 1: 1,600 1: 1,600 24 1: 8,000 1: 9,600 48 1: 1,600 1: 2,400 This experiment, as well as those that follow, was repeated several times, in each case duplicate animals being used upon each time interval. Minor variations occurred which could readily be ascribed to errors in technic, particularly in the intra- cutaneous injection, but the general type of reaction remained 40 , G. H. SMITH AND M. W. COOK constant with each of the factors used. A study of the results shows that the elimination of horse serum from the skin of nor- mal animals is not more readily brought about than the elimina- tion of goat serum. In fact, if curves based upon an average of the several animals are plotted for each of the several factors titrated, it is found that they coincide almost exactly through- out their entire course. We therefore felt secure in assuming that the normal mechan- ism of eliminating foreign protein from the skin operates equally well with either the horse or goat serum. This gave us sufficient warranty for attempting to detect a specificity of elimination due to an alteration in this mechanism brought about through immunizing treatment. To this end a series of guinea-pigs was immunized by repeated injections of horse serum. ‘Titrations, similar to those indicated above, were then made upon those animals, a parallel series of normal pigs being tested as controls. The results of such a test may be tabulated as follows. . But one conclusion can be drawn from this work, namely, that the immune state so alters the process of antigen absorp- tion from the skin that the specific antigen is removed more readily than in the normal state. Moreover, it is of especial interest that in an immunized animal the non-specific antigen is not eliminated as rapidly as is the same antigen from normal tissues, for in the animals immunized to horse serum the rate of dis- appearance of goat serum was delayed over that found in normal pigs. It would seem, therefore, that in immunized animals, a specific mechanism has been developed at the expense, in some measure, of the normal process. To test this point in partic- ular many repetitions of this experiment were made and in every case such a relationship appeared. Throughout the work with this series of animals it was re- peatedly noted that in the course of the preparation of the tissue extracts the portions of tissue removed from the immu- nized animals differed greatly in appearance. Those sections into which horse serum, the specific antigen, had been injected were much reddened and somewhat thickened, while those from SPECIFICITY OF INTRACUTANEOUS ABSORPTION 41 the same pig into which goat serum had been injected remained free from thickening and inflammation. A similar condition has also been noted in connection with our work upon the absorp- tion of antigen from the cutaneous tissues of sensitized pigs. TABLE 5 The absorption of horse serum and goat serum in immunized and normal animals a. Precipitin tests IMMUNIZED PIGS NORMAL PIGS Horse serum in skin | Goat serum in skin | Horse serum in skin | Goat serum in skin 6 1: 12,800 1: 12,800 1: 12,800 1: 12,800 12 1: 6,400 1: 12,800 1: 6,400 1: 6,400 24 600 1: 6,400 Be 837400) 1: 3,200 48 Lg 400 1: 6,400 1: 2,400 1: 2,400 Horse serum in Goat serum in Horse serum in Goat serum in circulation circulation circulation circulation 6 1: 1,600 Li: 50 Le 200 1 150 12 1: 6,400 is 100 1: 1,600 1: 2,400 24 1: 9,600 1 100 ee 33200 1: 2,400 48 1: 4,800 1 200 Ais 800 1: 1,600 b. Agglutination tests | Dysentery agelutinins Gonococcus Dysentery agglutinins Gonococcus in skin agglutinins in skin in skin agglutinins in skin 6 1: 12,800 1: 12,800 1: 12,800 1: 12,800 12 1: 6,400 1: 12,800 1: 12,800 1: 9,600 24 1: 1,600 1: 6,400 1: 6,400 1: 4,800 48 1 800. 1: 6,400 1: 1,600 1: 1,600 Dysentery agglutinins} Gonococcus agglu- |Dysentery agglutinins! Gonococcus agglu- in circulation tinins in circulation in circulation tinins in circulation 6° 1: 1,600 None 1 100 Rs 200 12 1: 2,400 Le 200 1: 1,600 1: 2,400 24 1: 9,600 1: 1,000 1: 6,400 1: 4,800 48 1: 12,800 1: 3,200 1: 2,400 1: (3,200 The specific antigen here, also, brought about inflammation such as never followed the injection of a non-specific antigen. Moreover, this reaction could not be due to an inherent primary toxicity, for the injection into normal pigs resulted in no reaction characteristic of either the horse or the goat serum. 42 G. H. SMITH AND M. W. COOK It is an established fact that the process of immunization produces an altered reactivity of the body cells. Experimenta- tion upon the reactions of the cutaneous tissues indicates that this change is specific and produces no effect upon the reacting properties of the cells for other antigens, except perhaps to lessen their activity. From the point of view of absorption of antigen, then, the immune state with the changes dependant upon it, is the result of a heightened reactivity for the specific antigen only, and does not stimulate the mechanism of elimina- tion of heterologous antigens. REFERENCE (1) Smrra AND Coox: J. Immunol., 1917, 2, 269. SPECIFIC REACTIONS OF THE BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS! G. R. LACY anp C. C. HARTMAN From the William H. Singer Memorial Research Laboratory, Pittsburgh, Pennsylvania Received for publication November 17, 1917 The problem of the reaction of the body fluids in pneumo- coccus infections has claimed our interest for some time. To- gether with the grouping of the pneumococci which was begun in our laboratory in the fall of 1916, the following studies were made to determine if possible, the extent and the way in which the body fluids reacted to the pneumococci and to antipneumo- coecic sera. The materials examined were obtained from pa- tients admitted to the Allegheny General Hospital. Sputum from persons presenting clinical manifestations of pneumonia were sent to the laboratory at the earliest opportunity and the erouping made according to the method of Dochez and Gillespie (1). When the type of pneumococcus to which the infection belonged had been determined, tests were begun for the purpose of demonstrating any defensive antibody formations that might be present in the body fluids. For this purpose, examinations were made of the blood serum, spinal fluids, urine, pleural fluid and, in those instances where autopsies were held, pericardial fluids. These fluids were tested for agglutinins, precipitins, precipitable substances, etc. The sera, urines and other materials were collected at varying intervals with reference to the time of onset of the infection and to the condition of the temperature curve. From some of the patients we obtained these materials as early as the second day of the illness and from then repeatedly until complete recovery. In some instances where death occurred early, only a single test 1 Read before the Pittsburgh Academy of Medicine, April 24, 1917. 45 G. R. LACY AND C. C. HARTMAN 44 a 0 0 0 0 |++++l0 0 0 AI dQouop 0 0 0 ++++[t+++4+]+++4]+4+4++]4++4+4+ SIoquINU e41Nz[NP) o|o |++++]++++)++++\++++4]++++4] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 t+++tl+t4+tl[++t+]+++4t] T pur al 0) 0 0 0 0 +++4+[44++4]4++4+4]4+4+4+4+ [pus Tie 1G Paech 1a! 68 €€ ONALLVd JO uqddWoNn BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 45 was made. In all instances where we could get sufficient ma- terial, the agglutination tests were made against at least four different cultures of types I, II and IV and one culture of type III. We invariably used the organism isolated from the pa- tient whose serum was being tested. These multiple cultures were used with the hope of finding one or more cultures that the serum would agglutinate. The method of procedure which we used is very well illus- trated in table 1, wherein the reactions of a typical example of each group except group II is shown. In this table the number of each culture corresponds to the number of the patient from whom it was isolated; 1.e., ‘‘culture 33” was isolated from “‘ patient 33,’ ‘‘19” from “‘patient 19,” ete. The patients were grouped according to the type of organism isolated from them. Patients 33 and 19 were typical group I and group II patients respectively and their sera agglutinated all cultures belonging to their respective groups and no others. All of the other patients were atypical as shown by the table. Pa- tients 11, and 12 and 25 were especially interesting in that each had two separate and distinct attacks of pneumonia before leav- ing the hospital and that each recovered fully from both attacks. These two patients we will take up somewhat in detail. Patient 11 had as his first infection a member of group IV and as his second a member of group I. He was exposed to a group I infection by having occupied a bed in the ward on two occasions of some days duration, adjacent to a patient from whom a pneumococcus, type I, had been isolated. His serum, as will be seen in the table, agglutinated all group I organisms or those corresponding to his last attack, and did not agglutinate any of the group IV organisms. Unfortunately his first culture had been lost before this phase of the work was taken up, so we do not know whether his serum would have agglutinated it or not. Patient 12 and 25 had as his first infection a member of group II and as his second a member of group I. We have no definite history of such intimate exposure in this instance, as in the case of patient 11, but he was kept in a large ward with patients suffering from infections with the other types of pneu- 46 G. R. LACY AND C. C. HARTMAN mococci. His serum showed agglutinins only for organisms be- longing to group II, or those corresponding to his first infection, and none for those belonging to group I—not even for his own last culture, no. 25 of table 1. We believe that if one were fortunate enough to follow a sufficiently large number of reinfec- tions or recurrences, much more light might be thrown upon the immunity or lack of immunity established by infections with pheumococci. Table 2 shows the results of the agglutination tests carried out with the sera from fifty patients. TABLE 2 GROUP NUMBER POSITIVE pete | Di cioiete aban soe md Sargon Seek ee © eee eee 21 15 71.4 Dc Ackte ster ooo 2.6 sc as ee OE ee ene 13 10 77.9 | Ie te de PE re me ree yer ns At ¢ 0 0 0 DIV 5 POE IS. al dled oe, See eee ee 10 5 50.0 Unclassified (clinically pneumonia)..... 6 0 0 Motels as. ccs pee ee ee eee 50 30 60.0 This differs somewhat from the report by Chickering (2) who found specific agglutinins in 100 per cent of no. I’s, 53.8 per cent of no. II’s and 55.5 per cent of no. IV’s. Of twenty-one patients with no. I infections six did not show specific agglutinins in their sera. Two of these died and four recovered; one of the ones who recovered had a blood stream infection and another an empyema necessitating a stay of some months in the hospital. Only two patients with fatal infections showed agglutinins in their sera and both had received anti- pheumococcic serum for group I before the test was done. Of the three patients with no. II infections who did not show spe- cific agglutinins in their sera—one ran a very mild course; an- other, a small child, developed an empyema from which it re- covered after many weeks; and the other one died. Of the five patients with group IV infections, showing no specific agglu- tinins in their sera, one had a very mild infection, two developed empyema and two died. BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 47 In the thirteen patients having infections with organisms of group II, whose sera were tested, there were three sera which showed that the infection was caused by organisms belonging to sub-group II, one to sub-group IIx, and the other two to sub- group Ila or IIb. In regard to the appearance and duration of the agglutinins it may be said that in all instances in which the serum was ob- tained during the fastigium the agglutinin tests were negative, except in the two patients mentioned above who had received antipneumococcie serum therapeutically. The agglutinins ap- peared only during or after defervescence. As would be ex- pected, the time during which they were demonstrable varied greatly. There was an apparent tendency for the agglutinins for organisms of group I to persist for a longer time than for the other groups (II and IV), and for those for group II to persist longer than those forgroupIV. We were unable to test any for group IIT. As has been noted before (Chickering (2), Lister (3) ) the agglutinins in instances of human infections correspond with the experimental findings resulting from immunization of ani- mals against the various groups of pneumococci in being spe- cific for the organism or group to which the infecting organisms belongs. That is, a patient having an infection with a group I organism shows agglutinins only for the organisms in that group (or sub-group); a patient having an infection with an or- ganism belonging to group IV possesses agglutinins only for his homologous organism and no other. We have tested a number of urines from patients whose sera gave strongly positive reactions and were unable to demonstrate at any time the presence of agglutinins. Agglutinins were not demonstrable in the spinal fluid in those patients having a meningitis, except in two who had received antipneumococci serum both intravenously and intrathecally. In several of the patients with meningitis we were unable to carry out aggluti- nation tests with their sera, so that we cannot say whether agglutinins were present or not in their blood. All of these patients died, so that/as a group they were not favorable for the development of agglutinins. 48 G. R. LACY AND C. C. HARTMAN In carrying out agglutinin tests on spinal fluids which con- tained a great number of pneumococci we noted that, upon the addition of the antipneumococcic serum, a precipitate was formed almost immediately. We did not appreciate the significance of this reaction until Blake (4) reported his observations upon the precipitin reaction on the urine. Since this precipitin reac- tion offered a rapid and accurate method of diagnosis, we took advantage of it and found it to be a reliable test, if a sufficient number of dilutions of both no. I and no. II sera were used. In order to explain our findings in the tests on spinal fluids and urines we made the following cross tests. Cultures were made in broth and after sufficient growth had occurred we centrifu- galized them and pipetted off the supernatant, clear fluid. Serum for groups I and II were added in varying dilutions to the spinal fluid, urine and broth separately. Precipitates oc- curred in all three with the corresponding serum. Mixtures of spinal fluid and urine, spinal fluid and broth and urine and broth showed no precipitates. Therefore, the precipitating sub- stance, ‘‘precipitin,’’ was shown to be present in the serum and precipitable substance ‘‘precipitinogen’”’ in the urine, spinal fluid and broth. We would advise, therefore, as a point of practical value—that, in cases of meningitis of indefinite etiology, a precipitin test with antipneumococcic serum be made as a rapid method for determining or excluding a pneumococcus infection with either group I or group II.. DISCUSSION The agglutinin tests on the sera of fifty patients showed that humans react to their acquired infections with the pneumo- coccus in a specific manner similar to that experimentally pro- duced in animals; i.e., the patient produces agglutinins only for the homologous organism or group to which the homologous organisms belongs. We did not obtain, however, as high a percentage of positive results as Chickering in the instances of group I infections. The very mild infections and the fatal infections were less apt to show specific agglutinins. BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 49 The agglutinins did not appear until during or after defer-_ vescence. They remained demonstrable in the serum for vari- able periods of time. There seemed to be a tendency for them to persist longest in group I, next in group II and shortest in group IV. We were unable to find any evidence of the presence of specific agglutinins in the urine, even in those patients having strong agglutination reactions in the serum. In the precipitin tests with antipneumococcic serum and urine we believe that the specific “‘precipitin” is in the serum and the precipitable sub- stance, “‘precipitinogen,”’ in the urine. In patients with pneu- mococcic meningitis, the precipitable substance may be found both in the urine and cerebro-spinal fluid. CONCLUSIONS 1. Specific agglutinins usually appear in the serum of patients during or just after defervescence. 2. The agglutinins in the serum seem to persist longest in group I; somewhat less in group II and for the shortest time in group [V. . 3. The urine contains specifie precipitable substances but does not contain agglutinins or precipitins. 4. The spinal fluid in cases of pneumococcic meningitis also contains specific precipitable substances similar to those in the urine, but neither agglutinins nor precipitins. 5. The precipitin reaction may be applied to the spinal fluid in pheumococcus meningitis as a practical means of rapid classi- fication of the infection. REFERENCES (1) Docurz anp Avery: Journal American Medical Association, 1913, 61, 727. (2) CuickreRING: Journal Experimental Medicine, 1914, 20, 599. (3) Lister: South African Institute for Medical Research (Publications—De- cember 22, 1913, 1). (4) Buake: Reported at the meeting of the American Association of Immunolo- gists, April 6, 1917. Vi =) ae ‘| , : ; ne ; | F a co ’ ‘; tae 4 _ \ ‘ : ; ok ; bite ~ , =n yh, ui = \ ‘ ‘ . < 1 STUDIES ON THE ANTITRYPSIN OF SERUM B. FUJIMOTO From the Forensic-Medical Institute of the Imperial University at Tokyo Received for publication November 23, 1917 INTRODUCTION As to the nature of the antitrypsin of serum, a great deal has been written, and even at the present day various views are held in regard to it. In our laboratory several tests were performed in order to see how it can be inactivated and which part of the serum exerts the antitryptic action. PREPARATION AND METHOD OF OUR TESTS In our experiments von Bergmann’s casein method was modi- fied a little to make the reaction finer. Namely: 1. Solution of casein: 0.5 gram of casein was dissolved in 50 ec. of ; NaOH solution and the solution was warmed a while. Then the solution was neutralized with 4 muriatic acid and finally diluted to 500 ce. with normal salt solution. This solution of casein in our case is, therefore, less concentrated than that usually employed; i.e., it is of 0.1 per cent instead of 0.2 per cent. 2. Solution of trypsin: 0.05 gram of trypsin was dissolved in 5 ec. of physiological salt solution, to which 0.5 cc. of 7 NazCOs solution was added. To make it 100 cc. (0.05 per cent) nor- mal salt solution was then added to the mixture up to 100 cc. thus making a 0.05 per cent solution. This solution is also much more diluted than that ordinarily used (0.1 per cent). 3. Solution of acetic acid: Acetic acid, 5 ec.; alcohol, 45 cc.; aqua destillata, 50 ce. 4. The doses in each test tube were always: Solution of casein, 2 cc.; solution of trypsin in graded doses; blood serum (1 per cent), 1 cc. ol THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 52 B. FUJIMOTO The reason why all of these solutions were used in less con- centrated form than ordinarily, is that the reaction may be more easily and clearly distinguished. These solutions being mixed, and after being allowed to stand half an hour in the thermostat (37°), the series were examined with the solution of acetic acid, which was poured slowly on the surface of reagents in the test tube. When the casein is not yet digested completely and a trace of it exists still in the reagents, a beautiful circle of gray precipitate is easily to be seen at the plane of contact be- tween the acetic acid and the reagents. In our following tables antitrypsin is not expressed in units; but if it is desired to ex- press antitrypsin content in units, it may be calculated in the same manner as in the original method, because the relation be- tween the concentrations of casein and that of the trypsin solutions is not changed, though both solutions are less con- centrated. I. INACTIVATING TESTS OF THE ANTITRYPSIN OF BLOOD SERUM 1. Inactivation by heating Inactivating experiments for the antitrypsin of blood serum have been performed by several investigators. Vandevelde (1) saw that the antitrypsin may be destroyed in greater or less degree by heating it at 55°, 60° and 65°C.; Achalme (2) proved that they became completely inactive by heating at 65° to 70°C. Jochmann and Kantorowicz (3) found that the antitrypsin of human serum may be completely inactivated by heating at 66°C. for half an hour. Kurt Meyer (4) has shown that the antitrypsin of serum is destroyed in but a slight degree (only one-fifth of it) by heating at 56°C. for half an hour. Kam- merer (5) compared the resistance of the antitrypsin of human serum with that of ox serum, and he saw that the former is more resistant than the latter, and that the latter is inactivated almost completely by heating at 65°C. for half an hour, but that it can scarcely be inactivated by heating at 56°C. for half an hour. Doblin (6) did not confirm these results, but he has, on the con- trary, proved that the antitrypsin is thermostable and of a STUDIES ON THE ANTITRYPSIN OF SERUM 03 colloid nature. In short, the temperature that is necessary to inactivate the serum antitrypsin has been described differently, and there is one investigator that denies its thermolability. In order to see whether the antitrypsin of serum is thermo- stable or not, and if it be thermolabile, at what point of tem- perature it may be completely inactivated, we have begun with the following tests. Before proceeding to the main tests, we have had to determine at what point the sera coagulate. If they begin to coagulate by heating, the factors of the reaction are changed by it and it be- comes hard to judge the results correctly, so that the coagula- tion of the sera must be avoided as much as possible. The tests with regard to this, were performed as follows. First, the undiluted serum was heated at various points; secondly, it was diluted with distilled water or with normal salt solution and then heated at various points. The results are shown in the following table. TABLE 1 Coagulation tests of rabbit serum in its different concentrations at 72°C. and 80°C. av 72°C. at 80°C. CONCENTRATION OF SERUM Not diluted per cent 100 +r “te Diluted with Aqua destillata Salt solution Aqua destillata Salt solution 50 + = hsp oe 20 — — + _ 10 — — — Fs = el ns _ 1 as Bi es = + = Coagulation; — = no coagulation. Thus it was decided that the coagulation of the sera can be avoided by diluting it. In our tests we have, therefore, diluted all sera to 1 per cent, and then heated them at the high tem- perature that is necessary for their inactivation. 54 B. FUJIMOTO a. Serum of rabbit. Sera of rabbits were heated at various temperatures, i.e., at 65°, 70°, 72°, 75°, 80°C. for from one min- ute to three hours. The results are shown in the following tables. TABLE 2 Rabbit serum heated at 65°C. Serum heated SEMIN ULES se. ae eee HIM INUMGESs nek lee. Ree IOsmimutest-e. os. Lee — a se SO MMINUtES.. 62 ere a > -- -— It | | | | GOlmmutes:. 2... oe eee 120 minutes......... Nite AER USOeMINUbeS ee... eee | | | | | +++++++ +4 +t+++4+4++ 41 +t+tt+4++ 41 ++t++h++ +1 + = digestion incomplete; — = digestion complete. TABLE 3 Rabbit serum heated at 70°C. 0.4") (0:5° | 0-6 4] 0:7 || 0:8-) 0:9.) 1.1) 1) at 25) as INorserlmky.. Wiehe: see eae oh See |S eS SS S| S | = Serum nob heated... nwck eee oe +/+}+]}4+]/4+]/+}4+/4+/]41]- Serum heated SAMINULES ¥. nsihsriel aS 6 +it/+)4+]/+]}4+/—-/]-!/]-|]- SAMINNUES se .-42 ssa eer Se ee fe a fe a en ee fe LO MINUTES os 22s ew ns + fp] ef] ey | Hf SH SS 30 MINUTES: 3 soe oe oe yep) = bh | HP a pe ES SS Ol aT BUtEs Ese: aim ee +/4+/+)+=]—-;-]-/;]-]-]- AZO-MINUGES 0.- 165 eetesc4s Bes ees eo} ee ee a a TABLE 4 Rabbit serum heated at 72°C. 0.5 | 0.6 | 0.7 | 0.8] 0.9} 1:0) 2-1 | 1.2 | 1.3 | 1-4 INORSeTUNY S320 oe. nes t+ye]—f]/—f]—-}—f}]—-f—-]—-] = perum not heated: :...0.........8 +/4+/)4+)4+]/4+]+}]4+]}4+/)4+/]- Serum heated SMILES Sere ae 6 Hee eR eek +}/ty+y4ty)+]+=;)-}]-]-] - DMINUeSsey eee eee eee t+y+ti4+s}4)y4+]4+}]—-];-)]-]- LO PMIMUbeS eee ee a +/+} +y4+s)/+}]—-};-—-}]-]-]- ANMMINIbEs Lee ss le Be +ytey+y) =] —-]—-]-}—-}]-] - SU THMUGES: | lacey ae ee ee +)t}—}—y—]—)] —] —-] =] = GOamimUtests apes oes +yt})+=]—-/—-}]/-]-/]-]-]- 120 TITLES Aci tee | STUDIES ON THE ANTITRYPSIN OF SERUM 55 TABLE 5 Rabbit serum heated at 75°C. DOSES OF TRYPSIN INGTSCRUTMeermeerei eo... s be || a P| a Serum not heated........ Es te esa ital lectee sto 3 stan | Serum heated for 3 NWI Oooo ee eee eee nee ey ec | es | ee EMPIMITIAUIGES irae fale Gis eis ce oe ees So ole: lc |) => |) || RS a 10; DOM So eee ee eee SN eel the |e a 3{0) TTUTNOOSS Oe nee ee eee as ep a all em 1 a ee At 80°C. it took only seven and one-half minutes to inacti- vate the serum antitrypsin completely. It may be concluded from the above tables that the sera of rabbits can be inactivated completely by heating at 75°C. for ten minutes, or at 80°C. for seven and one-half minutes, and that at temperatures lower than 72°C. the serum of rabbit cannot be inactivated completely even if it be heated for two hours. b. Serum of horse. Several experiments similar to the foregoing were performed with sera of horses. The results are given in table 6. TABLE 6 Inactivation tests with the sera of horses DOSES OF TRYPSIN 0.7 | 0.8 019) |) 9-00) a-0 e (Pies |e INTO) GICTATIT SS 5 rence eee Ge SIERO on LAPS aoe ep ae = |e | | | Mend NOL NEAEMS. 2.05056. oe ss se ees eee: +}/4+})4+/4+/4+/4+/4+/]- Serum heated at 50°C. for OpMINULES awe tases et tpt] tl] ty ste] st le] = “er BOMMAINMPES 2:8 - owes cae ke Spee +i+it+i/+/4+/4+/.+]- 6 LOnmInuUtesss sn seens. eh eee Se ee ae a | fi aes FATHULESEL) salted 55 AE ee) se sees SE = ye ee z POMGINIGES, oiiellars he os ow ot See Ee fat | Fa ee lees a cern tap MUNUMbES ee SHE |p ee ee s LO SATINUIGES| et pees ct o. Loa = |) = SSeS | Sj i= aoa i HUUMUIGES 5-2-5) ayn cess +}/—-/;/-—-]/-—-/|}-j;-]|-|- 56 B. FUJIMOTO As we see in above table, the sera of horses can be completely inactivated by heating at a lower temperature than those of rabbits. c. Serum of sheep. Further tests were undertaken in the same manner with the serum of sheep and the following results were obtained (table 7). TABLE 7 Inactivation tests with the serum of sheep DOSES OF TRYPSIN 0.7 | 08] 09 | 10 {1.2 | 14|16 | 18 INIOTSCLUMI ofan: tothe toca eee ee ey Sf S|] = | = | Serum not heated... ....:......:eeeneease +)/4+/}/4/]/4+)/)4)4/)4+)/)- Serum heated for ten minutes at G5cO ee ee ee ee p24 sare AM Er arabs MA ie Cee TO°W) Gb vic as ee oe ee ee Pee ea eg Ree poten eer = UD Cp GS ahs. BSS a ee +|/—|/=- —|—-|-|- SOS Re ie anne Sets dae eee ae ee +)/—-/-—-/;/-|-|]-|]-|- Though Dédblin maintained that the antitrypsin of sera is thermostable, it is certain that it may be inactivated by heat- ing; i.e., it is thermolabile. And the temperatures which must be applied for the inactivation differ in each case. Namely: Rabbit serum at 75°C. for ten minutes Horse serum at 65°C. for ten minutes Sheep serum at 75°C. for ten minutes Human serum at 65° or 66°C. for one-half hour (Jochmann and Kammerer) — Ox serum, higher than human serum (Kimmerer) Conclusion. It may be concluded that the temperatures, which must be applied for the inactivation of antitrypsin of sera, differ not according to individuality but according to spe- cies and are between 65° and 75°C. As sera never coagulate at 75°C. in their lower concentration, it would be better at first to dilute them and then heat them at 75°C. for ten to thirty minutes for their complete inactivation. Incidentally, a brief experiment for the inactivation of the human urine was undertaken. The results obtained are shown in table 8. STUDIES ON THE ANTITRYPSIN OF SERUM A TABLE 8 Inactivation tests with human urine DOSES OF TRYPSIN OL6r jaOem INOS meO-O0) 2.0) || 1.4 | 1.9°) 1,3) 14 ING) Wile. 5 <6 5a02 CRs =f || eee ee (en |r pe | Urine not heated, 30 per cent, lce.....} + | +] +/+]/+/4+/+/+]- Urine heated, 30 per cent, 1 cc. at MethOUT. |... 202. on 24) ser) SEU ea ee bs se a ste UE mo ee +)+)+)+)/+/+/]+] +] - x PO OURY as ects ote SE Ngee if |) es, ee | ey = Ee ae tor i oaant, fcd3% Avcdeeserschaces Safe ey eS PS ee The antitryptic action of urine cannot be affected so easily as that of sera; urine must be heated at 100°C. for one-half hour for its complete inactivation. 2. Inactivation by shaking We can find no record of the inactivation of serum antitrypsin by shaking. It is well known that complement may be inacti- vated by shaking at 37°C. for one-half hour. If the antitrypsin of serum also could be inactivated by shaking as well as com- plement, many interesting experiments might be undertaken as to its nature. One cubic centimeter of serum was diluted to 10 cc. with normal salt solution and then shaken at various temperatures. TABLE 9 Inactivation tests by shaking (rabbit serum) DOSES OF TRYPSIN EOS Oe Nu OO. I Ree | ai || a ING@ EGIING 5 5e8 See eee eee o _ — - — — — — = Serum neither heated nor Sha kehiaerne iets oe toe ber. - — — - bh + + + + — Serum shaken for two to 3 hours: at Be AC ee Ae ae ee ap it SF ae soul Seill Se ad = — AGL CMe ete ees ar = Fill lie allo S| bale Se = = Controlserum (notshaken) at Sib rie Hn oe yen ok 2 + ae (arse se |) are ll sr = == AOE Creeps te marcas holcayees ei Sos + ar ie + ar |) ae i) ar = = 58 B. FUJIMOTO Horse serum was similarly treated, the results being the same. Conclusions. Antitrypsin of serum cannot be inactivated at all by shaking. II. EXPERIMENTS WITH SERUM GLOBULIN AND SERUM ALBUMIN In regard to this problem, there are several different views. Landsteiner (7), Miiller (8) and Opie and Parker (9) reported that antitrypsin is mostly fixed to the albumin fraction of serum. Glissner (10) stated that the antitrypsin of serum is all fixed to the euglobulin, not to albumin. But Dodblin (6) and Kimmerer (5) maintained that both the globulin and the albumin fractions exert an antitryptic action, but the former less than the latter. Beside this, Kiimmerer (5’) has recently made interesting ex- periments, in which he has proved that the globulin fraction in- creases after heating serum at 56°C. for one-half hour, while the albumin fraction decreases, and that the antitryptic action of the albumin fraction may be more easily inactivated than that of the globulin fraction. In our laboratory different tests were performed bearing upon this question. To fractionate serum in globulin and albumin, we have employed several methods. 1. Serum treated with ammonium sulphate For the fractionation of 50 ce. of horse serum, there was added an equal volume of saturated solution of ammonium sulphate and the mixture was filtered. The precipitate was washed five times with a half saturated solution of the same salt. As ammonium sulphate precipitates casein, both fractions must be dialyzed for its elimination. However, as it takes a week or more to make them completely free from the salt by dialysis, their antitryptic action might, also, be weakened by the treatment. For this reason we were satisfied with a day’s dialysis, after which time the concentration of the ammonium — sulphate in the globulin and albumin fractions was so little that the casein was not precipitated by it. Fortunately it was al- ready proved by Kimmerer that ammonium sulphate does not STUDIES ON THE ANTITRYPSIN OF SERUM 59 prevent the action of trypsin in any way; and Doblin saw that the antitrypsin was not lost at all in the first week of dialysis. Afterward the same result as Déblin’s was obtained by us as to this property of trypsin (see the following chapter). It is, there- fore, clear that only a day’s dialysis exerts no influence upon the antitryptic action of serum. If the doses of serum globulin (100 per cent) in the test tubes are more than 0.2 ec., the globulin may be precipitated by TABLE 10 Tests with serum globulin and albumin DOSES OF TRYPSIN No serum.. i MGetsterseetee | P= —shSo Sea =e Peer cen cont Mice. veeeeeeees/ Le] eEE] +H] +] +l stir Serum globulin* ae S| a ee a Not eae ce per cent, I cc..... to} te} py] + (eaeper cent, 1 cc:. +y—]—f—F}—f—)] —]—-] —-— Heated (75°, one-half hour) 20° per cent, 1 cc.. ee es ae Serum albumin Not heated, 2 per cent, 0.5cc....| +] +] +/s=]|—/]—|—-|-|/- Heated (75°, one-half hour) 2 per cent; 0:5 COs jae. s6 see soe = +/—}—}/—/]—-/]—-]—-]|]-]- * The serum globulin was dissolved in normal salt solution in a volume equal to that of the original serum. acetic acid. Therefore, we must be careful in our experiments not to misjudge the reaction that is caused by the precipitation of serum globulin and not of casein. The precipitation of globulin may be easily distinguished by its coloring. As the antitryptic action of serum is not weakened by dialysis at all, we can say that the antitryptic action of serum is consid- erably weakened by precipitating with ammonium sulphate, and that the antitrypsin is to be found not only in the albumin frac- tion but also, in a small degree, in the globulin fraction. 60 ; B. FUJIMOTO 2. Serum treaied with carbonic acid By the preceding experiment it was shown that the globulin fraction of serum also shows antitryptic action. To confirm this fact through some other experiments, we have treated the serum with carbonic acid and 5; muriatic acid. To 5 ce. of horse serum were added 95 cc. of aqua destillata, which was cooled in ice, and then carbonic acid was led from Kipp’s apparatus into it until the precipitate no longer in- creased (about one-half hour). The precipitate consists solely of globulin. It was separated from the albumin portion by TABLE 11 Tests with serum globulin and albumin obtained by means of COz DOSES OF TRYPSIN 0.7 | 0.8 | 0.9 | 1.0 | 1.2 | 14 | 16 | 2.6 | 2.8 ING SCRUM 4206 fice coos se ee teen aa ie —|—-—-|/—-|;-|- Serum!) per cent, 1. Ce ..25 eee eee +/+)+/+}]+/4+)/- Serum treated with CO: Albumin part, 1 per cent, J ce:.:-) | ae) eo | aie 1.0 per cent...... SS sel SS eS 220) DED CEDbae-i1F: SP | == a el Globulinpart, } 5.0 per cent...... Se Wa eee 1 ce. } 10.0 per cent...... +/+/+/+=]-|]-|- Ex per cent...... +i/+/+{/+/+]/4+|{- : 50.0 per cent...... Se ee oa ee feo ol listen a alk means of a centrifuge, and it was washed more than five times with distilled water to render it free from albumin. Both the globulin and the albumin thus obtained were exam- ined as usual, the results being shown in table 11. Such an experiment was repeated with the serum of the horse and rabbit, and it was always followed by the same results. Of course, both the globulin and albumin fractions could be inacti- vated by heating. Thus it has been demonstrated that the antitryptic action of serum may be weakened only a little by precipitation with car- bonic acid, and that the globulin fraction undoubtedly has antitryptic action. pa OR Se ~ oy ae Ag Re ye -. s ice wee =. STUDIES ON THE ANTITRYPSIN OF SERUM 61 3. Serum treated with 35) HCl. Sachs employed this method to separate the complement of guinea-pig serum into two components. To 0.5 cc. of horse serum, there was added 4.1 ce. of 55) muriatic acid; after stand- ing an hour at room temperature it was separated by means of the centrifuge into two portions. The globulin fraction (..e., the precipitate) was washed with distilled water several times carefully. The results of the tests were mostly the same as we have seen above (table 12). TABLE 12 Tests with serum globulin and albumin obtained by Sach’s method DOSES OF TRYPSIN 0:5 | 0.6) | O67. O-85 (10h feuko! | stst| sina e7 PIGESE GUNN te yet. cick De « Wvintoie ere doce we a es ee (ee Sexumplapermcent. 1 Ck. so:.- 2-sece 0 a ee ee ee ee Serum treated with HCl Albumin) partel per cent, 1 -ces.4.2\osh stee | Mates ste | ete ete ( 5 per cent.., +] =/—]|—|]/—/]—-—|]- : 10 per cent..} + | +/+] =/—-|]-|- Globulin part, 1 ce. Solpen cont.) bo| erie Mee eee ea me (50 per cent..); + | + | +]+]+ pe eee Globulin part obtained according to Sach’s method shows an antitryptic action, but in less degree as in the case before. Conclusion. Both globulin and albumin fractions have anti- tryptic action, but the latter in less degree than the former. III. THE NATURE OF SERUM ANTITRYPSIN It has been shown by the above experiments that the anti- trypsin of serum is to be found in both the globulin fraction and the albumin fraction. To determine whether these proteins themselves have such antitryptic action or whether they are accompanied by some specific substances which exert an anti- tryptic influence, further studies were undertaken. 62 B. FUJIMOTO 1. Is the antitrypsin dialysable? In the first place, the globulin and albumin fractions of serum were dialyzed. As we know that globulin and albumin are not dialysable, if the antitryptic action of sera be lost by dialysis, there must be some specific acting substances; if it be not lost, antitryptic acting substances in sera must be undialysable and it TABLE 13 Dialyzation tests DOSES OF TRYPSIN INO SELUM 2k folasc- occ eee eas eee eer eeeROES +)}—/]—]—-/]-—-];-|]- Serum not dialyzed 1 per cent, 1 cc............ Jt}+)4+)]+},4+]4+)]- Serum (1 per cent) dialyzed for didays ll .€C2.\:..05 sk teeeiee eee eee +)/+y+ty ti +] =i] -—- PORWR; il CC's. 2. /cac: ee eee ere ee +i4+)/+)4+/4+)={/- | goidays, 4 ce... 2522-52 -e te see: eee eee +/+) +]/4+/4+)+=—- ge dideys, 1166 -<.302 Sk ce ere oe ee ee eee +)/+]+]}]+)})+)+{- Serum (1 per cent) not dialyzed but otherwise under the same conditions for tday:. 1 ces, 24at- Pelee ete Pee SS eamlae i) == |) = QiGays, ErGCs 202 ran oS)

. choke te eae Cee Se Se ea ea Albumin (1 per cent) dialyzed for DOAYS* 1sCOee sgancie aceon eee Se eel! sel he 4 days, 1 eels 4-2 Pee. ee eee +it+y+]/+]/4+)]+4+/- Albumin (1 per cent) not dialyzed, 1 cc. for 2:£0 A Gays. ot he oie ss speech eee ee: “eal eal ea S| ee might be said that antitryptic acting substance is nothing but protein itself. The following tests were always carefully con- trolled by undialyzed samples which were exposed otherwise to the same conditions. It is seen that the antitryptic action of serum as well as that of globulin or albumin is not weakened at all after one to four STUDIES ON THE ANTITRYPSIN OF SERUM 63 days’ dialysis. We can say, therefore, that the antitryptic sub- stance of serum at least is not dialysable. Human urine was similarly tested but in this case the antitryptic property disappeared completely after two days’ dialysis. ; TABLE 14 Dialyzation test with human urine DOSES OF TRYPSIN 0.3 | 0.4 | 0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 1.0 | 1:2 MAMVMTITEITN GC Pee PENIS clei) See yavors avacdeee Bes ae +/+/—/—;-—-/—-];-]/-—-]— Wrme 50) per cent, 0:6 ce.............. +i i+ty4+y)+)}4+]/4+)4+);)4+)-—- Urine 50 per cent, 0.6 cc. dialyzed for 2) CES HEbs Gace oe ee nena a eee +/+ ;—-}]/—/]—]—-]—-]-] =- Urine 50 per cent, 0.6 cc. (control) not dialyzed (only standing for PEGA S acess oats, sslveenes s SEaROR Yh eee Se se dl ae ar eae | sr ae) aS 2. Experiment with pepton A brief experiment was carried out with pepton, which is dialysable. But pepton itself was found to have no antitryptic action, and it neither prevented nor accelerated that action. 3. Ether extract of serum Serum was extracted with ether to see whether the substances that are soluble in ether, have an antitryptic action or not. O. Schwarz (11) stated that serum became inactive after the lipoids were extracted from it and that a mixture of the lipoids with protein exerts an antitryptic action. Kammerer (5’) refuted Schwarz’s theory by his experiment, in which he saw only a slight decrease of antitryptic action (only one-third) after ex- tracting the serum five to six times with ether. Five cubic centimeters of horse serum were treated with 50 ec. of ether in our experiment. The extract was redissolved in ether to make it pure and was made into emulsion with physio- logical salt solution. The results were as follows (table 15): 64 B. FUJIMOTO TABLE 15* Tests with ether extract of serum DOSES OF TRYPSIN 0:5 | 0.6 | 0:7 | 0:8 | 09° |) 1.0) 2 INOW SEGUINY 29S = Shi 2 2 are se rele Soe eee er eae a ee | = |) == Serum 1. per cent, 1 ce.......5+...2:9sgeeeee-eeal ae | 1 SE} Sete eee Ether extract 20 per cent? 1 \CCs..5. : a... ee Ee eer ee =i aS | Ht = || = HOspericent, 1 6C 122s 04.02 oe eee +) =};/—-—-}—-]—-}y-]- * See the note to table 10. It was established through the above experiments that the antitryptic substance is not dialysable and that the ether extract -of serum does not contain such a substance. We have, therefore, proceeded to the following tests. 4. Crystalline albumin It seems now probable that the antitryptic substance of serum may be nothing but the protein of the serum itself. This ques- tion is not yet definitely decided, and we can not find a record of any previous experiment which was undertaken to settle it. Taking advantage of the property of serum albumin to crystal- lize, we have undertaken an experiment with crystalline serum albumin and, also, with crystalline ovoalbumin. As albumin that is not denatured and that is free from any admixture can be obtained by crystallization, it seemed that an experiment with crystalline albumin might bring a decision with regard to this problem. To obtain the crystalline albumin, I have em- ployed the following method. Fifty cubic centimeters of horse serum and the same quantity of saturated ammonium sulphate solution were mixed and then filtered. To 50 cc. of the filtrate was added more saturated ammonium sulphate solution until further precipitation began, and then 1 to 1.25 ec. of saturated ammonium sulphate solution, to which acetic acid to 10 per cent had been added. After twenty-four or forty-eight hours (in the ice box), we obtained typical crystals of serum albumin; the crystals were separated STUDIES ON THE ANTITRYPSIN OF SERUM 65 from the fluid by centrifugation and washed five or more times with a half saturated solution of ammonium sulphate, which contains the salt and acetic acid in the same relation as above. The crystals were finally dissolved in distilled water or physio- logical salt solution. TABLE 16* Test with solution of albumin crystal DOSES OF TRYPSIN 0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 1.0 | 1.2 | 1.4] 1.6 ING SERUM eee ae be See On Oe ee +/—}—}]—};—}]—-);—-]-—- Senumeolepenicent, lice: «4. ....%...... +) ] +] eye ey ae ey K— Crystalline serum albumin, 1 cc. [OSE CAN se.5:6 cole hse eee +/+;)+/) +=); —-/—-/;—)];-]-— iQ), JOGTEGEI tess on 0 6.0 SDI eee +/)+/)+)/)+;+/4+)/)-/]-/- Crystalline ovoalbumin, 1 ce. 5) jOXeNP OSTA a Sbrotoit cs IOR Ee ae en Ree +)4+})+)]/+=;);—-—-]-—-]-]-]- NORMGTRCENGer en ete ee oa cae se +)4+)4+)+);)4+/+/-/]-)|]- * See the note to table 10. As the crystalline serum albumin is pure and free from any admixture, it must be accepted that the serum albumin itself has — an antitryptic action. Conclusion. There may be many kinds of substances in serum which exert an antitryptic influence, as Oppenheimer (12) has said; but there is no doubt that albumin itself exerts such action. It is still open to question, whether it is the serum protein alone that is the antitryptic agent or whether some other substances beside the protein are also antitryptic. Fur- ther quantitative experiments are necessary to solve this question. RESULTS 1. The temperatures that are necessary to inactivate sera, are different according to species, and are between 65° and 75°C. 2. The antitrypsin of serum can not be inactivated by shaking. 3. Both the globulin and the albumin fractions are antitryptic, but the former is less so than the latter, as Déblin and Kim- merer stated. 66 B. FUJIMOTO 4, The antitrypsin of serum is not dialysable. 5. Ether extract of serum has no antitryptic action. 6. Crystalline serum albumin is antitryptic. 7. It is not yet decided whether serum contains antitryptic substances beside the serum protein or not. It is a great pleasure to me to thank Prof. Dr. K. Katayama and Prof. Dr. 8. Mita for suggesting this problem and for their kind advice. REFERENCES (1) Vanpevetpe, A. J. J.: Uber die Wirkung der Erwirmung auf Protease. Biochem. Zeitschr., 1909, 18, 142. (2) AcHALME: Propr. pathol. de la trypsine. Annale de 1’Institute Pasteur, 1901, 15. (3) JocHMANN AND KantTorowicz: Antitrypsine und Antipepsine im mensch- lichen Blutserum. Zeitschr. f. klin. Med., 1908, 66. (4) Kurt Meyer: Uber Trypsin und Antitrypsin. Biochem. Zeitschr., 1909, Nr. 23. (4’) Kurr Meyer: Uber die Natur. d. Serumantitrypsins. Berl. kl. Woch- enschr., 1909, Nr. 42. (5) KAmMeprerR, H.: Studien iiber die Antitrypsine des Serums. Deutsch. Arch. f. kl. Med., 1911, 103. (5’) KAmMmerRER, H., anp Lupwic ANBRy: Untersuchungen iiber die Bezieh- ungen der Serumeiweisskérper zur Antitrypsinwirkung. Biochem. Ztschr., 1913, 48, 247. (6) Désiin: Untersuchungen iiber die Natur des Antitrypsins. Ztschr. f Immunitiatsforsch., 1909, 4, 229. (7) LANDSTEINER: Zur Kenntniss der antiferment. lytischen und agglutinier. Wirkung des Blutserums und der Lymphe. Zentralblat f. Bakt., 1900, 27. (8) Mituer: Uber das Verhalten des proteolyt. . Leukozyten-fermentes. Deutsch. Arch. f. kl. Med., 1908, 91. (9) Opin AND Parker: Leucoprotease and antileucoprotease of mammals and of birds. Journ. of Exp. Med., 1907. (10) GuLAssNzER: Antitrypt. Wirkung des Blutes. Hofmeister’s Zeitschr., 1904. (11) O. Schwarz: Die Natur des Antitrypsins im Serum und der Mechanismus seiner Wirkung. Wiener kl. Wochenschr., 1909, Nr. 33. (12) OppENHEIMER. C.: Die Fermente und ihre Wirkungen. 4. Aufi., 1913, I, 478. THE CONSTANCY OF THE PROTEIN QUOTIENT DURING INTENSIVE DIGESTION AND PROLONGED STARVATION SAMUEL HANSON From the Department of Biochemistry and Pharmacology, Rudolph Spreckels Physiological Laboratory, University of California Received for publication, January 8, 1918 It is usually assumed that the fluctuations of the protein quotient under normal and pathological conditions are due directly or indirectly to variations in the general nature and especially in the rapidity of the metabolic processes. Thus, Cervello (1) ascribes to a retardation of metabolism the rise in globulins he obtained under the administration of antipyrin. Hurwitz and Meyer (2) suppose that the increase in the serum globulins in certain actue infections is due to meta- bolic disturbances. Schmidt and Schmidt (8) are inclined to believe that the serum proteins have no relation to immunity and would rather attribute the increase in globulins to changes in the metabolism resulting from the toxemia and tissue waste of acute infections. | In view of the fact that alterations in the protein quotient have been so frequently attributed by various authors to dis- turbances in the metabolism, it was considered necessary to investigate the more precise relations of the globulin albumin ratio to difference in the rapidity of metabolism. From this standpoint it was shown in a previous communication (4) that the protein quotient of blood serum is not changed by the admin- istration of certain drugs which are known to retard or accelerate the metabolic processes. The present work is a continuation of the former experiments, and its object is to determine the influ- ence on the protein quotient of a disturbance of metabolism when such a disturbance is produced by periods of digestion alternating with prolonged periods of starvation. 67 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 68 SAMUEL HANSON Although many determinations of the serum proteins during starvation and digestion have been reported by various inves- tigators; yet, the inadequate methods for estimating the serum proteins, which were then available, have so precluded accurate and thorough work, that a wide disparity in the results obtained by different workers is rather the rule than the exception. The chief disadvantage in the older methods is that large quantities of blood are required for a determination. ‘This made frequent estimations in the smaller animals impossible, and as a result the normal variation was not known, and the intermediate effects remained undertermined. To illustrate the marked lack of uniformity in the findings of different investigators, the following reports have been taken at random from the literature. Salvioli (5) found that the quantity of globulins in the blood serum of two dogs starved for twenty to twenty-four hours is less than in two other dogs during a period of digestion. This change he attributed to individual variations, since he found no difference in the globulins of the same dog during fasting and digestion. Burckhard (6), however, obtained in the serum of fasting dogs a marked increase in the globulins and a considerable decrease ‘in the albumins: He explained these findings on the hypothesis that during starvation the globulins pass from the organs into the blood in order to cover the deficiency in the albumins. Lewinski (7) reported only a slight increase in the serum globulins of dogs after prolonged periods of starvation. Briggs (8) obtained a considerable decrease in the globulins of certain species of birds starved for a period of at least twenty- four hours. This work faces the criticism, however, in that the normal values were ascertained for one group of animals and the values in starvation for another; hence the comparison of the latter data with the former is not fully justified. METHOD AND MATERIALS The experimental conditions were very much the same as those of the previous work (4). Robertson’s micro-refracto- ee THE CONSTANCY OF THE PROTEIN QUOTIENT TABLE 1 Rabbit 1. Weight: June 8, 3332 grams; June 21, 2023 grams 69 k GLOBU-| 320- DATE TREATMENT aes sae rea aay aonar: seed : TEIN TEIN ae TIENT per cent|per cent|per cent|per cent\per cent LUTE (Cae ere ae Normal LEG een ete nos510 2080) 0625 DUES a site oss 205s, » Normal LS) Peon oe ee OMnle4.8 |) 210) 20L26 WVisivaelOQW is oe sass sss Normal ISOS PSe Shellie eee OF F230 R29 1G A Normal 163) 5 4735 ete 4-4) 2304) 0382 LWiGl7 12 a. ae Normal 1.55) SS Shiba sO). | 24.0) 0.32 IVER Semis. 2elsi> «ss « Normal 155 | eon ellen one | 2O0LOn | On26 Vietnygail Aaa ie eiete sec) Normal ay | 22550). |) Uoth || eal WP PaO) O77 ECG Normal LON 4 oe ele Omn Ronen | OeOn FOn24 June 8 Normal 15505220 Ossie eGe Orel 13.0) 0216 Jean 1) Sea ear Starved, 1 day 1.4 oma OR Gate 11620) 0220 | 0 Starved, 2days | 1.4| 6.0/0.7 | 6.7 | 10.5 | 0.10 IME AO shee ee oS. Fed, 2 hours 1:6) 4:35) 1A bee | 24:0) 0:33 CITY ee ee Starved, 1 day 1 On| oe OR lez Geen 1920) O24 ume wl Ane ys eyes. Starved, didays | 1.4 | 4.97/35 la652) . || 21:0) 10.27 JUL oe Ce Fed, 43 hours 1.4) 4:9) Tet 76.0) || 18.0) | 0:22 4G (Or ee Starved, 2days | 1725) Soi MAO 625 slel7 On Ons MERCY OO se tao S oot Starved, 5 days |> 1:4 | S25yiL2o6.6 17.0) 0820 ume e2 0 sey Pe yarns ae Fed, 33 hours 12251, 15)5OR | Te hOe teen SO) One AEE B AT ee a tat Normal 140) 4s 5s O85 5.2504 On Ona LUG 2-355 Ab enone bore Normal IAP “4 eSa Olden |S). One el 42 ON ORIG TABLE 2 Rabbit 2. Weight: June 8, 2261 grams; June 21, 1666 grams ate GLOBU-| p20- DATE TREATMENT ee gael me eas Sante penal TEIN TEIN ae TIENT per cent|per cent|per cent|per cent|per cent REC RY, 8 eB Icis.o2. 516 b> Normal 13s ih 2365 | SIRO 6: 20 e262 OulOeS> dunes Ue Ae ae eee Starved, 1 day 1 Aa pol a5 9 (6.65 |) 23200 |t0e29 JUTE Ce eae aa a Starved, 2days | 1:33) 520) })230) | 7.0) || 285010840 Meter OMe. ota... fesse Fed, 2 hours 4) |e el 2EOR | eSe On | son On mOnar dors) 11275 ee ao ee Starved, 1 day 1.4) )) soe Te S1e7 |) 68) 2520") 10233 PUNE EASE 5. Ba nis Starved, 3days | 1.2] 5.6|1.8 | 7.4 | 24.0] 0.32 itive 7: i Fed, 4 hours 1.3) S245 154 7628 | -20755 0526 (CT UP ee Starved, 2:days | We25) soe25 (eT) 6.9) 25-05" 0633 “JT Xs) 24 Stamved.,o days | 158 |) 4x7 eran | 6. 192370 W0r380 AUTTS VDE Ie aes ee Fed, 3 hours 1.3 MOI) IEA 6.1 23.0 | 0.22 AGS 21 Ree Normal 10s 42 4a RA 58.0) 24-0. 0832 divi DE} eRe ee Normal heal Aioed |), I Ne ecstsy || eal ay || (027/ 70 SAMUEL HANSON “metric method (9) was employed for the estimation of the serum proteins. Determinations were made in all cases through a fore-period of several days, to serve as a safe standard for comparison with the values to be obtained during the subse- quent period of starvation and digestion. For the analysis four cubic centimeters of fresh blood drawn from a manera ear vein was employed. _Rabbits were selected as the experimental animals. © During the fore-period the rabbits were fed once daily with a moderate quantity of alfalfa hay and crushed barley. Suffi- TABLE 3 BE! a NON TOTAL Sane wee : DATE TREATMENT ices coe ean ae peice avo. is . TEIN NU Ve..( per cent|per cent|per cent|per cent\per cent May O 7.0.24 6..000.4 Normal TRA 42 ee Oe 2) o) LOROMAOR24 May O60 840 200: Normal 12" |. 4.2 100.0 15.2. ||. 19502 Oaes May A0i82 94. ist Normal 1.2))) 4.4 | 1.0.) 5.4--| 18:5-)0%23 May 2 5...55204 Normal 152!) 74.4.)'91.1_)-5.5--| 1925 Oat Moly 42 Seis. ses. Normal 1.5 |, 4.1} °1.0|-5.1- | 19.5 |. 0024 Mays 2285 3... Urt.. Normal 1.3.) 45:10 1.11) .5.6- 4) 20:0) 0e4 May-l4- eo ees: Normal 1.24)—4.2 | -1.051-5-2—|-19-0-| 0824 May 1G tec. aad - Normal 13.) 4.6 |) 0.9. |: 5x55") 62010520 IMs MES eared aceon Normal 1:4.) 3.9) 1.4.) 5.34) 26.0 |)0235 1A) I i aL RSet cae Normal 13 | 44 Oost eal SOR oet VMsiy DUM RO eaten 52 Normal 1.3.| 4:4) 1:5 | 5.95 | 25.0 | 0.34 May 24 x: sinasers, a cetwis Normal 151A eb Slee 2a onder ele ON ROE Mia t29)3 Scns shins oe iie Normal 1.3:| 4:55) 0 Ls2elcoads 220-0 0s 26m June Ssa.c8. cade: Normal 123.|- 4.6.) 1.0, |-5s7—}-20-0-|50224 AWS aah SaaS or Starved, 1 day 1227 VASA Ta oe8 | ZOO ROE23 JumewlO ste... cee Starved, 2days | 1.2] 5.0} 1.0] 6.0 | 17.0} 0.20 JUBCUONM Feiss Fed, 2 hours 08'| 40° |e 1) 5-1 - |. 21.0. 0227 Juma Geek. sess Starved, 1 day Ma Ale e122 1} O29 - +) 20. 020326 Jide ML Mays ste: Starved, 3idays |) 1.3 * 5.2 91-11) 6.3. | £725-)/0228 JuMeULa Me... See: Fed, 33 hours WAN ALTO | 528 > 2020) 0223 BS Co ly GA ee Starved, 2days| 1.2] 5.0| 1.4] 6.4 | 22.0 | 0.28 Jthe 20...05.........| Starved, 5days |) 1.37/50 |i 1596-1 | 18:0) 0x22 Jee v20.4 sCo. coe sBat Fed, 2} hours 163} |~ 4564/5108) -5 26. | 1820-) O222 JUMeyAT esse geet ote Normal 1:2') 420 17039 \-5-0--| 18:0 | 0822 JURE 23 .\...060.2520..| ‘Normal 125 4 ae ee Pe 9522 |) 210M ORT Rabbit 3. Weight: June 8, 3332 grams; June 21, 2380 grams THE CONSTANCY OF THE PROTEIN QUOTIENT TABLE 4 Rabbit 4. Weight: June 8, 4165 grams; June 21, 3570 grams 71 GLOBU-| on | PRO- DATE TREATMENT aoe ene Bea | Beet we per TEIN a TEIN eae: ecbars per cent|per cent)per cent|per cent|\per cent Aprlldee sss. 2s. Normal Ld) eae 2; |.620.), 20.2) Or25 Apron 6.0.6... Normal 14a) oral 20. 166,)||. 18.9! (OF22 AprtGreke |. .22.). 5. Normal (EAD OP Zaeel eon iG | «22a OFZ Li he ee Normal Lebnioxaa at 2.1.66.) 18i9h 0522 BTN QOS 6c. sss’. s. Normal Wd Roe Ped 24 41!65.5.4),..,.22-S0 OL2e aune 1Ss-.2)......-+...) Normal LSE POBZa esl aluGh sil. 22h OLZ9 PMFC TT On tentet 3. 5b 5 ):. Starved, 1 day IEPA 25541 (POR ROLOW|aezone hOLoO me NO ease. e's 3. 2: Starved, 2-days| 1.2] 5.5| 1.4] 6.9.) 20 | 0.25 2 OS Oe _ Fed, 2 hours LAC). COMO Ne iGe 1 +! < L6-O Or20 Uwe ee ane ee ee Starved, 1 day 1eSHR4E Sale Gr enO24, || 25,0. O38 RUIN WA ras bev eae. starved, 3:days'| 1:1 | ‘4:80 1.57]. .6.3.|. 25-1)! 0:32 ume, AP ee Fed, 3 hours 123°) T4530 le GaeGe4 6.25.0 Oxss JUG I Ses as eaeCae Starved, 2days'| 1.1) 5.0} 1.5 -|- 6-5 | 23-0.30 JUINS MAAR Reo ce ee Starved, 6 days LO) ASS ee ON Gaal Zen Ok40 UIE BZ: arama ee Fed, 4 days IL PAI Gye Sal G6. 9alaeeo | 0235 JUIN, 7B sememcana eae Normal 12 | 42 eee oe Onl 2940840 ; TABLE 5 Rabbit 5. Weight: June 8, 3094 grams; June 21, not weighed agg ERO PRO- DATE TREATMENT ane er bes eae DOL ae! TEIN TEIN ee TIENT per cent\per cent|per cent|per cent|per cent gegotal U2 oe t.. zee Normal V4 eo O eleSe ol) ond 33 | 0.50 Pept eAe ies os. . Normal EY al ess ed er? eT | 36 | 0.56 POT Ufc jaan. ss Normal 1.4 | 4.3 2) bk 9} |. 6:2 31 | 0.44 /Nr over oe See Normal 1.4 | 42 eS) 6.0 30 | 0.42 AJ TiySs (Se ee Normal Dan leeean ORO 32 3) O47 ‘huis eo he Starved, 1 day 1 aFOSSue eer |) (6 30 | 0.43 dane 10. ce. 2. 4:545..|: Starved, 2.days'|. 1:2) 4.8) |.2-2) | 7:0 31 | 0.46 NS Os eek «agra. - Fed, 2 hours 19) AS Sbelo2e28 wile eOdale Sleny OL46 ANT e De en ee Starved, 1 day Ls2\ 4.9 22k | val 35 | 0.47 Jno ee ae ee ae Starved, 3 days Sample lost AJiiinven 11 Ee ee ee Fed, 22 hours 1342 Gono) i fol 385 | 0.54 NO fh ores As Sarena « Starved, 2idays |' 1:24 4.7..|.2.55 | 7.25 | 37 -| 0.54 June dA9. ee.) .c5.5. 6% Starved, 4 days Died fe SAMUEL HANSON TABLE 6 Rabbit 6. Weight: June 8, 2618 grams; June 21, 2023 grams daa RO- DATE TREATMENT aes eee onal SOEOr, one ea TEIN TEIN eae TIENT per cent|per cent|per cent\per cent|per cent JUTE! ISURAe seth ss : Normal 12 eae oboe 9 > 1) 7-4 Sie OnOsos JUN eee Nek. . 3. Starved, 1 day ese oe2enl eel 7.3 | -29s0nnOEeG ue wl OMRRe saan oe. Starved, 2 days | 1.2| 4.9 | 2:2 | 7:1 | 31.0-) 045 mel ON ets exc: Fed, 2 hours 159) 421 42223. 19624. “| 36408 BORDG Vibna yes Ae Pees Se ee Starved, 1 day 1.3) | 425. 10224 1.6.9 | 3520870253 ol jieun sl 2: Uy ae oa Starved, 3days | 1.3) 4.8 | 2.55 | 7.35 | 35.0 | 0.538 merase eee oy. ©... Fed, 23 hours ds osOm 222. | 7.2 | 30.59) One JUVE ets elackes <0: Starved, 2days | 1.2 | 5.0 | 2.05 | 7.05 | 29.0 | 0.41 Tue: ZOE es noes a Starved, 5days| 1.2) 5.0''| 2:2 | 7.2 | 30.5) 0:44 Jumen20 See ane ie ..3- Fed, 2 hours 122A SEO 222) aie? N30. bn | ORs! ditine Wle. aesnnon ane Normal 15529) Axoonle2e2e | O-oDNlooeom Oso UM CR23 eee thsstic Normal 1.3.) 4.2 19) WOR {| 3IR OF ORAS TABLE 7 Rabbit 7. Weight: June 8, 3094 grams; June 21, 2618 grams ; GLOBU-| p25. DATE TREATMENT aoe fered eee rae Toni, per TEIN TEIN cent TIENT per cent per cent|\per cent|per cent\per cent Mia eile cis tare pice Normal 123.) 4.284) ©29% || Gate |) 2830580839 Misty: 134.8: bck ee ss ‘Normal Be AZo 20) Gat {eZee ON eae Mutya Occ pikecprrcac Normal 1.4) 4.8} 2.0 | 6.8 | 29.0) 0.41 Ming HONS bes acct. Normal 15) “26a Toe eo ie 29R Om sOks 11 in ga a ar a Normal 15 | 4:9) Love e526) 1 S0k0sOeS4: IMlaiy LD: mehr tes ote ae Normal 125). 4269) 5 56st) 29255 t0rSs 11) Eid ee rege Normal 1-4. |) 4207) 208915680" 1330980250 Maye la mene Normal 14°) P4577) WGI) 2523" 1) S008 80243 IMiavalG see cee Normal VA | Ag8e) Wea G6 25" 4 2705 tO eso UN Guerre ben. cere Normal 132!) *5O 54 Dede IA ebe 42820 sORs ne Oe Ss es ees Starved, 1 day M3 Osae2. LZ 25) | 280 Oess “TT a ee ei a Starved, 2days| 1.3] 5.2|2.4 | 7.6 | 32.0 | 0.46 PUNE MOL es eect Fed, 2 hours 14] 455 122741629) | (3520) (0835 jini PAP ae se Seneione Starved, 1 day 1.3] 5.6] 2.75 | 8.35 | 33.5 | 0.49 dl UboVeh 3 BAe noeee ears Starved, 3 days | 1.1] 5.2) 2:5-1 7:7 | 32.0 | 0-48 jumevla ee cance t eer Fed, 13 hours QM NOeSe| oO! Weve: 932 20m| Oo! AJ Shee Geers ae eee Starved, 2 days| 1.0| 5.0] 2.5 | 7.5 | 33.0] 0.50 Jume 2h. o.20.5. Starved,6days| 1.0] 4.8/2.8 | 7.6 | 37.0 | 0.51 MAMIE 20 Pecos ears otoes: Fed, ? hour 120) || 48 °8288 187.6" *| 37-0 |) Oza THE CONSTANCY OF THE PROTEIN QUOTIENT 73 cient fresh water was always in the cages during the fore-period and during the period. At the beginning of the period, all the straw bedding and other possible food was removed from the cages. Where the interval of starvation exceeded two days, determinations of the serum proteins were made to show the intermediate effect of the fasting. At the termination of the starvation interval, blood was drawn for determinations. The animals were then given an excess of alfalfa hay, crushed barley, cabbage, and carrots. This. interval of feeding was allowed to vary from three-quarters of an hour to four hours and a half, in order that several gradations of the intensity of digestion may be included. All the food was removed again, at the end of this feeding interval, and blood was immediately once more drawn for determinations. A longer interval of fasting then followed, and so on. DISCUSSION From a perusal of the tables presented, it will be seen that the protein quotient remains normal during digestion periods alternating with prolonged starvation periods. The negative data for digestion are fully in accord with the findings of Rowe (10) who showed that high protein diets have no immediate effect on the globulin-albumin ratio. It will be recalled that the retardation of the nitrogenous metabolism as shown by the diminution in the elimination of nitrogen during starvation (11), is due to the intrinsic change, which consists of a gradual decrease in the exogenous metab- olism, accompanied but not equalled by an acceleration in the endogenous metabolism. Viewed in the light of these profound metabolic disorders, the constancy of the protein quotient during prolonged starvation is not an easy matter to explain. An analogous phenomenon is the constant and normal per- centage of glucose in the blood even during an extended period of fasting (12). It is assumed that this constancy of composition is maintained chiefly by the action of enzymes elaborated in the liver cells, which convert dextrose into glycogen, or glycogen 74 SAMUEL HANSON into dextrose, depending respectively on whether the glucose concentration in the blood is in the upper or lower limits of the normal variation. Is it not then also probable that a ‘similar mechanism serves to adjust the constancy of the protein quotient? (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) REFERENCES CERVELLO, C.: Arch f. exper. Path. u. Pharm., 1910, 42, 357. CrRVELLO, C.: Arch. f. exper. Path. u. Pharm., 1911, 64, 403. Hurwitz, S. H. anp K. F. Meyer: Jour. Exper. Med., 1916, 24, 515. Scumipt, E. S. anp Cart L. A. Scumipt: Jour. of Immunology, 1917, 2, 343. Hanson, S. ano I. McQuarrie: Jour. of Pharm. and Exper. Therap., 1917, 10, 261. Satviou1, G.: Arch. f. Physiol., 1881, S. 269. Burcxuarp, A. E.: Arch. f. Exp. Pathol. u. Therap., 1883, 16, 322. LewinskI, J.: Pfluger’s Arch. f. d. gesammte Physiol., 1903, 100, 611. Briees, R. S.: Jour. Biol. Chem., 1915, 20, 7. Rosertson, T. BrRArLsrorpD: Jour. Biol. Chem., 1915, 22, 233. Rowe, A. H.: Arch. of internal Med., 1917, 19, 499. Voit, E.: Hermann’s Handbuch, 1881, 6, 1, p. 89. TIGERSTEDT, Rosert: Lehrbuch der Physiologie des Menschen, 3 Aufl., Bd. 1, p. 111, Leipsic, 1905. (12) Howetu, W. H.: Text-Book of Physiology. 4th edition, pp. 888-889. THE IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT ALAN C. WOODS From the John Herr Musser Department of Research Medicine, University of Pennsylvania Received for publication January 12, 1918 In the course of a detailed study of the anaphylactic theory of sympathetic ophthalmia, a study has been made of the anti- genic properties of the uveal tract, especially the uveal pigment, of the eye. From the standpoint of the ophthalmologist the results are of little import beyond their possible correlation with clinical disease. From the standpoint of the immunologist, however, the results appear to be of more scientific interest, in so far as they indicate rather striking immunologic properties for a native body protein. It is, therefore, the purpose of this paper to report the results of this phase of our studies. The anaphylactic theory of sympathetic ophthalmia, se and advocated by Elschnig, assumes that sympathetic ophthal- mia is an anaphylactic uveitis brought about in the follow- ing manner. The injury to the uvea in the exciting eye, by trauma, intraocular tumor, etc., leads to a destruction or disin- tegration of uveal tissue. This uveal tissue is absorbed and acts as an antigen, producing a hypersensitiveness of the organ- ism and especially of the other eye. A reaction now takes place between the sensitized cells of the uvea of the second eye and the antigen circulating in the blood or lymph. This anaphy- Jaetic reaction or intoxication is manifested clinically as a sym- pathetic ophthalmia. This: theory, of course, assumes that the cells of the uveal tract, or some constituent of these cells, possess antigenic prop- érties in the homologous animal, and further assumes organ 75 76 ALAN C. WOODS specificity and lack of species specificity for the protein involved. Elschnig fully realized these assumptions and in his presenta- tion of the theory in its present form, presented experimental work to substantiate these assumed points. HISTORICAL Elschnig’s work (1), appearing in a series of papers, may be summarized as follows. Using as his methods the intraocular injection of sheep erythrocytes and cholera vibrio extracts, and as his indices the hemolytic titer and agglutination reactions of the blood serum, he established the fact that absorbtion from the eye could lead to immune body production. He then sought to determine three points, to give scientific support to the ana- phylactic theory. These points were: (I) Does uveal tissue possess antigenic properties? If so, in the homologous animal? (Il) What constituent of the uveal tract is responsible for such properties? (III) What are the antigenic properties as regards organ and species specificity? Elschnig immunized rabbits against various heterologous uveae, and against homologous uvea, and finally against so-called ‘‘chemically pure’ pigment from the uveal tract of various animals. Thus he obtained hetero-immune and iso-immune uvea serum, and pigment- immune sera. He then used the complement fixation reaction for the detection of immune bodies in these sera and the study of their properties. He found that heterologous uvea produced immune bodies, as would be expected. Homologous uvea, also, acted as antigen and produced immune bodies. The sera of animals immunized to uvea fixed complement with any uvea antigen, irrespective of species—the immune bodies were organ specific. As regards species specificity, the sera of animals immunized to an emulsion of whole uvea were species specific in their reactions. The sera of animals immunized to pigment, however, were not species specific. The reason for this apparent anomaly is evident. Emulsion of whole uvea contains two elements, (1) pigment which is not species specific and (2) such IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT a tissue as blood, smooth muscle and connective tissue which is species specific. Immunization with heterologous uveal emul- sion gives an immunity not only to non-species specific pigment, but also to the other species specific elements in the uveal tract. The species specificity of uvea immune sera is due to these last elements. Elschnig therefore concluded that uveal tissue could act as antigen in the same animal, and that the pigment was the constituent responsible for this property. The pigment was organ specific and not species specific. The remaining work on this subject may be quickly sum- marized, Weichardt and Kummel (2), using the doubtfully valuable epiphanin reaction, substantiated Elschnig’s findings. Likewise, with the epiphanin reaction, and complement fixation, Kummel (3), demonstrated uveal antibodies in a percentage of the sera of patients with sympathetic ophthalmia. Wissman (4), supported the anaphylactic theory by experimental work and showed uvea immune bodies in the serum of two sympa- thetic ophthalmia patients by the precipitin reaction, but failed to substantiate Kummel’s work with complement fixation, on the sera of these patients. Rados (5), substantiated Elschnig’s work only partially. Fuch and Meller (6), were unable to demonstrate uveal antibodies in the sera of patients with sym- pathetic ophthalmia. Von Szily (7), after lending much valu- able support to the theory criticizes the antigenic properties of pigment. OUTLINE OF WORK In our study of the various phases of the anaphylactic theory of sympathetic ophthalmia, we have studied primarily the immunologic properties of uveal tissue, and especially of uveal pigment. This has been done in two ways, first by a repetition of Elschnig’s work with the complement fixation reaction of the sera of immunized animals, and secondly this has been con- firmed by direct perfusion experiments upon the eyes of sensi- tized animals. 78 ALAN C. WOODS I. THE ANTIGENIC PROPERTIES OF UVEAL TISSUE AS SHOWN BY COMPLEMENT FIXATION This, as before mentioned, is substantially a repetition, in toto, of Elschnig’s work. Dogs were immunized, by intraperitoneal injections repeated every six days, against cow’s uveal emulsion, cow’s uveal pig- ment and dog’s uveal emulsion and uveal pigment.! In order to avoid non-specific fixation, the sera of the dogs selected for this work were all examined, in the complement fixation reac- tion, with the various uveal emulsion and uveal pigment anti- gens, before immunization was begun. Eight dogs giving completely negative reactions with all antigens in this prelimi- nary work, were selected for immunization. To immunize, intraperitoneal injections of uveal emulsion and uveal pigment were given to respective dogs at six day intervals. The initial injection was 5 ce. which was increased 1 ec. at each injection until 10 ec. was reached. A total of eight immunizing injections was given each animal. At suitable intervals after the last injection, the sera of these dogs was tested again against all the antigens. The results are shown in the tables 1 to 4. In order to quantitate the strength of the reaction, three different quantities of sera were used with every antigen. A +-+-+ reaction indicates complete, er practically complete, fixation of complement with all quanti- ties of sera, a ++ reaction similar fixation of complement in the tubes containing two largest amounts of sera, and a single + reaction indicates that such fixation was observed only in the tube containing the maximum amount of serum. The first two complement fixation reactions were done with the same antigens used in the preliminary reaction. The last reaction was done with freshly prepared antigens. All antigens were used in one-third the anticomplementary dose. All animals showed immune sera, differing, as might be ex- pected, in the various dogs. Dogs 16-96 and 16-98 showed the lowest immunity. i For the details of technique used to prepare the uveal emulsion and uveal pigment consult references (8) and (12). IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 79 Table 1 illustrates the complement binding phenomena shown by heterologous immune sera. As would be expected these sera gave complement binding with their specific antigens. More- over, the uvea-immune sera gave fixation of complement with pigment antigens. Pigment immune sera fixed complement with pigment antigens, showing here that the pigment itself is capable of acting as antigen. TABLE 1 Heterologous sensitization (cow’s uvea and pigment) FEBRU- ea) _|ARY 21, SERA ANTIGEN DECEMBER 14 ea nee aa FRESH 3 ANTI- - GENS (| Cow’s uvea | Negative el ele 16-89. Uvea immune... / ++) $/44+4+ _| Cow’s pigment | Negative +++\|4++)/+++ +) +4+]+++ Cow’s uvea Negative 16-90. U 1 aa eel : é : 6 20. Pires mamune Cow’s pigment | Negative +++] +4]4+4++ ++| ++/+++ Cow’s uvea Negative Cow’s pigment | Negative Period of immunization 16-91. Pigment immune BMG cit i {| Cow’s uvea Negative sates (ines eel Peace a a \| Cow’s pigment | Negative +++) +4]/4+++ Table 2 illustrates the complement binding reactions of the sera of dogs immunized to dog’s uveal emulsion and dog’s pig- ment—iso-immune sera. The iso-immune sera gave fixation of complement with their specific antigens, showing the presence of iso-immune bodies. In other words, uveal tissue possesses the power to act as a foreign protein—as antigen—in animals of the same species. This is the first, and cardinal point, to be proven in order to establish the anaphylactic theory. Further- more, the purified pigment of the uvea possesses this power of immune body production in the homologous animal, and uvea immune sera gives fixation of complement with pigment anti- gens. This substantiates Elschnig’s contention that it is the pigment which is the responsible factor for the peculiar antigenic properties of uveal tissue. 80 ALAN C. WOODS TABLE 2 Homologous immunization (dog’s uvea and pigment) SERA ANTIGEN DECEMBER 14 16-95. Uvea immune. 16-96. Uveaimmune. 16-98. Pigment im- FINN G5 hee evi syns ase 16-99. Pigment im- Dog’s uvea Negative Dog’s pigment | Negative Dog’s uvea Negative Dog’s pigment | Negative Dog’s uvea Negative Dog’s pigment | Negative Dog’s uvea Negative Dog’s pigment | Negative SERA 16-89. Cow’s uvea LINIMNUNES. -o eee 16-90. Cow’s uvea IMMUNE S - eee 16-91. Cow’s pigment (MMUNE.22- a. eee it 16-92. Cow’s pigment ATTN Cone eee eae 16-95. Dog’s uvea IMMUNE. oss ge see 16-96. Dog’s uvea ATMAMUING ea ee 16-98. Dog’s pigment { ATMUITUUN G2 5,28 eee 16-99. Dog’s pigment IMTS 2.5 = ho creree if \| Dog’s pigment TABLE 3 Organ specificity ANTIGEN DECEMBER 14 Negative Negative Dog’s uvea Negative Negative Dog’s uvea Dog’s pigment Negative Negative Dog’s uvea Dog’s pigment Negative Negative Dog’s uvea Dog’s pigment Cow’s uvea Cow’s pigment Negative Negative Negative Negative Cow’s uvea Cow’s pigment Negative Negative Cow’s uvea Cow’s pigment Cow’s uvea Cow’s pigment Negative Negative Period of immunization Period of immunization ae ae iek FEBRUARY 14 Pres GENS Soa Bist el (i aes ie = i ios (LS er ++] eae +] Negative| ++ aE 3 + aE ate ++ +] ° a een tale. aia = JAN- FEBRU- teed FEBRUARY 14 pete ARY 6 a +) eae a ca ++ Sisk + 4 +} Negative + ++) Negative|. ++ +| Negative} ++ seme |, sh aie “Ee SE ee SE | GEN eaatogee shF+] +a) | Ice saan Neca |r aber west aie seapaet Gren |E=55 = feasts | ais ois Seas SESRS5i| Sea eee IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT Sl Table 3 illustrates the organ specificity of the immune sera. Iso-immune sera gave positive complement binding reactions with heterologous antigens, and the heterologous immune sera gave complement binding with the iso-antigens. In other words the immune bodies were organ-specific, showing an affinity for anti- gens of the specific tissue, without regard to the species from which obtained. This substantiates Elschnig’s second point of the organ specificity of uveal tissue. The pigment-immune sera show this property also and it is evidently to the pigment constituent that this organ specificity of uveal tissue is due TABLE 4 Species specificity l JANUARY 31 eae SERA ANTIGEN AND FEBRUARY 14 aaSCie FEBRUARY 6 ANTIGEN 5 16-89. Cow’s uvea im- {| Cow’s liver = Saas + ++ (20.6 > Sea || Cow’s kidney " apse See Sear 16-90. Cow’s uvea im- Cow’s liver +++ +++ +++ MAUITENS «SPAM ALN Satins fie! de’ « Cow’s kidney| (= | +++ +++ +++ ©) 16-91.Cow’s pigment im- Cow’s liver S Negative | Negative ++ PAE eee tl. . Cow’s kidney 2 + | Negative + 16-92. Cow’s pigment im- if Cow’s liver + + | Negative PAINE hy EIN Ses. 2s || Cow’s kidney + “ob +4 Table 4 illustrates the question of species specificity. With the sera of animals immunized to cow’s uveal emulsion and cow’s uveal pigment, it was determined whether or not fixation of complement occurred with antigens of cow’s protein, other than uveal tissue. It will be seen by a study of this table that the uvea-immune sera give complete fixation of complement with antigens of cow’s liver and kidney extract. The reason for this is that already stated—that these dogs were immunized to the whole uvea, which contained not only the uveal pigment, but also some blood, smooth muscle and connective tissue of the cow. The fixation of complement with the liver and kidney 82 ALAN C. WOODS extracts is evidently dependent upon the immunization to this last factor. On the other hand, the sera of dogs immunized to cow’s pig- ment alone gave either weak or negative reactions with cow’s liver and kidney. The pigment was purified as much as possible, but it probably still contained traces of the other elements of the uveal tract—the blood, smooth muscle and connective tissue. The weak reactions occasionally observed are probably due to this impurity. However, the difference in the degree of fixation of comple- ment between the uvea immune and the pigment immune sera with the same antigens, is striking. Although in this phase of the work, our results are not so clear cut as those of Elschnig, it seems probable that the pigment acting as antigen, lacks at least in a degree, species specificity. These observations of the complement binding reactions of uvea and pigment immune sera corroborate Elschnig’s findings. Heterologous and homologous uveal tissues have the power of acting as antigens. In the case of homologous uveal tissue, the pigment is the factor responsible for its antigenic properties in animals of the same species. ; Whole uveal emulsion is both organ specific and species specific, the species specificity probably being due to the blood, smooth muscle, and connective tissue in the emulsion. The pigment, however, is organ specific and probably not species specific, and in this respect is analogous to lens protein and differs from other common body protein. These properties of uveal pigment—ability to act as foreign protein to animals of the same species, organ specificity, and lack of species specificity—are as before emphasized, funda- mental properties uveal tissue must needs possess to make the anaphylactic theory a possibility. _— niet IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 83 Il THE ANTIGENIC PROPERTIES OF UVEAL TISSUE AS SHOWN BY PERFUSION OF THE EYES OF SENSITIZED ANIMALS a. The reaction of the sensitized eye to perfusion with specific antigen With the confirmation of Elschnig’s fundamental work we next sought, both for the purpose of again confirming this work, and of more closely correlating these findings to the clinical disease, to determine whether these peculiar antigenic properties indicated by complement fixation, could be manifested in a direct anaphylactic reaction upon the eye, through general sensitization and vascular intoxication. To this end, we resorted to the perfusion of the eye. Figure 1, shows diagramatically the operation employed and the position of the inflow and out- flow cannulas. The apparatus employed has already been reported (8). Fresh defibrinated dogs blood to which sufficient Ringer’s solution was added to bring the red corpuscle content approximately to normal, was the perfusion fluid used. With the technique used we were dealing with what was to all purposes a living eye, maintained on an artificial circulation with defibrinated blood, oxygenated by an artificial lung. AI- though the dog’s heart always ceased to beat after the final ligature was placed above the heart, nevertheless the winking reflex persisted in the eyes often for an hour or more. In this experiment the eyes were perfused for three hours, constant observations being made throughout that period. It was found that when a sensitized dog was perfused with the defibrinated blood of a normal dog, there was no ocular reaction. Wheh, however, the specific antigen (the antigen to which the dog was sensitized) was added to the perfusion fluid, a prompt contraction of the pupil occurred, and as the perfusion continued, small hemorrhages appeared throughout the fundus. The contraction of the pupil was marked, usually from a dilated pupil 10 to 12 mm. in diameter to a pupil from 2 to 4 mm. in diameter. This observation is in direct accord with those of Dale (9) and Schultz (10) who observed the contraction of . sensitized smooth muscle in the presence of specific antigen, and THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 84 ALAN C. WOODS is another observation in support of the cellular theory of ana- phylaxis. The hemorrhages observed cannot be so easily SpiensesQe E o pattern’ meee es é i i g oe < > : PA TBE EPEC PELL i 3 ta sae - - - ml ue . Scat ; yee i, Inflow Cannula 4 FE 7; Brachio | Cephalic oe Ny, ~ : ~* Aorla Cannula Fia. 1 explained. Petechial hemorrhages have been observed over the peritoneum of animals recovering from anaphylactic shock, and it may be that these are analogous. From their method of IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 85 formation it appeared that they were caused by an alteration of the endothelium of the capillaries, allowing a diapedesis of red cells. In this experiment, the antigens used were horse serum and cows uveal emulsion. Before proceeding with the second phase of the perfusion work a number of perfusions were done to con- trol this observation. These controls are illustrated in table 5. One of these controls requires a little explanation. Normal dogs perfused with horse serum showed a slight contraction of the pupil, never over 3mm. Schultz (11) previously has shown that fresh horse serum possesses the power to produce a slight contraction of smooth muscle, and this is evidently what oc- TABLE 5 Contrgls RESULTS SENSITIZATION PERFUSION FLUID = Contraction of the HeMIOR : rhages in pupil fundi None DEB None None None D. B. + Uveal emulsion None None None D. B. + Horse serum Never over 3 mm.| None Uveal emulsion Dees None None D. B. is defibrinated blood of normal dogs. eurred with us. This contraction of the pupil in normal dogs, perfused with horse serum, was never over 3 mm. while the con-- traction in the sensitized dogs was from 8 to 10 mm. Normal dogs perfused with uveal tissue showed no contraction of the pupil. In these and in all subsequent perfusion experiments here reported, each individual observation was controlled by at least two perfusions. In those experiments where for any reason (clots, thrombosis) the nature of the reaction was not clear after two perfusions had been performed, subsequent perfusions were done until the presence or absence of the anaphylactic phenomena was established beyond question. From this it was evident, therefore, that the eyes may be sensitized as a part of general sensitization, and that anaphy- 86 ALAN C. WOODS lactic phenomena may be elicited in the eyes by means of antigen carried by the blood stream. ‘These anaphylactic phenomena consist in a marked contraction of the pupil, and in small extra- vasations of blood throughout the fundus. With the artificial condition, under which we were working, it is manifestly impossible to expect inflammatory phenomena of any kind, but the establishment of the fact that ocular ana- phylaxis, however manifested, may be demonstrated by a gen- eral sensitization and vascular intoxication, gives us the second important point in the establishment of a scientific basis for the anaphylactic theory. Moreover, this observation affords us a means of studying the anaphylactic properties of uveal tissue, in vivo, by direct observation of the eye. b. The antigenic properties of vveal tissue in the production of the perfusion anaphylactic reaction Having shown that an anaphylactic reaction can be obtained through the perfusion of the eye with specific antigen, this reaction has been used to determine the antigenic properties of uveal tissue. We first sought to determine whether whole uveal tissue possessed the antigenic properties necessary to make an anaphylactic uveitis a possibility. When this point was established, we sought to determine the constituent of uveal tissue responsible for its peculiar antigenic properties. Pre- liminary experiments with uveal pigment showed that there was much reason to believe, with Elschnig, that the pigment was the responsible factor. A pigment solution was finally prepared which was suitable for use in perfusion (12). Experi- ments similar to those in which uveal emulsion was used as antigen were then performed to determine whether the pigment is the responsible factor. The reactions given by the pigment were in every way similar to those given by whole uvea, except as regards species specificity. Moreover, dogs sensitized to homologous uvea, reacted to perfusion with pigment, establish- ing more conclusively the fact that the pigment is responsible for the peculiar antigenic properties shown by uveal tissue. IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 87 _ The results of this work are shown in the following tables (6 to 9). Table 6 represents the results obtained by sensitization and perfusion with heterologous (cow’s) uveal emulsion and uveal pigment. The anaphylactic reaction was observed in all of these perfusions, establishing the power of both uveal emulsion and pigment to act as antigens. TABLE 6 Heterologous sensitization (cow’s wea and pigment) RESULT SENSITIZATION PERFUSION FLUID Contrac- Hemor- tion of rhages in the pupil fundi Cow’s uveal emulsion D. B. + Cow’s uveal emulsion! Marked | Marked Cow’s uveal pigment D. B. + Caw’s uveal pigment | Marked | Marked D. B. is defibrinated blood of normal dogs. Table 7 illustrates the fundamental fact necessary in the anaphylactic theory—namely the power of homologous uveal tissue to produce an ocular anaphylactic reaction in a sensitized eye when carried there through vascular channels. It shows also that it is the pigment which is responsible for this remark- able antigenic property. Dogs sensitized to homologous uveal emulsion give an anaphylactic reaction when perfused with normal defibrinated blocd to which pigment is added, and dogs sensitized to pigment give the reaction when perfused with the pigment-containing blood. TABLE 7 Homologous sensitization (dog’s uvea and pigment) RESULT SENSITIZATION PERFUSION FLUID Contrac- Hemor- tion of rhages in the pupil fundi Dog’s uveal emulsion D. B. + Dog’s uveal emulsion| Marked | Marked Dog’s uveal emulsion D. B. + Dog’s pigment Marked | Marked Dog’s uveal pigment D. B. + Dog’s pigment Marked | Marked D. B. is defibrinated blood of normal dogs. 88 ALAN C. WOODS Table 8 illustrates the organ specific property of uveal emul- sion and uveal pigment. As shown before by the complement fixation reaction, uveal tissue and pigment are organ specific. The sensitization resulting from the introduction of uveal tissue into an animal, is specific for uveal tissue, without regard for the species from which the tissue is taken. Dogs sensitized to cow’s uveal emulsion react to perfusion with dog’s uveal emul- sion, and vice versa. Similarly dogs sensitized to cow’s pig- ment react to perfusion with dog’s pigment, and vice versa. There evidently results from sensitization with uveal tissue, a strong chemical affinity for similar tissue,—organ specificity. TABLE 8 Organ snecificity RESULT SENSITIZATION PERFUSION FLUID Contrac- Hemor- tion of rhages in the pupil fundi Cow’s uveal emulsion D. B. + Dog’s uveal emulsion| Marked | Marked Cow’s uveal pigment D. B. + Dog’s pigment Marked | Marked Dog’s uveal emulsion D. B. + Cow’s emulsion Marked | Marked Dog’s uveal pigment D. B. + Cow’s pigment Marked | Marked D. B. is defibrinated blood of normal dogs. Table 9 shows the species specificity reaction of both uveal emulsion and pigment. The same properties are shown here as were indicated by complement fixation, the uveal emulsion is species specific, while the pigment is not species specific. Dogs sensitized to cow’s uveal emulsion give an anaphylactic reaction when perfused with other cow protein, for here the sensitization is not alone with the pigment, but also with the other species specific protein contained in the whole uveal emul- sion. On the other hand, dogs sensitized to the pigment alone show no reaction when perfused with other cow protein. The last perfusion illustrated in this table is largely a control perfusion. Dogs sensitized to uveal emulsion give no reaction when perfused with other dog protein. Perfusion with the other elements of the uveal tract—blood, smooth muscle and connec- tive tissue, evokes no anaphylactic reaction. IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 89 To sum up the work thus far, both the complement fixation reactions of the immune sera, and the perfusion reactions, have shown that uveal tissue possesses the power to act as antigen in animals of the same species, and that the pigment is the con- stituent of the uvea responsible for this property. In its anti- genic action, uveal pigment is organ specific and not species specific. The sensitization resulting from the absorption of uveal pigment is specific. Other body protein can produce no anaphylactic reaction in animals so sensitized. Pigment alone may produce intoxication. TABLE 9 Species specificity RESULTS SENSITIZATION PERFUSION FLUID eee eae tion of | rhages in the pupil fundi Cow’s uveal emulsion | D. B. + Cow’s serum, liver and kidney} Marked] Marked extracts Cow’s uveal pigment | D. B. + Cow’s serum, liver and kid- | None | None ney extracts Dog’s uveal emulsion | D. B. + Dog’s liver, and kidney ex- | None | None tracts | c. Ocular sensitization from antigen absorbed from the other eye To complete our study of the antigenic properties of uveal tissue as manifested in the ocular perfusion reaction, and as an incidental point in our study of the anaphylactic theory, we sought to determine whether uveal tissue, absorbed in one eye, could create a hypersensitiveness to uveal tissue in the second eye. To demonstrate this we sensitized dogs by the injection of uveal tissue, both heterologous and homologous, into the vitreous of one eye. After a suitable period had elapsed for sen- sitization to occur, the injected eyes were enucleated, in order to remove any factor which could give a possible intoxication. One week after this, the remaining eye was perfused with the specific antigen. In every case, the perfused eye gave an ana- phylactic reaction, indicating that ocular sensitization had 90 ALAN C. WOODS taken place as a result of the absorption of antigen from the fellow eye. Table 10 illustrates this experiment. TABLE 10 Ocular sensitization e RESULTS SENSITIZATION PERFUSION FLUID Contrac- Hemor- tion of rhages in | pupil fundi Vitreous injection. | Enucleation left | D. B. + Cow’s | Marked | Marked Left eye. Cow’s | eye uveal emulsion uveal emulsion Vitreous injection. | Enucleation left | D. B. + Dog’s | Marked | Marked Left eye. Dog’s eye uveal emulsion uveal emulsion DISCUSSION As is well known, Uhlenhuth (13) and Haendel (14) have established the fact that animals can be sensitized to their own lens protein, and Rosenau and Anderson (15) have found that guinea-pigs can be sensitized to guinea-pig’s placenta. From the evidence brought out by the researches of Elschnig and this work, it seems that uveal pigment possesses these same anti- genic properties. However, we realize that we have not demon- strated the actual point of auto-sensitization, the sensitization of a dog to the pigment from his own eye. The technical diffi- culties in the way of such a proof are enormous. One attempt to demonstrate such auto-sensitization by general anaphylactic reactions has been unsuccessful. The eye of a dog was removed, the uvea excised and macerated, and injected to sensitize. Two weeks later the second eye was removed, the uvea excised and macerated, and this again injected intravenously under ether anesthesia, and this dog was observed for a fall in blood pres- sure, change in the coaguability of the blood, and drop in body temperature. No such signs were observed and we have little hope of demonstrating auto-sensitization by this method. A more delicate method must be devised. For the present we must content ourselves with the demonstration that dogs uvea can produce ocular anaphylactic phenomena in the dog. IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 91 CONCLUSIONS The pigment of the uveal tract of the eye possesses the prop- erties of acting as antigen in homologous animals, and in its immunologic reactions is organ specific and not species specific. These findings can be demonstrated by the complement fixation reaction with the sera of properly immunized animals, and by perfusion experiments on the eyes of sensitized animals. In the case of the perfusion experiments, the anaphylactic reaction is manifested by a marked contraction of the pupil, and the occurrence of small hemorrhages in the fundus. This reaction was used to study the antigenic properties of uveal pigment, and the results shown by complement fixation confirmed. REFERENCES (1) Etscunte, A.: Studien zur Sympathischen Ophthalmia. JI. Wirkung von Antigenen vom Augeninnem aus. Arch. f. Ophth., 1910, 75, 459. II. Die antigene Wirkung des augenpigments. Arch. f. Ophth., 1910, 76, 509. III. Teil Arch. f. Ophth., 78, 549. IV. (Elschnig, A. and Salus, R.) Wirkung des Augenpigments, Arch. f. Ophth., 1911, 79, 428. (2) Wetcuarpt, W., AND Kummet, R.: Studien itiber Organ-Spezifitit des Uvaeiweisses. Miinch. Med. Wochenschr., 1911, 58, 1714. (3) Kummet, R.: Versuche einer Serumreaktion der sympathischen Ophthal- mie. Arch. f. Ophth., 1912, 81, 486. (4) Wissan, R.: Ueber Versuche mit Augen-extracten. Arch. f. Ophth. 1911, 80, 399. (5) Rapos, A. : Ueber das Auftreten von Komplementbindenden Antikorpern nach Vorbehandlung mit arteigenen Gewebezellen, nebst Bemer- kungen tiber die anaphylaktische Entstehung, der Sympathischen Ophthalmie. Zeitsch. f. Immunititsf. 1913, 20 (orig.), 305. (6) Fucus, A., AND Metuer, J.: Studien zur Frage einer Anaphylaktischen Ophthalmie. Arch. f. Ophth. 1914, 88, 280. (7) von Satuy, A.: Anaphylaxie versuche mit sog. chemisch reinem Augen- pigment (von Rind, Schwein, und Kamnchen), nebst Pathologisch anatomischen Untersuchungen. I. Teil. Klein. Monatsbl. f. Augen- heilk., 1916, 56, 79. (8) Woops, A. C. : Ocular anaphylaxis. I. The reaction to perfusion with specific antigen. Archives of Ophth., 1916, 45, 557. (9) Daun, H. H.: The anaphylactic reaction of plain muscle in the guinea-pig Jour. of Pharm. and Exp. Therap., 1913, 4, 167. (10) Scuurrz, W. H. Physiological studies in anaphylaxis. I. The reaction of smooth muscle of the guinea-pig sensitized with horse serum, Jour. of Pharm. and Exp. Therap., 1910, 1, 549. 92 ALAN C. WOODS (11) Scuunrz, W. H. : Physiological studies in anaphylaxis. IV. Reaction of the cat towards horse serum. Jour. of Pharm. and Exp. Therap., 1912, 3, 219. (12) Woops, A. C. : Ocular anaphylaxis. III. The réle of uveal pigment. Arch. of Ophth. 1917, 46, 283. (13) Ustennurn P. : Zur Lehre van des Unterscheidung verschiedener Eiweis- sarten mit Hilfe- spezifischer Sera. Festschrift zum _ sechsigsten Geburtstage Robert Koch, Jena 1903. (14) Untennuth anp Harnpex. Untersuchungen iiber die praktische Verwert- barkeit der Anaphylaxie zur Erkennnng und Unterscheidung ver- schiedener Eweissarten. Zeitschr. f. Immunititsf., 1910, 4, orig., 761. (15) Rospenav, M. J. anp ANDERSON, J. L. : Further studies upon anaphylaxis. U.S. Public Health and Marine Service Hygienic Laboratory Bulletin, no. 45, 1908. THE EXAMINATION OF THE BLOOD PRELIMINARY TO THE OPERATION OF BLOOD TRANSFUSION ARTHUR F. COCA From the Department of Bacteriology in Cornell University Medical College and the New York Orthopedic Hospital Received for publication January 16, 1918 Before the operation of blood transfusion is carried out it is necessary to test the compatibility of the prospective donor’s blood with that of the patient; that is, it must be determined whether the blood corpuscles of the donor will remain intact in the circulation of the patient. It is altogether probable that no two human individuals possess blood having exactly the same chemical composition; indeed, the possible different com- binations of demonstrated different substances in the corpuscles alone have been estimated to be about 4000 (1). Fortunately, however, the differences in the blood plasma and most of those in the corpuscles, although they are, perhaps, the cause of certain unpleasant symptoms, such as a chill followed by a rise of temperature, do not constitute an incompatibility such as could contraindicate the use of the blood of an individual for transfusion. Incompatibility in the latter sense is always due to the presence in the blood of the one individual of isoag- glutinins or isohemolysins or both of these agencies against the corpuscles of the other individual. As the isohemolysins are never present without associated isoagglutinins and since the latter are sometimes found alone, it is necessary, for the purpose under consideration, to examine the blood of the respective individuals only for the presence of isoagglutinins. Through the studies of Landsteiner and of von Dungern it is known that there are two different isoagglutinins in human blood.! These have been designated with the capital letters A 1 A discussion of the large body of evidence supporting this assumption would lead too far from the practical purpose of this paper. 93 94 ARTHUR F. COCA and B. The respective agglutinable elements in the corpuscles are designated with the small letters a and b. An isoagglutinin, eg., “A” and its corresponding agglutinable substance ‘“a”’ cannot be present in the same person, as this would result in the agglutination of the individual’s corpuscles by his own plasma. All of the possible combinations of the substances A, B, a and b are found in human beings and individuals have been separated into four groups in accordance with the different combinations of those substances. This grouping is shown in table I. TABLE 1 Plasma’... v8 4. oe A and B (Plasma... :0.3-2 nee A SrepP iy esas pal ep satiaae omer gan eeue dt \'Corpuseles?:.9.255= ee b Pigsma 35515. eee B 7 { Plasma... 13245. Speer 0 Sn ee ba Slee Peer Eee a See da \ Corpuscles.........:. a and b The grouping is schematically shown in figure 1, which consists of four test tubes containing blood of the four groups of individuals. Since the plasma of individuals of group I contains both A and B isoagglutinins it is able to agglutinate the corpuscles of the individuals of all of the other groups all of which contain one or both of the agglutinable substances a and b. Similarly it is clear that a group II plasma will agglutinate group III or group IV corpuscles but not those of group I. Group III plasma will agglutinate group II or group IV corpuscles but not those of group I. Finally, group IV plasma is not able to agglutinate any corpuscles since it lacks both isoagglutinins. The proportional (percentage) representation of the four groups among human individuals has been estimated in three different studies as follows: VON DUNGERN AN W. L. Moss (3) HIRSCHFELD (2) OLMSTEAD (4) Group els a42 :.3.) 22 clse ee oo eee 36.0 43 46 Croup wll e en. a een peee accep eee 47.0 40 39 Groups ne se Sc sateen vee 11.0 i 13 Group Vs in. eee eee ote eee 5.7 10 2 These results show that by far the greater number (about a EXAMINATION OF BLOOD BEFORE TRANSFUSION 95 83 per cent) of human individuals are nearly equally divided between groups I and II. It is evident that donors that are of the same group as the patient should be preferred for the transfusion, but if such a donor is not available then one of another group must be selected whose corpuscles are not agglutinated by the patient’s plasma. In such a case the donor’s plasma will usually contain agglutinins for the patient’s corpuscles but this fact does not constitute a contraindication to the use of the donor’s blood for the trans- Fic. 1. Bhoop Grourina or HumAN INDIVIDUALS (SCHEMATIC) The corpuscles are represented by the circles, the plasma by the space sur- rounding the circles. fusion because the donor’s plasma is so diluted in the patient’s circulation that its agglutinating or hemolytic power is reduced below the point of injury to the patient’s corpuscles. The testing of patient’s and donor’s blood preliminary to transfusion may be done directly or indirectly. DIRECT TEST For carrying out the direct test it was formerly customary to obtain several cubic centimeters of blood by venepuncture from both patient and donor a small part of which was mixed with 96 ARTHUR F. COCA sodium citrate solution the remainder being allowed to clot. The citrated blood was washed with normal saline solution and in a dilute (1-10) suspension it was mixed with the clear serum of the other individual. After an incubation of one or two hours at 37°C. the presence of agglutination and hemolysis was determined macroscopically. Weil (5) took the first step in the simplification of this time consuming procedure by mixing the citrated blood of the two individuals in three proportions: namely, one part of A with nine parts of B; one of A with one of B; and nine of A with one of B. He found that the phenomenon of agglutination was not interfered with by the ‘presence of the even more numerous corpuscles not taking part in that reaction, and that the reaction could be observed macroscopically after the usual incubation period. Hemolysis, also, could be detected by the presence of hemoglobin in the supernatant fluid of the mixtures after centrifugation. A further modification in the interest of economy of bloed was contributed by Rous and Turner (6). These authors using the ordinary white blood cell mixing pipet drew up first, 10 per cent sodium citrate to the mark 1 and then blood from the finger to the mark 11 and blew out the citrated blood into a narrow test tube. The two bloods were then mixed with the use of the Wright capillary pipet in the proportions suggested by Weil, and the mixtures sealed in the capillary pipet by fusing the tip. After fifteen minutes the tips of the sealed pipets were broken and a drop of the mixtures examined microscopically in normal saline solution for agglutination. The following method, which is based on a suggestion of Dr. James Ewing, permits the mutual tests to be made with prac- tically a single drop of blood from each individual and with the use of little more than the usual apparatus employed for a white blood cell count. The patient and the donor or donors being in the same or adjoining rooms, a finger of each individual is prepared as for an ordinary blood count, for which purpose soap and water meet every requirement. Three ordinary glass slides are placed oe OS See RR es cag ey = EXAMINATION OF BLOOD BEFORE TRANSFUSION 97 in order as shown in figure 2. For convenience the right and left ends of all of the slides or of only slide 3 may be marked A and 5, the first letter referring to the patient, the second one referring to a donor. The stem of a white blood cell mixing pipet is filled with normal saline solution up to the ninth mark and the fluid thus measured (nine divisions) is deposited on slide 1 at the right (position B). The pipet is then washed out with a 10 per cent solution of sodium citrate, enough of this solution being left in the stem to fill the terminal division. The 1 9-Nacl ~1-Blood-B rs 9-Blood-A }- _J9-Bl 00d-B L 1-Citrate 1-Citrate 3 3-cit. Blood-A }- {| ae _j3-cit. Blood-B $-dil. Blood-B : = $-cit. Blood-A Fic. 2. Ssaowrna ARRANGEMENT OF SLIDES IN THE BLoop TEST. patient’s finger is punctured as for a blood-count and the blood is drawn up into the stem of the pipet to the mark 1. This mixture of blood and citrate solution is deposited on slide 2 at the left and a similar mixture of the donor’s blood with citrate solution is deposited at the right of the same slide, the pipet being thoroughly washed, between the two operations, first with saline solution and then with the citrate solution. One division of the donor’s citrated blood (B) is well mixed with the nine divisions of saline solution at the right of slide 1, the resulting mixture thus representing a 1 to 10 dilution of the donor’s blood B. If equal parts of this diluted blood and of the undiluted citrated blood of the patient A are mixed, the resulting mixture 98 ARTHUR F. COCA will represent 9 parts of blood A and 1 part of blood B in 10 parts of saline diluent. Such a mixture is made by drawing up with the white blood cell mixing pipet, first, 3 divisions of ‘ci- trated blood-A at the left of slide 2 and then 3 divisions of diluted blood-B, the two portions being then deposited together at the left of slide 3, stirred once with the tip of the pipet and covered with a ¢ inch cover-glass. At the right of slide 3 is placed a mixture of 3 divisions each of the two undiluted citrated bloods at right and left of slide 2, this mixture, likewise, being covered with a cover-glass. In making this second mixture, which, in a sense, is intended as the reverse of the first mixture, it is not necessary to dilute the patient’s citrated blood —-A— ; first, be- cause, in most instances, the corpuscular content of the patient’s blood is reduced sufficiently to meet the purpose of the dilution and secondly, because the quantitative relation in the mixture, as far as the ratio of donor’s plasma to patient’s corpuscles is concerned, is nearer that at the actual transfusion than if the patient’s blood were used in a 1 to 10 dilution. If, however, mutual tests are being made in two normal individuals in order, for example, to determine their group relationship, then blood-A must be diluted 1 to 10 before being mixed with blood-B in the second mixture. Agglutination usually begins at once and can easily be detected with the microscope as it has been described by previous authors. If agglutination does not appear the observation should be con- tinued for fifteen minutes, at the end of which time only the borders of the covered blood films will have begun to dry. To recapitulate: three ordinary glass slides are placed in order as shown in figure 1 and designated 1, 2 and 3, the third slide being marked left and right for identification with the letters A and B. With a white blood cell mixing pipet a mixture of 9 divisions of the patient’s blood and 1 division of 10 per cent sodium citrate is placed at the left on slide 2 and a similar mix- ture of the donor’s blood and the citrate solution is placed at the right of the same slide. A mixture of 1 division of the citrated donor’s blood and 9 divisions of normal saline solution is placed at the right end of slide 1. Three divisions of the citrated EXAMINATION OF BLOOD BEFORE TRANSFUSION 99 blood-A and 3 divisions of the diluted blood-b are drawn up into the pipet deposited at the left of slide 3 and covered with a cover-glass. Three divisions each of the two undiluted citrated bloods are similarly deposited at the right of slide 3 and covered with a cover-glass. These two latter mixtures are immediately examined microscopically for agglutination. If more than one donor is to be tested the series of slides must be duplicated or triplicated et cetera according to the number of prospective donors. In such case two divisions of the citrate solution are left in the end of the pipet before obtaining the patient’s blood the latter being then taken twice up to the mark 1. This amount suffices for the examination of three donors. INDIRECT TEST If the blood-grouping of patient and donors has been deter- mined it is unnecessary to carry out the direct test because the availability of the donor’s blood for the transfusion can be learned by referring to the constitution of the blood of the dif- ferent groups as shown in table 1 or in figure 1. A patient of group I, for example, requires a donor of group I, the blood of all other groups being incompatible. The indirect test is based on the foregoing principle and for its performance, as Moss pointed out, it requires the storing of sera of the two groups II and III. If the corpuscles of an individual are not clumped by either of these sera, that individual belongs to group I; if they are clumped by both sera they must be group IV corpuscles; if they are clumped only by the group II serum or only by the group III serum then they belong respectively to a group III individual or to a group II individual. The technical procedure of the direct test is applicable also to the indirect test. If the patient and a single donor are to be examined the mixtures on slides 1 and 2 (figure 1) are prepared exactly as for the direct test. Two slides—3a and 3b—must take the place of slide 3 in figure 1 since each blood is to be mixed with both stock sera. Three divisions each of the patient’s undiluted citrated blood are mixed respectively with three divisions of the two sera and placed under cover-glasses at the THE JOURNAL CF IMMUNOLOGY, VOL. III, NO. 2 100 ARTHUR F. COCA two ends of slide 3a and similar mixtures of the diluted donor’s blood with the two sera are placed on slide 3b. Clumping, if present, can be seen at once with the microscope and in a few minutes it becomes apparent to the naked eye. REFERENCES (1) von DuNGERN AND HirscHrELD: Muench. Med. Wochenschr., 1910, p. 741. (2) von DuNGERN AND HirscHFe.pD: Zeitschr. f. Immunitaetsf., 1910, 8, 526. (3) Moss, W. L.: Bull. Johns Hopkins Hospital, 1910, 21, 62. (4) MELENEY, STEARNS, FoORTUINE AND Ferry: Am. J. Med. Sci., 1917, 154, 733. (5) Weitz: Journal of A. M. A., January 30, 1915, p. 425. (6) Rous anp TuRNER: Jour. Amer. Med. Assoc., 1915, 64, 1980. EXPERIMENTS UPON THE PASSIVE TRANSFER OF ANTIBODIES FROM THE BLOOD TO THE CEREBROSPINAL FLUID JOHN A. KOLMER anv SHIGEKI SEKIGUCHI From the McManes Laboratory of Experimental Pathology of the University of Pennsylvania Received for publication January 24, 1918 Since the cerebrospinal fluid under normal conditions is gener- ally free of certain normal or natural antibodies or other con- stituents which may be present in the blood, as, for example, hemolysins for the erythrocytes of various animals, agglutinins for various micro-organisms, diphtheria antitoxin and comple- ment, the mechanism concerned in the production of the cere- brospinal fluid under normal conditions is regarded as an effectual barrier against the entrance of antigens, antibodies and various chemical substances into the cerebrospinal fluid. For this rea- son the presence of antibodies in the cerebrospinal fluid during disease is generally interpreted as indicating infection of the tissues of the central nervous organs with the production of antibodies by tissues in direct communication with the cerebro- spinal fluid or, by an injurious effect upon the choroid plexus and subarachnoid villi, facilitating the passage of antibodies from the blood to the cerebrospinal fluid. As shown by Wile and Stokes (1, 2 and 3) and Hauptmann (4) the cerebrospinal fluid may show the presence of syphilis reagin (the antibody concerned in the Wassermann reaction) and present abnormal chemical and cytological changes in the late primary or early secondary periods of that infection, with or without demonstrable clinical evidences of infection of the nervous system. For this reason Wile and Stokes very properly emphasize the importance of care- ful clinical examinations of the nervous system during the early stages of syphilis, coupled with a thorough examination of the 101 102 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI cerebrospinal fluid and regard the presence of syphilis reagin in the fluid as an indication of spirochaetic invasion of the tissues of the central nervous organs. As agglutinin for B. typhosus may be found in the cerebro- spinal fluid during typhoid fever without clinical manifestations of involvement of the central nervous organs, it would appear possible for 7. pallida or other micro-parasites lodged in tissues other than the nervous tissues to stimulate the production of antibodies to such extent as to reach a high degree of concentra- tion in the blood with a passive or forcible transfer of these anti- bodies into the cerebrospinal fluid in a manner analogous to the transfer or elimination of the syphilis reagin in the secretions of the mammary glands and kidneys. The object of our experi- ments was to determine whether antibodies introduced into the venous blood of normal experimental animals passed into the cerebrospinal fluid and more particularly if this occurred after the introduction of the syphilis antibody or reagin, in order to study by experimental means whether the presence of antibodies in the cerebrospinal fluid is to be accepted as indicating the presence and activity of the respective antigen or antigens in the tissues of the central nervous organs, or, whether antibody in high concentration in the blood may be passed into the cerebrospinal fluid by the uninjured mechanism governing the production of cerebrospinal fluid. EXPERIMENTAL Dogs were employed in all experiments because of the ease with which amounts of blood-free cerebrospinal fluid up to 2 ce. may be secured by spinal puncture. In experiments concerning the passive transfer of the syphilis reagin from the blood to the cerebrospinal fluid, normal dogs were bled from the carotid artery under ether anesthesia and an equal amount of human serum. from active cases of syphilis was injected intravenously. Was- sermann reactions were conducted with the blood and cerebro- spinal fluid of each animal prior to injection and at varying intervals afterward; in conducting these tests, the sera were used in amounts of 0.1 ec. and the cerebrospinal fluid in amounts of PASSIVE TRANSFER OF ANTIBODIES 103 0.5 cc. with each of three antigens; namely, an alcoholic extract of beef heart re-enforced with cholesterin; an alcoholic extract of syphilitic liver and an extract of acetone insoluble lipoids. The following protocols of several experiments express the results observed. a. The passive transfer of human syphilis reagin in the blood of a normal dog to the cerebrospinal fluid. Weight 6250 grams; preliminary Wasser- mann reactions with serum and cerebrospinal fluid negative with all antigens. Under ether anesthesia 150 ec. of blood was removed from the carotid artery and 210 ce. of human syphilitic serum yielding ++-+-+ Wassermann reactions with all antigens, were injected into the jugular vein. Wassermann tests with cerebrospinal fluid removed twenty-two hours later were weakly positive. Wassermann tests with blood removed two and four and one-half hours later were moderately positive, twenty-two hours after, the tests were weakly positive and seventy hours later, completely negative. The tests with urine secured four and one-half and twenty-two hours after the injection of syphilitic serum were negative in doses of 0.5 ce. urine with each antigen. b. A second dog weighing 5420 grams was bled 150 ce. under ether anesthesia and 250 ec. of human syphilitic serum yielding ++-+-+ Wassermann reactions with all antigens, were injected into a femoral vein. Preliminary Wassermann tests with blood and cerebrospinal fluid yielded negative reactions with all antigens. Cerebrospinal fluid removed four hours later yielded doubtfully positive reactions with each of the three antigens; fluid removed twenty- four hours after transfusion yielded negative reactions. Wassermann tests with blood serum secured two and four hours after injection yielded moderately positive reactions, while tests with serum secured twenty-four and seventy-two hours after injection were negative with all antigens. c. The passive transfer of human syphilis reagin in the blood of a dog to the cerebrospinal fluid after the preliminary injection of sterile horse serum into the spinal canal. Since the experimertts of Flexner and Amoss (5) with the virus of acute anterior poliomyelitis have indicated that the injection of sterile fluids into the spinal canal of monkeys facilitates the passage of the virus from the blood to the central nervous organs, due presumably to injury to the mechanism governing the pro- 104 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI duction of cerebrospinal fluid, we have conducted experiments similar ~ to those already summarized except that the animals received an intra- spinal injection of 1 cc. of sterile horse serum twenty-four hours before transfusion with human syphilis serum, followed by tests of the cere- brospinal fluid and blood for the syphilis reagin. Dog, weighing 9150 grams, was given an intraspinal injection of 1 ce. sterile horse serum. Wassermann tests with blood and cerebrospinal fluid were negative with all antigens. Twenty-four hours later 150 ce. of blood were removed from the carotid artery under ether anesthesia and 150 ec. of human syphilitic serum yielding ++-+-+ Wassermann reactions with all antigens, were injected into a femoral vein. Cerebrospinal fluid removed four hours later yielded moderately positive Wassermann reactions with all antigens; cerebrospinal fluid removed twenty-two hours after the injection of serum yielded com- pletely negative reactions with all antigens. Blood serum secured four hours after transfusion yielded moderately positive reactions and twenty-two hours after transfusion weakly posi- tive reactions. Tests made with the serum forty-eight hours after transfusion were negative with all antigens. While in this experiment the amount of human syphilitic serum injected was less per kilogram of body weight than employed in the former experiments, the amount of reagin in the cerebrospinal fluid appeared to be somewhat greater as judged by the degree of complement fixation with each of the three antigens; a similar result was observed in the following experiment. d. A dog weighing 7050 grams was given an intraspinal injection of 1 cc. of sterile horse serum. Preliminary Wassermann tests with cere- brospinal fluid and blood serum were negative with all antigens. Twenty-four hours later 150 cc. of blood were removed from the carotid artery under ether anesthesia and 250 cc. of human syphilitic serum yielding +-+-+-++ Wassermann reactions with each of the three antigens, were injected into a femoral vein. Cerebrospinal fluid removed three hours later yielded moderately positive reactions with each of the three antigens; fluid secured twenty- two hours after the transfusion yielded doubtfully positive reactions, after forty-eight hours the fluid yielded completely negative reactions. Blood serum secured three hours after transfusion yielded strongly positive reactions; serum secured twenty-two hours after yielded weakly positive reactions while with serum secured seventy hours after the reactions were completely negative. PASSIVE TRANSFER OF ANTIBODIES 105 e. The passive transfer of typhoid agglutinin (dog) from the blood of a normal dog to the cerebrospinal fluid. As in the experiments already mentioned a heterologous immune serum (human) was employed, it was not surprising that elimination of the syphilis reagin from the blood stream of the dogs was rapid and usually completed within seventy-two hours to such extent that the reagin could not be demonstrated in the blood serum by the Wassermann tests. We have immunized a dog with typhoid bacilli until the amount of agglutinin in the blood reached a high degree of concentration. During the process of immunization the blood serum and cerebrospinal fluid were examined at intervals for the presence of agglutinin. At the completion of the period of im- munization the animal was bled to death under ether anesthesia and the serum was injected into a second normal dog followed by examina- tions of the cerebrospinal fluid for the presence of agglutinins. A dog weighing 9450 grams yielded negative agglutination reactions with an emulsion of typhoid bacilli in final dilutions of 1: 2 with blood serum and cerebrospinal fluid. After a series of fourteen intravenous injections with typhoid vaccine at intervals of four and five days, the serum agglutinated in final dilution of 1: 5120 and the cerebrospinal fluid in final dilution of 1:8 (microscopic technic). The animal was now bled and 150 cc. of serum secured. From a second dog weighing 5000 grams 100 ec. of blood were taken from a carotid artery under ether anesthesia and 150 ce. of the typhoid- immune serum were injected into a femoral vein. Preliminary aggluti- nation tests with the serum and cerebrospinal fluid of this animal yielded negative agglutination tests in final dilutions of 1: 2 with an emulsion of typhoid bacilli. Cerebrospinal fluid removed three hours after transfusion yielded complete agglutination of B. typhosus in dilutions up to 1: 5; fluid removed twenty-four hours later agglutinated 1:4 while forty-eight hours after transfusion agglutination was not in evidence with the lowest dilution; namely, 1: 2. The blood serum removed three and again twenty-four hours after transfusion yielded complete agglutination in final dilutions up to 1: 160; forty-eight hours after transfusion the titer was 1: 40 and ninety- six hours later only partial agglutination occurred in final dilutions of 1:4 and 1:8. 106 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI DISCUSSION These experiments have, in our opinion, demonstrated two facts with special reference to syphilis reagin, namely, that large amounts of this antibody in the blood may result in the passage of small amounts into the cerebrospinal fluid with a normal condition of the central nervous organs and of the mechanism governing the production of the cerebrospinal fluid; secondly, that subarachnoid injection of sterile serum appears to facilitate the passage of antibody by injury to the mechanism governing the production of the fluid or by the production of localized congestion with increased transudation of blood constituents into the cerebrospinal fluid. Accordingly it would appear possible that in syphilis or other acute infections accompanied by a high concentration of antibody in the blood, small amounts of antibody may be found in the cerebrospinal fluid and that this finding does not of itself necessarily indicate infection of the central nervous organs. These observations, however, should have no further significance and do not by any means lessen the value of the studies of Wile and Stokes in syphilis, because they have found the reagin in the cerebrospinal fluid during the pre- roseolar period when the concentration of reagin in the blood cannot be regarded as having reached a point of high concentra- tion and, furthermore, found in some cases an increase of protein and cells in the cerebrospinal fluid, which cannot be explained at present on any other basis than actual infection of the nervous tissues with 7’. pallida. SUMMARY 1. The removal of blood from normal dogs followed by the intra- venous injection of human syphilitic serum in amounts varying from 30 to 50 ce. per kilogram of body weight was followed by the presence of small amounts of syphilis reagin (the antibody con- cerned in the Wassermann reaction) in the cerebrospinal fluid. 2. The reagin was found in the cerebrospinal fluid as early as three hours after transfusion with syphilitic serum; tests at shorter intervals were not made. The amount of reagin found in 0.5 ee. of cerebrospinal fluid was small in all experiments as based upon the degree of complement fixation with all antigens. PASSIVE TRANSFER OF ANTIBODIES 107 3. After irritation of the spinal meninges by the preliminary injection of sterile horse serum the amount of reagin gaining access to the cerebrospinal fluid after transfusion of syphilitic serum, appeared to be somewhat greater. 4. All traces of syphilis reagin in the cerebrospinal fluid of dogs following transfusion of human syphilitic serum apparently disappeared after 22 to 48 hours as determined by completely negative Wassermann reactions. 5. The intravenous injection of dog-typhoid immune serum into a normal dog in amount of about 30 ce. per kilogram of body weight, was followed by the appearance of traces of agglutinin in the cerebrospinal fluid within three hours after transfusion; fluid removed forty-eight hours later was free of agglutinin. 6. These experiments demonstrate the possibility of the passage of antibody from the blood into the cerebrospinal fluid without primary involvement of the central nervous organs or injury to the mechanism concerned in the production of cere- brospinal fluid, when the amount of antibody in the blood has reached a point of high concentration. While it is possible that in human syphilis the presence of traces of reagin in the cere- brospinal fluid may be due to the passive transfer of this sub- stance from the blood, as shown by Wile and Stokes the presence of the reagin with or without other changes in the fluid, as an increase of protein and cells usually indicates the presence and activity of 7’. pallida in the tissues of the central nervous organs. REFERENCES (1) Wie, Upo J., anp Strokes, Joun H.: A study of the spinal fluid with reference to involvement of the nervous system in secondary syphilis. Jour. Ciuian. Dis., 1914, September, 607. (2) WiuE, Upo J., AnD Stoxkss, Joun H.: Involvement of the nervous system dur- ing the primary stage of syphilis. The Jour. A. M. A., 1915, 64, 979. (3) Wir, Upo J., anp Stokes, Joun H.: Further studies on the spinal fluid with reference to the involvement of the nervous system in early syphilis. The Jour. A. M. A., 1915, 64, 1465. (4) Hauprmann, A.: Die Diagnose der “‘irith-luetischen Meningitis’’ aus dem Liquor befund. Deut. Ztsch. f. Nervenheilk., 1914, 51, 314. (5) FLExNER, S., AND Amoss, H. L.: The relation of the meninges and choroid plexus to poliomyelitic infection. Jour. Exp. Med., 1917, 25, 525. THE ISOLATION, PURIFICATION AND CONCENTRA- TION OF IMMUNE BODIES: A STUDY OF IMMUNE HEMOLYSIN 7 M. KOSAKAT From the Forensic-Medical Department of the Imperial University of Tokyo, Japan Received for publication February 8, 1918 The numerous attempts to isolate immune bodies from their original sera and to concentrate them have, up to this time, not given satis- factory results. To obtain pure immune substance free from serum protein is, on the one hand, a most important condition for the solu- tion of problems as to the biophysical and biochemical properties of immune bodies and it has, on the other hand, an even more important relation to the therapeutic use of antibodies, as, for example, in the possibility that by such means the uncomfortable results of serum dis- ease, which are said to be engendered by serum protein, may be avoided. Many workers have inquired into this problem with various methods, which will be classified according to their purpose in two groups. One group of methods endeavors to eliminate all of the serum protein ex- cept the immune substance from the immune serum, while the other seeks to extract only the immune bodies from the respective sera. 1. The elimination of serum protein With the purpose of facilitating the practical use of immune bodies the attempt to eliminate the serum protein has been chiefly performed with respect to diphtheria or tetanus antitoxin. a. Fractionating method. By saturation with magnesium sulphate or half saturation with ammonium sulphate the antitoxin in serum is precipitated with some kinds of globulin. Thus, by treating certain immune sera with ammonium sulphate, Pick (1) found that from diph- theria antiserum produced in the horse the antitoxin is precipitated with the pseudoglobulin whereas from the antiserum of the goat it is precipitated with the euglobulin. By this means the total amount of serum protein ca” be remarkably diminished and the concentration 109 110 M. KOSAKAI of antitoxin presumably can be expected. Pick states that by the isola- tion of those fractions it is possible to concentrate the protective power ten to fifteen times. This method of Pick was proved thereafter by many workers not to be practicable. Brieger and Krause (2) state that they succeeded in eliminating 75 per cent of total nitrogen from diphtheria serum, while preserving the original antitoxic power. The antitoxic serum is diluted with an equal volume of distilled water and it is then treated with ammonium sul- phate. The resulting precipitate is placed in a 10 per cent watery solu- tion of glycerin and treated with sodium chloride. The precipitate produced by the sodium chloride contains no antitoxin, while the fluid portion has protective power. Gibson (3) prepared a refined and concentrated diphtheria antitoxin in a similar way. By means of half saturation of ammonium sulphate, and treatment of the precipitate with a saturated solution of sodium chloride containing acetic acid, he could concentrate the number of units of antitoxin per cubic centimeter from 200 to 500. But in carry- ing out this process there was a loss of about 30 per cent of antitoxin units. Gibson applied this method of concentration not only to diphtheria antitoxin, but also with his co-worker Collins (4) to the con- centration of agglutinin. ; Frouin’s (5) method is more widely known. He added 2 to 5 per cent watery solution of glycerin to diphtheria immune serum and then saturated it with sodium chloride. The filtrate was placed on a water bath at 75° to 80°C. for ten to fifteen minutes. The coagulated serum was placed in 2 to 3 volumes of a half saturated solution of sodium chloride and macerated in the ice-box for twelve to thirteen hours. The maceration was repeated at least three times, each time with a fresh saline solution. After this manipulation the total decanted fluid was dialyzed and concentrated in vacuo at about 35°C. to the original volume. All of the above mentioned or other so-called fractionating methods, however, do not give constant and satisfactory results, though they have been more frequently employed in the practical preparation of antitoxin than the other methods of concentration that are immediately to be described. b. Evaporation and freezing also have been employed for the con- centration of immune bodiesas the works of Bujwid (6), Ernst, Coolidge and Cook (7) and Hata (8) show, but the use of these methods has not been continued, because by these methods not only the immune sub- stance, but also all the colloidal substances in serum are concentrated. woe A STUDY OF IMMUNE HEMOLYSIN Lit c. Reasoning from the assumption that antitoxin and other immune bodies are of non-proteid nature, many workers have attempted to eliminate the serum protein by trypsin digestion. Thus Préscher (9) treated diphtheria serum with extract of pancreas at 32°C. and expected to peptonize all of the protein substance associated with the immune bodies. Then by half saturation with ammonium sulphate he precipi- tated the antitoxin out of the solution of pepton. The last trace of pep- ton was eliminated by dialysis. Similar attempts, which were made by Belfanti and Carbonne (10), Pick and Brieger have not given satis- factory results, because the antibodies, also, appear to be attacked by trypsin, though their destruction is very slow. All of these methods designed to eliminate serum protein from immune serum are too incomplete to be used in general, because there are diffi- cu!ties in the technique of handling the blood proteins and our knowledge of serum proteins is still in confusion. Whether immune substances are of proteid nature or not is quite unknown. Though an increase in the globulin content of the blood as the result of immunization was proved by Atkinson (11), Moll (12) and recently by Hurwitz and Meyer (13), which may be indicative of the serum-globulin nature of immune body, Joachim (14) has observed that the increase is manifested in the non-protective fraction. Glassner (15), also, states that immunization can be accomplished without any essential globulin change. These methods have, therefore, only a practical interest their use having resulted, to some degree, in a concentration of the immune bodies, but they have no biological interest, because the purification of the im- mune bodies with these methods is not possible. It seems likely, then, that it will not be*possible to solve this great problem by any process of elimination of the serum protein. 2. Extraction of immune bodies from their original sera a. By employing their antigens. The question whether antigen and antibody once combined are again separable or not has been treated by many workers. When antigen has been mixed with antibody, the two unite at once if the conditions are favorable, but this union differs from a neutralization in a chemical sense inasmuch as it is not so difficult to separate the former combination again. Thus Calmette (16) made a neutral mixture of snake venom and its antitoxin again toxic, by heating it at 68°C. v. Wassermann (17) also obtained the same result with respect to a neutral mixture of pyocyaneus toxin and its anti- 112 M. KOSAKAI toxin. Morgenroth (18) extracted the toxic substance from the toxin- antitoxin mixture of both cobra venom and diphtheria toxin by means of weak hydrochloric acid. Hahn and Trommsdorf (19) treated agglutinated bacteria with 1/100 normal sulphuric acid and succeeded in separating the active agglutinin again. Landsteiner (20), and Landsteiner and Jagic (21) proved that agglu- tinated red blood corpuscles washed with physiological salt solution gave off some of the agglutinin again, when they were brought into physiological salt solution and that the quantity of liberated agglutinin is proportional to the temperature applied to the medium. They also ascertained that the same relation could be established with respect to precipitin. Bail and Tsuda (22), and Spaet (23) proved that immune bodies which combined with cholera vibrios can easily be separated from the latter in physiological salt solution at 40° to 42°C. and also that serum of various animals can separate that combination. Tsuda (24) sueceeded also in separating the so-called normal immune substances in normal sera from their union with the respective antigen, but he was not able so well to separate the immune bodies in immune sera. The preceding group of experiments demonstrate that a union of antigen and antibody is separable without the loss of their respective functions; however, they are far from effecting the purification and concentration of immune bodies. b. By employing «norganic surfaces. From the standpoint that the antigen-antibody combination may be explained as an adsorption in a merely physical sense, many workers have studied the phenomena of adsorption of immune bodies by inorganic surfaces. Thus Biltz, Much and Siebert (25) ; Andrejew (26); Landsteiner and Reich (27) have found that, in common with other colloidal substances, immune bodies also could be adsorbed by surfaces such as caolin and animal charcoal. These surfaces act selectively and there is some distinction with respect to the adsorbing power between electropositive and electronegative surfaces. Then, in logical sequence, the effort has been made to separate the adsorbed immune substances from the inorganic surfaces. But these experiments have always ended in negative or unsatisfactory results. Thus, the well known work of Jacqué and Zunz (28) has shown that after the antitoxin or the toxin has been adsorbed by such surfaces they are by no means easily separable zn vitro, although zn vivo a separation, in A STUDY OF IMMUNE HEMOLYSIN Pie greater or less degree, is observed. According to Zunz, however, if antitoxic serum is subjected to adsorption by animal charcoal, after a great deal of the protein substance in the serum has been diminished by the above mentioned method of Frouin, only a small quantity of anti- toxin can be separated again into physiological salt solution. Nicholas Ssbolew (29) states that when the immune serum of typhus or cholera is treated with iron hydroxide, the precipitate that forms contains all of the immune bodies, but attempts to separate the immune bodies from that precipitate by means of various physical or chemical processes succeeded only zn vivo, never in vitro. The use of these inorganic surfaces for the concentration of immune body can not give satisfactory results, because such surfaces adsorb not only the immune substances, but also the serum protein. Those agents have not a selective power to adsorb only the immune substance such as the specific antigens have. Therefore, even if a separation has succeeded, the purification must be quite questionable. Studies on hemolysin To solve the problem of the isolation, purification and concentration of immune bodies, it is most convenient to use hemolysin, because its biological reaction is more easily recognized zn vitro than that of other immune bodies such as precipitin, agglutinin or bacteriolysin. And since the antigen for hemolysin is red blood cells, we can use them as ideal surfaces for our purpose, because they can take up their ambocep- tors without adsorbing any blood protein. Moreover since this antigen- amboceptor union can exist without hemolysis taking place so long as complement is absent, the manipulation must be accordingly easy. But on the other hand, as red blood cells are very susceptible to slight physical or chemical changes in their environment, there are, in this respect some technical difficulties connected with their use. 1. Studies on normal hemolysin. Normal blood serum of many animals causes hemolysis to a greater or less degree when mixed with the red corpuscles of another species of animal. We call this hemolytic - substance “normal hemolysin.’ Bail and Rotky (30) showed that the blood cells of the horse or guinea pig, that were sensitized with normal active human serum could render hemolytic the physiological salt solution, in which they were afterward digested; that is to say, the nor- mal hemolysin of human serum which was combined with the red blood cells can be separated again into physiological salt solution. Bail (31) 114 M. KOSAKAI also reported a similar experiment with the blood corpuscles of sheep that had been sensitized with imactive pig serum. Similar attempts, however, with immune hemolysin have never been successful, though the biological property of immune hemolysin appears to be quite analo- gous with that of normal hemolysin. 2. Studies on immune hemolysin. Von Liebermann and Fennyvessy (32) have succeeded in the extracting immune hemolysin against pig blood cells from its union with the latter by means of 1/100 normal hydrochloric acid. They state that the pure immune hemolysin that is isolated by this method shows no protein reaction with most sensitive protein tests and moreover it does not go through animal membrane. But this isolation of immune hemolysin, as was pointed out by Pietro Rondoni (33), cannot be said in the strict sense of the word to be a true separation of immune body and antigen union, for hydrochloric acid destroys not only red blood cells but also the immune body, especially when the acid is of high concentration. Pietro Rondoni extracted active immune hemolysin from sensitized red blood cells by means of diluted NaOH solution, which has been added to physiological salt solution. He says that in that case there is no alteration of blood cells and immune hemolysin and that the extrae- tion of combined amboceptor is accomplished in a short time either at 0°C. or at 37°C. But this method has one and the same failing as that of von Liebermann, for alkali also effects hemolysis, though in a different degree from that of acid, and Rondoni was not able to purify the iso- lated hemolysin. With the encouragement of Professor Mita I have taken up the study of this problem and I have employed, as the most con- venient material, the immune hemolytic serum of the rabbit against sheep’s blood. The first question to be settled was whether the union of amboceptor and antigen is completely sep- arable or not. The method of Liebermann and Fennyvessy has its value only with respect to the isolation and concentration of immune hemolysin. That of Rondoni has also the disadvantage that alkali cannot be used in the higher concentrations on account of its hemolytic action. Furthermore, the experiment of Bail and his co-workers proves that the same manipulation employed for normal hemolysin cannot be applied to immune hemolysin. It occurred to me that in order to separate the immune hemoly- A STUDY OF IMMUNE HEMOLYSIN 1S sin (hemolytic amboceptor) and red corpuscle combination, it must be necessary to use some other agent beside sodium chloride in the medium, because it seemed very unlikely that antigen and amboceptor combined in ClNa solution could be separable again in the same solution, even if a little alkali or acid be added or if the temperature be modified. From this point of view I examined various salts and non- electrolytes and I found, after long investigation, that some kinds of sugar have a surprising property of separating the united blood cells and hemolysin. Technique By means of the usual repeated administration of the red blood cells of the sheep, I obtained from rabbits an immune hemolysin for my experiments the hemolytic power of which was 1: 10,000; which means that 1 ce. of a 1 to 10,000 dilution of the inactivated immune serum with 0.1 ce. of fresh guinea-pig’s serum could lake all of the blood cells in 1 ce. of a 5 per cent emulsion of washed sheep’s blood cells in two hours at 37°C. I employed this serum for my study, diluting it to 100 volumes with physiological salt solution. When 4 cc. of washed sheep’s blood is mixed with 5 ce. of this diluted inactive immune serum and left at room temperature for fifteen to twenty minutes, the red blood cells will unite with all the amboceptors in it. This fact is in accordance with the results of the work of Morgenroth (34), who found that red corpuscles ean combine with far more amboceptor than the minimal quantity by which the former will be completely laked. The antigen and amboceptor union thus obtained was repeatedly washed with physiologic salt solution, till the last trace of blood protein was removed. If, under this condition, it is possible to separate the combination, the amboceptor thus isolated must be pure. The union was digested in solutions of various chemical substances and the ex- tracted fluid after centrifugalization was tested as to its hemolytic power. The results that were negative will be omitted from this report. EXPERIMENTAL RESULTS With the above mentioned technique, two test tubes of pure antigen-amboceptor union were made and then digested at 37°C. A. In 5 ee. of physiological salt solution. THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 116 M. KOSAKAI B. In 5 ce. of 10 per cent watery solution of saccharose. After two hours, during which the tubes were shaken several times, both were centrifugalized and red tinged extracts were thus obtained. The hemolytic test was performed in the usual way, 1 ce. of a 5 per cent emulsion of washed sheep’s blood cells and 0.1 ec. of fresh guinea-pig’s serum being used in each tube. The results of this experiment are shown in table 1. TABLE 1 EXTRACT A (WITH SALT EXTRACT B (WITH SALT SOLUTION Junta coNMROG GHae SOLUTION UP TO 1 cc.) uP TO 1.0 cc.) SALT SOLUTION HOURS vP To 1.0 cc.) | 0.5cc. | 0.3cc. | O.1ce. | O0.5ec. 0.3 ce. 0.1 ce. 0.03 ec. | 0.01 ec. 05 | — | — | — | Compl.| Weak | Weak- | Weak | Weak 20 tee = = — | Compl. | Compl. | Very Almost | Very strong| compl.| strong 15 |) =) «= — | Compl. | Compl. | Almost | Compl. | Almost | compl. compl. 2.0 | — _ — | Compl. | Compl. | Compl. | Compl. | Compl. It is seen that the hemolytic amboceptor that had been united with the red blood cells was easily separated into the solution of saccharose, while the physiological solution was quite indifferent. Beside the saccharose I used also glucose and lactose, and found that these sugars also have the property of separating the amboceptor-corpuscle union. I should like to note here that these sugars must be pure and especially that the glucose must be that of Merck. The following experiments were performed with saccharose. It was important to determine whether the separation of anti- gen-amboceptor union by sugar could be made complete and it was necessary, therefore, to examine the various physical factors ° that favor the separating power of the sugar. Since the volume of the sugar solution used in our first experi- ment (table 1) for extracting the hemolysin from the sensitized corpuscles is the same (5.0 cc.) as that of the original hemolytic serum used for the sensitization of the cells, we can estimate the proportion of the absorbed hemolysin recovered by the digestion A STUDY OF IMMUNE HEMOLYSIN I 7; in the sugar solution by directly comparing the hemolytic titer of the original serum and of the extraction fluid B. This com- parison (0.1 ec.: 0.01 ec.) shows that 10 per cent of the absorbed amboceptor was separated from the sensitized cells during the digestion in the sugar solution. This calculation does not take into account the inhibitory action of sugar on the hemolysis (see under “‘Discussion’’) but as the amount of sugar present in the test mixtures was relatively small, as compared with the amount of salt present, its inhibitory influence may be disregarded. FACTORS THAT INFLUENCE THE SEPARATING POWER OF SUGAR The separation of amboceptor from its antigen combination with sugar depends upon various physical factors. 1. Influence of temperature Pietro Rondoni states that the separation of the antigen- amboceptor union by alkali is the same either at0°C. or at 37°C. But as Landsteiner showed that the quantity of the separated TABLE 2 Showing the effect of temperature on the extracting power of sugar solution. EXTRACT A EXTRACT B HOURS 0.5 cc. 0.3 cc. 0.1 ce. 0.5 cc. 0.3 ce. 0.1 cc. 0.5 Compl. | Weak Weak — _ -- 1.0 Compl. | Almost Strong _ | = compl. 125 Compl. | Compl. Almost Weak Trace = compl. 2.0 Compl. | Compl. Compl. Almost Strong Weak compl. agglutinin from agglutinated bacteria is proportional to the tem- perature of the medium, it is most probable that the separation of hemolytic amboceptor from the sensitized blood cells is also proportional to the temperature applied. With the usual technique two test tubes of pure antigen amboceptor union were made and digested: 118 M. KOSAKAI A. In 5ee. of 10 per cent saccharose, at 37°C. B. In 5 ce. of 10 per cent saccharose, at 5°C. After two and one-half. hours, during which the tubes were shaken several times, both tubes were centrifugated and the hemolytic power of the extracts was tested in the usual way. The results of this experiment are shown in table 2. The experiment makes it evident that the higher temperature Quantity ssolated hemolysen + wl ro/ Sneee ec AlN Temperature nee (Is?) eg as" oe g8>" 3 6 Cc Cuart I. SHOWING THE INFLUENCE OF TEMPERATURE ON THE SEPARATION OF HEMOLYSIN FROM SENSITIZED CORPUSCLES IN SUGAR SOLUTION Concentration of saccharose, 10 per cent; quantity of solution of siccharose, 20 cc.; time, fifteen minutes. favors the separation of the hemolysin. It is known that the blood cells are laked in physiological solution when certain high temperatures are applied; we call this phenomenon heat-hemolysis. According to Gros’s study, red blood cells are destroyed at 56°C. after eighteen minutes in 0.9 per cent ClNa solution. But the heat hemolysis, as he states, depends upon the concentration of the emulsion of blood cells. A concentrated emulsion of blood A STUDY OF IMMUNE HEMOLYSIN 119 cells is dissolved more slowly than a diluted one. In my second experiment temperatures up to 65°C. were examined, a 25 per cent emulsion (in this case, for only fifteen minutes) and the macroscopic appearance of the extract after centrifugation was about the same as that obtained at lower temperature. The experiment makes it seem probable that the heat hemolysis is inhibited by sugar. The results of this experiment are shown in Chart 1. Quantity of (solated hemalysin Jime> os £ Ss 2 2s z +s < hour Cuart Il. SHowING THE INFLUENCE OF TIME ON THE DISSOCIATION OF CORPUSCLE AND AMBOCEPTOR IN’ SUGAR SOLUTION Curvea: Concentration of saccharose, 10 per cent; quantity of solution of sac- charose, 20 cc.; temperature, 56°C. Curveb: Concentration of saccharose, 10 per cent; quantity of solution of sac- charose, 20 cc.; temperature, 45°C. If temperatures above 70°C. are applied, the red corpuscles will be altered in few minutes. If the same experiment is per- formed with agglutinin or bacteriolysin it may be possible to apply the higher temperature. After all, it is quite apparent that higher temperatures greatly favor the dissociation of the 120 M. KOSAKAI corpuscle-hemolysin combination and it also appears that the separation can be made complete, if the conditions are favorable, though in my experiment the separation has been carried out only up to 84 per cent. 2. Influence of time It might be thought that at the lower temperatures a longer period of digestion would result in the same degree of separation Quantity c. csoloted hemoly sr of Sucrose Ye Cuarr III. Showing THE INFLUENCE OF THE CONCENTRATION OF THE SUGAR ON THE DISSOCIATION OF THE CORPUSCLE-AMBOCEPTOR COMBINATION Curve a: Temperature, 56°C.; time, thirty minutes; quantity of solution of saccharose, 20 cc. Curve b: Temperature, 40°C.; time, thirty minutes; quantity of solution of saccharose, 20 ce. of the hemolysin as would a shorter period at a higher temperature. But within the limit of four hours the length of time, according to my experience, does not exercise any influence upon the separa- tion of the immune body. In this fact the dissociation of the hemolysin differs from other biological reactions, such as hemol- A STUDY OF IMMUNE HEMOLYSIN 121 ysis by heat. Whether, as seems improbable, any different effect would be obtained by a still longer digestion I am unable to say. A more prolonged digestion of the cells causes considerable hemolysis which, moreover, interferes with the further manipula- tion of the extract. 5. The influence of the concentration of saccharose It is, of course impossible to use a hypotonic solution, on ac- count of its hemolytic action. The isotonic solution of saccharose is theoretically a 7.8 per cent watery solution. I examined the effect of different con- centrations from the isotonic solution up to a 25 per cent solu- tion, but I found that concentrations above 20 per cent are also unavailable, because in such concentrated solutions there appears also partial destruction of the red corpuscles at higher tempera- tures. The result of this experiment, graphically presented in chart III shows that the concentration of the sugar does not exert any considerable influence upon its separating power, but it seems as though the isotonic or slightly hypertonic solution is the optimum for the action of the sugar. 4. Influence of the quantity of sugar solution It would be most convenient for subsequent study if all the combined hemolysin could be separated in a small quantity of sugar, but when the quantity of sugar solution is little, the separated hemolysin also is relatively small in amount; it seems as though the hemolysin becomes saturated, to a certain extent, in the solution. In such case a new solution should act effectively to separate more of the combined hemolysin, and the following experiment shows that this is true. With the usual technique one test tube of pure antigen-ambo- ceptor union is prepared. This is digested for two hours at 37°C. in 5 ee. of a 10 per cent watery solution of saccharose; after which it is centrifugated and a red tinged extract is obtained. (Extract A.) To the sediment of sensitized red blood cells a second por- 122 M. KOSAKAI tion of 5 ec. of 10 per cent saccharose is added and after 1 hour incubation also at 37°C., it iscentrifugated. The second diges- tion caused considerable hemolysis. (Extract B.) The hemolytic power of the two extracts was tested in the usual way. The results of this test are shown in table 3. TABLE 3 Showing the effect of a second treatment of the corpuscle-amboceptor combination with sugar solution EXTRACT A EXTRACT B HOURS —— eee 0.5 cc. 0.3 ce. 0.1 ce. 0.5 ce. 03ce. | O1ee. 0.5 Compl. | Weak Weak Compl. | Almost Weak compl. 1.0 Compl. | Almost Strong Compl. | Compl. Almost compl. compl. 15 Compl. | Compl. Compl. Compl. | Compl. Compl. 2.0 Compl. | Compl. Compl. Compl. | Compl. Compl. It is seen to be possible to extract a considerable quantity of hemolysin by a second digestion and it may also be possible by a third or more treatments to dissociate all of the combined hemolysin. But as even the second extraction caused more or less destruction of red blood cells, it was necessary to employ such a quantity of the sugar solution that by first treatment most of the united immune hemolysin will be separated. Experi- ments were carried out to determine what that quantity is and the result of this study is shown in chart IV. The experiment shows that it is necessary to use at least 20 ce. of sugar solution in order to extract almost all of the amboceptor that was combined with 4 ce. of red blood cells. To summarize the foregoing experimental results: when 20-30 ec. of a 10 per cent watery solution of cane sugar are mixed with 4 ee. of sheep’s red blood cells that have taken up the amboceptor in 0.05 cc. of immune hemolytic rabbit’s serum and this mixture is incubated at 60°C. (1: 10,000), after fifteen to twenty minutes almost all of the immune substance, at least five-sixths of it, will be transferred from the corpuscles into the sugar solution. Thus the separation of antigen-amboceptor union is practically A STUDY OF IMMUNE HEMOLYSIN 3 complete and in that separating action not only the quality of medium, but also its quantity and the temperature play impor- tant roles. Quantity ¢, cseialed hemolysin ‘ Quantity of 4 § a. 07S a Son wee se EO Gnd Solution of Susecrese ee, CuHart LIV. SHowINa THE INFLUENCE OF VOLUME ON THE DISSOCIATING POWER OF THE SuGAR SOLUTION ON THE AMBOCEPTOR-CORPUSCLE COMBINATION Concentration of saccharose, 10 per cent; temperature, 56°C.; time, thirty minutes. PURIFICATION AND CONCENTRATION OF THE ISOLATED IMMUNE HEMOLYSIN As shown above, I was able to isolate immune hemolytic ambo- ceptor from the union with its antigen. It was, however, very important to purify and concentrate the isolated hemolysin in order to investigate its biological properties. a. Elimination of corpusclar substances including hemoglobin It is true that the sugar extract of the sensitized corpuscles does not contain, beside the immune hemolysin, any serum pro- tein, but it does contain a small quantity of destroyed red blood 124 M. KOSAKAI cells. Therefore the extract has always a slight red nuance of hemoglobin. To purify the hemolysin this trace of corpuscular substance must be excluded. For the solution of this second difficulty I examined various means. 1. Dialysis. Though the hemolysin does not pass through parchment, hemoglobin, also, is not dialyzable; therefore we can re use dialysis in order to exclude hemoglobin. 2. Adsorption. By inorganic surfaces, such as caolin, hemo- Babin is completely adsorbed but at the same time the hone is also adsorbed. 3. All other methods except the following have ended in a negative result. 4. Successful method. It is well known that the hemolysin in immune serum can not be extracted with fat solvents such as ether, petroleum ether, aceton and chloroform and it is by no means susceptible to them, while complement is easily attacked by them. By shaking the sugar extract with ether in the manner to be described I have succeeded in removing from the extract all traces of the corpuscular substance thus obtaining the hemolysin in a pure condition. If, to the sugar extract, 5 to 10 volumes of pure ether are added, and the mixture is shaken in a separatory funnel for one to two hours, there appear three layers in it. The upper layer is of ether; the middle layer consists of a coagulated mass of corpuscular substance, which is colored red and looks gelatinous; the lowest layer is the sugar solution, which is now quite colorless and con- tains the same quantity of the extracted hemolysin as before the treatment with ether. When, instead of ether only, ether and hydrochloric acid are used together and likewise shaken in a separatory funnel, the exclusion of the hemoglobin is far easier, but in that case some of the hemolysin is destroyed, especially when the acid is con- centrated. If the exclusion of the hemoglobin by shaking with ether is not accomplished by one treatment then the lowest layer is separated into another funnel and treated again with fresh ether. After repeating this manipulation twice or thrice °the last trace of hemoglobin will be excluded. A STUDY OF IMMUNE HEMOLYSIN 25 b. Elimination of sugar and the trace of salt The solution thus purified with ether contains still a certain quantity of sugar and a trace of ClNa. These substances are quite easily eliminated by dialysis; i.e., the purified solution is poured into parchment and kept in running water for about two days. c. Concentration of purified hemolysin Finally the watery solution of the hemolysin is brought into an exsiccator and concentrated in vacuum to the required volume. d. The nature of pure immune hemolysin A further study is necessary to determine this. I can only report here that the isolated hemolysin preparation does not react so sensitively as true protein with protein tests such as that with sulphosalicylic acid or with ClNa + acetic acid. DISCUSSION 1. The influence of the medium upon the action of hemolytic serum The action of compound hemolysin depends upon the reaction of its environment or the quantity of ions init. It is well known that the action of hemolysin is prevented by a higher concen- tration of salts and von Liebermann (35), Michaelis and Skwir- sky (386), and Hecker (37) proved that both acid and alkali prevent the hemolytic action of hemolysin. Lisler (38) found that in a non-electrolyte medium the complex immune hemolysin cannot produce its characteristic effect. But as regards the explanation of these phenomena, all of the authors are agreed that alkali and acid and the absence of electrolytes act in such a way as to prevent the interaction of amboceptor and comple- ment upon the corpuscles. On the basis of his observation, however, that the combined antigen and amboceptor are separable in alkaline reaction, Ron- doni explains the above mentioned inhibitory phenomena as follows: those agents which prevent the hemolytic action of 126 M. KOSAKAI hemolysin act upon the combination of antigen and amboceptor instead of the combination of the latter and complement. This explanation appears to be correct, in the light of my own observa- tion, that the sugar solution or more correctly the non-electrolyte medium can separate the union of hemolytic amboceptor and its antigen. But Rondoni’s statement that the preventive power of alkali is the same at 0°C. as at 37°C., cannot hold for other agents, because my experiments show that at the higher temperatures the antigen-amboceptor union is better separated than at lower temperatures. Indeed it seems likely, that the other inhibitory agents also will act more effectively at higher temperatures and this belief is supported by the observation of Landsteiner, that the quantity of agglutinin separated from agglutinated bacteria is proportional to the temperature applied. The general conclusion appears justified that the combination of antigen and antibody of any kind is more or less influenced not only by the quality of the medium, but also by the degree of temperature applied. 2. Reversibility of antigen and amboceptor union Though it had been proved that the union of antigen and anti- body is reversible, as I have already stated, it had not yet been determined whether that reversibility is complete or partial. The experiments of von Liebermann and Fennyvessy have no bearing on the question of reversibility and those of Pietro Rondoni tell us nothing as to the degree of the reversibility. (See, also, the work of Morgenroth (39). ) My own experiments demonstrate that the reversibility of antigen and amboceptor is almost or quite complete, so far as immune hemolysin is concerned. Those antigen antibody unions, of which the reversibility was proved only in vivo, would probably be found separable also un vitro, if the quality of medium and its temperature were suit- ably modified. a7 ots A STUDY OF IMMUNE HEMOLYSIN 127 3. Indispensability of salt in the production of immune reactions It is quite evident that salt is indispensable to the carrying out of the biological reaction between immune hemolysin and red corpuscles. If this fact is taken together with the observa- tion of Bordet (40), that the agglutination of bacteria with spe- cific agglutinin does not occur in a medium free from salt, it may be said in general that an immune body can not exert its charac- teristic effect without salt. As to the reason for this, I should like to suggest that salt mediates the combination of antigen and antibody. 4. Difference between normal and immune antibodies This question is an old one and it has been often discussed and the difference between the two sorts of antibodies has been shown by several workers; for example, by Shibayama (41) by means of dialysis of hemolytic serum, by Eisenberg and Volk (42) from the standpoint of absorption of agglutinin and by Land- steiner and Reich (43) from the standpoint of separability of hemagglutinin and blood cells. To this evidence may be added the experiments of Bail that I have already referred to dealing with the separability of the hemolysin and blood cell combination. In view of all of these demonstrated differences between the normal and the immune antibodies, it must appear doubtful that the normal hemolysins are merely increased by immunization with red blood cells. 5. Concentration of the immune hemolysin Any method of isolation and concentration of immune bodies must be, in its principle, simple and certain and also there must not be any loss of immune body during the manipulation. The method of Rondoni is not applicable to a concentration because, with that method, a complete reversibility of the antigen and amboceptor union was not shown to be attainable. Though the method of von Liebermann and Fennyvessy was heretofore the most convenient, these authors succeeded only with the hemoly- sin against pig’s blood and there was necessarily some loss of hemolysin, because they employed a very strong hydrochloric 128 © M. KOSAKAI acid, which is able to destroy hemolysin even in its weaker concentration. As for my own method, since the reversibility is complete, there is no loss of immune body and though the method has hitherto been applied only to the immune hemolysin of rabbit against sheep’s blood, there can be no doubt that it would pro- duce the same results with other hemolysins. 6. The biological properties of pure rmmune hemolysin It has been shown that hemolysin does not pass through animal membranes and that it is not susceptible to ether; but as to its chemical nature or other physical properties further experiments are required. As, however, von Liebermann and his co-worker say, it is certain that hemolysin does not react so sensitively to the protein reagents such as sulphosalicylie acid or NaCl + acetic acid as is the case with the known protein. Nevertheless I hesitate to assert now that immune hemolysin or immune sub- stance generally are of non-proteid nature. 7. The pure isolation and concentration of other immune bodies As to the isolation of agglutinin or bacteriolysin a similar method will probably be available, though such attempts have succeeded hitherto chiefly in vivo. If the antigen is not formed; e.g., toxin, the purification of the corresponding antibody can- not be accomplished by the method under consideration. If antitoxin is of non-proteid nature, then, with the fractionat- ing method when the serum globulin is precipitated, it is possible that the immune bodies are adsorbed by the surfaces of smallest particles of globulin by a mere physical process so that a separa- tion of the antitoxin might be expected, though this was at- tempted by Pick with the use of alcohol and other chemical agents with negative result. As the works of Zunz and others show, it is difficult to separate immune bodies only, after they have been adsorbed by inorganic surfaces, because if the active substance is separated again, some serum protein is separated with it. The study of the separ- ability of immune substance from inorganic surfaces has, there- A STUDY OF IMMUNE HEMOLYSIN 129 fore, a different biological interest from the one that engaged our attention in this paper. SUMMARY 1. The reversibility of the antigen and amboceptor union is proved to be practically complete, so far as the immune hemolysin of the rabbit against skeep’s blood is concerned. 2. The isolation of hemolytic amboceptor from its antigen union is accomplished by a simple method: When the hemolytic power of the original immune serum is 1: 10,000, it is diluted to 100 times its volume with physiological salt solution; 5 ce. of this diluted serum is poured into 4 cc. blood cells, which are washed free from serum protein with physiological salt solution. A/ter 15 to 20 minutes at room temperature all of the hemolytic ambo- ceptor has been adsorbed by blood cells; and the antigen-ambo- ceptor union is thus obtained. After the sensitized corpuscular sediment is washed with physiological salt solution several times, till the last trace of serum protein has been removed, this pure antigen-amboceptor combination is mixed with an isotonic or shghtly hypertonic watery solution of saccharose, glucose or lactose and left at 55°C. for fifteen to thirty minutes, during which time the vessel is shaken several times. The sugar extract, which contains nearly all of the hemolysin used to sensitize the cells, is obtained by centrifugation. 3. In order to purify this sugar extract, which contains sub- stances from destroyed blood cells, it is placed in a separatory funnel and shaken for one to two hours with 5 to 10 volumes of ether, this treatment, if necessary, being repeated twice or thrice, till at last the solution becomes quite colorless. This colorless solution is dialysed in parchment against running water in order to eliminate the sugar and traces of ClNa. 4. The solution thus obtained is concentrated in vacuo to the required volume. I wish to express my indebtness to Professor 8. Mita for the facilities and encouragement, which he has extended to me in carrying out these experiments. 130 M. KOSAKAI REFERENCES (1) Pick: Beitrige z. chem, Physiol. u. Pathol., 1901, 1, 351. (2) BrreGER AND Krause: Berliner klin. Woch., 1907, Heft 10, p. 30. (3) Gipson: Journal of Biol. Chem., 1906, 1, 161. (4) Gipson AND Co.tuins: Journal of Biol. Chem., 1907, 3, 233. (5) Frovurin: Compt. rend. d.1. Soc. Biol., 1908, 65, 444. (6) Buswip: Centralblatt f. Bakt., 1892, Bd. 12, p. 287. (7) Ernst, CooLipGe AND Cook: Journal Boston Med. Soc., 1898, 2, 166. (8) Hara: Centralblatt f. Bakt., 1909, Abt. I, 48, 203. (9) Préscuer: Miinch. med. Woch., 1902, p. 1176. (10) BeLFANTI AND CaRBONNE: Centralblatt f. Bakt., 1898, 23, 906. (11) ATKInson: Journal of Exper. Med., 1901, 5, 67. (12) Mout: Zeits. f. exp. Path. u. Therap., 1906, 3, 325. (13) Hurwitz AND Meyer: Journ. of Exper. Med., 1916, 24, 515. (14) Joacuim: Archiv f. d. ges. Physiol., 1903, 93, 558. (15) GLAssNER: Zeits. f. exp. Path., 1905, 2, No. 1. (16) Catmerte: Annales d. |’Instit. Past., 1895, 9. (17) von WASSERMANN: Zeits. f. Hyg., 1896, 22, 263. (18) Morcenrotu: Minch. med. Woch., 1903, Nr. 2, p. 61. (19) Hann AND TrommsporF: Miinch. med. Woch., 1900, Nr. 18, p. 413. (20) LANDSTEINER: Wien. klin. Woch., 1902, Nr. 40. (21) LANDSTEINER AND JaGcic: Wien. klin. Woch., 1904, 17. (22) Barn AND Tsuba: Zeits. f. Imm., 1909, 1, 546. (23) SpaEet: Zeits. f. Imm., 1910, 7, 712. (24) Tsupa: Zeits f. Imm., 1909, 2, 225. (25) Brtrz, Mucu anp Srepert: Beitr. z. exp. Therap. v. Behling, 1905, H. 10, p. 30. (26) ANDREJEW: Arb. aus d. kaiserl. Gesundh., 1910, 33. (27) LANDSTEINER AND ReEicH: Centralb. f. Bakt., 1905, 39, 83. (28) JAguE AND Zunz: Arch. intern. d. Physiol., 1909, 8, 227. (29) SsBpoLew: Zeits. f. Imm., 1912, 13, 507. (80) Barn AND Rorky: Zeits. f. Imm., 1913, 17, 566. (81) Baru: Zeits. f. Imm., 1914, 21, 202. (32) von LigEBERMANN AND FENNyvEssyY: Centralb. f. Bakt., 1908, 47, 274. (383) Ronpont: Zeits. f. Imm., 1910, 7, 515. (384) Morcenrotu: Berl. klin. Woch., 1905, No. 50. (35) LIEBERMANN: Biochem. Zeitschr., 1907, 4, 25. (36) MicHarELis AND SKwirsky: Zeits. f. Imm., 1909, 4, Heft 5, p. 357. (37) Hecker: Arb. a. d. Inst. f. exp. Ther. i. Frankfurt a. M. 1907, Heft 3. (38) Ester: Zeits. f. Imm., 1909, 2, 159. (389) Morcenrotu: Miinch. med. Woch., 1903, Nr. 2. (40) Borpret: Ann. d. |’Inst. Past., 1899, 13, 225. (41) Suipayama: Centralb. f. Bact., 1902, Nr. 30. (42) EISENBERG AND VOLK: Zeits. f. Hyg., 1902, 40, 161. (43) LANDSTEINER AND ReicuH: Centralb. f. Bakt., 1905, 39, 712. A NEW METHOD OF ESTIMATING THE ANTITRYPTIC INDEX OF BLOOD SERUM T. BRAILSFORD ROBERTSON anp SAMUEL HANSON From the Department of Biochemistry and Pharmacology, Rudolph Spreckels Phystological Laboratory, University of California Received for publication February 18, 1918 I, INTRODUCTION The methods that are at present employed for the estimation of the antitryptic indices of blood sera yield only an arbitrary measure of antitryptic power, which bears no necessary relation- ship to the actual quantity or proportion of trypsin that is in- activated by the serum. ‘Thus in the Gross-Fuld casein method (1) the “antitryptic index’’ is expressed by the concentration of trypsin that is requisite to overcome the inhibiting action of a given quantity of serum to the extent of permitting the complete hydrolysis (removal of substances precipitable by acetic acid) of a given quantity of sodium caseinate in a given time. In the Loeffler plate method (2) the ‘“antitryptic index’”’ is esti- mated in terms of the concentration of trypsin that is required to overcome the inhibiting action of a unit volume of serum to the extent of permitting some digestion of heat-coagulated blood to occur in a given interval of time. In the method advocated by Mintz the antitryptic index is similarly estimated as the per cent of a given trypsin solution that is Just able to overcome the “paralyzing” action of an arbitrary unit of blood serum (3). While each of these methods yields results of qualitative value, they are not comparable with one another, nor do they yield any quantitative measure of the relative trypsin-inactivating powers of different sera. They would do so were (a) the quantity of protein digested in a given time always directly proportional to the concentration of trypsin acting upon it, and (b) the quantity 131 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 IBY T, BRAILSFORD ROBERTSON AND SAMUEL HANSON of trypsin bound by serum directly proportional to the quantity of antitrypsin that it contains. Both of these assumptions are, however, demonstrably invalid. In the first place the relationship between time of digestion and the quantity of protein digested is not one of simple pro- portionality, and in fact in the case of sodium caseinate (employed in the Gross-Fuld method of estimating the antitryptic index) the relationship between time and degree of hydrolysis is very accurately expressed by the logarithmic equation a a—-wZ log = Kt which is that which usually obtains in monomolecular chemical reactions (4). In the second place, the quantity of trypsin bound by different samples or quantities of serum is not directly proportional to their antitrypsin content, for in that event double the amount of serum which suffices to halve the activity of a given concen- tration of trypsin should reduce its action to zero. Now the _ experimental fact is that although it is an easy matter to halve the activity of trypsin by the addition of serum it requires a relatively enormous excess of serum to abolish its proteolytic activity altogether. Thus in rabbit 1 (table 1) the addition of 0.15 unit of serum neutralized one-half the specified quantity of trypsin employed, but the addition of 0.33 unit of serum only reduced the tryptic activity to one-third. Evidently in the inter- action of trypsin and antitrypsin we have, asinso many reactions involving biological antibodies (5) a “‘balanced”’ or incomplete reaction in which the station of equilibrium is determined by the relative masses of all of the reacting components. Since the velocity-constant (K) of the hydrolysis of sodium caseinate is, as usual in catalyzed reactions, directly proportional to the concentration of trypsin (4) we have in the value of this constant a quantitative measure of the concentration of free trypsin in the digest. We have accordingly measured the value of K in various mixtures of sodium caseinate, trypsin and anti- tryptic sera with the object of ascertaining the actual quantita- Pe ANTITRYPTIC INDEX OF BLOOD SERUM 133 tive relationships which obtain between the trypsin and anti- trypsin, on the one hand, and the quantity of trypsin inactivated on the other. The following was the procedure employed. II. PROCEDURE OF THE ESTIMATION A number of glass tubes 12 to 14 em. in length, having an inside diameter of 4 to 5 mm. and walls about 1 mm. thick are prepared and sealed at one end. The open ends are sealed by means of short lengths of sealed glass tubing of the same diameter as the tubes and inserted into short lengths of rubber tubing. The following are the reagents required: SUCMUMM IV VCNORIME) 25.) 5 sie each behets ee ee 0.048 N PRGCEUMCEA CIM 8th Ga ls 5 2 2 oe eaic So hou cs batieaonelo See eee eee 0.600 N Soelavi CM OPIG OL A 2.5555). 8 dsr sicninceste AES A ae 6.0 per cent SO CMUIMECAT ON AUC. bars ci. oyeu clsemisrs sane eae SEER oe ee acs 0.01 per cent Ae HEIN ea ge once erie hcaicie, § vkeapapavades. > sae ts Ooh e Pee eee 6.0 per cent Peaysin (GTUeblers)> 4s s.r). Was es setsa co shhgne dase aan 1.0 per cent Serum diluted with 6.0 per cent sodium chloride The acetic acid is prepared with sufficient accuracy by diluting 30 cc. of glacial acetic acid to one litre. This is kept as a stock solution. The casein solution is prepared from commercial casein (Himer and Amend’s or Merck’s C. P. “Nach Hammarsten’’) specially purified by washing in water, aleohol and ether (6). One and a half grams of the casein powder is introduced into a 50 ce. flask and 25 ce. of the 0.048 N sodium hydroxide solution is added. The casein gradually dissolves with the aid of frequent shaking and gentle warming. When clear and homogeneous it is filtered to remove any particles which may not have completely dissolved. The solution thus prepared is neutral to phenolphthalein ; it must be freshly made up on the day on which it is to be used. The trypsin solution was made by dissolving Gruebler’s trypsin in 0.01 per cent sodium carbonate. In preparing the diluted serum, fresh clear and only very slightly haemolyzed serum was employed. The serum and the requisite amount of 6 per cent sodium chloride to dilute it to the desired concentration are introduced into a glass tube, a glass bead dropped in and the mixture shaken thoroughly. 134 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON These various solutions having been prepared, the estimation of the antitryptic power of a given dilution of serum jis carried out as follows: Three of the glass tubes are labelled A, B and C. A glass bead is dropped into each of these. Into A is introduced 0.125 ec. of the diluted serum and into B and C 0.125 ce. of 6 per cent sodium chloride solution. One cubic centimeter of the 6 per cent casein solution is then added to each of the tubes and 0.125 ce. of the 1 per cent trypsin solution. The tube labelled C is im- mediately acidified by the addition of 0.125 cc. of the 0.6 N acetic acid which stops all tryptic action and precipitates the casein. The contents of all the tubes are mixed thoroughly by affixing a stopper and shaking. The tubes A and B are now incubated at 36°C. for one hour. After removal from the incubator these tubes are acidified in the same way as C and again stoppered and shaken. The liquid and precipitate in each tube are separated by centrifugalization and the refractive index of each of the fluidsis determined at the same time (i.e., at the same temperature) in a Pulfrich Refractometer, the angle of total reflection being read to within one minute, a sodium flame being the source of light. It has been previously shown that the refractive index of a solution of the mixed products of the tryptic hydrolysis of casein is identical with that of the original solution of casein from which the products were derived (7). Hence the difference between the refractivities of the fluids derived from the tubes B and C is proportional to the amount of casein digested in B by the action of the trypsin during the period of incubation. One gram of casein or its hydrolysis products dissolved in 100 cc. of dilute alkali changes the refractive index of the solvent by 0.00152. Hence, dividing the difference between the refractive indices of the fluids derived from B and C by the factor 0.0152 we obtain the percentage of casein digested by the trypsin (7). The difference between the refractivities of the fluids derived from A and C similarly yields the percentage of casein digested by the trypsin which has not been bound by the antitrypsin of the serum. It is necessary to recollect, however, that the glob- ANTITRYPTIC INDEX OF BLOOD SERUM 135 ulins of the serum are precipitated along with the casein by the addition of the acetic acid. It is for this reason that 6 per cent sodium chloride is employed as the diluent of the serum, for it is found that on the average, and within the experimental error of the determination, the refractive index of serum from which the globulins have been precipitated by acetic acid is equal to that of a 6 per cent solution of sodium chloride. Hence by using this concentration of sodium chloride as diluent the refractivities of the fluids derived from the tube containing serum and from those containing 6 per cent so- dium chloride in the place of serum are equally affected by the respective additions. The difference between the refractivities of the fluids from the tubes B and C, therefore is a measure of the hydrolysis attribut- able to the total amount of trypsin added, acting for one hour at 36°C., while the difference between the refractivities of the fluids from the tubes A and C is a measure of the hydrolysis due to the proportion of the trypsin that remains unbound by the antitrypsin of the serum, acting for the same period and at the same temperature. The order followed in mixing the reagents in the tube should not be changed. If the trypsin and serum are mixed before the casein is added, irregular and untrustworthy results are obtained. We have thus measured the relative hydrolysis in the presence and in the absence of the antitrypsin of the serum. We proceed to ascertain from these measurements the actual proportion of trypsin bound by the serum by computing the velocity constant of hydrolysis in the presence and absence of serum by applying the monomolecular formula a logy == eh Oh RTA where ¢ is the time in hours, a is the initial percentage of casein and z is the percentage of casein that has been digested. The value of K thus computed is directly proportional to the con- centration of active (unbound) trypsin in the digest. 136 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON III. EXPERIMENTAL RESULTS Illustrative results are shown in the accompanying tables in which A denotes the dilution of serum employed, the undiluted serum being taken as unity, and 7 denotes the proportion of the trypsin computed to have been neutralized by the serum. TABLE 1 Rabbit 1 Concentration of casein solution 4.36 per cent = > 3 ti = Xx K X 108 A L i= 1.26 148 0.33 0.68 4.00 1.52 186 0.25 0.60 6.00 1.68 211 0.20 0.54 5.87 1.78 228 0.15 0.51 6.93 2.04 274 0.10 0.41 6.95 2.55 383 0.05 * 0.17 5.00 2.79 444 0.03 0.04 4.83 2.86 463 0.00 0.00 Considering the number of factors involved in the determina- tion, there is a very evident tendency towards constancy of the ratio The significance of this fact will be clear from fT A ay the following considerations: If we consider the neutralization of trypsin by antitrypsin to be due to the formation of a proteolytically inactive compound, the simplest conception we can form of the process is that one molecule of trypsin combines with one molecule of antitrypsin to form the inactive compound. Applying the mass-law, there- fore, we should expect the following relationship to hold good: ka A= eT yar) =e | : where a is the number of molecules of antitrypsin contained in the specified proportion of wndiluted serum, 8 is the number of molecules of trypsin-antitrypsin compound which would’ be formed by the neutralization of the total quantity of trypsin employed, and k is the equilibrium constant of the reaction. ANTITRYPTIC INDEX OF BLOOD SERUM 137 Rearranging, this equation may be written: De =C (C2 B54 Ge 1) where C is a constant which is proportional to the molecular concentration of antitrypsin in the sample of serum employed. The experimental fact being that Ara = T) is constant, it is evident that 67 is negligibly small in comparison with A, or in other words the number of molecules of trypsin-antitrypsin compound which would be formed by the neutralization of all of the trypsin present would be negligible in proportion to the total number of molecules of antitrypsin contained in the serum. TABLE 2 TABLE 3 TABLE 4 Rabbit 1. Serum taken Rabbit 2 Rabbit 3 twenty-four hours subsequently (this A c A 4 animal was recetv- ing injections of 0.50 3.13 0.50 3.55 tr ° 0.33 3.32 0.33 3.25 ypsin) 0.17 3.29 0.25 3.14 A Cc 0.13 2.10 0.06 2.59 0.40 8.84 0.33 9.50 0.25 9.00 0.20 8.50 Evidently only a small fraction of the antitrypsin in serum is actually engaged in binding the quantities of trypsin employed and a very great excess of antitrypsin must be present in the serum. We should thus be inclined to identify the antitryptic fraction of serum with some quantitatively important fraction, for example, as several authors have suggested, the serum albumins (8). In the succeeding article, however, it is shown by one of us (Hanson) that the antitryptic index of serum, as meas- ured by the constant C, may be increased by immunization no less than 300 per cent without any definite increase in the .per- centage of albumins in the serum. It is evident, therefore, that if the binding of trypsin is indeed accomplished by the albumin 138 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON fraction the process of immunization against trypsin accomplishes, not an increase in the quantity of anti-trypsin, but an alteration in its physical or chemical condition of such a character as to enhance its affinity for trypsin. On the other hand it is possible that the neutralization of trypsin is in reality accomplished by ‘some other and hitherto unidentified constituent of the serum, which must, however, be present therein in disproportionate quantity to the actual molecular concentration (which is of course very small) of trypsin employed in these experiments. SUMMARY 1. A simple and accurate method of measuring the antitryptie indices of blood-sera is described. 2. It is shown that for varying proportions of antitrypsin (= A) added to a specified amount of trypsin (regarded as unity) the relation: d fe Age Ova holds good for any given serum, 7’ being the proportion of the trypsin neutralized by the serum and C a constant which is a direct measure of the number of molecules of antitrypsin con- tained in a specified volume of the serum. 3. The molecular concentration of antitrypsin in blood serum is in great excess of the molecular concentration of proteolytically active material in a 1 per cent solution of Gruebler’s trypsin. REFERENCES (1) R. Wertz: Arch. Intern. Med., 1910, 5, 109. (2) BERGMANN AND Meyer: Berl. klin. woch, 1908, 45, 1673. (3) S. Mintz: Folia Serologica, 1910, 6, 279. (4) E. H. Watters: Journ. Biol. Chem., 1912, 11, 267. (5) S. Arruentus: “Immunochemistry.’? New York, 1907. (6) T. Brartsrorp Ropertson: Journ. of Physical Chem., 1910, 14, 528. (7) T. BrartsrorD Rospertson: Journ. Biol. Chem., 1912, 12, 23. (8) S. G. Hepin: Ergebnisse der Physiol., 1910, 9, 433. E. RosEnTHAL: Folia Serologica, 1910, 6, 285. THE NON-INFLUENCE OF INJECTIONS OF TRYPSIN UPON THE PROTEIN QUOTIENT IN BLOOD SERUM SAMUEL HANSON From the Department of Biochemistry and Pharmacology, Rudolph Spreckels Physiological Laboratory, University of California Received for publication February 18, 1918 INTRODUCTION The relation of the serum proteins to infection and immunity has engaged the attention of many investigators. The earlier experimental evidence indicated that the production of antibodies is always accompanied by an increase in the globulins. Later work, however, has shown that immunity could be attained with- out any change in the quantity of globulins. Among the earlier workers may be mentioned, Hiss and At- kinson (1), and Ledingham (2), who obtained a marked in- crease in the globulins in the serum of horses immunized against diphtheria toxin. A great part of the recent work has been done by Hurwitz and Meyer (3), and Schmidt and Schmidt (4). These investi- gators found by extensive and varied experimentation that with small doses of antigen carefully administered, a high degree of immunity can be produced without a decided rise in the globu- lins. The antigens employed by Hurwitz and Meyer (3), were certain living bacteria, dead bacteria and fowl typhoid endotoxin. The antibodies to all of these antigens are carried down with the globulins on fractionating the immune sera. An antigen, how- ever, whose antibody is known to be in the albumin fraction has not been employed by Hurwitz and Meyer, Schmidt and Schmidt, nor, to the writer’s knowledge, by any previous in- vestigators. According to Kimmer and Mogulesko (5), pan- 139 140 SAMUEL HANSON creas trypsin and yeast trypsin are antigens of this sort. It was of interest to find whether or not a carefully produced immunity to this rather unique type of antigens is associated with a change in the protein quotient. This is the problem that incited the undertaking of the present work. METHODS AND MATERIALS The experimental conditions were in general similar to those described in a previous communication (6). Seven normal rabbits were selected as the experimental animals. Determina- tions of the serum proteins and of the antitryptic index were made through a prolonged fore-period. At the end of the fore- period the immunization of six of the animals was begun, while one served as a control. Solutions of Griibler’s pancreas trypsin puriss. sicc., freshly made up and filtered was the particular trypsin employed. The strength of the sample of this trypsin used was such, that the velocity constant of digestion of an 0.38 per cent solution freshly made up and filtered, acting in an equal volume of a 4.40 per cent sodium caseinate solution at a temperature of 37°C., was found to be 36 X 10-2 by Robertson’s refractometric method (7). The same sample of trypsin unfiltered showed only a slightly higher potency. This enzyme was administered subcutaneously and intra- venously in order that the slower and more rapid effects, re- spectively, may be elicited. The quantities and frequency of injection were also varied in the different rabbits so that the optimum rapidity of immunization may be at least more or less closely approached. In no case were the doses used so large as to cause severe reactions or even a loss of weight in the animals. The determinations of the serum proteins were made by Robertson’s micro-refractometric method (8). The estimations of the antitryptic index were carried out by the method described in the preceding article (9). The prin- ciples on which this method is based are briefly as follows: Since it is a fact that the refractive index of a sodium caseinate NON-INFLUENCE OF INJECTIONS OF TRYPSIN 141 solution is practically not altered by tryptic digestion, it is possible to estimate the extent of hydrolysis in such a solution containing trypsin by precipitating the undigested casein with a definite volume and strength of acetic acid, centrifuging down the precipitate and determining the refractive index of the super- natant fluid, which is a solution of the products of hydrolysis (6). The velocity constant (&) of the trypsin in question is then calculated from the monomolecular formula, a a—-xz = Kt log where a is the initial concentration of the substrate and x the amount digested during a period of time f. The velocity constant (k:) of the rate of hydrolysis is then obtained in a similar manner for a casein-trypsin solution to which a known quantity of serum has been added. Since K is the velocity constant of the rate of normal hydrolysis and K, is the velocity constant of the rapidity of digestion after a part of the trypsin was neutralized by the antitrypsin of the serum, * a = fraction of trypsin unneutralized x It has been observed (9), that the trypsin-antitrypsin reaction is reversible; hence it is practicable to calculate the antitryptic index (C) from the equilibrium equation, x — ies where x = the fraction of trypsin neutralized and s the amount of serum employed. DISCUSSION The experimental results presented indicate that the normal variation in the antitryptic index is on the average approximately 50 per cent in any particular rabbit and is considerably greater in different animals. Rabbits 1 and 3, and to a lesser extent 142 SAMUEL HANSON TABLE 1 Rabbit 1. Weight; July 17, 2862 grams; July 28, 2798 grams z | Qo ids eco al : Oe a pak ta TREATMENT a g 3 ae 3 a ae Pa) Z < ic} e ic} my ie) per per per per per cent | cent | cent | cent | cent Normal: fae tis sos, 5d 2 eter eee (=) 3) 122), 4.03) 125, I> bale 2 falOneng IN oye) |. SEe a ee Mee 2 cit f 511.3 | 4.2 | 1.8 |.6.0) 380 | 0:43))2%78 No rina all eresesy sisson SNe ote ere 7] 7 1-0 | 4.1 | 1.8] 5.9 | 30') 0-44)%3200 INOTMIAL Ass 2st bie s site se bce a oe C—O 27325 | 127 | 5.27) 33) OF48 Normal Seer oe on hse Cee 715) 125) 4:0: 1.2), 5:2 |) 23°) Os30aeas 0.2 gram trypsin subcutaneously onstheml 6th’ S.24 sce. 8 seen CAT 123713287 | WAN 2 27) Okan led 0.3 gram trypsin subcutaneously ongtherl Oth. Fee 4 eee 718) 1.4°)-4.1/] 1.5,|.5.6 | 23 .| 0-86) (3262 ' 0.4 gram trypsin subcutaneously Olgtteeae Gh dae- ee se ae eee 7-20) 1.67] 3:95] 1-35) 5.3] (26 0734) 78201 0.6 gram trypsin subcutaneously [| 7-23) 1.4] 3.5 | 2.8) 6.3 ke 0S Aas OR SORE Bt {| 7-26] 1.3 | 3.2| 1.7] 4.9 | 35 | 0.50) 2.70 ‘i, Sb. geet. {| 8- 2| 1.3) || 4:0) || 2.2 | 6.29) 5: Ors5ip2eso TABLE 2 7 Rabbit 2. Weight: July 17, 2894 grams; July 28, 2724 grams Zz 3 so | ¢6 5 4 a p SL ce) BE? aloe aieS TREATMENT = 3 5 ae A 2 alae Z z Zz & 6 | s#16e8! oe S z 2 5 |e *18 = om per per per per per | cent | cent cent cent cent IN Oriiiaal 7.2 24.) 2 eee eee 7.511.3| 4.2] 1.8] 6.0] 30.0] 0.43) 2.78 Normals 36724 5 hes ten Cre eee VT 1-34) 420 | 125 | 5.5 | 270) 0s37|e2e22 Normal .:-7 22. 4has on ae ee ae 7-9 | 1.4] 4.1 | 1.8] 5.9 | 30.5) 0.44 Normal. 22 52'.6 sce Soe ee 7-15] 1.4 | 4.5 | 1.4 | 5.9 | 24.0] 0.31] 2.55 0.2 gram trypsin on the 16th )| 7-17| 1.4 | 4.7] 1.2] 5.9 | 20.0) 0.26] 1.90 subcutaneously 7-18] 1.6 | 4.1-| 1.3 | 5.4 | 24.0] 0.31) 2.14 0.3 gram trypsin on the 19th (subeutaneously)................ 7-20) 1.4 | 3.4] 1.5 | 4.9 | 31.0) 0.44) 3.99 0.5 gram trypsin on the 22d (subcutaneously,)\.c--- +20 oc: 7-23] 1.5 | 4.0 | 1.4 | 5.4 | 26.0) 0.35] 5.00 0.6 gram trypsin on the 25th \ 7-26) 1.3 | 3.6 | 1.7 | 5.3 | 32.0) 0.47] 3.06 (subcutaneously) | $2 | 1-2 478+] 1.2 |-6.0.| 20.0|.0-251) 185 NON-INFLUENCE OF INJECTIONS OF TRYPSIN 1438 TABLE 3 Rabbit 3. Weight: July 17, 2610 grams; July 28, 2383 grams Z } 8d | 6 : a gee di TREATMENT S a g fe 3 Ze 4 z, 3 elke ol chanel OE eels per per per per per cent cent cent cent cent NOTIN ANIMAS Sls ie MA toktsh a cteeteno- 73) Vel 2n ASL el eOn i GuOn|Ro2 00846 IN| OTST IA LA a A iO (Adal Ae Oe liso Gaoole2o.0)) Oxssiecents MSCOTTVA RP Vet sh, de Nas loi ds sestle te de C0 Whe 3s) 404 IS G2 22950) 041193200 INIGIETED GLA, Pe ee Oe CO PALA ASS ale 18h | Ooi |aeo ele O.42 DOIG se ae a 7-15, 1 4.| 4.5 | 1.25) 5.75). 22.0) 0.27) 3.25 0.1 gram trypsin on the 16th (sub- EMPANeOUShY.). 0. 300.8. EL oe. 7-17 1.2 | 4.3 | 1.3.) 5:6 | 23.0) 0.32) 2.35 Same dose on the 17th and 18th. .| 7-18) 1.4 | 4.1 | 1.7 | 5.8 | 29.0) 0.41) 3.12 | 7-20] 1.4 | 4.3 | 2.0] 6.3 | 32.0) 0.46] 9.00 Daily on the 19th to 25th inclu- || 7-23) 1.2 | 3.9 |.1.9 | 5.8 | 33.0) 0.49] 8.61 sive 0.2 gram subcutaneously 7-26) 1.1.) 4.0%). 2.2) 1 6.24), 35.0) 0..55)'4580 || 8-2 | 1.2 | 4.8 | 1.6 | 6.4 | 25.0].0.33) 2.81 TABLE 4 Rabbit 4. Weight: July 17, 2270 grams; July 28, 2383 grams z oie aie gq ° & q ro) Sue eran Vet a deta OR TREATMENT r FI : 5 3% Bz Bz 2-2 po lee [BE PR ee per per per per per cent | cent | cent | cent | cent Normans: See 4 hee. Ps bee ed 7-3"|\ 1.3 | 4.25) 1°25) 5.5) | 24.01.0229 IN/@RTITGI |S Chol cle RO ORES eo ae Eee eae 7-5 | 1.8 | 4.4 | 1.38 | 5.7_| 23.0) 0.23) 2.50 NO UnI Dees eB te Se Bich ceed 7-7 | 1.3 | 4.0 | 2.1 | 6.1 |, 34.0) 0.52) 2.22 OTM SO eo Ras, hee es 7-9 | 1.05) 3.9 | 1.4 | 5.38.| 26.5) 0.36 IN[@ TTC Aa eee eee ere 7-15) 1.4 | 3.9 | 1.0 | 4.9 | 20.0) 0.26) 3.25 0.1 gram trypsin on the 16th (in- UAVEMOUSIN) IS. Get ocitede ace sh. *.| 7-17) 1.4 |, 3.7 | 1.8 | 5.0 | 26.0) 0.85) 3.52 Same dose also on the 17th and 18th) 7-18) 1.3 | 3.8 | 1.2 | 5.0:) 24.0) 0.32) 2:93 Daily 19th to 23d inclusive 0.2 | SRAM ee ae es. Scie: Bo tet ieee cies ol ae 7-20) 1.5 | 4.3 | 1.1 | 5.4 | 20.0} 0.26) 4.25 ~ (| -¥-23)-4.3 | 4.0 | 1:7 5.7 |-30:0) 0.42) 4°48 eipavh and 25th 0.4 gram (1n~ )| 7 96|.1.2.|/4.3°) 1.87 | 610 | 30.0) 0.42) 3606 Beayenously) || $2 | 1.2] 4.6| 1.3 | 5.9 | 22.0] 0.28] 1.88 144 SAMUEL HANSON TABLE 5 Rabbit 5. Weight: July 17, 2100 grams; July 28, 2270 grams Zz 5 Zz [e} Zz = TREATMENT Fi 5 8 E Zz Bil 33 < ° re) n a Z < oS per per per cent | cent | cent INformr al ay seb fas: e255. icrottsys Soferstel te (=o (ALS | 3.94) 1.6 Norra eee, rs Po eee teeta 7-5 | 1.3 | 4.4) 1.7 BN OTT MUSES get ease 2 acc volt aie ake ie wi Rare (ei Nl 3 4)0430) 6 INGE ANE Ee Ws cleo «tists ais ths coger 0-9) 10) 455 | 123 Normale 2h. is onc ae weave eros 7-15) 1.3 | 4.4 | 1.2 INormraley g4.ti.5 00 toi ens sree es i ee: a UR 74 INorinale etry: acc cats seeiasaeeee 7-18) 1.5 | 3.8 | 1.3 INormMaltes -frs28\. «Rite. Sees b—201 155 450110 Normale Se 327. 0, ...tkelo sesiose eerie 7-23) 1.3 | 4.0 | 1.7 INOrmal ee. a.%50) oe tnise ka eee 7-26] 1.2 | 4.3 | 2.5 INOKNIAL. £6le sais «artnet se bea oe Safe) 2) | 465d TABLE 6 Rabbit 6. Weight: July 17, 2383 grams; July 28, s BP ele TREATMENT = = 3 & = p Bee ah car | < 2} | I i=} Zz < S per per per cent | cent | cent Normal. Ja% tn cere nee cis Gao tele Agel lees Normal. (Sti: 7G wer ee mee oe aoe es 7-5 SN TAO) INOrMal%, scenes cS Nae eee ee tN le2 | Asses Da INommal tian aay fot aren pmo: 7-9 | 1.4 | 4.3 | 2.4 Normal 7>. etek fc eee seek ake eee 7-15} 1.3 | 4.8 | 2.0 0.1 gram trypsin on the 16th (in- ETAVYENOUSLY) J .4.02 ee 12 een oe (Wh 12 VARS 15 Also same dose daily to the 23d || 8-18) 1.3'| 3.8 | 1.7 inclusive AZO) 5 ofc eich Lo 7-23] 1.3 | 4.4 | 2.2 On the 24th and 25th 0.2 gram... 7-26] 1.3 | 4.6 | 2.2 8-2) (°1.3, | 4.7 | 1.6 2 [53/8 2x |22z| BS pF EFA BF | o per per cent cent 5.5 | 29 | 0.41 6.1 28 | 0.39 56 29 | 0.40} 3.99 5.8 22 | 0.29 5.6 21 | 0.27) 4:88 tise 23") (0) 29\fenoe peal 25 | 0.34!) 2.93 5.0 25 | 0.25] 4:69 edl 30 | 0.42) 5.00 6.8 | 37 | 0.58] 2.58 nerf 20 | 0.26] 1.85 2440 grams a4 Ze ea chen ay) per per cent cent 6.5 28 | 0.38 (fell 28 | 0.39] 3.99 6.5 34 | 0.49) 3.32 6e7 36 | 0.56 6.8 29 | 0.42) 3.82 5.8 26° 0535| 158 5505) 31 | 0.45) 2.26 i; Tee See 5.86 6.6 33 | 0.50) 4.43 6.8 | 32 | 0.48) 2.70 6.3 25 | 0.34) 0.95 rs NON-INFLUENCE OF INJECTIONS OF TRYPSIN 145 TABLE 7 Rabbit 7. Weight: July 17, 2610 grams; July 28, 2472.grams z eae les g io] & p eB Zz i] oO o TREATMENT 2 iS a ces] Ze - mt EB BP | 3az|Psz| 84 é z ea 6 | #8 |oos8| Sz BS z 2 oh SS 3 6) per per per per per cent cent cent cent cent Sooner 7-8: | 1 J) Aeoelenes, eseGnl SO. | Oat INC Ch TATU fers ciaci na, 5. stone Sse She sav om 7=5-| LeQuiv4acoailesaleono 28 | 0.40) 2.35 INL OMIA ly Sky See ee ea eee (HE Waterers OSU eects F 2222 load eee 7-9.| 4.0 13:9) | i730: |o144 “"oouid) «2a 7-15| 1.3 | 4.2] 1.6] 5.8] 28] 0.40] 1.80 ily 0.1 rypsi x - see a aa 7-17| 1.4| 3.7| 2.1|5.8| 36 | 0.57] 2.35 a . 7-18| 1.6 | 4.0] 2.0| 6.0] 33] 0.50] 1.50 clusive | | 7=20| 1:.4:| 3.7 | Le90le5:6 1) 84.) 0.51) 2255 7-23| 1.8 | 4.0 + 2.0 6.0] 33 | 0.50) 2.98 On the 24th and 25th 0.2 gram 7-261 1.3|3.9| 2.41 6.3| 38] 0.61] 2.05 [| $2] 1.2] 4.3] 2.1] 6.4] 33 | 0.49) 0.95 rabbits 2, 4 and 6, developed a certain degree of immunity to trypsin as shown by the rise in the antitryptic index. Rabbit 5 (the control) showed also a rise in the antitryptic index. Al- though this increse is considerably less than that shown by rabbits 1 and 3, it is yet beyond the limits of normal variation. Rabbit 7 failed to produce any immunity. Very interesting is the phenomenon that soon after the first one or two injections there is a marked fall in the antitryptic index. This decrease in the antitrypsin was possibly due to a neutralization of the antitrypsin by the trypsin injected, without an accompanying equal regeneration of the former for a time. The injections of trypsin in the doses used is not followed by any decided change in the protein quotient. This fact is in harmony with the negative results ofaimmunization against other antigens, cited at the beginning of this article. SUMMARY 1. The normal variation in the antitryptic index is very marked in any particular rabbit, and is even greater in different animals. 146 SAMUEL HANSON 2. An unmistakable rise in the antitryptic index has been pro- duced in at least two of the six rabbits immunized against trypsin. 3. The progress of immune body production against trypsin is very peculiar, in that at first there is a marked fall in the antitrypsin, followed soon, however, by rather a sudden rise which is at best only 300 per cent above the normal variation. At this level the antibody content remains approximately stationary for a comparatively brief period and is then supervened by a rapid return to the normal, in spite of continued periodical in- jections of trypsin. 4, Immunity to pancreas trypsin appears to cause no change in the protein quotient. This fact may serve as additional evidence in favor of the recently emphasized view that immunity 1s non-depend- ent on the concentration of the serum proteins. REFERENCES (1). Hiss, P. H., anp ATKrinson, J. P.: Jour. Exper. Med. 1900-1901, 5, 47. (2) Lepinenay, J. C. G.: Jour. Hyg., 1907, 7, 65. (3) Hurwitz, S. H., anp Meyer, K. F.: Jour. Exper. Med., 1916, 24, 515. (4) Scumipt, E. S., anp Scumipt, Cart L. A.: Jour. of Immunology, 1917, 2, 343. (5) Kamer, H., anp Mocutesxo, J.: Z. f. Immunititsf., 12, 16-28. (6) Hanson, S., anp McQuarriz, I.: Jour. Pharm. and Exper. Therap., 1917, 10, . 261. (7) Rospertson, T. Braitsrorp: Jour. Biol. Chem., 1912, 12, 23. (8) Rospertson, T. Brattsrorp: Jour. Biol. Chem., 1915, 22, 233. (9) Ropertson, T. BrarsrorD, AND HANson, SAMUEL: Journal of Immunology, 1918. 3. 131. ; EFFECTS OF INTRAVENOUS INJECTIONS OF A COLLOID (GELATIN), UPON RABBIT SERA GUY W. CLARK From the Department of Biochemistry and Pharmacology, Rudolph Spreckels Physiological Laboratory, University of California Received for publication February 28, 1918 The general purpose of the following experiments was to study the effects of intravenously injected gelatin upon the non- proteins and proteins in normal rabbit sera. The special phases considered were: 1. The effects of the injected gelatin upon the ratio of the serum albumins and globulins and upon the total amounts of proteins and non-proteins. 2. The time required for the gelatin to disappear from the blood stream. 3. The effects due to the gelatin attracting fluids from the body tissues, resulting in an excessive dilution of the blood (hydremic plethora). Four medium sized female rabbits (2.0—2.5 kilo), were used as experimental animals. For several days before and throughout the experimental period these animals were under uniform con- ditions as to feed, water, etc. The micro-refractometric method of Robertson (1) was used for the determination of the albumins, globulins and non-pro- teins. This method makes it possible to obtain accurate results with small amounts of serum, which is necessary when using the smaller experimental animals. The blood was removed from the vein of one ear and the injections made into a vein of the opposite ear. The analysis of the serum was made on alter- nate days, injections being given every day during the experi- mental period. The analytical work was carried out under 147 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2 148 GUY W. CLARK uniform conditions, about six hours being the average time for completing an entire set of determinations. Gelatin was selected as a colloid suitable for injection be- cause it is non-toxic and is not precipitated by N/50 acetic acid and subsequent boiling. It therefore appears quantitatively among the non-protein constituents. The chief disadvantage in using gelatin arises from the fact that the term ‘“‘gelatin” has only a relative meaning. Nelson’s ‘‘Gold Label” gelatin was selected for this work but before making up the solutions it was placed in a desiccator over CaCl, and allowed to dry several days. The gelatin then contained 7.5 per cent water and 1.5 per cent ash. From this product two solutions were made; a 10 per cent and a 20 per cent (10 grams of gelatin to 90 grams of water and 20 grams of gelatin to 80 grams of water, respectively). Not all of the water was added as such, since the solutions were acid and it was necessary to add several cubic centimeters of N/10 alkali to make the solutions neutral to phenolphthalein. After neutralizing and diluting to weight the solutions were sterilized in an Arnold steam sterilizer on three successive days to insure killing any spore-forming organisms. For injection the solutions were warmed on a water bath to 37 to 40°C. The accompany tables on pages 149 to 152, show that the serum proteins of an individual rabbit vary from day to day through quite a wide range and that this variability applies to a greater extent between different rabbits. The variability of the serum proteins has been observed by Schmidt (2) and others employ- ing the same methods of experimentation. From the same tables it is also evident that the injected gelatin showed no marked effect in altering either the ratio or the total amount of the serum albumins and globulins, nor the amount of non- proteins. Moll (8), in the course of investigations on the effects of im- munization upon the globulin content of their blood sera, in- jected gelatin into rabbits and obtained a very marked increase of the globulins in every case, in one instance his figures indicate an increase of 81 per cent. His method (4) for estimating the globulins depended upon a separation by salting out, followed EFFECTS OF INTRAVENOUS INJECTIONS OF A COLLOID 149 by successive washings with water, alcohol, ether, then drying at 110°C. and weighing. This method could hardly be expected to give quantitative results since the salt used to coagulate the proteins would be carried down in considerable amounts with TABLE 1 Ratio of globulins to total proteins ANIMAL NO. FORE-PERIOD PERIOD AFTER-PERIOD per cent per cent per cent ( 18.4 Ace 22.0 | 19.8 29.7 22.1 I J 18.1 14.6 BES es ye eres ots 143 4.1 18.2 22.2 21.4 IAVCLAR Ctra... «amos: 17.8 21.5 22.0 « 18.0 19.9 20.3 20.9 22.8 24.7 28.0 19.7 Lin S360 are 19.5 17.4 17.8 [ 19.3 PANVERAGCW. ss pg 7 a] Caan @ iad » “<) - ion a F 7 7 " : . oon 2 -, * Pacis =a t Py _ 4 [bad coe” © S - a, oe = - if _— THE INFLUENCE OF ACTIVE NORMAL SERUM (COMPLEMENT) UPON MENINGOCOCCE Il. THE BACTERICIDAL AND PROTECTIVE VALUE OF FRESH NORMAL SERUM ALONE AND IN COMBINATION WITH ANTIMENINGITIS SERUM FOR MENINGOCOCCI TOITSU MATSUNAMI anno JOHN A. KOLMER From the McManes Laboratory of Experimental Pathology of the Unwersity of Pennsylvania Received for publication March 29, 1918 In addition to specific opsonins and toxin neutralizing anti- bodies, potent antimeningococcus sera are generally regarded as possessing a distinct meningococcidal activity; the latter is ascribed to the presence of specific bacteriolysin requiring the presence of complement for lytic activity although this does not explain the total bactericidal activity inasmuch as the early ex- periments of Flexner (1) demonstrated that heated serum pos- sesses bactericidal activity and advanced the hypothesis that it is only necessary that the fresh or heated serum should injure the cocci in order that their intracellular enzymes should be ren- dered active and destructive for the microérganisms. The defibrinated bloods and sera of normal persons have been found by Davis (2), Flexner, (1), McKenzie and Martin (3) and others to possess meningococcidal properties; this activity was found most marked with fresh active sera and diminished but not totally removed by heating at 60°C. for thirty minutes. In- asmuch as the antimeningococcus sera ordinarily administered are free of complement and the cerebrospinal fluid in meningococ- cus meningitis apt to be free or contain but traces of this con- stituent of serum, we have sought to determine the bactericidal activity of the fresh serum of persons and guinea-pigs alone and in conjunction with anti-meningococcus serum for virulent men- 1 Aided by a grant from the Pediatric Society of Philadelphia. 177 178 TOITSU MATSUNAMI AND JOHN A. KOLMER ingococci, in an experimental study of the probable value of complemental bacteriolysis in resistance to meningococci. In our first communication concerning the influence of active normal serum (complement) upon meningococci (4) we have shown that the opsonic activity of fresh active antimeningitis serum is greater than the opsonic activity of the same serum after heating or the addition of tricresol, followed by exposure to room temperature for several days; also that the addition of active normal serum to commercial antimeningitis serum in- creases opsonic activity. It would appear, therefore, that the opsonic activity of antimeningitis serum is maximal when the serum is fresh and active, that is, contains the labile opsonin or complement, and that this constituent may be restored and the commercial immune serum reactivated to a certain degree, by the addition of fresh human or guinea-pig serum. In this investigation we have followed three main lines of study: 1. To determine the bactericidal activity of normal human and guinea-pig sera and various anti-meningococcus sera alone, and in combination, for virulent meningococci, by in vitro, plating and pipet methods. 2. A study of the agglutination of meningococci by normal and immune sera in relation to the bactericidal activity of these sera. 3. A study of the mouse test of Hitchens and Robinson (5) for determining the protective value of antimeningitis serum and the influence of normal sera alone and in conjunction with anti- meningitis serum upon virulent meningococci, as determined with this technic. The results of a large number of experiments are briefly sum- marized in this communication; with plating methods the bac- tericidal activity of various antimeningococcus sera alone as well as the sera of normal persons and guinea-pigs alone and in com- bination with antisera, has been found surprisingly low although these sera and particularly the whole blood of certain persons and guinea-pigs, have been found definitely bactericidal for menin- gococci with a delicate technic. The protection test of Hitchens and Robinson has, we believe, a definite and practical value al- INFLUENCE OF SERUM UPON MENINGOCOCCI 179 though not sufficiently delicate for the purposes of our experi- ments, namely, to determine the influence of complemental bacteriolysis alone as a factor in protection against virulent meningococci. Inasmuch as any animal protective test calls into consideration the influence of immunity principles in the body fluids of the experimental animal and thereby masking the possible influence of the addition of fresh normal serum (comple- ment) upon the bactericidal activity of the immune serum, we have given most attention to test tube experiments. THE BACTERICIDAL ACTIVITY OF FRESH NORMAL SERUM ALONE AND IN COMBINATION WITH ANTIMENINGITIS SERUM a. The bactericidal aciivity of anti-meningococcus and normal sera alone and in combination, as determined by plating methods. All experiments were conducted with strain 124 of normal men- ingococei? kindly furnished by Dr. George Robinson. Anti- meningococcus sera were obtained from various laboratories; normal human sera were obtained from ourselves and others of the laboratory staff by vein puncture, and from normal adult guinea-pigs by heart puncture. Bactericidal tests in vitro, and particularly with plating meth- ods, are generally unsatisfactory; strict attention to the reaction of the culture medium and numerous preliminary tests to deter- mine the proper dose of culture to employ in order to avoid too few or too many colonies in a plate with frequent repetitions of the main experiments, were necessary in this study. The pro- tocols of several experiments are given in tables 1, 2, 3 and 4 and the results may be summarized as follows: 1. As shown in table 1, active or fresh antimeningitis sera are more bactericidal than the same sera after inactivation by heat- ing at 60°C. for thirty minutes. The addition of 0.3 per cent tricresol to the sera increased bactericidal activity in dilutions up to about 1:20, but this preservative did not appear to appreci- ably injure or reduce the bactericidal activity of the serum itself as tested within one week after the addition of tricresol, although it reduces to some degree the opsonic activity of the serum. * This strain was isolated on June 2, 1917, on the hospital ship Solace. 180 TOITSU MATSUNAMI AND JOHN A. KOLMER 2. Complement free antimeningitis sera possess bactericidal activity as previously shown by Flexner and it is highly probable that complemental bacteriolysis exerts but a minor réle in the total bactericidal activity of immune serum. TABLE 1 The influence of heat and tricresol upon the bactericidal activity of antimeningococcus serum FINAL DILUTIONS OF ANTISERUM ANTIMENINGITIS SERUM (No. 2279) | | 1:80 1:160 1:320 | 1:640 | 1:1280 | 1:2560 180, 30) Ster-! Ster-| Ster-| Few ile ile ile 1:2 1:10 1:20 | 1:40 Fresh active....... 1500*; 900 |} 400 | 200 Five days after addition of 0.25 per cent tricresol.| Ster-| Ster-| Ster-| 20} 40) 240) 20] Ster-| Ster-| Few ile ile ile ile ile Fresh serum after heating at 60°C. for one-half hour eth 1600 | 1800 | 300) 180; 20) 30 10 | Ster-| Few ile Heated serum + active human se- EL 4 meee ee 1800 | 1600 | 1200 LUT LY 1800 | 1400 | 1200 | 1800 * Colonies per plate. In these experiments the culture (no. 124) was used in amounts of 0.5 cc. of a 1:500,000 emulsion of a twenty-four hour serum agar culture; 0.5 cc. of this emul- sion mixed wih 0.5 ce. sheep serum broth and plated in amounts of 0.5 ce. after two hours incubation, showed 600 colonies per plate. Dilutions of serum in amounts of 0.5 ec. of culture were incubated at 37°C. (water bath) for two hours and plates prepared with 0.5 ec. from each tube in 10 cc. of sheep serum dextrose agar. 0.5 cc. of human serum (1:5) plus 0.5 ec. emulsion of cocci and plated in amounts of 0.5 cc. after two hours incubation, showed 900 colonies per plate. 3. The bactericidal activity of practically all the antisera tested was low according to the results observed with the tech- nic employed although some bactericidal activity was generally apparent as compared with the controls; not infrequently the higher dilutions of serum were more bactericidal than the lower, these results being similar to those published by Jochmann (6). 4, As shown in table 4 active human and guinea-pig sera gen- INFLUENCE OF SERUM UPON MENINGOCOCCI 181 erally possess some bactericidal activity for meningococci; the bactericidal activity of these sera vary in different persons and animals and, as a general rule, human sera are more bactericidal than pig sera. 5. In practically all experiments however, the addition of equal parts of 1:5 or 1: 10 dilutions of fresh active human or guinea- pig sera to varying dilutions of different anti-meningococcus sera, was without appreciable increase of bactericidal activity; indeed TABLE 2 Bactericidal action of anti-meningococcus serum alone and in combination with equal parts of a 1:5 dilution of active guinea-pig serum PIG FINAL DILUTIONS OF ANTISERUM SERUM] CUL- SUBSTANCES AND | TURE CUL- |ALONE BP) 1;4 1:8 1:12) 1:16 | 1:24] 1:32 |] 1:48] 1:64 | 1:96 | ruRE Antiserum (com- mercial) A alone....| Ster-| Ster-| 540*/2160) Unc.| Unc.| Une.| Une.|Une. |Une. |Une. | Une. ile ile Antiserum Sr yeahs é serum....| Unc. | Unc. |6480 |9720} 4800} 8000) Unc.) Unc.| Une. |Unc. * Colonies per plate; Une. = too many colonies for counting. In this experiment a series of dilutions of the immune serum was prepared in sterile test tubes with sheep serum broth, in amounts of 1 ecc.; each tube was seeded with 0.1 cc. of a twenty-four hour serum dextrose broth culture of menin- gococci and incubated in a water bath at 37°C. for two hours, shaken and 0.5 ec. from each tube plated in 10 ce. of sheep serum dextrose agar. In the second series a 1:5 dilution of fresh sterile pig serum was used instead of sheep serum broth in making the dilutions, thereby rendering the final dilutions of immune serum the same in bothseries. Controls on the sterility of the immune and nor- mal sera, culture, etce., were included. The plates were counted after forty-eight hours incubation. ; the colonies of meningococci were generally more numerous than with the antiserum alone, suggesting that the addition of normal serum enriched the culture medium and thereby favored the more rapid multiplication of the surviving cocci. TOITSU MATSUNAMI AND JOHN A. KOLMER 182 OO00L | OOSF 0O00L | 00cE O0F9 | O0OL OO8O0T} 0009 0006 | 09& ANOTV | ANOTV log: any) -100 WOULS TVNYHON 0022 0061 0002 00GT 0008 OFS ‘oul) ou) OVE “UURY UD) Ser :T 49:1 GET OOTZ 0001 OOF OOOT OcOL 00€ ‘oun OO8OT 0082 o]L104g i) O0SG 006 006 006 OGL6 00€ ‘oun OFS89 O9GG O[L10}g ou) ‘ouy) 8:1 WOUASILINVY JO SNOILATIAG TVNIA 91949 By ieais tandos UBUINTY + Spm etn a: oUuoTe ([BIotoWUIOD) umes uetmny + SAONVISHOS wuInAOsTyUV WN.LOSsT}UW wn.LOsTyUy wnJostyuy wndostyUy umMJestyUy wINn.LOsT}UV wndestyuy UINnAosIyUy wndosTyUy WIN.IOST}UV S229 aS SSS INDW “TNH d Xa € QATaVL unas Bid-paumb pun UDwUny ]DUWLOU YIN UOUDULQUOI UL PUD BUOID DLASYUD YIN 8989} JOPIDLLa}IDG fo fiunwmwngs af kK 183 INFLUENCE OF SERUM UPON MENINGOCOCCI ° ‘s}WoUUTIOdxe [][v UT popnfOUT o10M BIOS PUB 9IN4[Nd UT s[oayUOD sSNoeuNy “UOT}EQNOUT SINOY FYITO-A4IOJ 10} JB OpVUI 919M S}UNOD {.1v3B og0I}xep wnsos dooys Jo 90 OT YIM poye[d yowo wody 090 GQ puB UdyeYS ‘sinoy 9014} OF “H2E 9B poyeqnour osoM soqny [TV TWUNTOS UBUINY [EUIOU OY} YIM UOTPVUIGUIOD UL IO OUOTE WUNIOS OUNWUTT YIM Solios You UT OUIeS OY} SABATY OOM UNIES OUNTMUUT JO SUOT}NIIp [BUY oY} ‘TooooOSUrTUDUE Jo SoINy[Nd YYoIG oSOIXOp WMMAVS MOY aNoj-£4U0My JO SuOTyn]Ip SutArea jo 09 G°0 YFTA popoes O19M Sogn} oy, “UNIS UBUINY 911048 YSoIJ JO SUOT}N]IP OL: T 10 ¢:] YITA JO YZoIq UMAds doays jo sjavd yenbe YIM popniip puv soynurur AyA1Y} OF “Hg 4B YWVq 19yVA B UT pozvoYy ATSnotAord ora BIOSTJUB IY} S}UIUTTIOdX9 Vseyy UT ‘GZ WNJOS SI}ISUTUOUL “que f¢:] jo°oo GQ umes Std fyyMoIS IvBe umaoas jo uorsuedsns (00‘00¢:1T JO 99 GQ UT pesn amjy[ng *g yueumtsedxyT ‘% WINANS STPISUTUOUI “que $G¢:] Jo 99 GQ) WNIOs UBUINY Y}MOIZ IvBe UNMIS Jo UOTSUdsNsS (Y0N‘00E: I jO *09 G'°Q UL posn oiNyND *¢ yueuTTIedx| "p uInJes SIPIBUIUOUNTZUB !G:] Jo 00 GQ Whies URUMY ‘oIny]ND YAorq UMaaS OOST:T JO “99 GQ UL posh oiny[Ny ‘fF queuTedxA 5 "e UNIS SyIsulueWMTUe fe: Jo ‘00 GQ Wnses UvUMY forng]ND YOIq WNIOS QOOT:T jo ‘00 GQ UI posn oiny[nyg *¢ yuoUTTIEdxA *Z wnies SIPISUIUOUTTYUB fG:] Jo ‘09 GQ UMIJOS UBUMY foINy[NO YoIq UMS CNET] Jo *09 GQ UI pasn onyynD °*z yuoUTTIedx] *[ uinies SIFIZUIMOUIT} Ue SOT: [ Jo ‘09 gg UNIeS UBUNY {ony[No YZorq UNAS OOOT:T JO “99 GQ UL pasn eIngND *T yuauTIEdxny 184 TOITSU MATSUNAMI AND JOHN A. KOLMER b. The bactericidal activity of anti-meningococcus and normal sera alone and in combination, as determined with a pipet method. In this technic which was modified after that of Wright (7) by Dr. George D. Heist and Prof. Lacey, the number of microérganisms exposed to the action of the serum for many hours was reduced to a minimum thereby rendering the test for bactericidal activity quite delicate. From five to eight dilutions of a culture of normal meningo- cocci (strain 124) prepared with sheep serum broth in a series of tubules, were employed with a many stemmed pipet; the various dilutions of culture were taken up in the respective sterile pipets and then expelled, leaving meningococci clinging to the interior TABLE 4 Bactericidal activity of active human and guinea-pig sera for meningococci* FINAL DILUTIONS OF SERA SERA 13:2 1:4 1:8 1:16 1:32 1: 64 Dy Mee aces eee |e 200) 1200 1400 1600 1800 3000 1D ee eee ene | 2700 900 1000 3700 3300 5000 1 Baa end A 2700 7500 4500 3700 2700 5500 PAG TD! a eee eee 7500 9000 9000 8000 5000 8000 * In these experiments 1 ec. of each dilution of serum was mixed with 1 ce. of culture and incubated for two hours on a water bath when plates were prepared. The culture alone showed 1200 colonies per plate before incubation for two hours in the water bath and 7000 per plate after this period of incubation. of the pipets.. Each pipet was now filled with serum, sealed and incubated for twenty-four hours when smears were prepared, stained and examined for meningococci. Numerous controls of each dilution of culture were included in which sheep serum water broth was substituted for serum. In these pipets in which all meningococci were killed, microérganisms were not found in the smears, whereas in those in which killing was incomplete or en- tirely absent, few or many cocci were to be found in the stained smears. After practice in the manipulation of the pipets, the results become fairly consistent and the method is well adapted as a ready and simple test for measuring the bactericidal activity of whole blood inasmuch as but a drop or two is required and this amount is easily obtained from a finger. INFLUENCE OF SERUM UPON MENINGOCOCCI 185 With this technic fresh normal human and guinea-pig blood were found distinctly meningococcidal, as shown in the protocols given in table 5; the blood of different persons and guinea-pigs were found to vary in bactericidal activity. As a rule human « TABLE 5 The bactericidal activity of normal human and guinea-pig blood for meningococct measured with a pipet method DILUTIONS OF CULTURE* FRESH WHOLE BLOOD ———— Undiluted 13/5 1:25 12/125. 1: 625 LIER DTN Snr aii a = = =z Tekno ae ee OSE ee ar ar ar ae ae 1d hWieNPiy ise CHR SLA RE eee — — — = = Elm any Assersye carlo e ee -- — _ = = le tremavnn Gio 25 eae ee = = = = = Errante Nel ay, che ces ove glen ciese a + + = = EUTTAD ATID He vey desde ed a eats _ — = = = nimea-pigide ye. eek ee as ais a ar ai SONS Nh 2 sr + = 7 = ORDINTROLET Bis noe eS I re ee + oF + ar ar * A twenty-four hour serum broth culture of normal meningococci (strain 124). + + = growth; — = no growth (sterile). TABLE 6 Whole blood more bactericidal for meningococci than serum i} RESULTS WITH BLOOD AND DILUTIONS RESULTS WITH SERUM AND DILUTIONS OF CULTURE* OF CULTURE SOURCE Undi- ' i 5 Undi- , , c F fated 1:5 1:25 1: 125 1625 ineeal 1:5 1:25 1:125 | 1:625 Homan.1.....) -+-T _ — _ = aL 4 = 2: ae imam 3.6225 \ — — _ — — ate ae sd = a Haman 6:....| - + = - _ a8 chs ae ails us * A twenty-four hour serum broth culture of normal meningococci (strain 124). ++ = growth; — = no growth (sterile). blood was found to possess a similar or slightly higher bacteri- cidal activity than guinea-pig blood. Whole blood of persons was found somewhat more bactericidal than the corresponding fresh active sera, as shown in the proto- cols given in table 6. 186 TOITSU MATSUNAMI AND JOHN A. KOLMER Fresh active human and guinea-pig sera were more bacteri- cidal than after heating at 56° C. for thirty to sixty minutes, although the latter are bactericidal to a degree, and showign thereby that the total bactericidal activity of normal serum is not dependent entirely upon complemental bacteriolysis. TABLE 7 The bactericidal activity of antimengitis sera alone andin combination with normal serum DILUTIONS OF CULTURE* COMPLEMENT FREE ANTIMENINGITIS SERA 5 Undi | 1:5 | 1:25 | 1:125 | 1:250 | 1:500 Antimeningococcus serum (A) undiluted| + — _ — = _ Antimeningococcus serum (A) 1:10......) + a = = _ -- Antimeningococcus serum (A) 1:100....) + _ ~ _ _ Antimeningococcus serum (A) 1:100 + human serum complementf........... — — = — ~ _ Antimeningitis serum (B) undiluted.....) + = = - _ _ Antimeningitis serum (B) 1:10.......... + - _ _ = = Antimeningitis serum (B) 1:100.........) + — - _ — Antimeningitis serum (B) 1: 100 + guin- ea=pic-complement: 2.20: 2M .eeee ee + - + + + os @ontrols idence eee ee + 7 + + — + Plater os ete ee ae ee one Une. | 2700 | 1620 |} 960} 900} 120 * Culture prepared by washing off twenty-four hour serum dextrose agar slant growths with 2 cc. of sheep serum broth, emulsifying and diluting with sheep serum water. + Plates were prepared by drawing culture into the capillary tubes employed, expelling the culture, and then washing the cocci adhering to the inner wall into petri dishes with twelve changes of sheep serum broth; the number of colonies gives an idea of the number of cocci adhering to the inner wall of a pipet and subject to the bactericidal activity of the various sera. t Four parts of antiserum 1:100 plus one part of undiluted normal serum. With this technic the bactericidal activity of antimeningitis sera Was more pronounced; in order to measure this activity, it was found necessary to use a larger number of meningococci than could be obtained in a twenty-four hour serum dextrose broth culture. The results of several experiments with antisera are shown in table 7; also an idea of the number of cocci exposed to the bactericidal substances of the sera. As shown in table 7 dilutions of antimeningitis sera up to 1: 100 INFLUENCE OF SERUM UPON MENINGOCOCCI 187 were found bactericidal and this activity was sometimes, but not always, increased by the addition of fresh normal serum and particularly human serum. AGGLUTINATION OF MENINGOCOCCI IN RELATION TO THE BACTERI- CIDAL ACTIVITY OF NORMAL AND IMMUNE SERA Inasmuch as agglutinins are regarded by many as aiding bac- teriolysis and phagocytosis, we have studied the agglutinin con- tent of horse antimeningitis and normal human and guinea-pig sera, for normal meningococci; furthermore laboratories engaged TABLE 8 Agglutination of meningococci* by normal human and guinea pig sera & ie 4 FINAL DILUTIONS ap ee BAB) 1:2 | 1:3 | 1:4 1:5 | 1:6 18 | 1:10 deg. C uM am SERUM: 4.5.5: sss + 55 + Se = = — ~ = Guinea-pig serum......... 55 a =F | Se | = = = = * Normal meningococci strain 1. } For sixteen to eighteen hours. in the production of antimeningitis serum commonly employ the agglutination test as a measure of antibody response and we have sought to determine as a by study the relation between the agglutinin, opsonin and bactericidan content of the immune and normal sera. The agglutinin content of all the various immune and many normal human and guinea-pig sera employed in this study, was determined with a macroscopic technic ;? likewise after the addi- tion of fresh normal human and guinea-pig sera to the immune sera to determine if these increased the agglutinating activity of the latter. Normal human and guinea-pig sera were found to contain traces of agglutinin for normal meningococci as shown in table 8 with one serum of each. 3 We are indebted to Dr. Shigeki Sekiguchi for assistance in conducting the agglutination tests. 188 TOITSU MATSUNAMI AND JOHN A. KOLMER The addition of fresh normal human serum to antimeningitis serum in amounts of 0.5 cc. of 1:5 dilution appeared, but not uniformly, slightly to influence agglutination as shown in table 9. The addition of tricresol to fresh antimeningitis sera to the point of 0.25 per cent, followed by exposure of the sera to room temperature for a week, was usually followed by slight decrease in agglutinating activity, which may be ascribed to the loss of labile substances or to slight decrease in agglutinin or both; heating antimeningitis sera at 60°C. for thirty minutes was TABLE 9 The influence of normal serum upon the agglutination of meningococci by antimen- ingococcus serum* ac FINAL DILUTIONS 1: SERA ‘gp als 60 | 80 | 100 | 120 | 160 | 200 | 240 | 320 | 400 | 480 | 640 deg. C Ree Antiserum alone): 22... +457: 37 | +) +) +) +) el tl -| -| -| -| - DO.) ae) ep ay aa] ae AP S| Se Antiserum+ human serum 0.5 ce. 37) +) +) +) +) +) 4+) 4+) -1 -| -| - (1: 5) See oS ier bere edparhariss| sr) =, = Autiserum + pig serum 0.5 ce. {| 37 | +) +) +) +) +) +) +) -| -| -| - (1: 5) a feel ea Val fe fen Mia eurf== [hh | * Strain of normal meningococci 124 employed. t Incubated for sixteen to eighteen hours. found to markedly decrease the agglutinating activity and this was not restored by the addition of normal serum. The results observed with a few of the sera tested are shown in table 10. Incubation of the mixtures of serum and culture in the macro- scopic tests at a temperature of 55°C. for at least sixteen hours was found to produce the most marked agglutination and usually higher than observed when the mixtures were incubated at 37°C. for the same period of time (the latter tests being conducted under sterile conditions to avoid the growth of contaminating bacteria which mask the reactions. . Despite a large amount of work with numerous repetitions of experiments, it is a difficult matter to draw deductions upon the i a INFLUENCE OF SERUM UPON MENINGOCOCCI 189 TABLE 10 ~The influence of tricresol and heat upon the agglutinating activity of antimeningocccus sera* FINAL DILUTIONS SERUM 1:10| 1:20! 1:40! 1:80 {1: 100|1: 320)1: 640|1: 1280 Peootresh and: active... ..2..s.....--- Se SES Se Se tee | Eb _ Gao yarvenwricresOlane ssh) Pica Soe +yey4+i]y4+)4+;,4+)]—- = Boo atten ME AbIN Ga c8 wi. so cole ole saa +/+)4]—-|]-|]-|]- = eT EReSI ANG ACUIVES. .. i. 5. os ene te +y/+)/+)])+;+!4+)]/2 ea PG merahuer GMeCLeSOli a... 2 <2 cs ccecsce wees + yey tye l tye] = Muborranten NeauINe: yest Sle eek ences 3s Sa er le || | | = Bios mesh sand aeulyers. 22 obra. Sone bias se - t+rt}y+)/+y+]}4+)- = Pas ater NEabln esse... si. x qpenicic ies. bacpels +/+ )])-|- | -- | —|— = * All tests were conducted with strain of normal meningococci 124 and at 55°C. for twenty hours. + Heating at 60°C. for thirty minutes. TABLE 11 A comparison of the agglutinin, opsonin and bactericidan content of various anti- meningitis and normal sera* Sit} ‘ CULTURE bse 5 : : EUSGOCTENC, RESULTS OF BACTERICIDAL TESTS CONTROLS Bea eons | 1:4 | 1:10] 1:20) 1:40| 1:80 |1:160|1: 320, 7CHPAL oe [SHHeS TESTS (20:01 Died in 17 hours | 13 | 0.01 | Died in 17 hours 12 | 0.005 Died in 17 hours | 12 | 0.005 | Died in 18 hours 12 | 0.0025 | Diedin17hours | 11 | 0.0025| Died in 18 hours 10 | 0.00125 Survived 6 days 11 | 0.00125, Died in 4 days 10 | 0.0006 Survived 6 days 11 | 0.0006 Died in 5 days uniform growths in fluid culture media have so far proven unsuccessful. 5. Highly virulent strains of meningococci yielded more regu- lar results than strains which had undergone some loss in viru- lence by reason of prolonged cultivation without animal passage (table 12); the protective tests were more regular in their results when highly virulent strains were employed. 6. For preparing the suspensions and dilutions of cocci it was not found necessary to use active guinea-pig serum; sheep serum broth was found satisfactory for this purpose. Comparative tests with heated and unheated (fresh) guinea-pig serum diluted INFLUENCE OF SERUM UPON MENINGOCOCCI 193 1:4 showed occasionally different results, the minimal lethal dose being less with the fresh active serum (table 13) and indi- cating the probable bactericidal effect of the active serum. 7. Our experiments have not been sufficiently numerous to warrant an opinion upon the relation between protection test in mice and such reactions in vitro as agglutination, bactericidal and opsonic tests for measuring the potency of antimeningitis serum, but we have tested different antimeningitis sera with the general result that those possessing the highest protective value also contained most agglutinin and opsonin. As previously stated the results of bactericidal tests in vitro were too irregular to permit comparison. 8. Even with a highly virulent strain or strains of cocci the results of protective tests on different days with the same anti- meningitis serum varied to some extent. This was expected because animal tests cannot yield strictly regular and mathe- matically exact results, owing to variations in susceptibility in addition to other factors. On the basis of our experience with the protective test of Hitchens and Robinson we believe the follow- ing conditions important: a. A highly virulent strain or strains of normal or para meningo- cocci or both should be employed—the virulence being maintained by passage through mice. b. A uniform suspension of cocci should be secured by wash ng off eight hour growths from twelve or more slants of serum dex- trose agar with sheep serum broth and mixing thoroughly. For measuring the polyvalency of a serum twelve or more different cultures of meningococci and parameningococci may be mixed and employed. c. Mice of approximately the same weight should be employed. d. In conducting a test for the protective value of a serum, one or two sets of mice should be injected with the suspension of cocci alone, the doses being arranged according to the approxi- mate virulence on the basis of former tests in order to determine the minimal lethal dose of the particular suspension being used in the protective tests. Owing to the likelihood of variation in re- sults with different suspensions, it is not advisable to determine 194 TOITSU MATSUNAMI AND JOHN A. KOLMER the minimal lethal dose with one suspension and conduct the protective tests with asecond. In other words the minimal lethal dose of each suspension used in the protective test should be de- termined at the same time and the results expressed according to the protection afforded by 1 cc. of serum over a period of 48 hours against multiples of the minimal lethal dose of culture. TABLE 14 The protection activity of antimeningitis serum alone and in conjunction with normal guinea-pig serum for virulent meningococct GUINEA-PIG WEIGHT CULTURE* ANTISERUM Saree RESULTS grams ce. cc cc 25 0.5 0 0 Died in 36 hours 23 0.25 0 0 Died in 2 hours 2 0.125 0 0 Died in 13 hours 10 0.0625 0 0 Died in 20 hours 20 0.5 0.5 0 Died in 24 hours 10 0.25 0.5 0 Died in 36 hours 22 0.125 0.5 0 Survived over 5 days 2a 0.0625 0.5 0 Died in 15 hours 24 0.5 0 OS Died in 22 hours 23 0.25 0 0.5 Died in 20 hours 23 0.125 0 0.5 Survived over 5 days 22 0 0525 . 0 0.5 Died in 20 hours 25 0.5 0.5 0.5 Survived over 5 days 20 0.25 0.5 0.5 Died in 24 hours 16 0..125 0 0.5 Survived over 5 days 16 0.0625 0.5 0.5 Survived over 5 days * Strain 124 normal meningococci suspended in heated guinea-pig serum diluted U4: 9. While this protection test in its present state has not yielded us results as close and delicate as may be obtained with a satisfactory toxin in determining the antitoxic value of diphtheria antitoxin, we believe the test to be one of value as a means for determining and measuring the approximate protective and curative value of antimeningitis serum. INFLUENCE OF SERUM UPON MENINGOCOCCI 195 b. The protective value of normal human and guinea-pig sera alone and in combination with antimeningitis serum. ‘The results of numerous experiments to determine the protective value of nor- mal active human and guinea-pig sera alone and in combination TABLE 15 The protective activity of antimeningitis serum alone and in conjunction with normal guinea-pig serum for virulent meningococct NORMAL : WEIGHT CULTURE* ANTISERUM GUINEA-PIG RESULTS SERUM grams cc. cc 15 0.01 0 0 Died in 14 hours 14 0.005 0 0 Died in 24 hours 12 0.0025 0 0 Survived over 4 days 12 0.00125 0 0 Died in 15 hours 16 0.0006 0 0 Survived over 4 days 16 0.01 0.5 0 Survived over 4 days 15 0.005 0.5 0 Survived over 4 days 13 0.0025 0.5 0 Survived over 4 days 12 0.00125 0.5 0 Survived over 4 days 12 0.0006 0.5 0 Survived over 4 days 16 - 0.01 0 0.5 Died in 15 hours 14 0.005 0 0.5 Died in 20 hours 14 0.0025 0 0.5 Survived over 4 days 138 0.00125 0 0.5 Died in 21 hours. 12 0.0006 0 0.5 Died in 19 hours r 18 0.01 0.5 0.5 Survived over 4 days 14 0.005 0.5 0.5 Survived over 4 days 14 0.0025 0.5 0.5 Survived over 4 days 13 0.00125 0.5 0.5 Survived over 4 days 13 0.006 0.5 0.5 Survived over 4 days * Strain 124 normal meningococci suspended in heated guinea-pig serum diluted 1:4. with different antimeningitis sera may be summarized as fol- lows; the protocols of several experiments are shown in tables 14, 15, 16, 17 and 18. In an effort to render the protection tests more delicate the technic of Hitchens and Robinson was modified in some experiments by using smaller doses of immune and normal serum. 196 TOITSU MATSUNAMI AND JOHN A. KOLMER 1. Normal active human and guinea-pig sera were practically without demonstrable protective value in these tests employing white mice. TABLE 16 The protective activity of antimeningitis serum alone and in conjunction with normal human serum for virulent meningococci ANTIMENINGITIS WEIGHT CULTURE’ SERUM HUMAN SERUM RESULTS grams ce. ce. cc 16 0.1 0 0 Died in 13 hours 17 0.1 ) 0 Died in 18 hours 19 0.1 0 0 Died in 19 hours 14 0.1 0 0 Died in 48 hours 18 0.1 0 0 Died in 13 hours 18 0.1 0.2 0 Died in 30 hours 23 0.1 0.1 0 Died in 24 hours 20 0.1 0.05 0 Died in 13 hours 19 (Qual 0.025 0 Died in 24 hours 16 0.1 0.0125 0 Died in 3 days 16 0.1 0 0.2 Died in 13 hours 18 0.1 0 0.2 Died in 24 hours 20 0.1 0 0.2 Died in 24 hours 23 0.1 0 0.2 Died in 3 days 20 0.1 0 0.2 Died in 13 hours 18° 0.1 0.2 0.2 Died in 24 hours 16 Opt 0.1 0:2 Died in 24 hours 19 0.1 0.05 0.2 Died in 3 days 20 0.1 0.025 0.2 Died in 13 hours 20 0.1 0.0125 0.2 Died in 24 hours * Strain 124 normal meningococci suspended in heated guinea-pig serum diluted 1: 4. 2. Occasionally the addition of normal human and guinea-pig serum to antimeningitis serum appeared to afford increased pro- tection as shown in the protocol of one experiment with pig serum in table 14, but a general review of all experiments showed that no protective influence could be attributed to the addition of the normal to immune sera. INFLUENCE OF SERUM UPON MENINGOCOGCI 197 As previously stated it is scarcely to be expected that an ani- mal test would bring out the increased protective value following the complementing of an immune serum if such actually does occur, because the body fluids of the test animal contains complement. TABLE 17 The protective value of active human serum (complement) alone and in combination with antimeningitis serum WEIGHT eth acacia Ft HUMAN SERUM CULTURE* RESULTS grams cc. cc. cc. 21 0.1 0 0.01 Survived. 24 0.1 0) 0.05 Survived. 26 0.1 0 0.1 Survived. 16 0.1 0 OR2 Survived. 20 Onl 0 0.4 Died in 24 hours 24 0.1 0.1 0.01 Survived 24 (Oyal 0.1 0.05 Survived 24 0.1 0.1 0.1 Survived 19 0.1 0.1 OFZ Died in 12 hours 21 0.1 0.1 0.4 Died in 12 hours 19 0 0 0.01 Died in 12 hours 17 0 0 0.01 Died in 12 hours 18 0 0.1 0.01 Died in 12 hours 16 0 0.1 0.01 Died in 12 hours * Strain 124 normal meningococci suspended in heated guinea-pig serum diluted 1: 4. We have also conducted combined in vitro-vivo tests by incu- bating in a water bath for one hour varying amounts of a sus- pension of virulent meningococci alone and in combination with antimeningitis serum and with antimeningitis and normal serum, followed by tests for virulence and bactericidal activity by in- traperitoneal injection into white mice; the results of these ex- periments were quite irregular, but the addition of normal serum did not appear to influence the results. 198 TOITSU MATSUNAMI AND JOHN A. KOLMER TABLE 18 The protective value of active guinea-pig serum (complement) alone and in combina- tion with antimeningitis serum WEIGHT at a freee CULTURE* RESULTS grams cc. cc. cc. 15 0 0 hyp sOXOL Died* in 15 hours 14 0 0 0.005 Died in 15 hours 12 0 0 0.0025 Survived* 12 0 0 0.00125 Died in 15 hours 16 0 0 | 0.0006 Survived 15 0.1 0 0.01 Survived 25 0.1 0 0.005 Survived De 0.1 0 0.0025 Survived 23 0.1 0 0.00125 Survived 19 0.1 0 0.0006 Survived 20 0.1 OL 0.01 Survived 23 0.1 0.1 0.005 Survived 26 0.1 0.1 0.0025 Survived 20 On 0.1 | 0.00125 | Survived 20 0.1 0.1 0.0006 Survived 19 0.1 0.1 0.01 Died in 12 hours 17 0.1 0.1 0.01 Died in 18 hours * Strain of normal meningococci 124 in heated pig serum 1: 4. SUMMARY 1. The bactericidal activity in vitro of different antimeningitis sera was found to be quite low. Fresh or active antimeningitis sera was somewhat more bactericidal than the same sera after inactivation by heating at 60°C. for thirty minutes. 2. Active normal human and guinea-pig sera are generally slightly bactericidal for meningococci. 3. The bactericidal activity of horse-antimeningitis and nor- mal human and guinea-pig serum, is largely independent of complemental bacteriolysis. 4. The bactericidal activity in vitro of normal and immune sera was best shown by a pipet method employing small num- bers of cocci and relatively large volumes of serum. b INFLUENCE OF SERUM UPON MENINGOCOCCI 199 5. The addition of active normal human and guinea-pig serum to antimeningitis serum sometimes increased the bactericidal activity of the latter. _6. Whole human and guinea-pig blood was found slightly more bactericidal than the sera alone. 7. Normal human and guinea-pig sera frequently agglutinates meningococci in final dilutions up to 1:4, but not in higher dilutions. 8. Antimeningitis sera containing the largest amounts of ag- elutinin were found to possess most opsonin and apt to prove most bactericidal in vitro. 9. The mouse test of Hitchens and Robinson was found under certain technical conditions to be of value as a means for deter- mining and measuring approximately the protective and cura- tive value of antimeningitis serum. 10. Normal active human and guinea-pig sera were practically without demonstrable protective value in mice infected with virulent meningococci, although the addition of these normal sera to antimeningitis serum appeared in some experiments to slightly increase the protective power of the latter. 11. The addition of active normal human or guinea-pig serum (complement) to antimeningitis serum cannot be expected on ihe basis of our experiments, to greatly augment bactericidal activ- ity because complemental bacteriolysis exerts but a minor réle in the relatively feeble bactericidal activity of antimeningitis serum, but the addition of normal sera definitely increases opsonic ac- tivity (10) and since it would appear that a large part of the curative properties of antimeningitis serum is to be ascribed to the presence of opsonin, it is suggested as worthy of clinical trial to complement the antimeningitis serum by the addition of active human or guinea-pig serum prior to intraspinal injection and particularly in the treatment of severe and serum-resistant infections. For this purpose human serum is superior to guinea-pig serum and may be obtained from the patient or a volunteer with the usual precautions for asepsis. 200 TOITSU MATSUNAMI AND JOHN A. KOLMER REFERENCES (1) Frexner, S.: Contributions to the biology of diplococcus intracellularis. Jour. Exp. Med., 1907, 9, 105-141. (2) Davis, D. J.: Studies in meningococcus infections. Jour. Infect. Dis., 1905, 2, 602-619. (3) Farruey, N. H., anp Stewart, C. A.: Cerebrospinal fever, 1916, Service publication no. 9, Commonwealth of Australia, 161-162. (4) Kormer, J. A., Toyama, I., anp Matsunamti, T.: The opsonic activity of fresh normal serum alone and in combination with antimeningitis serum for meningococci. Jour. of Immunology, 1918, 3, 157. (5) Hitcuens, A. P., anp Roprnson, G. H.: Standardization of Antimeningitis Serum. Jour. of Immunology, 1916, 2, 345-353. (6) Jocumann, G.: Versuche zur Serodiagnostik und Serumtherapie der epi- demischen Genickstasse. Deut. med. Wehnschr., 1906, 32, 788-793. (7) Wricut, A. E.: Handbook of the technique of the test and capillary glass tube. Constable and Company, London, 1912. 119. (8) Amoss, H., anp Woxtstern, M.: A method for the rapid preparation of an- timeningitis serum. Jour. Exp. Med., 1916, 23, 403. (9) Exser, W. J., AND Huntoon, F. M.: Studies on meningitis. Jour. Med. Research, 1909, 20, 372-541. (10) von LinGELSHEIM AND Leucus: Tierversuche mit dem Diplococcus intra- cellularis. Klin. Jahrbuch., 1906, 15, 489. (11) Kormer, J. A., Toyama, I., anp Matsunamt, T.: The opsonic activity of fresh normal serum alone and in combination with antimeningitis serum for meningococci. Loc. cit. THE RELATION OF THE MENINGOCOCCIDAL ACTIV- ITY OF THE BLOOD TO RESISTANCE TO VIRULENT MENINGOCOCCI TOITSU MATSUNAMI anp JOHN A. KOLMER From the McManes Laboratory of Experimental Pathology of the University of Pennsylvania Received for publication April 3, 1918 In a previous communication (1) we have shown the greater bactericidal activity of whole blood for the meningococcus as compared with serum and briefly described a simple and con- venient method for measuring the bactericidal activity of whole blood devised by Dr. Heist employing the many stemmed capil- lary pipet of Wright. Further experiences with this technic have demonstrated its simplicity; but a few drops of undefibri- nated blood are required and the results are usually quite sharp and convincing. As is true with all bactericidal tests, however, the results observed with the blood of one person or lower animal tested on different days not infrequently yields shghtly varying results, owing not only to probable fluctuations in the bacteri- cidan content of the blood but more particularly to fluctuations in the numbers of viable meningococci in the culture employed; with careful attention to technic however, the latter error can be reduced to a minimum. The object of the present investigation was to determine whether or not a relation exists between the meningococcidal activity of the blood and resistance to infection with virulent meningococci; furthermore whether or not the high natural im- munity or resistance of certain of the lower animals to the menin- gococcus is to be ascribed in part to a higher meningococcidal activity of their blood. Young and old guinea-pigs, rabbits and white mice were used in these experiments. Flexner (2) has found that guinea-pigs 201 1 Aided by a grant from the Pediatric Society. 202 TOITSU MATSUNAMI AND JOHN A. KOLMER weighing from 175 to 200 grams are highly susceptible to virulent meningococci as compared with pigs weighing 350 to 400 grams or more and that young pigs are more susceptible than mice; Von Lingelsheim and Leuchs (3) and Kolle and Wassermann (4) have also found young guinea-pigs highly susceptible. Betten- court and Franca (5) Elser and Huntoon (6) and Hitchens and Robinson (7) claim, however, that white mice are more highly sus- ceptible and the latter employ these animals in a protection test with antimeningitis serum. All investigators who have worked with rabbits found them highly resistant; likewise all appear to agree that young guinea-pigs are more susceptible than older and heavier animals. EXPERIMENTAL Our experiments were conducted with a single strain of men- ingococcus, the virulence and cultural characteristics of which was quite familiar to us; bactericidal tests were made with the whole blood of different series of young and old pigs, mice and rabbits followed by the intraperitoneal injection of the same culture in graded doses and according to the body weight of cach animal into each series of animals; all animals were kept under observation for two to four days and after death the blood of the heart examined for meningococci. The strain of meningococcus employed was highly virulent and all animals succumbing within four days invariably showed the presence of meningococci in the blood of the heart. In order to render the virulence tests with different animals strictly comparable, all were injected on the same day with the same emulsion of meningococci and, as stated above, according to body weight. By experiments of this nature we have sought to determine whether the resistance or non- resistance of a certain species of animal bore a relation to the meningococcidal activity of the whole blood of this species and whether variation in resistance of different species of animals bore a direct relation to variation in bactericidal activity of the whole blood. cali MENINGOCOCCIDAL ACTIVITY OF THE ELOOD 203 Bactéricidal tests In the bactericidal blood tests the many-stemmed pipet devised by Wright for measuring the coagulation-time and anti-tryptic power of the blood (8) was employed after the method devised by Dr. Heist. Each pipet measures about 9 em. in length and about 1 mm. or somewhat less in thickness and six are used at one time. In conducting the tests we have employed twenty- four hour serum broth cultures of a strain of meningococcus un- diluted and in five dilutions prepared with sheep serum broth, namely, 1:5; 1:25; 1:250; 1:500 and 1:1000. These cultures are arranged in sterile tubules and allowed to run by eapillary attrac- tion into the six sterile and numbered pipets respectively; the pipets are now blown out and each loaded to the saine level with blood secured by pricking the skin after cleansing with alcohol.. The pipets are now sealed by dipping the top in paraf- fin and incubated for twenty-four hours when a smear is made of each and stained for meningococci. With care in the technic contaminations are rare and the results quite sharp and regular. The numbers of microdrganisms exposed to the germicidal action of the whole blood are quite small being those which have ad- hered to the inner wall of each capillary tube. Controls were always included with each dilution of culture in which sheep serum broth was substituted for blood. Not infrequently the blood of persons and lower animals was able to destroy all cocci in the undiluted culture and in order to measure the bactericidal activity, denser cultures were prepared by washing off twenty-four hour serum dextrose agar cultures with 2 cc. sheep serum broth and preparing further dilutions of this emulsion. In order to estimate the number of viable meningococci in each dilution of culture adhering to the walls of the capillary tubes, plates were prepared by washing out the cocci with several changes of serum dextrose broth into sterile petri dishes followed by the addition of serum dextrose agar and counts at the end of twenty-four hours incubation. ‘ , ; oF > ~ / ‘ ' A NOTE ON THE RELATION BETWEEN PROTEOLYSINS AND HAEMOLYSINS ARCHIBALD McNEIL anp REUBEN L. KAHN From the Department of Bacteriology and Hygiene, New York University, New York City Received for publication May 23, 1918 This note is a report of studies undertaken with a view of determining whether specific proteolysins are produced in animals on protein injections, if a procedure simulating the production of specific haemolysins be adhered to. The question of the presence of specific proteolytic substances in the blood of animals injected with proteins, is an old one and no attempt will be made here to go over the range of literature. Suffice it to recall that the so-called Abderhalden reaction for eancer and pregnancy, which was based on the assumption of the presence of such proteolysins in the blood (2), is now discarded. More recently, Taylor and Hulton (1) also, studied this problem from a biochemical viewpoint. They injected various proteins into rabbits with a view of detecting the presence of proteolytic ferments in the blood of these animals, obtaining negative results. Although various attempts have been made from time to time to attack this problem from an immunological viewpoint (2), there is, to our knowledge, no work on record where a procedure was observed similar to that recorded in this paper. Generally speaking the procedure adapted for calling forth proteolysins in the animal organism was similar to that observed in the laboratories of the Department of Health of New York City, when attempting to produce haemolytic antibodies, except that some purified protein was substituted for red cells. After proper immunization with a given protein, the serum was tested for specifie proteolysins by digesting it with its protein antigen and guinea-pig complement for a definite interval at 37°C. and sub- sequently determining the increase in amino acid nitrogen. 295 296 ARCHIBALD McNEIL AND REUBEN L. KAHN Two purified proteins, edestin from hempseed and phaseolin from kidney bean (kindly furnished by Dr. Thomas B. Osborne) were used in these experiments. Two rabbits weighing approxi- mately 5 pounds each, served as the experimental animals. The injections were made intravenously. The following is a table of the quantity of protein injected and time of injections. RABBIT A RABBIT B DATE, 1915 SS. |= =e Quantity of Quantity of phaseolin injected | edestin injected mgm mgm October ™ (Qs 215i MT Aes ee ee eee 50 50 October: «7 /28s226 6 & See eet eee Stee ee 75 75 October SOs an 4 scenes Soe eee eee 100 100 INGVemben= elses see eee eee eae era: 125 125 INO VEMDer Cae cree oe ee Oe eet. oa ee 150 150 Both animals were bled under anesthesia, on November 9, six days after last injection. The sera were separated from the clots with much care to prevent hemolysis, and were divided into two portions, one-half being inactivated at 56°C. for a half hour, the remaining half being used in an unheated form. Our measure of ferment action consisted in the amino acid increase after incubation for six hours at 37°C. of mixtures of the rabbits’ serum with guinea-pig’s complement and the specific proteins. The amino acids were determined by means of the Van Slyke micro amino apparatus. The advantage of this procedure over other measurements of proteolytic action are well pomted out by Van Slyke and his co-workers in their recent paper on the Abderhalden reaction (3). ‘First, it is quantitative, and permits accurate results with the small amounts of material available. Secondly it is specific for proteolysis; it permits one to follow the chemical change which is characteristic of protein hydrolysis.” The principle of this gasometric method (4) for the determina- tion of amino-nitrogen, is based on the fact that aliphatic amino groups react with nitrous acid with the liberation of nitrogen gas. RELATION BETWEEN PROTEOLYSINS AND HAEMOLYSINS 297 Thus R.NH, + HNO, = ROH + H,O + N:. The quantity of nitrogen gas liberated serves as the measure of the amount of amino nitrogen present in the unknown solution. Another factor was essential for the accurate determinations of free amino nitrogen, namely, the removal of the proteins from the serum. The method adapted for this purpose was first sug- gested by Rona and Michaelis (5) and it has since been success- fully employed by other investigators (6) (7). Briefly stated the procedure was as follows: 2 ec. of serum were diluted to 20 ec. in a beaker and heated to boiling. 1.5 ee. of colloidal ferric hydrate were added drop by drop the mixture being shaken with each addition. Precipitation of pro- teins was then complete. 1.0 cc. of a 20 per cent MgSO, solution was added to coagulate the excess of iron. The solution was then filtered through a hardened paper into a 100 ec. evaporating dish, the filtrate being water clear. After the filtration was completed the precipitate was washed by means of a hot water wash bottle into the original beaker, about 20 ce. of water being used in the process. The mixture was again heated to boiling and the contents of the beaker were filtered into the first filtrate the same filter paper being used. Finally, the filtrate was evaporated nearly to dryness and it was redissolved in 0.5 ec. of water just previous to the amino nitrogen determination. The digestive mixture consisted uniformly of 2 cc. each 1 ce. of rabbit’s serum, 0.5 ec. of complement and 0.5 cc. of protein suspension containing 0.005 gram of protein. This quantity of protein was found to be sufficient in view of the fact that by means of the micro-amino apparatus one can measure accurately small fractions of a milligram of amino nitrogen. The following tabulation gives the procedure in detail with the results obtained. Tube A contained 1-cec. serum rabbit A (phaseolin), 0.5 ce. guinea- pig’s complement, 0.5 cc. phaseolin suspension in saline. Placed in the incubator for six hours, after which the protein was precipitated THE JOURNAL OF IMMUNOLOGY. VOL. II. NO. 4 298 ARCHIBALD MCNEIL AND REUBEN L. KAHN and the amino nitrogen determined. Quantity of amino nitrogen gas found = 0.310 ce. Tube B (control) contained same constituents of tube A, except that the amino nitrogen was determined immediately, without incubation. Quantity of amino nitrogen gas found = 0.310 cc. Tube C contained same constituents as tube A except that the serum was first inactivated for one-half hour at 56°C. The complement and phaseolin suspension were then added and placed in the incubator for six hours. Quantity of amino nitrogen gas found = 0.320 ce. Tube A’ contained 1.0 ce. serum rabbit B, 0.5 ec. complement, 0.5 ce. edestin suspension in saline. Placed in incubator for six hours, after which the protein was precipitated and amino nitrogen determined. Quantity of amino nitrogen gas found = 0.360 cc. Tube B’ (control) contained the same constituents as tube A’ except that the amino nitrogen was determined immediately, without incuba- tion. Quantity of amino nitrogen gas found = 0.345 ce. Tube C’ contained the same constituents as tube A’ except that the serum was first inactivated for one-half hour at 56°C. The complement and phaseolin suspension were then added and placed in the incubator for six hours. Quantity of amino nitrogen gas found = 0.360 cc. These findings indicate that the serum of rabbits immunized against protein, possesses no greater proteolytic activity than normal serum. CONCLUSIONS An attempt was made to find whether proteolytic substances are produced in rabbits on protein injections, if a procedure simu- lating the production of haemolytic substances in these animals, be resorted to. Proteolysis was determined by observing the increase in amino acid nitrogen after digesting mixtures of the immune serum, the specific protem and complement for a given period. The results gave no evidence of any increase in amino acids under these conditions, which would indicate that haemoly- sis and proteolysis are probably two distinct phenomena. RELATION BETWEEN PROTEOLYSINS AND HAEMOLYSINS 299 REFERENCES (1) TayLor anp Hutton: Jour. Biol. Chemi., 22, 59, 1915. (2) BRONFENBRENNER: Journ. Lab. and Clin. Med., 1915, 1, 79. (Reviews the literature. ) (3) Van Styke, VinoGrap-VILLCHUR AND Losers: Jour. Biol. Chem., 1915, 23 377. (4) Van Styrxe: This method is fully described in the following numbers of the Journal of Biological Chemistry: 1911, 9, 185; 1912, 12, 275; 1913-1914, 16, 121; 1915, 23, 407. (5) Rona anp Micuae is: Biochem. Zeitsch., 13, 121. (6) C. G. L. Wotr: Jour. Physiol., 1914, 69, 89. 7) Van SLYKE: VINOGRAD-VILLCHUR AND LOSEE: Loe. cit. THE INFLUENCE OF ARSPHENAMINE AND MERCURIC CHLORID UPON COMPLEMENT AND ANTIBODY PRODUCTION IKUZO TOYAMA anp JOHN A. KOLMER From the Dermatological Research Laboratories of the Philadelphia Polyclinic Received for publication May 31, 1918 Numerous investigations within recent years have indicated that certain drugs may induce a state of temporary immunity to trypanosome infections by stimulating the antibody producing tissues, the leucocytic mechanism or both, or, combine with antibodies and render the latter more active. Ehrlich and Shiga (1) have shown that mice infected with caderas and treated with one or more injections of trypan-red, developed a temporary immunity which could not be ascribed to an antibody response following infections with the parasites alone or to the presence of unexcreted dye, but rather to the presence of antibodies in response to the stimulating influence of the drug; later Ehrlich (2) demonstrated the same phenomenon with 7. brucez and Halberstaedter (3) in similar studies found the immunity highly specific, that is, mice infected with dourine and treated with trypan-red developed an immunity to dourine alone and not to other trypanosomes as 7’. brucet or vice versa. Cor- roborative evidence of the apparent effect of this and other drugs upon antibody production was given later by the extensive work of Terry (4) who found that a strong immunity against surra of India was obtained by injecting mice with dyes either alone or in combination with acetyl- atoxyl. That the action of the drugs is indirect rather than wholly trypanocidal, was seemingly shown by the fact that large intraperitoneal injections of surra and caderas were capable of infecting mice when introduced as early as twenty-four hours after the drug and before the latter had been wholly excreted. Further indications of the possible important relation of drugs to immunity is shown in the reports of severai homeopathic physicians as in that Watters (5), who claimed that the administration of calcium 301 302 IKUZO TOYAMA AND JOHN A. KOLMER sulphide increased the opsonic index to staphylococci; of Mellon (6) who found that the administration of baptisia influences favorably the production of group agglutinins for typhoid and other closely related bacteria and that veratrum viride increased the opsonic index to pneu- mococci; of Wheeler (7) who claims that phosphorus increases the op- sonic index of human serum to the tubercle bacillus; of Wesselhoeft (8) whose experiments were interpreted as indicating the curative effects of quinine in malaria, could not be ascribed entirely to its parasiticidal activity but probably in part to a favorable influence upon the pro- duction of anti-plasmodial antibodies; and of Hooker (9) who showed that the administration of phosphoric acid, arsenious anhydrid and mercuric chlorid homeopathically to normal persons, resulted in the elaboration of agglutinins and complement fixing antibodies for B. typhosus, B. paratyphosus A and B and B. dysenteriae. In several of these investigations the drugs alone were administered to healthy per- sons and the appearance of an increase of certain group antibodies in the blood serum was interpreted as an increase of normal or natural antibody and an indication of the possible stimulating influence of these drugs upon antibody producing tissues and a means of their curative value in certain diseases. Likewise the investigations of Arkin (10) concerning the influence of drugs upon phagocytosis may be mentioned in this connection; medicaments which have an inhibitory action upon oxidative processes as chloral, morphine and ether were found to depress phagocytosis while mercuric and other chlorids, colloidal metals, strych- nine, arsenic and others, were found to stimulate phagocytosis in vitro and 7n vivo. Following the introduction and encouraging results of arsenical compounds in the experimental chemotherapeusis of protozoan in- fections, several investigators have a studied their possible influence upon antibody production and particularly the influence of dioxydi- amidoarsenobenzol (salvarsan), with the result that a general impression exists that part of the curative influence of dioxydiamidoarsenobenzol in spirochaetic and trypanosome infections, is to be ascribed to the influence of the drug in stimulating the production of protective and curative antibodies in addition to its powerful parasiticidal activity. Aggazzi (11) found that arsenious acid, atoxyl and arsenophenylglycin increased the output of typhoid agglutinin; Friedberger and Masuda (12) claim that salvarsan increases the content of normal agglutinins and hemolysins in the serum; Boehneke (13) found that the adminis- tration of salvarsan may be followed by an increase of diphtheria anti- COMPLEMENT AND ANTIBODY PRODUCTION 303 toxin and of various bacteriolysins, opsonins and precipitins, but not of complement binding substances; Weisbach (14) also claims that the administration of salvarsan results in an increase of agglutinin and hemo- lysin, while Reiter (15) was unable to note any such influence, his experi- ments indicating that large doses of the drug lowets resistance to various bacteria. As further indications of the probable important relation of certain drugs to immunity, are several reports indicating that their administra- tion may be followed by an increase of complement in the serum. Weil and Duport (16) have reported that the intravenous administration of sodium bicarbonate to rabbits resulted in an increase of serum com- plement; Fenyvessy and Freund (17) claim similar results with the intravenous administration of calcium chlorid and Ciuea (18) found that the injection of tartar emetic and salvarsan was followed by an increase of serum complement in normal and trypanosome-infected animals, while the administration of atoxyl caused a decrease of com- plement in the serum of normal animals and in a proportion of try- panosome-infected animals. EXPERIMENTAL Since the results of these investigations have indicated that salvarsan may exert an important effect upon complement and spirocheticidal antibodies, we have conducted a series of experi- ments for the purpose of a further study of the probable influence of arsphenamine! (arsenobenzol) and mercury upon pales and complement production, as follows: 1. A study of their probate influence upon the production of immune antisheep and antihuman hemolysins and agglutinins for sheep and human erythrocytes in rabbits. 2. Upon the production of immune typhoid agglutinin in rabbits. 3. Upon hemolytic complement and normal antisheep hemoly- sin in rabbit serum. 1 Arsphenamine is the trade name proposed by the Federal Trade Commission for salvarsan and its substitutes. Throughout this study, the arsphenamine prepared by Dr. J. F. Schamberg, Dr. Geo. W. Raiziss, and Dr. John A. Kolmer in the Dermatological Research Laboratories of the Polyclinic and known as arsenobenzol, was employed in alkaline solution. 304 IKUZO TOYAMA AND JOHN A. KOLMER 4. Upon normal typhoid agglutinin and hemolytic complement in human serum. We have included a study with mercury bichlorid, because salts of mercury do not appear to have been previously employed in experiments of this nature, while their curative influence in syphilis is well known. Additional experiments similar to those recorded in this paper but with the employment of animals infected with trypanosomes are being conducted, inasmuch as the results may vary according to the nature of the stimulant (erythrocytes, typhoid bacilli or trypanosomes) used, but we have considered it advisable to record the work finished under the above plan. I. The influence of arsphenamine and mercury upon the pro- duction of immune hemolysins and hemagglutinins Experiment 1. In this experiment the sera of six large healthy rabbits were given preliminary tests for the presence of antisheep hemolysin and sheep agglutinin and 1 ec. of a 5 per cent suspension of washed sheep cells per kilogram of body weight injected intra- venously every three days followed two hours later by the intra- venous administration to rabbits 1 and 2 of arsphenamine in dose of 0.01 gram per kilo (equivalent to 0.6 per 60 kilos) and to rabbits 3 and 4, of bichlorid of mercury in dose of 0.0001 gram per kilo (equivalent to 0.006 gram or about j grain per 60 kilos); rabbits 5 and 6 were controls and received injections of cells only. All animals were bled from an ear three days after each injection of cells and drugs, the sera separated and inactivated and the titer of hemolysin determined in the presence of 1 cc. of 1:20 dilution of guinea-pig’s complement and 1 cc. of 2.5 per cent suspension of washed sheep cells; the agglutinin titer was de- termined with 1 cc. of 2.5 per cent suspension of sheep cells alone; both readings were made after incubation of 38°C. for two hours. The results of this experiment are shown in tables 1 and 2 and may be summarized as follows: 305 COMPLEMENT AND ANTIBODY PRODUCTION oe eS eee ee ee ee ! Peld | 91} OT -L} OF: ssey | OL:T sseyT!] OT:T 8897 | OL ‘Tf Ssey | OL:] sseT (Dade lira i pai ee rae JO1}U0D | 9g 991+) $6:T} OS:T} OL:T OT-T 8897 | OL: 1 S897 | OL: 1 S807 | OL:T sso'T A el ce en ease ‘Jor}UOD | ¢ 991 -T( OST.) OF :T] OOT :T 0S :T OL 1 S8e'T | OL :T Sser'T | OT: ssey | O00’ |°****Arnosou TO THOS Sy OS *T) OG :T [SZ ‘1 | OOT :T OOT :T OL 1 SST | OL:T SsoT | OT 1 ssey | T000'0 |°***-Amnosom OTH OTE iets SGT ‘1 (SCT :T SZ :1 | 99T :1 OOT :T OL -T S8oT | OL:T sso | OT :T sse7y LOUOT ae as “jozueqouesiy | Z 99T -T /99T =T S20 :T} OG :T O¢ :T OTT S8o'T | OL: Sse | OT ‘7 sse7 10°0 |" """"*"Jozueqouesry | | men Ot a worzoo fut worqo eur uorqoeful uworqzoelfur uorqoofur maceisa ane re WWnoy Pay puoveg FSI Are urunty arg Ox uaa shee ‘ON asoa TVWINY S9OuUd GNV STTGO JO NOILOACNI uaLay UdLIL NINILQIODVNaAH umn bbouay daosysyup aunwun fo uoyonposd ayy uodn fiunasau pun aurupuaydsiv fo aauanyfuy 6 WIV PFC |OOO9T :T |OOOOT | OOF :T | 991: |OOT:T 0 OT -T ssoT aie Sede WAP aos te te fetju0H | 9 OO0LE :T \OOSZT :T |O000T :T [00SZ:T jOOOL:1 | O¢:T 0 OT -T ssoT SE | alate Sic Pd 3 [OtMOO 5) 2S Q00S -T | 00S :T| 000¢ :T jO00E :T [ooOT :T SG ‘T 0S :T OL :T SST | 1000°0 |" ****Amnorow ‘sopyorg | F OOSS +I | 000S :T |OO00T :T JOOS: |O00T :1 SG ‘T OT -T S8eT | OL 1 ssey | 1000'0 | °° ****Arno sour OT Gore & OO0S -T | 000¢ :T | O00 :T jo0Sz:T [o00T :T OL :T 889T | OL 1 S8eT | OT :T ssorT LOMO es ees “lO ZueqouesIy | Z 000¢ :T | 000S :T |O000T : 1 |000¢ :t [ooOT :1 OT :T G3 :T OT -T ssory T0°0 [°° ""*""***Jozueqoussry | morjoef ar} uoroefut uoryefur | uorjoefut mor}o0fut mor49e fut aorjoef ul worqoofur Yueaeg | qIXIg DHE | Wanog | pays, puoveg a AISULUNPOIT || oar ama cave asoa pean TVWINY S9Oud GNV STIG0 JO NOILOGINI uaLavy UGLIL WALIONGH ee ee urshjomoy dasysyun ounuur fo uwononpoid ayy uodn fhaunasam pup auiunuaydsip fo aruanpzfuy I ATaAVL 306 IKUZO TOYAMA AND JOHN A. KOLMER 1. In the production of hemolysin neither arsphenamine nor bichlorid of mercury appeared to exert a stimulating influence, inasmuch as the amount of hemolysin produced in the drug treated animals was not higher than that in the contrdls; on the contrary they appeared to produce less hemolysin. It is possible that the doses of arsphenamine and mercury were too large for a three day interval of administration, but all animals appeared to stand the injections well and the majority gained slightly in weight during the course of the experiment. 2. Both drugs however, appeared to increase the output of hemagglutinin to a slight extent, as shown in table 2 by the more rapid production of this antibody in the drug treated animals. Experiment 2. In this experiment a series of six rabbits were bled and the inactivated sera were tested for normal antisheep hemolysin and hemagglutinin. Each animal was then given a daily intravenous injection of 1 ec. of a 1 per cent suspension of washed sheep cells per kilogram of body weight followed immedi- ately with an intravenous injection of the following to the first four animals: Rabbit 27. 0.004 gram arsphenamine per kilo (0.24 gram per 60 kilos). Rabbit 28. 0.001 gram arsphenamine per kilo (0.06 gram per 60 kilos). Rabbit 29. 0.00004 gram bichlorid mercury per kilo (0.0024 gram or about 4% grain per 60 kilos). Rabbit 30. 0.00002 gram bichlorid mercury per kilo (0.0012 gram or about g grain per 60 kilos). All animals were bled from the ear twenty-four hours after the injection of cells and drugs and the hemolysin content of the in- activated sera was determined by titration in the presence of 1 ce. of 1:20 guinea-pig’s complement and 1 cc. of 2.5 per cent sus- spension of sheep cells; the results of hemaggtutination tests were read after incubation at 55°C. for twelve to sixteen hours. The results of this experiment are shown in the tables 3 and 4 and may be summarized as follows: 1. Generally considered the daily administration of cells and drugs did not appear to result in a greater production of hemolysin COMPLEMENT AND ANTIBODY PRODUCTION Influence of arsphenanine a 307 TABLE 3 nd mercury upon the production of immune antisheep hemolysin HEMOLYTIC TITER AFTER INJECTION OF CELLS AND DRUGS 9 | DOSE Pe r= A a ra re | s 3 HALEN PER KILO 2.8 2 2 2 2 8 2 at 3 fe) 8 Sec) coc) Somes ae 5 oa | #8| 88) 38 ja | 28 $8 al < Ay cm 7) al Fe & 1) mM 27| Arseno- benzol. .|/0.004 1: 10!) 1: 20/1; 40) 1: 40 | 1: 40 |1 40 1: 500 | 1: 1000 28| Arseno- benzol. .|0.001 1: 20} 1: 40) 1: 40) 1: 100) 1: 125] 1: 250 | 1: 10000) 1: 25000 29| Bichlor. Less mercury|0.00004) 1: 10| 1: 20) 1: 20) 1: 100) 1: 125] 1: 500 | 1: 5000 | 1: 16000 30} Bichlor. mercury/|0.00002} 1: 20} 1: 40) 1: 40! 1: 166} 1: 250} 1: 2500} 1: 50000) 1: 50000 Less Sti Control.../ 0 1:10] 1: 10) 1:20) 1:40 | 1:40 | 1: 500 | 1: 16000) Died 32) Control... 0 1: 10) 1: 20) 1: 20) 1: 100) 1: 40 | 1: 500 | 1: 16000} 1: 50000 TABLE 4 Influence of arsphenanine and mercury upon the production of immune antisheep agglutinin AGGLUTININ TITER AFTER INJECTION OF CELLS AND DRUGS ¢ Borate |AOEM she sme lie oe akabeBBIS ABS) od al MAbs 2 28 23 33 2 ae ee ae a8 z £4 a 37 aA ie BA ue BF Less 27; Arseno-benzol |0.004 Gene Uses Sy eles SON eis i) ibe 220i ( ile 0) Less | Less Less 28} Arseno-benzol \0.001 Ibe eee Ie ai lorsye| Sey ale Sie alo sy | ier) are akg) 29| Bichlor. Less | Less mercury... .|0.00004) 1:3 | 1:3 | 1:3 | 1:10] 1:10] 1: 40) 1: 125) 1: 125 30} Bichlor. Less | Less | Less | Less mercury... -|0.00002) 1:3. | 1:3. | 153 | 0:3) 1220) 1:40) 1: 166) 1250 Less | Less oll) Control... .... 0 PS hes) P38) | LSP Oe 20) 14 OR Died Less | Less | Less Bel Control... 5... 0 Se LS ssi) ero elie O) eel, 140) ts. { i 308 IKUZO TOYAMA AND JOHN A. KOLMER than was observed in the controls receiving cells alone; rabbits 27 and 29 receiving the larger doses of arsphenamine and mercury produced less hemolysin than the animals receiving the smaller doses, indicating that the larger doses tended to depress anti- body production even though the general health and body weight of all animals was maintained at a normal level. 2. Likewise both drugs did not appear appreciably to increase agglutinin production; on the other hand the larger doses of both arsphenamine and mercuric chlorid appeared to limit agglutinin production, as was true of hemolysin production just described. Experiment 3. In this experiment a series of six large healthy rabbits were bled and the inactivated sera tested for normal antihuman hemolysin and hemagglutinin. Each animal was then given daily intravenous injections of washed human cells in dose of 1 ec. of a 5 per cent suspension per kilogram of body weight; immediately after these injections the animals were in- jected intravenously with the following: Rabbits 13 and 14. 0.001 gram arsphenamine (0.06 gram per 60 kilos). Rabbits 15 and 16. 0.00001 gram bichlorid mercury (0.0006 gram or about i35 grain per 60 kilos). Each animal was bled every three days and the inactivated sera titrated for hemolysin with 1 ce. of 1: 20 guinea-pig’s comple- ment and 1 cc. of a 2.5 per cent suspension of human cells; also for hemagglutinin employing a 1 per cent suspension of cells and reading the results after incubation at 55°C. for twelve to sixteen hours. The results are shown in tables 5 and 6 and may be summarized as follows: 1. The production of antihuman hemolysin was slight with all animals including the controls; neither drug appeared to influence hemolysin production. 2. Both drugs appeared slightly to increase the output of hemagglutinin, as shown by a somewhat quicker response on the part of the drug treated animals and the final higher titer of the serum of two (nos. 14 and 15). oa _ 4 | 7 : > ‘ _. POT Ie t COMPLEMENT AND ANTIBODY PRODUCTION 309 TABLE 5 Influence of daily injection of arsphenanine and mercury upon the production of NO, ANIMAL immune antihuman hemolysin HEMOLYTIC TITER AFTER INJECTION OF CELLS AND DRUGS DRUG DOSE jl EE peers) Preliminary After two After five After nine injection injections injections injections Arsenobenzol.. ./0.001 Less 1:10 | Less 1:10 | Less 1:10 | Died Arsenobenzol....|0.001 Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10 Bichlor. mercury...... 0.00001) Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10 Bichlor. j mereury...... 0.00001; Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10 Wontroll)...oc.< 2: 0) Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10 Controle. «oe 0 Less 1:10 | Less 1:10 | Less 1:10 | Died TABLE 6 Influence of daily injections of arsphenanine and mercury upon the production of — tet | ANIMAL BO. Or he Ww ee on SD immune antihuman agglutinin aT. AGGLUTININ TITER AFTER INJECTION OF CELLS AND DRUGS DRUG DOSE PER KILO! Preliminary After two After five [After nine injection injections injections injections Arsenobenzol......../0.001 Less 1: 10 | Less 1: 10 1:50 | Died Arsenobenzol......../0.001 Less 1:10: | Less 1: 10 aes ele 250 Bichlor. mercury... .|0.00001; Less 1:10 | Less 1: 10 LO} 500 Bichlor. mercury... .|0.00001| Less 1:10 | Less 1: 10 L216.) 1 166 Gontrolete.c; cic > 0 Less 1:10 | Less 1:10 | Less 1:10 | 1: 166 Monroe os =. han. - 0 Less 1:10 | Less 1:10 | Less 1:10 | Died The influence of arsphenamine upon normal antisheep hemolysin Experiment 4. In this experiment we have studied the influence of a single large dose of arsphenamine administered intravenously to rabbits, upon the total hemolytic activity (normal hemolysin and complement) of their sera for sheep cells and upon the normal hemolysin alone; as shown by Kolmer and Williams (19) the sera of about 70 per cent of normal rabbits contain antisheep hemolysin and about the same percentage (63 per cent) was found in the present series. 310 IKUZO TOYAMA AND JOHN A. KOLMER Each rabbit was bled immediately before, three hours and again eighteen hours after the intravenous injection of arsphena- mine in dose of 0.06 gram per kilo (equivalent to 3.6 grams per 60 kilos of body weight). The serum was separated and its hemolytic activity determined by titrating varying amounts of active serum with a constant dose of 1 cc. of 2.5 per cent sus- pension of washed sheep cells. Each serum was then heated at 55°C. for thirty minutes and its content of normal antisheep hemolysin determined in a titration employing 1 ce. of 1:20 dilution of guinea-pig’s complement and 1 ec. of 2.5 per cent sheep TABLE 7 The influence of arsphenanine upon the total hemolytic activity of the sera of normal rabbits = s RESULTS WITH SERA RESULTS WITH SERA peal anon ci tet ae ee pag AFTER £ } DOSE at PER KILO . . . . - . . . elal2i=(8l8lelslslsisislelsls/si8is o o o Oo So o o o o o o o Oo o o o o o gram | 606 | 0.06 |C|C|C|M| S| S|C/M/M] S| S| N| C|/M|M|M| S|N 607 | 0.06 |C|/M|M|M/M|S|C|M|M|M|M|s|C/M|M/|M|S]S 608 | 0.06 |C/M|M| S/N|N/|S|S/|N|N|N|N|M/M| S|N|N/N 611 | 0.066 |C/M|M/S|S|S/|M/M|S|S|S/S|C|M!S|S|S]S 618 | 0.066 |C|M|M/S|S/S/S|S|S|S/S|S/S|S|S|S}S|8 622 | 0.06 |M|/S/S|S/S|/S|S|S|S/S| S/S] S]S|S|S|S|S C = complete hemolysis; M = marked hemolysis; S = slight hemolysis; N = no hemolysis. * cells. The results in both tests were read after incubation in a water bath at 38°C. for an hour. The results of a few of these tests are shown in tables 7 and 8 and the whole has been sum- marized as follows: - a. The total hemolytic activity of the sera of 25 rabbits tested three hours after the preliminary tests and injection of arsphena- mine showed: No change with the sera of 10; a decrease with the sera of 15. b. Tested eighteen hours after the injection of arsphenamine, the sera of 22 of these rabbits surviving showed: No change with COMPLEMENT AND ANTIBODY PRODUCTION Salle the sera of 9; an increase of hemolytic activity with the sera of 2; a decrease of hemolytic activity with the sera of 11. c. The content of normal antisheep hemolysin in the sera of 30 rabbits tested three hours after the preliminary tests and injection of arsphenamine showed: No change with the sera of 27; a slight decrease with the sera of 3. d. Tested eighteen hours after the injection of arsphenamine, the sera of 25 of these rabbits surviving showed: No change with the sera of 22; a slight decrease with the sera of 3. From these experiments it would appear that a single large dose of arsphenamine tends to decrease the content of hemolytic TABLE 8 The influence of arsphenanine upon the antisheep hemolysin in the sera of normal rabbits RESULTS BEFORE RESULTS THREE HOURS RESULTS EIGHTEEN HOURS INJECTION AFTER INJECTION AFTER INJECTION DOSE — NO, PER KILO . S) 3 ) : ) SS) cS) . : ) ) 9 S 3 o o o 3 8 Co o Co S S o o o Saat ae |) Soles bP Sel rey | a |S Se S ke Wis |eSeS! |S | o o o o Co o o o o Oo o Oo Oo o So gram Fx, Pee or ss. | IS. | aol N | No SSeS | NUNS Sis |) Nahe IN; cee, 0.06 | S| N| N|N{|N/ S|NIN| NIN] S|] S|] Si NIN mo006! S|/S;)s|S/S|/S| sis] sr s|sis|s| sis feos | ON No N | NN | ON | ON CNS ON CN | NN | WN ON EN meaowe | S| N | oN | NNN [NE NO) ANUIONG), NUN ON | NaN my @.06; S| NIN; N|NIN| NI -N| NI NI N|N/ N| NN S = slight hemolysis; N = no hemolysis. complement, normal hemolysin or both and particularly the complement, for a period of at least eighteen hours following the administration of drug. Smaller doses of arsphenamine however, administered to persons, may produce an increase of hemolytic complement following a primary depression, as shown in the following experiment. II. The influence of arsphenamine wpon hemolytic complement of persons Experiment 5. The fresh active sera of 20 persons suffering with syphilis were titrated in varying doses to determine their ile IKUZO TOYAMA AND JOHN A. KOLMER content in hemolytic complement? with a constant unit of anti- sheep hemolysin and 1 cc. of 2.5 per cent suspension of washed sheep cells; each person was then given an intravenous injection of 0.4 to 0.6 gram arsphenamine in the clinic of Dr. Jay F. Schamberg in the course of their regular treatment. Blood was drawn one and eighteen hours later and the complement content was determined with each fresh serum and the same unit of hemolysin and dose of cells. The results observed with the sera of 19 persons secured one hour after the administration of arsphen- amine were as follows (a few being given in table 9 as examples) : TABLE 9 The influence of arsphenanine upon the hemolytic complement of adult syphilitic persons RESULTS WITH SERA RESULTS WITH SERA RESULTS WITH SERA Serons mgaonon | OFM EOS artax | mcneney roa oo a1 nN =| =~|/8 3 om 1A = - Ss Silal|ta = SE || S Oo o o Oo o o =) o Oo o —) o o o o o o o gram BB: 0.6 |C|M|M!|S|S|N/]|CIM/IM/| S| S|N|C|C|M] S| S/N 8.8. 0.6 |CiIM|S|S|N/N|C{/M|S/N|N/|-N/C|C!}C) C|M/S G. G. 0.6 |C|C|M!| S| Ni N| C)LCIM| S| SiN] C] CC} C) per Ja eise 0.6 1Ci GiC| S| 8S) S| CrRC|Mi S| S| S)C) CC) Sitsits PC. 0.6 | Cl Ci] Cc] C|M| S| C} C| €|CiM/| SC) CC) CiMiaiets S: HE: 0.6 C|C}C}M/ S| S| C}/C|C)M) S/N) C)C!Cc;} Cc) C|M C = complete hemolysis; M = marked hemolysis; S = slight hemolysis; N = no hemolysis. No change with the sera of 15; a decrease of hemolytic activity (complement) with the sera of 4. Seventeen of these persons returned to the clinic on the fol- lowing day, and their sera secured about eighteen hours after the administration of arsphenamine showed: No change with the sera of 5; an increase of hemolytic activity (complement) with the sera of 9; a decrease of hemolytic activity (complement) with the sera of 3. 2 As the majority of these sera also contained normal antisheep hemolysin and as this was not removed, the results indicate the influence of the drug upon the total hemolytic activity of the sera rather than upon complement alone. COMPLEMENT AND ANTIBODY PRODUCTION Bills: IV. The influence of arsphenamine and mercury upon immune and normal typhoid agglutinin Experiment 6. In this experiment the sera of a series of six rabbits were titrated for typhoid agglutmin and then each animal was given intravenously 1 ce. of a heat-killed monovalent typhoid vaccine per kilogram of body weight; each cubic centi- meter of vaccine contained two billion bacilli. The vaccine was administered every three days and it was followed two hours TABLE 10 Influence of arsphenanine and mercury upon the production of immune typhoid agglutinin | | AGGLUTININ TITER AFTER INJECTIONS OF VACCINE AND DRUGS DOSE | fe} De I S eI = i 5 4 DRUG PER gS a) i) 3 iS EB 3 z aq | ee | Bao| ae a eae les 4 & = BR a & iS RD gram Less | Less 7 | Arsenobenzol..../0.01 L744) 14) | Dred 0 0 0 Less | Less | Less | Less 8 | Arsenobenzol..../0.01 4S 4a Aa ae S64 all ST Oai 048 9 | Bichlor. Less | Less MOETCULY.. 46.. .: 0.0001} 1:4 | 1:4 | Died 0 0 0 0 10 | Bichlor. Less | Less | Less Mercury... .. 0.0001} 1:4 | 1:4 | 1:4 | 1: 768 | 1: 1024] 1: 6144] 1: 8192 Less | Less Mie Gontrol. 2... .. is 20 1:4 1:4 | 1:24 | 1: 1024) 1: 1024) 1: 6144; 1:12000 Less | Less Reet ontrol......... 0 Lira ler aS | Ue DAS ae 768 leios. [els 2043 hla 3072 later by the intravenous administration to the first two rabbits of arsphenamine in dose of 0.01 gram per kilo (equivalent to 0.6 per 60 kilos) and to the third and fourth rabbits, of bichlorid of mercury in dose of 0.0001 gram per kilo (equivalent to 0.006 gram or about 7 grain per 60 kilos). Each animal was bled three days after the injections and the inactivated serum was titrated for typhoid agglutinin in a macroscopic test, the results being read after incubation at 55°C. for twelve to sixteen hours. The results of this experiment are shown in table 10 and indi- THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 4 314 . IKUZO TOYAMA AND JOHN A. KOLMER cate that the influence of both drugs was an inhibition on agglu- tinin production; possibly this effect was due to the use of two large doses of both arsphenamine and bichlorid. Experiment 7. The sera of 24 persons suffering with syphilis were titrated for their content in normal typhoid agglutinin and the tests repeated one and again eighteen hours after the first dose of 0.6 gram of arsphenamine received in the treatment of their infection. The results of these tests showed no increase of normal agglutinin after the one injection of arsphenamine. SUMMARY 1. Numerous studies indicate that several drugs and partic- ularly arsenic compounds, may stimulate antibody production and that their curative effect in the treatment of disease and particularly syphilis and experimental trypanosomiasis, may be ascribed to this influence upon antibody production in addition to their direct parasiticidal influence. 2. In our experiments, the intravenous administration of arsphenamine to normal rabbits in doses varying from 0.01 gram to 0.004 gram per kilogram of body weight (equivalent to 0.6 to 0.24 gram per 60 kilos), did not result in an increased out- put of immune hemolysin to sheep cells, but rather suppressed hemolysin production; smaller doses did not appear to retard hemolysin production to sheep’s and human cells but likewise they did not result in an increase. The smaller doses, however, generally produced a slight increase of agglutinin for sheep’s and human cells. 3. Similar results were observed with mercuric chloride; large doses appeared to suppress hemolysin and agglutinin production while smaller doses tended to increase the production of hemag- glutinins but not that of the hemolysins. 4. A single large dose of arsphenamine (0.06 gram per kilo) administered intravenously to rabbits reduced the hemolytic activity of their sera within a period of twenty-four hours after injection, probably by an influence upon the hemolytic comple- ment; the administration of a smaller dose (0.6 gram) to syphilitic COMPLEMENT AND ANTIBODY PRODUCTION 315 persons as part of the treatment of their infection, was found generally to produce a depression in the hemolytic activity of the serum as tested one hour after injection followed by a general increase within eighteen hours. 5. Large doses of arsphenamine and mercuric chlorid to rabbits tended to limit agglutinin production for typhoid bacilli ; single doses of arsphenamine (0.6 gram) in adults did not influence the amount of normal typhoid agglutinin in their sera. 6. The general result of these experiments indicates that while massive doses of arsphenamine and mercuric chlorid tend to suppress antibody production and cause a decrease in hemolytic complement, smaller doses tend to increase the production of antibody (agglutinins) and augment the complement content after a primary decrease. It is probable that both drugs administered in the treatment of syphilis, owe part of their therapeutic efficacy to their influence upon increasing antibody production and com- plement.* REFERENCES (1) Enrutcu, P., anp Suiaa, K: Farben therapeutische Versuche bei Trypano somenkrankung. Berl. klin. Wochn., 1904, 41: 329, 362. (2) Enruicu, P.: Chemotherapeutische Trypanosomen Studien. Berl. klin. Wochn., 1907, 44: 233, 280, 310, 341. (3) HaLeerstarpter, W.: Untersuchungen bei expeirmentellen Trypanosomen erkrankungen. Centralb. f. Bakt., orig., 1915, 38: 525. (4) Terry, B. T.: Chemo-therapeutic trypanosome studies with special reference to the immunity following cure. Monograph No. 3, Rockefeller Institute. 5) Watters, W. H.: Homeopathy and immunity. North Amer. Journ. Home- op., 1909, 24: 460-472. (6) MELLO, R. R.: The effect of baptisia in the production of typhoid agglu- te «© tinins. Med. Century, 1913, 20: 261-166. (7) WurEetmr, C. E.: Recent experiments in the field of BOMmeC Hay. Brit. Homeopath. Jour., 1914, 4, 243. (8) WessELHorrt, C.: Sireioe a in regard to the action of quinine on the malarial plasmodia. New England Med. Gaz., 1913, 48, 64; 637. (9) Hooxmr, S. B.: The relation of drugs to immunity. New England Med. Gaz., 1914, August. 3 An exception to this general statement is the influence of both drugs upon the syphilis antibody or reagin concerned in the Wassermann reaction which tends to decrease during active treatment; the reagin however, may not belong to the category of antibodies. . 316 IKUZO TOYAMA AND JOHN A. KOLMER (10) Arkin, A.: The influence of strychnine, caffein, chloral, antipyrin, choles- sterol and lactic acid in phagocytosis. Jour. Infect. Dis., 1913, 13, 408-424. . (11) AcGazzi, B.: Ueber den Hinfluss einiger Arsenpraparate auf die Intensitat der Bildung von bakteriellen Antikorpern (agglutininen) beim Kan- inchen. Ztsch. f. Immunitatsf., orig., 1909, 1, 736-740. (12) FRIEDBERGER, E., anp Masupa, V.: Ueber den Einfluss des Salvarsans auf die Intensitiit der Antikérperbildung beim Kaninchen. Therap. Monatschr., 1911, 25, 288-291. (13) Boruncxe, K. E.: Die Beeinflussing der Intensitit der Immunkérperbildung durch das Salvarsan. Ztsch. f. Chemotherapie, 1912, orig. 1, 136-155. (14) Weisspacu, W.: Zur theorie der Salvarsani virkung. Ztsch. f. Immunititsf., 1914, orig, 21, 187-192. (15) Rerrer, H.: Beeinflusst das Salvarsan die Intensitiit der Antikorperbildung. Ztsch. f. Immunititsf., orig., 1912, 15, 116-144. (16) Wurtz, B., anp Dupvort: Effets des injections de bicarbonate de soude sur la teneur en alexine du milieu sanguin. Compt. rend. Soe. Biol., 1913, 74, 802-803. (17) Fenyvessy, B., anp Freunp, J.: Ueber kiinstliche Beeinflussung und Mes- sung der Komple rentwirkung im lebenden Tiere. Ztsch. f. Immuni- titsf , 1913, orig , 13, 666-681. (18) Cruca M.: L’action de quelque substances médica nenteuses sur le pouvoir alexique du serum. Bull. d.1. soe Path. Exot, 1914, 7, 626-632. (19) Koumer, J. A., AND Winurams, W. W.: Concerning natural hemolysins in rabbit serum Jour. Infect. Dis , 1913, 12, 96-102. PROCEEDINGS OF THE AMERICAN ASSOCIATION OF IMMUNOLOGISTS FirraH ANNUAL MEETING, HELD at THE New Mepicau LApor- ATORIES AND IN THE HYGIENE LABORATORY OF THE UNI- VERSITY OF PENNSYLVANIA, PHILADELPHIA March 29-30, 1918 The President, Dr. John A. Kolmer, in the Chair. 1. Report oF THE CouNcIL. EXECUTIVE SESSION. 2. THE ROLE oF IMMUNITY IN THE ConpuUCT OF THE PRESENT WAR John A. Kolmer (President’s address, see this volume, page 371). 3. EXPERIMENTS ON THE PRODUCTION OF ANTIPOLIOMYELITIC SERUM IN RABBITS Edgar H. Tsen (see this volume, page 213). 4. ActivE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS Harry L. Abramson (see this volume, page 437). DISCUSSION Arthur F. Coca: It is important to emphasize the concluding remarks of Dr. Abramson. The conditions under which Dr. Abramson was working to demonstrate immunity were highly artificial and the im- munity attained was possibly more than was necessary to constitute an absolute resistance to the natural infection as occurring in human beings. In the artificial experiments the mechanism of resistance at the natural atria of infection was circumvented. It would have been hard to foresee, from the guinea-pig experiments, how little antitoxin is necessary in order effectively to prevent the natural infection with diphtheria in human beings. Wiliam H. Park: I wish to say a word in line with Dr. Coca’s re- marks. In test animals, one species might react, and one might not react. In regard to protective sera: The rabbit gives very feeble protective serum; the monkey gives better serum, but the horse gives as good, or even a better serum than the monkey. Perhaps Dr. Abramson will say a word on that question. In my own experiments 317 318 PROCEEDINGS I first used virus from the brain, digested by trypsin, as suggested by Dr. Neustadter and then later untreated virus. Different species of animals show different periods of development of antibodies and also a different degree. John A. Kolmer: Has the serum been used in the treatment of human infections? Harry L. Abramson: 1 wish to add a word about rabbit poliomyelitis. I spent about six months inoculating rabbits with material from human cases and experimental poliomyelitis with rather discouraging results. Only a few of a large number of young animals inoculated intracere- brally, perineurally and intraperitoneally exhibited flaccid paralyses. The presence of paralysis was considered the only reliable evidence of a possible poliomyelitic infection. The pathologic findings in the cords of these animals were not like that of human poliomyelitis. The sera of these animals were not tested for the presence of immune substances. However, an attempt was made to determine whether injections into monkeys of emulsions of cords from rabbits that exhibited flaccid paralyses would prevent the experimental disease. In one in- stance an animal so treated was paralyzed after an incubation of four- teen days whereas the control animal was paralyzed in six days. It is problematical whether this prolongation of the incubation period in the treated animal was due to any degree of immunity conferred by the injections of rabbit’s eord or whether it was simply an instance of variability in incubation period. The test virus used was that of the 1916 epidemic, which is quite variable. As regards an anti-poliomyelitic horse serum, two animals have been subjected to injections of polio-virus emulsions over a period of three to six months. Both have neutralized in vitro a 5 per cent emul- sion of the highly virulent Rockefeller strain in the proportion of one part of serum to one part of virus. I have had no personal experience in the use of this anti-poliomyelitic horse serum in the treatment of the human disease. Dr. Neustadter of New York treated three cases, in all of which there were rather severe reactions. It is very difficult to draw conclusions as to the value of serum therapy in poliomyelitis because of the great difficulty in prognosis in this disease. Large numbers of unselected cases must be treated and the mortality figures compared with that of an equal number of un- selected cases in the same outbreak before one can make any positive statements on this matter. Rosenow’s anti-poliomyelitic horse serum was subjected to the im vitro neutralization test and was found to neutralize the Rockefeller virus in proportion of one part of serum to one part of virus. 5. Types or Mernrtncococct CoNCERNED IN EPIDEMIC INFECTIONS A Parker Hitchens and C. H. Robinson: The troops in this country are now experiencing the same conditions that the British troops experienced at the beginning of the war. The work of English investi- PROCEEDINGS 319 _ gators is here of value, especially that of Gordon on the meningococci. Gordon typed the different strains of these organisms as has been done with the pneumococci, into four different groups. This work has been followed in the present experiments, and 100 cultures, received from the first of the year, have been typed according to the Gordon method. The strains have been received from several concentration camps in this country. The cultures, from spinal fluids or from naso-pharyngeal swabs—were Gram negative cocci. They corresponded to a greater or less degree to Gordon’s types; namely, I, parameningococcus; Ue normal meningococcus: III and IV, C omparatively rarely found in this country. The grouping of the strains varied with different camps. Camp Jackson had 1 strain of type I; 9 strains of type II; 1 strain of type III. Great Lakes camp showed: type I, 1 strain; ‘type II, 28 strains; 2 Flavus; 2 undermined. It was shown that each camp had a predominating type, but whether this was a true epidemic was not determined. The work was undertaken with the hope that specific sera might be evolved to counteract the types. Gordon has had excellent results with his specific strains. He found that if specifie serum were given before the third day of the disease the mortality was only 9 per cent: if on the seventh day the mortality was 50 per cent. In testing the types the precipitin test has given valuable results, which were confirmed by cultural methods. Agglutination of strains seems to give distinct differences which follow the types as laid down by specific sera, but what the relationship is, is not yet conclusive. A rapid method of typing the meninogococci was sought, so as to expedite the application of specific therapy. * 6. THe INFLUENCE oF NorMAL HuMAN AND GUINEA-PIG’S Sut (COMPLEMENT) ALONE AND IN COMBINATION WITH ANTIMENINGITIS SERA OPON VIRULENT MENINGOCOCCI. F John A. Kolmer, Toitsu Matsunami and Ikuzo Toyama (see this volume, page 157). 7. THE RELATION OF THE MEnNINcococcwaL Acrrvenat OF THE Boop TO RESISTANCE TO VIRULENT MeniNGococer Toitsu Matsunami and John A. Kolmer (see this, volume, page 201). DISCUSSION. William H. Park: I would like to ask\a question, as to. the use of “Specific therapy” in the treatment of |cerebro-spinal meningitis Could one get a monovalent serum for.one type which would be appreci- ably stronger than the polyvalent serum: for the same. type?,, Lf,one or two types could be injected into the herse at. once and a nearly equally good antiserum produced for each, the advantage, would, be much greater than if one used a different horse. for.each|type. ,. Is there evidence that | 320 PROCEEDINGS a polyvalent serum had been produced? If not, it would be advantage- ous to produce a serum for one or two types. In the pneumococcus serum it is advisable to make antisera against distinct types, because the microdrganisms are not closely related; but in this case a polyvalent serum against two or more types would seem to be equally potent. Dr. Hitchens: Dr. Park’s point is very well taken and I agree up to a certain point. There might, however, be a greater possibility of reducing the death rate through the use of specific sera. It is possible that there may be immunological variations of meningococci within the serological types. It is well known that with the best baianced sera obtainable one finds cases of meningitis to resist their action. The meningococci isolated from spinal fluids in these cases are resistant in culture to the action of the polyvalent serum, although belonging to one of the well known fixed types. These may be serum-fast strains or not agglutinable strains. The Rockefeller Institute agglutination work would show that meningococci can not be accurately and sharply grouped within certain types, but follow the order of a chromatic scale. There are an almost innnumerable number of variations, al- though the variations are slight. The problem of getting a polyvalent serum is therefore difficult. It is felt, however, that work should fol- low the lines of making a polyvalent type serum and find out whether this would not confer more resistance in the refractory cases. It will be found necessary for this purpose to use many strains of meningo- cocci. One can not select single strains of the four types and immu- nize a horse to these four strains because there are an infinite number of variants within the types. In one case where the spinal fluid gave a specific reaction with the precipitin test, the patient did not do well with the polyvalent serum. With a specific serum the case recovered. In regard to Dr. Matsunami’s paper, those who are interested in active immunization against cerebro-spinal fever will welcome the technic of Dr. Heist. If even a low grade of immunity can be demon- strated, this technic will be of considerable assistance. William H. Park: Dr. Hitchens has not exactly answered my ques- tion. Can not one horse be injected with two or three types, as well as with one type alone? In New York we are using the combined serum, and we have obtained a reduction of mortality to 22 per cent. It will not do to take the time to type individual strains asin pneumonia. Has it been proved that one can not produce a serum from a single horse that would answer to several strains? If not, there must be a horse for each strain. This would be very confusing from the laboratory stand- point. H. Parker Hitchens: I thought I had replied to Dr. Park. If the number of antigens could be limited to four or five there would be no reason to think of specific sera.. It is not possible to produce sera equivalent to each four strains, as high as for a single strain. In an antiserum for all strains it may be as high as against a single strain. lf there were only five variants it would not be necessary to think of specific sera; but, if, as has been suggested, there isa chromatic scale of PROCEEDINGS aol -variation—forty to sixty different varieties, one can not obtain sera as high for individual variants, as if they were split into groups of four or five. There is, however, as yet, no positive evidence in support of this theory. Charles Krumwiede: I have been working on the precipitin test for the diagnosis of type. From my experience I anticipated considerable cross reaction between groups one and three and between groups two and four. Before progress can be made, an easier way of differentiation eliminat- ing the necessity of absorption to obtain an accurate determination of type will have to be found. However, the test might differentiate, with some degree of accuracy, groups one and three from groups two and four. Nevertheless, there is danger of a misleading reaction due to group antibodies even though the strain does not belong to any of the four groups. I have found, as has Dr. Kolmer, that horse serum is rich in normal opsonins and it is difficult for this reason to standadize ther- apeutic sera by determining the opsonic content. Attempts to dif- ferentiate between “normal” and “immune”’ opsonins by heating and reactivation have failed. 8. THe NATURE OF ANTIANAPHYLAXIS J. Bronfenbrenner and M. J. Schlesinger: The convulsion in eclampsia and epilepsy as well as the characteristic symptoms of asthma, hay fever and certain dermatoses have been interpreted by a number of investigators as anaphylactic phenomena. Our earlier experimental studies of anaphylaxis (1) led us to the conclusion that the latter are due to the liberation of proteolytic enzymes in the blood of the individ- uals and experimental animals affected. We found that the activation of these proteolytic enzymes is controlled by the mechanism based upon the balance of the proteolytic ferments and antiferments in the blood (2). The actual measurements of antitryptic index in the individuals during eclamptic and epileptic seizures, confirmed our view that a decrease of antitryptic powers of the blood leads to pro- teolysis 7n vivo, and vice versa, that recovery from anaphylaxis is ac- companied by a marked rise of antitryptic power of the blood (3). Since it was shown by several investigations that sensitized animals can be protected against anaphylaxis by special treatment, we started out to study the mechanism of this protection with the view of finding, if possible, a theoretical basis for the methods of prevention of clinical anaphylaxis in man. It is an established fact that the injection of a sublethal dose of antigen preliminary to the test injection prevents shock in sensitized animals. The explanation of this phenomenon offered by Friedberger was that the sublethal dose of antigen exhausts the serum of its specific antibody by combining with it, and thus prevents the formation of anaphylatoxin when a subsequent test injection is made. There are, however, some phenomena in antianaphylaxis which cannot be adequately explained on the basis of exhaustion of antibody. Thus, for instance, it was noticed that already a few days after the injection a22 PROCEEDINGS of a sublethal dose of antigen into hypersensitive animal, its hypersen- sitiveness partly returns and that, in general, the length of the anti- anaphylactic state depends upon the size of the vaccinating injection. Moreover, the experiments of Anderson and Frost suggested that anti- body is present in the blood of antianaphylactic animals long before they return to the state of hypersensitiveness. Our own experiments conducted with the view of determining whether the exhaustion of anti- body was the underlying mechanism in the experiments of Friedberger led us to the conclusion that such was not the case. We found that animals sensitized simultaneously against two proteins and receiving a vaccinating injection of one of them become resistant to the test in- jection of the second protein, though there could be no question of exhaustion of the second antibody. We found further that guinea- pigs receiving a large sublethal dose of any anaphylatoxin prepared in vitro become resistant to a subsequent test injection of several lethal doses of the same or any other anaphylatoxin. Moreover, direct experiments 7m vitro by means of the Abderhalden reaction show definitely that the antibody is not exhausted from the blood of sensi- tized animals receiving a sublethal dose of antigen. That the state of antianaphylaxis in experimental animals is not due to the changes in antibody concentration in their blood is also suggested by the fact that animals can be rendered resistant to anaphylactic shock by a number of nonspecific methods. ~Our analysis of the mechanism of the anti- anaphylaxis produced in experimental animals by a number of such nonspecific methods brought out the following conclusions. Intro- duction of certain poisonous substances, which may cause destructive changes in the tissues, in quantities not sufficiently large to kill the animal outright, is followed by the death of tissues immediately affected by poison. In this process the intracellular ferments are set free, and together with the ferments thrown out from the sur- rounding fixed cells as well as from the blood serum and leucocytes digest the dead material. The split products of such digestion exert antagonistic action, and retard or stop further activity of the proteolytie ferments. The time of appearance, the rate of their increase and the length of time during which these split products remain in the blood, determine the antitryptic titer of such blood. We have tested with this point in view the effects of ether, chloroform, alcohol, choral, morphine and scopolamine, and we found that the power of these substances to protect the animals against anaphylaxis is strikingly parallel with their power to increase the antitryptic titer in the blood of these animals. We found that the animals were protected against anaphylaxis only so long as the antitryptic titer resulting, from the treatment remained above the normal. The same relation between the power of substances administered to protect against anaphylaxis and to increase the anti- tryptic titer of the blood of treated animals was found also in the cases of treatment with BaCle, CaCl: as well as with lecithin and cholesterin. It was suggested by several investigators that starvation or exposure to low temperature may also protect the sensitized animals against PROCEEDINGS 323 anaphylaxis. Chemical studies by various investigators suggest very strongly that there is a certain amount of analogy between the changes in metabolism of animals during starvation and anaphylaxis. We found accordingly that the antitryptic index of the blood or starving guinea-pigs is above normal. As for the protective influence of low temperature, our measurements have shown that it 1s not connected with the antitryptic balance, but is merely due to the fact that the whole process of anaphylaxis becomes slow under the influence of cold and the symptoms become, accordingly, less acute. 9. StuDIES ON So-CALLED CELLULAR ANAPHYLAXIS W. P. Larson and E. T. Bell. DISCUSSION John F. Anderson: I wish to ask Dr. Bronfenbrenner whether my impression is correct that if the animal were sensitized to more than one protein there would be an inhibitory effect from the injection of a sublethal dose. J. Bronfenbrenner: In the experiments with animals sensitized to two antigens in which one of the antigens is injected in order to vaccin- ate against the effect of subsequent injection of the second antigen, one has to be very careful properly to select the vaccinating dose. If the amount of antigen injected is but a small fraction of the lethal dose, it may not protect sufficiently, but if the amount inoculated is sufficiently close to the minimum lethal dose, so that 1t may even cause slight ana- phylaxis, the animal is temporarily protected against at least one and a half or two lethal doses of the second antigen. Arthur F. Coca: Dr. Larson’s results contradict those of some pre- vious work of my own; that is, they seem to show that my conclusions, as far as they pertain to the percentage of residual blood after per- fusion, were wrong. This, however, does not invalidate my chief con- clusion, which was that the site of the anaphylactic reaction is in the tissue cells, because in the same paper evidence is presented proving that if as little as 50 per cent of the actively sensitized guinea-pig’s blood is removed by perfusion the residual 50 per cent will not, in half of the animals, contain an amount of the sensitizing antibodies sufficient to sensitize a single guinea-pig. The experiments in perfusion of pas- sively sensitized guinea-pigs were still more conclusive inasmuch as Weil had shown that the blood of such animals contains no demonstra- ble antibodies forty-eight hours after the sensitizing injection. Dr. Bronfenbrenner’s experiments illustrate what has been called nonspecific antianaphylaxis. For the purpose of Dr. Bronfenbrenner’s argument it is necessary to show that the desensitization observed is specific; otherwise the experiments can throw no light on the nature of specific “antianaphylaxis.”” In general it should be borne in mind that any humoral theory of anaphylaxis is at a disadvantage in not being able to explain the latent period in passive sensitization. 324 PROCEEDINGS G. H. A. Clowes: I can substantiate Dr. Anderson’s position. Six years ago he had tried desensitization against hay fever in patients sensitive to spring and autumn fever. Timothy and ragweed were used and it was found that the skin test was strongly specific. Patients desensitized to timothy were not altered in their reaction to ragweed. Immunity developed only against the agent employed. The matter of antitryptic reaction is related to the matter of surface tension. The soaps, lecithins, calcium chloride and barium chloride play a part in this question. The nonspecific interference is a matter of surface tension I myself have obtained a marked desensitization to hay fever with a marked antitryptic index. This was followed up with care on ac- count of its relation to the cancer question. The high antitryptic index was coincident with a maximum desenaitisation following a slight anaphylactic shock which fell appreciably a few months after the hay fever period. Anaphylactic effects appear to be due to increased per- meability of the protoplasn ic film. J. Bronfenbrenner: Of course the the time which is allowed to elapse between the vaccinating and the test injections in my experiments is a very essential element. The test injection must follow closely enough, so that the antitryptie index of the vaccinated animal is still above the normal at the time of the test injection. Another equally important factor is the method of injection. If the vaccinating injection is given int-aperitoneally or subeutaneously—the results can never be as sharp as in the case of intravenous inoculations. Dr. Coea is quite right in making a distinction between specific and nonspecific phenomena in antianaphylaxis, but in my experiments I did not intend to study this question. My problem was to see whether the mechanism of antianaphylaxis in the experiments of Friedeberger was that of exhaustion of antibody, and if not whether the mechanism in this case as well as in other instances of antianaphylaxis is the same. I suspected that there must be only one essential process (or set of processes) underlying all the phenomena of antianaphylaxis, be- cause it was found empirically that the specific anaphylaxis can be regularly checked by such nonspecific treatment as the administration of ether or the injection of BaCl. While the experiments of Bron- fenbrenner do not prove that the entire mechanism of the phenomena ean be reduced to the question of the control of ferment action in the blood, they show, nevertheless, conclusively that in all the cases of anti- anaphylaxis studied by him, the high antitryptic index of the blood was a part of the symptom-complex of antianaphylaxis in guinea-pigs, and vice versa; whenever the antitryptic index of the blood in the guinea- pigs pa raised (by whatever method) the animals were refractory to shock. Dr. Larson’s experiments seem to me exceedingly valuable. It was very interesting to see so clearly demonstrated that the perfusion of organs is very irregular. Dr. Larson himself has drawn conclusions as to the importance of his observation in relation to the question of cel- lular anaphylaxis as contrasted with humoral anaphylaxis. There is _ PROCEEDINGS 325 very little doubt but that the cells of the body take part in what we observe as final symptoms of anaphylactic reaction. What is the exact part played by the cells is not yet well understood, but it seems that in all of the experiments in which the behavior of the tissue was studied during the anaphylactic shock the presence of the serum in such tissues was not definitely excluded. While such was the feeling of several investigators all along, Dr. Larson’s experiments for the first time demonstrated that such was the case. I would like to ask Dr. Larson whether he can offer any suggestions as to the reason why the fluid with which the organ is perfused does not reach certain parts of the tissues. 10. EXPERIMENTAL POLLINOSIS IN GUINEA-PIGS Henry L. Ulrich. DISCUSSION G. H. A. Clowes: I do not feel competent to discuss Dr. Ulrich’s paper as a whole, but I am interested in the results. In experiments made with human material, numerous attempts have been made with a variety of means to induce sensitization in individuals that were not previously sensitive, but without result. This might give a clue to the work upon this mysterious condition. John F, Anderson: A good many years ago it was demonstrated that animals could be sensitized by instillations in the nose so as to respond with symptoms of anaphylaxis. Has Dr. Ulrich confused the sen- sitization to protein with the sensitization to pollen. These are two entirely different propositions and opened a field for large experimenta- tion. Arthur F. Coca: No doubt we all recognize this work of Dr. Ulrich as revolutionary. Despite the innumerable injection into human beings of various kinds of foreign protein no instance of the experi- mental production of a condition corresponding to hay fever has yet been reported. In a series of experiments with Dr. Cooke and Dr. Flood I was unable to produce the usual condition of experimental anaphylaxis in guinea-pigs with strong extracts of poilen. A. Parker Hitchens: I would like to ask Dr. Ulrich: what pollen was used and what method of extraction? All the pollen I have used was very badly contaminated with bacteria of different kinds. The bac- terial contamination might play a large part in the phenomena which resulted. G. H. A. Clowes: I found contamination by bacteria and minute insects. Reactions to these might be introduced. I never succeeded in getting reactions in normal individuals. Geo. H. Smith: We instilled pollen into the nose and mucous mem- brane of normal animals and also we tried to injure the mucous membranes to see whether they would respond more after injury. The animals were shut up in glass cages in which pollen was kept cir- culating by means of the electric fan. No sensitization was obtained. 326 PROCEEDINGS Henry L. Ulrich: The pollen was washed in acetone and the sterile pollen was injected into the peritoneal cavity. It never showed any growth. It was used dry in powder form. 11. A Skin ReaActTION TO PNEUMOTOXIN Charles Wess: This study has been instituted with the idea of using the endocellular toxin of the pneumococcus (hemotoxin) rather than the bacterial emulsion. The protein free pneumococci were dissolved in sodium cholate. A specific reaction was obtainable in guinea-pigs and also in persons suffering from lobar pneumonia. The guinea-pigs were sensitized with a sublethal dose of pneumotoxin. The animals survived the injection and there was a true pneumotoxin reaction, which was different from the reaction to pneumococcus protein or autolysates of it. A vaccine of dead pneumococci was tried and three animals reacted to it. It was found that heating pneumococci at 56°C. for one- half hour was not sufficient to destroy the endocellular pneumotoxin. The skin test would appear to be a true test for the presence of pneu- motoxin in the body of the animal. By various chemical tests it is found that the pneumotoxin is a true protein, and in human cases with lobar pneumonia there was a positive reaction demonstrable before the crisis and a negative reaction after the crisis. In children the reaction was most characteristic. Other persons tested, suffering with tuber- culosis, appendicitis, skin broncho-pneumonia or with acute or chronic infections not of pneumococcic origin, as well as healthy adults and chil- dren did not react. The reaction is regarded as similar to the tuberculin reaction and is indicative of a state of allergy to pneumotoxin. Sensitization to the toxin presumably takes place with its liberation (by the action of nor- mal body enzymes upon pneumococci normally localized in the lung alveoli) at the time of the prolonged chilling due to exposure. Failure to elicit the reaction during convalescence indicates the establishment of a temporary immunity or the disappearance of excess of toxin. This skin test does not seem to be of value as a method of serological type diagnosis but may aid in differential diagnosis between appendicitis or tuberculosis and pneumonia (especially in children). It is also of inter- est because of its bearing on the mechanism of the crisis. 12. THE INFLUENCE OF ARSENOBENZOL AND MERCURY UPON ANTIBODY PRODUCTION I[kuzo Toyama and John A. Kolmer: The possibility of certain drugs acting as antigens have been the theme of several studies by different workers. The antigenic action of drugs may account for acquired drug tolerance and also aid in resistance to infection by stimulating antibody production against microparasite apart from direct action of the drug on the parasite. The latter phase of the subject is the one particularly dealt with by the authors. Considerable evidence would point to the PROCEEDINGS 327 conclusion that many drugs exert a stimulating action on antibody production by the tissues. Such drugs as arsenous anhydrid, phosphoric acid and mercuric chlorid, administered by mouth, are found to act in this manner. Salvarsan appears to stimulate agglutinin: production, according to some workers. The present experiments were conducted on rabbits, to determine whether small daily doses of arsenobenzol and mercuric chlorid tend to increase antibody response to alien erythrocytes or to typhoid bacilli. After a series of experiments, five in all, on a large number of rabbits, and control animals, it was found that no increase of antibody production was shown, after injections of arseno- benzol or mercuric chlorid. On the contrary, it would seem that there is a lowering of antibody production, probably due to lessening of re- sistance by toxic effects. Further experiments, however, are in prog- ress, upon the action of these drugs in experimental trypanosome in- fections. It is felt that such work should be done upon human sera, as work upon the sera of lower animals may not be a true index in human eases. The importance of the subject demands careful experimental work in this direction. JOINT SESSION WITH THE AMERICAN ASSOCIATION OF PATHOLOGISTS AND BACTERIOLOGISTS 1. A CoNTRIBUTION TO THE BACTERIOLOGY OF B. FUSIFORMIS; ITS . MorpHoLocic PHASES AND THEIR SIGNIFICANCE Ralph R. Mellon. 2. THE VARIOUS IMMUNOLOGICAL REACTIONS IN GLANDERS G. Benjamin White: During an epidemic of glanders in a herd of ninety- five horses an excellent opportunity was afforded for observations on the results of various immunological tests. The following observa- tions may be reported. In all cases where glanders lesions were found at autopsies the horses had given positive subcutaneous mallein tests whereas the eye test and the complement fixation and agglutination tests were in some cases positive and in some cases negative. Horses giving positive eye tests always gave a positive subcutaneous test but were either positive or negative by complement fixation or agglutination. Complement fixation and agglutination tests were sometimes positive with a negative ophthalmic, negative subcutaneous and no lesions at autopsy. It was found that the subcutaneous injection of mallein never sen- sitizes a horse to the extent that it will react positively to a second subcutaneous test nor does such an injection produce sensitization of the mucous membranes. The eye test, therefore, in such cases cannot be changed from negative to positive with the subcutaneous injection of mallein even when the injection is repeated several times. Such in- jections, however, do cause complement fixation and agglutination to 328 PROCEEDINGS change from negative to positive. The injection of dead mallei bacilli — produces no general or eye sensitiveness but does change a negative complement fixation and agglutination to positive. It is felt, there- fore, that a positive subcutaneous mallein test shows the presence of a glanders lesion. The lesion, however, may be a healed, or inactive, one. The subcutaneous injection of mallein renders glandered animals insensitive to subsequent injections for a period of about four weeks. No horses were found refractory when tested five weeks after the first injection. The similarity between immunity reactions in glanders and those in tuberculosis is pointed out. DISCUSSION William H. Park: There is no absolute relation between the comple- ment fixation, agglutination and mallein tests in horses. Very few New York horses would pass all three tests. This has been a serious difficulty in determining what horses should be used for food. It has been found necessary to require both serum reactions to be positive or a definite mallein reaction, before reporting the horse as infected. This has been done and it has been found to admit sufficient horse meat to alleviate the food shortage. A. Parker Hitchens: It is my feeling that many more horses react to the subcutaneous test than those having glanders. I apply the sub- cutaneous test, as well as eye test and complement fixation, and I depend upon complement fixation. I have found that we must accept some horses that give positive subcutaneous tests. We have had no trouble with those animals. Reading of the subcutaneous test requirés considerable skill; the symptoms vary. Just what is a positive sub- cutaneous reaction and what is not, is open to question and depends largely upon the individual making the examination. G. Benjamin White: I quite agree with Dr. Hitchens that a horse may give a positive subcutaneous test and yet have no active glanders. A healed lesion may be present and it has been my experience that in such a case the horse develops no active glanders unless subjected to some particularly severe strain. 3. PERSISTENCE OF ACTIVE IMMUNITY IN THOSE IMMUNIZED AGAINST DIPHTHERIA William H. Park: In testing different species of animals, some have been found to have no immunity; some are entirely immune and again there is a group in which some are immune while others are not. Guinea- pigs have no natural immunity whereas horses almost always possess antitoxic immunity. The guinea-pig requires from four to eight weeks to develop immunity after toxin-antitoxin injections while the horse can be immunized very rapidly and nearly always produced consider- able antitoxin. The guinea-pig loses its immunity in about nine months; PROCEEDINGS 329 in the horse at the end of twelve months the developed immunity drops to the original level, i.e., 45 to } unit. Human beings differ from both horses and guinea-pigs in that some are immune while others are not. Dr. Zingher has made interesting studies in immunizing groups of babies. One group was of immune babies with immune mothers; these remained immune after the period of passive immunity from the mother had passed. Another group was from mothers that were not immune; these became immune after the injections. Dr. Zingher tested children in institutions and found that fourteen months after immunization, those that had received but one injection of antitoxin were as immune as those that had had received two or three injections. The development of the immunity, however, was slower. Results observed by Dr. Rosenberg are very encouraging; the very great majority of children remained immune. It seems as though the acquired immunity to diphtheria will prove to be as permanent as natural immunity. Of 404 immunized children tested, 396 remained immune. Two of these had never become immune. From these results it seems practical to immunize infants; this will give protection at the period of life when there is the greatest danger. DISCUSSION Alfred F. Hess: The use of toxin-antitoxin mixtures has made a great difference in the Hebrew infant asylum in New York City. For- merly there were about 15 cases of diphtheria a year in the institution with three or four deaths, but since the use of toxin-antitoxin was introduced there has been no diphtheria. Of 150 cases injected and followed for from one to two years, 98.75 per cent have remained im- mune. As regards the best time for immunization, early injection should be made because, while 80 per cent of infants have derived a passive immunity from the mother and are thus immune during the first few months of life, yet this immunity is for the most part lost in the later months of the first year. Immunization with toxin-antitoxin injections will, I believe, entirely do away with diphtheria in institutions. John A. Kolmer: What is the nature of the reaction following the injection of toxin-antitoxin? Is a single dose advocated or is it wise to give a second one? William H. Park: Children receiving only one injection develop immunity more slowly than those receiving two or three injections; 90 per cent of the former become immune by the end of six months whereas an equal percentage of the latter are immune by the end of the third month. For immediate protection antitoxin alone must be given. The toxin-antitoxin method is being used in the Navy as diphtheria is not always recognized at once and furthermore if the seamen are immunized one does not need to worry about carriers. In a few cases the injections cause a slight toxic effect. If the inert protein in the mixture can be controlled the procedure will become entirely harmless. John F. Anderson: Is there a quick deterioration in the mixture? THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 4 330 PROCEEDINGS William H. Park: So-called “‘standard”’ toxin that has become rel- atively stable through long standing, is used for the toxin-antitoxin mixtures; the mixtures can, therefore, be kept for months without appreciable change. 4, A SmmpLt—E Meruop For BLoop CULTURES John G. Wurts and S. W. Sappington. 5. A BACTERIOLOGICAL StuDY OF PostT-OPERATIVE PNEUMONIA Miriam P. Olmstead: One hundred and thirty cases of post-operative pneumonia have been studied and a pathogenic organism has been recovered from at least thirty-one, a percentage of 23.8. Two cases had a Pneumococcus I in the blood stream, one of these Pneumococcus I in pre- and post-operative sputum. One had a Pneumococcus II in the sputum, pre- and post-operative, and the urine was precipitated by Pneumococcus II serum. In two cases with an atypical II in pre- and post-operative sputum, the etiology was established, in one by a positive agglutination test of the patient’s serum, and in the other by a urine precipitin reaction with Type II serum. Pneumococcus III was established as the inciting factor in five cases; one had a positive blood culture, from one the organism was obtained by lung cultures at autopsy, the blood of two agglutinated the Pneumococcus III re- covered from the post-operative sputum, the urine of these two and one other, who had Pneumococcus III in the pre-operative sputum speci-_ men, gave a precipitin reaction with Pneumococcus III serum In eighteen cases a Pneumococcus IV was found to be the inciting factor. It was recovered from the blood once and seventeen cases gave a posi- tive agglutination reaction with strains from the sputum. In one of these cases the urine precipitated with Pneumococcus IV serum. A hemolytic streptococcus was found in the blood stream of one case, and a hemolytic streptococcus in the sputum of another gave a positive thread reaction with the patient’s serum. A mucoid streptococcus was found in sputum and chest fluid of one case. Some of these organisms were undoubtedly the inciting factors of other post-operative pneumonia cases, but in the absence of a positive blood culture or an immunological reaction, the significance of their presence in the sputum is uncertain. The most common inciting factor, as Whipple (Surgery, Gynecology and Obstetrics, 1918, xxvi, 29) has stated, is a Pneumococcus IV. It is at least suggestive that a pneumococcus, usually a IV, was recovered from either a pre-operative or a post-operative sputum specimen in ninety-seven of the one hundred and thirty cases studied, i.e., 74.6 per cent, while a pneumococcus was recovered from the sputum of only 32.2 per cent of all the surgical cases examined before operation, and of 50 per cent of the cases that subsequently developed post-operative pneumonia. PROCEEDINGS 331 DISCUSSION Alfred F. Hess: Did pneumonia develop among cases where the pneu- mococcus was absent from the pre-operative sputum; if so, how many were there? In what percentage of cases was there a discrepancy in the type of pneumococcus where pneumonia developed? John A. Kolmer: How many patients, with pneumococci in the blood stream, died? How many showing a precipitin reaction died? I am under the impression that persons showing the precipitin reaction have rather a bad prognosis. Miriam P. Olmstead: In reply to Dr. Hess I would say that pneu- monia developed in 44 cases from whose pre-operative sputum a pneumococcus was not recovered. In 14.9 per cent of the cases studied, there was discrepancy in the pre- and post-operative sputum findings. Type IV was present in 8 of these but the strains recovered were found to belong to different groups. It is probable that in at least some cases both strains were present in the sputum at the same time, but that, owing to the similarity of the colonies, only one strain was isolated. In reply to Dr. Kolmer, I would say that 2 patients with*pneumococci in the blood stream died; one had a type I, the other type III organism. There were no fatal cases among those whose urine gave a precipitin reaction. Augustus B. Wadsworth: I am greatly interested in this work. The fact that pneumonia and not a general pneumococcemia without local lesions developed in man is an indication of some degree of insusceptibil- ity or immunity against pneumococcus infection. In my experiments on rabbits pneumonic lesions developed only when the animal was partially immunized. Forthermore, the prevalence of pneumococcus infection of the upper respiratory tract suggests that different individuals acquire varying degrees of immunity from time to time. It is thus generally assumed that man possesses varying degrees of pneumococcus immunity, and the demonstration of the specific reactions of pneumococ- cus immunity in the serum of the normal healthy human subject is of interest. 6. Ture Active IMMUNIZATION AGAINST PNEUMONIA R. Kohn. 7. PrRopucTIOoN oF PNEUMOCOCCUS ANTISERUM AND THE CORRE- SPONDING CURVES OBTAINED BY PROTECTION AND AGGLU- TINATION TESTS G. Benjamin White: On account of the unusual demand for anti- pneumococcus serum every effort had been put forth to speed up and increase the production of this serum. The problem had three phases; first, to produce a potent serum; second, to bring the horses to produc- tion in the shortest possible time; and, third, to carry out the immuniza- tions with the least detriment to the horses. The basic plan of immuni- 332 PROCEEDINGS zation originated by Cole and his co-workers was followed. The first two phases have been studied with success and experiments on the third phase are now under way. The best type of horse is the medium heavy animal of the draft type with considerable spirit. A single strain of the type I organism appears to be sufficient. It is doubtful whether it is necessary to have a strain of high virulence. The stock cultures are carried on blood agar while the cultures used for injections are grown for sixteen to eighteen hours in meat infusion broth with a reaction of about plus 0.8. The cultures are centrifuged, the sediments are not washed but are emulsified in physiological salt solu- tion and injected into the horses intravenously immediately after prep- aration. Where killed cultures are to be given, the emulsions are devitalized by heating for three-quarters of an hour at 56°C. Various plans of immunization were tried and that suggested by Cole was found to give excellent results. A slight modification of this brought a horse to production in the record time of twenty-eight days. The dosage is regulated by the temperature reaction following the injection of the day before. When the horses have reached the stage of production, injections of live pneumococci are given once each week and the horses are bled six days after every second injection, 7500 ce. being taken at a time. No parallelism was found between the agglutinative and the pro- tective titers and the former is found to be less stable than the latter. All serum produced is released for distribution on its protective titer regardless of its agglutinative power. DISCUSSION ,. A. Parker Hitchens: I congratulate Dr. White on the results obtained in the rapid immunization of the horses. This is an especially impor- tant point at this time. In regard to bleeding, what was the plan of bleeding as regards the time of injections? What was the interval between the injections? John F. Anderson: We have been using a method of six daily injec- tions, in three courses, with seven-day intervals. The horse can be bled in thirty-two days. A standard has been made to protect against 0.1 cc., 0.2 ec. and 0.3 cc. of culture. George W. McCoy: I am glad to hear Dr. White bring out the point of the unreliability of the agglutinating power as compared with the protective value. We have had sera sent back marked N.G. because they did not agglutinate type I pneumococci although they possessed a high protective power. All commercial sera, before being sent out are tested by the hygienic laboratory. Augustus B. Wadsworth: Had the high protective serum any agglu- tinating power at all? George W. McCoy: It had no agglutinating power at all. Augustus B. Wadsworth, Was it controlled? George W. McCoy: Yes. PROCEEDINGS 300 _ William H. Park: In the South of France there was a great deal of pneumonia among the African troops. At one camp the men were vaccinated by Borrel with cultures from casesof pneumonia in the camp. In 18 days the pneumonia cases ceased to develop. The men in a distant camp were given the same vaccine, but the results were nega- tive. The supposition was that the pneumonias in the second camp was due to a different type. The cessation of cases may have been a coincidence, but it is striking that the cessation of the epidemic in the camp followed the use of the vaccine. G. Benjamin White: The horses are bled six days after the second injection. We have had four interesting instances of horses dying from pneu- mococcus infection during immunization. Three horses died from lobar pneumonia due to type I and at the time of death their serum showed both good protective and agglutinative power. Another horse died of a pneumococcus (type I) endocarditis with pneumococcus bacteremia during the course of immunization. The organism isolated from the blood was found to be serum-fast. A similar paradox was observed in the case of a horse immunized against paratyphoid B. Three weeks after the end of the course of immunization the animal was found to have a bacteremia and from the blood paratyphoid bacilli, the B type, were isolated. The organism was found to have acquired no serum-fastness. The serum of this horse at the time of death agglutinated both the hemotogenous strain and the stock strain in a dilution of 1: 5,000. 8. A Rapip StmuepLeE METHOD FOR THE DETERMINATION OF TYPE OF PNEtmMococcus IN SputuM OF LOBAR PNEUMONIA Charles Krumwiede, Jr. DISCUSSION O. W. H. Mitchell: I have gotten the same results by the method of extracting with sand. I am convinced as to the specificity of the reaction. The rapidity and simplicity of the method makes it an excellent one. Our sera are furnished by the N. Y. State laboratory. The tests have been checked by Dr. Wadsworth and invariably when the sputum has reacted definitely my report has been corroborated. Young rats have been found easier to procure than mice for the pneumonia work. Children keep them as pets and a few inquiries generally result in getting as many as are necessary. Half grown rats are preferable to mice. The peritoneal cavity is larger. 9. ON THE INFLUENCE EXERTED BY SALTS ON THE ELECTRICAL RESISTANCE AND PERMEABILITY OF TISSUFS B. H. A. Clowes: I have recently demonstrated by electrical conduc- tivity experiments that tumor tissues are more permeable than normal 334 PROCEEDINGS tissues in both plants and animals. I have also produced artificial membranes made by saturating filter paper with emulsions of oil in soap which exhibited variations in electrical conductivity when exposed to various antagonistic salts similar to those exhibited by laminaria and other marine organisms experimented with by Osterhout. Both laminaria tissue and emulsion membranes when exposed to the influence of NaCl exhibit a rise in permeability. If subsequently transferred to CaCk for a short period they exhibit a fall in permeability, and alternating variations in permeability within certain well defined limits can be effected in both cases by alternating treatments with NaCl and CaCh. These experiments are paralleled by experiments on the surface tension of soap films previously reported. If an aqueous soap or NaOH solution is allowed to flow from a capillary pipette through olive oil a given number of drops is obtained. If NaCl is added to the solution the number of drops is greatly increased.. If CaCl, is added to the solution the number of drops is diminished but if NaCl is mixed with CaCk in certain balanced ratios in which they oecur in the blood, sea-water, etc. the number of drops approximates that given by the original soap solution. These effects are all attributable to the influence of electrolytes on the state of dispersion and conse- quently the permeability of interfacial soap films. A practical demon- stration as to how the electrolytes in question may control the per- meability of emulsions is obtained by shaking a suitable emulsion of oil dispersed in water by means of soap with increasing proportions of CaCl. The conductivity which serves as an index of permeability remains approximately constant up to a critical point at which the emulsion of oil in water is converted into one of water in oil. At this point the resistance rises to an enormous extent owing to the transforma- tion from an emulsion which is permeable to water to one which is impermeable. NaCl, alkalis, etc. exert an effect the reverse of that of CaCl, promoting the permeability of emulsions and also of tissues. A further proof that these experiments on conductivity soap films and emulsions actually afford an index of the permeability to water and water-borne substances is afforded by introducing layers of a suitably constituted emulsion into long glass tubes, supported by filter paper and tightly fitting rubber tubes and passing various solutions through this emulsion diaphragm and determining the.rate of flow. Distilled water, sea-water and properly balanced mixtures of NaC] and CaCk flow through at nearly the same rate of speed. NaCl flows a great deal faster than the balanced solution and CaCl, considerably slower, the relative rate of flow corresponding remarkably with the number of drops obtained in the surface tension experiments. From the above experiments it appears probable that the mechanism controlling the permeability of the protoplasm is dependent upon an extremely delicately balanced emulsion of soaps, lipoids and fats and that proteins simply afford the mesh or net-work in the capillary spaces of which the influence exerted by electrolytes on the permeability of emulsions would be accentuated. Pathological changes are frequently PROCEEDINGS 335 ' attributable to disturbances in the balance of soap and fats in emulsions. For example, fatty degeneration appears to be simply the aggregation of fatty globules under the influence of surface tension changes to a point at which they become readily visible. Anaphylaxis and sensitization similarly appear to depend upon changes in permeability. A point particularly to emphasize is that the permeability of any given structure is not dependent simply upon the size of the pores of the filter but upon substances like soaps which lower the surface tension in the capillary spaces exerting an effect analogous to that of a lubricant. DISCUSSION James Ewing. Yesterday after listening to the philosophical pres- entation of Dr. Clowes I suggested that it would be well for the speaker to attempt an interpretation of his observations in terms intelligible to the pathologist and the immunologist. It has seemed to me that we were on the verge of seeing a great light, which never dawned. One wishes to know how physical systems have a direct bearing on patho- logical conditions such as fatty degeneration. Iam not prepared to admit that the fatty changes are as simple as the interactions of these systems of emulsions. This line of work corresponds to that done by Novy. The phenomena of anaphylaxis will probably be found to fall in line with all of these observations. I feel that I can congratulate Dr. Clowes on having made his experiments intelligible—to the pathologist. Henry Ulrich: I feel that surface tension’ is a vital factor in the formation of spores. B. subtilis will not grow on media with lowered surface tension. B. anthracis entirely loses its pathogenicity when grown on media with lowered surface tension. It will then produce no symptoms whatever. When blood serum is added to culture media, the tension is lowered. The tension has to be suited to the growth of the organism. This work of Dr. Clowes is very suggestive. J. Bronfenbrenner: In my opinion one can not emphasize too much the usefulness of physical methods in the study of biologic phenomena. Already several years ago Ascoli noticed the changes of surface tension in the mixtures of antibody-containing sera with suitable antigens, and he proposed to use the measurements of surface tension of such mixtures for diagnostic purposes. Though his results were not very sharp, there is no doubt that the application of more recent methods of measurement of surface tension will corroborate his conclusions. In my own experi- ence in the study of various questions in immunity, I was amply convinced of the necessity of applying physico-chemical methods to the study of these problems. In my study of the Abderhalden reaction, I noticed that the surface tension of the serum undergoing autodigestion always decreased, and vice versa, whenever the surface tension of serum was diminished, it underwent autodigestion. In collaboration with Dr. Fleisher I also noticed that the refractive index of serum increased during the process of autodigestion, thus indicating that the dispersion of the colloidal particles of the serum increases during this process. It 336 PROCEEDINGS is more than likely that the mechanism of the anaphylactic shock, or rather the nature of the cellular response to the humoral reaction in anaphylaxis, may be largely a surface reaction, affecting the permeability of vital cells, or of their electro-conductivity. There is also a great deal of evidence that the whole question of ferment action may be bound up with the question of colloidal dispersion. In general, the phenomena to which Dr. Clowes has called our attention are of fundamental impor- tance in the study of various biological problems. G. H. A. Clowes: I feel immensely interested in the observation about anthrax spores. That corresponds with my own findings. Any- one who has studied amoeboid movement, spore formation karyokinesis, budding, ete., side by side with emulsions will be convinced as to the importance of surface tension. I have produced from emulsions objects that looked exactly like leucocytes and I have shown them to pathol- ogists who thought they were leucocytes. Dr. Bronfenbrenner has mentioned the myostagmin reaction, which was Ascoli’s work. He let drops fall into the air instead of into oil. Ferment action appears to be definitely due to surface tension changes. Perfect contact must first be effected, then dispersion or aggregation may occur. ; 10. EXPERIMENTS UPon THE CHEMOTHERAPY AND CHEMOSEROTHERAPY or PNEumMococcus INFECTION John A. Kolmer, Edward Steinfield and Charles Weiss. 11. SruDIES ON THE ToxIcITy oF PNEUMONIC LUNGS John A. Kolmer, Charles Weiss and Edward Steinfield. DISCUSSION A. Parker Hitchens: Has Dr. Kolmer taken into consideration the great increase of toxicity in tissues that undergo autolysis? A piece of liver digested in sterile salt solution for twenty-four hours becomes extremely toxic. Ralph R. Mellon: Do these extracts of empyema fluid bear any re- lation to the aggressins of the disease? Charles Weiss: It is very difficult to decide whether or not the toxi- city of the pneumonic exudate is due to the fact that the tissue was undergoing autolysis. The hemolytic properties of the pneumonic exudate may not have been specific for pneumonia. They have been attributed to various fatty acids. Specific anaphylactic results were obtained by sensitizing guinea-pigs to normal and pneumonic exudates. The latter are rich in fibrin and the react’on may have been specific to the fibrin. But the experiments indicate that at least part of the toxicity and hemolytic activity of the exudate was due to the presence of toxins liberated from the pneumococci. PROCEEDINGS Bor James Ewing: This work is an important step in the right direction. I do not feel competent to offer any criticism, but I wonder why the work of immunologists turn always to specific toxins and away from products of tissue changes which the pathologist is interested in. I believe that immunologists have been held back by exclusive attention to Ehrlich’s theories. G. H. A. Clowes: Were these substances soluble in fats and lipoids or in water? This is of vital importance in determining théir char- acteristics. John A. Kolmer: We have no means of absolutely controlling this work in relation to aggressins. We tried to ascertain whether this toxic substance would retard phagocytosis of pneumococci in vitro and we found that it does so to a slight extent. Charles Weiss: I have isolated albumin, globulin, uric acid and lipoid substances from the extracts, but this work is still under way. H.G. Wells: There is an error in the last remark. Uric acid is formed only in the liver. Xanthin is probably what the speaker meant. John A. Kolmer: Mr. Weiss referred to the literature on the subject. 12. THr PROPERTIES OF PNEUMOTOXIN AND ITS PROBABLE ROLE IN THE PaTHOLOGY oF LOBAR PNEUMONIA Charles Weiss and John A. Kolmer. FINAL SESSION 1. THe EXAMINATION OF THE BLOOD PRELIMINARY TO THE OPERATION OF BLoop TRANSFUSION Arthur F. Coca (see this volume, page 93). DISCUSSION John A. Kolmer: I desire to describe briefly a microscopic method which I have employed during the past year with very satisfactory results and which is fashioned after a microscopic technic described by Lee. A small amount of blood is obtained from the finger of the patient and each of the donors in small and separate test tubes to supply a few drop of serum; also a few drops from each in small test tubes containing 1 ce. of a 1 per cent sodium citrate salt solution to supply a suspension of cells. The sera are separated and the cells washed once with the centri- fuge, although the latter is not absolutely necessary. In setting up the tests, hanging drop slides are employed as in the Widal reaction. Ona series or cover glasses, two loopsful of the recipient’s serum is mixed with one loopful of corpuscle suspension from each of the donors; in a second series, one loopful of the recipient’s suspension of cells is mixed with two loopsful of serum from each donor. The preparations are examined microscopically with the low power objective, fifteen minutes later. ~ 338 PROCEEDINGS Controls are included with the cell suspension of the recipient and each donor. Agglutination is well marked when it occurs and easily read. With this method no attempt has been made to group the bloods but it is extremely simple and it has been found very practical. I wish to ask Dr. Coca’s opinion of the practical value of these tests inasmuch as surgeons occasionally express themselves as believing that the tests are not essential to successful transfusion, although he personally does not share this view according to experience. Arthur F. Coca: The choice of methods may depend upon what ap- paratus is at hand. The procedure that I have described is quite as easy as that of blood counting and all of the apparatus required for it is available in any laboratory in which blood counting is done. Dr. Kolmer’s method requires much more time, much more blood and more apparatus than the one that I have described. His second series of mixtures is unnecessary to the purpose in view. So far as I am aware the deaths that have occurred as a result of transfusion have happened when the compatibility test had not been made. 2. THe. ISOLATION, PURIFICATION AND CONCENTRATION OF IMMUNE HEMOLYSIN M. Kosakai (see this volume, page 109). DISCUSSION William H. Park: Has Dr. Kosakai been able to separate antibodies from the bacteria? John A. Kolmer: I should like to ask Dr. Kosakai whether it is pos- sible, with his method, to separate the hemagglutinins from the hemoly- sins. It is desirable, particularly, where the anti-human hemolytic system is to be used, to produce a serum preparation that is free from agglutinins. Was Dr. Kosakai’s final solution free from agglutmin? M. Kosakai: I have not attempted to separate anti-bacterial anti- bodies from the bacteria with my method. I have not been able to separate the hemagglutinins from the hemolysins. 3. A Rapip StwetE METHOD FOR THE EXTRACTION OF PRECIPITIN ANTIGEN FROM BACTERIA Charles Krumwiede, Jr. (see this volume, page 1). DISCUSSION George H. Smith: Can this “precipitin antigen” be employed also for immunization? Charles Krumwiede: I have not tried to produce an immune serum with an antigen prepared in this way from bacteria. An extract made in this manner from meat did not stimulate the production of antibodies. PROCEEDINGS 339 4. A Metruop oF PREPARING BACTERIAL ANTIGENS J.C. Small (see this volume, page 413). DISCUSSION John A. Kolmer: Several years ago I experimented with bacterial antigen prepared by different methods, working with the typhoid colon group. ‘The results of the complement-fixation tests were similar to those of Dr. Small but with that method I was not able to differenti- ate between paratyphoid A and paratyphoid B. Were rabbits used for the production of the immune sera and if so was the serum used active or was it heated? In previous work I found that the serum of some rabbits when heated at 56°C. for one-half hour develops the prop- erty of fixing complement in a nonspecific manner; for this reason animals should be tested in a preliminary way before immunization is begun and those that do not show this phenomenon may be selected. Otherwise the nonspecific fixation may be avoided by heating the serum at 62°C. instead of 56°C. for thirty minutes as was done by Meyer and Boermer with the serum of mules. Charles Weiss: Has the speaker found that the results are different after the removal of the lipoids from what they were without this pro- cedure? In regard to the heating of antigen it has been found in work- ing with streptococcus antigen that heating destroys the anticom- plementary qualities but at the same time it weakens the antigenic properties of the preparation. G. H. A. Clowes: In reference to the question of heating is it the expe- rience of the members present that the quantitative activity of the serum is diminished by heating at 62°C.? John A. Kolmer: The heating of immune sera at 62°C. may cause slight deterioration of the specific antibodies present. The deterio- ration is, however, difficult to measure. Investigations have shown that nonspecific inhibition of complement is especially likely with bacterial antigens. M.A. Wilson: Our experience has been that there is no deterioration of the antigenic preparation if it is heated. Hassow von Wedel: I have used tubercle bacilli antigen that has been heated at intervals five or six times without noting any difference in its antigenic value. 5. A CONTRIBUTION TO THE StruDy. OF COMPLEMENT FIXATION IN TUBERCULOSIS M. A. Wilson (see this volume, page 345). 6. A CONTRIBUTION TO THE StTuDY OF COMPLEMENT FIXATION IN TUBERCULOSIS Hassow von Wedel (see this volume, page 35). 340 PROCEEDINGS DISCUSSION Paul Lewis: I have been much interested and instructed by these papers. The results may be valuable from a diagnostic point of view. I ean add some data from my own experience from work done within the last year. At first I used bacterial suspensions or the autolysate which latter I found, lost rapidly in anticomplementary effect. Test- ing with the same serum day after day, I found that one frequently does not get the same result on two days running. I think this may have been because the reaction between antiserum and antigen in those cases was a weak one and that the readings were only permanently positive with the strongest sera. Experiment showed that by prolong- ing the fixation period more durable reactions could be obtained. Four hours proved to be the maximum period which it was practicable to use owing to irregular deterioration of the complement in longer exposures. Using a four-hour incubation period I have titrated numbers of sera to determine the amount of complement fixed. In certain instances this may reach 2 cc. of the usual 1 + 9 dilution of guinea-pig’s com- plement. It has been found that the tuberculous sera can be well pre- served by mixing with an equal quantity of neutral glycerme. Pooled serum from a number of cases giving a strong deviation, thus preserved has been used over a period of six months to study the properties of various antigens. G. H. A. Clowes: In working on cancer cases and also on syphiltic cases eight years ago I attributed the development of an increase in complement fixing power in serum to changes in the colloidal particles, due to their being held in the ice-box for some time. I wish to ask what was the temperature of the ice-box used. Was the serum frozen? Hassow von Wedel: I am not able to answer Dr. Clowes’s question. The ice-box probably had a fluctuating temperature. G. H. A. Clowes: I have seen complement deviation variations as high as 10 to 1 It is easy to vary the surface of a particle by changes in physical conditions or by variations of the hydrogen ion content. This can frequently be determined by the ultramicroscope which shows variations in the size of the particles. William H. Park: I feel that, at present, this test should be used for primary cases; later we may be in position to use it on a diagnostic basis. John A. Kolmer: I wish to ask whether the peculiar property of human serum of developing fixation powers after standing was found also for the Wassermann and gonococcus-fixation tests. Miriam Olmstead: Did the guinea-pigs that gave a fixable comple- ment for the tuberculosis antigen also give a fixable complement for the gonococcus antigen? How many specimens giving positive gonococcus fixation were used for the tuberculosis tests? M.A. Wilson: We killed, at one time, 10 guinea-pigs and we found that -of the sera all were fixable in the meningococcus fixation; four were fixable in the gonococcus fixation and only one was fixable in the tuberculosis fixation. PROCEEDINGS 341 A Srupy oF CONTROLLED POSTMORTEM WASSERMANN REACTIONS: A SUPPLEMENTARY REPORT ON 400 CASES Stuart Graves: 1. Post mortem Wassermann tests confirmed ante- mortem Wassermann tests in 97 per cent of 68 controlled cases. A four plus positive reaction in a specimen obtained sixty hours post oe as a four plus reaction obtained ante mortem in the same one with anatomie and clinical evidence of syphilis. A a eee in a specimen taken twenty-two hours post mortem confirmed a negative ante mortem reaction. 2. In 91.2 per cent of the cases showing anatomic lesions of syphilis and presenting evidence of syphilis in their histories the sera post mortem gave positive Wassermann reactions. 3. The fact that only 2.5 per cent of the sera were anticomplementary or otherwise unfit for use compares favorably with the fact that 1.14 per cent of 6000 ante mortem specimens were similarly unfit. 4. Only 2.6 per cent of 378 cases showing anatomic evidence of syphi- lis gave negative reactions. 5. The reactions conformed to the anatomic and historic evidence in 304 of 378 cases or 80.4 per cent, which is considerably lower than it would have been if satisfactory histories and physical examinations had = recorded in Class V. There is no logical reason for supposing that acute infections or get tumors cause positive Wassermann reactions. 7. The positive reaction appeared in 2.7 times as Many negroes as whites, in 1.7 times as many males as females and in only 11 white fe- males or 6.5 per cent. 8. The Wassermann reaction, carried out on post-mortem blood according to the methods followed in this investigation, is practically as reliable a test for syphilis as when performed with ante-mortem specimens and is of great value in pathological anatomy and in medico- legal cases. DISCUSSION John A. Kolmer: I wish to endorse all that Dr. Graves has said in- asmuch as my own results coincide entirely with his experience. I am familiar with the published reports just described tending to discredit the practical value of the Wassermann reaction and I think it is parti- cularly unfortunate that reports published with this object in view should reach the practitioner of medicine. No one that has dealt with patho- logical findings would deny that a pathologist cannot exclude syphilis on the basis of autopsy findings alone. Spirochaetes might not produce much tissue reaction and yet suffice to produce antibodies, which would be indicated by the Wassermann reaction. In the last four or five years much has been learned of the technic of the Wassermann reaction rendering it a test of great diagnostic value. While we must welcome attempts to point out sources of error, criticisms should be carefully controlled. 342 PROCEEDINGS 8. OBSERVATIONS ON THE INTRASPINOUS AUTO-SALVARSANIZED SERUM THERAPY OF CEREBROSPINAL SYPHILIS Benjamin A. Thomas. 9. EXPERIMENTS UPON THE PASSIVE TRANSFER OF ANTIBODIES TO THE CEREBROSPINAL FLUID John A. Kolmer and Shigeki Sekiguch (see this volume, page 101.) DISCUSSION A. Parker Hitchens: The question whether the antibodies find their way from the blood to the spinal fluid is mportant, especially in cerebro- spinal fever. In severe cases of epidemic meningitis in the army camps the antimeningitis serum has been given by intravenous as well as by intraspinous injection. The results of this treatment are en- couraging. Dr. Kolmer’s work would show that specific treatment by the intravenous route should be of value in meningeal infection. J. Bronfenbrenner: It is still not settled definitely whether the anti- bodies normally pass directly from the blood into the spinal fluid. This passage is even more questionable in such pathological conditions in which the pressure in the spinal canal is much greater than in the blood vessels. The fact that the introduction of antimeningitis serum into the circulation seems to produce clinical improvement could possibly be due rather to another mechanism than the direct passage of antibodies from the blood into the spinal canal. Recent investigations of English workers lead them to the conclusion that the meningococcus produces a soluble toxin. Such a toxin may be forced out of the spinal canal into the circulation on account of the pressure in the spinal canal, and so cause a certain amount of general toxemia aggravating the specific symptoms of meningitis. The therapeutic serum may contain a cer- tain amount of antitoxin and to that extent may improve the clinical condition of the patient when introduced intravenously. A. Parker Hitchens: About every three hours during the severe stage of the disease the spinal fluid is removed, and with the rapid filling up of the canal again there may be some effect. John A. Kolmer: Our paper simply deals with the passage of anti- body from the blood to the spinal fluid under normal conditions. In the experimental animals it was found necessary to have a large amount of antibody in the circulation. 10. VacctnE DosaGE Joseph Head. 11. Tae Vaccine TREATMENT OF ACNE, WITH SPECIAL REFERENCE TO THE R6LE OF BACILLUS COLI Albert Strickler and Jay F. Schamberg: Thousands of cases of this disease never come to the physician at all; it is only when suffering from PROCEEDINGS 343 severe forms with great disfigurement, that the patients seek advice. The lesions are chiefly in the sebaceous glands, which are very active at puberty. Puberty is, in fact, the primary predisposing cause. Anemia and constipation are found to be pretty constant accompaniments of acne. The condition often develops after typhoid fever, and intestinal intoxication evidently plays a réle in the etiology. In the treatment of acne by vaccines, it was found that there is a complement fixation in 63 per cent of the cases with an antigen prepared from a colon bacillus isolated from the intestinal tract of the patient. The fixation is higher when the antigen is from the patient. It was resolved to treat 50 cases solely with vaccines prepared from an autogenous colon bacillus. These cases were controlled by cases treated with other methods such as vaccines from other germs, therapeutic and hygienic measures. The B. coli vaccines were found to possess better curative effects than any other mode of treatment. DISCUSSiON Jay F. Schamberg: I would like to emphasize one or two points. Vaccines prepared from B. acne and from staphylococci have been in use for many years, and In some cases they have given brilliant results; in others they have failed to respond. The present experiments were carried out in a hospital clinic upon a large number of patients. 63 per cent gave positive complement fixation with the strains of B. coli used for the vaccines. A large percentage of patients responded to a remarkable degree. It is likely that at puberty, when there is great developmental activity, there is liability to infection from the intestinal tract. The activity of the intestinal organisms may then produce noxious effects. The complement-fixation tests would seem to merimin- ate especially the colon bacillus. A. Parker Hitchins: It is animportant point to find out which strain of B. colt is the chief factor in producing the disease. Has any work been done on the sera in this respect? Albert Strickler: In regard to the present work, I personally welcome an active case as I feel that a good deal could be done for the patient. Formerly I rather dreaded to see such patient come, as the results of treatment were so often disappointing. Jay F. Schamberg: No.attempt has been made as yet to differentiate strains. The complement-fixation test merely showed that there is greater specificity to the B. coli from active cases. 12. Lipo-VACCINES Eugene R. Whitmore and E. Fennel DISCUSSION A. Parker Hitchens: This plan for the preparation of bacterial vac- cines is likely to be of immense value. If one can give the entire treat- 344 PROCEEDINGS ment in one day, the economic value of the procedure is obvious. At Camp Beauregard I saw a detachment injected with meningitis lipo- vaccines; the reactions were very severe, as were also those from triple typhoid vaccine. The meningitis vaccine caused severe headache and insomnia. All the symptoms disappeared the next day. I feel great confidence in the outcome of the work of Colonel Whitmore and Lieuten- ant Fennel. Lieutenant Fennel: There was a very severe reaction to the first vaccine we used. Since the improvement in technic the severity of the reaction was much lessened. 13. A Srupy OF THE IMMUNIZING PROPERTIES OF BACTERIAL VACCINES PREPARED AFTER VARIOUS METHODS M.W. Perry and Sara Levy. DISCUSSION Lieutenant Fennel: I wish to ask Mr. Perry whether he titrated the agelutinable culture by the standard set by Oxford. If so, the results will be of considerable value for the army work. John A. Kolmer: The culture was secured from a German university sixteen years ago and was standardized for agglutination work. It is highly susceptible to agglutination. It is not standardized by the Oxford method. A CONTRIBUTION TO THE STUDY OF THE COMPLE- MENT FIXATION REACTION IN TUBERCULOSIS. M. A. WILSON From the Research Laboratory of the Department of Health, New York City Received for publication May 14, 1918 I. ON THE STANDARDIZATION OF COMPLEMENT In our study of complement fixation in tuberculosis we have found a point of technique that has increased the efficiency of the test. It has to do with the standardization of the guinea- pig’s serum to determine the value of the complement. In this preliminary report, we shall describe the method or standardiz- ing the complement, the preparation of our tuberculosis antigen, and the diagnostic test, with such results as we have obtained thus far in our study. Technique All reagents are used in one tenth the classic Wassermann vol- umes. Fixation period, one hour, 37°C. Serum. The patient’s serum is inactivated for thirty minutes at 56°C. 0.02 ce. and 0.01 ec. of the undiluted serum are used in the test, and 0.04 cc. for control of anticomplementary action. Antigen. Two antigens have been used. One was made from 12 stock cultures of human tubercle bacilli, the other from strain 305 (used for tuberculin production.) Those antigens are suspensions in 0.9 per cent saline of dried bacilli, from which all constituents soluble in aleohol and ether have been removed. The bacilli were grown in glycerin-broth. The polyvalent anti- gen cultures were grown for three weeks, and the monovaient ones for three months. The broth cultures were killed by heating them in the Arnold sterilizer for one hour. The cultures were 1 Preliminary report. 345 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 5 346 M. A. WILSON then filtered through filter paper. The filtrate was discarded, and the residue was placed in absolute alcohol, in the proportion of one volume of residue to 10 volumes of alcohol. This mixture was shaken thoroughly by hand, and was placed in the ice-box for two weeks. It was then filtered through paper and the filtrate was discarded: The residue was washed in absolute alcohol and the sediment obtained by centrifugalization was washed in ether. After a further centrifugation the ether was discarded and the centrifuge tube containing the residue was plugged and placed in the dark at room temperature over night. By this simple procedure the residue was dried within twenty-four hours. The dried powder was emulsified in a large mortar with 0.9 per cent saline in the proportion of 1 gram of powder in 200 ce. of saline. This gave a concentrated emulsion convenient for storing as a stock antigen. The emulsion was heated for one hour at 80°C. The antigen was now ready for use, and it was standardized to be used in such a dilution that 0.1 ec. contained two standard fixation units and one fourth, or less, of the anti- complementary dose. The unit was determined by titrating varying amounts of the antigen with 0.01 cc. of a known positive tuberculosis serum, and two hemolytic units? of a complement known to be potent for tuberculosis fixation. The standard dilution of the two antigens employed is 1:50. This makes a final dilution of dried bacilli 1: 10,000. These antigens are not anticomplementary in the amount used in the test. They have given uniform and constant fixation reactions. The tests reported will show that they are specific and stable. They were made ten months ago and are perfectly efficient today. 2 As determined with the use of one-tenth of the standard amount (i.e., 0.1 ce. instead of 1 ec.) of 5 per cent sensitized sheep’s corpuscles. 3 In some instances a sample of antigen that has been standing in the ice-box for some time has been found to have increased in its anticomplementary action. This change is not accompanied by any deterioration of the ‘‘antigenic’’ property of the preparation and, as we have found, it can be removed by heating the diluted preparation for one-half hour at 56°C. The change is not of frequent occurrence; however, as a routine precaution, we heat all of our diluted antigen before using it in the tests. COMPLEMENT FIXATION REACTION IN TUBERCULOSIS 347 Complement. Guinea-pig’s serum, twenty-four or forty-eight hours old; pooled from six to ten pigs. Before the pooling, the serum from each pig is tested for its hemolytic strength, for anti- sheep amboceptor, for anticomplementary reaction with a hete- rologous serum and for fixability with the combination of tuber- culosis antigen and tuberculosis serum. This last test we emphasize as an essential, if uniform results are to be obtained with different lots of complement; it has proven beyond a doubt that although a guinea-pig’s serum may react perfectly in all other respects it may fail to be fixed by tuberculosis antigen and serum. Of the pigs we have tested 64 per cent failed in fixability, while they were perfectly good in other respects. If in a pool of six or ten complements, there are several strongly fixable, the presence of the negative complements in the pool may not appreciably affect the test; on the other hand, if most of the complements are negative the fixability of the pool will not serve to give a true reaction with the patient’s serum. In such a case a four plus reaction may drop to a two or three plus reaction and a two or one plus reaction may become negative. Tables 1 and 2 show the variation in fixability of the guinea- pig’s serum. Indicator for the fixation reaction. 0.1 ec. of a 5 per cent suspension of sheep cells sensitized with three standard units of antisheep amboceptor. Controls for diagnostic fixation reaction: Antigen—for anticomplementary reaction. for specificity. for potency. Serum—for anticomplementary reaction. for specificity. for natural antisheep amboceptor. Complement—for stability (system control). Sensitized cells—for stability (reading control). 348 M. A. WILSON TABLE 1 OONooPr wW ND | pra No. DATE BLED November November November November November November November November November November November November 13 13 13 13 13 13 13 13 13 13 13 13 COMPLEMENT FIXATION REACTION Tuberculosis No fixation No fixation Complete fixation No fixation Weak fixation No fixation No fixation Weak fixation No fixation Weak fixation Complete fixation Complete fixation Meningococcus Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation Gonococcus Complete fixation Complete fixation Complete fixation Complete fixation Complete fixation No fixation No fixation Complete fixation Complete fixation No fixation Complete fixation Weak fixation — TABLE 2 Showing the number of guinea-pigs serums efficient for tuberculosis complement fixation NUMBER OF PIGS BLED 10 3 11 8 129 NUMBER EFFICIENT FOR COMPLEMENT FIXATION Tuberculosis mPmeowW or POW WH WwW OF WwW re o> (or) Meningococcus Gonococcus WHDDOPADPORP KF wWOAAWwWHFE (o/) ie) aimboceptor. The pooled serum is titrated with cells sensitized with 3 standard units of The reaction is read at the end of fifteen minutes. units of the pooled complement are used. Two hemolytic | COMPLEMENT FIXATION REACTION IN TUBERCULOSIS 349 Results The results of tests of serums from 344 cases are given in table 3. TABLE 3 CASES : 5 __ | POLIOMY- : | PULMONARY BONE AND TUBERCU ELITIS HAVING NO ACTIVE JOINT LAR SPINAL SYMPTOMS DISEASES GLANDS FLUIDS OF TUBER- CULOSIS per cent per cent per cent per cent per cent Strongly positive.............. 67 12 35 0 0 Wed POSItIVe......56n0.5-5 4 25 10 23 0 0 "EDU Oe Ss oe GU ane ree 8 78 42 100 100 Conclusions 1. Not all guinea-pigs’ serums are efficient for tuberculosis complement fixation. 2. The serum from each guinea-pig should be tested for fixability with tuberculosis antigen plus tuberculosis serum before pooling the complement for diagnostic tests. II. OBSERVATION OF THE VON WEDEL REACTION Dr. von Wedel, working independently in our laboratory, found that some serums from active tuberculosis cases gave a negative complement fixation reaction when the test was made on the first day after bleeding, and when the same specimens were tested a week later, having stood in the ice-box during the interval, they gave a positive reaction. This occurred repeatedly on later bleedings from the same patient. The complement used for all tests had been previously tested for fixability with tuberculosis antigen plus tuberculosis serum: therefore, the negative reaction in the first test was not due to an inefficient complement. Serums from non-tubercular cases gave no fixation at any period after bleeding, and this fact rules out the question of non-specificity of the later positive reaction following the early negative phase. The controls for anticomplementary reaction in the patient’s serum and in the antigen were all completely hemolyzed. 350 M. A. WILSON This early negative phase was not demonstrated in the serums from all tubercular cases; but the percentage was so large as to be significant. Having personally observed the accuracy of Dr. von Wedel’s technique, and the many repeated tests he made to discover the possibility of an error, I was convinced of the verity of the re- action and of the necessity for making the later test before the tuberculosis antibody content of all serums can be determined. All tubercular serums tested, throughout the remainder of my study, will be given the early and later tests. A CONTRIBUTION TQ THE STUDY OF THE COMPLE- MENT FIXATION REACTION FOR TUBERCULOSIS HASSOW VON WEDEL From the Bacteriological Department of the New York University and the Bellevue Medical College Received for publication May 14, 1918 The complement fixation reaction for tuberculosis has occupied the attention of many investigators for the past seventeen years, and it has been studied mainly from the standpoint of its pos- sible value to clinical medicine in the diagnosis and prognosis of this disease. The methods commonly employed for diagnosing tuberculosis leave much to be desired and there is little doubt that many cases escape detection until marked damage has been done. A test, therefore, that will give a sure and early diagnosis, is of the utmost value both to the patient and to the general public. The results of the early work with this complement fixation test were of such a contradictory nature that they were of little practical value. However, the reports of recent investigators seem to promise that this test will be of marked value to the clinician. The following is a brief review of some of the more important investigations made during the past few years. Widal and LeSourd (1) appear to be the first who used the comple- ment fixation reaction in attempting to arrive at a more certain method of diagnosing tuberculosis. Bordet and Gengou (2) in 1903 demon- strated the presence of antibody capable of uniting with tubercle bacilli and fixing complement in the sera of tuberculous animals. Wasser- mann and Bruck (3) in 1906 also demonstrated the presence of an antibody to tuberculin in patients treated with tuberculin. Caulifield (4), Laird (5), Hammer (6), Calmette and Massol (7) using various forms of bacillary emulsions as antigens, obtained results which were 351 BSA HASSOW VON WEDEL very inconclusive, ranging from 33 per cent to about 97 per cent of positive results in cases of tuberculosis. Much (8), using various acid- fast bacteria as antigens, with sera from tuberculous and healthy per- ‘sons, obtained fixation in 77 per cent of the healthy cases, in other words, a large number of non-specific fixations. Frazer (9), usimg various antigens found that 96.6 per cent of sera from normal individuals gave no fixation with antigens made from living bacilli, and that with this antigen 42.3 per cent of sera from tuberculous individuals gave positive results. She states that the most reliable antigen is prepared from living human bacilli, and thinks that the complement fixation test made with living bacillary antigen is of more value in the diagnosis of tuberculosis than any other reaction thus far discovered. Dudgeon, Meek and Weir (10) state that all of their cases that had been treated with tuber- culin gave positive results. Products of the bacilli were found to be very satisfactory as antigens. With an alcoholic antigen, (11) prepared from tubercle bacilli they obtained 89.3 per cent of positive results. Bron- fenbrenner (12) using an antigen of tubercle bacilli grown by Besredka on egg broth, obtained a very high percentage of positive results, but also obtained quite a large percentage of non-specific positive results and 24 positive reactions with syphilitie sera. McIntosh, Fildes and Radcliffe (13) criticized Besredka’s (14) antigen and concluded after testing many antigens, that the living bacillary emulsion was best. Inman (15) and Kuss, Leredde, and Rubenstein (16) found the antigen non-specific. Stimson (17), using a variety of antigens, reported a small number of cases with but fair results. Corper (18), in 1916 using an autolysate as antigen and also a bacillary emulsion antigen, concludes that the complement fixation test for tuberculosis is not absolute, being positive in only about 30 per cent of all clinically definite cases of tuberculosis, both active and inactive. In 1917, Corper and Sweany (19) comparing their autolysate antigen with the bacillary anti- gen of Miller, (20) concluded that there is not a great deal of difference between the results obtained with these antigens and that it is impossible by means of this test absolutely to differentiate active from inactive tuberculosis. They prefer the autolysate to the bacillary antigen. They obtained 65 non-specific cross fixations out of 92 specimens of sera that gave positive Wassermann reactions. Slack, Burns, Castleman and Bailey (21) state that it appears from their observations that the complement fixation reactions are specific for tuberculosis, and that they obtain no cross fixation with positive Wassermann sera. Of recent investigators, Craig (22) and Miller (23) obtained the best results. Craig reports the results of 166 examinations on cases of pul- COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 353 monary tuberculosis, in which he used an alcoholic extract antigen of human tubercle bacilli. He obtained 92.6 per cent of positive results in active cases, and 66.1 per cent in inactive cases. Out of 150 cases of syphilis, which were free from lesions in the lungs, none gave positive reactions. All of 100 other examinations (various diseases) gave nega- tive results. However, Lucke (24) and other investigators report that alcoholic extract antigens, as used in Craig’s work, have many times proven inpracticable and worthless, and as his description of the prep- aration of his antigen was wanting in detail, the described preparation was frequently found impossible. Miller reports 100 per cent of positive results in active cases and 100 per cent of negative of results in non- tubercular cases. A great number of various antigens have been used and a great diversity of results has been obtained, reaching from 30 per cent efficiency up to Miller’s claim of 100 per cent. There seems to be no uniformity in the findings and conclusions of any two observers; even the most promising of all these investigations, that is, Craig’s and Miller’s are of limited practical value, as serologists have had great difficulty in using their antigens, and therefore have been unable to reproduce their results. The main object of this preliminary study of the complement fixation reaction for tuberculosis was, therefore, with the use of the perfected Wilson antigen, which is easy to prepare and which can be kept for a long period of time without becoming anticomplementary, to find a method that would give specific positive results in active tubercular cases and which would not give non-specific results in negative cases. I have made 1078 complement fixation tests on 200 specimens of blood serum taken from cases with no clinical history of tuber- culosis and from patients with active, inactive and primary pulmonary tuberculosis. This study was made in conjunction with Miss Wilson in the New York City Board of Health Re- search Serological Laboratory. The cases were practically all from the Westchester County Hospital and I have the complete clinical data on all the cases. The clinical data consist of age; past and present temperature, pulse and respiration records; sputum reports and physical symptoms with clinical diagnoses. 354 “ HASSOW VON WEDEL TECHNIQUE The technique employed was similar to that originally used by Wassermann with the following modifications. At first the tests were carried out in both one-quarter and one- tenth the original Wassermann volume, but as I found no differ- ence in the results, I have since continued to use the one-tenth Wassermann volume only. Complement. The pooled blood serum from six to ten healthy guinea-pigs was used as complement; in addition, we used serum from separate guinea-pigs untested for its complement fixation value; serum from separate guinea-pigs after having been tested for complement fixation value and serum from six to eight guinea- pigs, all of which had been specially tested for complement fixa- tion value. All these complements were titrated with 2.5 per cent sheep cells, sensitized with three units of anti-sheep ambo- ceptor; the unit was recorded at the end of fifteen mimutes. Exactly 2 units were used in the regular test. i Antigen. The Wilson antigen, which was used throughout this study, is simply a suspension of tubercle bacilli killed with heat, extracted with alcohol and ether, and dried. The com- plete technique of preparing this antigen is given by Miss Wilson in her paper on the study of the complement fixation reaction for tuberculosis in the current issue of this journal. Sheep’s cells. A 5 per cent suspension of sheep cells, which had been washed five times in sterile saline was used, after having been sensitized with an equal volume of amboceptor in the water bath for half an hour. Amboceptor. Three units were used in the tests. Fixation period. After the patients’ serum, complement, anti- gen and saline were mixed, the mixtures were incubated in the water bath at 37.5°C. for one hour. The sensitized cells were then added and the reading was made in exactly fifteen minutes. Six series of tests have been made for the purpose of comparing the results after the following different methods of incubation: one hour in the water bath; two hours in the water bath; two hours in the water bath followed by two hours in the ice-box; COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 355 and four hours in the ice-box. Apparently the one hour water bath incubation gives the most uniform results. The ice-box incubation gives by far the weakest fixation. The two-hours in the water bath followed by two hours in the ice-box gives almost the same results as the two hour water bath incubation alone which gave quite a number of anti-complementary reac- tions in the tests where we doubled the regular Wassermann amount of patients’ sera. These comparisons, of course, are too few to allow a positive statement as to which is the best method of incubation. A large number of comparisons will be made in the near future to determine this question. Results were reported as plus minus if any degree of fixation was observed; 1 plus if marked fixation in the first antigen tube; 2 plus if complete fixation in first tube; 3 plus if complete fixation in first tube and marked fixation in second tube, 4 plus if com- plete fixation in both tubes. During 1913, Dr. Cyrus W. Field and the writer carried out a series of 730 Wassermann reactions in the Bellevue Hospital Laboratories, using the regular amount of serum prescribed by Wassermann and also twice, three times, four times and five times that amount. These amounts of patient’s serum were tested in all of the 730 cases, our controls being carried out with double the amounts of serum used with the antigen. Discarding all those cases that were anticomplementary in the regular Wassermann amounts, and considering only those which ordi- narily would be considered as not anticomplementary, we found that we had no anticomplementary and no non-specific reactions with double the usual amount of serum. ‘Three times the usual amount of serum gave about 1 per cent of anticomplementary reactions, four times the usual amount of serum gave about 5 per cent of anticomplementary reactions and five times the usual amount of serum gave about 25 per cent anticomplementary reactions. As these results were so favorable, and as several other investi- gators have made favorable reports on the use of larger quanti- ties of patients’ sera, I have made all my tests since January 1 with the regular Wassermann amount and with double that 356 HASSOW VON WEDEL amount of patient’s serum; that is, 0.04 cc. of serum in the first antigen tube with 0.08 cc. of serum in its control tube; 0.02 ee. of serum in the second antigen tube with 0.04 ec. of serum in its control tube, and 0.01 cc. of serum in the third antigen tube. All the specimens of sterile sera from 95 known non-tubercular cases, gave negative results with 0.04 cc. of patient’s serum in the antigen tubes. Four contaminated specimens of serum from known non-tubercular cases, gave non-specific weakly positive reactions. Forty-nine sera from cases with active tuberculosis - gave strong positive reactions with double the amount of serum and only 43 gave strong positive reactions with the regular amount of serum. Ten sera gave positive reactions of 2 plus to 4 plus when double the amount of serum was used, and only doubtful or negative results with the single amount. Of the type II cases, 46 gave positive results with double the amount and only 41 with the single amount. Of the type IV cases, 6 gave positive results with double the amount and only 2 with the regular amount. We, therefore, had an appreciable number of cases which gave definite positive results with the double amount and negative or doubtful results with the single amount of patient’s serum. The patients were bled Thursday afternoons and the sera were separated from the clots Thursday evenings or Friday mornings and inactivated Friday mornings. Each specimen of serum was tested the morning after it was removed from the patient and this serum was retested week after week for from four to five weeks with an interval of seven days between each test, together with specimens from new patients each week. We, therefore, have a record in many instances of eight weekly complement fixation tests on the same specimens of sera kept under as nearly sterile precautions as was possible in the ice chest at about 8°C. By this proceedure, I made a very interesting observation which may possibly account for some of the wide discrepancies in the various complement fixation results reported by the different workers. The complement fixation results on sera from positive cases made the first day sfter taking the specimens from the patients were in a very large percentage of cases negative or COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 357 weakly positive; while in most instances, seven days later these same sera gave a strongly positive reaction and continued to give this strongly positive reaction week after week with unvarying regularity. None of the non-tubercular sera gave a positive reaction the second, third, or fourth week after the specimen had been obtained from the patient. Of all positive sera in our series, only 12 gave a 3 or a 4 plus reaction the first day tested. After being preserved in the ice- box under sterile precautions for seven days, 49 sera gave 3 or 4 plus reactions. Of 82 sera from tubercular cases of all types that gave negative or doubtful reactions the first day, the results on the seventh day were positive in 37. The reactions apparently did not change after the sixth day. TABLE 1 Results obtained with the same specimens of serum one day, seven days, fourteen days, and twenty-one days after the blood sera were removed from the patients SERUM NO. FIRST DAY* SEVENTH DAY FOURTEENTH DAY TWENTY-FIRST DAY 13 = 4+ 4+ 4+ 25 = 4+ Pee 3+ 14 = 4+ 2+ - 61 + 4+ 4+ 4+ 67t | = = = = | * This refers to the day after the blood was removed from the patient. + The serum was obtained from a normal non-tubercular individual. I have attempted to find cut on just what day the positive result first appears in most of these delayed reactions, but hereto- fore I have been able to carry out only two series of daily titra- tions and from these I could not draw any positive conclusions. This work, however, will be continued and when our final results are published, I hope that they will include information upon this important question. The phenomenon just described apparently bears no relationship to the type of case, as I find it occurring in the old active cases, in the primary cases and in the inactive cases. In order to see whether the phenomenon was possibly due to anticomplementary reaction, I have used four times the usual amount of patient’s serum in the serum controls without, however, 358 HASSOW VON WEDEL obtaining any more anticomplementary results with this large amount of serum at the end of the first week than on the first day. This phenomenon may be due to the presence of some inhibitory substance which disappears from the serum upon standing for several days. Stimson mentions the role possibly played by the inhibiting substances in the patient’s serum, which are stated by Caul- field, Calmette and his co-workers to occur in certain tubercle sera of tubercular individuals and which have the effect of pro-. ducing negative reactions in sera that contain anti-bodies. They, however, did not state whether these inhibitory substances disappeared from the sera upon standing. This is a point that will require rigid investigaition. Various specimens of serum taken from the same patient at different times gave complement fixation results which were comparable. I took duplicate specimens from one to six weeks apart on 31 cases. The results on 24 were identical. The re- maining 7 gave results that were closely alike; the slight differ- ences being, perhaps, explained by the variations in the physical condition of the patients. CLASSIFICATION OF THE CASES The types of tuberculosis cases have been variously classified by many investigators. Noted classifications were made by Williams, of Brompton Hospital, Cornet, L. Bard, Koeniger, Turban, (25-26), Meissen (27), and Walther L. Rathbun (28). Rathbun classifies tuberculosis as incipient, moderately advanced and far advanced. The following broad classification seems to be necessary both from the clinical and from the laboratory stand- point. Types II, III and IV must be separated, for they give entirely different clinical pictures and also entirely different com- plement fixation results. Type I. Primary cases; very few physical symptoms present; no tubercle bacilli found in the sputum or found only after the examination of many specimens by the antiformin method. Type II. Active cases; patient expectorating tubercle bacilli, the diseased area being walled off incompletely or not at all. COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 359 Type III. Active cases in the last stage; patient in a dying condition. Type IV. Partially inactive cases; that is, cases expectorating tubercle bacilli but having very marked fibrous formation with the consequent complete walling off of the diseased area from the body proper. Type V. Inactive cases. Type VI. Cases reported as suspicious but expectorating no tubercle bacilli and having no symptoms of tuberculosis. Type VII. Non-tubercular cases. 7 TABLE 2 Comparison of results in the various types of cases % a ! Ah & 2 Oe. AA AA 1 ed 1! Behera olor Se |) Se. eH aleeec| Beane Lone eee) |. ee | eee) emer E : Q Z DZ DZ S 5 Allen gE Gey WiGrrs eS SS» | ta 3 i=] a a & =) a a Mon Mong a n fe k& A QP hes gaRO eis) ae E, = aS < vr o op Oo 2 as) oP ° 4 5 a OZ OoZ< o ps a Z <8 Z & zat Aa we C) 3a & 2 eg ee a I aise Aap: Bp % 3 a r=} 2 2 D ro) R Ap a Za ae 27a Bm & Ansa Be aad a om ee] as D Bae Aes Ag > gam aye = = = =P Om RO cele! 20s saa aps BDz S 3) be pe ae gA Bae pm & pan p@A paw B Z ZA a ry cy ry Z Za Z Z 1 6 6 0 100 0 83 1 6 1 5 2 47* | 46 1 98 2 90 15 46 10 41 3 11 3 3 27 27 100 1 3 1 3 4 19 6 5 31 26 0 1 6 0 0 5 12 3 0 25 0 0 0 3 0 0 6 4 0 0 0 0 0 0 0 0 0 7 99 0 0 0 0 0 0 0 0 0 | * Two sera anticomplementary. In table 2 I have grouped the results of the complement- fixation tests according to the clinical type of the cases examined. Of 6 type I cases, all gave positive complement fixation re- actions, or 100 per cent of positive results. Of 49 type IT cases, 46 were positive, 1 was doubtful and 2 were anticomplementary, or omitting the 2 anticomplementary sera 98 per cent were posi- tive and 2 per cent were doubtful. Of 11 type III cases 3 were positive, 3 were doubtful and 5 were negative; or 27 per cent were positive and 27 per cent were doubtful. Of 19 type IV cases, 6 were positive, 5 were doubtful and 8 were negative; or 31 per cent were positive and 26 per cent were doubtful. 360 HASSOW VON WEDEL Of 12 type V cases, 3 were positive, 9 were negative; or 25 per cent were positive. Of 4 type VI cases, all were negative. Of 92 sterile type VII sera, all were negative. The contaminated type VII cases gave weak non-specific reactions. The clinical diagnoses were all made either by Dr. Rosenberg, who was formerly diagnostician for tuberculosis at the West- chester County Hospital, or by Dr. Slade, diagnostician for tuber- culosis for the New York City Board of Health. As Dr. Rosenberg left the Westchester County Hospital before my work was completed, some of the latter cases were diagnosed and classified by the internes in the hospital. All of these latter cases were examined and reclassified by Dr. Slade thus the above diagnoses were all made by expert diagnosticians. Dr. Slade’s diagnoses were made after my results had already been recorded. In attempting to determine whether the temperature, pulse and respiration records of the patients bore any direct relation- ship to the complement fixation reactions, I have compared my laboratory findings with the record charts, and I find that the cases giving a 4 plus reaction had an average temperature of 99.4; pulse, 94; and respiration, 27. The patients giving a 2 plus reaction had an average temperature of 99; pulse, 85; respiration, 24. Those giving a doubtful or negative reaction had an average temperature of 99; pulse, 95; and respiration, 25. 40 per cent of the 4 plus eases and only 10 per cent of the doubt- ful and negative cases had a temperature of over a hundred. In other words a large percentage of cases having a high temperature and respiration records gave strongly positive reactions; vice- versa, a large percentage of cases having low temperature and respiration records gave negative reactions. Forty per cent of the 4 plus, 10 per cent of the 2 plus and only 4 per cent of the doubt- ful and negative cases had a respiration record of 30 or over. I also attempted to see whether, possibly, the age of the patient had any effect on our reactions and found that the average age of the patients giving a 4 plus reaction was 36; the average age of those giving a 2 plus reaction was 43; the average age of those giv- ing a 1 plus reaction was 35; the average age of those giving a plus-minus reaction was 40. As all of the groups contained pa- tients both young and old, no conclusions could be made. COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 361 _ Miss M. A. Wilson has observed that the serum of a large per- centage of guinea-pigs is unsuitable for use in the complement fixation test for tuberculosis because the complement of these sera is not fixed, under the usual conditions of the test, with the sera of tuberculous individuals. A similar irregularity in the guinea-pigs’ sera, when employed in the Wassermann test, was reported by Noguchi and Bron- fenbrenner (29) in their article on the ‘Variation of the comple- ment activity and fixability of guinea-pig’s sera in Wassermann work.” They state that positive patient’s sera will often fix complement from some guinea-pigs and will not fix complement from other guinea-pigs, but that there is no definite relationship between the complementary activity and the fixability of a given specimen of guinea-pig’s serum. However, the irregularity observed by the latter authors was relatively infrequent as com- pared with that reported by Miss Wilson, who found that for the complement-fixation test in meningococcus and gonococcus infection and in tuberculosis, respectively about one-tenth, one-third and two-thirds of all guinea-pigs supply inefficient complement. With the purpose of further studying this phenomenon I carried out a number of tests with sera from actively tubercular patients, at first, in two series; one with pooled guinea-pigs’ sera that had not been tested as to fixability, the other with pooled guinea-pigs’ sera, each of which had been separately so tested and found satisfactory. The pooled serum was derived, in each case, from six or eight guinea-pigs. These parallel tests gave practi- cally the same results. In a few instances I obtained a 2 plus instead of a 1 plus reaction, or a 3 plus reaction with the tested complement and a 2 plus reaction with the untested complement; but this difference was not regularly encountered and in a few instances, in fact, better results were obtained with the pooled untested complement than with the specially tested complement. This experiment being inconclusive I then carried out a series of tests with sera from seven frank tubercular cases (six of which I had already examined) testing each guinea-pig’s complement * THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 3 362 HASSOW VON WEDEL separately against each positive serum. The results of these tests are shown in table 3. Serum 1, tested with complement from guinea-pig 1, gave negative results in the antigen tubes containing 0.02 cc. and 0.04 cc. of patient’s serum on the first and fourth days after taking the specimen from the patient. On the sixth day and again on the seventh day complement from guinea-pig 2 gave a 3 plus reaction with 0.02 ec. of patient’s serum and a 4 plus reaction with double that amount. Tests made on the ninth and eleventh days with complement from guinea-pig 3, gave negative results TABLE 3 Showing variations in flexbility of the complement of different guinea-pigs’ sera in the complement-fixation test in tuberculosis + + if: COMPLEMENT|COMPLEMENT|COMPLEMENT COMPLEMENT COMPLEMENT |COMPLEMENT no. 1, no. l, No. 2, No. 2, No. 3, No. 3, FIRST TEST | LATER TEST | FIRST TEST | LATER TEST | FIRST TEST | LATER TEST POOLED TESTED COM- PLEMENT FIRST TEST PATIENT’S SERUM NO. Say || 9 ce eS eee | ee eee 1 2} — — | — — |44 34 /4+ 34+]14 -— | — -— | 44+ 2+ 3} — — | — — {44 24/34 2+] — -— | —- -— |14+ = 4} — — | — — [44+ 44]4+ 44 ]}/14 - 44 3+ 5}2+ — |384+ 14 /4+ 44/44 44/44 -— | 44 -— | 4t 44 6} — — | — — |4+ 14/384 14] —- -— | —- -— ] = = 7) — — |2+ — [44 44/44 44] —- — | — — 184 14 * The results in the first column were obtained with 0.04 cc. of patient’s serum. 1 The results in the second column were obtained with 0.02 ec. of patients’ serum. t Complement was preserved in the interim with an equal amount of 18 per eent salt solution and kept in contact with ice. with all amounts of patient’s serum. On the fourteenth day, the pooled complement, made from sera of six tested guinea-pigs, gave results which were practically the same as those obtained when complement from guinea-pig 2 was used, that is, a 2 plus reaction with 0.01 ce. and 0.02 ec. of patient’s serum and a 4 plus reaction with 0.02 cc. and 0.04 ce. of patient’s serum. On the sixteenth day, complement from an additional guinea-pig which may be designated 4 gave aplus-minusand a 1 plus reaction. Patients’ sera 2, 3, 4, 6 and 7 gave almost identical results. Serum 5 gave a 1 plus and a negative result with complement ela a ieee a COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 363 -1, a 4 plus reaction with complement 2, negative results with complement 3, a 4 plus reaction with the pooled complement and a 1 plus reaction with complement 4, when the regular amount of patient’s serum was used. The results of this limited experiment confirm those reported by Miss Wilson, inasmuch as only one of the three individual guinea- pigs’ sera was found to be suitable for use in the test. In view of the fact that with this one serum (guinea-pig 2) the different patients’ sera reacted differently and since the otherwise un- satisfactory guinea-pig’s serum 3 was fixed with the double amount of one of the patients’ sera (serum 5) it would seem advis- able in testing guinea-pigs’ sera as to fixability, to test them with a strongly reacting patient’s serum, or, better, with two such sera and to use the latter in the usual quantity, not with the double quantity. H. J. Corper (19) states that in his series of cases, he used 92 sera with positive Wassermann reactions and of these, 65 gave cross fixation with all tubercle bacterial antigens. He, therefore, concludes that in the presence of a positive Wassermann reaction, the presence of a positive complement fixation test for tuberculosis is of no practical value. He also states (18) that, while the most reliable investigators concede that a suspension of living tubercle bacilli is the only one of the many antigens used, that is of specific value, the objections to the bacillary emulsion are the small leeway between the antigenic and the anticomplementary dose, the turbidity produced in the tubes and the fairly high percent- age of non-specific reactions. My own experiences are in disagreement with all of Corper’s conclusions. First, because in my series of cases there were 26 specimens from patients that gave positive Wassermann reactions but offered no physical symptoms of tuberculosis. None of these gave any cross fixation with the Wilson tubercu- losis antigen. Secondly, in no instance did I obtain a positive reaction using double the regular amount of sterile patient’s serum in my entire series of known non-tubercular cases. Thirdly I have never found the Wilson antigen to be anticomplementary in four times the dose used in the test, if the antigen is heated at 364 HASSOW VON WEDEL 55°C. for one half hour just before being used. Fourthly, my tests in the active tubercular cases gave 98 per cent of positive reactions. Fifthly, the very slight turbidity produced by the antigen in no way interferes with the reading of the results. CONCLUSIONS 1. The tubercle bacillus antigen of Miss Wilson is not anti- complementary in four times the amount capable of producing positive complement fixation with sera from the great majority of cases of active tuberculosis. 2. Pooled complement from at least six guinea-pigs should be used in making the tests, or the complement from single pigs should be tested for its complement fixation value with known positive sera. 3. Double the original Wassermann amount of patients’ serum should be used. 4. No report should be made until the sera has been tested, after having been kept under sterile conditions in the ice chest for from four to six days, preferably six days. 5. My results seem to indicate that if the afore-mentioned modifications of the original complement fixation tests are used, 100 per cent of non-tubercular cases will give absolutely negative results; nearly 100 per cent of the primary and active cases will give positive results with the exception of the dying cases; and about 25 per cent of the partially inactive and inactive cases will give only weak positive results. Before final percentage results can be arrived at, it will be necessary to make many more tests on a large number of sera from active, inactive and incipient pulmonary tubercular cases and a large number of control sera from non-tubercular cases. I hope to report on the results of about three thousand tests in the fall. I wish to take this opportunity to thank Dr. Wm. H. Park and Dr. Russell, chief of the staff of the Westchester County Hospital for their invaluable assistance and advice, without which this study could not have been made. COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 365 NAME May Cooper: s...c%- i... LORE BG) ia eee O70 fd i eae | SPUTUM REPORT* essoltreulie PH+tt++t+et+tteeteet+segs+ste TYPE OF DIAGNOSIS NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNKH HBR Yee CHART 4 Clinical and laboratory records of all our tuberculosis patients TEMPERATURE OO I 0S 98 130 135 100 136 RESPIRATIONT COMPLEMENT-FIXATION RESULTS ONE DAYy* |One week*| LATER® gia uh oi ama eaeal| resi «|v | ae] ee 0) ea es er 44 0 0 —| |4t] s+ +] - B+ [44 |2+ + | = [4+ [2+ [3+ | + +] — |4+ [2+ es ea ees lee 2+ | + [4+ |4+ Oa ae dee |e 24+ | — 4+ |4+ [3+ | + 2+ | — 4+ |4+ 4+ 4+ [a+ + 4+ 2+ 4+ \24 [4+ [3+ 4+ [2+ a+ [3+ 4+ l2+ [4+ [3+ = Aa 38+ 4+ 44 A+ 2+ [4+ |4+ [44 [4+ 2+ | — |4+ |/4+ —| [4+] f+ —| B+ ~ oe +) +] [8+ 4+ [2+ 4+ 4+ |4+ |4+ 4+ |4+ |44 |44 /44 /4+ 44+] + 2+ | + 4+ [4+ — | — [4+ [4+ [44 [3+ 4+ /2+ /44 |4+ 4+ 3+ | 0} 0 |4+ |4+ —| 4+] [4+ —/| f+] [4+ —| fat] 4+ +] — [4+ |4+ 2. 4+ —| [4+] 4+ —| [4+] [4+ —| jab 4+ jo+ 4+} 4+ - 366 HASSOW VON WEDEL CHART 4—Continued 2 COMPLEMENT-FIXATION RESULTS B18 i | o iy 2 % | One day* |One week*| Later® NAME a A & 2 pte ele lela|@ o | alae = : : + 2/25 | 99] 76) 24) — | — 2+ |= DHICIOS = cee ne ae =e: { ri Ge (eM el nb uD) ee i ne Pcl Far Mannix samara: + ole + 2 | 39 | 99 |108 | 24 4+ 4+ a ae { + | 2] 50] 98 | 84} 26} — | — |84+ |] + /2+ ] — ey hone a) eR alae o7 |72 |} 93,| — | — lap || eee Smit hieee cheers cece — 2] 35 | 98 | 84} 20] — |} — j8+ ] + May Hide {3}, ayer tue) 28 | RU MEVOSS cat tie e-werers ei { Tally eae ee z ee oa +o 2 haa aS dS = ac 7 2 | 34 |101 {122 | 28 == 2+ 4+ Mc wenieee eee see , in 9 102 {104 | 29 | | = lat 1b IGUSTIZO Sot tan cetct +} 21] 53] 98 | 86} 32 2+ 38+ 4+ Kiernanh erties nee + 2 Ne52 | 974-76) 20 — — |4+ /2+ Taylor { — | 2] 39 | 98 | 88 | 24 _ = (2+ | — SAN. aa ees ins 67-76 )\oo tee leet] eee Moeia { + 2;}40| 98 | 76} 22;—/}—}—|—-/]+] = Pete ee eRe roa ae 100 | s0|20| —| — lor | + : +] 3] 32] 99) 86} 29| +] = |44 /38+ Mirsainicheeremccccess { Get og ea] a) ge ge AaigetPerilla-ckg.c cee oe + 3 | 19-|102 {132 | 30 4+ 4+ 4+ Wood titre ee see + 3 | 44 100 }100 | 28; =} —}|—|]—]—] = Chas: Doyle=:.=:.4.'.8: { gid (ial coe se 8 eer bea etal a 3 SF | ea ee sly ae BANAT Feteig | au ee Kouwarte.. eee _ 3 | 34 | 99 |102 | 30 = — — st 3 97 | 76} 22} —|} —|]—]|— + 3 | 20 |104 |142 | 28 _ _— Rogneracancseane se one { ao ge ae + 4) 55} 98 | 72} 22) —|—/]}+] — Monahanesenne sneer +) 4 97 | 72} 20} —| —| —]| — are aE |e | i Wipe Ahrens. nc: ie bree orme + 4} 53 | 98 |112 | 26 — — _ +] 4] 48 | 98 | 90 | 28 _ —|+]-—- © Bennett#ea.-.. so. { a 4 Oeil ea leoanl =a lcs |-ae| oe is ; se 4 |} 31 |101 |110 | 24 _ _ _ PiraROussO-eee ese { moles pee ils aes teed i$ ee Deets eet es tae + | 4] 44] 98] 80 | 22 + 2. | aan Pohsamisarers too sscce + | 41] 26 | 99 |104 | 28 o 0] ac] ae COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 367 CHART 4—Concluded HI COMPLEMENT-FIXATION RESULTS ° 4 o = % | One day* |One week*| Later®? NAME a = 2 * < & ++ tor a n a mM P + 4} 30 |100 |116 | 29; —} —}] —] — SS { ill pe ee ilhet i — 4 | 30 |100 | 88 | 28] ac} ac] — ] — EVO a eae 59 2) a aH Mena Stokes..i. 0.0... 5) 4) 35] 99 | 86] 22}; =|] —]—] — i Spr cla) I Gil 86.1822) eee eed lee ge burturmmack.<........ { rd 4 eee ie he es le N — 4 | 41 | 98 | 92 | 24 — j2- )e) =] = SDIN7o eee AY pale | ee ee Pn — 5 | 32 | 99 {120 | 26 — —|—|-— Jeo |S) V0 | er { S BU BD 597 >. 76 al Seales Fs Ee cea ieee | WaerBennett: i... .... 2.0. _ 5 | 54 | 98 | 96 | 24 — —|/—-|-—- LEO Cae 5 ens Sale ee 5 | 44 | 97 |104 | 26 —/—-/;/—]|-]-—- 12-0700 Re aa 5 | 43 | 98 | 68 | 22; —| —|] —] — MMOPOUIE SSG. coc es — | 5 | 65'| 981 92 | 28 — a 25 The =e | 47 1-97 1-76: 122 — —/-|]- MNDOU AGS So lacest. creas = |) 5) 120) 99" || 725° 20 - —|—|— LSGT Sees oa nee - 5 | 49 | 98 1100 | 26 — —|/—-—|-—- Leno Fiorella......... { iz : BEEN NUEO |, Ze a si z % yer May Sedden............. — 5 | 50'| 98 | 88 |) 22 j2-- | = Wine Webbe 22h ese... - 6 | 34 | 97 | 80} 22 _ — — MMe =|) 6 |y35 | 98 | 90 | 24 = a Pees eet =" | = 6: | 402) 97 1-80) || 22 —;}/-;|}—-—-]-]- “SEG 0 ESS sO ae ea _ 6 |.22 | 99 | 70 | 24°) — | — |= = * Sputum report obtained within two weeks of the time the blood specimen was removed. { Temperature, pulse and respiration were taken the same afternoon the blood was removed. { Refers to the use of double the regular Wassermann amount of patient’s sera. § Refers to the use of the regular Wassermann amount of patient’s sera. “ Refers to the time the reaction was made after the specimen was removed from the patient. ° These results were obtained from two to six weeks after the specimens were removed from the patjents. 368 HASSOW VON WEDEL REFERENCES (1) Wipat anv LeESourp: Cited by Shennan and Miller. Edenburgh Med. Jour.; 1918, 10, 81. (2) Borpet AND GENGoU: Compt. rend. Acad. de sc., 1903, 187, 351. (3) WASSERMANN AND Bruck: Deutsch. med. Wehnschr., 1906, 32, 449. (4) CAULFIELD: Jour. Med. Research, 1911, 24, 122. (5) Larrp: Jour. Med. Research, 1912, 27, 163. (6) Hammer: Munchen. med. Wehnschr., 1912, 59, 1750. (7) CALMETTE AND Massou: Compt. rend. Soc. de biol., 1912, 73, 120. (8) Mucu: Miinch, med. Woch., 1912, 59, 685. (9) Frazer: Ztschr. f. Immunititsf., 1913, 20, 291. (10) Dupcron, Meek AnD WetrR: Lancet, 1913, 184, 19. (11) DupGron, MEEK AND WerR: Jour. Hyg., 1914, 14, 52, 72. (12) BRONFENBRENNER: Arch. Int. Med., 1914, 14, 786. (13) McInrosu, FILDES AND RapcuirFe: Lancet, 1914, 185, 485. (14) BesrepKa: Compt. rend. Acad. d. sc., 1913, 156, 1633. (15) Inman: Compt. rend. Soc. de biol., 1914, 76, 251. (16) Kuss, LAREDDE AND RUBENSTEIN: Compt. rend. Soc. de biol., 1914, 76, 244. (17) Strnson: Bull. Hyg. Lab. U.S. P. H. and M. H.S., 1915, no. 101, 7. (18) Corprr: Jour. Inf. Diseases, 1916, 19, 315. (19) CorpER AND Sweany: Jour. A. M. A. 1917, 68, 1598. (20) Miuuer, H. R., anp Zinsser, J.: Proc. Soc. Exper. Biol. and Med. 1916, 13, 134. (21) Stack, Burns, CASTLEMAN AND Battery: Jour. Amer. Med. Ass., 1917, 68, 1386. (22) Craic: Am. Jour. Med. Sce., 1915, 150, 781. (23) Mriuuer, H.R.: The clinical value of complement fixation in tuberculosis, The Jour. A. M. A., Nov. 18, 1916, p. 1519. (24) Luckre, Batpwin: Jour. Immunol., 1916, 457. (25) Turpan: Diagnosis of tuberculosis of the iung, Edition 1906, 42-49. (26) Turpan: Jour. of Amer. Med. Ass., Jan. 30, 1909. (27) Merssen: Handbuch der Tuberculose, Band 1, 743, published by Barth, Leipsig, 1914. ; (28) Ratrusun, W. I.: Tuberculosis Monograph, no. 4, March, 1917, N. Y. C., Dept. of Health. (29) Nocucut AND BRONFENBRENNER: J. Exp. Med., 1911, 13, 69. (30) Rupee. AND RickMANN: Zeit. f. Immunititsf., 1910, 6, 344. (31) Laus: Zeit. f. Immunititsf., 1911, 9, 126. (32) Scuutrz: Zeit. f. Immunititsf., 1911, 9, 709. (33) BorissJAK, SIEBER AND METALNIKOW: Zeit. f. Immunititsf., 1911, 12, 65. (34) Porter: Jour. Hygiene, 1911, 11, 112. (35) CauLFrireLp: Arch. Int. Med., 1911, 8, 440. (386) CAULFIELD: Arch. Int. Med., 1911, 8, 128. (37) DEBRE AND ParaF: Compt. Rend. Soe. Biol., 1911, 228 et seq. (88) Dr1LMANN: Zeit. f. Immunitatsf., 1911, 10, 421. (39) Larrpb: Jour. Med. Res., 1912, 27, 163. (40) Mouuers: Cent. Bact. ref., 1912, 54, Beiheft, 202. COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 369 (41) Hammer: Deutsch. Tierarzt. Woch., 1912, 20, 593. (42) Zweta: Berl. klin. Woch., 1912, 49, 1845. (43) CALMETTE AND Massou: Compt. Rend. Soc. Biol., 1912, 73, 122. (44) Meyer, Kurt: Zeit. f. Immunititsf., 1912, 14, 359. (45) LetuLue: These Fac. Med. Paris, 1912. (46) Aoxr: Zeit. f. Immunititsf., 1912, 13, 192. (47) Fraser: Zeit. f. Immunititsf., 1913, 20, 291. (48) DupGcron, Merek anp WEIR: Lancet, 1913, 184, 19. (49) Harris AND Lanrorp: Jour. Inf. Dis., 1913, 13, 301. (50) Bana anp ANDERSON: Cent. Bact. Orig. 69, 517. (51) AnMAND, DELILLE, Rist AnD VAaucHER: Compt. Rend. Soc. Biol., 1913, 74, 791. (52) WyscHELLESKyY: Zeit. f. Tuberk., 1913, 19, 209. (53) BunpscuuH: Zeit. Hyg. 1913, 73, 427. (54) FRANCESCELLI: Zeit. f. Immunitiitsf., 1913, 20, 309. (55) KINGHORN AND TwITCHELL: Zeit. f. Tuberk., 1913, 20, 11. (56) CALMETTE AND Masson: Compt. Rend. Soc. Biol. 1913, no. 28, 160. (57) Rorue AND BrerBaum: Deutch. Med. Woch., 1913, no. 14, 544. (58) Momose: Deutch. Med. Woch., 1913, no. 22, 1029. (59) BesrepKa: Zeit. f. Immunititsf., 1914, 21, 77. (60) Wwepensky: Revised by Wulffius., Zeit. f. Immunititsf. referat, 1914, 8, 931. (61) DEBAINS AND JUPILLE: Comp. Rend. Soc. Biol., 1914, 76, 199. (62) Inman: Comp. Rend. Soe. Biol., 1914, 76, 251. (63) McInvrosH AND Fiupss: Lancet, 1914, 11, 485. (64) Rapciirre: Lancet, 1914, 11, 488. (65) Duparon, MEEK AND WetrR: Lancet, 1914, 11, 72. (66) BRONFENBRENNER: Jour. Exp. Med., 1912, 15, 598. (67) BRONFENBRENNER: Soc. for Exp. Biol. and Med., 1914, 11, 92-93. (68) E1caHorN AND BuuMBERG: Jour. Agricultural Research, 8, no. 1. ‘J - ray we ie ot | bi 7 ale ' 7 i ros rth + . ; i mh ’ ip r a i . i : “ : a : a / « THE ROLE OF IMMUNITY IN THE CONDUCT OF THE PRESENT WAR! JOHN A. KOLMER Philadelphia When the history of the present great war is written a notable victory over the common enemy, disease, will be recorded as one of the greatest triumphs in this greatest of all conflicts. In all probability this triumphs over disease will also be recorded as the most important single factor in explanation of the stamina and long-sustained man-power of the involved nations; never before in the history of the world have so many men been engaged in combat with such freedom from internal deterioration due to disease not only among the warriors in preparation and at the line of battle, but also among the supporting civilian population; history records many instances of cessation of wars and sieges due to disease among offenders or defenders or both and a re- markable freedom from pestilence in the present conflict has undoubtedly played a prominent réle in permitting it to reach the dimensions of the greatest of all wars. This triumph over disease is due in most part to prevention by sanitary measures, specific immunization and improved meth- ods of treatment of the inevitable and unavoidable sick and injured. With the exception of small-pox, in which disease the science of immunity long ago contributed the most important and one essential means of prevention in the form of cow-pox vaccination, sanitary measures embracing the proper disposal of infectious material and the prevention of the spread of infec- tious diseases by the processes of isolation and quarantine and including the maintenance of individual resistance by proper 1 Presidential address at the Fifth Annual Meeting of the American Asso- ciation of Immunologists held at the University of Pennsylvania on March 29 and 30, 1918. 371 372 JOHN A. KOLMER food, work, rest and play, has played the most important rdle, with the science of micro-parasitology and immunity embracing a knowledge of the parasitic causes of so many of the acute in- fectious diseases and specific immunization of several by means of vaccines and sera, a close second worthy of the division of honor and credit. Mention has just been made of cow-pox vaccination in the prevention of small-pox; history shows that without this im- munological discovery and process great wars would be impossible and particularly one of the present dimensions involving so many countries and millions of men and offering splendid facilities for the rapid dissemination of the virus; the prevention of ty- phoid and paratyphoid fever by means of active immunization with vaccines while not as successful as cow-pox vaccination, must be credited with a great measure of success in the pre- vention of these diseases formerly so widely prevalent among armies; certain measures of success which in some instances are quitemarked, have also attended the prevention of bubonic plague, bacillary dysentery, cholera and rabies by means of active immunization. The prevention and treatment of tetanus and diphtheria with their respective antitoxic sera have proven most valuable im- munological procedures and particularly so in the prevention of tetanus at a time when the modern earth digging methods of war have widely distributed the bacillus and rendered practically every wound regardless of severity and location a real danger and menace to life; likewise in the treatment of epidemic cerebrospinal meningitis a potent antiserum has proven conclusively that it is the best means of treatment, its free and intelligent use resulting in a considerable reduction in the percentages of death and the disabling sequelae. In the treatment of that dreaded disease, pneumonia, ‘‘The Captain of the Men of Death,” the science of immunity has contributed a means for the serologic diagnosis of the type of pneumococcus present and prepared a serum for the treatment of type 1 infections which has proven its worth and right to a prominent place in the modern treatment of this disease. Still more recently the science of immunity has pro- THE ROLE OF IMMUNITY IN THE PRESENT WAR ite duced for the toxins of the gas-producing bacillus which has played havoc among so many of our wounded heroes in the past and present, a serum that bids fair to prove of value in the prevention and treatment of this dangerous infection. Immunological reactions are also proving of practical value at the present time in the diagnosis of several diseases and particularly the serological reaction in the diagnosis of syphilis, which disease menaces all peoples at present and particularly in the future, by reason of its wide dissemination and insidious nature rendering all persons regardless of age and sex vulnerable and liable to its attack. Furthermore in the treatment of this “Third Great Plague,” the newly developed branch of chem- otherapy in the field of immunity, has contributed a most remarkable remedy in the form of dioxydiaminoarsenobenzol or the popularly known ‘606,’ and our hopes for the present prevention of syphilis and protection of the future and unborn peoples, resides in large part in the treatment of the infected until they are rendered less infectious even if not completely cured, by the widespread and more free employment of this and other anti-syphilitic remedies. To this end all efforts made to lower its cost and thereby facilitate its use in the treatment of the poor and of large numbers of persons, are to be welcomed as commendable and a work of first rank importance. Therefore, while the science of immunity has contributed considerable that is of practical value in the diagnosis and treat- ment of various diseases of particular importance in relation to the present war, much and indeed more, remains to be accom- plished of which mention may be made of but a few of the more pressing problems as follows: The discovery of a test of effective natural immunity to pneumonia and meningococcus meningitis, if such immunity exists, comparable to the Schick test for im- munity to diphtheria, as a means of encouraging and facilitating active immunization with vaccines in the prevention of these diseases; a test for natural immunity to tetanus, which may be developed along the lines of the Schick test if some means can be devised for removing the danger of the spore; a means of specific immunization against measles, acute anterior polio- 374 JOHN A. KOLMER myelitis, syphilis and gonorrhea and an improvement of our means for active immunization against cholera, plague,dysentery and typhoid fever, not to forget that problem of problems, namely, the discovery of a means of specific immunization and treatment for tuberculosis. At out meeting last year the Association officially passed resolu- tions offermg to our federal government the services of our members and laboratories in the conduct of our great war; before and since then not a few of our members have enlisted for active duty in the federal service and at least one has given up his life as a sacrifice to duty; many and probably all members of our Association are more or less intimately associated in some work having a direct bearing upon the problems of health and disease and particularly those menacing or likely to menace the health of our armies abroad and at home; to all the Association would hold up in pardonable pride the accomplishments of the science of immunity in the past and wish all God-speed in their work for the present and future for the health and happiness of mankind for all time and everywhere. en heh: MODE OF* ACTION IN) VITRO AND THE PREPARATION OF HEMOLYTIC ANTIBODIES A. K. BALLS anp JOHN H. KORNS From the Department of Bacteriology, College of Physicians and Surgeons, Columbia University, New York Received for publication June 13, 1918 This work was undertaken with the idea, first, of studying the mechanism of amboceptor action in vitro, and secondly, of as- certaining, if possible. what part of the red blood cells is respon- sible for their antigenic property. Inasmuch as our work has now been brought to a necessary close, we are bringing together in this article our findings even though our conclusions at this stage cannot be far-reaching. Many of the general principles applying to the mode of action of hemolytic antibodies have been worked out by Muir (1) who showed that the combination of amboceptor and cells is a weak one, easily dissociable, and occurs according to a law apparently re- sembling the law of mass action. He showed, further, that amboceptor is not destroyed by hemolysis but remains bound to the receptors of the hemolyzed cells from which it can dissociate in the same manner as from the whole cells, such dissociation being more marked at incubator temperature. By analogy with the law of mass action, if the amount of un- hemolyzed cells is very large in proportion to the amount of hemolyzed receptors the dissociation from the hemolyzed por- tion will be practically complete, and almost the entire amount of amboceptor at any one time will be resident on the unhemo- lyzed cells. This was found to be so in experiments of which the following is an example. A relatively large quantity (1 ec.) of washed sheep cells, con- centrated by rapid centrifugation for ten minutes, was diluted 275 376 A. K. BALLS AND JOHN H. KORNS to 40 ec., and 80 units! of amboceptor added, followed by three units of complement. After incubating for thirty minutes a hemolysis of about 10 per cent of the cells had occurred. The liquid was centrifuged again at the same speed and for the same period as before, the volume of concentrated cells was measured, and the sediment was then made up to a 5 per cent suspension. A quantity of untreated cells was then taken, equal to the amount of cells remaining after the partial hemolysis just described, and to this 80 units of amboceptor and enough salt solution to secure a 5 per cent suspension were added. The amount of complement necessary completely to hemolyze 0.5 ec. of each of these two sus- pensions was determined and was found to be the same. We thus concluded that both samples of cells contained the same amount of amboceptor, showing first, that under these conditions, dis- sociation, as might be expected from chemical reasoning, is negli- gible from the unhemolyzed cells, and secondly, that amboceptor is not only set free during hemolysis, but is quantitatively unaltered. The results are noted in table 1. TABLE 1 ‘ Showing the action of amboceptor to be “ progressive ”’ COMPLEMENT 1:10 0.10 ce. 011 ce. 0.12 ce. 0.13 ce. 0.14 ee. 0 iireatedsecellsce...e2s- ose. +++ )+-++ sie +4+4+4+/4+4+4++4+ = Wntreatedscells' sss. c0 es +++ |)+++ siete ++++)++++ - ++-+-+ indicates complete hemolysis. -— Indicates no hemolysis. 1 Throughout this paper the ‘‘unit’’ of amboceptor is taken as the smallest amount of inactivated immune rabbit serum which hemolyzed completely 0.5 ce. of 5 per cent red cell suspension, with the addition of 0.05 ce. of complement, after one hour’s incubation at 37°C. and in a total volume of 2.5 cc. The “‘unit’’ of complement is the smallest amount of guinea pig serum which completely hemo- lyzes in one hour at 37°C. 0.5 ec. of 5 per cent red cell suspension in the presence of two (2) units of amboceptor, in a total volume of 2.5 cc. The 5 per cent red cell suspension is prepared by centrifuging washed sheep cells rapidly at a definite speed for ten minutes and then bringing up the measured sediment to the proper dilution with 0.85 per cent NaCl solution. PREPARATION OF HEMOLYTIC ANTIBODIES 377 - Should the amount of hemolyzed cells be large or should sey- eral washings with salt solution be made, the amboceptor on. the cells so treated will be smaller in amount than on the control series, showing a considerable loss by dissociation. If the hemolyzed receptors are heavily loaded with amboceptor, it is easy to conceive that this should in part dissociate, and undis- solved red cells being still present, that the dissociated ambo- ceptor should attach itself to these, causing their hemolysis. The action of amboceptor would therefore not cease with the hemolyzing of one cell, but would be continuous, and as the amount of hemolyzed receptors increased and the amount of unhemolyzed receptors decreased, the hemolyzed receptors would become less saturated with immune body, and conse- quently would split off less, thus causing the velocity of the reac- tion to decrease, a phenomenon well recognized in hemolytic work, for the last traces of unhemolyzed cells disappear very slowly. That this is not chiefly due to the deterioration of complement at incubation temperature can be shown by the fact that the system still contains that component, usually in considerable amounts. Furthermore, the velocity of deteriora- tion of complement in the presence of amboceptor is practically the same as in pure salt solution (as far as our rough immunologi- cal methods will permit us to measure). The products of comple- ment deterioration, so called ‘“‘ecomplementoid,”’ likewise do not inhibit hemolysis when used in quantities comparable to the amounts of complement ordinarily used in hemolytic work. On the other hand, laked cells, whether dissolved by amboceptor and complement, or by distiled water and subsequently made isotonic are capable of greatly HOIST the reaction. In table 2, A is guinea-pig’s serum heated to 56°C. for thirty minutes and represents “complementoid.” B represents 100 per cent cells laked with distilled water, made isotonic, and diluted with salt solution to ten times the original volume. Each tube contains 0.5 cc. of 5 per cent cells, two units of amboceptor and one unit of complement, the total volume being 2.5 ec. An addi- tional control for A not tabulated and containing 1 ce. of Bee. serum but no complement showed no hemolysis. . THE JOURNAL OF IMMUNOLOGY, VOL. II, NO. 5 378 A. K. BALLS AND JOHN H. KORNS This observation under B of table 2 coincides with the work of Muir (2), who showed that the receptors of the red cells are not destroyed by hypotonicity. They are, therefore, capable of binding amboceptor just as the whole cells do, and their addition to the system will cause a diminution in the amount of ambo- ceptor available for hemolytic purposes. Bordet (3) found that the stroma of hemolyzed red cells is capable of fixing antibody, and he was able, by injecting them into an animal, to produce antibodies hemolytic for the whole cells. Stewart (4) confirmed this but found the hemolytic prop- TABLE 2 Showing the absence of inhibiting effect on hemolysis by heated guinea-pig serum, and the presence of such effect by isotonic laked cell products AMOUNTS OF A OR A ADDED TO HEMOLYTIC SYSTEM (cc.) 0.1 0.2 0.3 0.4 0.5 0.6 A. Heated serum 1-10...... a ae es On Gat ef | lI cL LL el B. Laked cell products 10 DETRCON Greet txt eer =- + + + ae AMOUNTS OF B OR B ADDED TO HEMOLYTIC SYSTEM (cc.) 0.7 0.8 0.9 00 0 A. Heated serum 1-10.............. ++++)/++4+4]++44+]/4+4+4+4+]44++ B. Laked cell products 10 per cent..| Tr. ar’ irs AU. igre + Signifies perceptible hemolysis. Tr. a trace of hemolysis. ++-+4+ signifies complete hemolysis. erty of the antiserum less marked than the agglutinative prop- erty. By filtration of the laked cells through a Berkefeld filter, Muir (5) was able to retain all of the stroma and with it practically all of the receptors. 7 . > i. PREPARATION OF HEMOLYTIC ANTIBODIES 379 pended in salt solution, it possesses remarkably active ambo- ceptor-binding properties. As proof of this latter statement we briefly review one experiment. To 1 cc. of diluted stroma, rep- resenting not more than the equivalent of 4 cc. of 5 per cent cells,s 8 units of amboceptor were added and salt solution up to 8 ec. This mixture was kept at room temperature for thirty minutes and it was then centrifuged at high speed. To 2 ce. of the supernatant fluid, 0.5 ec. of 5 per cent cells and 2 units of com- plement were added. This mixture was incubated at 37°C. for thirty minutes. There was no hemolysis. The mixture was again centrifuged at high speed and to the supernatant fluid 0.5 ec. of sensitized cells were added. On incubating at 37°C. complete hemolysis promptly occurred. The amboceptor-binding properties of the stroma were found by us to be partially destroyed by heating to 65°C. for thirty minutes, markedly diminished at 70° and completely destroyed at 80°. These results differ from Muir’s (6) in that he found evidence of amboceptor-binding even after heating to 100° for one hour. Our results point to the protein nature of the anti- genic substance. Dried stroma on being resuspended in salt solution act in the same manner as the moist freshly-prepared material, while ox cell stroma is without effect, showing the species specificity of the reaction. Rabbits were then injected intravenously with the suspensions of stroma from sheep cells and it was observed that the produc- tion of both hemolysins and agglutinins was marked. Since the animals so treated receive very little protein in comparison with those injected with whole cells, it was thought possible to increase the quantity of injected material. Accordingly the stroma of approximately 25 ec. of concentrated sheep cells, sus- pended in about 10 ce. of salt solution, was injected each time, and apparently with little if any ill effect. Table 3 shows the details in the case of two of the rabbits. 3 The calculation was made on the assumption that all the stroma had been secured by our process. We intentionally overestimate the amount. K. BALLS AND JOHN H. KORNS A. 380 ‘S891 IO » 008- 009-T | 00&-1 O0Z-T | 006-1 00Z-T | OOT-T | OT-T | OI-T [0°07 °° ** "1097 oATYVUTQNISSV OO8-T |OSZI-T |OOSE-T |000Z-T OOTI-1 |O0&E-1 OOEE-T |OOSE-T |xOOT-T pale cee ~ ie Geel "19}1} OTPATOWO FY ye, 9 G ; p (2 Z T SORA IISI 00) tatey tan jo roquinN QZ6T | O96T | OL8T | SPGT | O88T | OOST | OSGT | OGAT | O96T | OL6T J SUUvIS UT FUSION : Cael 002-T 006-1 QOC=T OOS=T | 00G=1= |, O0T=E || OTSh) OTE 101} OATPVUTINIS . 009-1 |OOOT-T |OO9T-T |000Z-T |OOSZ-T |OOOS-T jOOSE-T |OOOS-T |OOSZ-T jOOOT-T |x0OT-T yn Joy} o1yATOUO TY g v & Gj T [ccc ttt uoroofur yo roquiny OPST | OSFT | SZST | O&FT | OSET | SZET | OFVET | OCFT J SUIvIS UT J YSIOM [pase Ke kSta oe Avy | 22 Av | st Sew | er Auw | 6 dew | 9 = z= = ells Seale D0, ONO) ORO Oras CG (OL: I) A UM = = = = =e = = = = = = = =a Stas Mallee raeser sages (Gi) touxe[y YITM dnois A190} UesA(T 4 4 3 Se es alhces eect orl ears less ies steel aie ae (21:1) proydsy, WHIM ni ies | ea me ome feat oe ee ice tects > ean) navy Cae ce 3 — |— |= |= J+ J++] ++) ft [44 | tt pr) & eed WM - ps — ms, —-+ Dy olverie) \s) vice) le) buleule: (Oeme\e o°R@ ‘UIn..as a BIBI 3 Ste sales el ete et acter tee |) aoa atect [estes Aaa (21:1) proydsy, WIM a —~ |-— |= |= |- |= | -4+] 44 +] 44 fOr) 2 eed WIM i Se ae eect cp el neem iaeeugepcaaler |e os (OL?) V Mt WHIM a = —-+ + CIO.O “YeOeOecM sys Ct o°@ ‘cWdIn.tes V BIB ceeealccatee ee| eae iea ca ate l ciecte alictonty ai cleat (OL :T) & PBT YIM SS a ee ee paras OL 1) ered aE Se ie cee Neee (ree cis erat aae|| oasis elas GT: De PIOHT AS EM = a = Seis as “o°e ‘undies proyds J, a eae a se th es eee dnois proyda y, 90 | v0 | £0 | 20 rol 500061000200 0|ez000| e100 | no 0 | 900'0 c10'°0 | e200 | ¢0°0 [o-@) LSaL ‘0O°V ‘NADIINV AGOL udd WOUdAS AO SUALAWIINGD 018090 penn era eee Scr a ee ee ee ath os ee ee st suoyniy wabyuy T HIAViL “qsoq Arejuoue|duI00I}UB OY} UL paesn UOTN{Ip usesTjuB Jo syunomy | 419 PREPARING BACTERIAL ANTIGENS | | — |- |-+| + — |-+] + | ++ — |-+] + | ++ — |+ |-++} ++ — |-+] + | ++ dnoid snovo003urueyy ‘pasn uorlsuedsns uss1yUe JO UOTINIIpP SozvoTpUy , OFZOT-T| OCTS- 1) O9SS:T| O8ZT: 1) OFD=T | OZE:T | OOT*T 08:1 pee STE aro eee cor aati ae er eee? STR Seiten: egeeretie sa tial Be fer Sha reece eager tr FEE rata eae et i ee en ee ae fearalca eee (08:1) Td WM sara eee ee (4:1) Id WHIM Erg Bee traces ee re Sane een ee ||| i oe ete verte ohegp hese Neairs se ‘cIn..es T ‘ouInoUg Pee Rogers ete. acai ord eae comets acon dina Parle ten Maat ey, Td ghee ae ae cag eae SP (ae ROO OT) Pa a 2) ‘uIMn.1es O9IN rahe erase cS aoe ae ie TO eee seme TGeanehl eat Sees ee ere sen Hed. ae eceen tenaa ve oueah es [leeches oiteices oetelre tebe la elem o°@ ‘crn..es Of IN “geutal| walle te ake Tecan fete oe eran ae Saracen ra, Were Mr awe erect Talis amochret cae? ee accai | Oras Learn Ore On o°@ ‘canes OLIN are me sacs (ST :1) O9IN WHIM ++ fo OL ET) Os WEA +H fo (ST ST) OLIN WHT ee ee Se). Gh Rian Pen 9 te Ce ear over ae o°R ‘cOn.1es IW 06:1 suoK NIC, UNseg 420 JAMES C. SMALL antigen was in this way determined. With this serum dilution as a constant in the next titration, the smallest amount of antigen giving complete fixation in this serum dilution was determined. This amount doubled became the antigen unit in subsequent comparative serum titrations. The above tables show the results obtained with these anti- gen preparations in direct and cross titrations of immune rabbit sera. TECHNIQUE OF SERUM TITRATIONS Here, as in determining the antigen unit, the anti-goat hemo- lytic system was used. The technique followed is more or less standard and will not be discussed in detail. The immune sera used were shown to contain natural amboceptor for goat’s cells so that it became necessary to remove this. Sensitized goat cells were used throughout the titrations. These were prepared by adding the previously determined amboceptor unit to a 5 per cent suspension of cells and incubating the mixture at 37.5°C. for one hour. After the cells were thrown down by centrifugali- zation the supernatant liquid was drawn off and replaced with an equal amount of fresh salt solution in order to resuspend the cells. The complement used was a 40 per cent solution of pooled guinea-pigs’ sera. In order to obtain the serum dilutions as represented in the tables, a series of tubes was set up with | cc. of normal salt solu- tion in each tube except the first, which contained 1.9cc. To the first tube 0.1 cc. of immune serum was added. After being thoroughly mixed 1 cc. of the dilution was carried across to the second tube and this process was continued down through the series to the last tube where 1 cc. of the dilution was discarded. Serum, complement, and antigen were added in order and the tubes were incubated for a half hour at 37.5°C. oD Sa Nn 10 c c c c Ww c c c tr c ce c 0.8 c c c c tr c c ce — | vst c c . 0.6 c c c c — c c 9) - Ati a avst iene 0.4 ale c c c — c c Cc = tr st c On st c c c a tr c c — w | vst 0.1 w | ale c c — -- w | ale |] — _ tr st 0.08 tr | ale c c — — st | vst | — — (He |) yyy 0.06 — | ale c c — -- tr st — — _ tr 0.04 — | vst c c - — _ Ww — = — — 9.02 _ st | vst c — _ _ tr _ _ -- TABLE 3 . von tonic saccharose solut . Uso nce of different salts on saponin hemolysis in a medium of > 2 Showing the influc PSPS PEL CRU RTE is AP aie cme: TOR mro[ ddd tri ti rr rss coat’ | (28 O12) 2) ©. B98 es Sasa pooET ||P)? 948 9 a EERE LE et ale ale ale st st st WwW WwW ‘Convaxk | on 00 9 0 0 nan | OR By rere ae OO! Ss mH fh rostn| © 10S © 2 ead pote edilinaed Hw 9 — » ostCHN) | +2 Mita: oR a ESE coo o oo OL ae & FOSteN Fa eee a ial anne ees i oHonn | (FS 5 me gi 2° Si eseee i el cicye icy = S10008H | tae hoa Metacare uae GW el | Sy i ER 4 F ex000'HD | OOF QNOROR Oe RO sae ese Oe greet Sto etree ea | | al OINLIO BN QS SOS Soar ||| | | ie | (>) mm ‘OcHM SRS) S80) OTC Oe ee eee = ° = Fa Xosy | O} OF O50 10; 0 9 OL ie Eas J a SON 2g EN SENS Si) Spo 3 SS Sh ‘ical 5 Oo +> w 2ONtN Ol BO OROFO Ose oe ine s coo oe Oo OL aH et 7ONM ep a eS eal ‘ © © 45 Pa ZONtN ON SS. oS St ie ern i of > > OF =>) B IM 810) OO. 8 OF 91010 Se ee ea i igey OOOO) oO TONON Oa = ipl OS > ‘ Eee Rl ECR CR Nr 2 oH tH IOM OPS KONO! OF O10 10. aaa coo ooo OL aH ae [O8N BAP eaPRFese | anon SLTVS SGADHNDMAOAANHODCOCSO #0 |NOMLOIOS "TVIEHON 9/0) 9 ees lh A sO ASOUVHOOVS AO } SI S0''O0) ON CN HCO 9B. G22 S31 C3 NOILAIOS LNAO udd g Sse oe SS oe 00¢‘29:T ASOUVHOOVS Serna eS ea Ss) Ss Soo) SOS oes Ss Ss“ 4O NOILQIOS INGO Sy Ns SNARE oe aT 5 RIEL ae udd g NI NINOdYsSoDar Ty tae’ ws TABLE 4 TORE novo | | alma sl aE he ale al eee Se | S ‘oom | FP SO ee es SS tt ‘~ S | = woman | FEET Prt rit b it ti iva) go mH he 2 Tan | SSS ie Pete) It I =) Ss oovovoD Oo De Fah te? bette > ‘OS3W PaeFe eee S| © Ss o » | S 'OS*CHN) | = 2) 2) hn eee 3 . >} 1S) HY = Oa CRO TO OOM Orome = OF = ose | aia ee | oo oO : TTT CAO Oo LOMO oon ~*~ Fat § onor™ | atria ee | “> a ooo0van 00S ZRH HOH = MOOO®HO a eq mae 2: || = 5 : ooo Sfolone ooo sss be & NOOODPHO aoa a8 == 2 OAC Same et S mma || © 8b Soe SiS EES pelt | = = OSS SIS tS > ft s O1ULIO BN aie ile lel < SIS) a & rattan | 02" PSe Poe Le Ola eae = S Ow mH OF = Nagst || C7 1S Pee Os Sheela a eee eke ny = Op S roNE PS Ses Sree ease LP le ile ” RS Saat 5 OheN|) ooo e Soe eae ON at ane SS o oo oO oO © oO 4 : ee = ‘ONS aBPEBBSS = aS fONPN |e 'e eee vere lela, > Sree iy Gr Oo SS 2 > S$ oo i 2 IgM OO EN ORS ita Bahia eas > s OO!) OL OMOon Omens sa © Igen aaa ai -sarll yy | em Om, ra fH = IOM 909 99 910 So pe Pe 3 = mH iS ION SOO ONS Oo (OA OwOGr Ri Wh IE | R Bras aonta ishran As SCmMaOHtNO lo ES i — a — a) AOVNOWMATOUTVNON | eats et ere SS Nt oo aso001D 40 pNHOWDONAHOHODRAAAASG NOILOTIOS ING) Wad 9G Ocqoooo nn Hse A etnA AA 000‘0¢ :1 #809015 40 Oo ©. So oro) So orosere SSS NOILOIOS INGO uad SAT aye ape : Menges Pct aS GoM trGsySooNr: |e nS ek ee ee Te 427 THE JOURNAL OF IMMUNOLOGY, VOL, Il, NO. 5 TABLE 5 Showing the influence of different salts on sapotoxin hemolysis in a medium of isotonic saccharose solution add g NI NIXOLOdyvs 428 108d Ge teat ede ie Hb ite ell) Abe ok 007M D9) OO: COO OP OTONOROm ermine £OO78N o.G 6 8 Oo. wo oo ose eum Sie rete la reyevce TOS | > b> ae H HS | I eam ab Sareea 7(ON)SH | 2 PSS Sasa we Eee See a Ae rOStN |) 2) Sire SNe Bess a3 0) 1 re ra ee 74 *OS*(FH.N) Pome ae ee [| {|| SF : costs | COCO SC CeCe eon a Ee YT | EO Om a> od smn | 9° OS SER ERE SY eae | es? ene | EE OREN SS = om + | omoon| O° SSR GEESE III I © 45 =e mmm | 9889 2.2 8 oo a Fee ¢ Sy mom Om soomno| COS SSEE EEE ET II enooxHo| coo oo REEERI III I musy| oo ooo ooo OO REE! | ume | 2 °° © 8 29 608 Eee nosxy| coco ooR8REBEEI II Ram ee rey ee onn| POC SSOKERESEEE II II Gat e one | ©.° °° Pa rire Ehime Sow a | onn| CC OCS RE SESS I 111 r=) Sony mw consen| °° SCR BREE E111 1S ¢ ‘ Onags 4 m| Po fe SS ears (iam ~~ So D7 a ion | °° 9 oe Eee | ee Che Saar mon | S99 OSB RSS 1 ee —_— nm ~~ wo in| eyo oe ee SS i i ~~ “ ©} (0) 1O MOMS) 12) ie {a TORN a > @ > | oot aA xo ieamacdao ipa eee WE Sea eS SOS a St et oS Oo SO Sor oaames at Oo asouvmoovsgo | | NHODONTOHRAAASS NOILQIOS INGO Hdd Q oooocornn nin wAwA nnn 0S2'9:1 ASOUVHOOVS Soo CO Sooo oO; OS SO Clore 40 NOILQIO0S INGO RiEia ie ee SORA Oren e Se ee oe oe TABLE 6 « ‘J uence of di infl Showing the ~ S » 1 ‘OdERN RE NS SF | oOo ~~ oo 0 8 ‘aH | PS SS sel al | — hay AI ES SL a ee : rgonE | CRON pre Sree Sia e COON eovv09s ooo FD OO One Dw Gn SUP eee 4) Ot OER) SP BE oe Sn won| PPR BEES | hs oe ee aera oo ran (CONDI | SUS VENOUS OV Rs eee esse | fferent salts on sapotoxin hemolysis in a medium of isotonic glucose so!ution oor Spam hoo rosa | eisesecanenes (i a ie lea FANQ2(F Hepp moots t OS*(FHN) o Em > 2 a nea 2 Sse ee Ee tOS*tN ee i a ei =) (lene TA tOstEN | SON OO 9 SO Ol OLonoroemin il s op wow soooHo | °° °° OZ REESE | eal < (PER ex0008H | SRS aoe lee bal OIULID Y Coo gle ge a acl | ad o CU Hy 1 fas poet | “°° SSS SS. Te 4 3) 42 tony | a eis ce ee eal) alive uh fe a tONSN | BONS at ees cea n Mle ly ip alata ail 5 a) fet con | °° RS eS vie ui ele ie slieal < ~~ ~ me onn| 2 O@FEBI LIT Vi ill Rae = 1 | SAP Perse ee inh cy bya elt > Sr = B game Oe SENS ae a a Wel ety ul oOo oO Se Jq@N = Pe ere wr lea aie | Sw | om pie? tts seers Fea can Sor moO on|oeger SSS 511111 | DOoDHA Deen ST See eee Se e 5 a Se eS CLS SS SuS Srers N st Oo CO aSo0021D Ao [OR Sty SOOO ECAH ASCO RCS 12 Se Ce NOILI TOS INA) ¥ad 9 ¢ oOo © COs Str rt et oe a te hrs! dc ah i ee 000‘0T:7] as09aT9 oooco sec © Co Soyo toto oC CFO HOUNO MD TOS SND) aS at aot eter tt udd 9 ¢ NI NIXOLOdVS 429 430 TANEMOTO FURUHATA seem to be indifferent. Ammonium sulphate is peculiar in its action inasmuch as in the larger quantities it alters and pre- cipitates the red blood cells and at the same time it inhibits the hemolysis by the saponin and sapotoxin. Furthermore, we see that salts of the same cation with different anions exert a differ- ent influence on saponin and sapotoxin hemolysis and that the same is true of the different cation salts with the same anion. This fact is interesting in view of the experiments of Hdéber (6) and also of Port, who pointed out differences between anion and eation in the influence of salts upon hemolysis. From the standpoint of the tonus of the medium, since the sugar solutions were isotonic, it is also important to study saponin hemolysis in isotonic solutions of the salts instead of in normal solutions. I, therefore, prepared the following isotonic solutions: per cent IN a NOB se ass ow diss weave savas etelw Bare ievese wtonhe ale later ae aide eyesore tere oias eres eae 1.30 |S Un eee rep en emma hater cain o 2.35 INa2S OS eek aoe cle oh crate ca ie ormeeve Bieistouel Sieieekaie Glepecaeiete nakest Lore totetete eto 1.63 RB GIs cis ecco espe ake ee tases Oasis e Balestier oan gene Caer eee eet 1.85 IN AIBEE :o: ass sase toc. Sree is Que tonera ths mers ois eels austere actors aie ee eee rotor 1.58 NaSON« 5 ceeds ese nos eines See ese ooo Ooe eh ea Dee EE eee 1.24 TE ki sig aces hie ete Ore: ein Discs Siecae ae wc wee eos ae Obs OF Cee een 1.14 In the medium of these solutions the saponin hemolysis was tested as in the previous experiment, and it was found that in all of them saponin can act more rapidly than in sugar solution. According to experiments of Poyarkoff (7), the action of sperma- toxin, studied in an electrolyte and a nonelectrolyte medium is much influenced by the viscosity of the environment. Poyarkoff found that when the viscosity is very much increased or de- creased by suitable mixtures of electrolyte and nonelectrolyte, the action of the spermatoxin, which was measured by the time taken to cause the motion of the spermatozoa to cease, is very much reduced. There is, accordingly, an optimal concen- tration of both substances. I have investigated the question whether this fact is also applicable to the action of saponin or sapotoxin. In this experiment the red blood cells of the rabbit were used. The results of the experiment are shown in table 7. es it ae A STUDY OF SAPONIN HEMOLYSIS 431 In table 7 the absolute quantity of saponin is the same in each mixture the variations affecting only the quantity of glucose and of salt. An increase in the quantity of sugar causes an increase of the viscosity of the medium. We see thus that as the viscosity of the medium increases there is an increased inhibition on of the hemolysis by saponin. I have not been able, however, to find any optimal viscosity for the saponin hemolysis such as that found by Poyarkoff for spermatoxin. The hemolytic power of saponin appears to be inversely proportional to the viscosity of the medium. The increase of viscosity may interfere with the velocity of diffusion of saponin into the red blood cells. TABLE 7 bAZA La eae baz? & Deo oe ae aS ees a HEMOLYSIS (1 HOUR OF 37°C.) BARE Zan 12 BAD 4 BOSD Pa a EVOrS i 2 oe ZAz Zoa A708 By Q i) =< 0 ° a m iD ofaga| zee zen | °2ad 1 9 4 4 bel 4 aeons ZOR maR BASDe BD S S = S Agana ons ong Aaaae BA = = = = GAMA w Oa hoa MAMQOn yo = Ss S oS 19 n n Ye) on a oo a wD 1.0 1.0 0 0 110) c c e c 10 0.8 0-2 0 9:1 c ce c ale 1.0 0.6 0.4 0 8:12 c c ec ale 1.0 0.4 0.6 0 US 33 c c ale st 1.0 0.2 0.8 0 6:4 c Cc vst tr 10 0 120 0 IS 5) c c st -- 0 0.8 0.2 1.0 4:6 c c Ww — 0 0.6 0.4 © One c c UE -- 0 0.4 0.6 1.0 2:8 c ale —_ — 0 On2 ls (ORS 1.0 13:9 c ale _ I have also investigated the question whether the relations described above with respect to rabbit’s blood hold as well for the blood of other animals. The results of these experiments are shown in tables 8, 9, 10 and Pt With red blood cells of the pigeon and of the sparrow hemolysis could not be produced with even a 1: 500 solution of jegosaponin. In these experiments, with all of the species of blood corpuscles that are susceptible to the hemolytic action of saponin and sapotoxin the adjuvant effect of electrolytes is apparent. We see, 432 TANEMOTO FURUHATA furthermore, differences in the resistance of the different species of corpuscles to saponin. Such differences have been noted by various workers. Thus, according to Meyer (8), the order of increasing resistance in corpuscles of the different species is: TABLE 8 Saponin and sapotoxin hemolysis with the red blood cells of the horse HEMOLYSIS (AT 37°C) Jegosaponin Sapotoxin 1: 5000 Salt medium | Saccharose medium Salt medium Saccharose medium 30 minutes; lhour |30minutes}) 1 hour (30 minutes}; lhour (30 minutes} 1 hour ey) c c c c c c tr tr 0.8 c c c ec c c — = 0.6 © e c ce c c _ _ 0.4 c c c c c c = = 0.2 E c c zac tr Ww — — 0.1 c c c c = tr _ _ 0.08 c c c c _ tr — _ 0.06 Cc c st c - tr _ — 0.04 c & tr st = — _— = 0.02 w c _ — — = _ = TABLE 9 Saponin and sapotoxin hemolysis with the red blood cells of the pig (HEMOLYSIS (AT 37°C.) Jegosaponin Sapotoxin 1: 5000 Salt medium Saccharose medium Salt medium Saccharose medium 30 ey lhour /30 minutes| lhour |30minutes) 1lhour |30minutes) 1 hour 120 c ec ce c c c c c 0.8 c c c C c c c c 0.6 en 5 c ce ec Cc c c c 0.4 c ec ce ce c c st st OR? c c c ce st vst tr tr 0.1 c c c ce Ww st = = 0.08 c ce c ec Ww Ww = = 0.06 c c st vst tr Ww = = 0.04 ale c _ - tr tr = 0.02 st ale — — {; — A STUDY OF SAPONIN HEMOLYSIS 433 horse < rabbit < pig < dog < sheep < ox. Meyer states, also, that this represents the order of the lecithin-cholesterin coefficients in the red blood cells of those species. However, according to Abderhalden, the order of the lecithin-cholesterin TABLE 10 Saponin and sapotoxin hemolysis with the red blood cells of the sheep. HEMOLYSIS (AT 37°C.) Jegosaponin Sapotoxin 1: 5000 “3 Salt medium Saccharose medium Salt medium u Saccharose medium 30 minutes) 1 hour (|30minutes| lhour /30 minutes; 1 hour |30 minutes 1 hour IO c c c c Cc @ c c 0.8 ec c c e c © c Cc 0.6 c c Cc c ale c tr. Ww 0.4 c c c c st c —- _ On? c é c c Ww ale — — 0.1 c c c Cc tr WwW — - 0.08 c c ale c — tr _ = 0.06 c c tr Ww — — — = 0.04 tr WwW — _ - — —_ — 0.02 tr tr _ _ | — - -— _ TABLE 11 Saponin hemolysis wiih the red blood cells of the guinea-pig HEMOLYSIS (AT 37°C.) Jegosaponin 1: 5000 Salt medium Saccharose medium 30 minutes 1 hour 30 minutes 1 hour 1.0 c c c c 0.8 c @. c c 0.6 c c c c 0.4 c c c c 0.2 c c ale ec 0.1 c e st ale 0.08 ale ye Ae tr st 0.06 vst c = tr 0.04 st c = = o =) i) | ee S| | | 434 TANEMOTO FURUHATA coefficients in the corpuscles is: horse < pig < rabbit < dog < sheep < ox. The results obtained by other*workers with respect to the resistance to the saponin hemolysis are as follows: Rywosch (10): Guinea-pig < rabbit < dog < pig < cat < ox < goat < sheep. Schauzenbach (11): Guinea-pig < man < horse < pig < ox < goat < sheep. Port (4): Rabbit < man < dog < pig < ox < sheep. According to my own experiments the order of resistance of the red blood cells against jegosaponin is as follows: Horse < guinea-pig < rabbit < pig < sheep < pigeon. SUMMARY 1. The hemolytic action of saponin or sapotoxin is, to 2 certain extent, inhibited in a nonelectrolyte medium. This phenomenon is, perhaps, attributable to the increase of viscosity of the medium, which makes the diffusion of saponin into red blood cells the more difficult. 2. Tons of various salts favor saponin hemolysis even in higher concentration, except (NH4):SO., which can alter the red blood cells in higher concentration. BaCl, and CaCl, are indifferent. 3. The resistance of red blood cells against saponin is different in the different species of animal. I desire to express my indebtedness to Professor S. Mita for his kind direction and encouragement during my experiments. REFERENCES (1) E1rsurerR: -Zs. f. Imm., 1909, 2, p. 159. (2) Ransom: Deutsche med. Woch., 1901, no. 3. (3) PorGEs AND NEUBAUER: Biochem. Zs., 1907, 7, 152. (4) Port: Deutsch. Arch. klin. Med., 99, 259. (5) Asantna AnD Momoya: Mitt. a. d. pharmazeut. Inst. d. Univers. Tokio, 395. (6) H6sErR: Biochem. Zs. 1908, 14, 209. (7) PoyarKorr: Compt. Rend. d. 1. Soe. Biolog., 1916, 79, 1150. (8) Meyer: Beitr. z. Chem. u. Physiol., 1908, 11. (9) ABDERHALDEN: Zs. f. physiol. Chemie, 1898, 25. (10) Rywoscu: Pfliiger’s Archiv, 1907, 116. (11) ScHAvuzENBACH: See work of Rywosch. ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS! H. L. ABRAMSON ann HERMAN GERBER From the Bureau of Laboratories, Department of Health, New York Received for publication March 29, 1918 The search for a method of specific inoculation against acute poliomyelitis is not a new one. Shortly after the announcement by Landsteiner and Popper (1), in 1909, of the successful trans- mission of this disease to members of the monkey family, work- ers in this country and Europe began to study the immunity problems of this disease. Flexner and Lewis (2), in 1910, demonstrated that monkeys could be immunized against poliomyelitis by repeated subcu- taneous injections with increasing amounts of a saline suspension of the crude unmodified virus. They injected animals over a period of two and one-half months. About ten days after the completion of the course of inoculations, the animals were in- jected intracerebrally with 2 cc. of a filtrate of a very potent virus, of which 0.05 to 0.1 ce. would prove fatal. The animal tested presented no sign of infection, whereas the control died of poliomyelitis. In a later communication (3), they state that artificial active immunity either by the injection of a single large dose or by series of increasing small doses over a period of time is not uniformly successful. In the former method, some of the animals would develop poliomyelitis as result of the subcutaneous injection, and in both, some animals: so inoculated would not resist the test intracerebral inoculations of rather large doses of a highly potent virus. In the blood serum of animals so im- munized, the presence of neutralizing principles for the potent virus was demonstrated in good concentration. ‘Read before the American Association of Immunologists at Philadelphia, March 29, 1918. 435 THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6 436 H. L. ABRAMSON AND HERMAN GERBER Landsteiner and Levaditi (4) attempted to devise a method for the prevention of poliomyelitis analogous to the Pasteur method for the prevention of rabies. They dried cords for as long a period as twenty-four days. However, while some animals so treated developed immunity without any ill effects as a result of the treatment, other animals similarly treated developed the clinical picture of experimental poliomyelitis. Romer and Joseph (5) thought they had produced immunity in monkeys by the intracerebral injection of a mixture of virus and serum that contained neutralizing substances for the virus. They found that a monkey so inoculated resisted apparently a subsequent suitable intracerebral injection of straight virus. However, that this is not invariably true is evidenced by the experience of Flexner and Lewis and Landsteiner and Levaditi, who had no difficulty in infecting animals that had been previ- ously intracerebrally injected with a neutralized mixture of serum and virus. In fact, Flexner and Lewis have had no dif- ficulty in infecting an animal that had previously resisted a suf- ficient intracerebral dose of straight highly potent virus. We can confirm these findings concerning neutralized and straight virus, and further than that we have re-infected, by suitable intracerebral inoculations, two animals that had been paralyzed over a year ago, and which presented residual palsies at the time of injection. These animals succumbed as promptly as control animals, which eceived a third of the dose. This fact will be pointed out later. Flexner (6) stated in 1910 that while the results in artificial active immunity thus far achieved were encouraging, our knowl- edge at that time was not sufficient to render those results of practical value. This statement was the stimulus for the work which has engrossed our attention for the past year and a half. The necessary requisites of any suitable method of artificial immunization are: 1. That the method shall protect against any reasonable ex- posure to the disease for which the individual is immunized. 2. That the inoculations in themselves shall be absolutely harmless. ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 437 3. That the method itself shall not be too cumbersome or prolonged, so as to render the production of immunity tco slow to be of practical value. These have been the conditions with which we have attempted to comply in our efforts to arrive at a suitable method. After an extended and rather discouraging experience with young rabbits, we turned our attention to the use of monkeys of the rhesus variety in our efforts to devise a practical method. The striking similarity in the characteristics of the virus of rabies and of that of poliomyelitis impressed us so as to cause us to work with the idea of attenuating a highly potent Rocke- feller strain. Attempts were made to this end with two methods, one with chemical means and the other by subjecting the virus to the action of heat. MODIFICATION OF POLIO VIRUS BY CONTACT WITH 0.5 PER CENT FORMALDEHYDE Cummings devised a method of anti-rabic treatment in which he put a 2 per cent emulsion of fixed rabic virus in contact with 0.5 per cent formaldehyde for four hours in the ice-box. At the end of this time, he dialyzed the formalin from the mixture through collodion sacs into distilled water until the cord emulsion failed to give test for formalin. The material was then inocu- lated daily into the rabbits to be protected in increasing doses. We applied this method in our attempt to chemically modify polio virus. We used a 10 per cent emulsion of the cords and brains of monkeys dead of highly virulent poliomyelitis virus and made it up fresh for each injection. This 10 per cent emulsion was kept in contact with 0.5 per cent formaldehyde for four hours. It was assumed that this contact would kill the polio virus. However, our experience proved to us without any chance for doubt, that it did not kill the virus. The protocol follows: Experiment. Macacus rhesi nos. 89 and 90 were injected subcuta- neously as follows: 438 H. L. ABRAMSON AND HERMAN GERBER March 28, 1917, 10 cc. March 30, 1917, 15 cc. Aprilt 2 O17, 5 ce: April, 5, 1917, 5" ee, On April 7, 1917, no. 90 appeared ill; refused food; seemed to be tremulous; showed no sign of paralysis. April 8, 1917, no. 90 died during the night. Postmortem examina- tion showed nothing of note in the abdominal or thoracic viscera. Section of cord showed some swelling and reddening of the gray matter. Microscopic examination of the cord showed slight perivascular cell infiltration; moderate diffuse cell infiltration; considerable nerve cell degeneration and neurophagocytosis; only slight changes in the meninges; that is, all the classical pathologic findings of poliomyelitis. Macacus rhesus 89 remained well until April 26, 1917, when it ap- peared to be ill. It would not feed well but it showed no paresis of any kind. The animal died during the night. Microscopic section of the cord showed the lesins of poliomyelitis. These animals presented a rather unusual type of infection, one dying within ten days and the other within one month after the institution of the subcutaneous injections. The type re- sembles very much that described by Flexner as the ‘“‘marantic”’ type of monkey poliomyelitis. In this type, animals may be sick over a longer period of time than the animals in the above experiment, yet they do not present flaccid paralysis. Our results with this particular method were decidedly not encouraging, and as monkeys were very scarce, we turned our attention to the possibility of attenuating the virus by heat. MODIFICATION OF VIRUS BY HEAT It has been demonstrated that poliomyelitis virus is rendered inert by exposing it to a temperature of 50° to 55°C. for one-half hour. With this fact as a basis, it was decided to expose a highly potent virus obtained from the laboratories of the Rockefeller Institute to heat in two ways. One method aims at a graded increase in the virulence of the material inoculated, paralleling aiter a fashion the use of increasingly more virulent cords in the- production of immunity to rabies. This method consists of the ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 439 subcutaneous injection on four successive days of 5 ce. of a 10 per cent emulsion in saline of brains and cords of monkeys that have recently been paralyzed by intracerebral injections of highly virulent virus, heated as follows: _PTCATIR CLES 28 22 Sa mi > ae ae PT 55°C. for one-half hour PASCLCLE CLR Ver vyo cla ahs. nits eittcape ie isiaicietanes Pelco 55°C. for one-half hour PACING As ers see bends eh MN tee 45°C. for one-half hour BURGER CAVE Nete's << suisis eee a> ee bidlasle cape aentee 37°C. for one-half hour On the fifth day, the animals received 5 cc. of 10 per cent emul- sion of the virus unmodified by heat. The other method of injection used consists in the subcutane- ous injection on ten successive days of 5 ec. of 10 per cent emulsion in saline of brains and cords from monkeys recently paralyzed, which had been heated uniformly to 55°C. for one-half hour.’ This method corresponds to the use by Semple of subcutaneous injection of killed rabic virus in the Pasteur treatment. Such a method as this immediately removes the dangerous possibility of producing poliomyelitis in the treated animal, inasmuch as the heated material is unable to infect a monkey when injected in large amounts by the intracerebral route. In other words, the virus is apparently killed. In testing the protective value of these two methods of vac- cination, we wished to determine two facts: 1. Are the animals so treated capable of resisting a multiple lethal dose of the virus intracerebrally? 2. Does the blood of animals so treated contain neutralizing principles for a highly potent polio virus? Three weeks after the completion of the series of subcutaneous injections, three animals of each of these series were bled from the heart. The serum obtained from these bleedings were put into contact with a 5 per cent emulsion of highly potent Rocke- feller virus in the proportion of one to one for two hours at 37°C. and twenty-two hours in the ice-box. Then 0.6 ce. of each mix- ture were inoculated intracerebrally into each of six normal ani- mals. At the time of these injections, the vaccinated animals were tested with the intracerebral injection of 0.15 to 0.3 cc. of 440 H. L. ABRAMSON AND HERMAN GERBER the same virus emulsion used in the neutralization test. Control animals were also inoculated intracerebrally with the same amount of the same emulsions used in testing the vaccinated animals and in the neutralization tests. Charts 1, 2 and 3 give in detail the results obtained by these methods of modifying the virus. It will be noted that Macacus rhesus 81 of the 5-injection series had received a similar series of injections, and that Macacus rhesus 80 of the 10-injection series had received 5 injections of killed virus two and one-half months previously. At this time we were unable to obtain new monkeys for controlling our test inoculations and for our neutralizing ex- periments. We thought that if immunity in poliomyelitis lasts no longer than that of rabies in rabbits, it would be well to sub- ject them to another series of injections. Chart -1 shows that of five animals treated by the 5-injection method, three survived the test intracerebral injection of 0.3 ce. of an emulsion of a highly potent virus, 0.05 to 0.1 cc of the fil- trate of which is fatal tomonkeys. The two animals, nos. 91 and 97, that succumbed to the injection exhibited paralysis on the sixth day after the test injection. The normal control animals nos. 19 and 23 showed paralysis on the fourth and fifth day repectively. Macacus rhesus 24, an animal that had survived an infection from virus of the 1916 epidemic, about one year ago, and which at this time presented a residual diplegia, was also injected in- tracerebrally with 1 cc. of the same emulsion. This animal ex- hibited paralysis of the arms six days after the inoculation and died on the seventh day. Animals that have survived an attack of the experimental disease are supposed to have a very high degree of immunity, yet here was such an animal that suecumbed almost as readily as a control to only three and one-third times the dose used in the controls. This result argues well for the virulence of the material used in the test. The virus neutralization table of this series shows that of the blood of the three animals bled, all show the presence of neutrali- zation substances, but in varying degrees of concentration. The serum of no. 81 neutralized completely in the proportion of one ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 441 aA0g® SV SI6I ‘OT Azenuer pot ‘ST6I ‘g Arenuvp pozAyered | LT6T ‘FI tequieseqd | 26 aAOqe SB SI6I ‘OT Atenuer potq 90 | LZI6T ‘Ef Jequieseq | 26 ez puv GT OU sforjzuoo snatA | SII ‘fF ArenuEP pozdyeivg | LIGI ‘FI 1equiecsq | 86 | 9°0 | LTGT ‘ST Jequieced | 96 DAOGB S¥ GY ‘OU [O1}ZUOD SNATA ‘Tjom—oatly | LTT ‘2z tequuaydeg | —¢ | 90 | LTT ‘9% Tequieydeg | 18 mr) z g on °} od ECSERE Bes SUUVNaU aWOOLDO NOILVIOQOONI ao giva | 5 | e q 2 aagta ig Pic Ha ® Hx addy Bins Heo ZO ie) Z 0 e a & snwia 139)9 fayooy quao sad ¢ yand [ 02 wnuas und J ‘uoysodorg *8ar1as 2a0QD ay} fo shayuou fo wnwas oY} fgg ‘uoynzypujnau snivA ‘soisyed jenpis -01 YPM Ory -oajut snoraoid WOIf patoaooay |/TET ‘TZ oquiedseq ped | LIGT ‘0% 1equiedeq_ | 00'T | ZT6T ‘pL Joquie.eq VG * jorquog |ZTeT ‘Tz toquIa0eq polq | LIGI ‘61 Lequiessq | ¢§°0 | LIGI ‘PI oquisseq &% [o1}UOD |LTGT ‘TZ Tequieseq ped | LIGI ‘Sf tequesoq | €°'0 | AI6T ‘pT Joquieseq 61 LIGI ‘1% Loquiasaq ped | LIGI ‘OZ tequieseq | ¢€°0 | LI6I “PI Jequteo0q | LT6T ‘EZ JoqMOAON | 16 TT24k— 9ATTYV auoN | €'0 | ZIGT ‘FT Jequisoeq | LT6I “EZ 1equIOAON | 96 LIGL ‘TZ equieseq ped | LIGT ‘OZ Toquiesaqy | €'0 | LIGT ‘FL Joqutsoeq | AT6T ‘ES JoqUIEAON | 16 [eS eA ouoN | $0 | LIGT ‘FI toquIoo0q | LT6T “EZ JoquIOAON | 06 TS “OU IJ [OLPUOL) LIGI ‘G 1040990 patd | LT6I ‘E 10q0990 | 0 | LTET “Lz raquiezdag GL []24—eATLV auoN | 2°0 | LI6T ‘2% tequiezdag | LT6T ‘ST Jlaquieydeg | Tg ‘92 ; : SUUVNaH aWOOLAO SISAIVUVd 3 OM Sema? -QNOONI ISG JO ALVA |NOILVTIOOONI GAILINGAGUd sosaHnu soovovn (worspnwa qua sad g U2 sn.na L9210fay90RT) P19 pu UWD.g pajpj0..a9h76 ‘pasn ipuaony “PayDOYUN PUD “DOLE ““OoSh sede 999 0] unoy fyDYy-aUo payway snirA *sarsas UoYseluy-g *squawmrsad xa U01I9}01q 1 LU VHO 442 H. L. ABRAMSON AND HERMAN GERBER toone. The sera of nos. 96 and 97 extended the incubation period ‘of four and five days of the controls nos. 19 and 23 to twenty-one and twenty-five days respectively. This prolongation of the incubation period of highly potent virus seems to us to be an indi- cation of the presence of neutralizing substances. We have been compelled to apply a rather severe test on account of lack of ani- mals. The proportion of 10 of serum to 1 of virus has been con- sidered sufficient for the determination of neutralizing principles. How great a quantity of these substances is developed by this process of injection, remains to be determined in experiments already planned, and which will be carried out as soon as monkeys are obtainable. No normal monkey serum plus virus control was used in these experiments, as it has been quite firmly estab- lished that the blood of normal monkeys does not contain neu- tralizing substances for poliomyelitis. Another series of three monkeys were subjected to the 5-injec- tion method, except that in these the inoculations were made up from glycerolated cords only, and that the intracerebral test in- oculation was 0.15 cc. of the 5 per cent Rockefeller virus. This was done with the idea that possibly cord material might produce a higher degree of immunity. The test dose was reduced to one- half, because 0.3 cc. appeared to be much more than enough to bring down the controls. Chart 2 tabulates the results of this series. This shows that two of three animals survived the intracerebral test dose of 0.15 ec. The third, no. 27, showed the first symptoms on the sixth day after the test inoculation. Both controls nos. 73 and 45 showed the first symptoms on the fourth day after the test inoculation. An additional indication of the potency of the testing virus is shown in the fact that monkey 10, which had survived an ex- perimental infection from the virus of the 1916 epidemic, with a residual paralysis of the right lower, was reinfected fatally with 0.5 cc. of 5 per cent Rockefeller virus. Despite the high degree of immunity conferred on the animal by recovery from the dis- ease, this animal succumbed readily to re-infection with cnly three times the dose of the control animals in this series, and to less than twice the dose used on the controls in the first series. * ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 443 Gp puUv EJ “SOU S[O1}WOD SATA SI6I ‘Z Yue pod ‘S161 ‘I youwpy pozAywavg | St6r ‘Et Axenaqeg | gf | ¢°0 | ST6I ‘gt Sxenaqag | 99 CH puB Gy “SOU S[O1JUOD SNATA SIGI ‘F Yorvyy porq . “SI6I “€ Youvyy pozAywreg | ST6T ‘6T Arensqog | TL | 90 | SI6T ‘ST Areniqag | 22% Cf PUB EJ “OU STOI}UOD SATA To“ pus oalTy | SI6T ‘6T Areniqaq | F | GO | STIGT ‘ST Azeniqaq | ge *99 one ox zee des | dq BAS SHUVNAY GNOOLAO NOILVTOOONI dO Gilva % q qd = q 4 aala a is 2 | BBO es oy oo ° : Zz Ssnsia sanjafayooar quaa sad g qund J 07 wnusas qund J ‘uoysodotg 891108 aaoqn ay} fo shayuow fo vias ayy ig “woynzyvsjnau snirA uot} -dayur snotaoid WOIf paldA0d0yT SIGT “OZ WIV PP | SI6T ‘ST youvyy G0 | S161 ‘IT yore Or JOLIWOD | ST6T ‘9z Arvnagey porq | STI ‘gz Arenaqeg | CTO | SIGT ‘6T Arenaqoag CF JO1}UOD |SI6I ‘9g Areniqay potq | giT6r ‘gz Areniqay | STO | SI6I ‘61 Arenaqay el ]]94 pues salty suON | STO | SI6T ‘61 Avenaqog SI6I ‘6c Aisnuee 99 SI6I ‘9g Arenigeay poztioyyg | ST6I ‘¢z Arenaqgeq | STO | ST6T ‘“6t Arwnaqaq | SI6r ‘6g Axenuee | 1z ]]94 pue sar[y ouON | ST'O | ST6I “61 Arenaqoq | ST6T ‘6g Arenuve | g¢ ‘99 a ee es a ae x ere ee hala rank ing el Be | onoGhe aicn ao imnvace Noamvanao.te aiisennme E : oq dd VAD no (uoisjnwma quad sad gG UL SNura saqjafayI0y) pod pajyn}0190h76 “pasn yorsaqoyy “PaIDIYUN PUD "DLE “Oo9% “Oo9F “OoGG 07 noy fypy-au0 paynay snsvA ‘saruas Uuoyoalur-g ‘*sjuamrsadxa W014NI9}01g @ LUVHO 444 H. L. ABRAMSON AND HERMAN GERBER Of the sera of the three animals of this series, one, no. 36, completely neutralized the virus in the proportion of one of serum to one of virus. The sera from nos. 27 and 66 prolonged the four day incubation period of the controls nos. 73 and 45 to° fifteen and twelve days respectively. As in the first series, the sera from the treated animals contain neutralizing substances, but in varying degrees of concentration. The controls, nos. 73 and 45, showed the first symptoms of the fatal disease in four days, though the intracerebral inoculation was only 0.15 cc. of a 5 per cent emulsion, or one-half the dose used in the first series. | Three other monkeys were subjected to the injection of in- creasingly virulent material. In these, the tests were not com- pleted, and therefore they were not included in the tables. The protocols follow: Experiment 1. Macacus rhesus 57 received subcutaneously injec- tions of 5 per cent Rockefeller virus as follows: February 13, 1917, 10 cc. heated te 455°C. for one-half hour. February 16, 1917, 10 cc. heated to 45°C. for one-half hour. February 20, 1917, 5 cc. heated to 37°C. for one-half hour. February 24, 1917, 5 cc. unheated 5 per cent virus. This animal was not bled for the determination of neutralizing sub- stances, and remained well till time of intracerebral test injection. March 20, 1917. Injected intracerebrally with 0.6 cc. of a 5 per cent emulsion of the Rockefeller virus. This was twice the dose used in the series shown on chart 2. March 26, 1917. Left hind leg is weak. Does not attempt to get up. Has tremors. March 27, 1917. Completely paralyzed. Abdominal respiration. Etherized to death. Postmortem examination shows typical changes of poliomyelitis in the cord. Comment: This animal was not protected against a very large dose of a virus of high potency. However, the vaccination in itself had no harmful effects. Experiment 2. Macacus rhesus 94, an animal that had some months previously been injected with an emulsion of rat fleas obtained from houses in which had occurred eases of poliomyelitis and that had shown no symptoms as result of such injection, was inoculated with a 10 per cent emulsion of Rockefeller virus as follows: ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 445 April 20, 1917, 5 ec. heated to 55°C. for one-half hour. April 23, 1917, 5 cc. heated to 45°C. for one-half hour. April 25, 1917, 5 ec. heated to 37°C. for one-half hour. April 30, 1917, 5 ec. unheated. This animal had not been well for some time. Had been coughing and was somewhat emaciated. On May 6, 1917, animal died. Postmortem showed generalized pulmonary tuberculosis. Microscopic examination of cord showed no changes indicative of poliomyelitis. Experiment 3. Macacus rhesus 100 was injected subcutaneously with a 10 per cent Rockefeller virus as follows: May 10, 1917, 5 ce. heated to 55°C. for one-half hour. May 11, 1917, 5 ec. heated to 55°C. for one-half hour. May 12, 1917, 5 ec. heated to 45°C. for one-half hour. May 14, 1917, 5 cc. heated to 37°C. for one-half hour.. May 15, 1917, 5 cc. unheated. May 29, 1917. Animal is well. As result of cardiac puncture to obtain blood for the purpose of testing for the presence of neutralizing substances, the animal died. Postmortem showed hemopericardium. Microscopic examination of cord showed no lesions of poliomyelitis. % June 1, 1917. A mixture of 0.2 cc. of the serum of Macacus rhesus 100 and 0.2 cc. of 5 per cent Rockefeller virus that had been in contact for two hours at 37°C. and twenty-two hours in the ice box, was in- jected intracerebrally into Macacus rhesus 51. This animal survived the inoculation and is alive and well. Comment: The vaccination in itself was harmless and the serum of vaccinated animal contained neutralizing substances. SUMMARY OF RESULTS OF 5-INJECTION METHOD Eleven animals were subjected to injections of the highly potent Rockefeller strain modified by heat as follows: First heated to 55°C. for one-half hour; second heated to 55°C. for one-half hour; third heated to 45°C. for one-half hour; fourth heated to 37°C. for one-half hour. The fifth injection was made up from glycerolated material from recently paralyzed animals and injected without previously being subjected to heat. These injections were administered subcutaneously on successive days. 446 H. L. ABRAMSON AND HERMAN GERBER Not one of these animals exhibited any untoward symptoms that were recognizable by careful observation as a result of the course of injections. Five out of eight of these animals tested intracerebrally re- sisted successfully the very reasonable test injection of from three to six fatal doses. The potency of the testing virus in 1910 was such that 0.05 to 0.1 cc. of a Berkfeld filtrate was sufficient to produce the fatal disease in monkeys. It has been passed through many additional generations since and it is safe to assume that the virulence has mounted considerably in the past eight years. ; The sera of seven of these animals tested in the proportion of 1 part serum to 1 part 5 per cent virus contained neutralizing substances. Three completely neutralized the virus in this pro- portion and four prolonged the short four to five day incubation period as shown in the controls to from twelve to twenty-five days. 10-INJECTION SERIES WITH VIRUS KILLED BY HEAT Chart 3 shows that of three animals vaccinated subcutaneously by daily injections of 5 cc. of 10 per cent emulsion of Rockefeller virus heated to 55°C. for one-half hour, on ten successive days, only one Macacus rhesus 1 survived the intracerebral test in- jection. This animal recéived 0.15 ce. of 5 per cent Rockefeller virus intracerebrally three weeks after the completion of the process of vaccination. On the fourteenth day after the test inoculation, this animal began to show weakness in the legs, which progressed slowly, so that on January 5, 1918, both arms and legs were completely paralyzed. After this, progress of paralysis ceased. On Janu- ary 10, animal was as lively as he could be under the circum- stances. ‘The eyes were bright and he eagerly ate fruit that was held up to his jaws. The animal, however, became so infected with body lice that he was etherized to death on January 20, 1918. Two others, Macacus rhesus 80, died one week after the test intracerebral inoculation of 0.3 cc. of 5 per cent Rockefeller Sa - — ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 447 ie ae ek ee eS ee LIGT ez sns_ | ‘Tg requioo0qq perp ‘LT6T -oyI SnNovoByT [O1}UOO SNATA ‘Te soquieoeq pozkyeied | LT6T ‘1g lequisv9q | €8 9°0 LIGBL {0% 1equIeD0q_ |Z ez sns SIOMOT [VNPTSOL TFTA OATTB -oyi snovovyyy [O1}UOD SNATA | ‘SI6T ‘6 Arenuve pozApeieg | LIGT ‘ZG toquieseq | OOT | 90 LIGBL ‘1% Lequieoeq, 1 g sns -oyt snovovyy [OL}UOO SNATA JOM pue oarTy | LIGT ‘L I8q0}90 | 66 9'0 LI6BI ‘9 19q039Q | 08 *09 g > 2 Sree: | Bree FS ane | ee Bes SUUVNAL WWOOLNO NOILVTOOONI dO Aiva 3 wield q z aa1a a be Ban (EB ads ) = i} ONG) UG Spe Zz y i} sna sajjafeyooy yuao sad g jund J 07 wn.«as quod J ‘UuoyLiodoig ‘sartas aaogn fo shiayuou fownsas hg “uouvzypsnau sn A “OAISN[OUL TEL ‘Og euNL-9z oUNL “MOY J[Vy-oUO LOZ “HCG OF popwod SNITA Usd od QT Jo SUOTPOLUT OA Jo Solos poATooOI PVH t Gapas [ ‘SOU IOJ [OIJWOD |ZTGT ‘8s Joquiovoq ped | LT6T ‘yz taquieoeqy | STO | LT6T ‘TS T9qures9q] GG LT@L ‘9z roquroseq, perl | LIGT ‘2Z Loquiovoq | GTO | LIGT ‘Te toqutoood | LIT ‘0g ToquIBAON | @ ST61 ‘og Arenueep pozt S39] -19Y}9 {patoaoovoy | puw SUI fo sishjeivg | S16 ‘g Aaenuee | GTO | LIGT ‘Tz raquiaoeq | LT6T ‘Og ToquIeAON | T 08 “ON OF OrZUO)Y FI 10q0990 pod | LIGT ‘ZI 1090790 | €'0 | ZT6T ‘L 1040990 #8 FI 1940990 peld | LIGI ‘ET 1090790 | € 0 | LTT ‘4 4040990 | LT6T ‘FI aoquiaydeg | Ost “00 ss c eo a NOILV'T agtLaTawod ae SSL V FLUE. Coron ene) SISK IV ad 8 ° " -OOONI ISaL 40 ALYa NOILVIQOONI DAILNGAGUd is La gq dq 24 aD (uorsynwma yuao dad GUL sn.tqa Layafayooy) P4oo pup UWAG paqnj0.1a0h yb “pasn 4011090 ynoy fyny-euo Lof "Do9G OF payway sniva pUd0 wad Of yun shop aarssavons uo suoyoalur Ua, "s7UWIUWIdLI WO2499}0Lq € LUVHO 448 H. L. ABRAMSON AND HERMAN GERBER virus and Macacus rhesus 2 died one week after an intracerebral inoculation of 0.15 ce. of 5 per cent Rockefeller virus. Macacus rhesus 84, the control for no. 80 also died 1 week after the intra- cerebral injection of 0.3 ec. of 5 per cent Rockefeller virus. Macacus rhesus 25, the control for nos. 1 and 2, died one week after the intracerebral inoculation of 0.15 ee. of 5 per cent Rockefeller virus. The blood of all three of the animals vacci- nated with killed virus contained neutralizing substances but in varying degree of concentration as virus neutralization table on chart 3 indicates. Macacus rhesi 99, 100 and 83 were each injected intracere- brally with 0.6 cc. of a mixture of equal parts of 5 per cent emul- sion of Rockefeller virus and serum obtained from Macacus rhesi 80, 1 and 2 respectively. Rhesus 99 is alive and well. Rhesus 100 began to show first symptoms on the eighteenth day after the inoculation. The next day complete paralysis of both legs. ‘This did not progress. The animal is alive with residual palsies of both lowers. Rhesus 83 showed first symptoms on the tenth day after the inoculation and died the same day of respiratory paralysis. The control animals in this series, Rhesi 84 and 25, which received 0.3 ec and 0.15 ce. of 5 per cent Rockefeller virus in- tracerebrally respectively, died of the infection with an incuba- tion of five and six days respectively. SUMMARY OF RESULTS OF 10-INJECTION SERIES Of three animals so treated, two succumbed to the intracere- bral test dose as promptly as the controls. The third exhibited paralysis after an incubation of fourteen days. This animal, Rhesus 1, recovered, but with complete paralysis of arms and legs. The infection of the control animal in this case had an incubation period of six days. Of the three sera tested, one, that of Rhesus 80, neutralized completely in the proportion of 1 part serum to 1 part 5 per cent Rockefeller virus. Another, that from Rhesus 1, delayed the incubation period to eighteen days with recovery, but residual paralysis. The serum from Rhesus 2 prolonged the incubation ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 449 period to ten days. None of these animals showed any ill effects from the method of injection. It would seem from the above results that the animals sub- jected to ten subcutaneous injections of virus killed by heating at 55°C. for one-half hour, did not offer as great a resistance to the intracerebral test injection of from three to six fatal doses, as those that received the 5-injection series. We, therefore, endeavored to increase this resistance by injecting virus heated to 50°C. for one-half hour for seven injections and finishing off with three injections of virus heated to 45°C. for one-half hour. Two animals were so injected. The protocols follow: Macacus rhesus 46. January 18, 1918. Received daily subcu- taneous injections of 5 cc. of 10 per cent Rockefeller virus in saline heated to 50°C. for one-half hour for seven days. January 26, 1918. Received daily subcutaneous injections of 5 ec. of 10 per cent Rockefeller virus of saline, heated to 45°C. for one-half hour for three days. January 29, 1918. Series of injections completed. January 30, 1918. Animal is quiet. Does not feed well. Seems tremulous on moving. January 31, 1918. Completely paralyzed. February 1, 1918. Autopsied. Sections show poliomyelitis. Macacus rhesus 7. January 18, 1918. Received a course of injec- tions similar to that of Rhesus 46. . January 29, 1918. Course of injections completed. February 6, 1918. Animal appeared sick. Did not eat his food. Was quiet. No sign of muscular weakness. « February 7, 1918. Was found dead in the cage. No sign of pa- ralysis had been observed. Autopsy showed pneumonic consolidation of lower right lobe. Section of cord showed slight but unmistakable signs of poliomyelitis. Comment: From this experiment, it is apparent that heating virus to 50°C. for one-half hour does not sufficiently attenuate to exclude the dangerous possibility of infection as a direct result of the injections per se. One animal developed frank symptoms of paralysis on the eleventh day after the beginning of the series of injections. The other died 19 days after the first injection of lobar pneumonia with lesions of poliomyelitis in the cord. 450 H. L. ABRAMSON AND HERMAN GERBER DISCUSSION We feel that this work opens up a field for the practical appli- cation of specific preventive measures in poliomyelitis. We have - subjected eleven monkeys to the 5-injection series without any ill effects to the animals as a result of the injections. The pos- sibility of harmful effects from this method, if applied to human beings, is, of necessity, less than when applied to monkeys, inas- much as this virus has been adapted to monkeys for the past nine years and by continuous passage through this animal, has become highly virulent for the monkey. The period of time required for this series of injections is a short one, only five days, which, of course, would render it highly practicable in time of epidemic, when a rapid method is to be desired. This method confers a substantial degree of immunity as shown by the resistance to multiple intracerebral doses of this highly potent virus and by the presence of neutralizing substances in the serum. it is a well established fact that persons who recover from poliomyelitis have in their blood neutralizing substances forthe virus of poliomyelitis and that in all probability, the presence of these substances is an indication of immunity to re-infection. The same can be said of monkeys that recover from the experi- mental disease. If this is so, then why should not animals that have been artificially made to produce such substances and some of which have been made to resist multiple intracerebral injec- tions of most virulent material, be considered immune? The natural disease is far less serious than is the intracerebral infec- tion of experimental poliomyelitis. In the former condition, the body fluids have an opportunity to combat the virus of the infec- tion at the portal of invasion. In the experimental disease, the defensive forces of the body are circumvented. It would, there- fore, be true that an animal body, which contained a large reser- voir of these anti-poliomyelitis substances artificially produced, would be able easily to take care of the comparatively mild infec- tion of human poliomyelitis. ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 451 REFERENCES (1) LANDSTEINER AND Popper: Zeitschrift f. Immunititsf. 1909, 2, 377. (2) FLEXNER AND Lewis: Jr. A. M. A., 64, May 28, 1910. (3) FLEXNER AND Lewis: Jr. A. M. A., 55, August 20, 1910. (4) LANDSTEINER AND LevapiTI: Compt. rend. Soc. de Biol., 1910, 68, 311. (5) R6mer anp JosepH: Miinchener Medizinische Wochenschrift, March 8, 1910, 505. (6) Fuexner, S.: Jr. A. M. A., 1910, 55, 1105. THE JOURNAL OF IMMUNOLOGY, VOL. II, NO.6 o “6 aes EXPERIMENTAL POLLINOSIS! PRELIMINARY REPORT HENRY L. ULRICH From Division A (University of Minnesota Medicine) City Hospital, Minneapolis Received for publication August 28, 1918 That hay fever is a form of hypersensibility to pollen proteins is conceded by everyone. The mechanism of this form of hyper- sensibility, likewise the phenomenon of its desensitization with pollen extract (phylactically or prophylactically administered) are still mooted questions. Cooke, Flood, Coea’s (1) definition of hay fever as a ‘‘clinical symptomatic expression of local hypersensitiveness” cannot hold. The skin and conjunctional reactions in hay fever subjects are certainly extranasal. Sewall’s (2) criticism that no local mani- festation of hypersensibility can occur without the background of a general hypersensibility is amply justified. Up to now animal experiments to produce clinical hay fever have not been successful. Koessler (3) reports passive anaphyl- axis in guinea-pigs. Cooke, Flood and Coca (4) and myself? have failed to produce active or passive anaphylaxis with pollen extracts. My experiments to produce a passive anaphylaxis were made with patients’ blood out of season. It is possible that Koessler’s experiments were made in season. At any rate the assumption (5) that there are no circulating antibodies in hay fever patients should be made with some reservation. In an effort to produce hay fever in animals, an entirely new method was used which proved satisfactory enough to Justify its publication. A group of animals were sensitized to pollen pro- _ 1Read before the annual meeting of the American Association of Immunolo- gists, Philadelphia, March 29, 1918. 2Unpublished experiments. 453 HENRY L. ULRICH 454 QYNIT- SAV UNS ONY PO StRUSMAS O SLL/ LOWOOLG T7709 QWIAT + SHVTKHIENY QV O eo NT TH PALYOIN SIS NIUAOF MLLFO -SINY TAI NY + atl Ly vy, L-YUOWS (§) wf Lunes (5) FJ Lon os 3 | Muang () HR/ Lhe UndA 3 ee ES i MUYY(S) O es Slhnio ‘ a ae weve Ze L- UMS (§) J LOSS 1 alot “ae L-LYf7-(6) Y SISONIT10d IV.LNSEWIaSdxS EXPERIMENTAL POLLINOSIS 455 tein by the injection of a suspension of pollen in salt solution, intraperitoneally. One series received 1 mgm. of pollen in 1.0 ce. of salt solution respectively; the other series received 25 mgm. suspended in 1.0 ce. of salt solution respectively. Ten to twelve days later insufflation of pollen in the nares was begun. This was repeated every other day for a long period of days. After each insufflation notes were made of the clinical behavior of the animals. The animals were watched usually for one hour after the in- suffation. Notes made were on each individual animal. The curves on the chart portray the indices of reactions on successive days of the groups. The curves, therefore, represent cage reac- tion and are merely charts of clinical symptoms measured graphi- cally. The pollen used in these experiments was collected during the 1917 season. It was cleansed and washed in pure acetone, dried and stored in sterile vials of two grams capacity. Another group of animals were sensitized to horse serum—l ce. intra- peritoneally. Ten days later instillation of horse serum in the nares was begun and repeated every other day for a long period. This group virtually was a control to the pollen series. Some experimental data with this method had already been published by Sewall (6). Another group of animals was not previously sensitized but was subjected merely to pollen insufflation at two days intervals. This group again was controlled by a horse serum group, the serum being instilled in the nares at two day intervals without previous sensitization by the intraperitoneal route. Lastly, a group of animals was subjected to insufflation of pure starch. This group was to act as a control to the pollen group in order to throw out the presumption that sneezing was caused by the local irritation of dust particles (7). Microscopically, the surface of starch granules and pollen granules are not at all alike. The question of irritation may be settled at this point by the fact that in these experiments, the pollen animals, which reacted by sneez- ing, always exhibited a latent period after insufflation. In the starch animals there were no evidences of sneezing or disagree- ‘ able sensations at any time. 456 HENRY L. ULRICH It may be also added, that a similar series of experiments were started at the same time with rabbits. The rabbit groups were conducted for six weeks. At the end of that time the experiments were stopped because at no time was there any evidence of hyper- sensibility obtained by the methods employed. In other words the nasal route as a method to demonstrate clinical hay fever (hyper- sensibility) in rabbits is not possible. The symptoms in guinea-pigs obtained by these methods were those closely resembling clinical hay fever inman. After a latent period of five to twenty minutes following exposure to pollen, sneezing, lachrymation, itching of the nares and the body were obtained, this being often followed by excitability, or depression, increased respiratory rate, deefication and urination. As the days progressed, nasal and bronchial stenosis were observed with rales in the chest, moist and rhonchus in type, but at no time was there prolonged expiratory breathing, so typical of asthmatic attacks. In the serum animals, similar clinical symptoms were observed but never so severe nor so consistent as in the pollen groups. During the period of the experiments, the animals thrived and grew in size and weight. No controls in growth or weight were made. Several of the females became pregnant and gave birth to healthy young. Experiment 1, chart A, pollenI. On November 24, 1917, three guinea- pigs average weight (575 grams) received 1 mgm. of pollen suspended in 1 cc. of salt solution, respectively. December 6, 1917, each animal re- ceived a nasal insufflation of pollen by means of a medicine dropper. The animal was held gently on its back by an assistant. One nostril? was held closed, the other mostril received the tip of the dropper, which contained a column of dried pollen. The animal thus inhaled the dust aided by a gentle compression of the head of the dropper. The other nostril was treated in the same way. This method was repeated every other day. Fifty-three exposures to pollen were made, covering a period of one hundred and six days. Including the period of sensitization the whole time involved was one hundred and sixteen days. The first insufflation called forth reactions resembling mild hay fever, sneezing, ’Guinea-pigs will not breathe through their mouths except in the last gasps of anaphylactic shock. . Rea os EXPERIMENTAL POLLINOSIS 457 itching of the nares and the body, and lachrymation occurred. A latent period varying from five to twenty minutes was noticed after each insufflation. After twenty minutes the symptoms usually were at their height. Spasmodic sneezing, besides the itching, lachrymation and nasal stenosis were most striking. At no time were true asthmatic symptoms obtained. Rales in the chest were made out at the height of some of the attacks. Usually after an hour’s time all evidence of clinical manifestations had ceased. Occasional sneezing was heard some hours after the exposure. A definite rise and fall of nasal irrita- bility was noticed during these fifty-three exposures. The height of the first rise occurred approximately at the thirteenth exposure and depth of the fall at about the twenty-eighth exposure. Another rise reached its height at the thirty-second exposure and dropped to its lowest at about the forty-third exposure. Another high tide was at the close of the experiments March 26, 1918. At each successive rise and fall there was a definite impression of waning sensibility (see chart). A refractory stage was impending but wasnever entirely established (?). On March 23, the blood of one animal was tested for precitins with negative results. Another animal on the same day received 1 ce. of pollen extract (0.5 per cent protein, refractometer method) intracardially with negative results.. The third animal on March 20 received a stronger extract (1.77 per cent of protein, refractometer method). This animal reacted with symptoms of acute anaphylaxis with recovery. Experiment 2, chart B, serum I. Inthe meanwhile a control group of animals was under observation. On November 13,1917, three guinea- pigs had received 1 cc. of Lederle’s normal horse serum intraperitoneally. This was repeated at two day intervals until forty-seven instillations had been given. The serum was dropped into nares off the needle that usually accompanies the Lederle’s package. No definite amount was given. Sometimes one, sometimes two or three drops were used in each nostril. Enough was instilled to insure bathing of the turbinates. The characteristic welling up of salivary secretion was noticed in these animals after each instillation. It was not until the third or fourth instillation that any semblance of hay fever symptoms were manifested. On the whole these animals exhibited the same complex of symptoms that were obtained in the pollen animals, but never so marked. At no time was asthma obtained. The group also showed the curves of in- creasing and decreasing reactivity. In fact at the forty-second in- stillation complete refractivity was obtained. This continued there- after and therefore after the forty-seventh instillation the animals were 458 HENRY L. ULRICH no longer subjected to horse serum by the nasal route. On March 6, one hundred and sixty-five days after the initial intraperitoneal exposure, two‘ of the animals received 1 cc. of horse serum intracardially. Acute anaphylaxis occurred with recovery in both animals. This group supported or illustrated more than any others the possible hy- pothesis put forth by several observers (8) that there may be a pro- tective mechanism as a part of the function of the nasal mucosa. Experiment 3, chart C, pollen II. Five animals, average weight 418 grams, were subjected intrapertioneally to 25 mgm. of pollen, suspended in 1 ce. of salt solution respectively. On January 8, 1918, insufflation of pollen in nares was begun and continued at two days intervals. This experiment was a repetition of experiment 1, with younger animals, a larger sensitizing dose and a larger number of animals. The first in- suffation provoked the characteristic symptoms. The reaction in these animals was more intense than those in experiment 1. Nasal and bronchial stenosis was obtained as early as the seventh insufflation. These animals also differed from the first in that no curves of refractivity were obtained. On the other hand, a steady rise of nasal susceptibility was noticed. On March 1, 1918, one of the highly susceptible animals was bled. The blood was tested for precipitin with negative results. March 31, 1918, one of the animals was injected intracardially with 1 ee. pollen extract (0.5 per cent protein by refractometer method). There were no symptoms of anaphylaxis. The blood serum of this animal previous to the intracardiac reaction was tested for precipitins with negative results. On March 20, 1918, two animals in this series were injected by the cardiac route with 1 cc. of a 1.77 per cent protein pollen extract. Both animals reacted with acute stormy anaphylactic symp- toms with recovery. Experiment 4, chart D, pollen III. January 13, 1918, three animals with an average weight of 604 grams were subjected to insufflation of pollen without previous sensitization of any kind. The insufflation was conducted at two day intervals just as in the other experiments. Janu- ary 19, 1918, after the fourth insufflation mild symptoms of sneezing, itching, and lachrymation was obtained. These animals were subjected to thirty-four treatments. No curves of refractivity were obtained but a steady rise of susceptibility was noticed. Precipitin tests were made on one animal on March 1, 1918, likewise another on March 21, 1918, and ‘The third animal had disappeared from its cage on February 15, and was not found again. EXPERIMENTAL POLLINOSIS 459 on March 26, 1918; intracardiac injections of 1 cc. of pollen extract in one animal for each date resulted in no symptoms. Experiment 5, chart E, serum II. These animals, five in number of an average weight of 400 grams, were used as controls for experiment 4. Horse serum was instilled intranasally without previous sensitization. The instillation was begun on January 23, 1918. No effect was noticed until after the fourth instillation. For nearly one month indefinite, individual reactions occurred, resembling the symptoms of the pollen animals but not so uniformly or clearly defined. During the next month, however, more definite symptoms were obtained. On January 6, 1918, three of the animals received 1 cc. of horse serum intracardially. Two responded with marked anaphylaxis with recovery, the third died in typical anaphylactic shock. On January 17, 1918, when the curve of sensibility was most marked up to this time, the two remaining animals were subjected to 1 ec. of horse serum intracardially and both died in typical anaphylactic shock. The series conclusively proved that hy- persensibility can be produced by the nasal route. Experiment 6, chart F, starch I. January 25, 1918; three animals, of an average weight of 390 grams, were subjected to insufflation similar to the method used in the pollen animals. The animals were used as a control to the pollen group in order to answer the question which might be raised whether the nasal irritation and symptoms were due to mere mechanical irritation of dust. The animals were exposed to 29 insula- tions on the same days as the pollen animals and at similar intervals. At no time did the animals manifest any symptoms of discomfort or annoyance by the treatments. CONCLUSIONS 1. Clinical manifestations of hay fever, pollinosis, can be pro- duced in laboratory animals. 2. Horse serum produces similar manifestations in the guinea pig but not as clearly nor as uniformly. At no time were true asthmatic symptoms obtained. 3. Evidence of refractive phenomena was obtained by rhyth- mic exposure of the nasal mucous membrane to foreign proteid. 4. Sensitization by the nasal route was established; first, by the manifestation of clinical signs of pollinosis after the fourth exposure; secondly, by the anaphylactic phenomena in the serum animals when serum subsequently was injected into the blood. 5. Non-protein dust gave no symptoms of hypersensibility. 460 HENRY L. ULRICH DISCUSSION That we can sensitize the guinea-pig to pollen by mere exposure of the nasal mucous membrane is of great importance to those who are interested in the membrane from a physiological point of view. Cooke and Van der Veer (9) have shown the influence of heredity on protein sensitization. Their. cases were those occurring entirely ‘‘spontaneously,” from exposure either to pollen or to food. It would be well to consider, in reflecting on the mechanism of hay fever, whether we are not dealing with a congenital or acquired defect of function of the nasal mucosa. It may be possible that the difference between the person not sen- sitive to pollen and the one sensitive lies in the rate of the diges- tion of the pollen proteids by the respective nasal mucosa. The sensitive membrane may have lost the faculty of rapid conversion of the proteins to innocous amino acids or may never have had it. This may be the explanation of the food types of sensitization through an intact mucosa. The rabbit is insensitive to nasal exposure to pollen. The guinea-pig is quite sensitive. The difference in reactions of the mucous membrane in the two species is suggestive of a difference in function phylogenetically. The animals in chart D, pollen 3, which had not been pre- viously sensitized, at no time showed precipitin reactions with the blood, nor anaphylactic symptoms after pollen protein had been introduced into the blood. This experiment approaches in some details what has been found in the human sufferer. Clowes (10) claims he has found precipitin and Koessler (11) reports passive transfer of immune bodies from the patient to the guinea. pig. Cooke, Flood and Coca (12) and I myself have never been able to reaffirm these claims. It would be of interest to attempt passive transference during the height of the hay fever season. Apropos of this: one animal highly sensitive, in chart C, pollen 2, was bled and its serum was injected peritoneally into a normal pig. Forty-eight hours later pollen extract introduced intra- cardially elicited no response. Again a group of (3) animals were injected with 3 ec. of human serum respectively, from a hay fever sufferer. Two days later these animals were subjected to insuf- Se EXPERIMENTAL POLLINOSIS 461 flation of pollen into the nares. Symptoms such as sneezing, itching and lachrymation occurred. The experiment was re- peated in another group (3), with another patient’s blood with entirely negative results. These observations will be repeated and elaborated at some future time. The possibility of demon- strating a passive transfer of immune bodies by a reaction such as that in the skin or mucous membrane, is, as far as I know, a new method of approaching this problem. Lastly, the fact that we can inject pollen grains intraperitoneally in the guinea-pig without untoward results suggests the idea of the possible use of pollen grains direct as a phylactic or prophylactic measure in the treatment of the disease. REFERENCES (1) Cooks, R.A., Fuoop, E. P., Coca, A. F.: Jour. of Immunology, 1917, 2, 217. (2) Sewatt, H.: The Journal of Laboratory and Clinical Medicine, 1917, 2, 875. (3) Korss er, K. K.: Forscheimer’s Therapeutics, 1914, 5, 685. (4) Vide (1). (5) Vide (1). (6) Vide (2). (7) ScHEPPEGRELL, W.: Archives of Internal Medicine, 1917, 14, 959. (8) Pacer, O.: Medical Record, 1915, 88, 470. Pacet, O.: Medical Record, 1917, 92, 668. Hermann, Cuas., N. Y. State Journal of Medicine, 15, 233. Kurca, A. C.: J. H. Bulletin, 1913, 24, 69. Sewatu, H: Arch. Int. Med., 1914, 13, 856. (9) Cooks, R. A., VAN DER Veer, J. A.: Jour. Immunology, 1916, 1, 201. (10) Crowes, G. H. A.: Proceed. Soc. Exp. Biol. and Med., 1912-13, 10, 69. (11) Vide (3). (12) Vide (1). PROMPT MACROSCOPIC AGGLUTINATION IN THE DIAGNOSIS OF GLANDERS OLGA R. POVITZKY From the Bureau of Laboratories of the Department of Health of the City of New York Received for publication September 16, 1918 The agglutination test for the diagnosis of acute cases of glanders is conceded by all workers to be of great value. Miess- ner (1) and others have shown that agglutinins in glanders are generally at their height between the fifth and eleventh day of the infection, after which time they begin to decline. At the end of two months, they may fall even below 1: 500. This test, if used alone, would fail to detect chronic cases and it should, therefore, be used as an adjunct to the complement fixation! and ophthlamic mallein? tests. No one test can be depended on alone as each one has its peculiar value in certain stages of the disease. If all these tests are combined, very few cases of glander can escape detection. Even though a case may appear to be nega- tive by one or two of the methods, it may be found positive by the third. The great objection to the employment of the macroscopic ag- glutination method used heretofore in the routine diagnosis of glanders, is the length of time required for the appearance of a reaction. While some tests can be read at the end of twenty- four hours, others do not show definite agglutination within forty-eight or seventy-two hours or even longer. This drawback led to the centrifuge method for which Muller (2) claims priority. 1 Specific amboceptors for the complement fixation test may be demonstrated seven to ten days after infection and they remain during the course of the dis- ease (6). 2 The ophthalmic mallein test is reliable twenty-one days after infection while the subcutaneous mallein test may be relied upon for diagnosis fifteen days after infection (6). 463 464 OLGA R. POVITZKY By it the process of the agglutination reaction is shortened con- siderably. Miessner (1) also Pfeiler (8) employed this method claiming it to be a great success. According to this method, the test fluid is prepared in the same way as in the old method—namely, a forty-eight hour growth of a suitable strain of B. mallei is suspended in salt solu- tion and heated at 60°C. for two hours. It is then diluted to a certain density with 0.5 per cent carbolic acid. The tubes are set up in the ordinary way, each tube containing 2 to 3 ce. of the titrated test fluid with varying quantities of serum to make the final dilution, 1: 400, 1: 500, 1: 800, 1: 1000, ete. After one hour incubation at 37°C. the tubes are centrifuged for ten minutes, after which they are allowed to stand at room temperature for one and one-half hours. The tests are read at the end of that time. The appearance of an irregular veil like clumping at the bottom of the tube with clearing of the upper portion is consid- ered an agglutination, while a dense white precipitate with a cloudiness of the upper part is considered a negative reaction. Though the centrifuge method possesses great advantages over the old method in point of time, it is difficult always to separate an agglutination from a sedimentation and the result is not always readable. Thus, Anthony and Grund (4) found this test unreadable in 12 per cent of their cases tested, even after an incubation of twenty-four hours. In order to eliminate the shortcomings of the former methods, I have worked out a modification by which an agglutination re- action, clear cut in appearance, can be obtained in less than two hours with positive sera in dilutions up to 1: 2000 or higher. The strain of B. mallei used was one called B. M. 5, isolated at this laboratory about five years ago from a case of human glanders. Fourteen other strains from various sources (chiefly horses) were tested also, but none have given invariably the prompt and clear cut reaction that-B. M.5 does. The speci- ficity of this strain was ascertained by testing it with a number of syphilitic, typhoid, streptococcus and pneumococcus sera; also with a number of sera from horses with equisepticus. One of the latter sera gave a positive reaction but the possibility of AGGLUTINATION IN DIAGNOSIS OF GLANDERS 465 glanders could not be ruled out. When passed through guinea- pigs, B. M. 5 became more virulent but, after successive passages, did not agglutinate so well; in fact, I found that it was not ad- visable to pass this strain through a guinea-pig, as most writers advocate to keep the strain agglutinable. Other strains, how- ever, have not proved so constant as this one. Next in importance to the native agglutinability and con- stancy of the strain, is the medium on which it is grown. The medium which has given the most satisfactory results is potato- glycerin-veal agar that is 2.5 per cent acid to phenolphthalein. It is prepared as follows: Wealeintisrons (UNAGTUStEG) sce... acts Acie eteeeeeenet 1000 ce. LM ERY Po Nee aE eae eS ARR ER ns ORS Lis 7 0d 30 grams INFO ES 68 Bee Cen an ae MERE SI: FCoE A pn 5 grams Re pporm (acl) so Aas teas cayetoue 3 sagcoye4 51.9.9 eyesore near 10 grams * Veal infusion: Chopped lean beef 10 pounds, water 10 liters. Soak over night. Heat for one hour at 45°C. then boil for one-half hour. Strain through cheesecloth. This infusion may be sterilized and stored for stock or used at once. Put in autoclave or Arnold to melt agar. Titrate at room temperature and lower the natural acidity, if necessary, to 2.5 acid (phenolphthalein) by adding normal sodium hydroxid. Do not add acid under any circumstances. Clear with egg and filter. Titrate again and if necessary, ad- just to 2.5 acid. Add glycerin (C. P.) 5 per cent and potato juice 5 per cent. Sterilize in autoclave for one-half hour at 15 pounds pressure. . The potato juice used in this medium is prepared by adding one pound of unpealed potatoes, sliced thin, to 1 liter of distilled water. Autoclave one-half hour at 15 pounds, strain through cheese-cloth and a thin layer of cotton. Bottle for stock and sterilize in autoclave one half hour at 15 pounds. This stock must be filtered each time before adding to medium. That too much stress cannot be laid on the careful preparation of the medium is shown by the following experience. At one time in the beginning of my work, the suspension of B. M. 5 grown on a freshly prepared medium did not give the proper reactions with the control sera. Keeping in mind that the strain 466 OLGA R. POVITZKY was old and in need of rejuvenation, I passed it through a guinea- pig and recovered the organism from the heart’s blood. Mean- while, a second fresh medium was made upon which the newly recovered organism was planted. This suspension reacted very successfully and I should have believed that the guinea-pig pas- sage was responsible for the success had I not planted the old organism on the second medium also. The suspension of this growth worked as well as the one just isolated from the guinea- pig. Therefore, the old culture is still used. Careful attention should be given to all the glassware used in . connection with cultures as well as the tests. It should always be neutralized. Bichlorid of mercury should be avoided as a dis- infectant for the pipettes, bottles, tubes, etc. The stock culture of B. mallet should be transplanted every ten to twelve days and incubated at 37°C. for two or three days. It should then be kept in the ice-box at 10°C. The suspension for the tests is prepared in the following man- ner: Forty-eight hour cultures are used to inoculate agar slants of the potato-glycerin-veal agar described above. A good growth should be insured by inoculating the surface generously. After forty-eight hours incubation at 37°C. the growth is washed off with 0.85 per cent salt solution and killed by heating at 60°C. for two hours. No earbolic acid is added to this stock suspension which is of considerable density. It can be kept in the ice-box for two months or more in 100 cc. bottles, corked and capped, if handled with aseptic precautions. The tests are carried out with a fresh dilution of the stock suspension made by adding 0.85 per cent saline solution. A sample dilution, which has been tested with known negative and positive sera, should always be kept in the ice-box and the fresh dilutions of the suspension compared with it on printed matter. The fresh suspension (dilution) is not filtered as the filtermg seems to hold back something essential to the reaction; moreover, this is not necessary as it can be shaken up to a per- fectly homogeneous fluid. The tests are carried out in the following manner: A primary dilution of the serum 1:40 is made. Each tube receives varying AGGLUTINATION IN DIAGNOSIS OF GLANDERS 467 quantities of this dilution to which 3 ce. of the bacterial suspen- sion (prepared as directed above) is added to make a final serum dilution of 1: 500, 1: 800, 1: 1000, 1: 1200, 1: 1600, 1: 2000. A known positive and a known negative serum are always used as controls with each test; also a control tube of the bacterial sus- pension without serum. The tubes, in copper racks are placed in a water-bath (37° to 42°C.) for two hours. With this technic a reaction up to 1000 or more may be ob- tained in ten to twenty minutes; a positive reaction always ap- pears in two hours. The tests may be kept over night in the ice-box and read again in the morning. The reaction, if com- plete is designated by two plus signs, incomplete by one plus, slight by plus minus, negative by adash. The reaction is so clear cut there is no difficulty in interpreting it. One can watch the bacilli clump through the tube, then fall to the bottom in a white granular mass, leaving the supernatant fluid crystal clear in a complete reaction. There is no necessity to look for “‘but- tons,” “‘veils’ or “films.” It is either an agglutination as we see it with other organisms or it is not considered a reaction. In reading the tests, a reaction is considered as positive which has double plus through 1: 1600; as suspicious with double plus in 1: 1200; as doubtful with double plus in 1:1000. Any reac- tions below 1: 1000 are considered as negative. It is essential to titrate the dilute bacterial suspension often with known positive and negative sera to be sure it is working. Tests should never be one with a suspension that has not been titrated the previous day. The Department of Health of the City of New York has been employing the agglutination test for the diagnosis of glanders tegether with the complement fixation and ophthalmic mallein tests. During the last year, over 2000 sera were examined, which might have given us very rich material for arriving at definite conclusions as to the comparative value of the sero- diagnostic and ophthalmic tests, checked by autopsy findings, had there been closer codperation between the city and state veterinarians and the laboratory. Such codperation should be based on a systematic method of recording full data in connection THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6 . 468 OLGA R. POVITZKY with each horse in a central file. For a thorough study of glan- ders in any region, each horse should be marked and registered under an identifying number, as are automobiles, only in this case the number should be a permanent one which should pass with the horse when it is sold. The traditional methods of horse dealers might make this a difficult matter, yet, for economic reasons a law to that effect rigidly enforced, would be thoroughly worth while. On only 180 horses was it possible to obtain complete data which I have endeavored to analyze. These horses fall into three groups as shown in tables 1, 2, 3. Street. railway horses: «.. oc. fare.2 de oe fo lok coe eeoke. © sv e0e snes Oe ne 111 Horses from ‘an infected stable.-)'.-0---e se tee ee eee 14 Horses for ‘slaughter: <5... Beet eee eee eee 55 "EOtalls coc jieciceon She tin neo ee oe ne ee eee eee 180 The street railway horses and those from a badly infected stable (125 in all) are considered together as at autopsy they all proved to have glanders with the exception of two. These two were in the infected stable and they were killed with the others because of their marked exposure to the disease. One of these horses, number 562, table 2, had been given vaccine; thus the number of 125 is reduced to 123 on which the percentages are based. The 55 horses for slaughter were among those that received careful tests to determine their fitness to be used for food. These horses came under the supervision of Dr. L. Price, veterinarian of the Department of Health and the data are the most satisfac- tory of all. Lately, we have been trying to confirm autopsy findings by pathological sections and this promises to be of great advantage. Many lesions which are suspicious of glanders do not prove so microscopically. The bacteriological inoculation of suspicious material into guinea-pigs for the Strauss reaction does not prove of much aid, as, according to Miessner (3), it is successful in only 25 per cent of the cases. My own experience is in accordance with this ee aa ee he ee a AGGLUTINATION IN DIAGNOSIS OF GLANDERS 469 estimate. It was only in acute cases when B. mallet was isolated from the blood (human) or in horses with an acute infection of glanders that the Strauss reaction was obtained in the guinea- pig after moculation with emulsified material from the lesions. All other inoculation tests resulted negative’y. In considering the horses, in tables 1 and 2, the nece sity of applying all three tests stands out prominently. Of the 123 horses which proved at autopsy to have glanders, 122 were shown to be positive, suspicious or doubtful by one, two or three of the tests. Only one case (horse 123) was negative with all three tests—a failure of about 0.8 per cent. The agglutination test showed 31 horses to be positive, 15 sus- picious and 32 doubtful, making a total of 78 horses or 64.2 per cent. Thus it failed in 35.8 per cent of the cases, probably be- cause most of the horses in this series were chronic cases. Three horses? were picked out by this test alone—the complement fixa- tion and eye mallein tests being negative. These may have been acute cases. The complement fixation test showed 74 horses as positive, 4 suspicious and 17 doubtful making a total of 93 horses or 75.6 per cent. This test failed wholly in 24.4 per cent of the cases. Four horses! were picked out by this test alone. On the other hand the mallein test (ophthalmic except for 9 horses in table 2 on which subcutaneous mallein was used) gave 39 horses as positive, 65 suspicious and 2 doubtful making a total of 106 horses or 87.8 per cent. It failed in 12.2 per cent, but 15 horses® were picked out only by this test. The grouping of suspicious and doubtful reactions with the positive ones is based on the interpretation that any serodiagnos- tic reaction not negative, should place an animal in the suspicious category to be subjected to close observation and frequent re- tests. That such a method is of great value is fully shown in tables 1 and 2. 3 Table 1, horses 76, 1218, 1504. 4 Table 1, horses 570, 1281, 1389, 2047. 5 Table 1, horses 19, 449, 462, 494, 612, 621, 1143, 1146, 1310, 1874 and 1390, Table 2, horses 730, 737, 751, 763. 470 OLGA R. POVITZKY Further analysis Positive by vallithree’ tests... 022.2 .22412 26a e ee 9 horses* Positive by agglutination and negative complement fixation. 6 horsest LALO) 0 Ive Ba Oe a ee eI mere ccm ke i el Cas 23 horses Positive by mallein and negative by complement fixation and SAP CRI ARION: ©. 5.2 .10.. dap et Dee eee eee ee oe eee 15 horsest * Table 1, horses 418, 506, 1776; table 2, 645, 683, 738, 746, 750, 757. {t Table 1, horses 163, 181, 605, 1218, 1360, 1504. t Table 1, horses 19, 449, 462, 494, 612, 621, 1143, 1146, 1310, 1374, 1390; table 2, horses 730, 737, 751, 763. 5 Of the 55 horses tested before slaughter for food purposes (table 3), 4 horses® proved to have glanders. The complement fixation test failed to detect this fact in all 4 cases while the ag- glutination test failed in 3 of them and the ophthalmic mallein in only 1 (horse 2001) which the agglutination reaction picked out. The agglutination test gave 3 false “suspicious’’ reactions,? the complement fixation none, and the eye mallein 9. In a set of horses such as are listed in tables 1 and 2, a large number of chronic cases may be expected and the agglutination tests positive in the least number of instances. On the other hand in a large number of sera examined during a whole year where different stables are tested there may be a great many cases of beginning infection or cases in the acute stages of the disease. . There were tested 1890 sera from various sources and among these the agglutination test gave positive and suspicious reac- tions in a larger proportion than the complement fixation. Un- fortunately the eye mallein and final outcome of the cases could not be obtained. Thus, with the complement fixation, about 14.5: per cent of the cases were either positive or suspicious while the agglutination registered 28.2 per cent. The doubtful reactions occurring with two tests were almost the same in number (complement fixation 300 and agglutination 291). ® Table 3, horses 2001, 2006, 3044, 3064. 7 Table 3, horses 3027, 3040, 3053. AGGLUTINATION IN DIAGNOSIS OF GLANDERS 471 With complement fixation test 2 TST NG ae ee ea AL erat) Ice A nd 254 SrCIOUB ER chi io. oi rheetley Ais FR AU pe pel pd boy 4. oes 21 “JOULE GTS SS BA a ene pO dW OC Rca Se td pd i a 300 1 EV LET 720 A SIN GOR Bc ae a 1315 With agglutination test iP DELICE Go a pee ee Ye TE OR 262 IMPICEOUS EE 2 2ec's fess at eee Cha bo neo ee ee 171 Metraet PEEEAM AEP. « o's GD ee rss oo oh! | 6 ee ee ae pete ate 291 Be eR TCO as os av SBYAS Gato R ABI o's 5.6 Vig EG Oe Cees 1166 With both tests 2 DELESTGL 25 Se Ree, 3 See ee eo eae ek ee 138 PSUS TOLCLO US menses ar A ier hack oS ch emacs As cla sarc vce tala oe 5 12 Ti) DEAE OSS See OR mt ers eek ls) RI 66 LS GISIES TES Cag GAR A ian eR INES, D806, oe aha ot 800 With either one of the two tests RSM Car 5A. M4, Sida Ss a 1h lerctlanitcy decetss Sehakcs \ alpape ee eee ele eae 378 SOTES DOE DUE a oe I ee ri dom ets tats 187 PNPUIMRRES CERES NO coc farce feave.S iasg atone pies Uke Sue w ad ORR ee 525 SRSA Ce Sse te IRE s Saal Ae oRenes 3 ath ORS a eee 800 These 1890 sera did not represent 1890 horses for about 15 per cent were retests. If it were not for these retests,a greater percentage of agglutination reactions would have been obtained for, after a lapse of time, the agglutinins disappeared in the horses retested, while the complement fixation antibodies re- mained, or appeared for the first time. Another factor which influences the sero-diagnostic tests is the subcutaneous injection of mallein. It is generally held that six to eight weeks should elapse before the blood is taken for tests after mallein has been injected subcutaneously. Recently, in a large number of horses that had been given subcutaneous mal- lein, many showed four plus complement fixation and positive agglutination reactions, when retested about six weeks later. At the time of the first test these same horses had shown either doubtful or negative results. These horses are being kept under observation and time will show whether the high titers in both tests were due to the subcutaneous injection of mallein or to a fresh infection. 472 OLGA R. POVITZKY SUMMARY 1. A method has been devised whereby a prompt clear cut macroscopic agglutination for the diagnosis of glanders can be obtained in two hours. 2. Fifteen strains of B. mallet were tested and one obtained which is constant in its agglutinability with positive sera. 3. The best culture medium is glycerin-potato-agar (2.5 acid to phenolphthalein) carefully prepared (see text). 4, Great care must be used in the preparation of the stock suspension of B. mallez (see text). For the tests, fresh dilutions in 0.85 per cent salt solution are made from the stock suspension. No carbolic acid is added to the stock nor to the dilutions. 5. The agglutination test is valuable in the routine examina- tion of horses. With this test nearly all early and acute cases can be detected and thus the spread of the disease can be prevented. 6. A negative reaction by a single agglutination test if not confirmed by the ophthalmic and the complement fixation tests does not prove the case negative; nor does a single negative re- sult either by the complement fixation or ophthalmic tests. -ar= slo eee 67 ——. The non-influence of injections of trypsin upon the protein quotient in blood serum: foie se ck ss so s yc sible ee ise S Dele he oe uae Pe Or een 139 —— and Robertson, T. Brailsford. Anew method of estimating the antitryp- tie index of bloodserum:, :..52. «oa l.sser tncrs fe nee eee > ea eee 131 Hartman, C. C., and Lacy, G. R. Specific reactions of the body fluids in PNEUMOCOCCIE IMFECHION. 2.0. 2c) so 5s sins wise ie os Ves oe ee See 43 Heist, George D., Solis-Cohen, Solomon, and Solis-Cohen, Myer. The bac- tericidal action of whole blood, with a technique for its determination. 261 Hemblysis, A. study Of saponin: «20665 2.60 = .4 Sos 5S ace ee eee 423 Immune bodies, The isolation, purification and concentration of........... 109 —— hemolysin, A studyouol:s> eee, om eases owe Seek eae eke ee 109 Immunity, Active, in experimental poliomyelitis................-..-...45+- 435 ——., The rdle of, in the conduct of the present war...........---++++++++:- 371 ——. The study of problems of, by the tissue culture method.......... 219, 233 Immunizing properties, A study of the, of bacterial vaccines prepared after various, methods... 62242 wate och bes > Rae Bale hic ee 247 Immunologic properties of uveal pigment, The..............-----...+--+-+ 75 Indigo test; On Von Dungern’s; forjsyphilis . asajeecer -.---1a<41205-- aera 11 Intensive digestion, The constancy of the protein quotient during, and pro- longed) starvation... .....). 01. 2's gmc Agaie mie ee =n a se eee eee 67 INDEX 483 Intracutaneous absorption, The specificity of.................... Sheps tits Intravenous injections, Effects of, of a colloid (gelatin) upon rabbit sera.. . Isolation, purification and concentration of immune bodies: a study of im- PAGE REM O Ly SUIT Netty ees eae ey Ne AL. BALL Latha ote ee Kahn, Reuben L., and MeNeil, Archibald. A note on the relation between proteolysinssand haemolysinst.g+..ssus ee tess sete etio dels noel See ntees ss ? ——-——. Archibald. Complement fixation with protein substances....... Kolmer, John A. The réle of immunity in the conduct of the present war... —— and Matsunami, Toitsu. The relation of the meningococcidal activity of the blood to resistance to virulent meningococei..................... —— and Perry, M. W. A study of the immunizing properties of bacterial vaccines prepared after varlous methods. 2)... os 2 vee cite e ew. oe ks ——and Sekiguchi, Shigeki. Experiments upon the passive transfer of antibodies from the blood to the cerebrospinal fluid.................... —— and Toyama, Ikuso. ‘The influence of arsphenamine and mercuric chlorid upon complement and antibody production. . —— and Weiss, Charles. Studies in pneumonia. VIII. “Abslaian eeictiow: to A SIVE URINE UO RAT RERMEO TREE NUS Ses 1K GR 1/08 (ek eeee CRE sive BRE en oUt enal i eta Ly SAP Re Rae PAM ——, Toyama, Ikuzo, and Matsunami, Toitsu. The influence of active nor- mal serum (complement) upon meningococci. I. The opsonic activity of fresh normal serum alone and in combination with antimeningitis serum OTIC IVETN EO COC Cina Say aeia takes ioosesocuat eines Faycici ae eae ort ei es Ge — and Matsunami, Toitsu. The influence of active normal serum (com- plement) upon meningococci. II. The bacterial and protective value of fresh normal serum alone and in combination with antimeningitis APMMELY KOR MINE IN ZOCOCEL J. hep Aaya A gets wes oectegs MUD Ane 4 cle Korns, John H., and Balls, A. K. On the mode of action in vitro and ‘fie preparation of hemolytie antibodiess./ 4/512. MN oe one. anna Kosakai, M. The isolation, purification and concentration of immune bodies-va study of immune: hemolysin) .)) 8. led a ee Pe one Krumwiede, Jr., Charles, and Noble, W. Carey. A rapid method for the production of precipitin antigen from bacteria: an attempt to apply it to the determination of the type of pneumococcusin sputum............ Lacy, G. R., and Hartman, C.C. Specific reactions of the body fluids in pneu- HAO COC CLE MMLC CLONAL Cpa e nak Repeal LhGbs eettats, ae cine erolonete) akere Se Retey eke t= MeNeil, Archibald, and Kahn, Reuben L. A note on the relation between PLOUCOKVSIN Sean GUMACIMOlY SING Hats co -are cietee ta ey~ ctael« eles tats MRepebouaPle) teers Complement fixation with protein substances................+-- Macroscopic agglutination, Prompt, in the diagnosis of glanders........... Matsunami, Toitsu, Kolmer, John A., and Toyama, Ikuzo. The influence of active normal serum (complement) upon meningococci. I. The opso- nic activity of fresh normal serum alone and in combination with anti- MEHINPIFIS, BELUM LOF MENINGOCOCCL: ...... 5.052 sti oe derssle a alae tee neee oles —and Kolmer, JohnA. The influence of active normal serum (complement) upon meningococci. II. The bacterial and protective value of fresh normal serum alone and in combination with antimeningitis serum for BUCH O COCO eke segs is isha ave vb SRI Sho ate « woe oh reph aphad Ws lee exh MRS Neer —— and Kolmer, John A. The relation of the meningococcidal activity of the blood to resistance to virulent meningococcl............--+-+++.++- THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6 . 301 395 157 157 177 484 INDEX Meningococci, The bacterial and protective value of fresh normal serum alone and in combination with antimeningitis serum for................ 177 ——, The influence of active normal serum (complement) upon........ 157, 177 ——, The opsonic activity of fresh normal serum alone and in combination with -antimeningitis serum for. . : 32:32. aSacs 4. been eee oe ee 157 ——, The relation of the meningococcidal activity of the blood to resistance tonvirul enibaseekyse ne ce eae 413 Mode of action in vitro and the preparation of hemolytic antibodies, On the 375 Noble, W. Carey, and Krumwiede, Jr., Charles. A rapid method for the production of precipitin antigen from bacteria: an attempt to apply it to the determination of the type of pneumococcus in sputum........... 1 Opsonie activity of fresh normal serum alone and in combination with anti- meningitis serum for meningococel, The... .:: : .......25.2225-7-=-e eee 157 Perry, M. W., and Kolmer, John A. A study of the immunizing properties of bacterial vaccines prepared after various methods................... 247 Pneumococcic infection, Specific reactions of the body fluids in............ 43 Pneumococcus in sputum, A rapid method for the production of precipitin antigen from bacteria: an attempt to apply it to the determination of the RY POOL «0. ses cise nade gol a doe a eas Bs Risiere ee eee ee 1 Pneumonia, Studies-ins : sce eebeotk jae ae okt Os eC eee eee ie ee eee 395 Pneumotoxin, A skin reaction t0:..c-5--e->--4-ss eee ee eee eee 395 Poliomyelitis, Active immunity in experimental..........................-- 435 Pollinosis,, Experimental: i022 22th: 30... «3:1... eee ieee ee eee 453 Povitzky, Olga R. Prompt macroscopic agglutination in the diagnosis of glanders: «icicle ois coed sea nt ae die deh = eee eels Reno eee ae 463 Precipitin antigen, A rapid method for the production of, from bacteria.... 1 Proceedings of the American Association of immunologists................-. 317 Active immunity in experimental poliomyelitis. Harry L. Abramson.. 317 — immunization against pneumonia, The. R. Kohn................. 331 Bacteriological study of post-operative pneumonia, A. Miriam P. Olmstead 4.5 HBA iis ross s:d wists Skee 4 sae ee ee eee 330 Contribution to the bacteriology of B. fusiformis; its morphologic phases and their significance, A. Ralph R. Mellon.................. 327 — to the study of complement fixation in tuberculosis, A. Hassow von: Wedel! pe ki cts Ste.x Sehr Ae SPO Se eee 339 to the study of complement fixation in tuberculosis, A. M. A. Wilson... | .<..)...k4bepe at eetten 25824. sabes dt cate ee eee 339 Examination of the blood preliminary to the operation of blood trans- fusion, The. Arthurvle Cotas iss conocer ot See eee eee 337 Experimental pollinosis in guinea-pigs. Henry L. Ulrich.............. 325. Experiments on the production of antipoliomyelitis serum in rabbits. Bidgar {H. ‘Tsent: 20.26: 2e 2s Shae S82 Se ee ie * AS Re. eee 317 upon the chemotherapy and chemoserotherapy of pneumococcus infection. John A. Kolmer, Edward Steinfield and Charles Weiss.... 336 upon the passive transfer of antibodies to the cerebrospinal fluid. John A. Kolmer and Shigeki Sekiguchi... . =... .<-.- Oy be. ss) rie AO lOWEs tt tee ete tt ce. aise ee oe 333 —— of arsenobenzol and mercury upon antibody production, The. Kuo; Loyama-and John Asolmer:. :2h0: stent ees eda eect ls Jecee 326 —— of normal human and guinea-pig serum (complement) alone and in combination with antimeningitis sera upon virulent meningococci, The. John A. Kolmer, Toitsu Matsunami and Ikuzo Toyama........ 319 Isolation, purification and concentration of immune hemolysin, The. iN. LGR ie Agape gehen a it St Seem RE le aga we RU Aled Mga 338 Lipo-vaccines. Eugene R. Whitmore and E. Fennel.................. 343 Method for preparing bacterial antigens, A. J. C. Small.............. 339 Nature of antianaphylaxis, The. J. Bronfenbrenner and M. J. Schlesin- CIE reese ois hte tetas es Eee she a tes rie alot ceenejenerens bite eters eine bon case 321 Observations on the intraspinous autosalvarsanized serum therapy of cerebrospinal syphilis. Benjamin A. Thomas........................ 342 Persistence of active immunity in those immunized against diphtheria. Wwvriliieyam ISL, TReay a See is 2s Fae a aie aN eo eae aA ar smn BER A nA a ae ee 328 Production of pneumococcus antiserum and the corresponding curves obtained by protection and agglutination tests. G. Benjamin White.. 331 Properties of pneumotoxin and its probable réle in the pathology of lobar pneumonia. Charles Weiss and John A. Kolmer.................... 337 Rapid simple method for the determination of type of pneumococcus in sputum of lobar pneumonia, A. Charles Krumwiede, Jr............. 333 —— —— —— for the extraction of precipitin antigen from bacteria, A. tren csmiGrunnwiede mmrsae yor eters fad Site ea as cme es eer ete 338 Relation of the meningococcidal activity of the blood to resistance to virulent meningococci, The. Toitsu Matsunami and John A. Kolmer 319 Réle of immunity in the conduct of the present war, The. John A. TACCOUITO ET a ‘ath gies a bee pch i Scusth Cin cae RAE IE aad: Oc RRS as 8 Av, 317 Simple method for blood culture, A. John G. Wurts and S. W. Sap- DIAN SO LIM ETS Se eae Ne et ke ete He al nic tate se cle aa eis Mech cigs Saas a os tte 330 Skin reaction to pneumotoxin, A. Charles Weiss...................... 326 Studies on so-called cellular anaphylaxis. W. P. Larsonand E. T. Bell 323 — on the toxicity of pneumonic lungs. John A. Kolmer, Charles Wielssranaalidwardgocelnteldr. ss. s.tti. ceca: os ccs e cists cee Hae sai emer 336 Study of controlled postmortem Wassermann reactions: a supplemen- tary report on 400 cases, A... ‘Stuart Graves... ........ 0.05000 ecenue: 341 — of the immunizing properties of bacterial vaccines prepared after various methods, A. M. W. Perry and Sara Levy................... 344 Types of meningococci concerned in epidemic infections. A. Parker Hiatchenssands@. Les obinsOns yeee tte eek cise nee one ee eee 318 Waceme dosage.” doseph Head! 3) e602 is awa ontlneeee ss cence 342 — treatment of acne, with special reference to the réle of Bacillus coli, The. Albert Strickler and Jay F. Schamberg........ . 342 Various immunological reactions in glanders, The. G. Benjamin White 327 Protein quotient, The constancy of the, during intensive digestion and pro- One edesbAnVvatlOnrey emanate mkt. rine cee entre eae ooeitieioreus eyemsienn se 67 —— ——,, The non-influence of injections of trypsin upon the, in blood serum 139 —— substances, Complement fixation with................-2..-++2+e+esee 277 486 INDEX Proteolysins and haemolysins, A note on the relation between.............. 295 Rabbit sera, Effects of intravenous injections of a colloid (gelatin) upon.... 147 Rabbits, Experiments on the production of antipoliomyelitic serum in...... 213 Relation between proteolysin and haemolysins, note on the................. 295 Robinson, T. Brailsford, and Hanson, Samuel. A new method of estimating the antitryptic index of blood serum... 326: sa ooer np cia ee 131 Saponin-hemoelysis; A study: of. ....<42./02h sens eee eee nee eee eee 423 Sekiguchi, Shigeki, and Kolmer, John A. Experiments upon the passive transfer of antibodies from the blood to the cerebrospinal fluid.......... 101 Serum (complement), The influence of active normal, upon meningococci, 157, 177 Skinvreaction to pneumotoxin, As). seeat ac 2 -ccloae = deheenee eek Cane 395 Small, James C. A method of preparing bacterial antigens................ 413 Smith, G. H., and Cook, M. W. The specificity of intracutaneous absorption 35 Solis-Cohen, Myer, Heist, George D., and Solis-Cohen, Solomon. The bac- terial action of whole blood, with a technique for its determination .... 261 Solis-Cohen, Solomon, Heist, George D., and Solis-Cohen, Myer. The bacterial action of whole blood, with a technique for its determination 261 Specificity of intracutaneous absorption, The....................---eceeees 35 Starvation, The constancy of the protein quotient during intensive digestion and! prolonged as. .¢22 sakes eee ers a aan Ce eee 67 Studies:in ‘pneumonia: ...,452 7860.4 fe sale gone sce Oe ee eee eee 395 =——— onthe vantitry psink Ofserums.2j-620.cc 1s orice DO eee oe oan 51 Suzuki, Yoshio. The study of problems of immunity by the tissue culture method. I. A study of the cells and blood plasma of animals which are naturally resistant and others which are susceptible to diphtheria and tetantis toxins 3.ire.t os lade Bronkiece ShisdoSoters se cee ene eee 233 — and Burrows, Montrose T. The study of problems of immunity by the tis- sue culture method. II. The tissue culture as a means for quantitatively estimating toxin and antitoxin and determining the distribution of anti- toxin in‘passively ammunized ianimals....5.:5, .41,.4n¢- 260 320 4260 ee 219 Syphilis; On: Von Dungern’s indigo test for..: 2% ...)..9.4-5-2- - -- ae 11 Tissue culture as a means for quantitatively estimating toxin and antoxin and determining the distribution of antitoxin in passively immunized ANIMAl Ss: ney ewe pach lon oratectehece yeu ae tegen ISR eee ee 219 — —— method, The study of problems of immunity by the......... 219, 233 Toxin and antitoxin, The tissue culture as a means for quantitatively esti- ; mating, and determining the distribution of antitoxin in passively im- Muinizéd. -animalses ty Ger..es-al dies 4 5.che-21) Sa ee aes A xa Oe 219 Toyama, Ikuzo, and Kolmer, John A. The influence of arsphenamine and mercuric chlorid upon complement and antibody production............ 301 — —, Kolmer, John A., and Matsunami, Toitsu. The influence of active normal serum (complement) upon meningococci. I. The opsonic ac- tivity of fresh normal serum alone and in combination with antimenin- Sibis, Serum! Lor IMENnINgOCOCCIERe en. a CeO eeeeee eeee 157 Trypsin, The non-influence of injections of, upon the protein quotient in blood ‘serumits 3» 5053). ds.Go see eae ee tee a 139 Tsen, Edgar T.H. Experiments on the production of antipoliomyelitic serum In: TAbbUS ech col acs shearek SAS eae Mes eer aerate cio ee eee 213 INDEX 487 Tuberculosis, A contribution to the study of the complement fixation reac- BTEURETL yee Pe Se hee At LAN Go Giah ae nn 0 fad ws SENG re alalel vig SI 345, Typhoid-antityphoid serum complex, Extracts of antibodies obtained from specific precipitates Of............ 6... s eee eee teeter tenet eens Ulrich, Henry L. Experimental pollinosis. Preliminary report Uveal pigment, The immunologic properties of...........-.--.--.+...50e05- Von Wedel, Hassow. A contribution to the study of the complement fixation Bete eTMEGTA UO GLCUI OSES: 05/55) to tents tafe bie oetale sual ietaene wie Gain. etsle in ysiel er War, The réle of immunity in the conduct Of thepresenite fae. Sec oa ate se Weil, M.O.R.C., Major Richard. .........-..-- 2... sees eee eee eee eens Weinstein, Israel. Extracts of antibodies obtained from specific precipitates of typhoid-antityphoid serum complex.........-..- + +065 0+e essere eees Weiss, Charles, and Kolmer, John A. Studies in pneumonia. VIII. A skin reaction to PMEWMO CORUM ous vas awe) s sic + ree eames os she aie « Wenner, John J. A note on bleeding guinea-pigs and on preserving sheep’s EEyMAEGCVES ss. asi nicaie(s Ue amc ood ass | alate ne # cle eeya Fagus o> ieinrs vin Whole blood, The bacterial action of, with a new technique for its determina- PATIO esa feat te re UNE sch thon oseleltytel chet steve nieteue at oravelaie Ml # mime Wilson, M. A. A contribution to the study of the complement fixation reac- PAM MIDE CULOSIS » «62's... Saline s sleaie oo es cle ee leininiels = Bliye sc mie ese aie spel Woods, Alan C. The immunologic properties of uveal pigment oc -* Shee tie AT yu Shag te Weta ms si CR The Journal of immunology 180 J6 Vv 23 cop.2 Biological & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY Te Laneks batabes Ready Lett ears wr . sete rsnina® bad an aiodiad * aru est see 1e Sr" Litt “ree SAcaaaed ++ mae aay ee eo 4 . whe Bi dm a te Cede aes Coase Surin * eee ae 4m oFeteaelt cr) haa rep Ce ee + 2 sn eat tee se en esr erarye + fi ; Alaa. . Sime 22 OO. or waurrssys 2 er Ee a SOIT 5 - 7 TASedes 2 ‘ Tata ona . oe - io Nota ata : . ated pie. * eet : see “htt sees state? -_ 9 me au sha ae y - Sen aa sets? oe ? Sete wha’ : coi BP aes 3 | | | | a = wat Tepieseysotety oe : ne Cee y ‘ fet, ese +e Wot Corners . ae stake 5 . res fe ne at 5 Ste aatie! yet Wah eet ha tatae, Ferhe8 oes.