oe = SSS Sree ied x ¥ x4 Breit Ruth rele eat oe eth ates roe ated Sete ch agate neces pik oho fia a =~ | . oe WA i Mse THE JOURNAL OF BIOLOGICAL CHEMISTRY FOUNDED BY CHRISTIAN A. HERTER AND SUSTAINED IN PART BY THE CHRISTIAN A. HERTER MEMORIAL FUND EDITED BY H. D. DAKIN, New York City. LAFAYETTE B. MENDEL, New Haven, Conn. E. K. DUNHAM, New York City A. N. RICHARDS, Philadelphia, Pa. WITH THE COLLABORATION OF J. J. ABEL, Baltimore, Md. P. A. LEVENE, New York. R. H. CHITTENDEN, New Haven, Conn. JACQUES LOEB, New York. OTTO FOLIN, Boston, Mass. A. S. LOEVENHART, Madison, Wis. WILLIAM J. GIES, New York GRAHAM LUSK, New York. L. J. HENDERSON, Cambridge, Mass. A. B. MACALLUM, Toronto, Canada. REID HUNT, Washington, D. C. J. J. R. MACLEOD, Cleveland, Ohio. WALTER JONES, Baltimore, Md. JOHN A. MANDEL, New York. J. H. KASTLE, Charlottesville, Va. A. P. MATHEWS, Chicago, Ill. J. B. LEATHES, Toronto, Canada. F. G. NOVY, Ann Arbor, Mich. THOMAS B. OSBORNE, New Haven, Conn. T. BRAILSFORD ROBERTSON, Berkeley, Cal. P. A. SHAFFER, St. Louis, Mo. A. E. TAYLOR, Philadelphia, Pa. F. P. UNDERHILL, New Haven, Conn. V. C. VAUGHAN, Ann Arbor, Mich. ALFRED J. WAKEMAN, New York. HENRY L. WHEELER, New Haven, Conn. - att a Hoan SF ae, cin VOLUME XI io BALTIMORE IgI2 ¥ er an a a ri PAC as Cie may ' r : = bg 7 * oP. TAMAUOL Te Bo, | | be LAOR he NS ol we ‘ ied A senile aa . om eshte ; ree fay Vo % TAS ta aries ee? Diol ws, pits mera we JES si. oF ; Hi a skape ahi fl e Read by title. Society of Biological Chemists XV1l agar-agar and water. The total nitrogen excreted during the period was 246.1 grams, leaving a balance of 80.2 grams retained. This pig was also found to be badly infested with worms and the experiment was discontinued. The weight at the end of the exper- iment was 23 pounds. Both of these pigs became very thin. A third pig weighing 48 pounds was fed during sixty days a ration of skimmed milk and starch. The milk was treated each day with 86 per cent phosphoric acid to make 0.6 per cent acid, and boiled for fifteen minutes, cooled and treated with enough milk of lime to just neutralize it. The pig ate this mixture well. The animal retained 329 grams of nitrogen, and increased in weight 17 pounds. These experiments indicate that the pig can grow to a considerable extent on casein as the sole protein, and that milk treated so as to disturb any specific complexes between organic and inorganic radicals is still capable of maintaining a fairly vigorous growth. THE URINE OF LATE PREGNANCY AND THE PUERPERIUM. Bry JOHN R. MURLIN anv H. C. BAILEY. (From the Departments of Physiology and Obstetrics, Cornell University Med- ical College, New York City.) From the maternity wards of the Bellevue Hospital and from the Emergency Hospital and School for Midwifery on 26th Street, were obtained continuous series of urines from the following cases. 1. Three normal pregnancies (ninth month) under control as to diet. 2. Three pre-eclamptic cases. 3. Two cases of eclampsia, one interpartum and one postpartum. 4. One case of pernicious vomiting with nephritis. The normal cases on creatine-free diets containing less than 35 calories per kilogram show creatine in the urine. The percentage of ammonia nitrogen in the best-fed case ran as high: as 12.2 per cent of the total. The amino-acid nitrogen by Henriques and Sérensen’s’ method runs as high as 7.9 per cent. The pre-eclamptic cases when placed on milk diet showed no high ammonia. * Henriques and Sorensen: Zeitschrift f. physiol. Chem., Ixiv, p. 120, 1910. XV1il Proceedings In the case of interpartum eclampsia, the ammonia was not above 6 per cent until after the convulsions. Afterwards it ran up to 30 per cent. The amino-acid N in this case just before labor was nearly 0.8 gram, being 6.6 per cent of the total. At the same time the undetermined nitrogen (possibly “peptid-bound” nitro- gen) amounted to 3:7 per cent. In the other case of eclampsia the ammonia fraction was high in the first urine received by us, but fell rapidly as the patient’s condition improved. The single case of pernicious vomiting seems to bear out the views of Underhill and Rand’ as to the effects of starvation. THE STORAGE OF FAT IN THE SALMON MUSCULAR TISSUE AND ITS RESORPTION DURING THE MIGRATION FAST.® By CHAS. W. GREENE. (From the Department of Physiology and Pharmacology, Laboratory of Phys- tology, University of Missouri.) The king salmon stores large quantities of fat in its tissues dur- ing its life in the ocean. When it enters the fresh waters of the rivers in the journey to the spawning ground it is now well known that it wholly ceases to take food and makes the Journey while fasting. Food material is stored in the salmon tissues during the ocean feeding period and this food consists almost if not entirely of fat. In the Columbia River those salmon caught at the lowest point at the mouth of the river have the greatest amount of stored fat in their tissues. The salmon fat is stored primarily in the muscles. These mus- cles are of three classes; namely, (1) The lateral dark muscle; (2) The great lateral pink muscle; (8) The small muscles of the fins and head. The fat is stored in each of these types of muscle in its own characteristic way. 1. The fat in the lateral dark muscle is in large drops chiefly within the fibers, but to some extent between the fibers. The fat drops between the fibers are relatively few in number and seldom exceed a diameter of 20 micra. The fat in the dark muscle within the fibers is in two characteristic regions, (a) between the sarco- 7 Archives of Internal Medcine, v, p. 61, 1910. 8 Abstract published by permission of the U. S. Commissioner of Fisheries. Society of Biological Chemists xix lemma and the muscle plasma where the drops often amount w as much as 6 to 12 micra in diameter, and (b) within the sarcoplasm. This intramuscular fat is in unusually large droplets in size, often varying from 6 to 8 micra in diameter and located primarily in Cohnheim’s areas. These larger droplets are associated with all sizes of fat droplets down to liposomes 2 micra in diameter. The liposomes are in longitudinal chains located ofttimes between the individual fibrillae, the liposomes corresponding fairly closely in position and number with the striations of the muscle sub- stance. 2. The fat in the pink muscle which represents the greatest muscular mass is wholly between the fibers, up to the time when the salmon stops feeding. The storage of fat in this region is enormous. The fat drops are of relatively large size, varying from the smaller ones only a few micra in diameter to drops as much as 100 micra in diameter. 3. In the smaller fin muscles which are in relatively constant activity, there is only a small amount of stored fat and that is chiefly intermuscular. From the time the salmon stop feeding until their death after . spawning the quantity of stored fat gradually diminishes. It is never wholly consumed even in fish taken after natural death. In the dark muscle the fat is gradually eliminated both from the inter and intrafibrous regions. It never wholly disappears from the substance of the muscle fiber but shows the extreme reduction at the time of dying. In the pink muscle the interfibrous fat is gradu- ally removed during the migration period and has practically dis- appeared when the fish have reached the spawning stage and at the death which follows. An observation of more than usual interest is found in the fact that a large quantity of extremely finely divided fat makes its appearance within the pink muscle fibers as soon as the fish stops feeding on entering the fresh water of the rivers. The fat is some- what greater in amount and the droplets are slightly larger in the smallest fibers. This intrafibrous fat is present in all specimens at all stages of the migration journey. Its quantity is remark- ably uniform. In fish from the spawning grounds which are approaching the spawning period this intramuscular fat begins to diminish in quantity. At the time of death, however, consider- xx Proceedings able quantities are still present in the smallest fibers though it has completely disappeared in the largest fibers. It seems evident that fat is thrown into the fibers of the great lateral muscle and kept there in strikingly uniform quantity and amount during the entire migration journey. It is suggested that this fat is utilized by the muscle as the source of the energy ex- pended during the migration fast. INTESTINAL ABSORPTION. By H. C. BRADLEY anp H. 8S. GASSER. (From the Laboratory of Physiological Chemistry, University of Wisconsin.) An emulsified mixture of olive oil and petroleum oil fed by sound to a dog leads to absorption of both fat and hydrocarbon. The thoracic lymph obtained by fistula contains both oils and in about the same relative proportion as in the emulsion fed. This sug- gests a mechanical absorption of droplets of fatty acid and hydro- carbon oil mixtures. Isolated loops of the intestine of dogs, cats, and goats, were per- fused with defibrinated blood from the same animal. The loops were either removed at the height of protein digestion or amino acid and peptone mixtures were introduced. Samples of the per- fusing blood were taken at the beginning and at intervals during the experiment. Proteins from these samples were removed by mercuric nitrate, or phosphotungstic acid precipitation, or by coag- ulating in a boiling saturated solution of sodium or potassium sul- phate. Tyrosine was not found in the concentrated filtrates from the protein precipitations although Millon’s test is definite in dilutions of 1:100,000. No definite evidence of other amino acids could be found in the perfusate. THE RELATIONSHIP OF THE SUPRARENAL GLANDS TO SUGAR PRODUCTION IN THE LIVER. By J. J. R. MACLEOD anp R. J. PEARCE. (From the Laboratory of Physiology, Western Reserve Medical School, Cleve- land.) That the failure of stimulation of the splanchnic nerve to produce evidence of hyperglycogenolysis, after removal of the corresponding adrenal gland, does not indicate that a hypersecre- Society of Biological Chemists Xx1 tion of adrenalin into the blood is the cause of the hyperglycogeno lysis, which otherwise follows such stimulation, is shown by the fact that after complete section of the hepatic plexus splanchnic stimulation is usually without effect on the blood sugar. Further evidence of the direct nerve control of the process of hepatic glycogenolysis is that hyperglycaemia follows stimula- tion of the hepatic plexus. For this peripheral nerve control to be effective, however, there must be adrenalin in the blood, for, after double adrenalectomy, stimulation of the hepatic plexus is without effect on the blood sugar. METABOLISM IN AN EXPERIMENTAL FEVER WITH SPECIAL REFER- ENCE TO THE CREATININE ELIMINATION. By VICTOR C. MYERS anp G. O. VOLOVIC. (From the Laboratory of Physiological Chemistry, Albany Medical College.) Fever was induced in rabbits (ten experiments) by inoculation with the bacillus of hog-cholera. Determinations of total nitro- gen, urea, ammonia, creatinine, creatine, chlorides, potassium and phosphates, together with the morning and evening temperature observations, were made in the urine during the fever period and a previous control period of four or more days. The creatinine findings were of particular interest. The elimination of this sub- stance during the fever was found to parallel very closely the body - temperature, likewise the total nitrogen and urea, though the percentage of creatinine nitrogen in terms of total nitrogen dropped slightly at the height of the fever (3.8 to 3.3 per cent). The max- imum temperature (about 42° C.) was always found to be accom- panied by the highest creatinine elimination, the percentage in- crease over the normal elimination averaging 36 per cent during this period. The elimination of creatine did not always accom- pany the fever, but when present was generally observed following the crisis of the disease. The view is expressed that the increased creatinine elimination still represents the normal endogenous pro- tein metabolism which is proceeding at an abnormal intensity due to the increased temperature, while the presence of creatine sug- gests the exhaustion of the normal glycogen store of energy, and perhaps measures the amount of abnormal endogenous protein metabolism. XXll Proceedings THE ROLE OF PROTEINS IN GROWTH. By THOMAS B. OSBORNE anp LAFAYETTE B. MENDEL. (From the Laboratories of the Connecticut Agricultural Experiment Station and the Sheffield Laboratory of Physiological Chemistry of Yale University.) The proteins satisfy several functions in the growing organism. A certain minimum is necessary for the maintenance represented by Rubner’s ‘Abnutzungsquote.”” With an additional adequate energy supply any excess of protein beyond this maintenance requirement may, in the adult, experience temporary storage or be devoted to dynamogenic uses; but in the organism capable of de- velopment it may contribute to growth. The perfection of a product containing the non-protein constit- uents of milk (protein-free milk) in a form adapted to the specific needs of growing rats has made it possible to examine the efficiency of individual proteins in respect to maintenance and growth respec- tively. The investigations have indicated the inadequacy of all prolamines, viz., zein, gliadin, and hordein in contrast with efficient proteins such as casein, lactalbumin, ovalbumin, edestin, glycinin, and glutenin, in promoting growth. Gliadin and hordein satisfy the needs of maintenance in young animals; zein does not. It will be noted that-all of the inadequate proteins are deficient in two or more familiar amino-acid complexes (Bausteine). Details of these experiments are presented in Publication 156, Part II, Carnegie Institution of Washington (1911). THE ROLE OF SURFACE TENSION IN THE DISTRIBUTION OF SALTS IN LIVING MATTER. By A. B. MACALLUM. (From the Laboratory of Biochemistry of the University of Toronto.) The author in previous communications to the Society had shown in a number of instances that the Gibbs-Thomson principle of surface concentration of solutes could reasonably account for the condensation of potassium salts found, by microchemical methods, to occur on certain surfaces and inclusions of living cells and structures, and he ventured to claim that this principle is the dominant force in determining the distribution of salts in living Society of Biological Chemists XXlil matter. inthe present communication he brought evidence which definitely shows, in one group of instances, that this is the case. . In Marine Suctoria which, by alterations of surface tension of the superficial membrane or film at points on their surface, are able to protrude or retract tentacles formed of protoplasm, the distri- bution of potassium salts in the organisms is consequently affected. When the tentacles are being protruded the potassium salts be- come localized in their films and very little may be found in the cytoplasm generally. When the tentacles are being retracted the potassium salts begin to diffuse throughout the cytoplasm, where it remains until it is excreted, or until the tentacles are again being protruded, when it is once more condensed in the excessively thin surface films of the tentacles. A COMPARISON OF THE EFFECTS OF SUBCUTANEOUS AND INTRA- MUSCULAR INJECTIONS OF ADRENALIN UPON THE PRODUC- TION OF GLYCOSURIA. By I. S. KLEINER anp S. J. MELTZER. (From the Rockefeller Institute for Medical Research.) The experiments were made on rabbits, all of which received no food for twenty-four hours previous to the injection but received 100 ec. of water by stomach tube shortly before the injection of adrenalin. The urine was collected for twenty-four hours after the injection. The intramuscular injections were made into the lumbar muscles; subcutaneous, in the lower part of the abdomen. The doses of adrenalin ranged from 0.3 to 1.0 cc. of 0.1 per cent solution. In forty-nine rabbits the average amount of sugar eliminated following intramuscular injection was 0.73 gram: in forty-nine animals receiving like doses by subcutaneous injection, the average sugar excretion was 1.20 gram. A difference in favor of the subcutaneous injection was noted for every dose tested, though the difference grew less as the dose diminished. The great- est differences were observed with 0.7 or 0.8 cc. of adrenalin solu- tion. Of the forty-nine intramuscular injections, thirteen were not followed by giycosuria: of the forty-nine subcutaneous injec- tions, only four failed to cause glycosuria. In eight experiments, a dose of 0.75 cc. of adrenalin solution was injected subcutaneously distributed-over four different places. XXIV Proceedings Four of the animals excreted no sugar: the average sugar excretion of the other four was only 0.56 gram. The average excretion of sugar following the subcutaneous injection of this dose at a- single point was 1.52 grams. The experiments show that methods which favor the absorption of other substances are ess favorable for the production of glycosuria by adrenalin. THE HOURLY CHEMICAL AND ENERGY TRANSFORMATIONS IN THE DOG, AFTER GIVING A LARGE QUANTITY OF MEAT. By H. B. WILLIAMS, J. A. RICHE anp GRAHAM LUSK. (From the Physiological Laboratory, Cornell Medical College.) A calorimeter of the Atwater-Rosa type was constructed by Dr. Williams, which is capable of measuring with great accuracy the heat of combustion of alcohol and the oxygen absorbed and the carbonic acid produced during its combustion, during periods of one hour each. A dog which had been fed 700 grams of meat at noon of the pre- vious day, was placed in the calorimeter between 10 and 11 o’clock in the morning, and his metabolism measured. The animal was given 1200 grams of meat at noon and placed in the apparatus again. The heat production and other factors of metabolism were determined during hourly periods for twenty hours. 1. It was found that the direct and the indirect calorimetry agreed perfectly. 2. It was found that the heat production rose largely, and that this increase in heat production was proportional to the nitrogen eliminated in the urine, and was in no way proportional to the quantity of material present in the intestine. 3. It was found that the carbon which was retained from the protein ingested, must have been retained in the form of glycogen, since the absorption of oxygen during the different periods cor- responded exactly with this assumption, whereas, if the carbon had been retained in the form of fat, the oxygen absorption would have been 10 per cent less than that found. Further experiments have shown that glutamic acid added to a standard diet does not increase the heat production in any way. Society of Biological Chemists XXV CHEMICAL ANALYSES OF THE ASH OF SMOOTH MUSCLE. By L. A. RYAN anp EDWARD B. MEIGS. (From the Wistar Institute of Anatomy and the Hare Chemical Laboratory of the University of Pennsylvania.) The ash of the smooth muscle of the bull-frog’s stomach has been analyzed for potassium, sodium, phosphorus, and chlorine. The methods ci analysis have been in general those described by Katz in the Archiv fiir die gesammte Physiologie, 1896, 1xiii, p. 1, and the striated muscle of the same frogs was analyzed by the same methods for the same elements. The quantities of the ele- ments found in the striated muscle were about the same as those reported by Katz in the article mentioned above. In the smooth muscle three analyses were made for potassium, sodium, and chlo- rine; and four for phosphorus. The following quantities, given as percentages of the weight. of the fresh muscle, were obtained: IV AVERAGE per cent per cent per cent per cent | per cent Potassium. 2% a | 0.306 0.343 0.346 | 0.332 Sodiumsieifo-ee: 2 y..| 01052. | 0:065~1' 0.080 0.065 Phosphorus...............| 0.128 | 0.133 | 0.146 | 0.149 | 0.139 Eiloane ee. ....|- 02099 | 0.120 Beko: | 0.117 Samples of the tissue analyzed as ‘‘smooth muscle’ were exam- ined microscopically, both in the fresh state and after fixation by various histological methods, and it was determined that from 70 per cent to 85 per cent of the volume of the tissue was made up of smooth muscle fibers; about 5 per cent was extraneous connec- tive tissue; and the remainder, interstitial spaces between the muscle fibers. The results of our investigation indicate that smooth muscle contains somewhat less potassium and phosphorus and somewhat more sodium and chlorine than the striated muscle of the same animal, but that the differences in these respects between the two tissues are not by any means so marked as has sometimes been supposed. . XXV1 Proceedings QUANTITATIVE MEASUREMENT OF OXIDASES. By H. H. BUNZEL. (From the Bureau of Plant Industry, U. S. Department of Agriculture.) Accurate measurements were made of the oxidizing power of potato juice towards a series of aromatic substances, such as tyro- sine, benzidin, hydrochinone, a-naphthol, guaiacol, and others, and a comparative study of the behavior towards these substances was made. Experiments are described on the susceptibility of the oxidases toward poisons and heat. The fact that the oxidizing power of the juice is limited, and the reaction comes to completion after several hours, as found in the case of pyrogallol, has been con- firmed for about half a dozen other substances, as has also the fact that the extent of the oxidation is directly proportional to the quantity of enzyme used. An entirely new fact has been brought out: The oxidizing power of the juice towards different substan- ces to be oxidized is not additive: If one uses two or three oxidiz- able materials in the same experiment, the result is not a summation of the individual oxidations where the oxidation by the same juice ~ is measured separately, but corresponds roughly to the result obtained in the case of the most rapidly oxidized substance. THE ESTIMATION OF DEXTROSE IN BLOOD AND URINE BY THE DIFFERENCE IN REDUCING POWER BEFORE AND AFTER YEAST FERMENTATION. By J. J. R. MACLEOD, C. D. CHRISTIE anp J. D. DONALDSON. (From the Physiological Laboratory, Western Reserve University.) After treatment of urine with 10 per cent blood charcoal (Merck) in the presence of 15 per cent acetone or 25 per cent glacial acetic acid, we have not found any adsorption of added dextrose to occur.° When the reducing power of such clarified urine is estimated by Bang’s method before and after twenty-four hours fermentation with fresh brewers’ yeast there is not infrequently more reduction after fermentation than before: ; ° Cf. Woodyat and Helmholz: Journ. of Exp. Med., vii. p. 598, 1910; Andersen: Biochem. Zeitschr., xxxvil, p. 262, 1911. Society of Biological Chemists XXV1l Before: After: 0.083 0.072 0.039 0.042 0.087 0.062 It is therefore impossible to estimate the amount of dextrose in urine by the difference in reducing power, as estimated by Bang’s method, before and after fermentation. These results raised the suspicion that yeast might produce some substance which, although not sugar, caused reduction with _ Bang’s solution. We have therefore compared the reducing power as estimated by Bang’s, Allihn’s and Bertrand’s methods of: (1) a 10 per cent suspension of yeast in water; (2) a 10 per cent suspension of yeast in dextrose solution and have found that there is always some reduc- tion by Bang’s method but none, or the merest trace by the other methods. For example, one per cent dextrose solution after fer- mentation for twenty-four hours gave 0.07 per cent dextrose (?) (Bang) and 0.002 (Allihn). These results probably explain the high residual reduction found by Lyttkens, et al., in blood after yeast fermentation, and from which they conclude that a considerable proportion of the reducing substance in blood is other than dextrose. Thus in blood plasma (dog) we have found: Percent dextrose before fermentation: f Bang..............-. 0. 166 Nobertrand.:....%../.-. 0.165 Percent dextrose after twenty-four { Bang................ 0.060 hours fermentation: \Bertrands.s 44./2... -trace A NEW METHOD FOR THE DETERMINATION OF HIPPURIC ACID IN URINE. By OTTO FOLIN anp FRED F. FLANDERS. (From the Laboratory of Biological Chemistry, Harvard Medical School.) The method is based on the hydrolysis of hippuric acid, the extraction of the benzoic acid with chloroform and titration with sodium ethylate.!* To 100 cc. urine add 10 cc. normal NaOH. Evaporate to dry- ness on steam bath. Boil the residue 43 hours in a 500 cc. Kjeldahl 10 Journ. of the Amer. Chem. Soc.. xxxiii, p. 161, 1911. XXVlil Proceedings flask, fitted with Hopkins condenser, with 25 cc. of water and 25 ce. of concentrated HNO;. Dilute to 100 cc., saturate with am- monium sulphate and extract with 50, 35, 25 and 25 ce. portions of chloroform. Shake the extract with 100 cc. of saturated NaCl con- taining 0.5 cc. of concentrated HCl per liter. Separate and titrate with 3; sodium ethylate. The method gives theoretical results on pure hippuric acid and excellent duplicates on urine. SYNTHETIC ACTION OF ENZYMES. By H. C. BRADLEY. (From the Laboratory of Physiological Chemistry, University of Wisconsin.) A comparison of the fat content and lipase activity of various vertebrate livers showed no apparent relation between the two. Fish livers which are evidently active in the storage and utiliza- tion of fat, may contain ten times as much fat as mammalian liver; on the other hand the latter may be nearly ten-times as active hipolytically as the former. A quantitative comparison of lipolytic activity appears of doubtful value in supporting the theory of the synthetic function of enzymes in tissues. Mammary glands from goats and other animals taken at the height of lactation have thus far failed to show the presence of lactase, the enzyme which should synthesize as well as hydrolyze lactose. Lipase of the active gland is much less abundant than in the liver of the same animal. The autolytic ferments of mammary and liver tissues are about equal as measured by the rate of diges- tion of a sodium caseinate solution. It has thus far been impossible to secure definite evidence of the synthetic function of enzymes in living tissues. NOTE ON THE INORGANIC CONSTITUENTS OF HUMAN BLOOD.!! By C. C. BENSON. (From the Laboratory of Biochemistry, University of Toronto.) Whole blood and serum were analyzed for sodium, potassium, magnesium, calcium and chlorine, using modern methods. The results of analyses show slight variations from earlier analyses. 11 Read by title. Society of Biological Chemists XX1X ON SPHINGOSIN.” By P. A. LEVENE anp W. A. JACOBS. (From the Rockefeller Institute for Medical Research.) Sphingosin was discovered first by Tudichum on hydrolytic cleavage of phrenosin. It was rediscovered later by Thierfelder on decomposition of cerebron. Regarding the chemical structure of the base there existed very little information beyond the knowl- edge of its empirical formula. This was presented as C,;H3;N2O2. The present work was undertaken with a view of elucidating the structure of the base. The work is still in progress, but the data already obtained lead to the conclusion that sphingosin is an unsaturated amino-alcohol of the olefine series. The data on which the conclusion is based are the following. The substance contains all its nitrogen in form of primary amino- nitrogen, it forms a triacetylderivative, which no longer contains the original primary amino group. It absorbs hydrogen in a pro- portion equivalent to one unsaturated bond. The substance ob- tained by the last process has the composition of dihydrosphingo- sin, C,,H3;NOo. It was analyzed as its triacetylderivative. In connection with this mention must be made of the fact that the substance obtained by Thierfelder in the filtrate from sphingo- sin sulphate and which is described as a nameless base, is in fact dimethylsphingosin, which is formed in the process of pre- paring sphingosin. This view is based on the results of the deter- mination of the methyl] groups present in the molecule of the name- less base. ON GLYCOLYSIS.” By P. A. LEVENE anp G. M. MEYER. (From the Rockefeller Institute for Medical Research.) The work of the previous year brought to light the facts that the action of tissue extracts on glucose was either altogether nega- tive, or consisted in a condensation of the monosaccharide into a more complex form. Thus, the problem of the study of glycoly- sis had not been furthered by those experiments, and it remained 12 Read by title. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 2. XXX Proceedings necessary to discover a method by which could be studied the chemical process of sugar degradation within the living tissues. It was concluded to employ for these experiments leucocytes obtained under aseptic precaution and to perform the experiments under perfectly aseptic conditions. There was noted a marked fall of the reducing power of a sugar solution kept in contact with the leucocytes. After the action of the leucocytes the sugar solu- tion could not be brought to its original reducing power by boiling with mineral acids. The analysis of the products of glycolysis under such conditions revealed the absence of carbonic acid, or of any volatile acids. The only substance obtained from the reaction mixture was lactic acid. This was identified as its zine salt. The value of the lactic acid obtained in the experiments was lower than that of the sugar decomposed by the leucocytes. Control experiments with leuco- cytes failed to discover lactic acid. Thus, it is for the first time definitely established that lactic acid is an intermediate product of glycolysis. ON THE PICRATE OF GLYCOCOLL.* By P. A. LEVENE anv D. D. VAN SLYKE. (From the Rockefeller Institute for Medical Research.) In a previous publication by one of us (L.) the picrate of glycocoll was described having the composition of CzH;NO2.CsH3(NOz)s, and the M.P. = 190° C. The present investigation brought to light the fact that this substance was a mixture of glycocoll picrate of the composition (C,;H;NOz2)2 CsH3(NOsz)3, and of picric acid. Digly- cocoll picrate is a very stable salt, which can be recrystallized with- out altering its composition and its melting point. The melting point of diglycocoll picrate = 201° C. (corrected); on the other hand it is possible to prepare artificial mixtures of the picrate and of picric acid which melt as 191° C. 13 Read by title. Society of Biological Chemists xR A METABOLISM STUDY ON A FASTING MAN. By PAUL E. HOWE anp P. B. HAWK. (From the Laboratory of Physiological Chemistry, University of Illinois.) A second seven-day fast on “‘Subject E”’ was reported. Thecon- ditions of experimentation were as follows; a preliminary period, of high protein ingestion, 21.86 grams of nitrogen per day, a fasting period with constant. water ingestion; a low protein feeding period, 5.23 grams of nitrogen per day for four days, and a final high pro- tein feeding period, 21.86 grams of nitrogen per day for five days. Observations were made on the changes in body weight and the total nitrogen, urea, ammonia, creatinine and creatine excretions. _The phosphates, chlorides, acidity and hydrogen ion concentra- tion of the urine were also determined but not reported. Data on indican, ethereal sulphates and fecal bacteria were reported in an- other connection. The effects produced by the ingestion of a high protein diet previous to the fast as compared with those obtained in the previous fast which was preceded by a low protein period, were; a high rate of protein decomposition during the fast with an unusually high nitrogen excretion on the third day; a higher creatinine nitrogen excretion for the first three days of the fast which approximated the creatinine excretion of the first fast for the last four days; the same percentage relations between the total nitrogen and urea nitrogen excretions; a smaller loss of body weight. During the feeding period there were negative nitrogen balances with slight gains in weight during the low-protein feeding period and positive nitrogen balances and marked gains in weight for the high protein feeding period. There was a gain in weight of 4.11 kg. in eight days with a nitrogen retention of but 4.52 grams on an intake of 108.2 grams of nitrogen as contrasted with a gain of but 3.08 kg. in weight and a retention of 25.20 grams of nitrogen on an intake of 139.3 grams of nitrogen for the first fast. The data indi- cate that this gain was largely of non-nitrogenous substances other than water. If we accept the percentage distribution of nitrogen as a criterion of normal metabolism there was a return to such a condition on the fourth day of feeding. XXXil Proceedings HYDROGEN ION CONCENTRATION OF FECAL EXTRACTS. By PAUL E. HOWE anp P. B. HAWK. (From the Laboratory of Physiological Chemistry, University of Illinois.) The acidity of fecal extracts was determined during a series of experiments upon the effect of water-drinking with meals and dur- ing a seven-day fasting period and the subsequent feeding periods which consisted of a low- and a high-protein period. The general type of the diet was the same in all cases. The Salm type of hydro- gen electrode cell was used. The feces were extracted, by means of centrifugation, with 0.5 normal solution of Na,SO,. A mixture of 0.2 mole of NasHPO; and 0.1 mole of NaH»PO, was used as the standard of comparison. A two gram sample of moist feces was used in each instance. The hydrogen ion concentration of the fecal extracts did not show any pronounced changes as the result of the ingestion of in- creased amounts of water, with meals—the results varying between 1 X 107° and 0.1 X 107° mole of hydrogen ion per liter. As the result of fasting there was in general a decrease in the hydrogen ion concentration, from an average of 5.3 X 107° for the normal fasting period to 1.2 x 107° for the two fasting stools. The hydro- gen ion concentration is different for different individuals on the same diet but in general rather uniform for each individual. CONNECTIVE TISSUES OF LIMULUS." By H. C. BRADLEY. (From the Laboratory of Physiological Chemistry, University of Wisconsin.) A chemical study of the cartilage-like connecting tissues of the gill plates of limulus, and the fibrous and tendon-like tissues to which the pedal muscles are attached. 14 Read by title. Society of Biological Chemists XXXill THE RESPIRATION CALORIMETER AND ITS USES FOR THE STUDY OF PROBLEMS OF VEGETABLE PHYSIOLOGY." By C. F. LANGWORTHY anp R. D. MILNER. (From the Office of Experiments Stations, U.S. Department of Agriculture.) In reconstructing the Atwater respiration calorimeter which was transferred at the death of Professor Atwater from Middletown to the laboratory of the Department of Agriculture, improvements were introduced in the devices for controlling the temperature of the water flowing through the calorimeter, which carries out the heat generated in the chamber, and for automatically recording the difference in temperature of the water as it enters and as it leaves the chamber. This calorimeter has recently been applied to a new line of inves- tigations concerned with the ripening of fruit. Several bunches of bananas were placed in the chamber and kept under observa- tion during ripening, the oxygen consumption, carbon dioxide excretion and heat elimination being determined. The data ob- tained indicate that physical and chemical factors of both theoreti- cal and practical value may be measured with the respiration calorimeter, and afford evidences of the adaptability of this instru- ment to the study of fundamental problems of plant life. A new respiration calorimeter especially constructed for the study of the problems here alluded to is nearly completed in which the size of the chamber is reduced and in which such recording and controlling devices have been introduced as to make the apparatus nearly automatic in its operation. ON THE EXCRETION OF FORMALDEHYDE, AMMONIA AND HEXA- METHYLENAMINE."® By HUGH McGUIGAN. (From the Laboratory of Pharmacology, Northwestern University Medical School.) When formaldehyde is injected intravenously it is oxidized with surprising rapidity. One hundred cubic centimeters of 1 per cent formaldehyde, injected into a 10 pound dog in the course of one . 15 Read by title. XXXIV Proceedings and one half hours, completely disappeared from the blood within thirty minutes after the injection. Only formic acid was present in the urine. Formaldehyde is also excreted into, but less rapidly oxidized in, the intestine. In other instances free formaldehyde was found in the urine, only when large doses were given. A like amount of hexamethylamine, injected in the same manner, could be found in the blood several hours after the injection. It was also found (formaldehyde test) in the urine, bile, intestines, eye, saliva, bronchial secretions, amniotic fluid, eggs (hen) and sweat (human). Free formaldehyde is at least much harder to detect in these fluids. The tests in most cases were negative. Ammonia is not excreted by the lungs. Combined with formal- dehyde, however, it is found in the bronchial secretions. From the similarity of the alkaloids to ammonia it was thought that, per- haps morphine, which is not excreted normally by the kidneys, might pass through if administered with formaldehyde. No posi- tive result on this point has been obtained. ‘ GLYCOLYSIS, AS MODIFIED BY REMOVAL OF THE PANCREAS AND BY THE ADDITION OF ANTISEPTICS."® By H. McGUIGAN anp C. L. von HESS. (From the Laboratory of Pharmacology, Northwestern University Medical School.) Repetition of previous work showed that mixtures of extracts of normal muscle and of pancreas, with toluol or chloroform added as an antiseptic, caused no or only slight glycolysis, but not more than normal muscle extract alone. The above suggestions are open to two criticisms: 1. (Suggested by Woodyatt, also by Oppenheimer) if, by Cohn- heim’s theory, there be a pancreatic internal secretion, the normal muscle would probably contain enough of it to exert maximal glycolysis, hence no change will be obtained by the addition of pan- creatic extract. However, muscle of pancreatectomized dogs, | similarly tested, gave no action on glucose either with or without the addition of pancreatic extract. e 16 Read by title. Society of Biological Chemists XXXV 2. Aseptic glycolysis of yeast or of blood is greatly inhibited by antiseptics in the concentrations used in the muscle experiments. The foregoing method, which involves the use of antiseptics, de- stroys normal glycolysis to such a degree that the results obtained by it can prove neither the presence nor the absence of an internal secretion of the pancreas. It further indicates that normal gly- colysis is due more to cellular than to enzyme activity. EFFECT OF THE QUANTITY OF PROTEIN INGESTED ON THE NUTRITION OF ANIMALS: VI. ON THE CHEMICAL COMPOSI- TION OF THE ENTIRE BODY OF SWINE." By A.D. EMMETT, W. E. JOSEPH anp R. H. WILLIAMS. (From the Laboratory of Physiological Chemistry, Department of Animal Husbandry, University of Illinois.) Three lots of young pigs, four in a lot, were fed on low, medium, and high protein planes. One pig of the low protein lot and two from each of the medium and high protein lots were subjected to detailed slaughter tests as soon as the medium or standard fed pigs reached marketable weight and condition. The various parts of the animals were analyzed. It was found: (1) That, out of the four pigs kept on the low protein plane, three died before the close of the experiment. These pigs grew slowly, were drowsy, lacked vigor and became stiff in their joints. The pigs of the other two lots were thrifty, in good condition and grew normally. Blood counts showed no definite differences. (II) That the average daily gains for the 174 days of individual feeding were 0.33, 0.85, 0.90 pounds respectively for the low, medium, and high fed lots. From the standpoint of economy of gains, the medium pretein-fed lot made the best show- ing. (III) That the chemical data for the entire bodies of the five slaughtered pigs showed the medium and high protein-fed lots to be remarkably similar in their percentages of water, fat, protein, ash, and phosphorus. The pig of the low protein lot had a low percentage of fat and a high one of ash and phosphorus. Compar- ing the data from the five animals with the average of two repre- sentative pigs, slaughtered at the beginning of the experiment, 17 Read by title. XXXV1 Proceedings the percentage increase of dry substance, protein, fat, ash, and phosphorus was lowest in the low protein-fed pig. In case of the medium and high protein-fed pigs the percentage of increase of the nutrients was practically the same. In the majority of instances the differences between the averages within the lots were greater than those between the averages of the lots. (IV) That the aver- age chemical composition of the bodies of the five pigs, at about 200 pounds live weight and in good marketable condition, was on the fresh basis: water, 45.74; protein, 14.38; fat, 37.20; ash, 3.84; and phosphorus, 0.673 per cent. THE EFFECT OF QUININE ON CULTURES OF PNEUMOCOCCI.'8 A PRELIMINARY Reronr By O. H. BROWN. (From the St. Louis University School of Medicine.) Numerous reports of apparent specific curative effects of large doses of quinine in penumonia led me to use it in a small number of cases of this disease. While the results were highly satisfactory, I afterward concluded that they were probably accidental because of the apparently trustworthy claims made by other clinicians that quinine failed to show any beneficial result in their cases. I have however carried out tests upon the antiseptic power of quinine and its salts of citric, sulphuric and salicylic acids upon pneumococci in vitro. Sterile tubes of human blood bouillon were prepared, half of them containing quinine in the form and percentage (0.05-0.1 per cent) which I desired to test. Inocu- lations into plain bouillon (control) and into quinine bouillon were made. Plate cultures on blood agar made immediately and afterward at varying intervals showed the growth or destruction of pneumococci in such tube. Thirty strains of pneumococci from various sources have been tested. The results showed: (1) Pure quinine is more destructive to pneumococci than are its salts. (2) The time required for 0.1 per cent quinine to kill pneumococci varies from twenty minutes to four or five hours. (3) Other organisms, such as streptococci and staphylococci, are destroyed only by much longer exposure to quinine. 18 Read by title. Society of Biological Chemists XXXVII MAINTENANCE AND GROWTH. By THOMAS B. OSBORNE anp LAFAYETTE B. MENDEL. (From the Laboratories of the Connecticut Agricultural Experiment Station and the Sheffield Laboratory of Physiological Chemistry of Yale Uni- versity.) In connection with the authors’ feeding experiments with isola- ted food-substances it has been found that diets which are satis- factory for the maintenance of full-grown animals are entirely inadequate to induce growth in ungrown individuals. The sus- pension of growth on a maintenance diet here referred to is not that caused by an insufficient supply of energy, but is a retarded development associated with the chemical make-up of the diet. These chemical features of the diet essential for proper growth involve not only the type of protein, but likewise certain non-pro- tein components (presumably the inorganic ingredients). Dwarf- ing, in the sense of maintenance of both weight and size, can readily be brought about in young animals; and the capacity to grow can be maintained unimpaired by such stunted individual for many months. The non-protein constituents of the diet can be pre- pared from the protein-free portions of cow’s milk (protein-free milk) in a form suitable to permit proper growth.. The experi- mental records of rats, selected as the animals for study because they manifest the utilization of a suitable diet speedily by measur- able changes in size, are presented in Publication 156, Part LI, Carnegie Institution of Washington (1911). THE STUDY OF ENVIRONMENT. By WILDER D. BANCROFT. (From the Department of Physical Chemistry, Cornell University.) When studying the effect of environment on an organism, we must distinguish three distinct things: the direct effect of new external conditions involving no adaptation; the adaptation of the organism to the new conditions; and the possible inheritance of the adaptations. The botanists have not made these distinctions. They consider the change of curvature of tendrils with change of temperature as a case of non-adaptive response, whereas it has no more to do with adaptation than the shortening of a fishing-line when it is wetted. XXXVI111 Proceedings The problem of the inheritance of acquired characters has been complicated unnecessarily by the arbitrary limitation that the character must be inherited for four or five generations after the organism has been brought back to the original surroundings. Since an organism which responds readily to a new environment will also revert readily when brought back, this definition has probably excluded most of the cases in which the inheritance of acquired characters could be shown. The biologists seem never to have realized that inheritance is primarily a hysteresis phenom- enon and should be studied as such. THE SYNTHESIS OF THIOTYROSINE. By TREAT B. JOHNSON. (From the Sheffield Laboratory of Yale University.) A knowledge of this new amino-acid was especially desirable, in order to acquire a more definite conception of the true nature of sulphur combinations in proteins. The acid has been prepared by the application of a new, general method for the synthesis of a-amino acids and its chemical properties are now being studied. The most important characteristic of the acid, so far observed, is the fact’ that it does not give Millon’s test. On the other hand, it gives, on warming the concentrated sulphuric acid, as charac- teristic color reaction as the Millon’s test is characteristic for tyrosine. This study is one of a projected series on new sulphur combinations which has been planned for the Sheffield Laboratory. THE RELATION OF OHIO BOG VEGETATION TO THE CHEMICAL NATURE OF PEAT SOILS. By ALFRED DACHNOWSKI. (From the Department of Botany, Ohio State University.) Analyses are submitted showing that several types of vegeta- tion of varied growth-form occur upon a habitat essentially similar in range of chemical composition. The prime conditions determining distributional relationships and succession are not the mineral salts in the soil but biochemical processes. The vari- able composition of peat renders it necessary to determine experi- mentally what organic substances are absorbed and of value or injurious in nutritive metabolism. _ Society of Biological Chemists XX X1X PHYTOCHEMICAL STUDIES IN CYANOGENESIS. By C. L. ALSBERG ann O. F. BLACK. (From the Bureau of Plant Industry, U. S. Department of Agriculture.) The relation between the nitrates in the soil, nitrification dur- ing drought, and cyanogenesis in sorghum, based on experiments done at the Arlington Farm in the course of the past summer, is discussed, and an incidental error in the common method of deter- mining hydrocyanic acid in plants is pointed out. THE NITROGEN EXCRETION OF THE MONKEY, WITH SPECIAL REFERENCE TO THE METABOLISM OF PURINES. By ANDREW HUNTER anp MAURICE H. GIVENS. (From the Department of Physiology and Biochemistry, Cornell University.) A female monkey (Cercopithecus callitrichus), weighing 4.7 kilo- grams, was maintained for forty days on a daily ration of 200 ce. whole milk, 200 grams bananas, and 20 grams peanuts. The urine was collected every forty-eight hours. For the first sixteen days the average daily excretion of N was 1.83 grams, distributed as follows: urea, 1.59; NH, 0.028; creatinine, 0.065; allantoin, 0.015; purines, 0.0027; undetermined, 0.13 grams N; or, urea 86.9; NHs, 1.5; creatinine, 3.5; allantoin, 0.82; purines, 0.15; undetermined, 7.1 per cent of total N. Uric acid could not be detected. During the remainder of the experiment attention was devoted particularly to the metabolism of endogenous and exogenous purines. On seven normal two-day periods the excretion of allantoin N ranged from 27,0 to 31.8, that of purine N from 4.7 to 10.3 mgs. On five periods, each interpolated between two normal ones, doses of 0.5, 0.5, 1.0, 1.0, and 2.0 grams sodium nucleate were adminis- tered. Of the purine N thus fed 90, 56, 41, 24, and 29 per cent respectively of the theoretically possible was recovered in the form of allantoin and urinary purines. Of the amount so recovered 79 to 98 per cent took the form of allantoin; after the second dose of 0.5 grams 2 per cent, and after 2.0 grams 9 per cent appeared as uric acid. In normal periods allantoin accounted for 71—87, in nucleate periods 77-86, per cent of the total purine-allantoin N. In respect of the ratio between allantoin and purine excretion the x] Proceedings species examined resembles the lower mammals rather than man. On the other hand we did not meet with the almost quantitative conversion of exogenous purines into allantoin, which has been reported for the dog. THE DEFINITION OF NORMAL URINE.'® | By JOHN H. LONG. (From the Laboratory of Physiological Chemistry, Northwestern University Medical School.) Our notions as to what is a normal urine have undergone many changes in the years which have elapsed since the first attempts were made to establish standards. The same individual, at one time on a high protein diet and again on a low protein diet, will excrete urine which may be markedly different in many ways, and yet both be normal. : Improved methods of examination have shown that hyaline casts are much more frequently present in the urine of healthy men than was suspected a few years ago; and it must be admitted that traces of albumin occur in the urines of men, who, from all ordinary points of view are perfectly well. The statement as to what constitutes normal urine must take cognizance of these facts and of the further fact that for each individual there seem to be agencies at work which modify the nitrogen distribution, the acidity and the neutral sulphur in ways which we cannot account for. In a certain sense each individual has his own standard of normality. SHOULD THE TERM PROTAGON BE RETAINED. By WALDEMAR KOCH. (From the Laboratory of Physiological Chemistry and Pharmacology, Univer- sity of Chicago.) Data were presented which indicated that the preparations referred to as protagon contain at least three substances: a phos- phatid containing cholin, a cerebrosid-containing sugar, a complex combination of a cholin-free phosphatid with a cerebrosid to which an. ethereal sulphuric acid group is attached. The term protagon cannot therefore be said to have any chemical significance. The de- tails will be presented in a more extended publication. 19 Read by title. Society of Biological Chemists xli OXIDIZING ENZYMES IN CERTAIN FUNGI PATHOGENIC FOR PLANTS. By H. S. REED AND H.S. STAHL. (From the Laboratory of Plant Pathology, Virginia Agriculture Experiment Station, Blacksburg, Virginia.) The oxidizing ability of the plant extract is often altered as a result of the invasion of parasitic fungi. The extracts of apples invaded by Sphaeropsis malorum show no oxidizing powers what- ever. Apples attacked by Glomerella ruformaculans show on the contrary a somewhat increased oxidizing ability. When grown in pure culture on synthetic media Glomerella develops oxidizing enzymes in certain media but not in others. MODIFIED COLLODION MEMBRANES FOR STUDIES OF DIFFUSION.?° By WILLIAM J. GIES. (From the Laboratory of Biological Chemistry of Columbia University, at the College of Physicians and Surgeons, New York.) Lipins and many substances which dissolve in ether, alcohol and similar solvents can be dissolved, in large proportions, in U. S. P. collodion solution without inducing prceipitation of the collodion. Such mixed solutions, when treated in any of the usual ways for the production of collodion membranes, yield composite homogeneous products. If the proportion of added substance is not too large it is wholly incorporated uniformly in the resultant composite membrane. Lecithin, cholesterol, lard, olive oil, rubber, alcohol-ether soluble protein, organic pigments, ferric sulfocyanate and many other substances have been incorporated homogeneously in such modified collodion membranes. Mem- branes prepared in this way show interesting differences in per- meability in diffusion experiments, according to the general nature of the incorporated materials. Such membranes promise to afford valuable means of studying cell permeability and osmosis in general under biological conditions. I am proceeding with various types of experiments with such modified collodion mem- branes in the hope of extending our knowledge in these particular directions. 20 Read by title. x]il Proceedings A METHOD FOR DIFFERENTIATING BETWEEN ‘METABOLIC”’ AND RESIDUAL FOOD NITROGEN OF THE FECES.*! By MORRIS 8S. FINE. (From the Sheffield Laboratory of Physiological Chemistry, Yale University.) It is hardly necessary to point out that the nitrogen of the feces is in great part composed of bacteria, unabsorbed intestinal secre- tions, etc. If the quantity of this “metabolic” nitrogen were known, the nitrogen of the food actually escaping absorption could readily be estimated. Investigators have sought a measure of the “metabolic” nitrogen in the feces obtained during starvation or from a digestible non-nitrogenous diet; or the attempt has been made to differentiate by chemical means. Asa rule, such methods do not take into account the fact that the indigestible materials, e.g., cellulose and hemicellulose such as are present in cereals, legumes, etc., show a marked tendency to increase the elimination of fecal material. That this is a consideration of some import- ance is shown in a paper from this laboratory, now in press. The following procedure is believed to offer certain advantages over those hitherto proposed. From the fecal nitrogen accruing from a given diet is subtracted the corresponding value resulting from a non-nitrogenous diet, yielding practically the same amount of feces. Such a non-nitrogenous diet may be conveniently obtained by adding agar-agar to non-nitrogenous food whose calorific equiv- alent does not differ materially from that of the diet under investi- gation. The result thus obtained represents the amount of nitro- gen of the latter diet which has escaped utilization. BIOCHEMICAL ANP BACTERIOLOGICAL STUDIES OF THE BANANA.*! By E. MONROE BAILEY. (From the Connecticut Agricultural Experiment Station.) An earlier study” has been extended. Enzymes concerned in ripening processes have been investigated, and in addition, bac- teriological and chemical examinations of the fruit in various stages of maturation have been made. Amylase, sucrase, raffinase, pro- 21 Read by title. 2 Journal of Biological Chemistry, i, p. 355, 1906. Society of Biological Chemists xlili tease, lipase, and peroxidase were detected. Tests for maltase, dextrinase and lactase were doubtful or negative. The inner por- tion of the pulp of sound fruits appears to be sterile, but the regions of the inner coats of the peel may be sparsely inhabited by bacteria. As ripening progresses, starch disappears and the content of alcohol- soluble sugars and dextrine increases. Maltose could not be de- tected. PREPARATION OF CREATINE AND CREATININE FROM URINE. By STANLEY R. BENEDICT. ESTIMATION OF CREATININE. By STANLEY R. BENEDICT. CREATINE ELIMINATION IN THE PREGNANT DOG. By J. R. MURLIN anp H. I. MULLER. THE IODINE CONTENT OF THYROID GLANDS OF SHEEP FED MAINLY UPON MARINE ALGAE. By ANDREW HUNTER anp SUTHERLAND SIMPSON. RECOVERY OF ALCOHOL FROM ANIMAL TISSUES. By P. J. HANZLIK.” CHANGES IN THE COMPOSITION OF BLOOD AND MUSCLE FOL- LOWING DOUBLE NEPHRECTOMY AND BILATERAL URETERAL LIGATION. By H. C. JACKSON. 23 Journal of Biological Chemistry, xi, p. 61, 1912. anes wet = Gime ©. }: STUDIES IN NUTRITION. V. THE UTILIZATION OF THE PROTEINS OF COTTON SEED. By LAFAYETTE B. MENDEL anp MORRIS S. FINE. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven, Connecticut.) (Received for publication, September 25, 1911.) The influence of cotton-seed on the well-being of cattle has been extensively investigated in this country, the protein of this mate- rial being 88 per cent! utilized by steers or sheep. It was of interest to learn to what extent this substance was utilized by dogs, the alimentary canal of which more closely resembles the human diges- tive tract. Such experiments are of special import, inasmuch as cotton-seed flour bids fair to become an important article in the human dietary. As far as we are aware, an investigation of this nature is not on record.? EXPERIMENTAL PART. Product Employed. The cotton-seed* flour of these experiments was a deep yellow impalpable powder, containing 7.4 per cent nitrogen. Fraps* found similar samples to have 4.0 to 6.5 per cent crude fiber. Cotton-seed flour contains some pentosans but no starch.° 1 Cf. Fraps: Texas Agricultural Experiment Station, Bull. 128, 1910. 2 Correspondence with Dr. C. F. Langworthy and Dr. Marion Dorsett, of the United States Department of Agriculture, also fail to reveal any lit- erature on this subject. - 3 Obtained from the Southern Cotton Oil Company, Charlotte, N. C. 4 Fraps: loc. cit. 5 Fraps: loc. cit. THE JOURNAL OF BIOLOGICAL CHEMISTRY XI, NO. 1. I 2 Utilization of Cotton Seed Proteins Metabolism Experiments. In Table 1 are recorded three experiments on the utilization of cotton-seed flour. The usual method of procedure® prevailed. The daily supply of cotton-seed contained 2 to 3 grams of crude fiber. The cotton-seed feces of dogs 5 and 6 were hydrolyzed according to the method outlined in a previous paper,’ and yielded a daily average of respectively 5 and 3.5 grams of hemicelluloses. The diets of these two dogs, therefore, included 7.5 and 6 grams of indigestible non-nitrogenous substances. This, however, cannot account for the manifestly poor utilization of the cotton-seed nitro- gen. The coefficients of 67 to 75 per cent for cotton-seed contrast TABLE 1. Cotton-seed Flour. Doe 5 | Doe 6 | Doc 7 | PERIOD XIX PERIOD XX j PERIOD IX (4 days) (4 days) | (3 days) | Cotton-seed Feed- | Cotton-seed Feed- | Cotton-seed Feed- ing | ing | ing grams | grams | grams | Cotton-seed Cotton seed | Cotton-seed Flour 45 Flour 45 | Flour 45 | Sugar 25 | Sugar 25 | Sugar 20 | Lard 20 | Lard 20 | Lard 25 Composition of daily diet {| Water 225) Water 225 | Agar 3 | Bone Ash 7 | Water 175 Estimated Estimated | Estimated | calories 410! calories 410 calories 440 Nitrogen output. | Daily Averages | Daily Averages | Daily Averages Urine nitrogen, gm......... 2.61 2.61 2.55 Total nitrogen;gm:. ...:..:-- | 3.51 3.70 3.45 Nitrogen in food, gm....... .| 3.32 3.32 3.59 Nitrogen balance, gm........ —0.20 —0.38 +0.14 Feces. Weight air dry, gm.......... 2325 23.9 31.7 Nitrogen, gm...............| 0.91 1.09 0.90 Nitrogen, percent..........| 3.87 4.57 2.84 Nitrogen utilization, per Cent.:. -. 62 eee am 72.6 67.2 74.9 6 Cf. Mendel and Fine: This Journal, x, p. 303, 1911. : 7Cf. Mendel and Fine: Jbid., x, p. 339, 1911. Lafayette B. Mendel and Morris S. Fine 3 strikingly with those of 88 to 93 per cent for meat diets containing comparable or greater amounts of such indigestible materials. (See Table 2.) There is of course the possibility that the cotton- seed flour employed in this study contained some constituent® which either inhibited secretion or promoted premature evacua- tion—conditions which would result in poor utilization. TABLE 2. Utilization with Reference to Indigestible Materials in the Diet.* Daily Averages. waa |e Zz estes bce |) se } izsa | == 4 =n pa Bea | ba EB a Z | NATURE OF INGESTA Bae ies A p ag OZ, | Ra z PA oe Paik) aR | a P & [= & a z < grams | grams | grams per cent | per cent | 3 Spa: | 72:6 Cotton seed 3 Gimitoeon Ore | L6 7 J seh 4b S26 | 7520 | | 5 P xviii 4 vee. 1-53.38. 90.5 Go| xb 4 Meat | 6 6 | 3.3 | 89.2 | 91.0 7 lexwa 6 G33 93.5 | 5 | xv 4 |) Meat | 6M e138"! 33 | 91.6 | 6 4 Bone ash, 5 grams | G7 43 | 3.3 | 87.7 | 89.2 4 Agar, 2 grams Le ke ISUaTSo eSoeo" | aEIE ee es ‘nie * This problem will be treated in detail in a subsequent paper of this series. 7 Including respectively about 5 grams and 3 grams hemicelluloses which escaped digestion. t Including approximately 4 grams indigestible hemicelluloses (exclusive of agar), i.e., the aver- age of }gramsand 3 grams. An actual determination of the hemicelluloses of the feces of this experiment was not made. 8 Cf. Crawford: Journ. of Pharm. and Exper. Ther., i, No. 5, p. 519, 1910. ms OS ee urea Mia as elses heer ‘pro te Sebi nk ha elec a ent Wiyraw Wad iw: & cat . ‘ oN iets piper My crest sah bd 2 by Fy bear: yr eee) a A t wnat oaths, Siting hh ae Videbin, Adee ( i r 7 G ae x dnal,'¢9 + Sn S meet, b ipa “i Si. SOteae GR: ees Away QS aie? oy Se :* ar: Be Sepa ae ; 7 ue ee oye ak wh ransrhnd 29 Side Dads 102 Vote: : Li hie cant ete at : A fis ioe cst mits do ‘on Liege tute ae ee a = a 7 = ie A ‘e _ . : Oe ; a. 2 cous we) = ei. « D > : ’ ti =u arate — 1" . - oS ee sa r Se ara “teas T0525 Be TE! ee i ay Boh eran Of ro y = ic z STUDIES IN NUTRITION. VI. THE UTILIZATION OF THE PROTEINS OF EXTRACTIVE-FREE MEAT POWDER AND THE ORIGIN OF FECAL NITROGEN. By LAFAYETTE B. MENDEL anp MORRIS 8S. FINE. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven, Connecticut.) (Received for publication, September 25, 1911.) CONTENTS. The utilization of the proteins of extractive-free meat powder......... 5 Earlier studies with this and related materials.....,.............. 5 LOSSTE SEPT, [OCULAR ee peer co ah loan: 2a ae 6 SRT ESDGe CiPpTE To Dein oes oe aie A nr ee 6 WRAP MRI eZ PCTIMENGS. cals 44s ccs ec ai ee ee nt 6 MMP aRICIMECANIMILTOLEN... .. oases eee nase see eee ke eee 10 JRtnce SUG i on hae amet ” oS ee 1] “DOPE SEED) je eg eee ae ie eo Oe 15 Influence of indigestible non-nitrogenous materials upon the MRE MERPALIStICS Of the feces: .762een 22k 6. ee we 15 Influence of poorly utilized highly nitrogenous matter upon ihe witrogenistatistics of the feces. 1.22. ).0...2:...0.6.0.0.. 18 Simultaneous influence of both types of materials upon the maropenssisiainties Of the feces. :. 22.52. ss 6. be ee eee 19 Effect of previous thorough evacuation upon utilization Be ook 21 Estimation of ‘‘metabolic’’ products in the feces.............. 21 The utilization of the vegetable proteins.............. Lhd et se 23 THe UTILIZATION OF EXTRACTIVE-FREE MEAT POWDER. EARLIER STUDIES. Forster was the first to conduct an investigation with this mate- rial. His immediate problem was the question of salt metabolism, but incidentally we note that the nitrogen was 91 to 96 per cent available. During the past twenty years, considerable attention has been paid to the comparative utilization of fresh meat and 5 6 The Utilization of Proteins dried meat preparations, for example, ‘‘soson,”’ “‘somatose,”’ “tropon,” and the meat residues from meat extract factories. Passing over the literature previous to 1901, we may dwell briefly upon the results obtained by Prausnitz, which are in general accord with those of the earlier workers. The average coefficient of digestibility of dried meat was 90 per cent against a coefficient of 93 per cent for fresh meat. Moreover the nitrogen concentra- tion of the dried-meat-feces was 1.35 to 1.76 per cent higher than the fresh meat feces. These facts make it probable that a portion of the dried meat had escaped absorption. Prausnitz also showed that dried meat was less readily digested in artificial gastric juice than fresh meat. He accounted for these phenomena on the assumption that a not inappreciable length of time elapses before the dried meat particles are sufficiently “hydrated” to permit the digestive enzymes to operate. Max Voit found similar although less striking differences. Considerable work has also been accomplished with dried blood preparations, but a consideration of these investigations would lead us too far afield. EXPERIMENTAL PART. Product Employed. The meat residue! employed in the present studies was a light brown impalpable powder, containing 13.2 per cent of nitrogen, 8.9 per cent of ether extract, 2.5 per cent of ash, and 7.0 per cent of inoisture. Metabolism Experiments. Tables 1-8. During these experiments, the methods described in a previous paper? were followed. The utilization of the nitrogen of meat powder is distinctly, although slightly lower than that of fresh meat. The relatively high nitrogen concentration of the meat powder feces is indicative of a loss of this material through the excrement. These points are concisely presented in the accompanying brief tabular summary. 1 Obtained from Armour and Company. 2 Mendel and Fine: This Journal, x, p. 303, 1911. Lafayette B. Mendel and Morris S. Fine 7 Summary of the Data-on Nitrogen Utilization (see Tables 1-3) MEAT POWDER FRESH MEAT (AVERAGES) DOG Papoee a Nitrogen in feces ogee. | Nitrogen in feces tet ——————e——————EEE per cent per cent per cent per cent 1 91.3 2.98 94.0 1.94 4 89.3 3.81 94.5 2.04 4 | 91.0 3.87 93.7 2 36 TABLE 1. Extract-free Meat Powder. SUBJECT, DOG 1 PERIOD V ate lay se PERIOD VII Weight at beginning, 14.6 kg. (4 days) Meat Bede (4 days) Weight at end, 14.6 kg. Meet Feeding Feeding Meat Feeding aE grams grams grams Meat 300 | Meat Pow- Meat 300 der 80 Lard 60 | Lard 60 | Lard 60 “4° “yas Agar 5 | Agar 5 | Agar 5 (G osition of daily diet pesnceres é Boneash 15 | Bone Ash 15]! Bone Ash 15 Water 300 | Water 500 | Water 300 Estimated Estimated Estimated — Nitrogen output. Urine nitrogen, gm......... Total nitrogen, gm.......... Nitrogen in food, gm........ Nitrogen balance, gm...... Feces. Weimhiaimary. 4.00.02... 2 = Nitrogen, per cent.. Nitrogen Eiliation, per calories 1070 calories 860 | calories 1070 i Daily Averages Daily Averages Daily Averages 8.81 9.38 10.44 +1.06 29. Hos on c oO oun 94.5 8.76 9.68 10.53 +0.85 31.0 0.92 2.98 91.3 — 8.47 9.15 10.46 +-1.31 * Food almost entirely forced. TABLE 2. The Utilization of Proteins Extract-free Meat Powder. SUBJECT, DOG 4 PERIOD V | *Beavaha PERIOD VII Weight at beginning 4.9 kg. (4 days) MESA Powder (4 days) Weight at end, 5.1 kg. Meat Feeding Feeding Meat Feeding grams grams | grams Meat 150 | Meat Meat 150 Powder 39 Sugar 25 | Sugar 25 | Sugar 25 Starch 5 | Starch 5 | Starch 5 ve - : | Lard 20 | Lard 25 | Lard 20 Composition of daily diet. | Bone Aake10 |: Wea 3. leAgear 3 Water 200 | ‘‘Salts’’ 4 | “Salts” 4 Water 260 | Water 200 | Estimated Estimated Estimated calories 570| calories 510) calories 570 Nitrogen output. Daily Averages | Daily Averages | Daily Averages Urine nitrogen, gm......... 4.23 3.95 4.37 Total nitrogen, gm......... 4.50 4.50 4.70 Nitrogen in food, gm........ 5.40 5.138 5.40 Nitrogen balance, gm......., +0.90 +0 43 +0.70 Feces. Weight air dry, gm.......... 15.0 14.4 14.0 Nitrogensgmen ssc 0.27 0.55 0.32 Nitrogen, percent.......... 1.79 3.81 2.29 Nitrogen utilization, per CONG) Sree ee 95.0 89.3 94.1 Lafayette B. Mendel and Morris S. Fine TABLE 3. Extract-free Meat Powder. SUBJECT, DOG 4 Weight at beginning, 5.1 kg. Weight at end, 5.2 kg. Composition of daily diet. Nitrogen output. Urine nitrogen, gm Total nitrogen, gm.......... Nitrogen in food, gm...... ..| Nitrogen balance, gm...... | Feces. Weight airdry, gm.......... INitrogenhoniecseins es Nitrogen, percent.......... Nitrogen utilization, per CON recy ees Wise usss,b es PERIOD XI | (5 days) Meat Feeding ee SS grams Meat 150 Sugar 25 Starch 5 Lard 20 Agar 8 **Salts”’ 4 Water 200 Estimated calories 570 | 9 PERIOD XII (Sidon) PERIOD ‘Xu Meat eee Mee ee ng grams) grams Meat | Meat 150 Powder 40) Sugar 25 | Sugar 25 Starch 5 | Starch 5 Lard 25 | Lard 20 Agar 8| Agar 4 “Salts’’ 4 | Bone Ash 8 Water 300) Water 200 Estimated Estimated calories 510 | calories 570 Daily Averages Daily Averages Daily Averages 4.28 4.62 5.20 +0.58 4.51 4.98 5.26 +0.28 4.46 4.77 5.22 +0.45 16.0 0.31 1.96 94.0 10 The Utilization of Proteins ON THE ORIGIN OF FECAL NITROGEN. In previous papers® of this series we have followed the current custom of basing the data for nitrogen utilization upon the rela- tion of the nitrogen appearing in the excrement to that of the ingesta. This procedure would be strictly correct only in case the fecal nitrogen consisted entirely of food residues. As a matter of fact, there is abundance of evidence in the literature to demon- strate that fecal nitrogen in great part emanates from ‘metabolic products.’’4 Obviously an adequate understanding of the source of fecal nitrogen and the conditions influencing its excretion is essential for the proper interpretation of experiments on nitrogen utilization. In the earlier papers referred to we have at times pointed out that an apparently poor utilization was probably in- duced by the indigestible matter—cellulose, hemicellulose—inher- ent in the experimental material. The influence of such materials upon utilization has not always been fully appreciated. Rubner, and later Wicke, did indeed call attention to the unfavorable effect of cellulose upon the utilization of bread nitrogen; but in these cases it is difficult to decide in what measure the insufficiently ruptured cells are responsible for the low coefficients of digesti- bility, and to what extent the latter is to be attributed to the cel- lulose per se. This question is not satisfactorily answered by the poor utilization of meat obtained by Hoffmann when coarsely cut straw was added to the diet. Such coarse particles probably unduly irritated the digestive tract, resulting in increased secre- tion and peristalsis. Lothrop demonstrated an increased elimi- nation of fecal nitrogen when bone ash was added to the diet. In the present paper the nitrogen of the excrement under a variety of conditions is discussed briefly from the historical aspect ;* data purporting to show to what extent indigestible non-nitroge- nous substances may influence the amount and character of the feces are presented; and a plan of experimentation is proposed, 3 See footnotes 20-24, pp. 23 and 24. 4 By this term is understood intestinal secretions, cast off cells, bacteria, etc. For aconsideration of the important réle of bacteria in this respect and the literature related thereto, see MacNeal, Latzer and Kerr: Journ. of Infect. Dis., vi, p. 123, 1909. 5 For a more detailed review reference is made to Tsuboi (see bibli- ography). Lafayette B. Mendel and Morris S. Fine II with which it seems possible to approximately determine to what degree the nitrogen excreted in the feces is derived from undigested or indigestible nitrogenous constituents of the ingesta. Were this known, the term “utilization”? would be eminently appropriate. EARLIER STUDIES. Feces in Starvation. Man. The accompanying table presents oft quoted data® obtained from the professional fasters, Cetti and Breithaupt, and from certain patients. Daily Nitrogen Excreted through the Feces in Starvation. gram CRIN coo a gas 5 ROE Ee EP RIE Ae anand i 0.32 BIEL P@LY DE coc ese Ren ERC API Pe oUF ack cll ne 0.12 Pavent, (stenosis: of oesophagus).......2565:.--2..-.+-...- 0.46 Re erect te hs ire om ee RR Seed Seeks ois 0.22 Nea shh eni Cae wart acct hs os, eae eR a tucees 3 pa eee es. 0.17 JMTHEYPE ICIS Lier SiS lel at Anes ie UP ed 0.26 Dogs. Bidder and Schmidt, and Voit early observed that during starvation black pitch-like feces were obtained from dogs. The latter obtained daily 2 grams of feces (= 0.15 gram of nitrogen) from a dog of 30 kilos. The studies of Miiller offer further illus- trative data. Daily Feces Obtained from Starving Dogs (Miiller, 1884). BODY WEIGHT] * arta FECAL NITROGEN Sere uTO none. WEIGHT grams per cent yrams gram 43 4.8 5.0 0.24 0.0056 30 2.4 8.0 0.19 0.0063 30 1.4 8.0 0.11 0.0037 23 2astey em 5.3 0.15 0.0065 7 OR7 7.6 0.05 0.0071 MCTARC 06:3... og Re eran Onn ie 0.0058 6 Taken from Schmidt and Strasburger (see bibliography), p. 115. 12 The Utilization of Proteins Benedict has pointed out that the amount of feces formed during starvation is probably much smaller than is indicated by earlier studies. Fasting feces are in great part derived from retained fecal matter, resulting from the food immediately preceding the period of inanition. This is owing to diminished peristalsis con- sequent upon the withdrawal of food. With Nitrogen-Free Diets. The accompanying table embodies results obtained by Rieder. Nitrogen Eliminated through Feces on Nitrogen-free Diet (Rieder). 1 a | FECES SUBJECT WEIGHT FECAL NITROGEN FOOD DRY grams per cent gram Man..: .... | 13:4 4.08 0.54 485 grams cakes of starch, sugar and fat. Manse Gy! 5.69 0.87 159 grams eakes of starch, sugar and fat. Mans... 7 13.4 5.85 0.78 147 grams cakes of starch, sugar | and fat. Dog es es370 3.67 0.11 | 70 grams starch. | 140 grams starch. Doge 620 3.85 0.22 Rubner (1879) reported similar results. Tsuboi fed dogs for periods of six to nine days on cakes made of starch, fat and sugar, and obtained data, which are in accord with the above. Nitrogen Eliminated through Feces on Nitrogen-free Diet (Tsubor). be FECAL NITROGEN i tah WE | Starch | Sugar Fat ; grams [ per cence gram | grams grams grams 2.6 5.1 0.14 | 0 0 0 5.8 4.1 | 0.24 | 70 12 50 12.9 4.4 | 0.57 | 200 25 80 There can of course be no question as to the source of fecal nitrogen in the above experiments. Lafayette B. Mendel and Morris S. Fine 13 With Meat Diets. The most interesting work bearing upon the nitrogen of the feces obtained with meat diets and the relation of the amount of meat ingested to the nitrogen thus eliminated was contributed by Miiller. Inflwence of Meat Diet on Fecal Nitrogen in Dogs (30-35 kilos) (Miiller). ] FECES WEIGHT MEAT aE FECAL NITROGEN NITROGEN UTILIZED grams grams per cent gram per cent 0 2.0 7.96 0:15 500 5.1 6.50 0.30 98.2 i000 ype 6.50 0.55 98.4 1500 10.2 6.50 0.67 98.7 1800 _ 10.3 6.50 | 0.70 98.9 2000 ile i 6.50 | 0.80 98.8 2500 15.4 6.50 1.00 98.8 | It is clear from this summary that the nitrogen of the feces does not increase in proportion to the amount of meat eaten. That the fecal nitrogen incident to a meat diet is essentially of metabolic origin,’ is very convincingly brought out by Fritz Voit. After a loop of the intestine had been isolated, a dog was fed with meat. It was found that the contents of the loop resem- bled the feces in appearance and nitrogen content. Moreover, when calculated to unit surface the absolute amount of dry sub- stance in the loop compared favorably with that of the feces. Equally significant is the recent study of Mosenthal, who also worked with isolated intestinal loops. This author estimated that the succus entericus contained nitrogen equivalent to 35 per cent of the nitrogen ingested, and 300 to 400 per cent of the nitrogen of the feces. Nitrogen equivalent to at least 25 per cent of that of the intake must therefore have been reabsorbed. From the foregoing there can be no doubt that the feces resulting from a thoroughly digestible food such as meat are almost solely of ‘‘metabolic origin.” Prausnitz has attempted to give this more widespread application. 7 By an ingenious microscopical method, Kermauner (see bibliography) showed that in man but one per cent or less of the ingested meat reappeared in the feces. 14 The Utilization of Proteins Composition of Feces on Various Diets (Prausniiz). —— r finn Stor Leet : | | FECES—DRY NUMBER PERSON MAIN FOOD eT Nie | etka is en | Nitrogen a Ash Seo ee se | | | per cent per cent per cent 1 H. Rice 8.83 | 12.4 | 15.4 2 H. Meat |. 8.75 fw 16.0 14.7 3 M. Rice 8.37 | 13/2) ||) eo 4 M. Meat | 9.16 16:07 Vy t2e2 5 W.P. | Rice 8.59 | 15.9 12.6 6 Vee Meat | 8.485 | 07 25 | eave th eba Rice 8.25 | eul4eo 8 Ea. Meat 8.16 | \ See 9 EPL Rice 8.70 | | 16.1 10 ifijere Meat 9.05 | > 18 11 aC: Rice 8.78 | 18.6.1) Sa2ee | (vegetarian) | Average 8.65 | 16:4 | ts 12 M. | Mixed diet 6.76 lense 12.0 13 | H. Mixed diet | 6.63 | 25.8 | 14.9 14 | H. | Mixed diet | 6.07 | 0c. i ieee | | ) The excreta from the above diets (Nos. 1 to 11) contained no starch, and the composition of the feces did not alter materially as the character of the food changed. Such feces Prausnitz con- sidered “normal feces.”” When, however, the food contains mate- rial of a less digestible nature, the composition may change. Where this indigestible material is cellulose the nitrogen content of the feces is lowered (Nos. 12 to 14); if a nitrogenous substance, the nitrogen content might be expected to be raised. Schierbeck recognizes three types of individuals: (1) those that consistently have feces with low nitrogen concentration (about 4 per cent) whatever the nature of the diet may be; (2) those that under these conditions have feces of high nitrogen percentage (6-7 per cent); and (3) those in whom coarse food yields feces of low nitrogen percentage, and readily absorbed material produces feces with nitrogen concentration as high as 8 per cent. We are inclined to agree with Benedict that during starvation the formation of feces is reduced to a practically negligible quan- tity. When a material such as meat is eaten whose protein utilization, estimated according to the usual custom, is at least 95 Lafayette B. Mendel and Morris S. Fine [5 per cent, the resulting feces are for the most part of metabolic origin. It has been shown that the feces from such a diet represent a very small portion of the originally secreted intestinal juice, the latter having been absorbed in great part before reaching therectum. Ob- viously the degree to which this secretion is reabsorbed will depend upon the rate of peristalsis, which in turn ts influenced by the mass and character of material in the intestine. Hence, if to a meat diet an indigestible or less digestible material is added, thus stimulating peristalsis, more metabolic products’ must escape reabsorption If we deal with a non-nitrogenous material, e.g., agar, bone ash or crude fiber, the percentage nitrogen of the feces will of course be lower. If the comparatively indigestible material is highly nitrogenous ‘ike protein, the nitrogen concentration will be higher; and if both types of indigestible materials are present, the percentage of nitrogen may be indistinguishable from that found in. meat-feces. Illustrative data follow. EXPERIMENTAL PART. The conduct of these experiments did not differ essentially from that of trials described in previous papers. The quantities of ‘meat and indigestible non-nitrogenous materials can be learned from the tables; the amounts of water, sugar and lard approxi- mated those employed in previous experiments. The influence of indigestible non-nitrogenous materials wpon the nitrogen statistics of the feces is illustrated in Tables 4 and 5. In Table 4 the contrast is made between feces resulting from meat and feces accruing from an identical diet to which 3 grams of agar plus 7 grams of bone ash had been added daily. In Table 5 a similar contrast is drawn between meat- and meat-crude-fiber feces. The data are briefly summarized in Table 6. The increase in absolute fecal nitrogen due to the addition of indigestible materials to the diet is manifest, although the nitrogen intake did not vary. Thus the fecal nitrogen of (1) is increased 60 per cent by the addition of 10 grams of indigestible non-nitrogenous sub- stances, and that of (3) is augmented 133, 133, and 192 per cent? 8 Possibly also food residues and products of digestion. » Too great a quantitative significance should not be placed upon these figures, as an accurate isolation of pure meat-feces is almost impossible even when special precautions are taken. 16 TABLE 4. The Utilization of Proteins Influence of Agar + Bone Ash upon the Feces Resulting from a Meat Diet. Daily Averages. FECES z 7 ne Q a o& 2 D ° NATURE OF INGEST: 2p § S on 2 9 | x z NATURE OF INGESTA ay & & eet Zz Aa} A Cy ie s 5 Pao Sis) eats yt Z i grams | gram _ per cent | per cent Yl So alex f Meat, sugar, lard = 4.6 \| 4.5 | 0.22| 4.95] 95.2 2|5 5 | xxviii | to 4.9 gm. nitrogen {| 3.4 | 0.16 | 4.62] 96.6 3/5/5 |i Agar 3 gm.) | 13-2 | 0.29 | 2.22 | 94.0 415] 4| iii Reahooeee Rone Ach 14.5 | 0.36 | 2.48 | 92.7 5|5| 41] iv en 15.5 | 0.40 | 2.60} 91.8 6/515 | viii J] 15.0 | 0.35 | 2.32 | 92.7 Average ofiland®@.........| 4.0 | 0.19 | 4.78 | 95.9 Average ofS to. jasenn alla WOnd7 oe 40elmngee : : : cee \Meat, ete., as for Dox 5, A ; : ; Meat, etc., with indiges- a a, tible materials, as for 11|}6|4{ vi Dog 5 12) 6 15) wa 3 Average of Vand 8......... Average Of P1012. ~ 2.62. - |. 13 |7| 4] xx J| 3.2] 0.19 | 5.93] 95.8 Meat d for Dog 5 14| 715 | xxviii } Bo eee eer Salk eSe 8s i015 ancora 15 | 7 = 1 Meat, etc., with indiges- 12.5 | 0.28 | 2.26 | 91.4 1a iia pea fea ible quatemnlactenier 12.8 | 0.23 | 1.81 | 93.0 1d a ae ah Dogs ; 12.7 | 0:26 |/2.08 |gamt 18 7/5 vi [| 12.8 | 0.24 | 1.86 | 92.6 | { | | } Average of 13and 14.......| 3.6 | 0.17 | 4.88 |, 96.8 | 1.99 92.8 Average of 15to 18........ | 12.7 for kh Lafayette B. Mendel and Morris S. Fine TABLE 5. Influence of Crude Fiber upon the Feces Resulting from a Meat Diet. Daily Averages. NATURE OF INGESTA NUMBER PERIOD Meat, etc., = 3.3 gm. nitrogen 7% XViil As above + 6 gm. crude fiber* 3|6| 4 | xvii Meat, etc., asfor Dog5d 4|6|4| xix Meat, etc., + 6 gm. crude fiber as for Dog 5 | 5|7|4| xvi Meat, etc., as for Dog 5 6|7| 4] xviii Meat, etc., +- 6 gm. crude fiber as for Dog 5 AVERAGE Ofila ONOe rn one INTO REYES Uplda5 cn eons 6! i grams 10.0 FECES HI Weight Air Dry Nitrogen gram 0:4T| 0.02 10.1 | 0.30 1 9.10513 0.34 | 1.5 5. 2. Nitrogen per cent 97 .03 32 | L7 NITROGEN UTILIZATION per cent 99. 4t 90.5 96.0 —— Meat, etc., + 2 gm. agar + 5 gm. bone ash tO | 0.28) || 2.59) | 918 The same HS OL 2Gue2n27 ce 92e 8 The same + 6 gm. filter; paper. Nike (AOR 271 leno) Oi Gr Meat, etc., + 2 gm. agar + 5 gm. bone ash LOESE Osos) axon |" Oo-S The same 10.0 | 0.31 | 3.07 | 90.6 The same +°6 gm. filter paper | 18.0 | 0.40 | 2.23 | 87.7 Meat, etc., + 2 gm. agar| +5 gm. bone ash 8.5 | 0.23 | 2.74 | 93.3 The same 9.5 | 0.26 | 2.70 | 92.2 The same + 6 gm. filter paper 18.0 | 0.38 14 | 88.3 Average of 7, 8, 10, 11, 13, 14, 10.2 | 0.28 | 2.79 | 91.6 Average of 9, 12,15......... 17.9 | 0.35 | 1.97 | 89.2 is * Newspaper (0.1 per cent nitrogen) was thoroughly disintegrated under water. + These values are abnormally low owing to poor separation of feces of successive periods. Thvy are not included in the averages. t Omitted from the averages. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI NO. aly 18 The Utilization of Proteins TABLE 6. The Influence of Indigestible Non-Nitrogenous Materials upon the Nitrogen Statistics of Meat-Feces. (Summary of Tables 4and5). Daily Averages. oe ee EEE : gies : ra , nee FECES g P be z Pr: s of , bz B eer a ro] me ~ HpAo p z ap SAS) a m a z 5 3 7 z A PNAS se Nitrogen | Nitrogen 5 tl pete a e | ge8s le z % z Z & et z grams if grams grams gram per cent per cent 1 6* 4.6 0 3.8 0.2004) abr 95.7 2 IPRA 4.6 10 NBS 0.32 2.3 92.7 3 | 3t 3.3 0 a7, 0.12 6.5 96.4 4 3t 3.3 6 9.5 0.28 3.0 91.0 5 6§ one a 10.2 0.28 2.8 91.6 6 3|| (2140) 13 17.9 0.35 2.0 89.2 =e * Cf. Table 4, Nos. 1, 2, 7, 8, 13, 14. ** Cf. Table 4, Nos. 3 to 6, 9 to 12, 15 to 18. ¢t Cf. Table 5, Nos. 1, 3, 5. { Cf. Table 5, Nos. 2, 4, 6. § Cf. Table 5, Nos. 7, 8, 10, 11, 13, 14. || Cf. Table 5, Nos. 9, 12, 15. by the addition to the diet of 6, 7, and 13 grams respectively of such materials. The low nitrogen concentration of the feces of (2), (4), (5), and (6) is characteristic of diets of thoroughly util- ized materials including much indigestible non-nitrogenous matter. The nitrogen concentration, however, is not sufficiently diminished to compensate for the increased volume of feces—hence the above increment in absolute fecal nitrogen and the correspondingly low- ered coefficients of digestion. Illustrations of the influence of poorly utilized highly nitrogenous matter wpon the nitrogen statistics of the feces are especially con- spicuous in certain data already published!? and which are repro- duced in Table 7. The nitrogen concentration of the phaseolin- feces is 6.1 per cent against 2.3 per cent for that of feces resulting from a meat diet fed under conditions identical with those attend- ing the phaseolin feeding. A similar though less striking example 10 Mendel and Fine: This Journal, x, p. 433, 1912: Table 24 (phaseolin) ; Table 25 (pea globulin); Tables 10-11 (soy bean). Lafayette B. Mendel and Morris S. Fine 19 is offered in the case of the pea globulin experiment. Nos. 5 to 8 of this table disclose how closely the nitrogen concentration of feces accruing from diets containing both poorly utilized highly mitrogenous materials and indigestible non-nitrogenous materials may simulate the corresponding value for meat feces. = TABLE 7. The Influence of Poorly Utilized Highly Nitrogenous Materials upon the Nitrogen Statistics of the Feces. Daily Averages. I fe | FECES | 2 y fe NATURE OF INGESTA | nea a | ‘onmtiza- a (Air Dry) Nitrogen | Nitrogen | ; a a; | grams grams grams per cent | per cent 1 | Phaseolin 5.2 20.0 1721 Gi - 76.9 2 | Meat* 5.2 10:9 4| 0.26 2.3 | 95.0 3 | Pea Globulin 4.8 15:6" |) 0.56 | 3:6 88.3 4 Meatt 4.8 14.7 0.37 255 92.4 5 Soy Bean 4.6 15.9 0.57 3.6 87.6 6 Meat 46 3.8 0.15 3.8 96.9 a Soy Bean She 14.0 | 0.65 4.7 80.2 8 Moat B45 0.4f | 0.02t bee 3083 99.4 * Average of fore and after periods. t Average of fore and after periods. t See second note to Table 5, this paper. Obviously the fecal nitrogen concentration by itself is not a safe criterion! by which to judge the digestibility of a material. The nitrogen of the voluminous meat-cellulose-feces may be almost entirely of metabolic origin and yet be present in relatively low concentration; whereas a soy bean diet may yield feces composed in great part of highly nitrogenous undigested food residues, the nitrogen concentration,” however, being comparable to that of meat-feces. 11 Tsuboi (see bibliography), p. 80, likewise believes that one should be conservative in drawing conclusions from this one factor. 12 Tsuboi (loc. cit., p. 81), has made a similar statement. He points out that in Rubner’s studies, peas were poorly utilized (72 per cent) and yet the nitrogen concentration of the feces was 7.3 per cent, thus according closely with that of 6.9 per cent for the nitrogen concentration of feces from meat which was 97 per cent utilized. 20 The Utilization of Proteins Benedict has called attention to the difficulty encountered in satisfactorily isolating feces accruing from a particular diet, owing to the lagging behind of fecal material from the preceding diet. Our own experience testifies to this difficulty. It was especially pronounced where the experimental, preceding and succeeding TABLE 8. _ Influence of Thorough Evacuation upon Nitrogen Statistics of Feces Daily Averages. FECES z me As A ° F 5 = = gg s 4 NATURE OF INGESTA © 3 os 5) S oo Ba z a 2 2 a fe Boi Ze ———— = grams | gram | per cent) per cent Meat, etc., + 10gm. Agar ( = 4.6 gm. Nitrogen 15.7 | 0.52 | 3.33 | 88.6 FN Gi \\ a3 | Sabi Meat, ete., + 10 gm. Bone Ash 13.8 | 0.26 | 1.90 | 94.3 Meat, etc., +10 gm. Agar | 13.5 | 0.40 | 2.95 | 91.4 Mein, ete., +10 gm. Agar | 15.5 | 0.53 | 3.42) 88.5 6) PSs exxv. Meat, etc., + 10 gm. Bone} Ash Meat, etc., + 10 gm. Agar Meat, etc., + 10 gm. Agar Meat, ete., + 10 gm. Bone Ash Meat, etc., + 10 gm. Agar PAW ERAG CO; tea ete See VAVENAQC Offe ON Gt ee eee AVERAGE OS AON Oe ee Lafayette B. Mendel and Morris S. Fine 21 diets were all composed of thoroughly digested materials and the resulting feces were not adequate stimuli to peristalsis. This difficulty was obviated in a measure when the experimental period was preceded and succeeded by a 2-3 day period of a meat diet including 10 grams of bone ash daily. This lag and the effect of previous thorough evacuation upon util- ization is illustrated in Table 8. The first period for each dog (Nos. 1, 4, 7) was preceded by a period of wheat gluten, which is very well utilized.8 After thorough evacuation, it is clear (Nos 3, 6, 9) that the apparent utilization is considerably improved. Estimation of “Metabolic” Products in the Feces. Investigators have sought a method whereby the prominent part taken by alimentary waste products in the formation of feces could be determined with some degree of accuracy. This would enable one to estimate what proportion of the feces is due to undi- gested food residues. Processes have been proposed which involve treating the feces with pepsin-HC] or dilute alkali. Data thus obtained are of doubtful value. Equally unsatisfactory are those procedures which involve subtracting from the experimental feces the equivalent of fecal material obtained during starvation or on a thoroughly digested non-nitrogenous diet. The plan generally followed in the present work, namely the comparison of experimen- tal feces with feces obtained from a control meat diet is likewise not always free from objection. None of the above methods take into account the influence of undigested masses upon the degree of reabsorption of the intestinal juice. We propose the following plan“ which seems to avoid most of the above shortcomings: 1. Determine the volume and nitrogen of feces resulting from the material under investigation. 2. Determine the fecal nitrogen resulting from a nitrogen-free diet to which has been added an amount of indigestible non- 13 Cf. Mendel and Fine: This Journal, x, p. 324, 1911. 14 Tsuboi has applied a similar principle to certain results reported by Rubner. The nitrogen eliminated on a starch diet was subtracted from that excreted in feces of comparable volume resulting from diets of wheat bread and maccaroni. The food nitrogen actually escaping utilization could thus be computed. 22 The Utilization of Proteins nitrogenous matter’ that will yield approximately the same vol- ume of feces as was obtained in (1). 3. Subtract the fecal nitrogen of (2) from that of (1). This excess of nitrogen is presumably due to undigested or unabsorbed nitrogenous matter of the food material. An experiment with a nitrogen-free diet including indigestible non-nitrogenous matter follows: A 6 kilo bitch was fed for four days on a mixture of 35 grams of sugar, 45 grams of lard, 200 grams of water and 10 grams of agar. On this diet 13.2 grams of feces with a nitrogen concentration of 2.44 per cent were obtained daily. There were thus eliminated through the feces 0.32 gram of nitrogen daily, which was obviously of metabolic origin. This result makes it prob- able that the feces from meat diets, containing similar amounts of indiges- tible non-nitrogenous matter, (see Table 6) are likewise made up entirely of alimentary waste—proof in itself that meat nitrogen is 100 per cent utilized. From the single experiment above reported and from Table 6, No. 6, we may conclude that a thoroughly digested material may yield 13.2 to 17.9 grams of feces and yet the nitrogen (0.32-0.35 grams) thus eliminated will be of ‘“‘metabolic” origin. Hence in feces of comparable volumes" al] nitrogen in excess of 0.32-0.35 gram may be attributed to the nitregen of the food. This prin- ciple is applied in Table 9. ‘“Utilization,’”’ as the term is employed in the last column of this table, exactly expresses our meaning. The actual utilization of soy bean nitrogen!’ is 90.3-92.8 per cent and that for the crude bean protein is 91.8 per cent. If anything the latter value is low, as 24.6 grams of meat-feces would probably contain more than 0.35 gram of nitrogen. 15 The choice of indigestible adjuvant is a matter of some moment, as these materiais may vary in their ability to stimulate peristalsis. 16 This of course applies only for dogs of approximately the same weight (5 to 7 kilos.) as those in these experiments. 17 Soy bean is reported (Wolff-Lehmann: Landw. Fiitterungslehre, cited by Schulze und Castoro: Zeitschr. f. physiol. Chem., xli, p. 455, 1904) as having 10 per cent of its nitrogen present as non-protein. The latter may be more thoroughly utilized than the protein constituents, and thus the utili- zation calculated for the total nitrogen intake would be greater than is actually the case for the soy bean protein. Excepting the soy bean and cotton-seed flours, the preparation of the materials employed in this series of studies renders contamination with nitrogenous non-protein matter unlikely. Lafayette B. Mendel and Morris S. Fine 23 The Utilization of the Vegetable Proteins. About the thorough utilization of the proteins of wheat!* there is no question. The probability that those of barley!® and corn?° are equally available was pointed out in previous papers of this series. With regard to the legume proteins?! we must for the present conclude that the presence of indigestible non-nitrogenous materials cannot entirely account for their low coefficients of diges- TABLE 9. Utilization as Estimated from the Portion of Fecal Nitrogen Derived from Food Residues. Daily Averages. x FECES on NITROGEN UTILIZATION Ba all> 628 Zz > 29 a | < NATURE OF INGESTA a x aR a Bo Ae Anes 28 fe) 4» ) 98 Ordinaril oe 5 a > : R 2 & 2 é z 8 Sarre Utilization Z a ~ a gram | grams gram gram per cent | per cent 1 | Nitrogen-free diet including 10 gm. | agar 0.0 | 38.2} 0.32] 0.00 | 3 | Meat diet* 3.3 | 17.9 | 0.35 | 0.00 89.2 100.0 6 | Soy beant 4.6 | 17.2} 0.68] 0.33 85.3 4) 92.8 3 | Soy beant 3.3 | 13.0 | 0.64] 0.32 81.1 90.3 1 | Crude bean protein§; 3.4 | 24.6 | 0.63) 0.28 81.5 91.8 * Cf. Table 6, No. 6 this paper. ¢ Cf. Mendel and Fine: This Journal, x, p. 433, 1912. Tables 5-10. ¢t Cf. Mendel and Fine: loc. cit. Tables 11 to 13. § Cf. Mendel and Fine: loc. cit. Table 21. tibility. These proteins appear to be less readily affected by the digestive processes than those of barley or corn. This resistance is even more pronounced in the case of the cotton-seed protein.” Nevertheless, future research with the isolated proteins nay modify our opinion with regard to these two last classes of materials. The lack of animal extractives in vegetable materials has at times been thought to be the cause of the apparently poor utili- zation of plant foods in comparison with those of animal origin. 18 Cf. Mendel and Fine: This Journal, x, p. 303, 1911. 19 Cf. Mendel and Fine: Jbid. x, p. 339, 1911. 20 Cf. Mendel and Fine: [bid., x, p. 345, 1911. 21 Cf. Mendel and Fine: Jbid. x, p. 433, 1912. 22 Cf. Mendel and Fine: Jbid. xi, p. 1, 1912. 24 The Utilization of Proteins The evidence bearing upon this seems to be inconclusive. Bischoff showed that meat extracts did not appreciably influence the diges- tibility of bread. Thompson came to the opposite conclusion. Effront noted that meat extracts exert a favorable influence upon the availability of vetegable diets, but this could not be confirmed by Wintgen. The fact that the proteins of wheat, and probably those of barley and corn also, are thoroughly utilized lends support to the view that the secretory influences of the extractive materials play a minor réle in the ultimate utilization. It was pointed out in an earlier paper that certain wheat preparations evoked intense nausea in man, and necessitated forced feeding in the dog experi- ments, but were, nevertheless, thoroughly digested. This would suggest that psychic secretion does not influence the ultimate utilization to any great extent.2? It would be interesting to study the relation of gastric secretion to ultimate utilization by means of a sequestered stomach. Studies on the digestibility of vegetable proteins in vitro are not lacking. Rothe*! found that in 24 out of 26 of his experiments, the coefficients of digestibility were above 90 per cent. The coefficients for the legumes averaged about 95 per cent. In these studies, 2 grams of material were acted upon by 250 cc. of concentrated gas- tric juice for forty-eight hours. As yet, it is uncertain to what extent in vitro experiments of this kind can be held comparable to studies on the utilization of proteins from the alimentary tract. It may be noted that artificial digests are not contaminated with “metabolic” products, and this may explain the high coefficients of digestibility obtained in such trials compared with those result- ing from experiments in vivo. Studies on animals where this factor has been taken into account yield results not very unlike those obtained in artificial digestion experiments (see Table 9). A study of the literature on the availability of vegetable materials reveals an interesting situation. In instances where the protein has been very poorly utilized, the carbohydrate, on the contrary, has rarely been less than 95 per cent digested. This is illustrated in the accompanying table, containing data gathered at random. 23 Cf. Schmidt und Strasburger: ‘“Die Fazes des Menschen,”’ p. 17, Berlin 1910. Also Osborne and Mendel: Carnegie Institution of Washington, Publication No. 156, p. 5, 1911. 2 Rothe: Zeitschr. f. physiol. Chem., li, p. 185, 1907, (contains the literature). Lafayette B. Mendel and Morris S. Fine 25 Table Illustrating the Simultaneous Poor Utilization of Protein and Good Utilization of Carbohydrate. COEFFICIENTS OF DIGESTIBILITY NATURE OF FOOD i. a he a a ae | aay Protein Carbohydrate es ee : | per cent [ per cent Migneyspedtne ss. SITE. 77 4 Wait”® Wihtterbe@ansS-n sss. 45..2 466s enk sess as hee 78 96 Wait Ah DELS coe hl Sh ar rr i an 70 87 Wait Rice, barley, and vegetables........ 76 97 Oshima?é Riceyand barleyse= eek. eo Nt 67 99 Oshima Cookedsnicemesee es a. 62 76 99 Oshima [DCU EE crag 2, ot rr na a aR 60 97 | Oshima It is possible that the starch has been completely utilized and the carbohydrate of the feces is in reality hemicellulose. BIBLIOGRAPHY. THE UTILIZATION OF MEAT POWDERS AND ALLIED MATERIALS. ABDERHALDEN und RUERL: Zeitschrift fiir physiologische Chemie, Ixix, p. 301, 1910. EvuincERr: Zeitschrift fiir Biologie, xxxiii, p. 190, 1896. Forster: [bid., ix, 1873. (cited by Atwater and Langworthy: U. S. Dept. of Agriculture, Office of Experiment Stations, Bull. 45, 1897. HILDEBRAND: Zeitschrift fiir physiologische Chemie, xviii, p. 180, 1894. ImaBucui: Jbid., lxiv, p. 1, 1910. Maas: Medizinische Klinik, No. 8, 1906. MUuuEr: Miinchener medizinische Wochenschrift, xlvii (2), p. 1769, 1900. Neumann: Jbid., xlv (1), p. 72, 1898 and xlvi (1), p. 42, 1899. NEUMEISTER: Deutsche medizinische Wochenschrift, xix, pp. 866 and 1169, 1893. Piautu: Zeitschrift fiir didtetische und physikalische Therapie, i, p.62, 1898. Pravusnitz: Zeitschrift fiir Biologie, xlii, p. 377, 1901. SaLkowskI: Biochemische Zeitschrift, xix, p. 83, 1909. ScHMILINSKy und KLEINE: Miinchener medizinische Wochenschrift, xlv (2), p. 995, 1898. Srrauss: Therapeutische Monaishefte, xii, p. 241, 1898. Voir: Zeitschrift fiir Biologie, xlv, p. 79, 1904. *% Wait: U. S. Dept. of Agriculture, Office of Experiment Stations, Bull. 187, p. 53, 1907. 26 Oshima: Jbid., Bull. 159, 1905. 26 The Utilization of Proteins ORIGIN OF FECAL NITKOGEN AND CONDITIONS INFLUENCING ITS EXCRETION. Benepict: ‘‘The Influence of Inanition on Metabolism,” p. 347, Carnegie Institution of Washington, 1907. BippER und Scumipt: ‘‘Die Verdawungsséfte und der Stoffwechsel,’’ 1852. BiscHorF: Zettschrift fiir Biologie, v, p. 454, 1869. Errront: Frinfter internationeller Kongress fiir angewandte Chemie, p. 97, 1904. (Cited by Wintgen.) HorrMann: (Cited by Voit: Sitzwngsberichte der bayerischen Akademie, ii, (4), 1869). KERMAUNER: Zeitschrift fiir Biologie, xxxv, p. 316, 1897. Lotrurop: American Journal of Physiology, xxiv, p. 297, 1909. MosEnTHAL: Journal of Experimental Medicine, xiii, p. 319, 1911. Mier: Zeitschrift fiir Biologie, xx, p. 326, 1884. (Cited by Tsuboi.) Prausnitz: Ibid., xxxv, p. 335, 1897. RiepeEr: Ibid., xx, p. 378, 1884. (Cited by Tsuboi.) RuBNER: Ibid., xv, p. 115, 1879; also Jbid., xix, p. 45, 1883. Scumipt und STRASBURGER: ‘‘ Die Faeces des Menschen,’’ p. 115, Berlin, 1903. ScHIERBECK: Archiv fiir Hygiene, li, p. 62, 1904. Tuompson: Zentralblatt fiir Physiologie, No. 17, p. 814, 1910. TsusBor: Zeitschrift fiir Biologie, xxxv, p. 68, 1897. Voir (C.): Ibid., ii, p. 308, 1866. (Cited by Tsuboi.) Voir (F.): Ibid., xxix, p. 325, 1893. (Cited by Tsuboi.) WickeE: Archiv fiir Hygiene, xl, p. 349, 1890. WINTGEN: Ver6ffentlichungen aus d. Gebiete des Militdrsanitétswesens, xxix, p. 56, 1906. CHEMISTRY OF THE DOG’S SPLEEN.! By HARRY J. CORPER. (From the Department of Pathology, University of Chicago.)? (Received for publication, December 3, 1911.) In conjunction with experimental work on the histological and chemical changes in the spleen during autolysis, results of which are to be published later, it was found desirable to analyze spleens and also to study their purine enzymes. The results of these studies are given in this paper. The dog’s spleen has never been analyzed in a complete manner chemically, but has frequently been used for the study of its enzymes and individual elements. The spleen of man and other animals has, however, been studied more fully. The water con- tent of the spleen is given as about 78 per cent,? 70 per cent to 77 per cent,* Schulz® finding 79.5 per cent. in human spleen. 12.5 per cent of the spleen is blood content. The fat and lipoid con- tent, which is probably variable, is given as about 11 per cent to 14 per cent of the dry weight. The phosphorus content of cow’s spleen varies with age, the spleen of fetuses being richer in phosphorus (up to 2.4 per cent of the dry weight) while adult spleens have less phosphorus (1.3 to 1.4 per cent of the dry weight). The sulphur content of the cow’s spleen is fairly constant in the different periods of life,” ranging from 1.8 per cent to 2.2 per cent 1 This work has been aided by a grant from the Rockefeller Institute for Medical Research. 2 A portion of this work was done in the Physiological Laboratory of the University of Illinois. 3 Oppenheimer’s Handbuch der Biochemie, ii, 2, p. 172, 1909. 48. Frankel: Descriptive Biochemie, Wiesbaden, 1907. 5 Schulz: Pfliiger’s Archiv, liv, pp. 555-573, 1893. 6 Oppenheimer’s Handbuch der Biochemie, ii, 2. p. 172, 1909. 7 Zeitschr. f. Biol., xxxi, pp. 400-413, 1895. 27 28 Chemistry of the Dog’s Spleen of the dry weight. Schulz® examined the spleen of a man thirty- nine years old and found a water content of 79.5 per cent and a sulphur content of 0.78 per cent of the dry weight. The iron content is also variable, differing markedly in differ- ent periods of life,°and the iron is held mostly in organic molecules.!® The proteins of the spleen have been very little studied, investi- gation along this line being mainly confined to the study of the nucleoproteins and purines. Mandel and Levene" hydrolyzed the spleen nucleoprotein and obtained glutamic acid, glycocoll, alanine, aspartic acid, proline, phenylalanine, tyrosine, lysine, arginine, histidine, adenine, guanine, cytosine and thymine. The ammonia content of the organs of the body varies, being greater than normal during hunger, as determined by the Folin method. Grafe” examined one spleen, that of a horse and found 9.5 mg. of ammonia per 100 grams of spleen. The fresh spleen contains neither albumoses nor peptones. Levene® found as end products of the autolysis of the spleen, alanine, leucine, amino- valerianie acid, aminobutyric acid, and a-pyrrolidine carboxylic acid, phenylalanine, aspartic acid and tyrosine. Adenine and guanine were replaced by hypoxanthine and xanthine, and thymine, cytosine and uracil, which are present in spleen nucleic acid, were found as thymine and uracil after autolysis of the spleen. Jones obtained from autolyzed pig spleen guanine and hypoxanthine, but no adenine, while in place of the thymine and cytosine which is found after hydrolysis, he obtained uracil. Later Jones found that these differences between the purines obtained after auto- lysis in different experiments were due to the fact that spleens from different animals were being studied; and he determined the presence of guanase in large amounts in the cow’s spleen, although this ferment was entirely absent from the pig’s spleen. 8 Schulz: loc. cit. 2 Oppenheimer’s Handbuch der Biochemie, ii, 2, p. 172, 1909. 10 Capezuolli: Zeztschr. f. physiol. Chem., |x, pp. 10-14, 1909. Burow: Biochem. Zeitschr., xxv, p. 165, 1910. 11 Mandel and Levene: This Journal, ili, p. xxiii, 1907-08. 2 Zeitschr. f. physiol. Chem., xlviii, p. 300, 1906. 13 Levene: Amer. Journ. of Physiol., xi, p. 437, 1904; xii, p. 275, 1904-05. ‘4 Jones: Zettschr. f. physiol. Chem., xlii, p. 35, 1904. ‘Ss Jones: Ibid., xlv, p. 84, 1905. ‘Harry J. Corper 29 Schumm! autolyzed spleens from cases of myelogenous splenic leucemia and obtained free guanine, xanthine, and hypoxanthine. Schittenhelm,!’ as a result of his investigations, states that the spleen of the cow contains a hydrolytic ferment which changes adenine of hypoxanthine, and guanine to xanthine, and an oxydase which forms xanthine from hypoxanthine but does not destroy uric acid. He analyzed one sterile dog spleen, which had been autolyzed in the ice chest and then kept in alcohol for one year; it yielded on hydrolysis xanthine and hypoxanthine but no adenine nor guanine. Wells and Corper!® have previously made note of the fact to be reported in this paper, that the dog spleen contains no uricolytic ferment, and observed that the human spleen con- tains no xanthine oxidase. Batelli and Stern,!® using their method of determining uricolysis by gaseous exchange, were also unable to demonstrate uricase in the dog spleen. Jones and Austrian*° found that normal dog spleen contained guanase, adenase and xanthineoxydase and was to this extent similar to cow spleen. Burian and Schur! obtamed 0.16 gram of purine N from 100 grams (moist weight) of calf spleen, which was divided into 0.046 gram of free purine nitrogen and 0.101 gram of combined purine base nitrogen. Kossel” found in the horse spleen 0.175 per cent purine nitrogen. Jones and Winternitz** observed that upon autolysis of the swine spleen there was a conversion of hypoxan- thine into xanthine in the absence of air. Pohl** states that the normal spleen of starving dogs contains no allantoin, but that after autolysis allantoin appears. He doesnot, however, give his experiments with autolyzed spleen. This does not agree well with the absence of uricase noted by other authors. 16 Schumm: Hofmeister’s Beitrdge, vii, p. 175, 1905. 17 Schittenheim: Zeitschr. f. physiol. Chem., xlv, p. 84, 1905. 18 Wells and Corper: This Journal, vi, p. 321, 1909. 19 Batelli and Stern: Biochein. Zeitschr., xix, p. 219, 1909. 20 Jones and Austrian: Zeitschr. f. physiol. Chem., xlviii, p. 110, 1906. *t Burian and Schur: Pfliger’s Archiv, |xxx, p. 309, 1900. 2 Zertschr. f. physiol. Chem., vi, p. 422, 1882. 78 Jones and Winternitz: Jbid., xliv, p. 1, 1905. * Pohl: Arch. f. exp. Path. u. Pharm., xviii, p. 367, 1902. 30 Chemistry of the Dog’s Spleen EXPERIMENTAL DATA. Analysis of Dog’s Spleen. MerxHops.—The tissues were preserved for analysis in the ice chest in ten parts or more by weight of 95 per cent alcohol, samples for water content having been taken, and when ready for analysis the alcohol was filtered off to be mixed with the rest of the alcohol and ether extracts. The total ether extract was examined quanti- tatively for cholesterol by Ritter’s method® and for lecithin by the method suggested by Koch and Woods.%2 The combined residues after ether extraction were then pulverized and extracted in a shaking machine with N free water, containing alternately. traces of alkali (sodium carbonate) and acid (acetic), and after bringing to faint acidity the combined watery extracts were concentrated to 1 liter and filtered hot, thus constituting the water-soluble fraction. This fraction was analyzed for purine content by the method of Kriiger and Salomon,?’ and was also analyzed for a tannic acid precipitable fraction and a fraction not precipitable by tannic acid; and its phosphorus content was deter- mined by the Neumann method as described by Koch and Woods. The remaining residue, after removal of ether and water extracts, was analyzed for its total nitrogen content by the Kjeldahl method, and for iron, phosphorus and sulphur by the method described by Koch and Mann.?8 Purine nitrogen was also determined on this residue after hydrolyzing, by means of 5 per cent sulphuric acid, using the Kriiger and Salomon method. The Hausmann fractions were also determined on this tissue residue according to the modifi- cation described by Osborne and Harris.?9 NoRMAL SPLEEN A. Results of Analyses.—Three small spleens weighing 19.25, 31.5 and 21.5 grams, with a moisture content of 76.82 per cent, 76.50 per cent and 76.64 per cent respectively, making a total dry weight of 16.87 grams, were taken. Total ether soluble material weighed 2.609 grams, or 2° Zeitschr. f. physiol. Chem., xxxiv, p. 461, 1903. 26 Koch and Woods: This Journal, i, p. 203, 1906. 27 Hoppe-Seyler-Thierfelder: Handbuch d. physiol. u. pathol. chem. Ana- lyse, 8th edition, p. 188, 1909. 28 Koch and Mann: Archives of Neurol. and Psychiatry, iv. p, 20, 1909. 29 Osborne and Harris: Journ. of the Amer. Chem. Soc., xxv, pp. 323-325, 1903. Harry J. Corper 31 15.47 per cent. Absolute alcohol insoluble part of this amounted to 0.2783 gram, containing 0.00082 gram of phosphorus after Neumann oxidation. The cholesterol determination was lost in this analysis. The lecithin phosphorus found in 1 gram dry weight tissue was 0.00269 gram, figured as lecithin=0.0694 gram. Non-lecithin phosphorus in this fraction = 0.0005 gram in 1 gram dry tissue. ‘Water Soluble Fraction: Gran Nitrogen precipitable by tannic acid in 1 gram dry tissue=0.00567 Nitrogen not precipitable by tannic acid in 1 gram dry tissue = 0.00396 Water soluble N in 1 gram, total =0.00972 Water soluble phosphorus in 1 gram dry tissue =0.0052 Purines found in waiter soluble fraction, only a doubtful trace. : Tissue Residue (Insoluble) Fraction. One gram of tissue (after extraction of soluble constituents) yielded 0.0035 gram of iron, 0.0070 gram of sulphur, and about 0.0084 gram of phosphorus. One gram of dry, tissue residue (used 8.4 grams for analyses) yielded 0.00345 gram of purine nitrogen. Total nitrogen determination (using 0.28 gram tissue residue) yielded 0.1447 gram of nitrogen in one gram of tissue residue. NoRMALSPLEEN B. One large spleen with a moist weight of 134 grams, and a moisture content of 75.59 per cent, making a total dry weight of 32.71 grams was taken. Total ether-soluble material weighed 3.8155 grams or 11.65 per cent. Absolute alcohol insoluble part of this = 0.6861 gram, containing 0.0140 gram of phosphorus (after Neumann oxidation.) One gram dry tissue contained 0.015 cholesterol by the Ritter method (using one-half of the total ether extract for the analysis). The lecithin fractions were lost in this analysis. Water Soluble Fraction : Nitrogen precipitated by tannic acid in 1 gram dry tissue=0.00412 Nitrogen not precipitated by tannic acid in 1 gram dry tissue = 0.00393 Water soluble N in 1 gram dry tissue, total.............. = 0.00804 Water soluble phosphorus in 1 gram dry tissue.......... = 0.00346 Purines found in water soluble fraction = a doubtful trace. Tissue Residue (Insoluble) Fraction. One gram of tissue residue (freed from soluble constituents) yielded 0.00455 gram of iron, 0.00714 gram of sulphur, and 0.00459 gram of phosphorus. One gram of dry tissue residue (used duplicates of about 6 grams each) yielded 0.00292 gram of purine nitrogen. Total nitrogen determination (using about 0.25 gram tissue residue in duplicates) yielded 0.1639 gram of nitrogen in one gram of tissue residue. NorMAuL SPLEENC. Onelargespleen witha moist weight of 82 grams and a moisture content of 76.94 per cent making a total dry weight of 18.91 grams. Total ether soluble material weighed 2.859 grams or 15.11 per cent. 32 Chemistry of the Dog’s Spleen Absolute alcohol insoluble part of this = 0.2021 gram, containing 0.00339 gram of phosphorus (after Neumann oxidation). Cholesterol determination was far too low, due to loss occasioned by the Ritter method, which will be discussed in a future paper. The lecithin phosphorus found in 1 gram dry tissue = 0.00241 gram, figured as lecithin = 0.0622 gram. Non-lecithin phosphorus in this fraction = 0.00015 gram in 1 gram dry tissue. Water Soluble Fraction: Nitrogen precipitated by tannic acid in 1 gram dry tissue=0.0012 Nitrogen not precipitated by tannic acid in 1 gram dry tissue=0.00338 Total water soluble N in 1 gram dry tissue................ =0.00458 Water ois phosphorus in 1 gram dry tissue............. = 0.0027 Only doubéful trace of purines was found in the water soluble extractives. Tissue Residue (Insoluble) Fraction.—One gram of tissue residue yielded 0.0122 gram of iron, 0.00745 gram of sulphur, and 0.00489 gram of phosphorus. One gram of dry tissue residue (used about a 6 gram sample) yielded 0.00428 gram of purine nitrogen. Total nitrogen determination (using about 0.26 gram, tissue residue in duplicates) yielded 0.1629 gram of nitrogen in 1 gram of tissue residue. Hausmann Fractions on tissue residue C.—Duplicate analyses were obtained with the following result-, from 1 gram of dry t ssue residue: Amid N =0.01349 and 0.01329 gram N — mean =0.01339 Humus N =0.00938 and 0.00754 gram N — mean =0.00844 Diamino N- =0.03546 and 0.03244 gram N — mean =0.03395 Monamino N =0.1007 and 0.0969 gram N — mean =0.09880 Total N =0.15458 Purines and Purine Enzymes in Dog Spleen. I. Hyprotysis oF Dog SrpLEEN.—1091 grams of dog spleen (moist weight) were hydrolyzed by means of 5 per cent sulphuric acid and the purines isolated in the pure state and weighed, with the following results, figured on the basis of 1 gram moist weight of the original spleen tissue used: Guanine =0.00109 gram (weighed as such). Adenine =0.00062 gram (weighed as pierate). Hypoxanthine =0.00015 gram (weighed as silver nitrate combination). Xanthine =0.00004 gram (weighed as such). Total products =0.00190 gram. No uric acid was found, although tested for. j - A purine nitrogen figure was obtained (using one-eightieth of the total material) and yielded in the figures of 1 gram moist weight of the original spleen tissue, 0.00126 gram of nitrogen. Harry J. Corper 38 Nitrogen figure obtained from the total amount of isolated purines was 0.000899 gram per 1 gram moist splenic tissue. If. Avurouysis oF DoG SPLEEN IN THE ABSENCE OF AIR.—552 grams of dog spleen (moist weight) were autolyzed for about a month, part of the time at 37° C. and using toluene as antiseptic and to keep out air. The autolysate was examined quantitatively for liberated purines and gave the following re- sults, on the basis of 1 gram moist weight of original spleen tissue used: No guanine, adenine or uric acid was found. The major portion of the purine product was Xanthine =(.00168 (weighed as such, and free from uric acid). Hypoxanthine =0.00002 gram (weighed as silver nitrate salt). Total products =0.00171 gram. Calculated purine nitrogen in these products = 0.00063 gram N, indicat- ing that about half the total purine nitrogen was present in the form of free purines. III. Avronysis or DoG SPLEEN IN THE PRESENCE OF AIR.—617 grams of dog spleen (moist weight) were autolyzed for about a month, part of which time a current of air was passed through the autolyzing mixture, and the temperature frequently brought to 37° C., toluene being used as anti- septic. The autolysate was examined quantitatively for purines, with the following results, on the basis of 1 gram moist weight of original spleen tissue used: Neither guanine or adenine were found. The major portion of the purines was Uric acid = 0.001695 gram (weighed as such. Repurified from con- centrated H2SQx,). Xanthine = 0.000094 gram (weighed as such). Hypoxanthine =0.000004 gram (weighed as hypoxanthine silver a nitrate). Total products =0.001793 gram Calculated purine nitrogen in these products =0.0006 gram or about half the total purine nitrogen. IV. XANTHINE-OxiDAsSE oF DoaG SpLeEeN.—Fifty grams of finely ground fresh dog spleen, mixed with three volumes of toluene water, was allowed to stand at room temperature over night, and strained through cheese cloth the following morning. To this spleen extract thus obtained was added 0.1355 gram of xanthine in solution, and the mixture was placed in a bottle with sufficient toluene to prevent putrefaction, and connected with another bottle containing toluene and water, through which was drawn the air that then passed through the digestion mixture, and all the bottles were kept at about 40° C. for twenty-four hours. After autolysis 0.0969 gram of uric acid was recovered, which gave a positive murexide test, and upon repurification from H.SOs, pure uric acid was obtained, thus differ- entiating it from xanthine. THE JOURNAL OF BIOLOGICAL CHEMISTRY, XI, NO. I. 34 Chemistry of the Dog’s Spleen V. Uricase oF Doa SpLEEN. (Hxperimeni 1.)—Sixty-three grams of ground up dog spleen, and 50 grams of ground up dog’s liver as control (high uricolytic power of dog liver having been demonstrated) were each mixed with three volumes of saturated toluene water and strained through cheese cloth after standing over night at room temperature. To the spleen extract thus obtained was added a solution of 0.1735 gram of uric acid, and to the liver extract was added 0.1616 gram of uric acid. Both were then kept at about 40° C. with a supply of air passing through them, for 48 hours and they were then analyzed for uric acid. From the spleen extract was recovered 0.1919 gram of uric acid, while the liver extract did not contain a trace of uric acid. Experiment 2.—Eighty grams of ground up dog spleen were allowed to stand at room temperature over night, and was then strained through cheese cloth. The extract was divided into two equal portions: to No. I was added 0.1655 gram uric acid in solution, and to No. II, after boiling fifteen minutes was added 0.1962 gram uric acid in solution. Both were then kept at about 40° C. for twenty-four hours, during which time they re- ceived a supply of air bubbling through them and were then analyzed for uric acid content. From No. I was recovered 0.1687 gram uric acid (which wher recrystallized from H.SO, yielded 0.1638 gram) and from No. II which had been boiled, was recovered 0.1856 gram of uric acid. SUMMARY. 1. Average of analyses of three normal dogs’ spleens resulted as follows: A moisture content of about 75 to 77 per cent. A content of ether-soluble materials between 11.6 and 15.5 per cent of the dry weight, which was made up of about 1.5 per cent. cholesterol and between 6 and 7 per cent lecithin (leaving but 2 to 6.5 per cent neutral fats.) The total soluble nitrogen ranged between 0.45 per cent and 0.97 per cent of the dry weight, divided about equally between that precipitable and that not precipitable with tannic acid. A water-soluble phosphorus content of about 0.27 to 0.52 per cent. No purines were found in the water-soluble fraction, at least — not in sufficient quantity from the amounts of tissue used to be recognized as such. The insoluble part of the tissue contained about 0.26 to 0.98 per cent of dry weight as iron, 0.53 to 0.60 per cent of dry weight as sulphur, and about 0.39 per cent of dry weight as phosphorus, Harry J. Corper 35 with a purine nitrogen content of 0.24 to 0.35 per cent of the dry weight. The total nitrogen content of the insoluble part was about 11.0 to 13.4 per cent of the dry weight, which was distributed as follows: Amid N, 8.60 per cent; Humus N, 5.76 per cent; Diam- ino N, 21.87 per cent; and Monamino N, 63.71 per cent. In contrast to normal liver, examined by Wells*® by the same method, the insoluble residue of the spleen contained a slightly larger percentage of its dry weight as iron, about the same per- centage of sulphur and a greater percentage of phosphorus. The Hausmann nitrogen fractions were also slightly different; more amid nitrogen, about the same or a little more humus nitrogen, less diamino nitrogen and about the same percentage of monamino nitrogen. 2. The purines obtained from 1 kilo (moist weight) of dog spleen after hydrolysis were: Guanine, 1.09 gram, adenine, 0.62 gram; hypoxanthine, 0.15 gram, and xanthine 0.04 gram. Analyzed purine nitrogen content in one kilo moist weight = 1.62 grams. . 3. The purines obtained from the autolysate from 1 kilo (moist weight) dog spleen, after autolysis in the absence of air, was 1.69 grams of xanthine and 0.017 gram of hypoxanthine. + 4. The purines obtained from the autolysate from 1 kilo (moist weight) of dog spleen, upon autolysis in the presence of air, were 1.69 grams of uric acid, 0.09 gram of xanthine and 0.004 gram of hypoxanthine. 5. Of the purine enzymes, evidence was obtained of the presence of xanthine oxidase, adenase and guanase, while uricase was lack- ing. The conversion of hypoxanthine to xanthine during com- parative anaerobic conditions indicates the presence of an oxidiz- ing enzyme with this particular function; whether it be the xan- thine oxidase itself, or a special hypoxanthine oxidase. 39 Wells: This Journal, v, pp. 141-142, 1908. j i ee ee 7A Ge : ‘ > 52 ; ee om! - ; i : . = | é ' i ee) = C * ; re : 4 7 | ’ * aa # us : Qe . at a Um) J j ©. > | : : } + A ty . = + A . 3 ore LD FAS i c= es . ie ; et Bi “2 5 hfe ra a ee P i < = Taek: BS My at rs - ery cars 4 spies 4 ~ , r ‘ é grt AD aS DS ™* - a ae ibis a ay , es tats yas se 5 we i 4T ri pet ray ERRORS IN THE QUANTITATIVE DETERMINATION OF CHOLESTEROL BY RITTER’S METHOD: THE INFLU- ENCE OF AUTOLYSIS UPON CHOLESTEROL.! By HARRY J. CORPER. (From the Pathological Laboratory of the University of Chicago.)? (Received for publication, December 3, 1911.) While carrying out analyses of normal and autolyzed spleens reported in previous papers’ the Ritter method‘ for determining cholesterol quantitatively was found inadequate. The results were so variable for the amounts used that it was thought advisable to investigate the steps in the method in order to find the source of error, and also to find out whether the method actually could be used for the quantitative determination of cholesterol in tissues. At least one other investigator, Helen Baldwin,> has had difficulty in the quantitative determination of cholesterol by this method. The quantitative methods for cholesterol determination suggested up to 1908 (a discussion of these methods is given by Glikin®) practically all depend either upon the saponification of the fat to be examined, the cholesterol being recovered as such, or upon the esterification of the cholesterol and the determination of the iodine figure or saponification number. More recently Windaus’ has suggested a new method for the determination of cholestero] and cholesterol esters, precipitating 1 This work has been aided by a grant from the Rockefeller Institute for Medical Research. 2 A portion of this work was done in the Physiological Laboratory of the University of Illinois. 3 Cf. preceding article on The Chemistry of the Dog’s Spleen and reference 23, page 44. 4E. Ritter: Zeitschr. f. physiol. Chem., xxxiv, p. 461, 1903. 5 Helen Baldwin: This Journal, iv, p. 213-219 (218), 1908. 5 W. Gilkin: Biochem. Centralbl., vii, pp. 289-306; 357-77, 1908. 7A. Windaus: Zeitschr. f. physiol. Chem., |xviii, pp. 110-117, 1910. 37 38 Cholesterol Determination them by means of digitonin. Lapworth® reports good results using this method. In order to more clearly understand where loss in Ritter’s method might occur, the literature was searched for studies on the action of the various chemicals used in the method upon choles- terol in forming salts, or actual decomposition products and the properties of these compounds. Lindenmeyer® prepared sodium cholesterolate, which he purified from chloroform by cooling the solution on ice, and found it to be insoluble in water, and that it was only slowly decomposed by water, weak alcohol hastening this decomposition. Obermiiller!? prepared - potassium cholesterolate and found the properties similar to those of the sodium compound. Darmstadter and Lifschutz# were able to obtain oxidation products of cholesterol by heating cholesterol with alcoholic potash with a reflux con- denser for eight hours, the yield being 20 to 25 per cent. Lif- schutz2 divided the products into three phases: (1) Oxy- cholesterin ethers; (2) Oxycholesterins and (3) Dicarbonséure (Chollanséure), the first two being soluble in all the ordinary solvents except water, and the last being soluble in water and alkalies and precipitating as white flocks on acidification. Schulze and Winterstein® noted a drop in the melting point of cholesterol exposed to the light, apparently due to oxidation, as an atmos- phere of CO, prevented this change. Lifschutz“ states that one hour’s cooking with half-normal alcoholic potash does not alter cholesterol. E. Schulze recovered cholesterol from substances in which potash did not liberate it so that it could be extracted by ether, by heating with benzoic acid in sealed tubes and forming an ester insoluble in alcohol and ether. 8 A. Lapworth: Journ. of Path. and Bact., xv, pp. 254-61, 1911. ®Lindenmeyer: Erdmann’s Praktische Chemie, xc, pp. 321-332, 1863. 10Obermiiller: Zeithschr. f. physiol. Chem., xv, pp. 37-48, 1891. 11J,. Darmstaédter and J. Lifschutz: Ber. d. deutsch. chem. Ges.,xxxi, pp. 1122-27, 1898. 2 J. Lifschutz: Zeztschr. f. physiol. Chem., 1, pp. 436-40, 1906-07. 13K. Schultz and E. Winterstein: Jbid., xliii, pp. 316-19, 1905; xlviii, pp. 546-48, 1906, 4 J. Lifschutz: Ibid., lviii, p. 175, 1908. 16° EK. Schulze: Zeitschr. f. anal. Chemie, xvii, p. 173. Harry,) fe 1 Corper 39 Very little reliable work has been done on the changes occurring in cholesterol during autolysis, probably in part because of the fact that no reliable quantitative cholesterol method has been avail- able, and partly because the change occurring, if such does occur, is naturally a slow one. Windaus,!® method for determining cholesterol and cholesterol esters, and Lifschutz’s!’ recent investiga- tions upen the determination of the presence of oxidation pro- ducts of cholesterol, should make research fruitful along these lines. Moore! failed to find any change in the cholesterol content of the liver in autolysis under toluene for forty-two days at 37°C. (Cholesterol analyzing 0.038 per cent before and 0.0372 per cent after autolysis). He also found no significant difference between the cholesterol content of a normal area (0.64 per cent) and an infarcted area (0.58 per cent) in a human spleen, and strongly objects to the reasoning of Carbone,!® who believes that cholesterol originates from lecithin by decomposition, and of Waldvogel?® who claims to have established the same by digesting lecithin with sterile liver juice, and who also found an increased cholesterol content in pathological livers as compared to normal (normal cholesterol content being 0.42 per cent, pathological—acute poisoning—being 24.46 per cent according to these analvses.) EXPERIMENTAL PART. Meruops. Ritter puts 50 grams of fat into a porcelain dish, adds 100 ce. of alcohol, brings it to a boil on the water bath, and then adds 8 grams of sodium dissolved in 160 cc. of 99 per cent alcohol with constant stirring. (The sodium alcoholate is pre- pared according to the method described by Kossel and Kriiger.?! These authors bring the absolute alcohol to a boil under a reflux condensor and carefully add the metallic sodium to it while boil- ing. They state that 10 cc. of a 5 per cent sodium alcoholate 16 Windaus: loc. cit. © 17 J. Lifschutz: Zeitschr. f. physiol. Chem., lili, pp. 140-48, 1907. 18. Craven Moore: Medical Chronicle, xivii, pp. 204-40, 1907-08. 19 Tito Carbone: Arch. ital. de biol., xxvi, p. 279, 1896. 20 Waldvogel and Mette: Muinch. med. Woch., lili, p. 402, 1906. 21 A. Kossel and M. Kriiger: Zeitschr. f. physiol. Chem., xv, p. 321, 1891. 40 Cholesterol Determination solution thus prepared will saponify 5 grams of mutton tallow, and 15 ce. will saponify 5 grams of butter fat).22 The alcohol is then evaporated off on the water bath and about one and one- half times as much salt as fat used is added, and enough water so that most of the contents of the evaporating dish goes into solution. This is then dried on the water bath with constant stirring, and then at 80°C, in a drying oven. It is pulverized, put into a sul- phuric acid dessicator for a short time, then into an extraction thimble, and is extracted in a Soxhlet apparatus with ordinary ether for nine hours. The ether extract is then put into a separa- tory funnel and shaken out with water to remove glycerin. The ether extract is dried, dissolved in hot alcohol, precipitated by means of water, precipitate dried at 100-120°C. and weighed. Experiments on the Effect of Sodium Alcoholate upon the Quantita- tive Yield of Cholesterol by Ritter’s Method. In order to test the loss occasioned by the steps in the method, the following experiments were carried out, using pure cholesterol instead of a complex fat mixture. EXPERIMENT 1. The amount of cholesterol which can be recovered from salt mixture, being mixed in alcohol solution, dried, and extracted by means of absolute ether in a Soxhlet apparatus, and the ether extract shaken out by means of water. (a) Used 0.1016 gram cholesterol and recovered 0.1026 gram. (b) Used 0.1020 gram cholesterol and recovered 0.1030 gram. EXPERIMENT 2. Amount of cholesterol which can be recovered from salt mixture after treating with sodium alcoholate, evaporating to dryness, dissolving residue in ether and shaken out by means of water. (a) Used 0.1057 gram cholesterol mixed with 5 ec. of 5 per cent sodium alcoholate and recovered 0.1048 gram cholesterol. (b) Used 0.1007 gram cholesterol mixed with 10 ec. of 5 per cent sodium alcoholate and recovered 0.1016 gram cholesterol. (c) Used 0.1005 gram cholesterol mixed with 40 cc. of sodium alcohol- ate and recovered 0.0950 gram cholesterol. (d) Duplicate of (c). Used 0.1003 gram cholesterol and recovered 0.1020 gram cholesterol. EXPERIMENT 3. Amount of cholesterol which can be recovered after solution in absolute alcohol, saponification by means of 10 cc. of sodium 22 The sodium alcoholate employed in the following experiments was prepared according to this method, 5 per cent strength being used. Harry J. Corper 41 aleoholate, evaporated to dryness, mixing with salt (10 to 15 grams), and extraction by means of absolute ether in a Soxhlet for 9 hours, shaking out ether extract with water, etc. (a) Used 0.1013 gram cholesterol and recovered 0.0390 gram. (b) Used 0.1012 gram cholesterol and recovered 0.0437 gram. EXPERIMENT 4. Similar to Experiment 3 but using only 5 cc. of sodium alcoholate. (a) Used 0.1016 gram cholesterol and recovered 0.0847 gram. (b) Used 0.1042 gram cholesterol and recovered 0.0858 gram. EXPERIMENT 5. Similar to Experiment 3 but using 40 cc. of sodium alcoholate. (a) Poured on salt in absolute alcohol solution, after saponifying. Used 0.1014 gram cholesterol and recovered 0.0415 gram. (b) Duplicate of (a). Used 0.1016 gram cholesterol and recovered 0.0390 gram. (c) Evaporated to dryness after saponification and emulsionized by means of water, mixed with salt, dried, extracted with ordinary ether, etc. Used 0.1032 gram cholesterol and recovered none. (d) Duplicate of (c). Used 0.1046 gram cholesterol and recovered none. (e and f) Not evaporated to dryness after saponification, emulsionized with saturated salt solution, dried, ground up, extracted with absolute ether, etc. Used 9.1023 gram cholesterol and recovered 0.0163 gram. Used 0.1002 gram cholesterol and recovered 0.0150 gram. EXPERIMENT 6. Amount of cholesterol that can be recovered after saponification by means of 40 cc. of sodium alcoholate, evaporating as nearly dry as possible, dissolved in ether and in water and these extracts poured on salt and dried and then extracted in a Soxhlet by means of ab- solute ether, etc. (a) Used 0.1010 gram cholesterol and recovered 0.0770 gram. (b) Used 0.1003 gram cholesterol and recovered 0.0700 gram. As a result of the above experiments we can conclude that even an excess of 5 cc. of 5 per cent sodium alcoholate added to choles- terol, will prevent its complete extraction from a dried salt mixture by means of ether. That the trouble lies in the use of an excess of sodium alcoholate is further shown by the following experiments: In order to simplify matters the following abbreviation is used for the various steps. A. Cholesterol dissolved in absolute alcohol and heated on water bath for three days. B. Saponifying with sodium alcoholate. 42 Cholesterol Determination C. Mixing with NaCl and drying: D. Extracting in a Soxhlet with absolute ether. E. Shaking out the ether extract with water in a separatory funnel. F. Allowing the ether to evaporate off at room temperature, and drying the residue at 100° C and weighing. TABLE OF RESULTS. a Cholesterol. COMBINATIONS USED RECOVERED LOSS a ee Set Eien 4 = = : pure gram gram gram Aaya Aosta Mert are Zhe. 0.1068 0.1075 0.00 Bech: pee ees hess 0.1001 0.1013 0.00 AFC e Dal. Hes raises 0.1001 0.0969 0.0032 1c a Oe Be gs 1 ee 0.1096 | 0.0955 0.0141 AvBICTD Eh see «i: 0.1061 0.0675 0.0386 In the above experiments the exact amount of sodium alcohol- ate used was not noted, as they were carried out before the first set cited. Now if an excess of sodium alcoholate thus affected the yield of cholesterol by the Ritter method when pure cholesterol was used, what would be its effect upon the cholesterol yield from tissues? In this case we are unable to tell the exact amount of fats and esters present, and therefore the amount of sodium alcoholate necessary to saponify them. If we use too small an amount our result will be high, due to the unsaponified fats and esters remain- ing as such with the cholesterol; and if we use too much the choles- terol yield will be low. To test these points the following experi- ments were carried out in the ether and alcohol extract from a steer spleen. The steer spleen weighed about 850 grams (moist) and yielded an alcohol and ether extract weighing 29.13 grams, which was dissolved in a liter of absolute alcohol and divided in 50 cc. samples (5 per cent. of the total extract) for the following analyses for cholesterol. Cholesterol. EXPERIMENT 1. ‘The extract after evaporation was dissolved in 10 cc. of absolute alcohol and warmed, 10 cc. of 5 per cent sodium alcoholate was added and the mixture warmed several hours, evaporated to dryness, redissolved in absolute alcohol, poured on 10 to 15 grams of salt, dried, extracted in a Soxhlet with absolute ether. Harry J. Corper 43 Results. (a) Yielded 0.0263 gram cholesterol. (b) Yielded 0.0195 gram cholesterol. EXPERIMENT 2. Identical with Experiment 1] except that the saponified mixture was poured directly on the salt (without evaporation). Results. (a) Yielded 0.0162 gram cholesterol. (b) Yielded 0.0380 gram cholesterol. EXPERIMENT 3. Identical with Experiment 2 except 30 cc. of absolute alcohol was used as a solvent before saponification. Results. (a) Yielded 0.0215 gram cholesterol. (b) Yielded 0.0528 gram cholesterol. EXPERIMENT 4. Identical with Experiment 2, but used 40 cc. of sodium alcoholate and evaporated the saponified mixture as nearly to dryness as possible, redissolved and poured on salt. Results. (a) Yielded 0.0072 gram cholesterol. (b) Yielded 0.0066 gram cholesterol. (c) and (d) were not evaporated after saponification before adding to the salt. (c) Yielded 0.0050 gram cholesterol. (d) Yielded 0.0100 gram cholesterol. EXPERIMENT 5. Identical with Experiment 1 except that only 5 cc. of sodium alcoholate was used. Results. (a) Yielded 0.1773 gram cholesterol. (b) Yielded 0.1600 gram cholesterol. EXPERIMENT 6. The object of this experiment was to compare an old preparation of sodium alcoholate (five months old) and of dark brown colo: with the freshly prepared compound. It is practically a duplicate of Experiment 5 but using 5 cc. of old sodium alcoholate. Results. (a) Yielded 0.1624 gram cholesterol. (b) Yielded 0.1840 gram cholesterol. EXPERIMENT 7. Resembled Experiment 5 except in that the mixture was not evaporated to dryness after saponification but poured directly on the salt after standing 24 hours. Results. (a) Yielded 0.1306 gram cholesterol. (b) Yielded 0.1312 gram cholesterol. EXPERIMENT 8. Identical with Experiment 1 but used 3 cc. 5 per cent sodium alcoholate. (The resulting product was only slightly oily in ap- pearance). Results. (a) Yielded 0.2180 gram cholesterol. (b) Yielded 0.2155 gram cholesterol. EXPERIMENT 9. Identical with Experiment 8, but used only 1 cc. of sodium alcoholate. (The product was not crystalline but oily). Results. (a) Yielded 0.3543 gram cholesterol. (b) Yielded 0.2956 gram cholesterol. As a result of the above experiments we can conclude that the best yield of cholesterol is obtained from the alcohol and ether 44 Cholesterol] Determination extract of the spleen when about 5 ce. of 5 per cent sodium alcohol- ate is used to saponify 1.5 gram of the ether-aclohol extract; we must, however, expect an error in the quantitative results on this amount of extract of from 10 per cent to 25 per cent. Cholesterol in Autolysis. Remembering the possibility for analytical error by the Ritter method as shown above, the following experiments can carry no great weight, but will merely be cited to show that there is no marked change in the cholesterol content of the spleen (dog) during autolysis. EXPERIMENT 1. Sixty-five grams of ground dog spleen were mixed with 0.338 gram of cholesterol suspension (made by dissolving the choles- terol in a minimum amount of absolute alcohol and pouring it into 0.9 per cent sodium chloride solution), toluene was used as preservative and the mixture allowed to autolyze for ninety-two hours at room temperature. Recovered 0.591 gram of cholesterol or 0.253 gram above the amount added which must have come from the spleen. (Extraction, etc., was carried on here as in the case of the steer spleen analyses.) EXPERIMENT 2. Sixty-five grams of ground up dog spleen was mixed with 0.314 gram cholesterol (suspended in 0.9 per cent NaCl-toluene water). The mixture autolyzed at room temperature for fifty-three hours. Recovered 0.5566 gram cholesterol or 0.242 gram of cholesterol over the amount added. As a few successful analyses for cholesterol were obtained while carrying out the autolysis experiments reported in previous papers”™ they may be put into tabulated form for comparison with the two above mentioned experiments. For the sake of convenience the figures will be given in the form of the amount of cholesterol found in one gram dry weight of spleen (on a basis of 23 per cent of solids in fresh spleen). In Experiments 1 and 2 above only the cholesterol content of the splenic tissue is given (that obtained by deducting the cholesterol added). 3 Work to be published in the Journal of Experimental Medicine upon correlation of chemical and histological changes. Harry sj -Corper AS | | CHOLESTEROL PER | SPLEEN TIME AUTOLYZED | GRAM DRY WEIGHT | SPLEEN : gram Normal (Spleen B. Ref. 3).......... 0.00 0.0150 Experiment 2: (above) ...0...5.05..4) 53 hours 0.0162 Experiment 1 (above)............... 92 hours 0.0169 *Six days autolysis (spleen H, Ref. 3) 6 days 0.0216 *Two days invivo autolysis (Spleen I, Rea De ase as ra | 48 hours in vivo 0.0217 *These two spleens were analysed at the same time; the normal spleens above being analysed at an earlier period. The last two cholesterol figures though differing from the first three by about 20 per cent are still within the error limit of the method. In conclusion we can say then that within the limit of error of the Ritter method for cholesterol, this constituent of the tissues does not markedly change in amount during autolysis. GENERAL SUMMARY. 1. A source of error was found in the quantitative estimation of cholesterol by the Ritter method, in the fact that the presence of an excess of sodium alcoholate over that necessary for the saponification of the fats and esters, prevents a complete extraction of the cholesterol from the salt mixture by means of ether. 2. This error may vary from 5 per cent to 20 per cent in the case of a normal tissue when there is an excess of from 1 ce. to 3 ec. of a 5 per cent sodium alcoholate solution used in the saponi- fication of 1.5 grams of the alcohol-ether extract. 3. The Ritter method for the quantitative determination of cholesterol in tissues should be used only with certain restrictions and precautions in mind. 4. No marked change was found in the amount of cholesterol present in the dog spleen after in vitro and in vivo autolysis of short duration. 5. The steer spleen contains about 0.4 per cent of its moist weight as cholesterol. a Ss ae) a 0) - =a t P A Ma on 7 ee il, A eh g aye aes Tee ws Le Oy TiC atin ver’ as aR a e wl —— , - Tengen a ' = ik ret f - Be Ta tr : yan eth { i | rey ty aise Cae | eek TANTS oda wot e RT > - \d i by : a anon writs (+ Cat el. t: reg dec Dy Vail he Cont Pian by. ahd qi ri) Mod oe a i ee . ¥ Heike jh preet ji = ore < ; ¥ 7 ve aks! 1 ‘oe A, al. Senne niaeel- | ‘| ra! . was) desl isi cet ae Ir Met? ni we ay Bama a5 . uta? aes) Feb a a ri ga ru ‘Ce bi ARATE - 85 %, ma u Mae yh ‘har . aer SAS teal wha aoe = : ase abl Ni , os } : SANG a2 uur ie Vea stoste we sop +k yu? DiT9 a Oe) aso t aes Sith as pe “4 Dae ia: . + a rete one a io 4 a 1s hA ‘alee nob - : rg % f ~y a ont i A i iy. coal ao o\P metre “se " 2 ‘ aS act se aa on he ¥ Pr eenit rae Hara 5 RE) et a nteinnge A ome a dt $5 FF aed ; ae at — a. 4) 5 = pa eit tel5 K: Hi undiyts, ‘ : (i Anil sthapen hg) teare LVS Ls 5 evn eee b ts 4 Lae ‘ jwtay.” Peyirthal) £ bien tp AG Andl haa v olay siegiod sh 5 B 5 ¥ “ te tis BT. i] en eoleya. a0) ¢ trey at) \ Sat ate Fahit ae oitd OP +53 asiewale Sole e Pa. sath ra eae ait isa Nahsgy GF slain id cesta HM reat it Rave ; Bith sy ‘ie expe ‘heAisioe oth ae i 3, ed P, 2 ciaeall op a iy” 1 rye? ad sive THE HAEMAGGLUTINATING AND PRECIPITATING PROPERTIES OF THE BEAN (Phaseolus). By EDWARD C. SCHNEIDER. (From the Department of Biology of Colorado College, Colorado Springs, Colorado.') (Received for publication December 4, 1911.) The extracts of a number of kinds of seeds are capable of pro- ducing in vitro an agglutination and sedimentation of the red blood corpuscles of various animals. This peculiar property is largely confined to species of the Leguminosae and to a few Sol- anaceae, although an occasional member of other families may possess it. The property was first noted among certain toxic seeds; the several species of Ricinus, Abrus pecatorius, and Croton tiglium.2 In recent years the list has been enlarged by a careful search for haemagglutinin bearing seeds. Landsteiner and Raubit- schek® found this property in extracts of beans, Phaseolus, peas, Pisum, vetches, Vicia, and lentils, Hrvuwm; and v. Eisler and v. Portheim‘ report its presence in five species of Datura. Mendel? added the following: sweet pea, Lathyrus odoratus; lentil, Lens esculenta; yellow locust, Robinea pseudacacia; five species of Vicia, Wistaria Chinensis, Caragana arborescens; senna, Cassia Mari- landica; and sweet rocket, Hesperis matronalis. He also found among beans that the haemagglutinins are absent in the Lima 1 Most of the work here reported was done in the Sheffield Laboratory of Physiological Chemistry of Yale University. The writer wishes to ex- press his hearty thanks to Professor Lafayette B. Mendel for the sugges- tion of the problem and for his kindly interest. 2 For the early literature on these see Jacoby: Biochemische Centralblatt, i, p. 289, 1903. 3 Landsteiner and Raubitschek: Centralblatt fiir Bakteriologie, 1 Abtei- lung, xlv, pp. 660-67, 1907. 4V. Eisler and v. Portheim: Zeitschrift fiir Immunitdtsforschung und experimentelle Therapie, i, p. 151, 1908. 5 Mendel: Archivio di fisiologia, vii, pp. 168-177, 1909. 47 48 Haemagglutinin of the Bean bean. Wienhaus® reports that this property occurs in the soy bean, Glycine or Soja hispida; and Assmann? found it is the seeds of Canavalia ensiformis, Datura stramonium, and three species of Lathyrus. The agglutinative property is not necessarily coincident with the toxie activity of seeds. It varies greatly in the seeds known to contain haemagglutinin and does not manifest itself equally well with the blood of different kinds of animals. Among laboratory animals Mendel’ reports the blood of the rabbit to be most sus- ceptible, and those of the pig and the sheep the most refractory. The extract of a number of the seeds noted above reacts well with rabbit’s blood but gives negative results with all other bloods tested. The reaction is strongest with suspensions of serum-free corpuscles. Landsteiner® found the normal blood serum of many kinds of blood capable of checking the process but that agglu- tination occurred readily when washed corpuscles were used. Several workers have suggested methods for obtaining purified preparations of the agglutinins from the crude extracts. Land- steiner and Raubitschek'® found that (1) the addition of a little acid produced a precipitate which contained only a trace of the agglutinin, the chief portion remaining in the filtrate. (2) When aleohol was added an agglutinative precipitate was obtained. It was also observed that when this precipitate was redissolved there was no loss of power. (3) The agglutinin was also salted out with the proteins on saturation with ammonium sulphate. From the extract of beans Wienhaus" separated a mixture of proteins to which he has applied the name of Phasin. Ten grams of bean meal were extracted with 500 grams of 0.9 per cent sodium chloride solution for twenty-four hours and then filtered. To the filtrate an equal volume of alcohol was added. A voluminous precipitate of albumin and globulin was secured in which the agglu- tinin is held quantitatively. On drying this precipitate in a * Wienhaus: Biochemische Zeitschrift, xviii, pp. 228-60, 1909. * Assmann: Pfliiger’s Archiv, exxxvii, pp. 489-510, 1911. 8’ Mendel: Loc. cit. See Raubitschek: Hamagglutinine pflanzlicher, Provenienz und thre Anti- kérper; Kraus and Levadite’s Handbuch der Technik und Methodik der Immunitdtsforschung, p. 625, 1911. 10 Landsteiner and Raubitschek: Loc. cit. 1 Wienhaus: Loc. cit. Edward C. Schneider 49 vacuum he secured a white powder which yielded to physiological salt solution all of the agglutinin and some inactive proteins. He suggests that he hopes later to free the ‘“‘Phasin” from proteins by digestion. Landsteiner’ employed the characteristic of erythrocytes that causes them to give up to the suspension fluid, when gently heated, the agglutinins with which they are combined. To this end he agglutinated, in an ice chest, sensitive serum-free corpus- cles with purified bean extract for several hours. The corpuscles were then washed with cold isotonic salt solution in a centrifuge until no trace of agglutinin was found in the washing solution. The agglutinated corpuscles were next suspended in a small amount of salt solution and stirred for an hour at 45° C. With precautions to avoid cooling they were then centrifugalized. By this means he obtained a clear but often red colored solution containing the agglutinin. This he found he could further purify by dialysis or with ammonium sulphate. Thus far the nature of these vegetable haemagglutinins has not been satisfactorily determined. lLandsteiner and Raubitschek conjecture it to be a protein by analogy with the very pure ricin isolated by Osborne, Mendel, and Harris.% The latter investi- gators separated the proteins of the castor bean, Ricinus zanzv- barensis, by dialysis and fractional precipitation with neutral salts and found the physiological properties, toxic and haemagglutina- tive, to be associated with the coagulable albumin. The agglu- tinative action was absent in the globulin and proteose fractions, and very active in the albumin fractions. SEPARATION OF THE PROTEIN CONSTITUENTS OF THE BEAN. In view of the experience of Osborne, Mendel, and Harris an attempt has been made to separate the haemagglutinin of the Scarlet Runner bean, Phaseolus multiflorus, Willd. A preliminary examination of a number of varieties of beans was made for the purpose of determining which is richest in haemagglutinins. 12 See Raubitschek in Kraus and Levadite’s Handbuch der Technik und Methodik der Immunitdtsforschung, p. 625, 1911. 13 Osborne, Mendel and Harris: American Journal of Physiology, xiv, pp. 259-86, 1905. THE JOURNAL OF BIOLOGICAL CHEMISTRY, XI, NO. 1. 50 Haemagglutinin of the Bean Among these were the dwarf wax-podded varieties Burpee’s Kid- ney, Wardell’s Kidney Wax, Red Kidney, Dwarf Champion, and Early Six Weeks; and the climbing wax-podded variety Golden Champion of P. vulgaris, L.; also the Scarlet Runner, P. multi- florus, Willd. The extracts prepared from equal weights of bean meal were almost equally active. The Scarlet Runner seed is much larger than the seeds of the other varieties which favored the removal of the seed coat. The Scarlet Runner beans were first passed through a very coarse grinder. Much of the seed coat was thus broken away from the substance of the coty- ledons and was blown out with an air blast. Afterward these cracked beans were ground to a coarse meal and treated with benzine to remove the oil. Following this the coarse meal was ground to a powder and 1 kilo of it was extracted with 5 liters of a 2 per cent sodium chloride solution that had been previously heated to 60° C. After frequent stirrings for two hours it was placed in a cold room over night and then filtered perfectly clear. The extract was dialyzed in running water for thirty-six hours. The precipitate I which separated was filtered from the solution Banddried. Unfortunately precipitate 1 was dried so slowly that more than two-thirds of it was changed into an insoluble protean. Solution B was further dialyzed three days and yielded a heavy precipitate 1] which, when dried, was more than 85 per cent soluble. The solution while dialyzing tended to become acid in reaction and required frequent neutrali- zation. It was protected against decomposition with toluene. Solution B1, which remained after filtering off precipitate 1], was again dialyzed four more days and yielded a small amount of precipitate 111. Precipitates 1 and 1] were the globulin phaseolin; and 111 was probably the other globulin, phaselin, separated by Osborne from the kidney bean. The proteins remaining in solution B2 (obtained from B1 on filtering off precipitate 111) were salted out by saturating with ammonium sulphate. This procedure yielded precipitate 1V and solution B3. Solution B3 was then dialyzed in running water until free from salts when it was found it did not contain a trace of the haemagglutinin. Precipitate 1V was dissolved in a small volume of water and the clear solu- tion C was then saturated with magnesium sulphate and weakly acidulated with acetic acid. The small amount of precipitate V, which will be called albumin, was then filtered from solution Cl. The albumin precipitate, V, was redissolved and precipitated with mag- nesium sulphate and then dissolved in a very small volume of water. From this solution the salt was removed by dialysis. The solution was next evap- orated at a low temperature and yielded 0.3 gram of albumin. 144Qbsorne: Journal of the American Chemical Society, xvi, p. 635 and p. 707, 1894. Edward C. Schneider 51 Solution Cl was dialyzed for several weeks until free from salt, then was evaporated in low dishes at 48°C. Almost a gram of proteoses was secured from this solution. THE ACTION OF THE BEAN PROTEIN PREPARATIONS ON BLOOD. Tests were always made with defibrinated rabbit’s blood diluted (1:5) with 0.9 per cent sodium chloride solution. One cubic centimeter of this blood mixture was placed in a small and very narrow test tube; and 2 cc. of the protein preparation, dissolved in the salt solution, were added. The time of the visible beginning of agglutination and the condition at the end of two, four, and twelve hours were noted. Preliminary tests with the phaseolin and phaselin (preparations I, II, and III) revealed the presence of haemagglutinin. Prepara- tion II was most active but none of the globulin preparations exhibited the property in a striking degree. Believing that these proteins adsorbed the haemagglutinin, an attempt was made to purify precipitate II. About half of preparation IT was dissolved in 0.9 per cent sodium chloride solution. One-half of this solution was dialyzed until it yielded its phaseolin, preparation Ila, and the other half of the solution was saturated with magnesium sulphate. The resulting precipitate was redissolved in water and on dialysis yielded preparation IIb. Preparations Ila and IIb were less active than preparation II. This weakening in activity by purification indicates that the haemagglutinin in these preparations is held there by adsorption. The albumin, purified from precipitate v, was also active but less so than the globulins. The degree of activity of the albumin and globulins is given in Table I. The proteose preparation was found to be rich in haemagglu- tinin. It produced strong agglutination when present in blood dilutions of one part to 100,000 and more. Wienhaus” found his crude product ‘‘Phasin’” completely agglutinated rabbit’s blood in the ratio of 1:7000 in fifteen hours; and with cat’s blood, which reacted still better, in dilutions of 1:11,000 in eighteen hours and 1:60,000 in twenty-three hours. Assmann" also working with a 1s Wienhaus: Biochemische Zeitschrift, xvili, p. 232-33, 1909. ® Assmann: Pfliiger’s Archiv, cxxxvii, pp. 489-510, 1911. 52 TABLE I. Haemagglutinin of the Bean Agglutination Trials with Protein Preparations. MILLIGRAMS ADDED TO1 CC. | FIRST INDICATION PREPARATION USED REMARKS OF BLOOD OF AGGLUTINATION MIXTURE No.1. “Phascaliae 5.000 2 minutes Complete in 2 hours 1.000 25 minutes Partial in 12 hours 3.000 At once Complete in 2 hours 0.600 2 minutes Complete in 4 hours No- Il Phascelinee | 0.300 5 minutes Complete in 12 hours 0.200 15 minutes Complete in 18 hours | 0.150 | Trace | 0.100 Negative | 0.600 4 minutes Complete in 4 hours No le. Phocealnn 2 | 0.300 10 minutes Complete in 12 hours 0.200 (?) Partial in 12 hours | 0.150 Negative 1.500 2 minutes Complete in 12 hours No. IIb. Phaseolin. 0.750 (?) Complete in 18 hours 0.300 Negative : 3.000 4 minutes Complete in 4 hours No. III. Phaselin.. . ; ara { 9.600 (?) Trace Nov Albumin) |) 37200 4 minutes Complete in 12 hours 0.610 Negative 0.300 At once Complete in 2 hours 0.060 Atonce | Complete in 2 hours PrGieGhels Gs 0.020 2 minutes Complete in 4 hours 0.015 15 minutes Complete in 12 hours 0.0075 19 minutes | Complete in 18 hours 0.0037 (?) | Partial in 18 hours “Phasin” preparation obtained agglutination of diluted rabbits blood in 1:35,000. The rapidity of action and the dilutions of the proteose preparation that are effective are given in the latter part of Table I. This preparation is very soluble and gives the dis- tinctive proteose tests. One milligram of the proteose dissolved in 1 cc. of salt solution added to 5 ce. of undiluted defibrinated rabbit’s blood produced almost instantaneous agglutination; and asolid clot-like mass of corpuscles settled out leaving a clear serum in less than half an hour. Table II gives further testimony as to the power of the haemagglutinin associated with the proteose. Edward C. Schneider 53 TABLE II. | : COMPLETE VISIBLE AGGLUTINATION AGGLUTINATION] AND SEDIMENTATION MILLIGRAMS OF PROTEOSE ADDED {| 1.0in 1 cc. 0.9 percent NaCl | Atonce | 20 minutes 0.5in 1 cc. 0.9 per cent NaCl At once | 45 minutes 0.4in 1 cc. 0.9 per cent NaCl At once 1 hour 0.3in 1 cc. 0.9 per cent NaCl At once 2 hours 0.2in 1 ec. 0.9 per cent NaCl | 2 minutes 6 hours 0.1 in 1 cc. 0.9 per cent NaCl | 4minutes | Incomplete in 6 hours* 5 ce. of 1:5 blood * Further observation was impossible because it then stood over night at room temperature. All but the last gave a firm clot-like mass in the time recorded. It seemed probable, in view of the observation that the haemag- glutinin was largely confined to the proteose preparation, that all proteoses might cause agglutination. Hence tests were made with Witte’s peptone upon the diluted rabbit’s blood corpuscles but these were entirely negative. DOES AUTOLYSIS ACCOUNT FOR THE HAEMAGGLUTININ? The presence of the haemagglutinin in the proteose preparation also suggested that it might be a product of the hydrolysis occur- ring in the solution during the period of extraction and later. To settle this point a fresh extract was prepared as rapidly as possible and immediately tested for its agglutinating power. A portion of the extract was also immediately heated for five minutes at 82° C.—a temperature the haemagglutinin withstands for thirty minutes without injury!7—the coagulated proteins were filtered off and the filtrate was then tested for the relative amount of haem- agglutinin. There was slightly less in this than in the original extract as is shown-in Table III. This is very likely due to a slight adsorption by the coagulated proteims. The remaining portion of the original extract, after the addition of toluene, was set aside in a cool room for thirty days. Its agglutinating power was again determined on the eighth and thirtieth days. There was not a decided change in agglutinating power as will be observed 17 Wienhaus: Loc. cit. 54 Haemagglutinin of the Bean TABLE III. EXTRACT naioeen! AFTER HEATING | | EXTRACT WITH 0.9 PERCENT, FRESH EXTRACT FIVE MINUTES AT | pimtineebc: wad THIRTY DAYS Nac? | | Soles Pe a eae oLD = ‘ | a —— | = ee = Undiluted | Complete | Complete Complete _ Complete 1:100 | Complete | Complete | Complete _ Complete 1:200 Complete ' Complete / Complete | Complete 1:300 , Complete | Complete Complete ' Complete 1:400 | Complete Partial | Partial Complete 1:500 | Partial Slight | Partial | Complete 1:600 | Partiai | Negative | Slight Partial *Two cubic centimeters of extract and 1 ce. of 1:5 blood used in each test. Agglutination recorded at end of twelve hours. in Table III. It was also found that active agglutinins may be secured by extracting the bean meal at 80° C. From these obser- vations it would seem that autolysis does not account for the haemagglutinin in the proteose preparation. Digestion trials. Obsorne, Mendel, and Harris!® showed that the toxicity and agglutinating power of their pure preparation of ricin could be impaired or destroyed by pancreatic digestion pro- longed two or three months. Wienhaus,!* on the other hand, in digestive trials with pepsin, trypsin, and papain made on his “‘Phasin” for periods ranging from three to seven days failed to show any destructive action. The haemagglutinative proteose preparation was subjected to various digestive trials with trypsin, erepsin, and mixtures of these two, in water and in sodium carbonate solutions for a period of twenty-eight days with practically negativeresults. The diges- tive mixtures were tested with fresh blood fibrin and Witte’s peptone several times during the period and found to be active. Wienhaus calls attention to the fact that his ‘‘Phasin” acts as a protein and he expresses the opinion that it is a protein or enzyme- like substance. Since Wienhaus’ digestion trials were so very short and the effective trials of Osborne, Mendel, and Harris were so prolonged the failure of the proteose preparation to respond to digestive agents in the time allowed still leaves the question of the 18 Osborne, Mendel and Harris: American Journal of Physiology, xiv, p. 284, 1905. 129 Wienhaus: Biochemische Zeitschrift, xviii, p. 256, 1909. Edward C. Schneider 55 digestibility of these haemagglutinins open. A more prolonged series of carefully controlled digestive trials is planned for the near future. IS THE HAEMAGGLUTININ A FOOD STORED FOR THE USE OF THE GROWING SEEDLING? If the haemagglutinin of the seed is a proteose it should readily be utilized by the growing seedling in the early growth after germination. It is also probable that preliminary to the translo- cation of the protein from the cotyledons to the growing tissues of the seedling further haemagglutinin may be formed from the proteins by the action of the enzymes evolved during germination. It certainly is surprising to find the haemagglutinins in the proteose portion of the seed, inasmuch as proteoses and peptones are not commonly normal constituents among the reserve proteins of seeds. They are of course present to some extent during germination. It may be noted here that Osborne? found a small amount of pro- teoses when he studied the proteins of the kidney bean. Hedid not determine whether the proteoses were a normal constituent of the seed or a product of autolysis during extraction. Landsteiner and Raubitschek” showed the agglutinin to be absent from green beans. A future study must determine when the haemagglutinin enters the seed and an attempt be made to learn its source, whether it is formed in the seed or brought to it to be stored. To determine if the haemagglutinin is utilized by the seedling and whether it is increased in amount during germination a study was made of seedlings and cotyledons at frequent intervals, from the beginning of germination until the depleted cotyledons fell from the seedling. Two series of observations were made, one with plants grown in darkness and the other with sturdy plants grown in the light. For the determination of haemagglutinin content the seedlings were hastily washed and the cotyledons separated from the seedling close to the stem, and then cotyledons and seed- lings dried separately. When dry each was ground to a powder and known weights extracted with constant proportions of a 0.9 20 Osborne: Journal of the American Chemical Society, xvi, pp. 758-64, 1894. 21 Landsteiner and Raubitschek: Loc. cit. 56 Haemagglutinin of the Bean per cent sodium chloride solution. Three kinds of beans were used, the Scarlet Runner, Wardell’s Kidney, and the Early Six Weeks. The cotyledons of the last two are lifted above the soil by the grow- ing stem of the seedling while those of the Scarlet Runner remain underground. The underground habit of the Scarlet Runner made it difficult to secure from the late stages cotyledons that had not undergone decomposition to some extent. The data obtained from the three kinds of beans were wholly concordant through- out each series. Repeated tests with colorless seedlings and with the green leaves and stems of those grown in the light failed to show the slightest trace of agglutinative power. Hence the haemagglutinin as such is not carried into the seedling or, at least, not in sufficient amounts to be detected. Roots, stems, and leaves were also examined separately from plants of many sizes, all being negative. The agglutinin is not a normal constituent of the organs of the vegeta- tive plant. As a type of the results obtained, those with the cotyledons of the Wardell’s Kidney bean are given in Table IV, p. 57. Dur- ing the early days of growth the agglutinative action for equal weights of cotyledon is only slightly lowered, which indicates that the haemagglutinins are withdrawn gradually along with other stored foods. Later there is a more rapid disappearance of the agglutinin. Extracts prepared from cotyledons that fell from the seedlings of Wardell’s Kidney bean and Early Six Weeks bean, grown in darkness, gave no agglutinative response. From seed- lings of all three varieties when grown in light, and the Scarlet Runners grown in darkness, it was impossible to get depleted coty- ledons wholly free from the haemagglutinins. With each, however, there was a very marked quantitative reduction in this property. It follows, therefore, that the haemagglutinin of the bean is utilized or destroyed, along with other stored foods, by the developing seedling. THE PRECIPITATING REACTION OF BEAN EXTRACTS. When the clear extract of any of the several beans examined in this study was added to rabbit’s blood serum a flocculent precipi- tate always appeared. The reaction usually occurred slowly. Edward C. Schneider 57 TABLE IV. Wardell’s Kidney Bean During Germination and Early Growth. Grown in Light. WEIGHT OF TWENTY GREATEST DILUTION AT WHICH LENGTH OF SEEDLING COTYLEDONS AGGLUTINATION WAS OBTAINED grams centimeters 6.000 0 1:550 0.918 9.0 1:350 0.606 13:2 1:300 0.300 17.8 1:100 0.262 + Undiluted Grown in Darkness. 1.000 9.4 1:400 0.508 13.2 1:200 0.245 18.3 1:150 0.215 35.5 Negative* * Cotyledons that had fallen from seedlings. For some minutes, and often more than an hour, after the addition of the extract to the blood serum the mixture remained clear. It then gradually became cloudy and opaque, finally the white floccu- lent precipitate appeared. The entire reaction may be completed within a few minutes when strong extracts are used but will require five or more hours with dilute extracts. This precipitating reaction is not constantly associated with the agglutinative property of seed extracts. It was found to be absent in extracts from such agglutinin containing seeds as the Wistaria Chinensis, the hairy vetch, Vicia vilosa, and the pea, Pisum sativum. From the sweet pea, Lathyrus odoratus, an extract was obtained that gave a slight clouding of the serum but it failed to produce a precipitate. A fresh extract prepared from Scarlet Runner bean meal was heated repeatedly at various temperatures for five-minute inter- vals; and after each period of heating the coagulated proteins were filtered off and the filtrate then tested for the agglutinating and precipitating properties. Both properties continued practi- cally undiminished up to a temperature of 80° C. At 83° C. the precipitating power was destroyed in ten minutes. Table V 58 Haemagglutinin of the Bean TABLE V. CONDITION OF EXTRACT PRECIPITATE IN 8ERUM AGGDUR NON OF CORPUSCLES Tha] Re de el Heavy in 50 minutes Strong After heating at 80° for five minutes. .| moderate in 4.5 hours | Strong After heating at 83° for five minutes. .| Trace in 7 hours Strong After heating at 85° for five minutes. .| Negative Strong After heating at 87° for five minutes. .| Negative Strong After heating at 91° for five minutes. .| Negative Strong After heating at 94° for five minutes. .| Negative Negative shows a trace present after five minutes at 83°. The agglutina- tive power was weakened above this temperature, but withstood five minute exposures to 91°, and was wholly destroyed at 92°. Table V contains the data obtained from one series of heat tests. The protein preparations separated for the study of the agglu- tinins have also been tested for the precipitin reaction. The glob- ulin preparations II, Ila, and IIb were rich in it while I contained a trace. The albumin (V) gave a negative test, and 3 mgm. of the proteose preparation in 2 cc. of serum failed to give the reaction. It was also found that after serum had been added repeatedly to extract until no more precipitate formed that the mixture retained its agglutinating power practically unaltered. These several differences warrant the conclusion that the precipt- tating and agglutinating properties of the extracts of beans are due to different constituents of the seed. Or we may better express it that rabbit’s blood contains a precipitin for certain of the bean’s proteins. Wienhaus* made certain observations which are of interest in this connection. He found his “Phasin’”’ did not react with serum taken from hen’s blood. On adding the preparation to a clear fluid collected from the joint of a diseased knee a heavy precipi- tate was obtained. After immunizing rabbits to the phasin it was impossible to obtain a precipitate in the blood serum on the addition of phasin. He points out that this is contrary to the experience of Jacoby and others when they immunized animals to ricin, abrin, and crotin, inasmuch as these substances gave a pre- cipitate when added to the immune sera. It would seem from Wienhaus’ work that the agglutinating and precipitating proper- ties are both lost for the blood on immunizing the animal. 22 Wienhaus: Loc. cit. Edward C. Schneider 59 The precipitating property does not occur in extracts of bean plants. It disappears from the cotyledons, as does the agglutinin, with germination and the growth of the seedling. SUMMARY. 1. The proteose prepared from the Scarlet Runner bean was found to be a very active haemagglutinating agent. Other bean proteins contained some haemagglutinin but this was shown to be adsorbed by them. 2. The haemagglutinin is not a product of autolysis. 3. The haemagglutinin gradually disappears from the cotyle- dons, simultaneously with the stored food, as the seedling develops. 4. The agglutinative property does not occur in the extracts of the roots, stems, or leaves of the bean plant. 5: The addition of the clear extract of beans to rabbit’s blood serum produces a flocculent precipitate. This reaction is not coin- cident with the agglutinating property of all haemagglutinin con- taining seeds and appears to be chiefly associated with the phaseo- lin in the bean. Since this paper was written there has been brought to my attention an abstract, in the Zentralblatt fiir Biochemie, xii, p. 391, 1911, of a recent paper by v. Eisler and v. Portheim. They regard the haemagglutinin as a protein and prove it to be a reserve substance that disappears from the embryo during germination. ft? 2 Ve flee Mia cade BG ao. "> Ma! SUCRE eSeet ob ee aes HLS area tS OD Se wi Seo Sroe ’ - . MP Hil ‘ tf siti Pein oehs Sieaat , oc er ee ae fie iy tigaton M Putt eA LAD itCaean eaHs ON THE RECOVERY OF ALCOHOL FROM ANIMAL TISSUES. ’ By PAUL J. HANZLIK. (From the Pharmacological Laboratory of the Medical School of Western Reserve University.) r (Received for publication, December 7, 1911.) In the course of some experiments on the absorption of alcohol, I have found it advantageous to introduce certain modifications in the current distillation method. The procedure which is herewith described differs only in a few points from those already in use, but these points are rather impor- tant. They involve (1) the digestion of the tissue with phosphoric acid to liberate the alcohol and to facilitate its distillation; (2) the automatic filtration of the distillate to remove the volatile solid products which would interfere with the specific gravity determination, and (3) a more delicate ring modification of Anstie’s test, to determine the completion of the distillation. The method is as follows: The organ or tissue (15 to 300 grams) is placed in a distilling flask of 1000 ce. capacity and 50 per cent phosphoric acid (5 to 25 cc.) is added, together with 300 cc. of water. The flask is then connected with a condenser. In the distal end of the condensing tube a plug of dry absorbent cotton is packed _ rather firmly. The flask is heated over a direct flame or sand bath until the distillate no longer gives a bluish or light green color with Anstie’s test made by contact. A distillate of 200 cc. usually suffices with small quantities of alcohol. The distillate, which is free from insoluble matter, is stirred by gentle rotation of the flask and carefully measured and its specific gravity determined with a pyknometer at 25° C., in the usual manner. The alcohol percentage is calculated from the alcohol specific gravity tables of the United States Pharmacopoeia. 61 62 Recovery of Alcohol from Tissues The modifications were adopted for the following reasons: 1. The addition of phosphoric acid: This has two advantages: it liquefies the tissues and thus hastens the distillation, and it makes the recovery of the alcohol more complete, increasing the yield by about 1.3 per cent. The quantitative data will be discussed later. 2. The automatic filtration of the distillate through cotton: Dis- tillates from animal tissues, particularly from the alimentary tract, contain some white flaky material consisting of fatty acids, indql, skatol, ete. They are especially abundant if the tissue is acidified. These falsify the readings of the pyknometer, and, there- fore, had to be eliminated. Various modes of chemical treatment were tried to effect their removal, but with no practical success. Sodium hydroxide pre- vented the volatilization of the fatty acids but the residue foamed so much that it was impossible to distil it. Other alkalies such as calcium hydroxide and sodium carbonate had the same disadvan- tage. In neither acid nor alkaline media was the volatilization of indol and skatol prevented. Redistillation was ineffective. The object was finally accomplished by simple filtration of the distillate through filter paper or cotton. This was combined with the distillation in the manner described above in order to avoid loss of alcohol by repeated handling of the distillates and to make the process as practical as possible. In this way the distillates always appeared free from any insoluble matter and the pyknom- eter weighings were not affected. 3. The contact modrfication of Anstie’s method. A practical quali- tative method of detecting alcohol was needed, to insure the com- pleteness of the distillation. All the more common methods were tried, but the bichromate-sulphuric acid test: (also known as the Anstie test), modified so as to make it more delicate, was found to be the most useful. If solutions containing very small quantities of alcohol are used, the green color obtained becomes too diffuse and can not be recognized. To avoid this, the ‘{‘contact-test’’ was performed as follows: The solution containing the alcohol (about 1 ce.) is placed in a test-tube; then the bichromate-sulphurie acid solution? (about 0.5 ec.) is introduced by means of a pipette beneath ! Merck: Reagentien Verzeichniss, Darmstadt, 1903. 2 Five-tenths of a gram of potassium bichromate dissolved in 75 grams of concentrated sulphuric acid. Paul J. Hanzlik 63 the alcohol layer taking care not to mix the solutions. At the point of contact a blue or light green ring will develop depending upon the concentration of the alcohol solution. After standing for a short time the ring becomes more intense but gradually fades away owing to diffusion and the establishment of equilibrium between the two liquids. The other tests were performed in the usual manner. In all cases blank tests on distilled water were simultaneously performed. The results obtained are shown in Table I. Positive reaction is designated by +; negative by —. TABLE I. STRENGTH OF ALCOHOL USED TEST BLANK 1:2000 1:5000 1:€000 1:10,000 | [— Bichromate-sulphu- - + ue spk | ef ric acid (contact)| (yellow)| (blue) | (light blue)| (green) | (light green) Todoform. 5-5... . - + == = = Ethyl benzoate | | (Berthelot)....... _ + ae ae & Ammonium molyb- Gates sre ss. _ | -- — | a oe % It can readily be seen that the bichromate-sulphurie acid test was the most sensitive. The color ring usually appeared quite promptly (within five minutes). With high dilutions of alcohol (1: 10,000) a light green ring appeared inabout ten minutes. Next in order of sensitiveness was the iodoform test. In: high dilutions, however, it was difficult to differentiate the odor of iodine from iodoform and the result of a search for crystals was often unsuccess- ful. Least sensitive and reliable of all were the ethyl benzoate and molybdate tests. In high dilutions, it was impossible to differentiate the odor of benzoyl chloride from that of ethyl ben- zoate, while the molybdate gave an almost indistinguishable faint blue tint with the lowest dilution of alcohol. Variable and uncertain results are to be expected in tests which require the olfactory sense. On the other hand, if properly carried out, an objective test, such as the bichromate-sulphuric acid test, is apt to give more constant and certain results under otherwise varying conditions. 64 Recovery of Alcohol from Tissues Quantitative control tests: The procedure here described, was tested out in the following manner. Blood and intestines of cats and dogs were used. The blood was intimately mixed with the alcohol. The viscera, deprived of their circulation, were ligated at both ends and injected with different quantities of alcohols of known strengths. The total quantity of alcohol varied between 0.6 and 5.1 grams. The material was allowed to remain different lengths of time before the recovery of alcohol was begun. The results are presented in Table II. TABLE II. | ABSOLUTE | ABSOLUTE - seg tile geet Seger eco aee ha cee aon eee grams | grams per cent 1 1.0245 | 1.0270 | 100.24 | Half of whole Phosphoric intestine acid — 3 5.1100 | 4.8816 95.53 | Blood 100 ce. and) Phosphoric half of intestine} acid 4 0.6108 | 0.6300} 103.14} Blood 100.ce. Phosphoric acid 5 5.0900 | 4.9400. 97.05 | Blood 150 ce. Phosphoric acid 6 0.6108 ; 0.6200} 101.50 | Intestine 15 cm. | Phosphoric acid 10 1.4139 | 1.4080 99.58 | Intestine 15 em. | Phosphoric acid nol 1.2568 | 1.2719 | 101.20 | Intestine 15 cm. | Phosphoric _ acid 12 1.2568 | 1.2640) 100.57 | Intestine 15 em. | Phosphoric acid a ae Average | 99.85 it 2.0360 | 1.9950 97.99 | Whole intestine | Water alone 9 1.0180 | 1.0099 99.20 | Intestine 15 cm. | Water alone oe ES Average | 8 1.0180 | 0.9986 98.10 | Intestine 15 cm. |Sodium hydrox- Grand i. average } Paul J. Hanzlik 65 An inspection of the table shows that distillation with phosphoric acid gave the highest results. The amount of alcohol recovered above that when water alone (Experiments 7 and 9) was used was about 1.3 per cent, and above that when sodium hydroxide (Ex- periment 8) was used about 1.8 per cent. Inasmuch asthe individ- ual results under varying conditions were quite comparable with the average (99.85 per cent), it would seem justifiable to conclude that the procedure is suitable for quantitative purposes. The best results were obtained when the quantity of alcohol did not exceed 2 grams. The results obtained agree favorably with those reported recently by Bacon? with the aid of the refractometer. The average of his six experiments was about 97.4 per cent with quantities of alcohol ranging from 0.95 gram to 8.0 grams in variable strengths. There were no animal tissues involved in the residues from which the distillates were obtained. Hamill‘ has reported a quantitative method for the determina- tion of alcohol in perfusion fluids and tissues: The alcohol is recovered by distillation and estimated volumetrically in the dis- tillate by sulphuric and chromic acids. Good results are claimed to have been obtained with quantities varying from 0.5 to 1 part per mille. Nicloux’s® method, also based on the principle of oxi- dation of alcohol by sulphuric acid and potassium bichromate, is said to be accurate for small quantities, the limit of error being about 5 percent; in more experienced hands somewhat less. CONCLUSIONS. The modifications in the method of alcohol estimation in animal tissues, which are described in this paper, give results which are accurate within 1 per cent. The bichromate-sulphuric acid test reveals the presence of alcohol in dilutions of 1:10,000. I wish to thank Professor Sollmann for his suggestions and criticisms in this work. 3 Bacon: Circular No. 74, Bureau of Chemistry, U. S. Department of Agriculture, July 14, 1911. 4 Hamill: Journ. of Physiol., xxxix, p. 476, 1910. 5 Abderhalden: Handbuch der biochem. Arbeitsmethoden, ii, p. 7, 1909. THE JOURNAL OF BIOLOGICA™ CHEMISTRY, XI, NO. 1. ) hitbas ,: Jtatatede wales mak oe eth: ; rash kee lags ii He ABtbiinhsic of TTR ae aang gas 3 es : inguy 9 ‘ey? Ge i is i aif ene ; . iv Pas La ‘ “ pitt}. FP rah ; cS ' — “ nn from guanine. These workers reasoned from anal- ogy that the compound which they obtained must be 2-oxypurine but they did not offer any direct proof of its structure. It may be for this reason or through an oversight that Fischer in the intro- duction to his book, Untersuchungen in der Puringruppe, page 49 (1907) states that 2-oxypurine is still unknown. The writer has synthesized 2-oxypurine from 2-oxy-5,6-diamin~ opyrimidine and finds that it agrees in all respects with the descrip- tion given by Tafel and Ach of their compound. When 2-oxy-5,6-diaminopyrimidine’ (IV) was heated with formic acid a monoformyl] derivative was obtained. This yielded a potas- sium salt which, when heated, gave off water and changed to the 1 Amer. Chem. Journ., xli, p. 58, 1909; Zbid., xlv, p. 79, 1911; This Journal, ix, p. 161, 1911. * Ann. d. Chem. (Liebig), |xxiii, p. 328, 1850. 3 Ber. d. deutsch. chem. Gesellsch, xxx, p. 2228, 1897. 4 Tbid., xxx, p. 2218, 1897. 5 [bid., xxxiv, p. 1170, 1901. § Johns: Amer. Chem. Journ., xlv, p. 82, 1911. 67 68 Researches on Purines potassium salt of 2-oxypurine. As good yields were obtained, reasonable quantities of 2-oxypurine can be made in this manner. 2-Oxy-purine is characterized by the fact that it crystallizes with one molecule of water so firmly bound that it does not escape at 110° C. On heating at 130° C. the crystals become anhydrous. Neither hypoxanthine nor 8-oxypurine crystallize with water. Of the salts of 2-oxypurine, the picrate, nitrate, and hydrochloride are easily prepared. When 2-oxy-5,6-diaminopyrimidine is boiled with acetic anhy- dride it forms chiefly a monoacetyl compound together with some of the diacetyl compound. When the potassium salt of the mono- acetyl compound is heated it yields the potassium salt of 2-oxy- 8-methylpurine(V). Thispurineforms a picrate and a nitrate, both of which have rather definite decomposition points. These salts may therefore be used to identify this purine. Work on the preparation of 2-oxy-1-methylpurine is almost completed and this compound will be described in a later paper. N=. CH HN — CO N = CH | OC C—NH HC C—NH HC C—NH lat ASS | [SNe | spa CE CH CO Nee bath eee HN——C—N N——C—N N——C—NH I II 1 | N: =s ONE No CH | oC CNH, — > OC C—NH | l Sc.cx, | ecient HN——CH HN——C-—-N EXPERIMENTAL PART. Formyl-2-oxy-5 ,6-diaminopyrimidine, CsHgQ2Ns. Four grams of 2-oxy-5,6-diaminopyrimidine’ were dissolved in 10 ce. of 85 per cent formic acid.. The solution was heated on 7 Johns: Loc. cit. Carl O. Johns 69 the steam bath for an hour after which it was evaporated to dryness. The residue was treated with a little aleohol and evaporated again to remove the last traces of formic acid. It was then dissolved in dilute ammonia, a trace of insoluble material was filtered off and the filtrate was evaporated to dryness. A yield corresponding to 90 per cent of the calculated was obtained. The portion used for ‘analysis was recrystallized from water and was obtained as a powder composed of aggregates of very minute crystals. These were easily soluble in hot and sparingly soluble in cold water, and almost insoluble in alcohol. Calculated fo. CsHsO2Na: Found: i Se 2.2 c= 2 os ta'es «wee 36.36 36.33 iS ie | 2-Oxypurine, ie co. - | ~ 2 NH——C—N When formyl-2-oxy-5,6-diaminopyrimidine was heated at 160°—-170° C. it blackened considerably and, although it was partly changed to 2-oxypurine, the reaction was unsatisfactory. A good yield of the purine could be obtained by heating the potassium salt of the formyl compound. This salt was made by dissolving the formyl compound in a small volume of water containing a little more than one molecular proportion of potassium hydroxide. Alcohol was added gradually to this solution to precipitate the salt which separated as a white powder. Five grams of this potas- sium salt were heated at 150°-160° C. for an hour. Water was liberated, leaving a light brown crust. This was dissolved in water and the solution was decolorized with blood coal whereupon it was acidified with acetic acid. On standing over night, the solu- tion gave a precipitate of small globules which in turn were found to be aggregates of very minute prisms. These were recrystal- lized from water. The yield was 70 per cent of the calculated. The crystals contained one molecule of water which they did not lose at 110° C. When heated at 130° C. they became anhydrous. Analyses of samples dried at 110° C. gave the following results: 70 Researches on Purines I. 3.7600 grams of substance lost 0.4400 gram at 150° C. II. 2.6800 grams of substance lost 0.3100 gram at 150° C. III. 0.7792 gram of substance lost 0.0907 gram at 150° C. Calculated for Found: CsHiON, . H.0: I Ili III HOM eee alae fc. 11.68 11.70 11.57 11.64 i Re oe, a 36.36 36.60 0.2144 gram of anhydrous substance gave 0.0593 gram of H2O and 0.3475. gram of COs. Caleulated for CsH4ONa: Found: Cae Parenloe Fo eo HOES Te SOG Te ee oe ee F 44.11 44.20 A eee cen oe ss 'e's cra Bye 0 2 eee eee ROME 2.94 3.07 ee ee MME os o,0 shoes ne Solemn oateaereeits 41.17 41.28 The properties of 2-oxypurine agreed in all respects with the description of the purine which Tafel and Ach* prepared from guanine. Salts. The Hydrochloride. CsHsONy.2HCl. One-half gram of anhy- drous 2-oxypurine was dissolved in 10 ce. of hot 20 per cent hydrochloric acid. On cooling rapidly the hydrochloride separated as slender prisms, but when the solution was cooled slowly rectangu- lar plates were obtained. The precipitate weighed 0.3 gram. 0.1203 gram of substance gave 0.1641 gram of AgCl. Calculated for CsH,ON,«. 2HCI: Found: Clete nn eS! oS AR, Peer ee 33.97 33.73 The Nitric Acid Salt. Cs3H,ON4s.2HNO3. One-half gram of the anhydrous 2-oxypurine was dissolved in 10 ec. of warm 20 per cent nitric acid. Clusters of slender prisms separated rapidly on cooling. The yield was 0.5 gram. Calculated for Found: CsHsONs .2HNOs:: I II SINE year eet soi csal's sree ee er oteiene 32.06 32.10 32.06 The Picrate. CsHsON,z.CeH2(NO2)30H. This was made by adding a cold saturated solution of picric acid to a hot solution of the purine. On cooling, hexagonal and lenticular shaped prisms were obtained. These were easily soluble in hot and difficultly 8 Tafel and Ach: Loc. cit. Carl O. Johns 71 soluble in cold water. They turned brown when heated above 210° C. and effervesced at 245° C. Calculated for CsH:ON, . CeH2(NO2),0H: Found: IMs o Sides Bet eae ae rete PARIS 26.85 26.82 Acetyl-2-oxy-5 ,6-diaminopyrimidine. CesHsO2Ny. Six grams of 2-oxy-5,6-diaminopyrimidine were dissolved in 60 cc. of acetic anhydride by heating at 140° C. in an oil bath. This solution was then evaporated to dryness on the steambath. A little dilute ammonia was added to neutralize the last traces of acetic acid. After evaporating to dryness the residue was washed with a little water. The yield was greater than the calculated for monoacetyl- 2-oxy-5, 6-diaminopyrimidine, a mixture of the mono- and diacetyl compounds having formed. This mixture was moderately solu- uble in hot water from which it crystallized in slender prisms with square ends. The calculated per cent of nitrogen for a monoacetyl compound is 33.33, for a diacetyl 26.66. The mixture contained 31.67 per cent of nitrogen. N= aa 2-Oxy-8-methylpurine, Sos Ne C.CH; ae 7 NH——C—N Three and one-half grams of acetyl-2-oxy-5,6-diaminopyrimidine were dissolved in 8 cc. of water containing 2.5 grams of potassium hydroxide. About 200 cc. of absolute alcohol were added. Fin- ally ether was added until the precipitation was complete. The potassium salt deposited as a thick oil which solidified on standing. The salt thus obtained was heated in an oil bath at 240° C. It melted partially and foaming ensued as water was liberated. The heating was discontinued when steam ceased to escape. The reaction product was a brittle crust. This was dissolved in cold water and the solution was acidified with acetic acid and evapor- ated to dryness. The potassium acetate was removed by washing with a little cold water. The yield of the crude purine was 90 per cent of the calculated. The yield varied widely in several experi- ments, the variation being probably due ‘to the proportion of di- acetyldiamino pyrimidine present. The purine was purified by 72 Researches on Purines dissolving in dilute ammonia and clarifying with blood coal. The filtrate was boiled to remove most of the ammonia whereupon it was acidified with acetic acid. The 2-oxy-8-methylpurine separated slowly from the solution in the form of small slender prisms with tapering ends. These were soluble in about 40 parts of boiling water and slightly soluble in cold water and 95 per cent alcohol. They turned brown at 285° C. but did not decompose completely below 310° C. 0.1780 gram of substance gave 0.0671 gram of HO and 0.3123 gram of CO» Calculated for CeHsONa: Found: I II III 1S PARAS ASS 6. 4'5.q:0 cin REE 4.00 4.18 CRA A RG opias a 00. coe nee ite 48 .00 47 .87 DS Ee Cie Coy Gln ae 37.33 37.39 37.32 37.32 SALTS. The Nitric Acid Salt. CsHpONs. HNO3. One-half gram of 2-oxy- 8-methylpurine was dissolved in 3 cc. of 80 per cent nitric acid by warming gently. On cooling, the salt separated rapidly in minute lenticular crystals. When the solution was cooled slowly the crystals were more compact and had truncated ends. When heated rapidly in a capillary tube the salt began to turn dark at about 170° C. and decomposed suddenly at 205° C. with enough force to throw the substance out of the tube. Calculated for CsHsONs.HNO3: Found: IN soe aera erele isis «ons» 9 aidate(sraiclanel sr oe ko eategs 32.86 32.94 The Picrate. A cold saturated solution of picric acid was added to a hot solution of the purine. Sheaf-like clusters of slender prisms were deposited as the solution cooled. These were moder- ately soluble in hot water. They began to turn dark at about 210° C. and decomposed with violent effervescence at 250° C. Caleulated for CeHeONs . CeH2(NO2)20H: Found: ING Sieeersetiaeke fis es 2 os Sa astaakeiete 25.85 26.02 RESEARCHES ON PURINES. ON 2-OXY-1-METHYLPURINE. FIFTH PAPER.! By CARL O. JOHNS. (From the Sheffield Laboratory of Yale University.) (Received for publication, December 14, 1911.) Five of the six isomers of 2-oxymonomethylpurine have been described. The first of these was obtained by Emil Fischer who made 2-oxy-7-methylpurine? (X) from 2-iodo-7-methylpurine. The same purine was also prepared by Tafel and Weinschenk.* The latter workers also prepared 2-oxy-3-methylpurine! (VIII). In all of the above cases the starting material was a purine. The remaining isomers have been synthesized from pyrimidines. 2-Oxy-6-methylpurine® (IX) was prepared from 2-oxy-4-methyl- 5,6-diaminopyrimidine. 2-Oxy-9-methylpurine® (XII) was _pre- pared from 2-oxy-6-methylamino-5-aminopyrimidine, while 2-oxy- 8-methylpurine’ (XI) was made from 2-oxy-5,6-diaminopyrimidine. The sixth isomer of this series, 2-oxy-l-methylpurine (VII) has now been synthesized from 2-oxy-3-methyl-5,6-diaminopyri- midine (IV). The reactions involved in this synthesis are as follows. The potassium salt of nitrocytosine, 2-oxy-5-nitro-6- aminopyrimidine® (II) was methylated by the means of methyl iodide. The yield of a monomethyl derivative was 70 per cent of the calculated quantity. Three different monomethyl deriva- 1 This Journal, xi, p. 67, 1912. 2 Ber. d. deutsch. chem. Gesellsch., xxxi, p. 2854, 1898. 3 [bid., xxxiii, p. 3376, 1900. A,[bid:, P.id3i 2. 5 Johns: Amer. Chem. Journ., xli, p. 65, 1909. 6 Johns: This Journal, ix, p. 161, 1911. 7 Johns: This Journal, xi, p. 67, 1912. 8 Wheeler and Johnson: Amer. Chem. Journ., xxxi, p. 591, 1905; Johns: Ibid., xlv, p. 81, 1911. 73 74 Researches on Purines N == (Ne N—CNH, CH;-N—CNH, | | il | CH; N—CH ~ HN—CH NCH I II jig 8 ae NX N=CNH, aie es | OC CNH, OG ~CNO; OC, CNO; Foes! | || CH;-N—CH CH; N—CH HN—CH IV Vv VI ') CH;;: N—CH NCH N=C:-CH; OC C—NH oc C—NH a C—NH m= [ioe tl CH CH | | Wa Vi | Vi N=C—N CH;- N—C—N HN—C—N VII VIII IX NCH Meg NCH OC C—N-CH; OC. C—NH oc C—N ioe tiem TTD CH C:CH CH | latent A Vi 5 Ma HN—C—N HN—C—N HN—C—N‘°CH; x XI MAT tives are possible in this reaction, namely, 1-methylnitrocytosine (III), 3-methylnitrocytosine (1), and 6-methylnitrocytosine (V1). To determine which one of these was formed the reaction product was heated with sulphuric acid in a sealed tube. This treatment removed an amino group, giving a methylnitrouracil melting at 255° C. and containing one molecule of water of crystallization. Carl O. Johns 7 Hence formula VI was excluded. The two isomers of methyl- nitrouracil have been investigated by Behrend and Thurm.°® 1-Methylnitrouracil crystallizes without water and melts at 263° C. while 3-methylnitrouracil (V) crystallizes with one molecule of ‘water and melts at 255°C. Our compound was identical with the latter and hence our methylated product was 2-oxy-3-methyl- 5-nitro-6-aminopyrimidine (I), or 3-methylnitrocytosine. When this compound was reduced with freshly precipitated ferrous hydroxide it gave a good yield of 2-oxy-3-methyl-5,6-diamino- pyrimidine (IV), which, in turn, reacted with formic acid to give a formyl derivative whose potassium salt, when heated, lost water and formed the potassium salt of 2-oxy-l-methylpurine (VII). 2-Oxy-1-methylpurine crystallizes beautifully from water in flat prisms and these contain 2 molecules of water of crystallization. They effloresce in the air and become anhydrous over sulphuric acid. An aqueous solution of the purine gives difficultly soluble precipitates with platinic chloride and picric acid. The picrate decomposes at 214° C. Work on the preparation of other purines from 2-oxy-3-methyl- 5,6-diaminopyrimidine is in progress. EXPERIMENTAL PART. 2-Oxy-3-methyl-5-nitro-6-aminopyrimidine. N=CNH, OC CNO, CH;°'N—CH This is the chief product where nitrocytosine” is methylated as follows: Five grams of nitrocytosine, 2-oxy-5-nitro-6-amino- pyrimidine, were dissolved in 50 ce. of water containing 2 grams of potassium hydroxide. Five grams of methyl iodide were added and the mixture was heated in a sealed tube at 100° C. for one hour. On cooling, the methylated product separated in slender prisms. The yield was 64 to 74 per cent of the calculated. The 9 Ann. d. Chem. (Liebig), cccxxiii, p. 163, 1902. MOMGOC Cit. 76 Researches on Purines compound was moderately soluble in hot and difficultly soluble in cold water, and difficultly soluble in hot alcohol or hot glacial acetic acid. It erystallized from water in beautiful, slender prisms which in some cases were about | to 2 mm. thick and 3 em. long. These began to turn brown at 260° C. and melted with decomposition at 274° C. Calculated for CsHsO3Ns: Found: Nis ice Ss SS SPR Sth) ols. yh ga Semyrs erties < 32.93 32.72 2,6-Dioxry-3-methyl-5-nitropyrimdine. HN—CO OC CNO, | CH,-N—CH The position of the methyl group in the pyrimidine obtained by methylating nitrocytosine was ascertained as follows. One and five hundredths grams of the methylated product were heated with 18 cc. of 25 per cent sulphuric acid in a sealed tube at 140 to 150° C. for one and one-half hours. On cooling, 0.9 gram of crystals deposited from the acid solution. This substance was recrystallized from water and obtained in the form of slender prisms. These prisms melted at 255° C. and contained one mole- cule of water and in all other respects agreed with the properties of 3-methylnitrouracil as described by Lehman" and by Behrend and Thurm.2 In this experiment the ammonia produced by heating the methyl! derivative in the sealed tube was determined by making the acid solution alkaline and distilling the ammonia into 7 acid, 62 cc. being required. Calculated for the loss of NH» in CsH6O3Ns: Found: 25g. Bie ee, Se 8 IAS EN 8.23 8.27 Analyses of the crystals obtained from the sealed tube gave the following results: 0.7303 gram of substance lost 0.0710 gram of HO at 120 to 130° C. 11 Ann. d. Chem. (Liebig), 253, p. 77, 1899. 12 Loc. cit. Carl O. Johns 77 Calculated for C;HsO4Ns.H20: Found: TEI AO) 8 pS i SS nes i ee ee 9.53 9.72 Caleulated for CsHsO4Ns: Found: IN Pe i ic nc cin b's bis eee SEE Se Les dss 24.56 24.83 2-Oxy-3-methyl-5,6-diaminopyrimidine. N == CNH, OC CNH, | CH;:N —CH Ten grams of 2-oxy-3-methyl-5-nitro-6-aminopyrimidine were dissolved in a mixture of 200 cc. of concentrated ammonia and 100 ee. of water. To this solution was added a warm, almost saturated, aqueous solution of 120 grams of crystallized ferrous sulphate. Reduction took place rapidly with the liberation of heat. The sulphate was precipitated by the addition of 140 grams of crystallized barium hydroxide dissolved in hot water. After shaking thoroughly the excess of barium hydroxide was removed by the means of carbon dioxide. The reaction was allowed to proceed over night whereupon the precipitate was filtered off and washed with hot water. The filtrate was concentrated to about 30 cc. The diamino compound crystallized in small, stout, anhydrous prisms. These were easily soluble in hot and moder- ately soluble in cold water and almost insoluble in alcohol. They began to turn dark at about 220° C. and decomposed slowly when heated above that temperature. The yield of the substance isolated was 82 per cent of the calculated, exclusive of a small quantity that remained in the mother liquor after further con- centration. Caleulated for CsHsONs: Found: Joo) ee Ds 7 eres 40.00 39.77 Formyl-2-oxy-3-methyl-5,6-diaminopyrimidine, CsHsQ.Ns. Eight grams of 2-oxy-3-methyl-5,6-diaminopyrimidine were dis- solved in 20 ce. of 85 per cent formic acid and the solution was evaporated to dryness on the steam-bath. The resulting residue was taken up in water, the solution was filtered to remove a little 78 Researches on Purines insoluble substance, the filtrate was made slightly alkaline with ammonia and evaporated to dryness. The yield of formyl com- pound was almost quantitative. The compound was very soluble in hot and easily soluble in cold water. From a concentrated aque- ous solution, it crystallized in masses of colorless, slender, dis- torted prisms. Caculated for CeHsO2Nq: Found: Ny cscs Ses. . TA ee one aaa 33.33 33.30 2-Oxy-1-methylpurine. CH; N—CH | | | OC C—NH ~ | | Ys N=C—N The potassium salt of formyl-1-methyl-2-oxy-5,6-diaminopyri- midine was made by dissolving 4.5 grams of the formyl com- pound in 8 ce. of water containing 3 grams of potassium hydroxide. Two hundred cubic centimeters of absolute alcohol were added but the salt did not precipitate. Finally ether was added, grad- ually, until the solution became turbid. On stirring, the salt began to crystallize. More ether was then added until crystal- lization was complete. Five grams of salt, dried at 80° C., were obtained. This salt was heated in an oil-bath at 160° C. until water ceased to escape. A brittle crust remained. This was dis- solved in water, the solution was neutralized with acetic acid and clarified with blood coal. On concentrating to about 15 cc. the purine began to crystallize from the hot solution. The presence of potassium acetate appeared to render the purine easily soluble. It crystallized slowly in small, flat prisms. These contained two molecules of water and effloresced in the air. When dried over sulphuric acid for two days they became anhydrous. The anhy- drous substance dissolved in about eight parts of boiling water. It was slightly soluble in hot alcohol and easily soluble in hot glacial acetic acid. From the latter it crystallized in small stout prisms. The aqueous solution gives precipitates with silver nitrate Carl O. Johns 79 and platinic chloride. The anhydrous purine decomposed slowly without melting when heated above 280° C. The portion used for analysis was recrystallized from water and the crystals were dried for two hours on filter paper. 0.6950 gram lost 0.1350 gram of H.O at 130 to 140° C. 0.7359 gram lost 0.1421 gram of H2O at 130 to 140° C. Calculated for CsHsON4.2H20: - Found: it ato re 19.34 19.42 19.29 Nin Soe. ere 30.10 30.18 0.2178 gram of anhydrous substance gave 0.0812 gram of H.O and 0.3822 gram of COQ:. Calculated for CeHsONa: Found: OPT ices ore ce ove ee caine eo meiooee 48 .00 47.75 lol. oo ooo. o Sa Oe ee eco Seen: 4.00 4.14 IS oon 3 biel o.crnettho.eig tae RE ae eke eee 37.33 31.23 The Picrate, CsH,ON,.CgH2(NO.);0H. A cold saturated solution of picric acid was added to a hot solu- tion of the purine. On cooling, clusters of small, slender prisms deposited. These were moderately soluble in hot and difficultly soluble in cold water. They melted with decomposition at 214° C. Calculated for Ci2HsO8N7: Found: 25.85 25.88 OM ee ‘ i bas, : Lig ry ; ee j dati. 4 * theory oa eile nell =o amet gated! or qngae ad ee a par: qutarvi: incall ohise \(Bihicikana aan oe gre tpn 5 ee! aed a ate 4-40 i ties Giksw Oc6t.0 nae _ st Seat &? OEE fa, Coe. ie are Tah odaahiny i ins é Pata, ik biLet, Se ee F "Oe ; F ae. 1 f re OG a, aa RF 2 tina Riv ety re a ai ’ 7 : a Cone.) owt ‘OM agp Ciel. ve dioeiatadsie, vi melita i) ie a oe Bitz cal wnere ‘capa eae — vest 4. each gattonda toby pay “hale en eo r 5 cal Wega aee SAE 8 : > es CH;CO-N-CH,COOH — : | BEEN CS:NH, III NH—CO NH—CO Fol | esr Sa OS | | CH,CO-N——CH, NH—CH, IV V In other words, the product of the reaction is 2-thio-3-acetyl- hydantoin (IV), which can be converted quantitatively into the 2-thiohydantoin (V), by hydrolysis with concentrated hydro- chloric acid. This method of synthesizing 2-thiohydantoin was also described by Komatsu,” but his interpretation of the mechan- ism of the reaction was entirely incorrect. Komatsu also exam- 1 Journ. Amer. Chem. Soc., xxxiii, p. 1974. 2 Memoirs Coll. Sci. and Eng., Kyoto University, (Japan), iil, p. 1. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 2. 97 98 Hydantoins ined the behavior of potassium thiocyanate towards alanine and states that this amino-acid reacts with the rhodanide in the presence of acetic anhydride, giving the corresponding thiohy- dantoic acid (VI), which is converted into 2-thio-4-methylhydan- toin (VII), by digestion with hydrochloric acid. NH, COOH NH—CO | | CS (CHS | NH—CHCH; NH—CHCH; VI : VII It seemed very probable to the writer that glycocoll and alanine would react in a similar manner with potassium thiocyanate, under the conditions employed by Komatsu, giving the corre- sponding acetylated thiohydantoins. However, it was not im- probable that these two acids might behave in a different. manner and that in the case of alanine Komatsu actually was dealing with a thiohydantoic acid derivative. The analytical results obtained by analysis of the barium salt of his acid, however, do not offer strong evidence that Komatsu’s compound possessed the struc- ture which he assigned to it. He found 29.6, 29.5 and 30.3 per cents of barium and concluded therefore that he was dealing with a hydrous barium salt, which theoretically contains 30.73 per cent of barium. He did not determine the percentage of water of crys- tallization. I now find that alanine reacts smoothly with potassium thio- cyanate, in the presence of acetic anhydride, forming 2-thio-3- acetyl-4-methylhydantoin (VIII). In fact, I obtained no evi- dence of the formation of a thiohydantoic acid (VI)-as described by Komatsu. The constitution of my acetyl compound was estab- lished not only by the analytical determinations, but also by the fact that the same compound was formed when. acetylalanine® was used for the synthesis instead of alanine. When the acetyl- thiohydantoin (VIII) was digested with hydrochloric acid it was converted quantitatively into 2-thio-4-methylhydantoin VII. This hydantoin has previously been prepared in this laboratory.‘ ’ Fischer and Otto: Berichte, xxxvi, p. 2114. 4 Wheeler, Nicolet and Johnson: Amer. Chem. Journ., xlvi, p. 456. Treat’ B. Johnson 99 by a different method. Komatsu® showed that this thiohydan- tion .can be desulphurized by mercury oxide giving 4-methyl- hydantoin. Alanine therefore reacts with potassium thiocyanate in a similar manner as glycocoll and the chemical changes involved are to be represented as follows: NH.CH(CH;)COOH + (CH;CO),0 = CH;CONHCH(CH;)COOH — CH,CONHCH(CH,)COOH ———> CH;CONCH(CH;)COOH —> eo SCN CSNH, NH—CO NH—CO | | Cs | rd ers) | CH;CO:-N——CHCH; NH—CHCH; VIII VII These nitrogen-unsubstituted 2-thiohydantoins are represen- tatives of a new class of hydantoins. A knowledge of their chem- ical properties is especially desirable since it is probable that such cyclic groupings may be involved in the molecular structure of sulphur proteins. Like the thiopolypeptides they contain the thioamide grouping, —-CS-NH—, and as the writer has previously indicated,® thioamides probably functionate in the natural synthe- sis of sulphur proteins from simpler substances. We shall continue our investigation of this interesting class of sulphur compounds and will discuss in future publications their biological significance. EXPERIMENTAL PART. The Action of Potassium Thiocyanate on Alanine, NH2CH(CHs)- COOH: 2-Thio-3-acetyl-4-methylhydantoin. NH—CO | CS CH;CO:N——CHCH; SLi bihees(eip § This Journal, ix, p. 331. 100 Hydantoins This thiohydantoin can be prepared easily in the following man- ner: Dissolve 2 grams of alanine and 2 grams of finely pulverized, cry potassium thiocyanate in a mixture of 9 cc. of Kahlbaum’s acetic anhydride and 1 cc. of glacial acetic acid by warming in a water-bath. Connect the flask with a return condenser in order to avoid the absorption of moisture from the air. On warming, there is an immediate reaction and within two minutes a clear yellow solution is obtained. The liquid is then heated, at 100°, for about twenty-five minutes, cooled and then poured into about five volumes of cold water when the greater proportion of the above acetyl hydantoin will separate in a crystalline condition. The yield is about 2.5 to 2.7 grams and in every experiment tried the sub- stance melted, without further purification, at 163 to 165° to a clear oil. The hydantoin crystallizes from 95 per cent alcohol in stout prisms, which melt at 166° to an oil. On cooling, the oil solidified in the capillary tube and on heating again it melted at 166° as before. ANALysiIs: Sulphur determination (Carius): 0.1377 gram substance gave 0.1939 gram BaSOx,. Nitrogen determination (Kjeldahl): Calculated for CsHsO2N28: Found: IN aa RE, so coy See Red. LAs ae teh 16.27 16.33 Stith. cts. ..s"sj,acg eee Wiese areas ae 18.8 19.30 Hydrolysis of the Acetylthtohydantoin with Hydrochloric Acid. 2-T hio-4-methylhydantoin. NH—CO CS NH—CHCH; The acetyl] derivative was suspended in about ten to fifteen parts of concentrated hydrochloric acid and the mixture warmed on the steam-bath. The hydantoin completely dissolved and after evap- oration of the acid practically a quantitative yield of this hydan- toin was obtained. The substance was purified for analysis by recrystallization from hot 95 per cent alcohol. It separated, on cooling, in beautiful, hexagonal tables or plates which melted at 161° to a clear oil. Treat B. Johnson IOI ANALYSIS (Kjeldahl): Caleulated for H;ON28: Found: Nac ce docs 6 Go re ot ae eee 21.53 PALE The hydantoin was identical with the hydantoin obtained from alanine by Wheeler, Nicolet and Johnson.’ The Formation of 2-Thio-3-acetyl-4-methylhydantoin from Acetyl- alanine, CH;CO NHCH (CH;3) COOH :% The acetylalanine was warmed with the required proportion of potassium thiocyanate under the same conditions as when ‘alanine was used. After heating one-half hour to complete the reaction, and finally cooling, the liquid was then poured into cold water. A yellow solid separated at once and after purification by crystallization from alcohol it melted at 165 to 166°. A mix- ture of this substance with the above acetyl hydantoin, prepared from alanine, melted at exactly the same temperature. 7 Loc. cit. 8 Fischer and Otto: Loc. cit. 7 = oe iene nate et ‘eam aN ‘a agit, a im) aay ne ih oe ow soc sida ROS TLY ma RRs oat Bi © ay. Ce bys Bere & c ighhaell jaeeg OE: Ae Bac SSE gt = bara. sotoiad | } ee ‘ inyt sigstue Bie scalar el bina ies i 1) TAI. ¥ Ase ) st - ed 4) s t ty ; * ¢ 3 ‘ a a i, ea 2 é. ime 2 pitt Behe it gedinan: oat bag us uy 15, satel Th gaits et a sly dimwlved : bealty's a’ ' vial : ie =i . vith oft Ea) Fs ie ints any eiciyrern:: i ae eas - as -_ ies ee - hah 1 GIS nu ae ‘at Res ae ae SSORIIERES - 4 ie ap \" aR Zz fe a tive 3 f annie Pats, eae KROHN rT we aot bs its, “Bile ; sek Be abled andthe parirete by May GH pt nainalalé iv pais aes seca qd Laks abana ri 2g itty en hid mt ai Su ‘kh Been Fy pie welistataty ee ndotiay inggIONK. notre ; fil sleiye Rp aptadey f . arity tions as Dptioat osinals @ = fences: > Sa) a SA ASLT OT) SER 33 $ > p! > ta sie epathet 7 ay Saas a 7 =P aoe = arse > Orr pS ; ur} , bigs ‘ia ‘ ; ‘iY is ? a * = LJ > - io - i+ ea: et Poa >» io i - NOC laf] in er iw wa oa thirnlxiu ure Te e fs yy id phi rat a < TAD i FASTING STUDIES: VI. DISTRIBUTION OF NITROGEN DURING A FAST OF ONE HUNDRED AND SEVENTEEN DAYS. By PAUL E. HOWE, H. A. MATTILL ann P. B. HAWK. (From the Laboratory of Physiological Chemistry of the University of Illinois.) (Received for publication, January 13, 1912.) If the literature of fasting be searched it will be found that the records indicate that adult dogs may live for periods ranging from thirty days to fifty days without partaking of food provided free access to water is permitted. At the tine the data embraced in the present paper were reported! the longest normal fast on record so far as we are aware, was that reported by Falck? in which the dog used as subject fasted sixty days. The ninety-eight-day fast reported by Kumagawa and Miura? cannot be considered as a normal fast, inasmuch as the animal was subjected to the influence of phlorhizin. The reports of fasting tests in which human beings have served as subjects afford data on authentic fasts ranging in length from two to fifty days. The most complete data on short fasts have been furnished by Benedict.4 Of the longer fasts those on Beauté,* 1 Howe, Mattill and Hawk: Boston Meeting, Soc. Biol. Chem., Dec., 1909; Proceedings Soc. Biol. Chem., July, 1910. * Falck: Bettr. Physiol., (Stuttgart), 1875. Quoted by Pashutin in Path- ological Physiology, 1902. 3? Kumagawa and Miura: Arch. f. Physiol. u. Anat. (physiol. Abt.), p. 431, 1898. 4 Benedict: Carnegie Pub. 77 (1907). 5 Cathcart: Biochem. Zeitschr., vi, p. 199, 1907. 103 104 Elimination of Nitrogen Durinz Fasting Tosea,® Schenck,’ Succi,® Cetti,? Breithaupt,!® “E’’ and “H,’’# Tanner,” and Merlatti® are the most important. The longest of these are the thirty-day fasts of Succi, the forty-day fast of Tan- ner and the fifty-day fast of Merlatti. DESCRIPTION, PLAN, ETC. The methods of analysis employed in our investigation were the same as those used in fasting studies already reported from this laboratory.“ In conformity with the custom in this laboratory in experiments of this character the dog used as subject was not catheterized but was allowed to urinate at will. This explains to a degree the cause of the irregularity in the urine volumes which obtains in the early part of the experiment, notwithstanding the fact that the urine was measured at a uniform time from day to day. We prefer this mode of urine collection to the catheterization procedure because of the attendant danger of infecting the animal during the latter process. The diet used in the preliminary period of the experiment was as follows: Constituent. Grams. Meats eiieaas 2.0 2 RA k. SU RE Beek na ee oe oe 400 Grackeriduste 325 22 Gide So IAS 4a ie oer ee dee 100 TEE Gl eee se 25. we ots, Len ay ie etre eae aM ae a a ea ee 45 Bone, asleep oe |e ae oe os ee 12 Te en ee A. a ne a AR Oe an Ben tT i I eS a 700 § Van Hoogenhuyze and Verploegh: Zeitschr.f. physiol. Chem., xlvi, p. 415, 1905-06. 7 Brugsch and Hirsch: Zeitschr. f. exp. Path. u. Therap., iii, p. 638, 1906. 8 Luciani: Das Hungern, Leipzig, 1890; Ajello and Solaro: La riforma medica, ix, 2, p. 542, 1893; E.and O. Freund: Wiener klin. Rundschau, xv, pp. 69 and 91, 1901. 9 Lehman, Miller, Munk, Senator, Zuntz: Virchow’s Archiv, cxxxi, supp}., 1893. 10 Td.; ibid. 11 Howe, Mattill and Hawk: Jour. Amer. Chem. Soc., xxxiii, p. 568, 1911. 12 Lusk: Science of Nutrition, 2d Ed., p. 55, 1909. 13 Merlatti: Luciani’s Das Hungern, 1890. 14 Howe and Hawk: Jour. Amer. Chem. Soc., xxxiii, p. 215, 1911. Paul E. Howe, H. A. Mattill and P. B. Hawk 105 The above diet contained 15.796 grams: of nitrogen. Approx- imate nitrogen equilibrium was secured after an eight-day feeding interval. The balance for this period was as follows: Income. Grams GO GREER IRR R ery sinus. ora Sart eek ARR stich « 15.796 Outgo IEC OS IRIN RM e ec ocr. at SSE MEME UEEATS fon oc erases 0.372 ISLES 2 Ui jiend techie nee ge rt errata Ee OS oe er 2 Re ere 0.359 Wir cenwas ln ESheiaess Choe. Fo) SEE Woe abe ea UNE cide 0.147 Olirh aXe se SS ic cha ea es het te aoe ee eR aa a a reat ey Ot 15.588 —16. 466 +15.796 = 0.670 The subject of the experiment was our fasting dog “Oscar,” an adult Scotch collie weighing 26.33 kg. at the opening of the fast. Inasmuch as his preliminary diet contained 15.796 granis of nitro- gen per day he was receiving about 3.75 grams of protein per kilo- gram of body weight during the preliminary period. It was our intent at the start of the investigation to fast the ani- mal to the pre-mortal rise in nitrogen excretion, then to bring the dog back to the normal condition by means of careful feeding and subsequently to fast hima second time. In other words we wished to make a study of ‘‘repeated fasting”’ similar to the one already reported from this laboratory.” DISCUSSION OF RESULTS. It was evident from the very beginning of the fasting period that “Oscar” was not being influenced by the fasting régime in as pro- nounced a manner as were other fasting dogs in adjoining cages. There was a less rapid loss of weight, a less precipitate destruction of body tissues as shown by the nitrogen output and a conserva- tion of bodily vigor and energy not noted in the case of any of the other dogs. As time passed each of the associated dogs in succes- sion reached the “‘pre-mortal rise” in nitrogen excretion. At the 15 Howe and Hawk: Loc. cit. 106 Elimination of Nitrogen During Fasting forty-eighth day, when the experiment upon the initial fast of the last dog terminated, ‘‘Oscar’’ was so full of vigor that he jumped into his cage from the floor. In performing this act it was neces- sary for him to project his body upward to a height of about three feet. He had been in the habit of jumping into his cage each day after being weighed but the practice was discontinued after the fifty-eighth fasting day in order to protect the dog from pos- sible injury due to coming in contact with the sharp corners of the cage front. He continued to jump out of his cage up to and in- cluding the one hundred and first day of the fast. It was appar- ently quite a task for him at this time in his weakened condition to maintain his equilibrium after leaping from his cage to the floor. _ In order to avoid injury, he was, therefore, not permitted to per- form this feat after the one hundred and first fasting day. The animal, however, continued to wag his tail vigorously and fre- quently barked when we approached his cage, all of which seemed to indicate that he was in “‘good spirits” even up to the very end of the experiment. It is an interesting fact that one of our dogs was fasted to the pre-mortal rise, then subjected to an intermediate equilibrium feeding period during which time the dog regained its original body weight and then fasted a second time to the pre- mortal rise while “Oscar” was undergoing his initial fast and jump- ing in and out of his cage daily, and-no sign of the pre-mortal rise being apparent. When the one hundred and seventeenth day was reached it was decided to terminate the fast. The dog now weighed 9.76 kg. as against 26.33 kg. at the opening of the fast, a loss of about 63 per cent in body weight. It was then June 2 and the animal had been fasting continuously since February 6. At the commence- ment of the experiment we had expected to be able to finish two fasts and the intermediate feeding period in the four months which had passed. The great length of the first fast had nullified this arrangement. It was therefore decided to initiate the second fast at the opening of the next college year. During the summer the dog passed the time on a Kansas farm under close observation. He was brought back in the fall and upon examination was found to weigh somewhat more than he did at the commencement of his first fast. He also seemed to be stronger, more energetic and in better all round physical condition Gener be Paul E. Howe, H. A. Mattill and P. B. Hawk 107 than he had been before he was subjected to the one hundred and seventeen-day fast. After being brought into nitrogen equi- librium he was then subjected to a second fast. The data from this “repeated fast”? will be presented in a subsequent paper.! Distribution of Nitrogen. The general data for each individual day of the one hundred and twenty-five days of the experiment are given in Table I, pp. 108-111. To facilitate discussion’ and comparison the data have been placed in Table II, p. 112 in the form of four-day averages. The course of the excretion of various forms of nitrogen has also been represented in graphic form in Fig. I, p. 115. The per- centage values are given in-Table III, p. 113. Totat NitroGen. If we examine Table II, p. 112, it will be seen that the average daily output of nitrogen during the eight- day feeding period was 15.588 grams. This value was lowered to 6.231 grams for the first four-day fasting interval whereas the three succeeding four-day periods showed progressively decreas- ing nitrogen values, the figures being 4.471 grams, 4.028 grams and 3.216 grams respectively. From this point the output of nitrogen fluctuated irregularly until the twenty-first period, or eighty-first day of the fast at which time a more uniform level was assumed and fairly well maintained throughout the remainder of the fast. The slight rise in the nitrogen excretion upon the last day of the fast cannot be considered as the beginning of the pre-mortal rise. In the first place the pre-mortal rise is invariably preceded by a slight decrease in the nitrogen excretion, a condition not observed in this experiment. A much more potent argument against con- sidering the slight increase in the nitrogen of the final fasting day as an indication that the pre-mortal rise had been established is found in the fact that at no time had the daily output of creatine- nitrogen exceeded that of creatinine-nitrogen. In all the fasting studies made in this laboratory, where the animals have fasted to the pre-mortal rise, we have noted in every instance that the out- put of creatine increases during the final stages of the fast and finally a few days before the fall in the nitrogen output which 16 Reported before American Physiological Society, Baltimore, December, 1911. PURINE N | grams 0.050 0.056 0.069 108 Elimination of Nitrogen During Fasting TABLE I. General Data. & a Zz : 3 8 » A 3 a Bx z iS ga os Zz Fs z > of 2Z | &% 2 z = a sa ~ bE 3 x g 2 < EA a =) om & = = ro) <3 ° | oP A S) a 2 a 2 > n & < oO o Preliminary Period—700 cc. water per day. | | kgs. ce. | grams | grams | grams | grams | grams 1 474 | 10135 | 10.504} 8.795| 0.479| 0.271 | 2 963 | 1026 | 17.354 | 14.908 | 0.608| 0.421] 0.470 3 | 448| 1018 | 6.050} 5.165} 0.390] 0.187 4 | 1330 | 1026 | 24.770 | 21.477 | 0.948 | 0.666] 0.484 5 26.36 400 | 1030 | 9.900! 7.881) 0.341} 0.193 | 0.236 6 | 26.36! 360! 1031 | 8.920! 7.698| 0.330) 0.311 | 0.108 7 1028 | 23.192 | 20.479 | 1.000| 0.622} 0.331 8 1028 | 24.136 21.070 | 0.986 | 0.608 | 0.430 ea = ee 13.434 | 0.635 1 2 3 4 5 6 7 8 9 10 11 22 12 29 13 22 14 | 22 15:9 22 16 17 22. 18 21 19 21 20 21 21 21 22 21 23 Fasting Period—700 cc. 219 | 10235 520 | 1008 - 980 | 10055 | 637 | 1003 318! 1011 660 | 10095 a 10065 790 | 1008 350 | 1003 422 | 1007 3.320 4.099 1.657 2.432 4.678 3.444 5.966 1.269 2.970 4.284 3.477 2.672 3.179 1.341 1.934 3.722 2.805 4.860 1.075 2.407 water 0.257 | } 0.018 0.010 0.006 0.013 9.020 0.023 | 0.004 | | J ALLANTOIN N grams 0.031 0.066 0.025 0.082 0.014 0.007 0.100 0.025 0.008 0.003 0.019 DAY OF EXPERIMENT Paul E. Howe, H. A. Mattill and P. B. Hawk 109 BODY WEIGHT VOLUME OF URINE TABLE I—(Continued). GRAVITY SPECIFIC ToTaL N UREA N AMMONIA N CREATININE N Ze Q z ‘2 & < Q om is) PURINE N Fasting Period—700 cc. water per day—Continued> kgs. grams 5.817 - co | ao a> e oo I to or wo _ ow a I o oO te 1) Ld oo ~~ ~1 oo o Co oO ao v=) = [es] a a ~1 (=) w ~ —_ Go 3 | oo > to ~ a Se Oe EE eee eee Oo oe oe OO grams 4.738 6.768 3.355 5.165 4.439 5.978 3.917 4.905 4.072 5.357 5.917 5.668 grams 0.347 0.587 0.258 0.390 0.399 0.453 grams 0.389 0.557 0.274 0.393 0.395 0.507 0.310 0.389 0.139 0.388 0.420 0.356 0.291 0.396 ALLANTOIN N grams ee 0.00 | 0.00 0.00 0.00 | | | | t 0.00 0.00 0.00 0.00 grams grams 0.066 0.065 | 0.469 | 0.223) 0.034] 0.005 0.036 0.362 | 0.174! 0.071 | 0.004 0.024 0.391 | 0.142 | 0.032 | 0.002 0.036 0.284 | 0.138 |} 0.045 | 0.002 0.028 700 cc. water per day. r = 0.118 | 0.032 0.016 0.159 | 0.022 0.018 0.107 | 0.021 0.010 110 ~— Elimination of Nitrogen During Fasting TABLE I—(Continued). | a a | Z Z = = 5 “ oa 2 a 4 % ae | = eee ee | Z, < Z z a 5 Se | Se jememi ee | a < E B z ¢ saSA ~ 2 ™ og *« go = « < - < 28 5 | 52 | Bo 5 E = a 2 5 3 eee vik. eae > 2 || & » E < 3) & fu < 700 cc. water per day—Continued. kgs. IPP ica: grams | grams | grams Pa grams | grams | grams 67 1444 625 | 10045 | 3.566 | 2.862] 0.329] 0.163 | 0.00 |) 6S 14.13*| 192*] 1019 | 2997) 2,446] 0.249] 0.135 | 0.00 69 14.20 180 | 10135 | 2.580 | 2.136] 0.218 | 0.117 | 0.00 | 70.131 |$ 0.030 70 14.21 480 | 10035 | 1.805 | 1.480] 0.146 | 0.070 | 0.004 7 13.98 740 | 10055 | 3.947 | 3.179 | 0.352 | 0.148 | 0.047 72 13.87 600 | 1006 | 3.573.| 2.885 | 0.309 | 0.134 | 0.019 73 13.70 612 | 10035 | 2.571 | 2.089 | 0.230 | 0.106 | 0.012|| | | 14 13.61 573 | 10045 | 3.266 | 2.671 | 0.275} 0.120 } 0.013 | ¢ 0.036 |} 0.031 75 13.62 428 | 1004 | 4.113 | 3.544] 0.156 | 0.073 | 0.016 76 13.38 675 | 1006 | 3.942 | 3.207 | 0.323 | 0.144 | 0.028 77 13.27 565 | 10045 | 3.024 | 2.488 | 0.242 | 0.111 | 0.025 | 78 13.60 568 | 10045 | 2.609 | 2.173 | 0.212 | 0.089 | 0.025 79 13.00 630 | 1005 | 3.588 | 2.932] 0.307 | 0.119 | 0.039 | 0.048 | } 0.027 80 13.02 367 | 1004 | 1.854 | 1.548 | 0.136 | 0.060 | 0.025 81 13.07 398 | 1005 | 2.278 | 1.885 | 0.176 | 0.081 | 0.012 82 12.80 711 | 1004 | 3.842 | 3.156 | 0.303} 0.140 | 0.015 83 12.65 602 | 10035 | 3.436 | 2.841 | 0.278] 0.121 | 0.013 84 12.65 475 | 1004 | 2.002 | 1.672 | 0.153 | 0.072 | 0.015 | } 0.037 |} 0.040 85 12.46 655 | 10065 | 3.578 | 3.017 | 0.261 | 0.109 86 12.37 550 | 10055 | 2.784 | 1.822} 0.469] 0.092 | 0.013 87 TORS 479 | 1003 | 1.956 | 1.603} 0.163} 0.066 | 0.021 88 12.20 680. | 10035 | 3.534 | 2.911 | 0.253 | 0.119 | 0.061 89 12.26 44x | 10025 | 1.841 | 1.541 | 0.119 | 0.062 | 0 034 | } 0.047 |} 0.048 90 12.14 587 | 10035 | 2.931 | 2.458 | 0 188 | 0.091 | 0.043 91 11.97 670 | 1003 | 2.891 | 2.427] 0.200 | 0.094 | 0.035 92 12.02 465 | 10025 | 1.950 | 1.638 | 0.132 | 0.062 | 0.017 | } ) 93 11.78 630 | 1003 | 3.186 | 2.703 | 0.216 | 0.105 | 0.020 94 11.78.| 530 | 10025 | 2.080 | 1.750 | 0.141 | 0.068 | 0.012 / 0.034 0.065 95 11.65 620 | 1003 | 2.928 | 2.559 | 0.171 | 0.090 | 0.027 96 11.52 557 | 1003 | 2.582 | 2.209] 0.153 | 0.079 | 0.032 |} 97 11.52 486 | 10015 | 1.930 | 1.601 | 0.125 | 0.052 | 0.02% 98 11.50 467 | 1002 | 2.124 | 1.86%] 0.131 | 0.057 | 0.032 99 11.32 597 | 1004 | 3.308 | 2.816 | 0.195 | 0.089 | 0.037 100 11.19 550 | 1003 | 2.079 0.152 | 0.060 | 0.027 101 11.18 412 | 10025 | 1.983 0.132 | 0.059 | 0.021 102 11.11 490 | 10025 | 1.853 0.144 | 0.052 | 0.028 103 11.00 573 | 10025 | 2.297 0.181 | 0.061 | 0.031 104 10.92 5 | 10025 | 1.822 0.148 | 0.055 | 0.023 105 10.85 1002 | 2.532 | 2.18i | 0.181 | 0.065 | 0.038 106 10.79 10015 | 2.316 | 1.793] 0.307} 0.055 | 0.030 107 10.82 10015 | 1.724 | 1.307] 0.242] 0.043 | 0.024 108 1003 | 2.878 | 2.247] 0.336] 0.065 | 0.039 * Big loss in weight and low urine volume due to fact that no water was given the dog on sixty-seventh day. Paul E. Howe, H. A. Mattill and P. B. Hawk 111 TABLE I—(Concluded). : peepee etapa Be | | alee: se a ic} fe Z es - El rs cS) rs z a Fan ionst - & A < =) i.) 4 E ga a Z rt Z i 5] & 68 a4 & 4 a FI Bui) [eee Zz iv > BS i < ° is >I is] Z) mK a Sees a a = < F a <8 ° oP n° 5 fe = = EI = 4 a m > o = p % Es a} cs < 700 cc. water per day—Continued. a: See E ; ial Eee kgs. ce. grams | grams | grams | grams | grams | grams | grams 109 10.59 575 | ' 1002 2.263 | 1.760) 0.330} 0.055 | 0 030 110 10.48 622 | 1002 2.385 | 2.014) 0.173 | 0.058 | 0.030 111 10.38 601 | 1003 2.444) 2.068} 0.173 | 0.055 | 0.043 112 10.18 635 | 1003 2.665 | 2.255 | 0.198 | 0.056 | 0.043 113 10.23 473 | 10025 | 2.244 | 1.915 | 0.138 | 0.046) 0.044 114 10.15 510 | 10015 | 2.211] 1.883) 0.140} 0.048 | 0.041 115 10.05 514 | 10025) 2.368) 2.003 | 0.163 | 0.049 | 0.028 116 10. 02 550 | 1002 2.390 | 2.038 | 0.106 | 0.047) 0.032 117 9.76 596 | 1003 2.780 | 2.371} 0.174) 0.046] 0.042 precedes the pre-mortal rise the creatine-nitrogen excretion rs found: to be greater than that of creatinine-nitrogen."" If the data for the creatine-nitrogen and creatinine-nitrogen excretion of the present experiment be examined it will be noted that the creatine-nitro- gen output at no time exceeded that of creatinine-nitrogen. This fact precludes any possibility of considering the slightly increased nitrogen output of the last fasting day as the beginning of the pre- mortal rise in the nitrogen excretion. The pronounced increase in the nitrogen output for the sixteenth period was due to the fact that the daily water ration was increased from 700 cc. to 2100 ce. for each of the days of this period. The influence of this high water ingestion has already been discussed by us in another connection.!® The conclusion drawn from this increased nitrogen excretion when taken into connection with the creatine, purine, and allantoir data hereinafter discussed was to the effect that the high water ingestion had caused increased protein catabolism. This augmented output of nitrogen is neatly represented in Fig. I, p. 115. Urea-Nitrocen. For the most part the urea excretion ran closely parallel with that of total nitrogen. This fact is especially 17 Howe and Hawk: Loc. cit.; Howe, Mattill and Hawk: Proceedings Amer. Soc. Biol. Chem., July, 1910. 18 Howe, Mattill and Hawk: This Journal, x, p. 417, 1911. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 2. Elimination of Nitrogen During Fasting TABLE II. Nitrogen Distribution—Four-day Averages. I12 5B a eae eee ha = Z Boa ee 4 Bi Sees i | grams grams eae | 15.588 13.434 1 6.231 | 5.102 2 4.471 | 3.513 3 4.028 | 3.270 4 | 3.216] 2.544 5 | 3.410} 2.787 | 6 | 3.854] 3.162 | 7 | 8.474) 2.845 | 8* | 3.172) 2.577 9 2.608 2.130. 10 3.209 2.604 11 | 2.682 2.205 12 | Depse le. sar 13 3.391 2.896 14 2.983 | 2.555 15t | 2.642) 2.253 16t | 3.669 | 2.917 17 2 752 12.217 18 2.832 | 2.310 19 3.380 | 2.797 20 | 3.291 | 2.700 D1 eS] SEBS Q ee S5 i. 22 2.580 2.028 23 | 2.799) 2.334 24 2.536 | 2.163 | ig 2b... | Dasa ye 2 122 26 2.120 | 27) ©\|.:2.498sied 3760: 28 | 2.492] 2.022 29 CARY PAPA IE 30§ =| 2.390 2.038 31§ | 2.780 | 2.371 * Average for five days. t Average for two days. AMMONIA N CREATININE N Z a re) a A z | a] Zz z z Z = 2 2 & z 4 & < 2 A Ft =) 3 Z ° & = iS) MS = So ' . grams grams grams grams 0.051 0.032 0.112 0.027 0.029 0.060 0.036 0.123 0.039 0.160 0.032 0.107 0.042 0.147 t Period of high water Ingestion. § Single days. Paul E. Howe, H. A. Mattill and P. B. Hawk 113 TABLE III. Percentage Nitrogen Distribution. | | FOUR DAY PERIOD CREATININE N CREATINE N ALLANTOIN N Preliminary Period. 8 Day \ percent | percent per cent per cent per cent per cent | per cent Average {| 86-11 | 4.07 | 2.63 | 2.20] 0.42} 0.22 | 4.27 Fasting Period. : 1 81.89 | 1.94 4.75 1.56 1.00 0.50 8.38 2 78.59 | 7.54 6.00 1.99 0.40 0.54 4.97 3 81.16 | 9.93 5.46 1.04 | 0.60 0.52 127, 4 79.09 | 8.18 6.25 0.78 Pin |, 1.18 4.14 5 81.69 | 6.42 5.85 0.50 | 0.50 0.12 4.95 6 82.05 | 6.59 5.86 0.31 0.38 0.05 4.80 7 S91 |"6:16 | 6.33 0.12 0.37 0.06 5.07 8* 81.25 | 6.72 | 6.68 0.00 0.41 0.22 4.73 9 81.66 | 6.21 | 6.40 0.00 0.50 0.27 4.95 10 81.14 | 6.64 | 7.01 0.00 0.47 | 0.22 4.52 ll e220) |*"6. 11 | 6.52 0.19 0.67 0.22 4.06 12 83.14 | 6.80 | 4.65 0.28 | 0.63 0.21 4.37 13 85.40 | 7.84 | 4.98 0.09 0.44 0.32 0.91 14 85.57 | 9.31 | 5.76 0.00 0.50 0.37 15t 85.28 | 8.55 5.75 0.00 0.57 0.42 _ 16t 79.49 | 10.25 4.61 1.25 0.08 0.84 3.46 17 80.57 | 9.19 5.05 0.69 0.36 0.44 3.71 18 81.57 | 8.51 4.17 0.42 0.92 | 0.21 4.20 19 82.74 | 7.16 3.19 | 0.44 0.21 0.18 6.06 20 82.05 | 8.24 3.50 0.88 | 0.27 0.18 4.89 21 82.64 | 7.82 B5t | |. 0.66 s0e08 0.25 4.91 22 78.61 | 10.11 3.290 (0.627%) “oror 0.35 | 6.71 23 83.39 | 5.18 3.25 1.54 0.32 0.36 5.97 24 85.29 | 6.51 3.19" 19 0:75 0.28 0.47 3.51 25 85.37 | 6.07 iy {ae OP) 4.51 26 7.31 2.74 | 1.27 4.48 27 83.21 | 10.31 2.500 | 1.28F 2.83 28 81.14 | 10.15 2.33 | 1.44 4.93 29 84.91 6.75 2.07" 1.64 4.64 30§ 85.27 | 6.95 1.97 | 1.34 4.48 31§ 85.28 | 6.26 | 1.65 | 1.51 5.29 * Average for five days. t Period of high water ingestion. + Average for two days. § Single day. 114 Elimination of Nitrogen During Fasting TABLE IV. Body Weights, Creatinine Coefficients, Urine Volumes and Water Balance. q 2a ae | 5 i~| ee /$.8 | Bs Pr | ae | sB6e 6 é aR | RES | Ge82| g82 | gc | 38 | e228] 28 | BR Me 2 ue cae | ge ga | ats Bae nase ae Preliminary Period. g Day, kgs. kgs. per cent | grams cc. cc. | per cent Average! 26.33 | | 0.410 | 15.6 | 708 700 101.2 Fasting Period. 1 24.69 | 0.41 1.56 | 0.296 | 12.0 | 556 700 79.5 2 23.80) 0722 | 0.84 | 0:268 |.11.2 4) 470 700 Ciel 3 29 -O0sieOve3 | 0.87. | 02220) 1) GiGi 847 700 49.6 4 F484) || Oo IIAY OLS) =! 240i 9.0 | 649 700 92.7 5 21.49 | 0.22 | 0.84 | 0.199) 9.3 | 538 700 76.9 6 QOn Ga Onl9), | ON 720 082267) S059 5st 700 83.0 G 20.00 | 0.14 | 0.53 | 0.220! 11.0 | 533 700 76.1 8* 195317) 0014 | 0.53) |) 02212) 10) 506 700 (2.3 9 18.69 | 0.16 | 0.61 0.167 |} 8.9 | 467 700 66.7 10 17.98 | 0.18 | 0.68 | 0.225 | 12.5 78 700 82.6 11 SOR ORLS 10a ie OR iiom mele 554 700 79.1 12 16.99 | 0.10 | 0.38 | 0.132). 7.8 | 470 700 67.1 13 16.44] 0.14 | 0.53 | 0.169; 10.3 | 521 700 74.4 14 15.98 | 0.12 | 0.46 | 0.172) 10.8 | 449 700 64.1 15t 15.78 | 0.10 | 0.388 | 0.152) 9.6 | 475 700 67.9 16t 15.32 | 0.46 1.75 | 0.169 | 11.0 |1825 2100 86.9 17 14.44 | 0.22 | 0.84 | 0.139 | 9.6 | 657 700 93.9 18 UZ29BN Ola) | O46) OPIS: |) Sko 491 700 70.1 19 13.62 | 0.09 | 0.384 | 0.108! 7.9 | 544 700 UU ead 20 13,00) | FOG), | 0.6159) OFS) hy StS 609 700 87.0 21 12°65 10-09 = |--0-34 4) OL11G) |p 8. 7g 519 700 74.1 22 12.43 | 0.06 | 0.23 | 0.085; 6.8 | 540 700 (ila 23 11.97 | 0.12 | 0.46 | 0.091 7.6 | 596 700 85.1 24 11.65 | 0.08 | 0.30 | 0.081 7.0 | 561 700 80.1 25 DS2a0208 | O30) |)” OLOGST Ne Get 527 700 |. 75.3 26 11.00 | 0.08 | 0.30 | 0.058; 5.3 | 497 700 71.0 27 10.82 | 0.05 | 0.19 | 0.055; 5.1 540 700 Cele 28 10.38 | 0.11 0.42 | 0.058) 5.6 | 581 700 83 .0 29 10.05 | 0.08 | 0.30 | 0.049} 4.9 | 533 700 76.1 30§ 10.02; 0.03 | 0.11 0.047 | 4.7 | 550 700 78.6 318 9 76710.26. | 0.99) 3), 020461) 4-7) 396 700 83.7 * Average for five days. + Average for two days. t Period of high water ingestion. § Single day. 115 Paul E. Howe, H. A. Mattill and P. B. Hawk sou {M AGOE INI LVI" ININILW3Y9 SAVO 241 - ‘| qunory @sN 900 INISVI 116 Elimination of Nitrogen During Fasting clearly shown in the figure. Some interesting relationships are observed when the data for the percentage of the total nitrogen out- put which was excreted in the form of urea (Table III, p. 113) are examined. During the feeding period a trifle more than 86 per cent of the total nitrogen had been excreted as urea-nitrogen. Under the influence of the fasting metabolism this value decreased to an average of 81.5 per cent for an interval of forty-eight days. At this point in the fast the urea values rose to a value somewhat above 85 per cent for twelve days, then sank to the low level for about a month, finally coming back to the 85 per cent level for the final portion of the fast. Our data therefore show in general a decreased output of nitrogen in the form of urea during the fast above that excreted during the period of normal feeding. How- ever. these data do not substantiate the claim of numerous inves- tigators!® that the percentage output of urea decreases as the fast progresses. In this experiment the region of low urea values was during the first part of the fast whereas the region of high values occurred during the more advanced stages of fasting. We have verified the truth of the claim mentioned above in connection with certain experiments on fasting men®° in which a gradually decreas- Ing percentage output of urea-nitrogen was observed from the beginning to the end of the fasting interval. However, in the case of fasting dogs we have never succeeded in demonstrating a similar course for the urea output.” While it is true that the urea values during the first part of the fast were somewhat lower than at later stages of the test, it will be observed that the variation was not marked. The urea values may be looked upon, therefore, as more or less uniform in the case of fasting dogs. On the other hand in the case of fasting men there is a marked decrease in the percentage output of urea as the fast progresses, the decrease in one of our experiments being from 89.6 per cent to 79.2 per cent in six days, and in another instance the drop being from 86.2 per cent to 75.8 per cent in a similar interval. 19. and O. Freund: Loc. cit.; Brugsch: Zeitschr. f. erp. Path. u. Therap., i and iii, 1906; Osterberg and Wolf: Biochem. Zeitschr., v, p. 304, 1907; Underhill and Kleiner: This Journal, iv, p. 165, 1908; Schéndorf: Pfliger’s Archiv, cxvii, p. 257, 1907; Cathcart: Loc. cit. 20 Howe, Mattill and Hawk: Journ. Amer. Chem. Soc., xxxili, p. 568, 1911. 21 Howe and Hawk: Loc. cit. Paul E. Howe, H. A. Mattill and P. B. Hawk 117 The reason for the difference in the course of the urea output of fasting dogs as compared with fasting men may lie, as expressed in another paper from this laboratory,” in the fact that the dog, whether normally nourished or fasting is deriving its energy pri- marily in each instance, from nutritive material: of the same char- acter, 1.e., fresh lean meat when normally fed and muscular tissue when fasted. On the other hand man is accustomed to a cooked diet of a much lower protein content than the dog consumes and his cells are therefore confronted by very unusual conditions when asked to catabolize body tissue as they must perforce do in the course of a fasting test. It is not at all beyond the realm of pos- sibility that the differences just mentioned may account at least in part for the fact. that the course of the fasting excretion of urea- nitrogen is different in the organism of the dog from that observed in the human organism. AmMoNIA-NITROGEN. In common with the values for total nitrogen‘and urea-nitrogen the excretion of nitrogen in the form of ammonia underwent a sharp decline at the opening of the fast from the value as determined for the period of normal feeding. The actual figures for the average daily output were 0.635 gram for the feeding interval as against 0.121 gram for the first four-day period of the fast. The values for the two succeeding four-day periods were 0.337 gram and 0.400 gram respectively, but from this point for an interval of about two and one-half months the excretion of ammonia-nitrogen was fairly uniform. There were low values for ~ the ninth and eleventh periods and a very high value for the six- teenth period (water) but apart from these variations the general level of the excretion was about 0.25 gram. At this time (twenty- third period) there came an abrupt decrease in the ammonia values the data indicating an average daily output of about 0.160 gram for the remaining periods of the fast with two exceptions. The high ammonia value mentioned as occurring in the sixteenth period was due to the influence of the increased quantity of water fed the animal during each of the days of that period. It will be remembered that the usual daily allowance of 700 cc. was increased to 2100 cc. on each of these four days. The increased output of ammonia we interpret as an index of stimulated gastric function. 22 Howe and Hawk: Loc. cit. 118 Elimination of Nitrogen During Fasting This feature of the fast has been discussed in another connection™ and the interpretation offered is right in line with other interpre- tations from this laboratory which have had to do with the influence of water upon the ammonia excretion of the normally nourished individual.” CREATININE-NITROGEN. During the period of normal feeding the average daily output of creatinine-nitrogen was 0.410 gram. This value underwent a gradual decrease as the fast progressed, as is indicated by an output of 0.296 gram during the first period, one of 0.152 gram during the fifteenth period about two months later and one of 0.046 gram upon the one hundred and seventeenth day. This downward tendency of the creatinine output may be followed very nicely in Fig. I, p. 115. It is also of interest to note that the body weight curve as plotted in this figure runs in general parallel with the curve representing the creatinine output. In other words as the body lost in weight the creatinine output decreased. These facts are in line with the claim that the creatinine output is a func- tion of the amount of active muscular tissue in the body.» We have made similar observations in connection with other fasting studies.**, If Table I be examined it will be observed that the output of creatinine-nitrogen for the sixtieth day was far above the average output for the experiment up to thattime. Thiswas the day upon which the water ingestion of the animal was increased 200 per cent above the usual quota, as before mentioned. In experiments previously made in this laboratory in which men and animals have been subjected to the influence of fasting or water drinking the creatinine output has without exception been decreased under these conditions. It is interesting, therefore, in the present instance where we have the influence of the two factors (fasting and water drinking) exerted simultaneously upon the same individual, that the creatinine output should be increased rather than decreased. This feature has been more fully discussed in a previous paper to which reference has already been made. 23 Howe, Mattill and Hawk: This Journal, x, p. 417, 1911. 24 Fowler and Hawk: Journ. Exp. Med., xii, p. 388, 1910; Wills and Hawk: Proceedings Amer. Soc. Biol. Chem., This Journal, ix, p. xxix, 1911. 25 Folin: Amer. Journ. of Physiol., xiii, p. 66, 1905; Shaffer: Zbid, xvi, p. 252, 1906; McCollum: Amer. Journ. of Physiol., xxix, p. 210, 1911. 26 Howe and Hawk: Loc. cit.; Howe, Mattill and Hawk: Loc. cii. Paul E. Howe, H. A. Mattill and P. B. Hawk 119 At the very end of the fast it will be noted by examining Fig. I, p. 115, that the curves for the excretion of nitrogen in the forms of creatinine and creatine approach rather closely to each other but do not cross. The relation of this fact to the absence of any pre-mortal rise in the nitrogen excretion is discussed elsewhere in this paper in connection with the excretion of creatine and total nitrogen. The percentage of the total nitrogen which was excreted as creatinine (Table III, p. 113) decreased gradually as the fast pro- gressed. CREATINE-NITROGEN. During the course of the eight-day pre- liminary period there was a daily average of 0.343 gram of creatine- nitrogen excreted by the dog. This creatine excretion is of course directly traceable to the fact that the dog is a “high protein” ani- mal. As we have already said, the diet of the dog in question con- tained 3.75 grams of protein per kilogram of body weight. In other words a man of 70 kg. weight if fed on the same level would be ingesting over 260 grams of protein per day. . This is a protein ingestion about two and one-half times greater than that sug- gested by the Voit standard’ and about five times greater than that suggested by Chittenden.2* The normal human organism does not excrete ingested creatine to any degree unless such an organism be living upon a high protein level similar to that above men- tioned.2? The gradual increase in the creatine output accompany- ing an increase in the diet is very nicely shown in the study of “repeated fasting’ recently reported from this laboratory by Howe and Hawk. In the present instance upon the first day of the fast there was, of course, a very sharp drop in the creatine elimination, only 0.109 gram being eliminated whereas the average daily output for the first four-day period was 0.097 gram as against a daily average of * 0.343 gram for the interval of high protein feeding. From this point the creatine excretion underwent a gradual decrease until the eighth period at which time the urine of the dog was found to be practically creatine-free. This period of low creatine values con- tinued for nearly one month or until the end of the fifteenth period. 27 Lusk’s Science of Nutrition, 2d Ed., 1909. 28 Chittenden: Physiological Economy in Nutrition, 1904. 29 Folin: Hammarsten Festschrift, p. 15, 1906. 120 Elimination of Nitrogen During Fasting At the opening of the sixteenth period it will be observed that creatine again appeared in the urine in large quantities, the excre- tion of this constituent being greater than at any time in the whole experiment subsequent to the second period or eighth day. In other words in a total of one hundred and nine fasting days the period in question showed the highest creatine values. This is all the more striking when we recall the fact that it follows imme- diately after an interval during which the urine was to all intents and purposes creatine-free. By referring to Fig. I, p. 115, the course of the creatine-nitrogen excretion may be very conveniently followed. It will be seen that the curve gradually descends during the early part of the fast and in the eighth period, after about one month’s fasting, it assumes the low level mentioned, and continues at this low level until the opening of the sixteerith period as before mentioned. This period of very low creatine values is represented on the figure by a vir- tually straight line thus accentuating the following rise of the six- teenth period. It will be remembered that this sixteenth period was the inter- val during which the daily water quota of the animal was increased from the usual one of 700 ce. to one three times as great, 7.e., 2100 ec. In previous work from this laboratory*® upon the influence of water drinking upon the creatine excretion it has been demon- strated that the ingestion of large volumes of water by normally nourished men was followed by the appearance of creatine in the urine. The creatine data of the experiments mentioned have been offered as the first direct experimental evidence in support of the hypothesis that the increased nitrogen output which fol- lows water drinking is due to a stimulation of protein catabolism and not to a simple flushing of the tissues. Bearing these findings in mind the high creatine values of the water period of this fasting _ study are very significant. Here we have an animal which has been fasting for nearly two months, receiving a daily ingestion of water amounting to 700 cc. Under these conditions the urine volumes averaged about 500 cc. indicating that the tissues and organs of the dog must have been pretty well flushed during each day of the fasting interval. Moreover the urine was practically 3° Fowler and Hawk: Loc. cit.; Howe and Hawk: Unpublished. Paul E. Howe, H. A. Mattill and P. B. Hawk 121 creatine-free as has already been mentioned. At this very oppor- tune moment of minimum creatine values the water ingestion of the animal was increased 200 per cent, and coincident with this in- creased water intake comes the augmented creatine output. Cer- tainly it is perfectly logical to conclude in this connection that the water was the active factor in bringing about the increase in the quantity of creatine eliminated. We very naturally look to the muscular tissue when we inquire as to the origin of the creatine. The water has evidently been instrumental in causing a true cat- abolism of protein material. As discussed in a previous paper,*! however, when we attempt to show a definite relationship between the total nitrogen figures and those for creatine-nitrogen our cal- culations indicate a discrepancy. The total nitrogen output was increased 3.188 grams during the water period, a nitrogen quota equivalent to 98 grams of flesh if we take the value 3.25 per cent for the nitrogen percentage of flesh. Taking creatine in a similar way and using 0.123 per cent as the creatine-nitrogen value of flesh we find that the increased creatine-nitrogen of the water period aggregated 0.182 gram, a value equivalent to 148 grams of flesh. There is thus a discrepancy of 34 per cent between our total nitrogen and our creatine-nitrogen figures if we consider that each type of value represents the complete disintegration of muscular tissue. This being true, we were forced to the conclusion, as already discussed elsewhere, that creatine may be removed from mus- cular tissue and excreted in the urine without its removal of necessity being accompanied by the complete disintegration of that tissue. In support of this contention we would offer certain other evidence obtained in connection with fasting experiments carried out in this laboratory. In these tests the creatine content of muscle was much decreased as the result of fasting, a decrease of over 60 per cent being noted. The nitrogen content of this same muscle was however but slightly lowered. This low creatine value for a muscle which still retains its original nitrogen quota practically unaltered is a very significant finding, and emphasizes again the inaccuracy of considering the total amount of creatine excreted as having arisen from the complete and permanent disintegration of muscular tissue. It is evident then that creatine may be removed 31 Howe, Mattill and Hawk: This Journal, x, p. 417, 1911. 32 Howe and Hawk: Journ. Amer. Chem. Soc., xxxiii, p. 215, 1911. 122 Elimination of Nitrogen During Fasting from tissues which are still functioning within the body. Mendel and Rose* have recently reported an increase in the creatine con- tent of the muscles of fasting rabbits and hens. In this connection they have objected to certain of our interpretations. The matter has received further consideration from us in a recent article.# After leaving the water period the curve for the excretion of creatine-nitrogen descends abruptly and from this low plane be- gins a somewhat gradual rise to the end of the fast. Coincidently with this rise in the creatine-nitrogen curve it will be noted that the curve representing the output of creatinine-nitrogen descends. They approach very close to each other but do not cross. This fact is of great significance when taken into consideration with other creatine and creatinine data collected by us in recent fasting studies. In every instance in which our animals have been fasted to the so-called pre-mortal rise in the nitrogen excretion, the crea- tinine-nitrogen output has decreased during the later stages of the fasting interval and this decrease has been associated with a much more pronounced increase in the output of creatine-nitrogen. When these values were plotted it was noticed that the curves for the excretion of creatine and creatinine, in every case, crossed a few days before the fall in the nitrogen output which preceded the pre- mortal rise. This ‘“‘creatine crossing” occurred with great regular- ity at practically the same point with respect to this fall in total nitrogen output and it is believed to be a sign of more than ordi- nary significance. It will be studied further in this laboratory. On the basis of our knowledge regarding the relationship between the ‘creatine crossing” and the ultimate death of an animal we estimate that ‘Oscar’ would have been able to live at least one hundred and thirty days without food. When the elimination of total nitrogen was discussed in a pre- vious paragraph attention was called to the fact that there was a slight increase in the nitrogen output upon the one hundred and seventeenth or final day of the fast. This slight increase in the nitrogen excretion is not believed to be connected in any way with the pre-mortal rise in nitrogen excretion inasmuch as it was pre- ceded neither by a decreased output of nitrogen nor by the phe- nomenon we have termed the “creatine crossing.”’ 33 Mendel and Rose: This Journal, x, p. 255, 1911. 34 Howe, Mattill and Hawk: This Journal, x, p. 417, 1911. Paul E. Howe, H. A. Mattill and P. B. Hawk 123 An examination of Table III, p. 118, will show that there was not only an actual increase in the output of creatine-nitrogen dur- ing the final stages of the fast but also an accompanying increased percentage output aswell. In other words the increased output of creatine was accompanied by a less pronounced increase or by a decrease in the output of total nitrogen. PuriNnE-NiTROGEN. These values and those for allantoin-nitro- gen were determined for the first ninety-six days of the one hun- dred and seventeen-day fast, and reported in connection with other data on the allantoin and purine output of fasting dogs.* The data are included in the tables of the present paper in order that the records may be complete on this exceptionally long fast. A brief summary of the findings in this connection will be given at this time. The purine values were somewhat irregular during the fast but there was nevertheless a decided tendency toward a decreased output as the fast progressed. For example if we com- pute the average output for the first half of the fast and compare this with the average output for the second half of the fast we ob- serve that the output was considerably decreased during the second half of the fast. Scaffidi®® has recently reported a decreased out- put of purine-nitrogen by a dog during the course of a sixteen-day fast. On the other hand Schittenhelm,*’ and Underhill and Kleiner*® found the course of the excretion to be irregular. In the interest- ing work of Hunter and Givens*® on the purine excretion of the coyote the course of the elimination of this form of nitrogen was not studied inasmuch as composite urine samples were utilized for analysis. By referring to the data for the water period in Tables I and II, pp. 109 and 112, it will be observed that the purine-nitrogen out- put decreased in a very marked manner during the time of high water ingestion. This phenomenon is discussed later. ALLANTOIN-NITROGEN. The output of allantoin-nitrogen during the fast was irregular. However, if we compare the output for the first thirty days of the fast with the output for the last thirty 35 Wreath and Hawk: Journ. Amer. Chem. Soc., xxxiii, p. 1601, 1911. 36 Scaffidi: Biochem. Zeitschr., xxxili, p. 153, 1911. 37 Schittenhelm: Zeitschr. f. physiol. Chem., \xii, p. 80, 1909. 38 Underhill and Kleiner: This Journal, iv, p. 165, 1908. 39 Hunter and Givens: This Journal, viii, p. 449, 1910. 124 Elimination of Nitrogen During Fasting days we find that this output has decreased 40 per cent, 7.e., from 0.4 gram to 0.24 gram. None of the previous investigators of the fasting output of allantoin have observed such a marked decrease during the final stages of the fast as we have recorded here. Upon the days of high water ingestion the output of allantoin was increased in a very pronounced manner. Previous to this water period the average daily excretion of allantoin-nitrogen had been 0.011 gram. The high water intake caused this daily value to be increased more than three-fold on the first day of its ingestion, the values remaining quite uniformly high throughout the period. It will be recalled that the values for the purine-nitrogen excretion were decreased during this interval of copious water ingestion in which the allantoin-nitrogen values were increased. Furthermore it has been shown by Rulon and Hawk“ that the uric acid output is decreased under the influence of an increased water ingestion. It is well known that purine bodies may be oxidized to allantoin and furthermore that the allantoin excretion of an animal may be in- creased by purine feeding. It therefore seems fair to conclude that the large volume of water introduced into the body of this fasting dog has markedly stimulated the oxidation mechanism and consequently such substances as would under ordinary con- ditions go to augment the purine-nitrogen output have been transformed into allantoin and are excreted in this form. If we consider the total output of nitrogen of purine origin (purine-nitrogen + allantoin-nitrogen) we observe that it is in- creased during the interval of high water intake. This may be taken as further evidence in the support of the hypothesis that at least a part of the increase in the total nitrogen output observed to follow copious water drinking is due to a true protein catabolism rather than to a flushing of the tissues. As before mentioned in connection with the discussion of the output of creatine, the in- creased nitrogen excretion during the water period was equivalent to 98 grams of flesh. If we calculate the purine-nitrogen value" for this 98 grams we find it is 0.059 gram whereas the actual increase in this form of nitrogen was but 0.032 gram. in other words, we cannot account for 46 per cent of the theoretical quantity of 40 Rulon and Hawk: Journ. Amer. Chem. Soc., xxxii, p. 1686, 1910. 41 Hall: The Purine Bodies in Foodstuffs, Manchester, 1902, p. 29, Table IV. Paul E. Howe, H. A. Mattilland P. B. Hawk 125 purine-nitrogen. The cause of this discrepancy may be due partly to the method of analysis employed and partly to the fact that the allantoin was further oxidized with the resultant formation of other nitrogenous substances. UNDETERMINED NitTRoGEN. The data for the actual output of this form of nitrogen indicate considerable irregularity as would naturally be expected in a fasting organism. In general the values were higher during the first half of the fast than they were during the later stages of the fasting interval. When we consider the percentage output of undetermined nitrogen as given in Table III, p. 113, we observe a much greater uniformity from period to period than is ordinarily the case. The average daily percentage value was about 4.6 per cent which was a trifle higher than the value for the period of normal feeding preceding the fast, 7.e., 4.3 per cent. Other fasting tests already reported from this laboratory” have also shown very uniform undetermined nitrogen values for fasting periods. In the instances cited, however, the level of the feeding periods was considerably above the fasting level, whereas in the present experiment this variation was not noted. Body Weights, Creatinine Coefficients, Urine Volumes and Percentage Water Elimination. The dog weighed 26.33 kg. at the start of the fast whereas his weight at the end of the fast on the one hundred and seventeenth day was 9.76 kg. He had lost approximately 63 per cent in body weight. The daily loss was greater during the first portion of ‘the fast than during the later stages, as would logically be expected. There was in general a gradual decrease in the daily loss in weight up to the time the water ingestion of the animal was increased from 700 cc. to 2100 ce. This high water intake occurred during the sixteenth period. If Table IV, p. 114, be examined it will be observed that the average daily loss for over a month previous to this time had been 0.10 to 0.15kg. With the advent of this inter- val of copious water ingestion, however, the daily loss increased to 0.46 kg. this being the highest daily loss sustained by the animal at any time during the fast. The loss in weight was still marked (0.22 kg.) during the period following the time of high water intake 42 Howe and Hawk: Journ. Amer. Chem. Soc., xxxiii, p. 215, 1911. 126 Elimination of Nitrogen During Fasting but from that time up to the last day of the fast the average daily loss was in general progressively decreased. The loss upon the one hundred and seventeenth fasting day was 0.26 kg. This big loss was due partly to a high urine volume and a mild diarrhoea. The creatinine-coefficient was 15.6 for the preliminary feeding period. At the opening of the fast the coefficient dropped to 12.0 and decreased slowly and irregularly from this point to the eigh- teenth period. For the remainder of the fast the coefficient de- creased rather more sharply, the tast ending with a coefficient of 4.7. The average daily urine volume during the preliminary period was 708 cc. a volume which was practically identical to the daily water ingestion. As the fast opened the urine flow very naturally fell somewhat maintaining an average of about 500 ce. for the inter- val up to the time of high water ingestion, 7.e., the sixteenth period. The average urine volume for the later part of the fast was some- what higher than it had been at earlier stages in the fast, the value being about 550 ce. as against 500 cc. The volumes were more uniform from day to day and from period to period in the later portion of the fast. With few exceptions, however, the daily urine volumes showed a satisfactory uniformity throughout the fast when we take into consideration the fact that the animal was not catheterized. The urine was acid in reaction throughout the fast. Differential leucocyte counts were made throughout the fast a report of the findings having already been presented.* SUMMARY. The subject of the fast was a Scotch collie dog (‘‘Oscar’’) weigh- ing 26.33 kg. at the opening of the fast. The fast was one hundred and seventeen days in length thus constituting by many days the longest fast on record. The dog gave evidence of being possessed of wonderful vigor and stamina. This was indicated by the fact that he was able to jump out of his cage so late as the one hundred and first fasting day. At the end of the fast of one hundred and seventeen days the animal was carefully fed and ultimately brought back to his original 48 Howe and Hawk: Proc. Am. Soc. Biol. Chem., 1911. Paul E. Howe, H. A. Mattilland P. B. Hawk 127 body weight and subjected to a second fast. The data from this second fast will soon be published. The urine of the animal was examined eeetitciely for total nitrogen, urea, ammonia, creatinine, creatine, allantoin and purine- nitrogen. The total nitrogen content of the feces was. also deter- mined. During the pre-fasting interval, the dog was fed a diet containing 3.75 grams of protein per kilogram body weight. He also received 700 ce. of water per day during the feeding interval as well as during the fast. The body weight loss aggregated about 63 per cent for the one hundred and seventeen day fast, the actual weight being 26.33 kg. before the fast and 9.76 kg. on the one hundred and seventeenth day. There was no indication of a pre nani rise in the nitrogen ex- cretion. The ‘‘creatine crossing, ”’ 7.e., the point in a fast at which the putput of nitrogen in the form of ne exceeds that in the form of creatinine was not inevidence. This fact is interpreted as indicating that the dog would probably have been able to fast a total of at least one hundred and thirty days if he had not been fed upon the one hundred and seventeenth day. At the end of the fifty-ninth fasting day, the water ingestion of the dog was raised to 2100 cc. per day for an interval of four days. This caused an increase of 77.5 per cent in the total nitrogen output for the first day, urea, ammonia, creatinine, creatine, and allantoin being simultaneously increased, whereas purine was decreased in quantity. The creatinine coefficient was 15.6 for the period of normal feed- ing preceding the fast, 12.0 at the opening of the fast and 4.7 on the one hundredth and seventeenth fasting day. The percentage nitrogen distribution was in general similar to that reported by us in connection with shorter fasts on dogs. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 2. ge * ns aa ct < oe Daf At ae? , nite base Perere 7 . bovine befatcas a} ab A jap age Slee bo oat Res winger gary 7 ’ . ; . esse vincent gee - staideterp snd peataays a 7 - 4 - er ie ions a We 7) F G ett id 21 as) eee ae 2 OB — , a ae Th ua 2s tt eae 4+. ehipieegs! ares Aine. 7 ce , >. ‘ s Pune 44% ey sae es cb , eee or ae. : bs oi _ = =* ior é % =a. ’ ~ Ae ee 3 3 L if if = 7s3 stan my te . aretin = iy 7" isa wilds : ieoth eels dL igrecki's ane 2 ' " ; an nat ieee Ae = ad oa Bier - SP % > ae ™ > _— thy STUDIES ON WATER DRINKING: XIII. (FASTING STUDIES: VIII.) HYDROGEN ION CONCENTRATION OF FECES! By PAUL E. HOWE anp P. B. HAWK. (From the Laboratory of Physiological Chemistry of the University of Illinois.) (Received for publication, January 18, 1912.) INTRODUCTION. The reaction of feces has been determined qualitatively by v. Koziczkowski?? who noted the color change of litmus in aqueous extracts of feces. Hemmeter*® also observed the reaction in fecal extracts and noted the varying reactions of the same extract to different indicators. The use of various indicators for the deter- mination of the degree of acidity of the intestinal contents has been reported by Mattes, Macfadyen, Nencki and Sieber® and also by Hemmeter,® and the results show that the true reaction varies slightly, but is in general not far from neutral. The results of the authors already mentioned as having made a study of the reaction of feces with the use of various indicators and of v. Oefele,? Schmidt and Strasburger® and Lynch?® show the re- 1 Reported before the American Society of Biological Chemists at Balti- more, Dec., 1911. 2 vy. Koziczkowski: Deutsch. med. Woch., 1904, No. 33. 3 Hemmeter: Arch. f. d. ges. Physiol., |xxxi, p. 151, 1900. 4 Mattes: Berliner klin. Woch., 1898, p. 539 (XVI Kongress fiir Innere Medicin, Wiesbaden). 5 Macfadyen, Nencki and Sieber: Arch. f. exp. Path. wu. Pharm., xxviii, p. 311, 1891. ® Hemmeter: Loc. cit. Ty. Oefele: Statistische Vergleichstabellen zur pract. Koprologie, Jena, 1904 (from Schmidt and Strasburger). 8 Schmidt and Strasburger: Die Faeces des Menschen, Berlin, 2d Edition, 1905. ® Lynch: Copologia, Tesis, Buenos Aires (from Schmidt and Strasburger 1896, p. 52). 129 130 Hydrogen Ion Concentration of Feces action of normal feces to be approximately neutral, 7.e., acid to indicators which give changes in color at hydrogen ion concentra- tions of 1 X 1077 or less and alkaline to indicators changing color at hydrogen ion concentrations greater than that value. Similar conditions have been observed by us in our work upon the reaction and hydrogen ion concentration of feces, in which we used methyl orange, litmus, lacmoid, rosolic acid and phenolphthalein papers to test the reaction. Methods for the determination of the titrat- able acidity of feces have been proposed by Rubner,!® who ex- tracted with water and titrated with baryta water; by Blauberg," Boas,” and J. Miiller® who titrated with 4} sodium hydroxide (hydrochloric or sulphuric acids for alkaline stools) using phenol- phthalein or litmus paper as an indicator. Langstein' observed that the titratable acidity depended upon the indicator used. The neutral, slightly alkaline or amphoteric (Lynch, Schmidt and Strasburger,!* Hecht,!’ Nothnagel'®) reaction observed in the normal feces of an individual on a mixed diet, may give place to a distinctly acid or alkaline reaction, according to the kind of food ingested. The reactions of the normal stools as shown by the use of different indicators have been explained for the neutral stools as due to the presence of carbonates, and perhaps other gas-forming substances, and phosphates (Mattes,!? Hemmeter?”’). For acid stools the presence of large quantities of fatty acids has been suggested which may result from excessive carbohydrate fermentation or from the poor utilization of the ingested fat. The 10 Rubner: Zeitschr. f. Biol., xv, p. 159, 1879. 11 Blauberg: Experimentelle und kritische studien tiber Sduglings Faeces, Berlin, 1896, p. 42. 12 Boas: Diagnostik und Therapie der Darmkrankheiten, Leipzig, 1898, p. 103. 13 J. Miiller: Uber die Reaktion der normalen Sduglings Faeces, Diss. Rostock, 1907. 14 Langstein: Jahrbuch fur Kinderheilkunde, lvi, p. 330 (cited by Hecht). 15 Lynch: Loc. cit. 16 Schmidt and Strasburger: Loc. cit, p. 106. 17 Hecht: Die Faeces des Séuglings und des Kindes, Berlin, 1910. p. 20. 18 Nothnagel: Beitrége zur Physiol. und Path. des Darms, Berlin 1884, p. 79. 19 Mattes: Loc. cit. 20 Hemmeter: Loc. cit., p. 156. Paul E. Howe and P. B. Hawk 131 alkaline stools which accompany pronounced putrefaction are due largely to the resulting ammonia (Schmidt and Strasburger) 2! Especial attention has been given to the reaction of infant stools and the data show that with the ingestion of mother’s milk an acid stool results, whereas with the ingestion of cow’s milk the reaction is alkaline (Blauberg,” Hellstrom,” Langstein,2* Schlossmann,25 J. Miiller®). Schlossmann explains this phenomenon as due to the higher fat content, with relation to the protein, which is found in mother’s milk. A quantitative examination (Blauberg) of the feces resulting from these two diets shows that the acidity is due largely to volatile fatty acids. Hedenius?? has shown that with the same diet (carbohydrate), he obtained an acid stoolor an alkaline stool according to the age of the infant (two months or seven to ten months). The reaction of the feces of adults fed upon cow’s milk is similar to that of the infant stool resulting from a similar diet, 2.e., neutral to slightly alkaline (Rubner), although Lynch shows that the reaction may be slightly acid under a conditions. A pathological condition of the intestinal tract may result in a change in the reaction of the feces. Those diseases associated with the poor utilization of the fats or in an increased fermentation give rise to an acid stool, while those diseases which result in an increased putrefaction are accompanied by stools having an alka- line reaction. Miller? and Schmidt and Strasburger report the reaction of fasting feces as slightly acid. The exact hydrogen ion concentration of feces, so far as we have been able to find from an examination of the literature, has never been determined. 21 Schmidt and Strasbutger: Loc. cit., p. 107. 22 Blauberg: Loc. cit. 23 Hellstrom: Archiv fiir Gyndkologie, 1901 (cited by Hecht). 24 Langstein: Loc. cit. 25 Schlossmann: Zentralbl. fiir Kinderheilkunde, ix, No. 7, 1906. 26 J. Miller: Loc. cit. 27 Hedenius: Arch. jf. Verdauungskrankheiten, viii, p. 379, 1907. 28 Miller: Berliner klin. Woch., xxiv, p. 433, 1887; Virchow’s Archiv, exxxi (supplement), 1893. 132 Hydrogen lon Concentration of Feces EXPERIMENTAL. Three men, C, V and E served as subjects in this investigation. Subjects C and V were used in the study of the effect of water drinking with meals. Subject C was twenty-nine years old and weighed 60 kg.; while subject V was twenty-four years old and weighed 58 kg. Subject E, who was used in the fasting test, had been the subject of previous experiments in this labor- atory.?° The investigation upon subjects C and V was divided into five periods, the diet remaining uniform. The periods are given in Table I, p. 1385. Examinations were made of the stools of the last four periods. The fasting experiment was divided as follows: a four-day pre- liminary period of high protein intake, a fasting period of seven days, a period of four days in which the subject ingested a low protein diet and a final period of four*®® days during which the diet was the same as that ingested during the preliminary period. The diet was the same in character for each of the three subjects, The two men in the water drinking experiments ingested 400 ce. milk, 100 grams of graham crackers, 15 grams of peanut butter and 25 grams of butter with each meal. The diet of subject E will be given in a forthcoming paper from this laboratory.*! The fecal extract, to be used in the determination of the hydrogen ion concentration was prepared as follows, 50 cc. of $ NaSO, solution being used for the extraction; exactly 2 grams of moist feces was weighed out, by difference, into a mortar, and about 5 ec. of the $} Na,SO, solution added after which the feces were worked up with a pestle until the sample was in a fine homog- eneous suspension. The remaining portion of the 50 cc. of the 29 Howe, Mattill and Hawk: Journ. Amer. Chem. Soc., xxxili, p. 568, 1911; Mattil] and Hawk: Jbid., xxxiii, p. 1978, 1911. Unpublished experi- ments. 30 Tn the case or the urine the period was five days in length (see Sherwin and Hawk, unpublished). 31 Sherwin and Hawk: Loc. cit. 32 The comparison was made upon the bases of 2-gram samples of moist feces. The variable moisture content of the feces seemed to preclude such a standard but careful consideration indicated that no other satisfactory basis was apparent. Paul E. Howe and P. B. Hawk 133 Na,SO, solution was then added and the whole thoroughly mixed together. This suspension was centrifugated and the supernatant liquid taken for the determination. The solutions prepared under these conditions were usually colored a light yellow to a dark brown and showed very little sediment upon standing. The “true acidity” of the fecal extract is defined as the hydro- gen ion concentration while the ‘‘titratable acidity” is the quantity of acid or alkali of known strength required to produce neutrality with respect to some indicator. 'Two methods of determining the true acidity are available, by the use of a series of indicators or by the aid of the hydrogen electrode. The indicator method* is very satisfactory when clear solutions can be obtained; however, when colored or turbid solutions are to be examined this method either fails or loses its accuracy. In the case of feces it is prac- tically impossible, by filtration through paper or by centrifugation, to obtain an extract which is not colored and turbid. The method of centrifugation offers the best means of preparing the fecal extract free from all of the heavier particles, and adapts itself especially to routine and clinical work. The determination of the hydrogen ion concentration by means of the hydrogen elec- trode offers the most accurate method of obtaining the true acidity of fecal extracts. In our work the determinations were all made upon fecal extracts prepared by centrifugation and the hydrogen ion concentration was determined with the hydrogen electrode.* The determination of the hydrogen ion concentration by means of the -hydrogen electrode depends upon the difference of potential which exists between two hydrogen electrodes dipping into'solutions of different concen- trations. Knowing the difference of potential and the hydrogen ion con- centration of one solution we can calculate the hydrogen ion concentration of the other solution according to the Nernst formula, REA; et any ee i nF Cs 33 Friedenthal: Zeitschr. f. Elektrochem., x, p. 118, 1904; Salm: Zettschr. f. physikal. Chem., lvii, p. 471, 1907; Michaelis and Rona: Zevtschr. f. Elec- trochem., xiv, p. 251, 1908; Walpole: Biochem. Journ., v, p. 207, 1910. 34 We wish to thank Drs. E. W. Washburn and Grinnell Jones of the lab- oratories of physical chemistry and electro-chemistry for their courtesy in loaning us apparatus and in aiding us with many helpful suggestions. 134 Hydrogen Ion Concentration of Feces : F : . : 2 C; ; This expression can be simplified in this case to t = KT log or where 2 R = Fx 04343 and is equal to 0.0001983, a is the difference of potential between the two electrodes in solutions whose concentrations are C; and C2, Ris expressed in joules per degree, 7’ is the absolute temperature, n the valence is equal to unity and F is equal to 96,540 coulombs. The differ- ence of potential was measured by means of the Poggendorf compensa- tion method. A Lippmann electrometer was used to indicate the zero po- tential. The apparatus was sensitive to changes of 0.001 volt but readings were only recorded to 9.1 volt as this was sufficient for the purposes of this experiment. The difference of potential between the two solutions was ordinarily small, consequently a Weston cell was introduced into the circuit of the concen- tration cells to increase the voltage of thatcircuit. Intaking readings this cell was placed first in series with and then against the concentration cell, thus giving a check on the readings. The Weston cell was compared with a standard Weston cell both before and after a series of readings. The standard of comparison was a solution containing 0.2 mole Na,HPO, and 0.1 mole NaH2PO,.** This solution as has been shown by both Washburn and Henderson,** gives a hydrogen ion concentration of approximately 1 X 1077; 7.e., it is neutral. The feces were extracted with a 3 solution of Na,SO,. Such a concen- tration of Na2SO, was selected in order that the concentration of the sodium ions should be approximately equal in both the standard and the unknown solutions, thus tending to reduce the solution-potential betweenthem. The sodium sulphate solution served to neutralize the effect of the variations in electrolyte content of the feces and to carry the current and thus prevent changes in the concentration of the hydrogen ions around the electrodes. A saturated solution of sodium sulphate was used as the connecting solution between the feces extract and the standard phosphate solution. The form of apparatus was that described by Salm.*7_ This consisted of two U-tubes 18 mm. in diameter, the arms of which were 60 and 80 mm. in length respectively. The long arm of each U-tube was closed with a three- hole rubberstopper which held the platinum electrode, a tube for conducting the hydrogen to the electrode; one end of which was drawn out into a capil- lary and bent at an angle of nearly 180° and another glass tube bent at an angle of 90° which permitted the escape of the hydrogen from that arm of the U-tube. The two solutions were connected through the short arms of 3¢ This solution has already been used in this laboratory for another pur- pose (see Hawk: Arch. Int. Med., viii, p. 552,1911). 36 Washburn: Journ. Amer. Chem. Soc., xxx, p. 31, 1908; Henderson: Amer. Journ. of Physiol., xxi, p. 173, 1908. 37 Salm: Zeitschr. f. physikal. Chem., lvii, p. 471, 1907. Hydrogen Ion Concentration of Feces. Paul E. Howe and P. B. Hawk TABLE I. Water Drinking. 135 (Hydrogen ion concentrations are expressed in moles per liter). NUM- SUBJECT V NUM- SUBJECT C apc Nl eas Weight of | H:O | Ht Weight of | H:0O Ht ea F eces, Feces eialata pe Feces. Feces Somes Moderate Water Drinking (ten days). grams per cent grams per cent 3} 101.5 196 721% 108 4/.°-153.5 S2.4°) 1.7 1 103.5 16021 1:9"X = —- ] o CHOH CH — | + 3H,0 CHOH , CHOH CHOH CHOH | CH,OH + HOOC 3 CyHs; CH,00C 2 Cus 8 It had been observed in the preparation of mannid distearate that the amounts of this ester and of the mannitan distearate varied considerably in different preparations, and the variation was reciprocal, 7.e., when the amount of mannitan distearate was great, that of mannid distearate was small, and vice-versa. The explanation of the variation and of the absence of mannid distearate in the above preparation seems to be that when the sulphuric acid mixture is poured into the salt solution, the excess of water present completely hydrolises the mannid distearate to the mannitan form, while when the esters are extracted from the sulphuric acid mixture directly with ether, the change takes place to only a limited degree, owing to the small amount of water present. ® Griin: Loc. cit. 148 Mannite Esters of Stearic Acid Formation of mannid distearate. Mannid distearate CH,OH + HOOC-CyH- CH,00C - CrHss CHOH CH | iio CHOH CH = O | >o + 4H,0 CHOH CH CHOH CH CH,OH + HOOC-CyHs CH,00C- CrHss Formation of isomannid distearcte by heating mannid distearate. Isomannid distearate CH,00C - CyHss CH,00C -CrHs CH CH | yo yo CH CH | + HO CHOH CH teas vs CHOH CH CH,00C - Cy Hs CH,00C- CyHss A closer study of the reaction was now undertaken with the object of improving the yield. The reaction mixture left after extraction of the esters was first examined. It was diluted with water and treated with powdered barium carbonate and hydrate until the sulphuric acid was removed, filtered, the precipitate of barium sulphate and carbonate washed with cold water, the fil- trate and washings carefully neutralized and evaporated to small bulk. A considerable amount of a compound identified as the barium salt of ethyl sulphuric acid (derived from the ester) was recovered from the residues, but never any mannite compounds. This was remarkable, since according to the amount of esters obtained about 60 per cent of the mannite should have been present in the residues The reactions which take place when mannite is treated with concentrated sulphuric acid at different temperatures were next studied. This work has proved to: be unexpectedly complicated W. R. Bloor 149 and, as it is still incomplete, a full report will be reserved for a later publication. The following facts are fairly well established. When mannite is dissolved in concentrated sulphuric acid at tem- peratures below 50° C., there is formed mannitan- (and possibly mannite-) disulphuric acid ester. At temperatures over 50° C., mannid-disulphuric acid ester is formed and, at the same time, there is a considerable condensation of the mannite molecules with the formation of a substance of the nature of a mannite ether, which has little or no power of forming esters. The formation of this substance at temperatures above 50° seemed to account for the low yield and accordingly a number of experimental syntheses were conducted at 38° to 40°, and the time increased to about twenty- four hours, the results of which experiments showed a yield of 85 per cent of the theoretical. This modification was adopted in the preparation of most of the material for the digestion and feeding experiments. The method of synthesis of mannitan distearate as finally carried out was as follows: Ten grams of mannite was dissolved in 200 grams of concentrated sul- phuric acid warmed to 40°, 30 grams of stearic acid stirred in, the flask stoppered and placed in an ordinary incubator at about 38° +2. The mix- ture was shaken from time to time until the stearic acid was completely dis- solved, and then left in the incubator over night. Next day it was poured in a fine stream with stirring into about a liter of cold saturated ammonium sulphate solution and after thorough stirring, set aside for the esters to. separate out; after which it was filtered on a Buchner funnel and washed two or three times with saturated ammonium sulphate solution. The fil- tered mass was pressed as dry as possible, then treated with hot benzol on a water bath. After washing two or three times with hot water to remove any remaining sulphuric acid, the water was siphoned off and the benzol solution was allowed to cool, depositing the mannitan distearate, and retain- ing the excess of stearic acid in solution. After filtering from the benzol the ester was purified as before by precipitation from hot alcohol. Digestion Experiments on Mannid and Mannitan Distearate with Lipases. The following lipase-containing materials were einployed: 1. Pancreas powder from pigs’ pancreas prepared according to Dietz.!° 10 Dietz: Zeilschr. f. physiol. Chem., lii, p. 286. 150 Mannite Esters of Stearic Acid 2. Glycerin extract from pigs’ pancreas preparéd according to Kanitz." 3. Water extract of pancreas—made by shaking the finely divided, fat-free pancreas with water, leaving over-night and using the turbid supernatant liquid. 4. Human pancreatic juice.” 5. Castor bean powder prepared as follows: Large, fully ripe, castor beans were selected, the shells removed and the beans ground as fine as possible in a mortar. The thick paste was trans- ferred to a wide mouthed bottle and extracted over night with a mixture of equal parts of alcohol and ether. Next morning the alcohol-ether mixture was shaken up and after standing a few seconds was poured off into another bottle, and with it the finer portion of the bean powder, which was now allowed to settle out. The coarse portions in the other bottle were drained, removed to a mortar, ground up again and returned to the bottle with the fine settlings from the alcohol-ethermixture. A second ‘extraction was made, this time with ether alone. At the end of the extraction the mixture was well shaken and after standing a few seconds the ether was poured off, carrying with it in fine suspension, most of the bean powder. Theremainder was ground up, and again shaken out with ether as before. Only the fine powder carried out in suspension in the ether was used forthe work. It was filtered free from ether, dried and kept in a.tightly stoppered bottle. The powder so prepared was very active and retained its activity undiminished for several months. This plant lipase differs from animal lipase in that it works best in a weakly acid medium, 4. It also requires the presence of a small amount of acid for activation. The experiments were carried out in loosely stoppered test tubes holding about 50 cc., with the liquids saturated with chloroform, which according to Kikkoji’* best prevents the action of bacteria with the least harm to the enzymes. Care was taken to obtain and preserve a good emulision—the protein of the lipase prepara- tion serving in most cases as the emulsifying agent. Parallel experiments were conducted, using cotton oil, both as a check on 11 Kanitz: Zeitschr. f. physiol. Chem., xlvi, p. 483. 12 Obtained through the kindness of Dr. Benj. R. Symonds from a case of pancreatic fistula in the general hospital at Salem, Mass. An account of the case is contained in the Thirty-fifth Annual Report of the Salem Hospital, 1909, p. 29. 13 Connstein, Hoyer and Wartenberg: Ber. d. deutsch. chem. Gesellsch.. XXXV, p. 3988. 14 Kikkoji: Zettschr. f. physiol. Chem., \xiii, p. 109. W. R. Bloor 151 the activity of the enzyme and for comparison. Blank experi- ments on the reagents alone were carried out with each experiment, under exactly the same conditions, and corrections made accord- ingly. At the end of the time allowed for digestion the tubes were filled with absolute alcohol, shaken to loosen the digestion mixture and emptied into small beakers. The tubes were rinsed out with two tubes full of alcohol and one of ether, the washings added to the liquid in the beakers and the whole titrated with normal alco- holic alkali, using phenolphthalein as indicator. The end point chosen was the first rose color which lasted for one minute. The results of the experiments were as follows: Mannid Distearate (M.P., 61° C.). With human pancreatic juice. (1) Twogramsmannid distearate melted with 5 cc. hot water and shaken until emulsified, cooled (the emulsion remained), mixed with 5 cc. of the human pancreatic juice, again shaken, and kept at 37° C. over night. Titration: 0.75 cc. -¥- alkali: correction for blank = 0.35 cc.; weight of stearic acid set free = 0.11 gram, corresponding to 0.14 gram mannid di- stearate; digestion = 7 per cent. (2) One gram of ester was treated with water as in (1) and to it added 5 ec. pancreatic juice and fifteen drops of ox bile; left over night at 37°C. Titration: 0.90 cc. = alkali: correction for blank = 0.35 cc.; weight of stearic acid set free = 0.16 gram, corresponding to 0.2 gram ester; digestion = 20 per cent. (3) Two grams of mannid distearate, 5 cc. pancreatic juice, 5 cc. water, ten drops 5 per cent soap solution; left over night at 37° C. Titration: 1.30 ce. cc. % alkali: correction for blank = 0.35 cc.; weight of stearic acid set free = 0.28 gram, corresponding to 0.35 gram ester; diges- tion = 17.5 per cent. (A) Experiment with cotton oil to test the activity of the pancreatic juice. Two grams cotton oil, 5cc. water, 5cc. pancreatic juice. The whole well emulsified by shaking and left over night at 37° C. Titration: 4.4 cc. * alkali: correction for blank = 0.4cc.; weight of fatty acid as oleic acid = 1.128 grams, corresponding to 1.17 grams olein; digestion = 58 per cent. (4) A mixture was prepared by melting together 2.5 grams of cotton oil and 5 grams of mannid distearate, in the hope that a mixture of lower melt- ing point would digest better. Two grams of this mixture, emulsified with 5 cc. of pancreatic juice and 5 cc. of water, were kept at 37° C. over night. Titration: 3.35 cc. 4 alkali: correction for blank=0.35cc. Stearicacid = 0.85 gram. 152 Mannite Esters of Stearic Acid If 58 per cent of the cotton oil had digested as in (A) above there would be left about 0.41 gram as originating from the ester, corresponding to an ester value of 0.52 gram, a digestion of 28 per cent. (5) Two grams of the same mixture with 5 cc. of water and 5 cc. pan- creatic juice, shaken to a permanent emulsion, and left overnight at 37°C. Titration: 2.7 cc. -¥ alkali: corrected as in (3) shows a digestion of 0.35 gram of ester or 17.5 per cent. (6) One and four-tenths grams of mixture, 5 cc. of water, 5 cc. of pan- creatic Juice shaken to permanent emulsion and left overnight at 37° C. Titration: 2.0 cc. ¥ alkali: digestion 0.252 gram of ester = 18 per cent. Experiments 1, 2 and 3 indicate that mannid distearate is slowly attacked by human pancreatic juice especially in the presence of bile. The results of Experiments 4, 5 and 6 are of doubtful value as evidence for digestion of the mannid distearate, for if the cotton oil were completely digested instead of the assumed 58 per cent, this showing would be eliminated. Castor bean lipase. (1) One gram of mannid distearate, 0.5 gram of castor bean powder, 4 cc. 75 sulphuric acid, 5 cc. water, shaken to a good emulsion (the protein of the bean is the emulsifying agent in this case) and let stand over night. Titration: 1.3 cc. -¥ alkali —0.4 cc. (blank) = 0.9 ec. = 0.26 gram stearic acid = 0.35 gram of ester; digestion = 33 per cent. (A) Five cubic centimeters cotton oil, 0.5 gram castor bean powder, 4 ec. gy sulphuric acid, 5 cc. water, shaken to good emulsion, let stand over night at 37° C. Titration: 9.0 cc.— 0.4 ce. (blank) = 8.6 cc., * alkali = 2.44 grams oleic acid = 2.54 grams olein; digestion = 55.2 per cent. Mannid distearate digests fairly well with the castor bean lipase. Digestion Experiments with Mannitan Distearaie (M.P., 124° C.) Mannitan distearate when freshly prepared and moist will form an emulsion with water which will pass through a filter while hot, and is permanent on cooling. This material, spoken of as “ester suspension” throughout these experiments, was prepared from the crude salted out mass (see p. 149) in this way: After draining on a filter and washing several times with water, it was stirred with excess of alcohol and let stand an hour or two and filtered. The washing with alcohol was repeated until the washings were no longer W.R. Bloor 153 colored. Washing with water was then resumed and continued until the washings were free from sulphates. The moist substance so obtained, free from fatty acid and salts, was stirred into boiling water until 100 cc. of water contained 20 grams of the ester, cal- culated to dry weight. On cooling it was about the consistency of thick cream and if evaporation is prevented may be kept in this form for several weeks without any separation. With castor bean powder. (A) Test of activity of powder with cotton oil. Forty-eight hours at room temperature (18 to 20° C). Four and six tenths grams (5 ce.) cotton oil, 5 cc. 7) acid, 5 cc. water shaken to good emulsion. Titration: 9.71 ce. — 1.5 ce. (blank) = 8.21 cc. * alkali = 2.35 grams. oleic aicd = 2.47 grams olein; digestion = 53 per cent. (1) Ten cubic centimeters of the ester suspension (containing 2 grams of dry ester) were mixed with 0.5 gram of the bean powder, 2 cc. of 74 sul- phuric acid + 5 cc. of water. Left at room temperature over night. Titration: 1.5 cc. ¥ alkali— blank = 1.5 ce. of alkali; no digestion. (2) Ten cubic centimeters of the ester suspension together with 0.5 gram castor bean powder, 2 cc. 75 sulphuric acid, 4c. water; at room tem- perature for twenty-eight hours. Titration: 8.6 cc. 74 alkali — 3.3 ec. (blank) = 5.3 ec. +, alkali, corre- sponding to 0.14 gram stearic acid or 0.18 gram ester; digestion = 9 per cent. (3) Ten cubic centimeters of ester suspension, 0.5 gram castor bean powder, 2 ce. 75 sulphuric acid, 5 cc. water; forty-eight hours at room temperature. Titration: 11.2 ce. — 9.91 cc. (blank) = 1.29 ce. 7% alkali = 0.037 gram stearic acid = 0.047 gram ester; digestion = 2.5 per cent. These results indicate that mannitan distearate is not saponified by the castor bean lipase. With pancreas powder. (A) Test of the powder with cotton oil: 0.5 gram pancreas powder, 5 cc. water, | cc. of 0.5 per cent NasCO; solution, 5 ec. cotton oil; forty-eight hours at 37° to 38° C. Titration: 105.6 ce. — 31.9 ce. (blank) = 73.7 75 alkali = 2.09 grams oleic acid = 2.174 grams oil; digestion = 46 per cent. (1) 0.5 gram powder, 5 cc. water, lec. of 0.5 per cent NasCOs, 10 cc. ester suspension (2 grams), forty-eight hours at 37 to 38° C. Titration: 38.1 ce. — 31.8 cc. (blank) : 6.3 ec. 4 alkali = 0.193 gram stearic acid = 0.24 gram ester; digestion = 12 per cent. (2) The same amounts of material left six days. Titration: 40.9 cc. — 25.00 ce. (blank) = 15.90 ce. 74 alkali = 0.45 gram stearic acid = 0.56 gram ester; digestion = 28 per cent. (3) The same mixture + 5 cc. of bile left eight days at 37 to 38° C. 154 Mannite Esters of Stearic Acid Titration: 77.74 ce. — 30.45 cc. (blank) = 47.29 cc. 74 alkali = 1.33 grams stearic acid = 1.69 grams ester; digestion = 84 per cent. Mannitan distearate is therefore slowly attacked by the enzyme contained in pancreas powder. Glycerin suspension of pancreas. (A) Preliminary test of glycerin extract with cotton oil. Ten cubic centimeters cotton oil neutralized with 4.5 cc. +5 alkali (previously determined as advised by Kanitz):'° 5 cc. glycerin suspension of pancreas; forty-eight hours at 37 to 38° C. Titration: 226.5 ec. — 37.74 cc. (blank) = 188.76 cc. 74 alkali = 5.36 grams oleic acid = 5.57 grams olein; digestion = 60 per cent. (1) Ten cubic centimeters ester suspension (2 grams), 2 cc. of 7% alkali, 5 ce. glycerin extract. Forty-eight hours at 37 to 38° C. Titration: 52.56 cc. —35.58 ec. (blank) = 16.98 cc. #5 alkali = 0.482 gram stearic acid = 0.61 gram ester; digestion = 30 per cent. (2) Ten cubic centimeter ester suspension, 5 cc. glycerin extract, 5 ce. ox bile, forty-eight hours at 37 to 38° C. Titration: 56.15 cc. — 39.16 ec. (blank) = 16.99 75 alkali = 0.482 gram stearic acid = 0.61 gram ester; digestion = 30 per cent. The mannitan ester digests about one-half as well as cotton oil with glycerin extract of pancreas. Water extract of pancreas. (A) Testing with cotton oil. Five cubic centimeters water extract, 5c. cotton oil, 5 cc. fy Na2CO;: forty hours at 37 to 38° C. ‘ Titration: 9.04 cc. — 2.98 cc. (blank) = 6.06 cc. ¥ alkali = 1.71 grams oleic acid = 1.78 grams olein; digestion = 39 per cent. (1) Ten cubic centimeters water extract of pancreas, 10 cc. ester sus- pension + 10 cc. 7) NasCOs: five days at 37 to 38° C. Titration: 3.81 cc. — 2.98 ce. (blank) = 0.83 cc. ¥ alkali = 0.24 gram stearic acid = 0.30 gram ester; digestion = 15 per cent. 1s Kanitz: Loc. cit. W.R. Bloor 155 The results of the digestion experiments are summarized below. Mannid distearate. | ED a PERCENT- DIGESTIVE AGENT | Aes AGE TIME OF DIGESTION | REMARKS MENT DIGESTION ; 1 a Twenty-four ' Cotton oil under hours ' similar condi- 2 20 Twenty-four | tionsis digested hours(withbile)' to the extent of 3 26 Twenty-four | 58 per cent. hours (with cot-| Human pan- fonsil) ! SreE cure 4 17.8 | Twenty-four _ hours (with | cotton oil) | 5 18 | Twenty-four hours (with | cotton oil) 1 33.2 | Twenty-four _ Cotton oil under Castor bean || hours similar condi- Mipase™ cs tions, 55.2 per | cent digestion. Mannitan distearate. | | 1 None , Twenty-four Cotton oil under hours _ similar condi- Castor bean 2 9 | Twenty-eight | tions 48 hours powder..... | hours | 53 per cent di- 3 2.5 | Forty-eight | gestion. || | hours | | 1 30 | Forty-eight | Cotton oil under Glycerin ex- hours similar condi- tract of pan- 2 30 Forty-eight tions, 48 hours, CYCAS ec. 2: | hours (with 60 per cent di- bile) gestion. 1 15 | Five days Cotton oil, under similar condi- Water extract | tions, 2 days, of pancreas.. | 39 per cent di- / gestion. 156 Mannite Esters of Stearic Acid CONCLUSIONS. 1. Mannid distearate shows, with human pancreatic juice, a digestibility of about one-third that of cotton oil; with castor bean lipase about one-half that of cotton oil. 2. Mannitan distearate does not seem to be attacked by the lipase of the castor bean; but with the various pancreas prepara- tions, a digestibility of from one-fourth to one-half that of cotton oil was obtained. FEEDING EXPERIMENTS. Feeding experiments were made only with the mannitan distear- ate, since it alone could be easily prepared pure and in large quan- tity. No attempt was made to feed the crystallized ester alone, since it has been conclusively shown that even in the case of normal food fats, those of high melting point (as for instance tristearin) are utilized with great difficulty when fed by themselves; but when dissolved in the liquid fats they are well utilized by the animal organism.!6 For this reason the mannitan ester was fed (1) as the ‘ester suspension” used for most of the preliminary digestion work with the pancreas derivatives, and which for these feeding experiments was made thicker so as not to make the food too liquid; and (2) dissolved in cotton seed oil. Mixtures with cotton oil containing different amounts of the ester and varying in consistency from soft lard to hard tallow were used. The crude ester was prepared as described on p. 149 and the actual amount of ester in the sample fed was determined in each case by precipitation from hot alcohol and weighing the dried precipitate. The animals experimented on were cats which were prepared for the experiment by starving for two days before the feeding. In most cases bone ash was given with the food, both to mark the feeding periods and to ensure well formed feces; but to make sure that all the undigested ester was recovered, the feces were collected from the time of feeding till one day after the bone ash had passed. The combined feces were ground in a mortar, then 16 Arnschink: Zeitschr. f. Biol., xxvi, p. 434 W. R. Bloor 157 extracted three or four times with boiling alcohol, which removed all but traces of the undigested ester. The alcoholic extracts were allowed to stand over night, the precipitate collected on a weighed filter, washed with cold alcohol, dried and weighed. Mannitan distearate, as noted on p. 144 is practically insoluble in cold alco- hol, so that this simple procedure gives sufficiently accurate results. EXPERIMENT 1. Young cat, weight 1 kilo, fed 40 grams of a mixture con- sisting of 22 grams hashed lean meat, 30 cc. of the ester suspension (contain- ing 2.44 grams of ester) and 6 grams bone ash. Weight of ester fed = 1.70 grams; recovered from feces = 1.58 grams; retained 0.12 gram, or 7 per cent. 7 The ester suspension is not well utilized and the other experi- ments were made with the ester dissolved in cotton seed oil. EXPERIMENT 2. The same cat fed the whole of a mixture consisting of 3.5 grams ester melted with 3 grams of cotton oil (yielding a tallow-like product), 12.7 grams of hashed lean meat, 3 grams bone ash and 1 cc. of blood (to increase the palatability). Weight of ester fed = 3.5 grams; recovered from feces = 2.04 grams; retained 1.46 grams, or 41.7 per cent. EXPERIMENT 3. A full grdwn male cat, weight 3.5 kilos was fed 64.5 grams of a mixture consisting of 3.8 grams of ester dissolved in 10 cc. of cotton oil (the solution when cold had the consistency of soft lard and melted readily in the fingers), 50 grams hashed lean meat, 10 grams bone ash, 3 grams blood—in all 76.8 grams. Ester fed = 3.16 grams; recovered from feces = 2.00 grams; retained 1.16 grams, or 36.7 per cent. EXPERIMENT 4. Prolonged feeding of ester, using the same cat as in Experiment 3. In this experiment the ester was fed to the cat every aay for six days. Feces were collected daily at 10 a.m. and the amount of unused ester deter- mined. To find out whether the bone ash had any effect on the amount of ester absorbed it was omitted on certain days. As may be seen from the results, the bone ash seemed to aid absorption, since the bone ash feces con- tained as a rule less ester than the others. Aside from the bone ash, the food was the same every day and consisted of 3:8 grams of ester in 5 grams cotton oil, 50 grams hashed lean meat, 4 cc. of blood and on the days when it was fed 6 grams of bone ash. First day—Fed as above with bone asn. Second day—No bone ash. Feces collected weighed 30 grams containing 1.9 grams of unabsorbed ester. BETES TH LSTVITTETe | Ae ee Anite 2 1.9grams 158 Mannite Esters of Stearic Acid Third day—6 grams bone ash. Feces collected weighed 22.4 grams containing 1.69 grams of unab- sorbed ester. Eater retammpaiea. 6& ...: > hics 0 See ee ee ee 2.11 grams Fourth day—No bone ash. Feces collected—14.5 grams, contained 2.27 grams of ester. ister retainediees:: oo. Sek eee Moen ae re eee ee 1.53 grams Fifth day—No bone ash. Feces collected 16.5 grams contained 1.90 grams of ester. Bstenregnumedss 2022 0:5 ee SPS eee ee 1.9 grams Sixth day—Bone ash in feed. Last feeding day. No defecation. Seventh day—Feces not collected until next day. Eighth day—Feces collected, weight 35.6 grams containing 3.9 grams esters. Ester retained from two days feeding, 3.7 grams. Ninth day—Feces collected, weight 13 grams. Contained no ester, therefore elimination is completed. To recover any ester which may not have been previously extracted, the extracted feces were combined and boiled out several times withsmall portions of alcohol. The extracts were combined cooled and the preci- pitates weighed. There was recovered in this way a total of 1.5 grams of ester from the week’s feces. The balance for the six days was as follows: Totalesterfed:. =\6.X N vA as co Mi ~% wetkgersil| a < ' si ter of cH, 4 ss real ire 34 is 7 of yas Sak NE : 5 = y- P i _ aa ~ + pase, Seat mre Be Shes ta ‘. 7 7 a van soe 2 ‘Prib, dau Cs phe hid ON THE REFRACTIVE INDICES OF SOLUTIONS OF CER- TAIN PROTEINS. VI. THE PROTEINS OF OX-SERUM; A NEW OPTICAL METHOD OF DETERMINING THE CONCENTRATIONS OF THE VARIOUS PRO- TEINS CONTAINED IN BLOOD-SERA. By T. BRAILSFORD ROBERTSON. (From the Rudolph Spreckels Physiological Laboratory of the University of California.) (Received for publication, February 6, 1912.) 1. INTRODUCTION. The influence which is exerted by the various proteins of blood- serum upon the refractive indices of their solutions was first systematically investigated by Reiss... This observer fractionated the proteins of horse and human sera in the following way: The globulins were precipitated by ammonium sulphate, “ Euglobulin” being precipitated by 32 to 36 per cent saturation of the serum with ammonium sulphate, “‘Pseudoglobulin I” by 36 to 39 per cent saturation, and ‘‘Pseudoglobulin II” by 42 to 50 per cent saturation. Each of these fractions was dissolved in distilled water and reprecipitated twice, the final solutions thus obtained being purified by prolonged dialysis (four to six weeks) until the water outside the dialysor contained no trace of sulphates. The euglobulin fraction proved to be insoluble in distilled water, that is, it was precipitated from its solution on dialysis. For the pur- pose of refractometer measurements it was dissolved in a dilute salt solution and the refractive indices of the globulin solution and of a salt solution of the same concentration were separately deter- mined. The difference between the two refractive indices afforded 1E. Reiss: Beitr. z. chem. Physiol. u. Path., iv, p. 150, 1903; Arch. f. exp. Path. u. Pharm.., li, p. 18, 1903. 179 180 Refractive Indices of Proteins of Ox-Serum a measure of the effect of the globulin upon the refractive index of its solution. The quantity of globulin in each of the solutions employed was estimated by precipitation with alcohol and drying and weighing the precipitate thus obtained. In this way the effect due to 1 per cent of each of the proteins in question can readily be computed. The following were the results obtained by Reiss, employing solutions of the various globulins defined above: One gram of ‘“Euglobulin’’ changed the refractive index of 100 cc of solution by 0.00230. One gram of “ Pseudoglobulin I’’ changed the refractive index of 100 cc. of solution by 0.00224. One gram of ‘‘ Pseudoglobulin II’ changed the refractive index of 100 cc. of solution by 0.00230. As Reiss himself points out, the differences between these figures are within the experimental error of the estimate, and from his results we may conclude that 1 gram of any of the globulins of serum, when dissolved in 100 ce. of water or dilute salt solution, changes the refractive index of the solvent by about 0.00230. In a previous communication? I have extended and confirmed the results of Reiss in so far as they apply to the “ Euglobulin”’ or the globulin fraction which is insoluble in distilled water. I pre- pared euglobulin by precipitation from diluted serum withCO, and, after careful purification, dissolved carefully estimated amounts in measured volumes of dilute KOH. In this way I found the value of a ( =change in the refractive index of the solvent due to 1 gram of protein) for this protein to be* 0.00229 +0.00024. This value of a is, within the experimental error, identical with that obtained by Reiss for each of his different globulin fractions. We may therefore assume that in all probability the value of a for the globulins of serum, irrespective of the mode of precipitation, to be that which I have just cited. In measuring the value of the constant a for the albuwmins of blood-serum, Reiss proceeded as follows: The filtrate obtained by half-saturating serum with ammonium sulphate and filtering off the precipitated globulins was acidified 2 T. Brailsford Robertson: This Journal, viii, p. 441, 1910. 3 T. Brailsford Robertson: Die physikalische Chemie der Proteine, Dresden, p. 323, 1912. T: Brailsford Robertson. 181 by the cautious addition of i1,80, and the precipitate which was thus produced was allowed to stand until it became crystalline.* This substance after separation from the mother liquor, was redissolved and recrystallized twice and finally dissolved in dis- tilled water and dialyzed for from four to six weeks until salt-free. The value of a for this substance (‘‘crystalline serum-albumin’’) proved to be 0.00201. The mother-liquor, after the deposition of the ‘crystalline serum-albumin” contained another protein characterized as “amor- phous serum albumin.” This solution was dialyzed for from four to six weeks and the value of a for this protein was determined as in the previous cases; it proved to be 0.00183. Continuing his investigations, Reiss arrived at the remarkable conclusion that the specific refractivity of the mixed proteins in serum is actually less (0.00170 to 0.00175) than the specific refrac- tivities of any of the constituent proteins of the mixture, thus rendering it impossible to estimate from the value of a for thé mixed proteins in blood-serum the relative proportion of globu- lins and albumins contained therein. This appeared to me to be a very important conclusion, meriting further investigation, for the following reasons: I have found® that that change in the physical and chemical condition of a dissolved protein which immediately precedes coag- ulation, and may be induced by the addition of a coagulating (dehydrating) agent to the solution, is accompanied by a decrease in the specific refractivity of the protein. Now there is much reason for believing that the proteins in circulating blood or in unaltered blood-serum are not merely present therein in the form of a mixture, but in the form of a chemical complex possessing recog- nizably different physical and chemical properties from those of the constituent proteins out of which it is built up.6 It appeared 4 Hofmeister: Zeitschr. f. physiol. Chem., xiv, p. 163, 1889; xvi, p. 187, 1891; A. Giirber, Sitz. d. physik. med. Ges. zu Wiirzburg, p. 143, 1894; cited after Schulz, Die Krystallisation von Eiweissstoffen, Jena, p. 13, 1901: A. Michel: Verh. d. physik. med. Ges. zu Wiirzburg, xxix, 3, p. 28, cited after Schulz: loc. cit., p. 13; H. T. Krieger: Dissertation, Strassburg, 1899, cited after Schulz: loc cit., p. 11. 5 T. Brailsford Robertson: Die physikalische Chemie der Proteine, Dresden, 1912, chapters 10. and 13. §T. Brailsford Robertson: ibid., pp. 126-133. 182 Refractive Indices of Proteins of Ox-Serum possible that in the building up of this complex the individual proteins composing it might sustain a loss of refractive power and that the phenomenon observed by Reiss might be susceptible of this explanation. Accordingly, the following investigations were undertaken. 2. EXPERIMENTAL. In endeavoring to measure the refractivity of the mixed proteins in blood-serum itself we are confronted with the difficulty of deter- mining the refractivity of the solvent in which these proteins are dissolved, that is, the proportion of the difference between the refractivity of the serum and that of distilled water which is due to the non-protein constituents of serum. For reasons which will be referred to later, I did not consider the estimates of the refractivity of the non-protein constituents of blood-serum which were adopted by Reiss to be wholly satisfactory. Accordingly I preferred to assume, until further evidence should demonstrate that assumption to be false, that the solvent in which the proteins of ox-blood serum are dissolved may be regarded, for the purposes of refractometer measurements, as being essentially % sodium chlo- ride.’ As we shall see, the results of my measurements show that this assumption is justified. Fresh ox-blood serum was prepared by whipping and centri- fugalizing the freshly-drawn blood.’ This serum was diluted with varying proportions of distilled water. The refractive indices 7 The refractive powers of dilute equivalent solutions of the chlorides of the mineral bases found in serum are nearly equal. Since only traces of KCl and CaCl, are present in serum we may safely take the refractivity of the saline constituents of serum to be that of ¥ NaCi. In assuming that the refractivity of the non-protein constituents of serum is likewise identi- cal with that of a NaCl solution we are assuming that the fats, sugars, etc., which are normally present only in small amounts in the serum derived from systemic blood, take only a negligible part in determining the refrac- tivity of serum. 8 A source of error which may possibly invalidate some estimates of the concentrations of solid constituents in blood obtained from slaughter-houses may be pointed out here. When blood is allowed to clot and the clot is left in the cold-chamber of a slaughter house to contract and express serum very considerable evaporation may occur under the condition of dessica- tion which is necessarily maintained in such rooms. Serum thus prepared may be found, on analysis, to contain from 10 to 12 per cent of proteins. T. Brailsford Robertson 183 of these solutions and of distilled water and of % sodium chloride solution were measured by means of a Pulfrich refractometer at about 20° C., employing a sodium flame as the source of light. The results which were obtained are expressed in the accom- panying table. The value of ™ ( =refractive index of the sol- vent in which the serum-proteins are dissolved) is calculated upon the assumption that the non-protein portion of serum may be regarded, for the purposes of refractometer measurements, as being essentially % NaCl and upon the further assumption (which I have experimentally verified) that the difference between the refractive index of a sodium chloride solution of this or lower con- centrations and that of distilled water is directly proportional to the concentration of the sodium chloride solution: TABLE 1. i | 2= REFRACTIVE a | SOLUTION INDEX OF INDEX OF n—n, SOLUTION SOLVENT Distillediwater) 2022.-..6..-5.....% 1.33410 MPI AC RE EDN lot ea no os ance wisn 1.33567 1.33410 0.00157 5 ce. of serum + 20 cc. of water... 1.33759 1.33441 0.00318 10 ce. of serum + 15cc. of water... 1.34139 1.33463 0.00666 15 cc. of serum + 10 ce. of water... 1.34511 1.33504 0.01007 20 cc. of serum + 5 cc. of water... 1.34872 1.33536 0.01336 25 ce. of serum + Occ. of water... 1.35224 1.33567 0.01657 Dividing each of the values of n—m tabulated above by the number of cubic centimeters of serum in 25 cc. of the corresponding mixture of serum and water we obtain: TABLE 2. n—m SOLUTION CUBIC CENTIMETERS OF SERUM IN 25 cc. SOLUTION 5 ec. of serum + 20 cc. of water ................ 0.00064 10 ce. of serum + 15 cc. of water ........... .... 0.00067 15 cc. of serum + 10 ce. of water ................ 0.00067 20 ce. of serum + 5cc. of water................ 0.00067 25 cc. of serum + Occ. of water................ 0.00066 184 Refractive Indices of Proteins of Ox-Serum From the constancy of this ratio, since, of course, each cubic centimeter of serum contains the same quantity of serum-protein, we may conclude that the value of a for the mixed serum proteins is independent of the dilution. In order to determine, from these results, the absolute value of a for the mixed serum proteins it was only necessary to determine the percentage of total proteins which was contained in the serum under investigation. This determination was performed in the following manner: Two samples of the serum, measuring respectively 2.95 and 3.00 cc. were accurately delivered, drop by drop, into a small hardened nitrogen-free filter-paper (5.5 cm. diameter) which was at the same time kept filled with absolute alcohol. The filters and contained protein were then washed in alcohol and ether and dried for two or three hours at 40°. _They were then analyzed for nitrogen by the Kjeldahl method. From the quantity of nitrogen thus found, the percentage of protein in the original serum was calculated on the assumption that the nitrogen-content of the serum- proteins is 15.9 per cent.’ The results follow: PERCENTAGE OF PROTEINS SAMPLE N IN SAMPLE aay SITE ce. mg | 2.95 39.5 8.4 3.00 41.0 8.6 Meanpemae As he TR by Beas ore 8.5+0.1 From this result the concentration of serum-proteins in each of the mixtures enumerated in Table I can readily be computed. We can compute the value of a from those of n—mn, in Table I by adding together all of the observed values of n—7, and dividing this sum by the sum of the concentrations of protein in the mix- tures employed.!° The possible error of this estimate may be computed if we recol- lect that each determination of the angle of total reflection is 90. Hammarsten: Arch. f. d. ges. Physiol., xxii, p. 431, 1880; E. Abder- halden: Zettschr. f. physiol. Chem., xxxvii, p. 495, 1903. 10 For the rationale of this procedure consult T. Brailsford Robertson: This Journal, viii, p. 507, 1910. _—_—<— SS T. Brailsford Robertson 185 liable to an error of = 1’ corresponding, for solutions of these refractivities, to an error of between + 0.00008 and + 0.00009 in the determination of n—7,. Proceeding in this way we find that for the serum proteins dissolved in diluted or undiluted serum the value of a ( =change in the refractive index of the solution due to 1 per cent of protein) is 0.00195 = 0.00002. This result, it will be observed, is very different from the above- cited result obtained by Reiss (a =0.00172). The value of a for the mixed proteins, instead of being less than that for any of its constituents, would appear, as might have been expected, to be intermediate in magnitude between the value of a ( =0.00183) for serum-albumin and that of a ( =0.00229) for serum globulins. The protein-complex into which the individual proteins of unaltered serum would appear to be built up is decomposed by acids; consequently, it appeared of importance to ascertain whether the refractivity of the mixed serum-protein is changed by the communication of an acid reaction tothe serum. Accordingly, samples of the same serum as that employed in the experiments cited in Table I were diluted with 4 hydrochloric acid instead of with distilled water. About 12.5 cc. of % acid suffice to communi- cate a neutral reaction to 100 cc. of undiluted serum; hence all of the mixtures of serum with + HCl which were employed in these experiments were acid in reaction. The values of m ( =refrac- tive index of the solvent) are calculated in the same way as those in Table I, with the aid of the further assumption (the truth of which I have experimentally verified) that the change in the re- fractive index of water due to the addition of HCl is proportional, within the limits of concentration considered, to the concentration of the HCl. The following were the results obtained: 11W. B. Hardy: Journ. of Physiol., xxxiii, p. 251, 1905 (appendix) ;T. Brailsford Robertson: Die physikalische Chemie der Proteine, Dresden, 1912, pp. 126-133. > 12 T. Brailsford Robertson: This Journal, vii, p. 351, 1910. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 3 186 Refractive Indices of Proteins of Ox-Serum TABLE 3. n= REFRACTIVE|721 = REFRACTIVE SOLUTION INDEX OF INDEX OF | am—nN1 SOLUTION SOLVENT mint aa teers ee Tis Distilled*watenwees~. ... fic. 8 | 1.33410 NaC) eee. See | 1.33567 1.33410 0.00157 ToL CL ecient so | 1.83481 1.33410 0.00071 5 cc. of serum + 20 cc. of 7 HCl.| 1.33815 1.33498 | 0.00317 10 ce. of serum + 15 cc. of zo HCl. 1.34155 1.33515 | 0.00640 15 ce. of serum + 10 cc. of 4} HCl.) 1.34519 1.33533 | 0.00986 20 ce. of serum + 5 cc. of 7) HCl.) 1.34881 1.33550 0.01331 25 ce. of serum + Occ. of #y HCl., 1.35224 1.33567 0.01657 Dividing each of the values of n— n, tabulated above by the num- ber of cubic centimeters of serum in 25 cc. of the corresponding mixture of serum and 34 acid we obtain: TABLE 4. nm—-mM1 SOLUTION CUBIC CENTIMETERS OF SERUM ; IN 25 cc. SOLUTION 5 cc. of serum + 20 cc. 7) HCl.........:....... 0.00063 10 ce. of serum + 15 cc. 4 HCl. ............... 0.00064 15 cc. of serum + 10 cc. 4 HCl.........2........ 0.00066 20 cc. of serum + 5cc. 4; HCl................. 0.00067 25 cc. of serum + Occ.) HCl................. 0.00066 From the constancy of this ‘ratio we may conclude that acidifi- cation of serum does not alter the refractivity of the serum proteins. Computing, from the results cited in Table 3, the value of a for the mixed proteins in acidified serum we obtain: a = 0.00193 + 0.00002, a result which is identical within the experimental error, with that obtained for the proteins in neutral serum. In seeking for a reason for the discrepancy between my results and those cited by Reiss the initial assumption upon which my estimates of a are based, namely that the non-protein constituents of serum may, for the purpose of refractometer measurements, be regarded as % NaCl, at once suggests itself as a possible source of error. Reiss’ estimates of the refractivity of the non-protein constituents of serum (0.00270 to 0.00292 in different experiments) T. Brailsford Robertson 187 is much higher than mine (0.00157, cf. Tables 1 and 3). If non- protein constituents other than mineral salts occur in serum in sufficient amounts to appreciably influence the refractivity of the serum, then their refractivity would be added to that of the pro- teins in my estimates and the values of a obtained above would be too high. Accordingly, the following experiments were under- taken: Four samples of serum were taken of which the first (sample 1) measured 100 cc. and the remainder (samples 2, 3 and 4) 80 cc. each. The proteins from sample 1 were precipitated immediately by the addition of 10 volumes of absolute alcohol. To sample 2 were added 20 ce. of 7 HCl and the proteins were then immediately precipitated by the addition of 10 volumes of absolute alcohol. To samples 3 and 4 were added 20 cc. of 74 acetic acid and 7y hydrochloric acid, respectively. These mixtures were allowed to stand at room temperature for about two hours before precipitating the pro- teins from them by the addition of ten volumes of absolute alcohol. These precipitates were washed three times in the volume of absolute alcohol originally employed for the precipitation and then three times in the same volume of ether, the precipitates being allowed to settle after each washing and the supernatant fluid drawn off by means of a syphon. The thick suspensions of protein in ether which were finally obtained were dried at 40° over sulphuric acid for twenty-four hours, pulverized and sifted and then dried over sulphuric acid at room temperatures for over three weeks. The solutions of these proteins in distilled water are not sufficiently transparent for refractometer readings. On adding a little alkali, however, they at once clear up, i per cent solutions being of a clear pale yellow color. Accordingly solutions were prepared in the following manner: Three grams of each of the above preparations were separately dissolved in 300 ce. of X KOH, by first stirring up the protein with a little of the solvent until it formed a paste and then adding the remainder of the solvent and stirring rapidly for about an hour. One hundred cubic centimeters of each solution were diluted to 200 cc. making 0.5 per cent solutions. The refractive indices of these solutions were determined by means of the Pulfrich refractometer at 20° C., employing a sodium flame as the source of light: 188 Retractive Indices of Proteins of Ox-Serum TABLE 5. Pieaain eV Seis SOLUTION n = pabtn Mal eS i iim SOLVENT DUE ae be EAb ae OF PROTEIN oN ey KOH 8 5. oo ees peter eee ees 1.33426 Oier tant Pe! Semple.” ° rgasarf | 0-00180+0,00011 Ces por sont PAGEL of Seana raseaty | 0-00189-+0.00011 Ooiter esnt Ree ot Saapls S08 i gaeoiy | 0-00189-+0.00011 ise "crip s yaya meer atece Men BITE NECA ot The values of a enumerated in the third column of this table are identical with one another and also, within the experimental error, identical with those determined above for the proteins in unaltered, diluted, or acidified serum. Now from the method of preparation it is obvious that these proteins must have been free from appreciable non-protein contamination other than, pos- sibly, small amounts of inorganic bases.!* We may therefore conclude (1) That acidification of serum does not alter the refrac- tivity of the serum proteins; (2) That the value of a for the serum pro- teins 1s 0.00195 + 0.00002; (3) That the refractivity of the non-pro- tein constituents of serum may be regarded, without introducing any appreciable error, as that of a % sodium chloride solution; (4) That Reiss’ estimates of the refractivity of the non-protein constituents of serum are nearly 100 per cent too high. His excessive estimates of the refractivity of the non-protein constituents of serum are sufficient to account for the low value of a for the mixed proteins which was obtained by Reiss. It is more difficult to assign a cause for his excessive estimate of the refrac- tivity of the non-protein constituents of serum, since, in the arti- cles to which I have referred, he does not specifically describe how he arrived at this estimate. From the context“ however, 18 Cf. T. Brailsford Robertson: This Journal, vii, p. 351, 1910. 144 In his second paper (Arch. f. exp. Path. u. Pharm., li, p. 20, 1903) he thus describes his method of determining the value of a for the mixed pro- teins in horse-serum: ‘‘Eine zweite Bestimmung wurde in etwas anderer T. Brailsford Robertson 189 it appears that he estimated the refractivity of the non-protein constituents of serum by removing the proteins with the aid of heat-coagulation and determining the refractivity of the protein- free fluid which was thus obtained. This procedure is obviously open to the objection that during the heating of the protein solu- tion changes of unknown magnitude (such as hydrolyses, etc.) may occur, leading to the formation of substances not normally present in serum and affecting the refractivity of the fluid. Fur- thermore, it is possible that in the blood-serum substances of a mucoid or proteose-like nature occur which are not coagulable by heat. In estimating the value of @ for “amorphous serum albumin” by Reiss’ method such proteins would be present in the solution and the observed refractivity (since they would be pre- cipitable by alcohol) of “amorphous serum albumin” would, in reality, be the sum of the refractivities of serum albumin and the mucoid or proteose. Such substances, however, would not be coagulated by heat and their refractivity would, employing Reiss’ method, be estimated along with that of the non-protein constit- uents of serum. I have also measured the refractivity of the serum-albumin or albumins in the filtrate obtained by adding an equal volume of sat- urated ammonium sulphate solution to serum and filtering off the globulins thus precipitated. In this experiment we are confronted Weise an Pferdeblutserum vorgenommen. Eine genau abgeniessene Menge desselben wurde auf etwa das zehnfache verdiinnt, die ausgefallenen Globu- line durch abgemessenen Zusatz von Natriumkarbonat wieder zur Losung gebracht und aus den Brechungskoeffizienten des nativen und des ver- diininten Serums unter Abrechnung der Lichtbrechung des zugefugten Natri- umkarbonats der Verdiinnungsgrad berechnet. Sodann wurde ausprobiert, wieviel Tssigsiure einen bestimmten Quantum des verdiinnten Serums zugesetzt werden musste, um das Eiweiss in der Hitze zum volligen Aus- fallen zu bringen. Eine dementsprechend gestaltete Mischung wurde im zugeschmolzenen Glasréhrchen etwa 10 Minuten auf 100° erhitzt. Nach einigem Stehenlassen wurde das Réhrchen ge6ffnet und ein Tropfen der obenstehenden Fliissigkeit—der nur noch Spuren Eiweiss enthielt—refrak- tometrisch untersucht. Der Eiweissgehalt des nativen wie des verdunnten Serums wurde durch Wigung (Fallen mit Alkohol, einstundiges Erhitzen auf 80°) bestimmt. Aus den so erhaltenen Zahlen berechnete sich der Anteil des Brechungskoeffizienten: Fur 1 proz. Eiweiss auf 0.00175, fiir die Nichteiweisskérper des gesamten Serums auf 0.00292.” 1 Cf. O. Hammarsten: Ergeb. d. Physiol., i, Abt. 1, p. 354, 1902. 190 Refractive Indices of Proteins of Ox-Serum by the difficulty that the refractivity of serum-albumins dissolved in half-saturated ammonium sulphate may not necessarily be the same as their refractivity in distilled water or in serum, since, as I have shown elsewhere," the refractivity of a protein, upon succes- sive additions of a coagulating (dehydrating) agent to its solution tends to diminish some time before manifest coagulation occurs. It occurred to me, however, that if this were the case with serum- albumin dissolved in half-saturated ammonium sulphate, then the value of a should alter upon dilution of this solution and finally approach its value in distilled water; accordingly the following experiments were undertaken: To 250 ce. of fresh centrifugalized ox-serum were added 250 cc. of satu- rated ammonium sulphate solution, and the mixture was filtered through hardened filter paper. The entire process of filtration occupied about three and one-half hours, the filter being changed from time to time as it became clogged, in order to secure as rapid filtration as possible and thus avoid con- centration of the filtrate through evaporation. Samples, measuring 25, 33.3, 50, 66.7 and 75 cc., respectively, of this filtrate were diluted to 100 cc. by the addition of distilled water.!7_ It will be observed, therefore, that the proteins in the serum were diluted, in each mixture, to exactly the same extent as the saturated ammonium sulphate solution. To 250 ce. of distilled water were similarly added 250 cc. of saturated ammonium sulphate and samples measuring 25, 33.3, 50, 66.7, and 75 cc. of this mixture were similarly diluted to 100 ce. The refractive indices of these serum and ammonium sulphate solutions were measured at 24° C. inaPulfrich refractometer, employing a sodium flame as the source of light. The following were the results obtained: 16 Cf. Previous communications of this series, This Journal; also Die physikalische Chemie der Proteine,’’ Dresden, 1912, chapter 13. 17It is necessary thus to specify exactly the methods of dilution em- ployed, on account of the volume-change which occurs on diluting strong ammonium-sulphate solutions. T. Brailsford Robertson IQl TABLE 6. Pas T (n—n1) X DILUTION OF a SOLUTION n LW THE ORIGINAL SERUM Distilled water........ 1.33364 4 Saturated Am.SO,...| 1.34419 Seralbumins dissolved 0.00134+0.00008 | 0.01072+0.00064 ANVAD OVC Wee sec ayo. 1.34553 2 Saturated Am,SO,...| 1.34753 Seralbumins dissolved | in above.........«...| 1.34940) 1 Saturated Am2SO,... Bo) 0.00187+0.00009 | 0.01122+0.00054 Seralbumins dissolved 0.00279+0.00009 | 0.01116+0.00036 IMeAWOVes.2--........| 1.35667 3 Saturated Am,SO,... | 1.35993 | Seralbumins dissolved + | 0.00376+0.00009 | 0.01128+0.00028 iueAUOWe. o..... ... 1.36369 | 3 Saturated Am.SO,...) 1 oa | Seralbumins dissolved 0.00416+0.00009 | 0.01110+0.00024 OWE. skye Le >: . 1.36695 | 3 Saturated Am.SO,...| 1.37098 Seralbumins dissolved 0.00483+0.00009 | 0.00966+0.00018 IMPADOWEH-Aeeen a x. | Lo ol0Sl | The average of the first five values of the product (n—m) X dilu- tion of the original serum is: 0.01110 + 0.00042, all of them being identical within this experimental error, while the value of this product for the serum albumins dissolved in one-half saturated ammonium sulphate is considerably less. We may, therefore conclude that the refractivity of the serum albumins dissolved in three-eighths saturated or more dilute solutions of ammonium sulphate is independent of the dilution, and, consequently, that the value of a for serum albumins dissolved in three-eighths saturated or more dilute solutions of ammonium sulphate is the same as its value for serum albumins dissolved in distilled water or in serum. The total refractivity of the serum albumins plus the non-pro- tein constituents in the ox-serum employed is, therefore, 0.01110 += 0.00042. The total refractivity of the proteins plus the non-protein con- stituents in ox-serum!® is 0.01815 + 0.00017. 18 Cf. the value for a for the mixed proteins of serum and the value of n—n for -¥ NaCl in Table 1. 192 Refractive Indices of Proteins of Ox-Serum The difference between these two estimates is the total refrac- tivity of the globulins in serum, namely: 0.00705 + 0.00030. Hence the percentage of globulins in ox-serum? is: 705 30 299.7 3.1 + 0.1 The total concentration of proteins in the ox-serum employed was, as we have seen, 8.5 = 0.1. Hence the concentration of the albumins was 8.5 + 0.1—3.1 + 0.1 = 54 + 0.1. The total refractivity of the serum albumin plus the non-pro- tein constituents of the ox-serum employed was, as we have seen: 0.01110 += 0.00042. Hence the total refractivity of the serum albumins alone was: 0.01110 + 0.00042—0.00157 = 0.00953 + 0.00042. Hence the value of a for the serum albumins is: 0.00953 + 0.00042 eal 01 = 0.00177 + 0.00008 an estimate which is, within the experimental error, identical with Reiss’ estimate of a for “amorphous serum albumin.’”?° This result is very striking, for Reiss, by adding acid to his ammonium sulphate solution of the albumins of horse-serwm and allowing the precipitate thus produced to stand, was able to sepa- rate the serum albumins into two fractions; the one, the “‘crystal- line” fraction possessing a much higher value of a (= 0.00201) than the other, the ‘‘amorphous”’ fraction (0.00183). Yet my estimate of the value of a for the mixed albumins is identical with or even slightly lower than Reiss’ estimate of the refractivity of the ‘‘amorphous”’ fraction.24 It would appear, therefore, that erystallizable serum-albumin does not exist in appreciable quan- tity in ox-serum, a conclusion which finds striking confirmation in the fact that Krieger? was unable to obtain crystalline albumin from ox-serum. 19 Cf. estimate of a for serum-globulins in Introduction. 20 Cf. Introduction. 71 Reiss himself states that his preparation of ‘‘amorphous’”’ serum albu- min was probably contaminated with ‘‘crystalline’’ serum-albumin. 22 H. T. Krieger: Dissertation, Strassburg, 1899, cited after Maly’s Jahres- bericht, p. 14, 1899. T. Brailsford Robertson 193 From the above data we may conclude that the percentages of the globulins and albumins in the ox-serum employed were as follows: (ti GL 200) U0 11 i ella a eae ee vy del ee 3.1 + 0.1 per cent Eeeumatemisnms: = 2. s,s 8 5.4 + 0.1 per cent I have also determined the percentage of the “insoluble” or CO.-precipitable globulin (euglobulin) in ox-serum in the following way: Three samples of fresh centrifugalized ox-blood serum, measuring 100 cc. each, were diluted with distilled water to 1000 cc. and CO, was bubbled through the mixtures at a quick rate. The time of passage of the CO, was purposely varied among the three samples, the least time of passage being 1 hour and the longest 2 hours. The precipitate of insoluble globu- lin which was thus produced was allowed to settle in tall glass cylinders and the supernatant fluid drawn off by means of asyphon. The precipitate was then washed in a liter of distilled water; this washing was repeated.2?7 The subnatant precipitates were then dissolved by the addition of 10 cc. of KOH and the volumes made up ineach case tol100cc. The refractive indices of these solutions were measured in the usual manner with the following results: TABLE 7. SOLUTION n n—n he Pass 0.00229) mse KOH 1.33387 1 1.33560 0.00173 +0.00008 2 1.33560 0.001730 .00008 0.76+=0.04 3 1.33560 00.0173+0.00008 J Hence the percentages of the various proteins contained in the ox-serum employed were as follows: 23 The error introduced by contamination with the other proteins of serum must, after this washing, have been negligible, as may readily be calculated in the following way: The total refractivity of the substances dissolved in serum is 0.01815. The volume of the subnatant suspensions obtained in the above processes of washing was always less than 100 cc., hence after dilution and two washings the refractivity of the dissolved substances in the fluid in which the CO, globulin was suspended must have been less than 1/1000 of 0.01815, that is, less than 0.00002. The experimental error in the determination of the refractivity is + 0.00008. 194 Refractive Indices of Proteins of Ox-Serum “Insoluble? globulins. . 2... 6 Pee a. eee 0.76+0.04 *Solublue?? 1G HWlans:......: : C-H;CH» COOH gms. benzoic acid inlce. 7 solution) j ~~ Ss ae x=weight of phenyl acetic acid. Weight of sublimate — x = weight of benzoic acid. To test the applicability of the method 1 gram of phenaceturic acid was added to 100 ec. of urine, of which the hippuric acid con- tent was known, and the general method outlined for hippuric acid followed. The table shows the results obtained on an aliquot corresponding to 25 ce. of urine and 0.25 gram added phenace- turic acid. TITRATION OF | WEIGHT OF SUBLIMATE IN WEIGAT | PHENYL- WEIGHT OF WEIGHT OF WEIGHT OF | CUBICCENTI-|BENZOIC ACID ACETIC ACID BENZOIC ACID PHENYL- SUBLIMATE METERS FROM FROM PHEN- ACETIC ACID N NaOH. HIPPURIC ACID) ACETURIC Bs EON BY TITRATION 30 a ACID Atsibe ei 0.3726 58.15 0.2030 | 0.1761 0.1988 0.1738 Be vaa8) ss. 0.3700 57.75 0.2030 0.1761 0.1981 0.1719) | While the above values agree remarkably well it must be remem- bered that the method is not without its limitations. The differ- ence in molecular weight between benzoic and phenyl acetic acids is small and any impurities in the sublimate will materially affect the final values; a variation of 0.1 cc. of 7 NaOH will mean a difference of 0.0068 gram of phenylacetic acid or, when multiplied 26 Mitteilungen d. land. Instii. Breslau, iv, 1909. H. Steenbock 209 by the conversion factor 1.4191, 0.0096 gram of phenaceturic acid. However, it is believed that this method for phenaceturic acid is far superior to fractional crystallization which up to the present time has been the only one in use. Where examination for unconjugated benzoic acid in the urine is to be made the method outlined was found entirely applicable, since phenaceturic like hippuric acid is practically insoluble in benzol. SUMMARY. Dakin’s method for isolating benzoic acid was found to yield quantitative results when followed by sublimation. Hippuric acid and phenaceturic acid occurring together can be determined respectively as benzoic and phenylacetic acid by sublimation followed by titration. No salts of non-conjugated benzoic acid or of phenaceturic acid were found in cow urines. aed ° ie igs 4, >> ee ont . it ‘tae ee ’ a Sagat ed ere HiPDe, R premenne cere = e et Bis an a ee atte er Giese a ee aera -; «oii ee So! sapghlppacpiapals eaten ed tae ve honbenct aus : “sate el Dia beset aul : “Si eR: “Lp blige a ete ano ‘i eR tit We at. 3 eitetbend Hani Bae dS a2 7 a a Hn ile als Yo ssum ; é —? ee eae LTS Ge tpl = heh ~~ _ TEARS ere i inate: 4g} AGH sig ? 6 agmatine oo iat Dalat) Bae RU ER : ae in ee ie PECAN gta... “ alee PPPS alel s | ea Fos, Sense Be oY! Seat Pda hearts i —- Nasoitavi ih pewobop hor i : : ila: af iF. a5 eee! baiegnincieaiog: Ri ait be sae ¢ ya : : € y 2 }, VEG Shs: Peet : *: WES - i oe. mine aa. VED, pe : = 0 - a ae gf rn % Sy Sn i » ye “. 4 - ae agree ree D 1 °“ A Sao i ‘= iT hi P 7 r = = a re a a 4 ae tio) ; = P< - vy j 5 hd é . ‘ rf 326 we a 1 “i < es d Coe 7 + tot = ¢ - ety 7 +a . , on a7 7 a Laie if 7 . = V. — od 7 ives cud : 7 ‘ : " " 7 ee Ui 3 “ar i= To <1" -. (a5) i. —" a ar a 7 ba Ci bet >" = art t 4 2 reas tle) ~ See bermh te Vey Fa Wh crs i yl yee > ; ; : ‘ om] i <> 1.) 7 ~Al pat foe i > : z : i i | ee ‘yy - ae 7 7 ay ‘ i fi : a s <1 _ v ~ Uae i a sore Arn y ei Ser > ¢ - * Mildly CGE °.10 ia oR NOTE ON A CASE OF PENTOSURIA PRESENTING UNUSUAL FEATURES. By J. H. ELLIOTT ann H. S. RAPER. (From the Department of Pathological Chemistry, the University of Toronto.) (Received for publication, February 12, 1912.) Since pentosuria was first recognized by Salkowski in 1892, up- wards of thirty cases have been described. The general features of the cases have been quite similar and may be summed up briefly as follows: The majority of the cases have been met with among the Jewish races. There has usually been no marked general dis- turbance, and the abnormality has often been discovered by acci- dent or when the patient has been under treatment for diabetes owing to faulty diagnosis. Examination of the urine has shown that it possesses reducing properties. With Fehling’s solution reduction takes place after a somewhat long latent period (about half a minute); the solution turns yellow, no cuprous oxide being precipitated at the time, but only after standing. With Nylander’s reagent there is only slight reduction. The urine is not fermented by yeast; it gives a strongly positive orcin reaction, and when heated with phenylhydrazine gives an osazone which melts in the crude state at about 150°C. Inallthegenuine casesso far described the urine has been optically inactive. The present case, for which we are indebted to Dr. C. J. Currie of Toronto who brought her to the notice of one of us, was, in all the above respects, like those previously described. It was only in an attempt to estimate the pentose by Neuberg’s method! that any abnormality was met with. In only one instance, toour knowledge, has the pentose been actually isolated from a case of pentosuria; and that was by Neuberg in 1900.2 Using twenty liters of urine, Neuberg was able to obtain about twenty grams of the pentose asthe diphenylhydrazone. This on decomposition with formalde- 1 Neuberg and Wohlgemuth: Zeitschr. f. physiol. Chem., xxxv, p. 38. 2 Neuberg: Berichte, xxxiii, p. 2248. SANUS 212 Case of Pentosuria hyde yielded the sugar itself, which was found to be inactive ara- binose. In our case, on attempting to estimate the pentose by means of diphenylhydrazine no insoluble hydrazone was obtained. An attempt was therefore made to isolate the pentose following the process exactly as described by Neuberg, but was fruitless. It was also found impossible to separate any crystalline derivative other than the phenylosazone, proceeding by way of the precipi- tate produced in the urine by basic lead acetate and ammonia, which contains all the reducing substance. That the substance is a pentose is shown by analysis of the osazone, but we have been unable to obtain any evidence that it is arabinose. The other two pentoses which give an osazone melting above 160° C., are ribose and arabino-ketose. The latter so far as we are aware, has not been encountered up to the present time in the examination of animal tissues or fluids. The fact that d-ribose had been shown by Levene and Jacobs to be present in some of the nucleic acids, suggests that there is some possibility of the pentose in the present case being inactive ribose. The clinical history of the case is as follows: Mrs. H., aged 32, a Russian Jew. Married, two children living. One child died at five and one-half months of pneumonia, one of scalds. Previvus illness: At about four years of age had a swollen hand with discharge of pus from opening on the dorsum, finally some bone discharged and the opening healed, leaving a transverse scar. A similar condition about the knee healed with no discharge of bone; at the same time also there was a discharging sore near one ankle and this the patient says still discharges at tines. One year ago a child two years of age fell into a tub of boiling water and died an agonizing death. Patient was much upset, became very nervous and went to Philadelphia to live. In the summer of 1910 she was admitted to the Jefferson Hospital with a swollen leg. There were places on the leg like boils but no pus was present. She was told her veins were inflamed. While in the hospital she was told there was sugar in the urine and that she had a mild form of diabetes. She returned to Toronto in the autumn of 1910 and sought advice regarding the glycosuria. She was not losing in weight, general appearance good, though she was of a nervous type and much worried and anxious about herself. At times she showed a tendency to magnify her symptoms and reported to her physicians many minor aches and pains. She complained of being always tired and at times weak; occasionally her hands were haavy to raise or move. She perspired freely in the axillae under examination. The general physical examination yielded nothing of importance. Pulse and temperature normal. No polyuria or thirst. J:.H..Elhott:and Hi.cS: Raper aire Examination of the urine. A routine examination of the urine showed that some reducing substance was present. The typical latent period before reduction was obtained with Fehling’s solu- tion. Nylander’s solution was only slightly reduced. The urine was -not fermented by yeast. Orcin reaction (Bial-Salkowski method), strongly positive. With phenylhydrazine a crystalline osazone was readily obtained. It erystallised out on cooling and melted in the crude state at 148 to 150°. On recrystalli- zation from 20 per cent alcohol the melting point rose to 161 to 163°, finally on crystallization from 10 per cent alcohol contain- ing a little pyridine it was obtained in long glistening yellow needles melting constantly at 163 to 164° C. ANALysIs: The osazone was dried in vacuo over P20; at 35° C. (1) 0.1431 gram gave 21.1 cc. moist nitrogen at 16° and 753.2 mm. Hg. = 17.08 per cent nitrogen. (2) 0.1548 gram gave 23.1 cc. moist nitrogen at 16° and 744.5 mm. Hg. =17.09 per cent nitrogen. Theory for C:;H2.N,O; requires 17.07 per cent nitrogen. Estimation of the pentose. A forty-eight-hour quantity of 2650 ec. was collected. One hundred cubic centimeters of this gave 0.168 gram crude phenylosazone corresponding to 0.0769 gram pentose or 2.04 grams in the two days. On attempting to estimate the pentose by Neuberg’s method’, using 1000 cc. of urine con- centrated to 60 cc. under reduced pressure at 35° C., no insoluble diphenylhydrazone such as arabinose gives, was obtained. Since the method for the estimation of arabinose in urine was worked out by Neuberg using normal urine to which varying amounts of the pentose were added, and since Neuberg has suggested that the pentose is excreted in combination with urea, it was conceiv- able that such a ureide might not react with diphenylhydrazine in the same way as the free sugar. This might therefore account for the failure to obtain the diphenylhydrazone in the above experi- ment. An attempt to isolate the pentose as diphenylhydrazone was therefore made using the process adopted by Neuberg for this purpose.* A liter of urine was concentrated to 50 ec. at 35° under reduced pressure then poured into 360 cc. of hot 95 per cent alcohol. 3 Neuberg: Berichte, xxxiii, p. 2243. 214 Case of Pentosuria Cooled, filtered from salts, the latter were then dried in the air, ground up with ether, dried and extracted in a Soxhlet apparatus for eighteen hours with 95 per cent alcohol. The extract was added to the main alcoholic solution and the whole evaporated to 40 cc. under reduced pressure. This residue was poured into 100 ce. of hot 96 per cent alcohol, cooled, filtered and the filtrate boiled with animal charcoal for a few minutes. The fluid was again filtered and concentrated at 35° C to 44 ec. Three cubic centimet- ers removed for estimation of the pentose by titration, showed that 0.63 gram was present assuming the pentose to be arabinose. The alcoholic solution was therefore heated with 1 gram of di- phenylhydrazine dissolved in a few cubic centimeters of alcohol in a boiling water bath for an hour. No diphenylhydrazone separated onstanding. (The diphenylhydrazine used was a freshly prepared specimen which easily gave the. characteristic hydrazone when heated with arabinose from gum arabic.) A second attempt was made using 4,700 cc. urine following the same procedure, but was likewise unsuccessful. Taking advantage of the fact that the pen- tose is precipitated by basic lead acetate and ammonia, an attempt was made to isolate the pentose in this way. One liter of urine was precipitated with lead acetate, filtered, and the filtrate precipi- tated with basic lead acetate and ammonia. This removed all the reducing substance. The precipitate was decomposed with hydrogen sulphide and the solution’ so obtained concentrated to about 100 cc. under reduced pressure at 35° C. It was neutralized to Congo red with sodium hydroxide, then to litmus with barium hydroxide, evaporated to a syrup in vacuo and poured into 200ce. of hot 95 per cent alcohol. The precipitated salts were filtered off after standing and the filtrate evaporated to small bulk under reduced pressure. This solution reduced very strongly and gave the characteristic osazone with phenylhydrazine but on treatment with diphenylhydrazine p-brom-phenylhydrazine or p-nitro-phenyl- hydrazine gave no crystalline hydrazone. Since some authors have reported instances of more than one case of pentosuria in a family, the urines of the available blood rela- tions have been tested. They were the son, daughter and brother, but ali proved negative with Fehling’s solution, the orcin reaction and phenylhydrazine. J; He Elhott and "iia Ss. Raper 215 Nore. Since sending the above results for publication we have been able, through the kindness of Dr. P. A. Levene, who fur- nished us with a specimen of d-ribose, to compare the properties of the phenylosazone of this substance with those of the phenyl- osazone of the urinary pentose. A specimen of the osazone of /-ara- binose (Merck) was also prepared for comparison. The osazones of the two active pentoses, which on theoretical grounds should have identical properties, melted at 162.3°, crystallized in the same manner and appeared to possess the same solubility in 10 per cent alcohol. The urinary pentosazone appeared to be less soluble in 10 per cent alcohol, and crystallized much better than the osazones from either of the active pentoses. No appreciable alteration in the melting point was obtained by mixing the urinary pentosazone with the osazones of the two active pentoses. * —" ba ee worden +p Sti iene me gait we erent te thor pam i A ites af aan on ipitine, to CR ars ) — Re Sa mee ac ee th pa ee er ae : : aad ees wernt a ae Se aT Ot ar ae ed ee : A BRIEF INVESTIGATION ON THE ESTIMATION OF LECITHIN. By R. C. COLLISON. - (From the Ohio Agricultural Experiment Station, Wooster, Ohio.) (Received for publication, February 12, 1912.) Methods of determining lecithin in animal tissues are somewhat unsatisfactory, due to the lack of a routine method whereby leci- thin itself can be separated and quantitatively estimated. By present methods it is determined along with related phosphatides, and the quantity reckoned from a phosphorus estimation by multi- plication by a convenient factor. The brief investigation, the results of which follow, is a com- parison of several methods with a view to gaining evidence as to their accuracy. In this laboratory, lecithin, in the strict sense, is estimated by an extraction with anhydrous alcohol and ether, evaporating the solvents and drying the resulting extracts, taking up with anhy- drous ether, filtering and determining the phosphorus in the ethe- real solution. W. Koch! states that owing to the difficulty of working under anhydrous conditions, this method is open to objection and that his method of separation of the lipoids with acid-chloroform- water is on this account preferable. This latter method is long and laborious when used in a routine way, and involves an undesir- able correction for non-lipoid phosphorus clinging to the precipitate. The other method is more easily workable and its accuracy in so far as any method for lecithin is accurate is questioned only on the ground of the practicability of working under anhydrous con- ditions. The following comparisons were made in an effort to learn whether or not it is practicable to work under essentially anhydrous conditions, and whether solvents, not absolutely water- free, appreciably affect the results. 1 Journ. Amer. Chem. Soc., xxxi, Dec., 1909. 217 THE JOURNAL OF BIOLOGICAL CHEMISTRY, XI, NO. 3 218 Estimation of Lecithin Five to eight gram samples of brain and liver contained in alun- dun extraction capsules were used in this work. They were thor- oughly dried in a vacuum according to the method of Schackell,2 being previously mixed in the capsule with sand to facilitate drying. Three sets of samples were treated as follows: 1. Extracted with anhydrous reagents, the dried extracts taken up with anhydrous ether and filtered. 2. Extracted with 95 per cent alcohol and U.S. P. ether, the dried extracts taken up with anhydrous ether and filtered. 3. Extracted with 95 per cent alcohol and U. S. P. ether, and the lipoids separated by acid-chloroform-water according to Koch. After complete desiccation the capsules containing the samples, the latter well covered with lipoid-free absorbent cotton, were divided into three sets as above, placed in small bottles or flasks and covered with alcohol, absolute and 95 per cent respectively. The bottles were loosely stoppered with glass stoppers, placed in a water bath and the alcohol boiled gently for two hours. In this way extraction takes place at the boiling point of the solvent. The alcohol was then decanted and the capsules and samples washed several times with alcohol of the appropriate strength. This extraction and washing was repeated four times, making eight hours boiling in all. The capsules were then dried at a low temperature in hydrogen and extracted in the usual way for 16 hours with ether, using anhydrous or U. S. P. material as above indicated. The samples were now removed from the capsules and pulverized in an agate mortar. After returning to the capsules they were again extracted for 16 hours with ether of purity previously desig- nated. The ether and alcohol fractions were then freed from the sol- vents by evaporation at a low temperature, and the two fractions combined. The absolute alcohol used in this work was prepared in this laboratory and was practically anhydrous. The anhydrous ether was made by redistillation from sodium of the ordinary anhydrous ether. 2 Amer. Journ. of Physiol., xxiv, June, 1909. R. C. Collison 219 The extracts from sets 1 and 2 were now dried at 50° C. in hydro- gen, taken up with anhydrous ether and filtered through asbestos. The ethereal solutions were transferred to 250 ec. Kjeldahl flasks, the ether evaporated and the residue decomposed with nitric and sulphuric acids. The phosphorus was determined, either by pre- liminary precipitation with magnesia ‘mixture, or directly with molybdate solution, taking care that in the latter case an excess of ammonium nitrate was used and the precipitate was digested from four to six hours at 60° C. The extracts of set three were treated according to the method of Koch with acid-chloroform-water, lipoid phosphorus being determined on the precipitate and the necessary correction made for non-lipoid phosphorus. The filtrates were perfectly clear. Percentages of lipoid phosphorus. Averages of three to five determinations. BRAIN LIVER METHOD TSE ae = F =e ee sans | First Second First | Second | series | series | series | series AS eae ee meeane permits as Set I. Anhydrous reagents, filtered. 0.176 | 0.215 | 0.110 | 0.092 Set II. Crude reagents, filtered. .... 0.202 | 0.225 0.112 0.112 Set III. Crudereagents, acid-chloro- | | | ll foorecter ten eB 0.201 0.223 0.118 | 0.110 In comparing these results it should be borne in mind that in any method for lecithin estimation, taking for granted complete- ness of extraction and accuracy of technique, the probable error lies in the direction of results that are too high, due to inclusion of inorganic or other forms of non-lipoid phosphorus. The lower results, other things being equal, are probably more nearly accu- rate. Although the variations in the results by the three methods are not great, those obtained by acid-chloroform-water treatment are slightly but uniformly higher than those obtained with anhydrous reagents and are probably, by that amount, too high. It would therefore seem, that taking ordinary precautions to | secure anhydrous reagents, the straight extraction method is preferable to the acid-chloroform-water separation. 220 Estimation of Lecithin With the purpose of further shortening the method, it was thought that possibly where anhydrous reagents were used, the filtering of the extract with anhydrous ether was unnecessary. To test this point, two sets of determinations were made on brain, liver and muscle. All samples were extracted with anhydrous reagents by the method given above. The dried extracts of one set were then taken up with anhydrous ether and filtered. The other set of extracts were analyzed for phosphorus directly with- out filtering. The results are given in the table below. Percentages of lipoid phosphorus. Averages of four.determinations. | EXTRACTS TAKEN UP WITH ANHYDROUS ETHER AND | FILTERED | | | EXTRACTS NOT FILTERED PHOSPHORUS RUN DIRECT Bait eeee eneaics 0.248 | 0.249 LING? caateeniae e- 6 | 0.153 0.153 Muscles sees cs. | Oz0470 | 0.044 These results indicate that taking up the extracts with anhy- drous ether and. filtering is an unnecessary step in the determina- tion, since practically identical results were obtained in both cases with all three tissues. The most satisfactory method found therefore and the one in use in this laboratory, is that in which the combined alcohol and ether extracts are analyzed for phosphorus without previous filtra- tion with ether, provided the necessary precautions are taken to secure dry reagents. Crude reagents seem to render soluble certain forms of phos- phorus which are not successfully separated from the hpoids by the acid-chloroform-water treatment. The author wishes to express his thanks to Dr. E. B. Forbes for making this investigation possible. THE PURINES OF MUSCLE:.! By C. B. BENNETT. (From the Rudolph Spreckels Physiological Laboratory of the University of California.) (Received for publication, February 21, 1912.) The vast amount of work done on the purine bases in mam- malia has shown that the two purines commonly found in fresh glands’are adenine and guanine, and that hypoxanthine and xan- thine, previously reported as having been found in various organs, occur only as the products of enzymatic changes of adenine and guanine, and are not found in the fresh glands.2, Muscle tissue, however, even when perfectly fresh, always shows the presence of more hypoxanthine than of any other purine, and the question therefore arose as to where this hypoxanthine came from. Schit- tenhelm*® maintained that it was probably due to the conversion of adenine from nucleic acid by the ferment adenase, while Jones‘ said that the hypoxanthine was formed in the muscle itself. This work was therefore started in the hope of obtaining more light concerning muscle purines. GUANINE AND ADENINE. The fact that guanine occurs in muscle tissue in small amounts has long been known. Adenine, however, is usually considered to be absent,® though Mendel and Leavenworth® showed that large amounts are obtainable from embryo calf muscle. ' Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of California. 2 Oppenheimer’s Handbuch der Biochemie, 1, p. 610, 1908. 3 Schittenhelm: Zeitschr. f. physiol. Chem., |xiii, p. 248, 1909. ; 4 Miller and Jones: Jbid., lxi. p. 393, 1909; Rohdé and Jones: this Journal, vii, p. 237, 1909-10. ° Cf. Frankel: Descriptive Biochemie, p. 124. 5 Mendel and Leavenworth: Amer. Journ. of Physiol., xxi, p. 99, 1908. 221 222 Purines of Muscle In this work it has been found that fresh rabbit muscle, after extraction with water at 65 to 70°C., upon digestion with dilute sulphuric acid yields guanine but not adenine. If the extraction be made with cold water, the residue after digestion yields both guanine and adenine: and further it has been shown that this dif- ference is due to the extraction of adenine from the muscle by the hot water. Three typical experiments are cited below in support of these statements. 1. A white female rabbit was etherized, then bled to death, and after skinning and cleaning, was immediately placed for fifteen minutes in dis- tilled water at 60°C. The muscle was then stripped from the bones and ground three times. Four hundred and sixty grams of this hashed muscle, mixed with 700 ce. of distilled water, were gradually heated to 60°C., then filtered through cloth, the residue being squeezed as dry as possible, and ground again. This process was repeated twelve times, using 600 cc. of dis- tilled water each time, and heating each extraction slowly to 60°C. before filtering. The final residue, the moist weight of which was 371 grams, was digested with 1500 cc. of 3 per cent sulphuric acid at 100°C. for twenty-four hours. On treatment of the fluid with copper sulphate and sodium bisul- phite in the usual way, 31 mgs. of guanine were obtained. No trace of ade- nine was found, although a little of some other purine was present, probably hypoxanthine. The guanine obtained was purified through its silver salt and gave characteristic reaction when evaporated with strong nitric acid and treated with sodium hydroxide. 2. A white female rabbit was killed by bleeding after etherization, and its muscles removed and hashed. Five hundred and sixty-four grams of the hashed muscle were extracted with equal weight of distilled water for ten to fifteen minutes and the extract filtered through cloth. This process was repeated four times. The residue remaining was then digested for twenty- four hours with five times its weight of 3 per cent sulphuric acid. It yielded 65 mgs. of guanine and 143 mgs. of adenine picrate, equivalent to 53 mgs. of adenine. There was also a little hypoxanthine(?) present. 3. Alarge black male rabbit was killed by bleeding afteretherizing, and its muscles hashed. Nine hundred and forty-one grams of the hashed muscle were extracted five times with its own weight of distilled water. It was then extracted twice with like amounts of water, heated gradually to 55 to 57°C. The combined warm water extracts were acidified with 45 cc. of, concen- trated sulphuric acid and digested for 24 hours.at 100°C. It yielded 27 mgs. of guanine and 62+mgs. of adenine picrate, equivalent to 25+ mgs. of adenine. Although adenine is found in the vegetable kingdom as in tea leaves in a free condition, it has almost never been found free in the tissues of higher animals, except in such abnormal tissues as cancers. Usually it is present in animal tissues as a constituent C.°B. Bennett Toa of thymus nucleic acid. It would seem probable therefore that in the muscle also, adenine is present as a part of thymus nucleic acid. Only a portion of the guanine however, could thus be united since some guanine was always found in the tissue after the extraction by warm water had removed all of the adenine-con- taining molecules. All of the figures obtained in this work indi- eate that there is more guanine in muscle tissue than adenine. The following data are taken from Mendel and Leavenworth. From thirty embryos were obtained: 0.093 gram guanine. 0.065 gram adenine. 0.029 gram hypoxanthine. ( Guanine hydrochloride..... 0.137 gram Adenine picrate............0.185 gram Hypoxanthine nitrate......0.057 gram It may be partly due to this fact and also to the fact that gua- nine more readily crystallizes out of its solutions that the presence of guanine in muscles is better established than that of adenine. Since striated muscle contains, inaddition to the striped muscle, fibers, connective tissue and blood vessels, with their smooth muscle fibers, the following data obtained by Sieber and Dzierz- gowski’ on the purine content of the lungs are of interest; for in spite of the fact that the connective tissue of the lungs is not iden- tical with that of muscle, it is convenient to look upon lung tissue as approximately equivalent to muscle tissue minus its striped muscle fibers. These workers found by Kossel’s method that for every 100 grams of lung tissue, they obtained 60 mgs. of xanthine, 164.7 mgs. of guanine, 126.9 mgs. of adenine, 179.6 mgs. of hypo- xanthine. Varying yields however were obtained with different methods. The high percentage of hypoxanthine present may possibly be due to the smooth muscle fibers of the blood vessels, ete., for Saiki® found that thé purines of smooth muscle, as is the case with striated muscle, included hypoxanthine in prepon- derating amounts. To determine directly whether connective tissue contains adenine or not, I have digested connective tissue obtained from ox-tails and from the tendon of Achilles of the ox with sulphuric acid; 135 grams of connective tissue from the ten- don of Achilles, digested with 675 ec. of 3 per cent sulphuric acid, yielded 10.5 mgs. of guanine, 25.5 mgs. of adenine picrate, equiv- alent to about 10 mgs. of adenine. It seems certain, therefore 7 Sieber and Dzierzgowski: Zeitschr. f. physiol. Chem., |xii, p. 259, 1909. 8 Saiki: This Journal, iv. p. 483, 1908. 224 Purines of Muscle that some of the adenine and guanine in muscle must come from the connective tissue in it. It is possible that the large amounts of adenine and guanine obtained by Mendel and Leavenworth? in embryo pig muscle, and also the high percentage of guanine found by Kossel!° in embryonic calf muscle, may really be due to the very high percentage of connective tissue in these embryonic muscles rather than to any specific feature of the muscles them- selves. It is interesting to note that Saiki" found a trace of ade- nine in extract of smooth muscle—a substance not usually noted in extracts of striated muscle—and immediately afterward pomts out that smooth muscle contained very much more connective tissue than does striated muscle. In order to further identify the guanine obtained from the various exper- iments, the picrate was made. Wulff!? stated that guanine picrate begins to decompose gradually at 190°C., but the guanine picrate I obtained decomposed at nearer 260°C. Guanine was then prepared from the fresh thymus gland of the calf, by digesting it with 3 per cent sulphuric acid by volume, precipitating with copper sulphate and sodium bisulphite. The solution obtained by breaking down the copper purine was concentrated and the guanine precipitated from the hot solution by making it fairly strongly ammoniacal. After filtering off the mother-liquor the guanine was dissolved in hot dilute sulphuric acid and again precipitated while hot with ammonia in excess, filtered, and again this process was repeated It was then changed into the silver salt and washed with ammonia, con- verted into the hydrochloride and evaporated carefully over a water-bath until just dry. This hydrochloride was then dissolved in hot water and filtered through hardened paper, and the picrate made by adding to it a solution of picric acid. This picrate also decomposed gradually from about 255 to 260°C. The following table compares the behavior of adenine and guanine picrates on heating. Adenine picrate. At 240°C. a trifle more brown than at first, and kept growing darker very slowly indeed. At 275°C. still brownish yellow. At 280° to 281°C. decomposed by turning rapidly black and melt- ing. Guanine picrate. At 200°C. began to grow slightly more brown. At 212°C. much more brown. At 235°C. orange. At 240°C. chocolate. At 252°C. brown. At 258° to 260 C. turned slowly quite black and melted. 9 Mendel and Leavenworth: Loc. cit. 10 Kossel: Zeitschr. f. physiol. Chem., viii, p. 404, 1883-4. 11 Saiki: Loc. cit. 12 Wulff: Zeilschr. f. physiol. Chem., xvii, p. 468, 1893. C.B. Bennett 225 In this work it was found more convenient to make the picrates of ade- nine, guanine or hypoxanthine by adding a saturated solution of picric acid to the acid salt of the purine rather than the sodium picrate so often advised. A slight excess of the mineral acid did not interfere in any way with the reaction and it was found much easier to distinguish the purine picrates from the crystals of picric acid than from the needle-like ervstals of sodium picrate in those cases in which for some reason or other the solution had to be very strongly concentrated. When picric acid is added to even a strong solution of guanine hydrochloride, no clouding is usually observed but in a very short time the extremely insoluble guanine picrate separates out as well defined crystals which soon sink to the bottom of the solution. When picric acid is added to adenine hydrochloride or sulphate, there is an.immediate clouding of the solution due to the formation of very fine crystals of the adenine picrate, and these crystals do not settle out at all readily. The difference therefore between the two is well marked, although both picrates when once formed are extremely insoluble.* When dried at — 105°C. the guanine picrate always appears as sparkling crystals which do not readily give up their water of crystallization even when heated above 110°C, while adenine picrate,dried, shows no brilliancy and apparently read- ily gives up all its water ata little over 100°C.!* Cuanine picrate seems to crystallize in several forms. The tree- or fern-like form, described so well by Wulff,!* has a distinctly redder appearance than adenine picrate, but those in the forms of long needles or platelets have the same color as adenine picrate. The redder, tree-like crystals were carefully picked out with a fine pair of forceps and placed in another receptacle, and when recrystal- lized from a little distilled water, came out as the lighter-colored platelets. The melting point determinations were all made on the yellower type. Usually guanine picrate dissolves very slowly in ammonia while adenine picrate dissolves almost immediately. Hypoxanthine also forms picrates of two forms—it sometimes appears as silky threads but more often as short, thick, six-sided crystals.!7_ As its solubility, however, is very much greater than either that of adenine or guanine, there is very little danger of contaminating the picrates of the two latter with hypoxanthine. . HYPOXANTHINE. By the digestion of fresh meat with dilute sulphuric acid we find that hypoxanthine is a normal constituent of muscle, but obviously we cannot thus learn anything concerning the condition of this hypoxanthine while in the living tissue. Rohdé and Jones!® 138 Bruhns: Zeitschr. J. physiol. Chem., xiv, p. 588, 1890: Wulff: loc. cit. 14 Wulff: Loc. cit. 16 Bruhns: Loc. cit. 16 Wulff: Loc. cit. 17 Wulff: Loc. cit. 18 Rohdé and Jones: Loc. cit. 226 Purines of Muscle stated that hypoxanthine is probably formed in the muscle itself without the action of adenase, while Scaffidi,!® Mendel and Leaven- worth,?? Kennaway,% and Krukenberg” speak of “‘free purines’ or “free hypoxanthine’ of muscle. Most of the investigators using the term “free purines” seem to mean by that any purine not directly connected with a protein, or with a coagulable pro- tein (see Scaffidi), but we know that there are several substances in commercial meat extract, such as inosinic acid, carnine, inosin, ete., which are neither compound proteins nor in any sense free hypoxanthine, although easily yielding the latter on decomposi- tion. No reference was yet found which really considered care- fully whether the so-called “free hypoxanthine” found in the muscle was free or not, for it is obvious that a method like that of Kruger-Schmid” as ordinarily carried out cannot be relied upon for such determinations. The first problem was therefore to see how much inosinic acid was present in fresh meat. The principal methods given for the preparation of inosinic acid are briefly as follows: Liresie’s Metuop.** Thecold extract of fresh meat was boiled to coagu- late the proteins, and then the filtrate was evaporated to a very small vol- ume at a low temperature to crystallize out the inosinic acid. Hatser’s Metuop.” The commercial meat extract was boiled with abso- lute alcohol and the inosinic acid in the insoluble residue was precipitated as the silver salt after the elimination of all the phosphates with barium hydroxide. Levene’s modification” of this consisted in extracting with 95 per cent alcohol instead of absolute. ; BauEr’s Meruop.?” The water solution of commercial meat extract, after clearing with animal charcoal, was freed from phosphates with barium acetate and hydroxide, and the inosinic acid was precipitated in an alkaline solution with basic lead acetate. 19 Scaffidi: Biochem. Zettschr., xxxili, p. 247, 1911. 20 Mendel and Leavenworth: Loc. cit. 21 Kennaway: Biochemical Journal, v, p. 188, 1910. 22 Krukenberg: Untersuchungen aus dem physiologischen Institut der Universitat Heidelberg, iii, p. 217, 1880 (ref: from Mendel] and Leavenworth). *3 Hoppe-Seyler-Tierfelder, Handbuch d. physiol-chem. Analyse, vii, p. 435 (ref. from Frankel’s Descriptive Biochem.). 24 Liebig: Ann. d. Chem. u. Pharm., \xii, p. 257, 1847. 25 Haiser: Monatsh. f. Chem., xvi, p. 190, 1895. *6 Levene and Jacobs: Ber. d. deutsch. chem. Gesellsch., xli, p. 2704, 1908. 27 Bauer: Beitr. z. chem. Physiol., x, p. 345, 1907. C. B. Bennett 225 HAIsER AND WENZEL’s Metuop.?3 The commercial meat extract, after freeing from phosphates, was neutralized and the inosinic acid precipi- tated with basic lead acetate. All but the first of these methods start with the commercial meat extract and not with the fresh meat, and the only one which starts with the meat itself has been found to yield uncertain results.?9 The following method in which boiling, long evaporations, and acidity are avoided, was therefore adopted in this work: A rabbit was etherized and killed by bleeding, and its muscles were sepa- rated as quickly as possible. The meat was then ground twice, and extracted with its own weight of cold, distilled water five times, allowing the meat to soak ten or fifteen minutes for each extraction. The meat was filtered through cloth, and squeezed as dry as possible each time, and the same cloth was used for all the filtrations. When the extractions are made with hot water, gelatin and other undesirable substances are also extracted which are difficult to separate from the inosinic acid. The united filtrate was then heated to 65°C. to coagulate some of the proteins and filtered through paper. To the clear filtrate a saturated barium hydroxide solu- tion was added to precipitate the phosphates and sulphates, and also most of the remaining proteins. The precipitation of proteins by barium hydrox- ide has already been used by Peters*®® in his preparation of thymus nucleic acid and has been found to be an extremely convenient reagent in this work. When the further addition of the barium hydroxide caused no more pre- cipitation, the solution was warmed to 45 or 50°C. with constant stirring, to make the precipitate form a coagulum which was easily filtered off. The filtrate was then exactly neutralized with dilute acetie acid and basic lead _ acetate immediately added to precipitate the inosinic acid, until all pre- - cipitation just ceased. Anexcess of basic lead acetate is to be avoided as it dissolves the precipitate. After the precipitate had settled the liquid was decanted into a filter and the precipitate washed once or twice by decan- tation with distilled water, pouring all the water through the filter. Finally all the precipitate was also carefully washed into the filter. Then the paper with the lead precipitate in it, was placed-in a beaker containing a little water ard was beaten into a pulp, special care being taken that no large lumps of the lead precipitate remained. After slightly warming this mix- ture, some clear saturated solution of barium sulphide was added drop by drop while constantly stirring until the solution just began persistently to tarnish a well-cleaned silver coin when a drop of the solution was left on the coin one minute. A large excess of barium sulphide is to be avoided as the barium sulphide constantly but slowly changes to barium carbonate °8 Haiser and Wenzel: Monatsh. f. Chem., xxix, p. 157, 1908. 29 See Bauer, Loc. cit. 30 Peters: This Journal, x, p. 373, 1911. ‘ 228 Purines of Muscle when exposed to the air, and this precipitate is therefore sure to contam- inate the final product. The solution was then warmed to 60°C. and filtered warm. The filtrate was usually perfectly clear and apparently contained very little besides barium inosinate. After thoroughly washing the preci- pitate with warm water, the united filtrate was placed in a beaker with per- pendicular sides and mixed thoroughly with five times its own volume of 95 per cent alcohol, and then left covered for twenty-four hours. In this way the barium inosinate was precipitated quantitatively at the bottom of the beaker. All the liquid that could be safely decanted was very carefully decanted off. When this operation was well done, it saved a great deal of time as the subsequent filtration was always very slow. The remaining liquid and the precipitate were then placed into a Buchner funnel provided with filter paper, and suction was applied. All the precipitate that passed through the filter, and some was sure to go through at first, was again placed on the filter until a perfectly clear filtrate was obtained. This filtration was always extremely slow, but fortunately did not require much of the operator’s attention. It was finally washed with absolute alcohol and ether and allowed to dry at room temperature. The dry material was then separated as much as possible from the funnel and filter paper and the twolast were carefully washed with a little hot water to dissolve out all the barium inosinate adhering to them. The major por- tion of the precipitate was then transferred to the same water and after carefully breaking up all lumps, was placed over asteam bath and constantly stirred. As soon as most of the material had dissolved, it was filtered hot through a small filter-paper. Should the material not readily dissolve in the amount of weter present, a little more may be added, but a great excess should not be added. All the soluble products should easily dissolve without heating over 80°C. After filtering, the filter paper was washed free of the precipitate on it with a little hot water and this precipitate again digested over steam a short time and again filtered hot. The warm filtrate was allowed to cool slowly and was finally placed in the ice-chest. After twenty-four hours or more, the crystals of barium inosinate were filtered off and allowed to dry at room temperature. If the volume of mother-liquor was not too small, it was found advisable to concentrate it strongly at 50°C. and then again leave in the ice-chest for a second crystallization. Haiser*! in his method warns us of the danger of adding an excess of barium hydroxide and states that all the inosinic acid may be lost as an insoluble basic barium inosinate by doing so. In the article by Haiser and Wenzel® this statement is very much modified,. but from the following experiment it seems evident that the caution is wholly needless. Barium inosinate was dissolved in a little hot water and barium hydroxide, a saturated solution, was added 31 Haiser:. Loc. cit: 32 Haiser and Wenzel: Loc. cit. C. B. Bennett ; 229 in gradually increasing amounts. No precipitate, however, formed either in the cold or on heating moderately. Then a clear solu- tion of saturated hydroxide was taken and some pure dry barium inosinate added to it and very gently heated. All the salt dis- solved in the solution. The solution was then carefully filtered through hardened paper, just neutralized with acetic acid and treated with a solution of silver nitrate. The characteristic silver inosinate precipitate formed, which immediately dissolved on adding ammonia. Of course, when the carbon dioxide of the air was allowed to act on the solution of barium inosinate in barium hydroxide, some barium carbonate was formed, but that was the only precipitate present even after standing several hours. It seems evident therefore that no loss of inosinic acid need be feared from an excess of barium hydroxide. In attempting to get a good method for the preparation of ino- sinic acid various reagents were tried. The silver, copper and mercury salts were found unsuitable. In concentrated solution of barium inosinate, neutral lead ace- tate gives a precipitate, but not when dilute, while basic lead ace- tate precipitates even from very dilute solutions. One milligram of barium inosinate, dissolved in 15 ce. of water gave a character- istic precipitate after ten or fifteen minutes. A great excess of either neutral or basic lead acetate completely dissolves the pre- cipitate. The basic lead salt is insoluble in cold water and prac- tically insoluble in hot. It settles rather quickly and so can be easily washed by decantation. It filters better than the silver salt and is not affected by light. An excess of carbon dioxide, however, should not be present. Basie lead acetate and ammonia is perhaps a still better pre- cipitating agent for inosinic acid but as this reagent precipitates also all the purine bases, gelatin, creatine, ete., which would ordi- narily not be precipitated by basic lead acetate alone, the use of ammonia or any other alkali is to be avoided when the purity of the inosinic acid is a consideration. In the purification of barium inosinate when the substance does not read- ily purify by repeated recrystallization with hot water, the crystals were dissolved in cold, very dilute, sodium hydroxide (about 0.3 per cent), after completely freeing the alkali from all carbonates with barium hydroxide. In this alkaline solution the barium inosinate dissolves rather readily, and 230 Purines of Muscle may be recovered again after filtration by almost neutralizing the solution with dilute acetic acid. As the barium inosinate is naturally extremely faintly alkaline to litmus, the yield is larger if the neutralization is not quite completed. It is interesting to note that Haiser® states that he obtained from inosinic acid a substance agreeing in nearly all particulars with Steudel’s hypoxanthine, but differing from his description in that it was precipitated with basic lead acetate. In this Weidel* seems to agree with Haiser, and cites the fact that Stadeler® precipitated part of his xanthine with basic lead acetate. On the other hand Rohdé and Jones** recommended the use of this basic acetate to clear the solutions of substances other than purines, arid the same reagent was used in a method also previously fol- lowed by Neubauer*’? and others for the-same purpose. Later even Haiser**® used the same reagent to separate the inosinic acid fraction from the carnine fraction without, however, withdrawing or explaining his first statement so far as I know. It seemed wise, therefore, to obtain a better knowledge concerning this question. Hypoxanthine was therefore prepared from the commercial Liebig’s meat extract by precipitating it as the copper salt, and then as the silver salt which was next dissolved in boiling nitric acid according to Neubauer’s method. The silver nitrate salt of hypoxanthine which separated out was digested in weak ammonia, then broken down with hot dilute hydrochloric acid and the acid then evapor- ated away on a water bath. The salt was then dissolved in wate and carefully precipitated as the free base by just neutralizing the solution with sodium hydroxide. ; Ten milligrams of this dry base was carefully dissolved in 10 ce. of distilled water by heating, forming a 0.1 per cent solution, and to the cooled solution basic lead acetate prepared exactly as directed in the U. 8S. Pharmacopeia, page 267, was added, but no precipitate formed. This, however, was a super-saturated solution from which crystals of hypoxanthihe separated out when left ' 33 Haiser: Loc. cit. 34 Weidel: Ann. d. Chem. u. Pharm., clviil, p. 353, 1871. 35 Stadeler: Ann. d. Chem. u. Pharm., cxvi, p. 102, 1860. 36 Rohdé and Jones: Loc. cit. 37 See Balke: Journ. f. prakt. Chem., clv (n. s. xlvii), p. 552, 1893. 38 Haiser and Wenzel: Loc. cit. Cy. Be Bennett 231 over night without the acetate. Emil Fischer states that hypo- xanthine is soluble in 65.5 parts of boiling water, in 1415 parts of water at 19°, in 1370 parts of water at 23°29 Even when 10 mgs. was dissolved in 5 cc. of water, a 0.2 per cent solution, no preci- pitate formed with the lead salt, but when a 0.4 per cent solution of hypoxanthine was tested, the basic lead acetate did cause a precipitation. The precipitate dissolved again when the solution was sufficiently diluted. As the total purine content of the meat is generally thought to be about 0.2 per cent*® it seems evident that when the weight of the water used for extraction equals the weight of the fresh meat to be extracted there is no danger of the basic lead acetate precipitating any of the free hypoxanthine. This then. gives a convenient method of separating hypoxanthine from inosinic acid—a method already used and published by Haiser and Wenzel.*! Two attempts to determine quantitatively the amount of ino- sinic acid in rabbit meat were made with the process given above. The first time 0.48 gram of the barium salt, CioHiBaN,POg+ 734 HO, air dried, was obtained from 530 grams of meat, and the second time over 0.60 gram from 580 grams of meat. This points strongly to the conclusion that a large part of the hypo- xanthine must be in some other form besides that of inosinic acid if the total hypoxanthine content is anywhere near 0.2 per cent. Indeed the fact that Balke*® by Neubauer’s method in which he also used the filtrate from solutions treated with basic lead acetate, could get large amounts of hypoxanthine, proves the same conten- tion. DISTRIBUTION OF INOSINIC ACID. Turning now to the distribution of inosinic acid in the different kinds of animals, we find that shortly after Liebig’s ** discovery of inosinic acid in 1847, more or iess work was done t6 determine if the muscles of all animals contained this acid or not. In 1848 39 Ref. from Beilstein Handbuch d. org. Chem., Ergadnzungsband, iii, 1708. 40 Scaffidi: Loc. ctt.; Burian and Hall: Zeitschr. f. physiol. Chem., xxxviil, p. 336, 1903. 41 Haiser and Wenzel: Loc. cit. #2 Balke: Loc. cit. 43 Liebig: Loc. cit. 232 Purines of Muscle Gregory published the statement that he was unable to find inosinie acid in ox-heart, in pigeons, and in codfish, although he obtained fairly large amounts from some other animals. A little later Schlossberger* was unable to find any in human flesh. Since that time a vast amount of work has been done on the structure of inosinic acid by Haiser, Levene, and others, but very little seems to have been done concerning its distribution, except for the work of Creite,** so that Bauer*’ in 1907 still quoted Gregory’s statement concerning the absence of inosinic acid in certain animals. As it appeared a rather significant fact if inosinic acid were really absent in certain animals while present in their near relatives, a test was undertaken on the pigeon, which was the only one of Gregory’s animals easily available at the time. The method followed was in general that of Haiser and Wenzel*® with the excep- tion that meat, instead of meat extract, was used, and the washing of the lead salt was carried on considerably more, as this part of the work was done before the method given above for the prepara- tion of inosinie acid, was devised. Five pigeons were etherized and killed by bleeding, immediately cleaned, and the meat freed from bones was ground and extracted with water at 60°C. several times. The united filtrate was treated with saturated barium hydrate until no more precipitate formed. The barium precipitated all the phosphates and most of the proteins in the solution. The filtrate was made neutral with dilute acetic acid and then very faintly alkaline with ammonia. Basic lead acetate was added until complete precipitation took place, but an excess was carefully avoided. The precipitate was then washed by decantation in tall cylinders by chang- ing the water every three or four hours for several days. The acid was now liberated with hydrogen sulphide in the cold; powdered barium carbonate was added, and the whole heated over steam. The filtrate was next evapor- ated to a very small volume at about 50°C. The resulting crystals were recrystallized until iong beautiful platelets over a millimeter in length were obtained, which were in every respect exactly like those obtained from Liebig’s meat extract by Haiser’s earlier method. 44 Gregory: Ann. d. Chem. u. Pharm. \xiv, p. 100, 1848. 45 Schlossberger: Ann. d. Chem. u. Pharm., \xvi, p. 80, 1848. 6 Zeitschr. f. rationelle Med., xxxvi, p. 195. 47 Bauer: Loc. cit. 48 Haiser and Wenzel: Loc. cit. C. B. Bennett 2353 The yield was only about one-tenth of a gram and responded to all the usual tests for the barium salt of inosinic acid. This method probably does not give even approximately quan- tatative results, but it leaves no doubt that inosinic acid is present in the pigeon, and all statements concerning the absence of this acid in birds or mammals, should, I think, be held in considerable doubt until some better and more quantitative methods be used. Attempts were made to see if the smooth muscle tissue also con- tained inosinic acid. As Saiki** had already pointed out that hypo- xanthine constitutes there also the largest portion of the purine bases, the presence of inosinic acid was naturally expected. For this purpose bladders from oxen just killed were taken, carefully trimmed from loose connective tissue, red muscle fibers of the neck, fat, ete., cut open and washed, then ground and treated exactly as was the rabbit meat. It was found however that on heating the cold water extract the coagulum would not separate from the mother liquor as in the case of striated muscle, but formed a fine milky precipitate which almost prevented filtration. This was found to be due to the lack of acidity of the extract of smooth muscle, for as shown by Halliburton and others,®® the extract of striated muscle is always slightly acid, while that of smooth mus- cle®! is not. A few drops of weak acetic acid were therefore added and the solution vigorously stirred, which made the subsequent filtration very much easier. On adding basic lead acetate until precipitation completely ceased, it was found that the solution always contained a fine suspension of some lead precipitate which would not settle even after hours of standing. As the basic lead precipitate of inosinic acid settles almost completely under simi- lar circumstances, this milky solution was decanted off. The heavier precipitate was treated with fresh water and the solution again decanted off after all the heavier precipitate had settled. This was repeated until most of the fine precipitate had been washed away. The subsequent treatment of the lead precipitate was conducted exactly as in the case of rabbit meat. No inosinic acid could be identified from the ox bladders, although the attempt 49 Saiki: This Journal, iv. p. 483, 1908. 50 Halliburton: Journ. of Physiol., viii, p. 133, 1887. 5! Vincent: Zeitschr. f. physiol. Chem., xxxiv, p. 417, 1901-2. 234 Purines of Muscle was made three times, using respectively 560 grams, 810 grams, and 2660 grams of the fresh bladder meat for the trials. Some crystalline substance was indeed obtained but it did not give the inosinie acid tests tried above. Whether there is any inosinic acid in smooth muscle or not, is therefore an open question but it seems evident that the amount must be much less than in the striated muscle fibers. CONCLUSION. From the facts previously stated, we conclude that muscle tissue contains small amounts of adenine and guanine, very likely in the form of thymus nucleic acid, and that some of this, perhaps all of it, is found in tissues other than the striated muscle fibers themselves. More guanine, however, is present than adenine. The inosinic acid of striated muscle fibers represents only a fraction of the total hypoxanthine present, but we are not at all certain whether the remaining hypoxanthine is free, that is, uncom- bined with any complex organic radicle, or not. Inosinie acid is probably present in the striated muscles of all warm blooded animals, although in apparently very varying amount. It may be absent in smooth muscle tissue. Since the inosinic acid can be so easily obtained by extraction with cold water, it seems hardly probable that it is confined to the nuclei of the muscle cells. This supposition is still more strength- ened by the fact that the nucleic acid in other tissues are found, after death at least, united with proteins in the form of nucleopro- teins, which are rather insoluble in cold water and which must be first digested with some alkali or acid to liberate the nucleic acid. This is well shown by the method of Steudel and Brigl® for the prepara- tion of guanylic acid, also by Peter’s method for the preparation of thymus nucleic acid.. Inosinie acid, however, is obtained in fairly pure condition by precipitating the cold water extract of meat with basic lead acetate. Inosinic acid therefore is probably a nucleic acid only from the chemical standpoint and not from the histological. 52 Steudel and Brigl: Zeitschr. f. physiol. Chem., \xviii, p. 40. 1910. THE INFLUENCE OF COCAINE UPON METABOLISM WITH SPECIAL REFERENCE TO THE ELIM- INATION OF LACTIC ACID. By FRANK P. UNDERHILL anp CLARENCE L. BLACK. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven, Connecticut.) (Received for publication, February 27, 1912.) The introduction of cocaine into the organism is followed by such well defined symptoms that an almost specific influence upon the nervous system is indicated. In the main, it is to this aspect of its action upon the body that the very extensive literature! regard- ing this drug relates. Definite knowledge of the effect of cocaine upon general metabolism is meagre although the picture presented by the cocaine habitué is sufficiently characteristic to lead one to infer that ultimately at least the nutritional rhythm must be altered. The widespread employment of cocaine as an ingredient of various types of proprietary remedies and the large number of cases of cocainism makes pertinent at this time an inquiry into the in- fluence upon metabolism of the drug under discussion. The observation of Araki? that lactic acid appears in the urine in unusually large quantities after cocaine injections considered in connection with the findings of Wallace and Diamond? that cocaine causes vacuolization of the liver cells of rabbits suggested the pos- sibility of a disturbance in intermediary metabolism. In the present paper the relation of cocaine poisoning to lactic acid out- put is shown and the influence of the nutritive condition of the animal upon this type of acidosis is discussed. It is also demon- strated that in spite of the marked symptoms characteristic of 1 Cf. Richet: Dictionnaire de physiologie, iv, p. 1, 1900. ? Araki: Zeitschr. f. physiol. Chem., xv, p. 335, 1891. 3 Reported at the 19th Annual Meeting of the American Physiological Society, New York, 1907. 235 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 3 236 Influence of Cocaine upon Metabolism chronic cocaine poisoning general metabolism is only slightly changed from the normal even though the quantity of drug admin- istered is sufficient to finally cause death. These observations serve as a further illustration of the tenacity with which the organ- ism adheres to the fundamental laws underlying its metabolic processes; in other words, another example of the “factor of safety”’ principle is encountered in cocaine poisoning. THE INFLUENCE OF COCAINE UPON METABOLISM, AS INDICATED BY ITS EFFECT UPON NITROGENOUS EQUILIBRIUM AND PROTEIN AND FAT UTILIZATION. Methods. The experiments were planned so that the animals (dog and rabbit) employed were kept upon a fixed diet and cocaine administered subcutaneously at a time sufficiently long after a meal to avoid the danger of food being vomited. During the first period of the experiments the drug was given once daily, later the animal was kept under the influence of cocaine the greater portion of each day by repetition of the injection. Lactic acid was estimated by the Ryffel‘ procedure. The Folin method as modified by Steel’ was employed in the deter- mination of ammonia in the urine of rabbits. The other deter- minations were carried out according to the well known meth- ods usually employed in this laboratory. Urine was collected in twenty-four hour periods by catheterization (dogs) or by pressure on the bladder through the body wall (rabbits). Unless otherwise noted all urines of dogs were acid in reaction to litmus. The rab- bits’ urines were alkaline throughout. Description of experiments. Experiments 1 and 2. In these observations dogs 50 and 51 were kept for several days previous to the investigation upon the diet arranged for the experimental trials in order to bring them as nearly as possible into a condition of nitrogenous equilibrium. A fore-period was followed by an interval during which the animals received daily subcutaneous injections of cocaine hydrochloride (Kahlbaum’s crystalline pro- duct) dissolved in water. In addition to a constant diet through- out the experiment the animals received, also, a fixed water intake. ‘Ryffel: Journ. of Physiol., xxxix, p. v. 1909-10. 5 Steel: This Journal, viii, p. 365, 1910-11. Frank P. Underhill and Clarence L. Black 237 Protocol of Experiment 1. Dog 50, weighing 12.8 kilos, was normal in every respect except that she was extremely deaf. The diet consisted of 200 grams meat, 80 grams cracker meal, 40 grams lard, 10 grams bone ash, and 300 cc. water. The total nitro- gen intake amounted to 7.40 grams nitrogen daily with sufficient fat and carbohydrate to yield approximately 70 calories fuel value per kilo of body weight. Each day food was given at 9:30 a.m. and the first cocaine injection at 3:30 p.m. On October 20 the cocaine period was begun. Just before the cocaine injection the rectal temperature was 38.6 C. and two hours later had risen to 39.0° C. The pupils showed extreme dilatatior. October 21. In the morning the dog seemed normal and ate food with evident relish. Temperature before cocaine administration was 38.6° C. and had risen to 40.0°C. twohourslater. About 45 minutes after the injection the animal exhibited peculiar movements of the head which were constant. The dog was extremely restless. The pupils were greatly dilated. October 22. The dog was apparently normal at meal time. Symptoms after cocaine injection similar to those of previous days. October 22. Symptoms unchanged. October 24. Rectal temperature at 9:30 a.m. = 38.8° C., just before injection at 3:30 p.m. = 38.6° C.; at 4:30 p.m. = 40.9°C. At 4:30 p.m. the-heart action was very slow but strong. Arhythmic beating was in evidence. There was extreme dilatation of pupil. The animal was very much excited and the head was constantly moved up and down. Usually the animal was too deaf to pay attention to any sound, but at this time it would respond to a call. October 25. Inthe morning the dog appeared normal and devoured food as usual. ‘ Temperature at 9:30 a.m. = 38.8° C., just before injection; at 3:30 p.m. = 38.8° C.; at 5:00 p.m. = 41.1°C. The movements of animal were more pronounced and there was much more excitation after cocaine administration than had been observed at any previous time. The peculiar irregularity of the heart was again in evidence at 5:00 p.m. although previous to the injection, the beat was normal. October 26. The appetite of animal was ravenous. Temperature at 9:30 a.m. = 38.6° C., just before injection; at 3:30 p.m. = 38.6°C., at 4:00 p.m. = 41.6°C., at 5:00 p.m. = 40.9°C. It was apparent that the animal had become much more sensitive to the cocaine since the usual daily injection was followed by greatly augmented symptoms of excitation. These lasted for a period’of two hours. October 27. The dog devoured food with apparent relish. Temperature at 3:15, just before injection = SRibes. \at 345 .= 41.2°C.; at 4:15 = 41.6° C.; at 4:45 = 409° C.; at 5:15 = 39.8°C. The symptoms of excitation and pupil dilatation appeared within fif- teen minutes after cocaine administration. Apparently the peculiar head movements were caused by an attempt to push the head out of the cage 238 CAINE DATE 1910 October 15 16 17 18 19 Average per day... 26 Influence of Cocaine upon Metabolism TABLE 1. Experiment I—Dog 60. Fore Period. (Daily Nitrogen Intake = 7.40 grams.) DAILY DOSE OF CO BODY WEIGHT Ammonia Nitrogen Specific Gravity Total Nitrogen ee 275 128 12.6 260 | | 128 agra 270 128 | 12.6] 210 128 | 12.4) 170 | i 128 | 12.5) 240 | | | + 1.026) 5. 1.030) 6. | (5.0) | 0.30 (4.5) 1.040) 6.66 * Figures in brackets indicate percentages of total nitrogen. Lactic Acid Weight Total Nitrogen Ether Extract 3.30 6.09 | 15.0 | a | 49 lees be? | 35.0 51.0) | 20.0 | 30.0) 41) | [vod ite pai 8. 0717.83 Frank P. Underhill and Clarence L. Black 239 TABLE 1—Continued First Cocaine Period—Continued B URINE ; m1 FECES 2 | eae | | 8 3 Weight | Bed. ees} g E ai 2 DATE ; 5 a = > - 2 g ro} a 6) = 3 3 = * ee i | z 8 s mais pu ee | Se pealas a|s capa 2) eS |e eee ied 1910 ae gms. gms. October | 27 12.3)165 |; 1.045) 6.12) 0.27 (4.4) 28 12.3) 160 | 1.046) 6.54) 0.30 (4.5) 29 r22170' | 1-040) -6:.12/'0733) 71 (5.3) 30 12.2] 165 | 1.041! 6.06) 0.30 4.9) | Average perday... 1.032} 6.19} 0.30} 63 | 1.62 | pee. | Second Cocaine Period. November | | | 1 1.040) 8.25 | 0.36} 84 46 0) 21.0 (4.3) 2 1.050) 6.84 | 0.35 | 83 |52.0)32.0 (51) y 256 | 11.3)120 | 1.052} 5.94 | 0.31 | 79 |53.0)33.0) 38 |3.55)/13.95 (5.2) 4 256 | 11.2) 125 0.31; 80 (5.0) Average per day... 46 |0.88) 3.48 240 Influence of Cocaine upon Metabolism Balanecs Fore Period. grams grams Nitragen in food: <2. .2...... 37.00 Etherextractinfood........ 323 .20 Nitrogen in excreta: Ether extract in feces ........ 6.09 Urmes c= s.'..5 5 ae 30.78 Recess... sees 3.30 34.08 Bat utilized’ ..4 <5 ene Si (a — Fat utilization = 98 per cent. Nitrogen balance........ +2.92 Per day eee.) « +0.58 Nitrogen Utilization = 91 percent. First Cocaine Period. grams grams Nitrogeninfood............. 81.40 Etherextractinfood........ 711.04 Nitrogen in excreta: Ether extract in feces....... 17.83 Urine... / eose-< 68:10 oe Bees ee eo OZ 76.17 Fat utilized. 2.5. 2422. (42 (69o021 Fat utilization = 98 per cent. Nitrogen balance.......... +5.23 Per dayanermte. <8:5...4 +0.47 Nitrogen utilization = 90 per cent. Second Cocaine Period. grams grams Nitrogeninfood............. 29.60 Etherextractinfood.......... 258 56 Nitrogen in excreta: Iether extract in feces......... 13.95 Urine: 2s aeeoe ce 2k ECeSo Eee. O00 30.76 Patitilized) 2) i. oe 244 61 Fat utilization = 94 per cent. Nitrogen balance.........— 1.16 Pera eee =. 22 — 0.29 Nitrogen utilization = 88 per cent. toward the light. During the remainder of this period which was concluded on October 31 no new features developed. It was planned to begin the second cocaine period on October 31 by giving two injections of the drug, at 12:00 m.and4:00 p.m. respectively. The first injection caused vomiting which contaminated the urine. This period was therefore, commenced on the next day, November 1. On this date cocaine in doses of 128 mgrms. each was administered at 3:00 p.m. and 5:00 p.m. Just previous to the first injection the temperature was 38.5° C., at 5:p.m.. 40.0° C., at 6:00, p.m., 40.9°C. The dog was inastate of extreme activity during this time. November 2. Cocaine was injected as on November 1. The conditions of the animal had, however, undergone a marked change since all movements were executed in a weak and uncertain manner. ee Frank P. Underhill and Clarence L. Black 241 TABLE 2. Dog &1. Fore Period. (Daily Intake of Nitrogen=4.72 grams) 2 | URINE > FECES = 8 | i | ¢ i Weight | > co) DATE 6 5 = ae S 2 2 o = ee, 3 | 2 z= ro) a Oo Ss | s 3S s = a Es ° © Z = < Fz cs} es) = ie (ee eee) . | Sa | 8 een el | 2 epee la je PE) a 1910 mgms.| kilos| cc. gms.| gms mgms.| gms.| gms. | seat gms.| gms. November | 30 8.3} 120 1.040/4.56) 0.21 45 *(4.6) | | December; | uae 8.2] 175! 1.035/4.25| 0.23} 46 |52.0.23.0 56 | (5.1) | 2 8.2| 165) 1.036) 4.23 0.23) 51 |17.0)10.0, 42 ‘ | | : ou | 3 8.2| 165| 1.036)4.24) 0.]7| 51 |29.0,18.0) 38 |2.10) 5.46 (4.0) 4 1.040)4.21| 0.16, 45 |20.0,11.0 5 1.030) 4.20 | 15.0 10.0 Average | per day. 4.28} je 12.0 é wi | 3 ilo Nal Cocaine Period. | 7 i | Ph, a co 1.026) 4.44 11.0) 5.0) 55 } | 7 200 1.030, 3.84 23.0/11.0 51 | | 8 155| 1.035,4.35) 0.16 9 150} 1.034) 4.29 | 10 155} 1.035) 3.8 42.0/22.0| 46 | _L * Figures in brackets indicate percentages of total] nitrogen. 242 Influence of Cocaine upon Metabolism TABLE 2—Continued Cocaine Period—Continued 1) URINE Zz : - Tet ote] Lee & =f + » DATE 2 E a iz = 3 a) 2. | an es a MS k: Lal b § a z = be Bsa 8 5 rs A Galea 3) ca < a a 1910 eae kilos | cc. December 11 123 | 7.6) 250} 1.030 12 |123 | 7.6} 130} 1.040 | 13 {123 | 7.5) 140) 1.035 14 | 123 | 7.5) 135| 1.037 15 |123 | 7.5/'145| 1.033 Average per day.|123 | 7.6| 168) 1.033) 4. Balances Fore Period. grams grams Nitrogeninfood............. 28.32 | Etherextractinfood......... 243 .60 Nitrogen in excreta: Ether extract in feces......... 5.46 Wigs cconcvvacs ARO) —_——— Feces! cia syae. 2.10 27.79 HatuGilizedeeeneeesece ree 238.14 Nitrogen balance.......... +0.53 Fat utilization = 97 per cent. IReridiay Aenea: cc) -s.56 +0.08 Nitrogen utilization = 92 per cent. Cocaine Period. grams Nitroreninfoodseeeres - 4 tte 20, Nitrogen in excreta: Urine... eee 02 GCes! iy. seer 3.89 43.91 Nitrogen balance...... +3 .29 Ber d'ay2e epee toasts race +0.33 Nitrogen utilization = 91 per cent. grams Ether extractinfood.......... 406 .00 Ether extract in feces...... 32.54 JAHRDUHNVARC | Geko Gc ao hod 373.46 Fat utilization = 91 per cent. Frank P. Underhill and Clarence L. Black 243 November 3. The dog showed signs of diminished appetite. Conditions remained unchanged. November 4. Conditions about as usual. Animal appears weak. November 5. The dog died twenty-five minutes after the first cocaine injection. Just before death the dog was in a state of extreme activity. This was rapidly followed by a period of partial paralysis culminating in respiratory failure. Further data concerning this experiment may be found in Table 1, pp. 238-240. Protocol of Experiment 2. Dog 51. A fox terrier bitch of 8.3 kilos was placed upon a fixed diet composed of 125 grams meat, 60 grams cracker meal, 20 grams lard, 10 grams bone ash and 150 cc. water for a period of 10 days previous to the actual fore period of the experiment. The, nitrogen content of this diet amounted to 4.72 grams; the fuel value was approximately 69 calories per kilo body weight. November 30. On this date the fore period of six days was begun. December 6. The cocaine period was commenced by the injection of 123 mgms. cocaine at 3:00 p.m. No rise in temperature could be observed. The only symptoms noticeable were salivation and pupil dilatation. December 7. About one-half hour after the administration of cocaine the dog became markedly excited, the bodily movements not being under perfect control. Pupil dilatation was extreme and the arhythmic heart beat was evident. Each day up to December 12 the symptoms of excitement etc. were noticeable but unchanged in character. December 12. Shortly after the cocaine injection the animal became completely paralyzed in the hind-quarters. The jaws and tongue were kept constantly in motion as though the animal was tasting something unpleas- ant. The dog remained in this condition for several hours during which she appeared deaf and blind. December 13. The animal seemed normal although somewhat weak. The weakness became more and more noticeable and on December 15 the experiment was terminated. For other data associated with this animal see Table 2, pp. 241-242. DISCUSSION OF RESULTS. From the details of the protocols and tables submitted it is apparent that the most obvious symptoms arising from cocaine injections in the doses given are distinctly of nervous origin. A significant influence is also exerted upon the heat regulating mech- anism whereby the temperature is quite markedly increased for a short period after which there is a gradual return to the nor- mal. With daily doses of 10 mgms. of cocaine hydrochloride Reichert: Centralbl. f. d. med. Wissenschaften, 1889, p. 444. 244 Influence of Cocaine upon Metabolism per kilo of body weight no appreciable influence can be detected upon the course of nitrogenous metabolism nor upon the utiliza- tion of protein and fat although body weight shows an appreciable decline. When injections of 15 mgms. cocaine per kilo are daily admin- istered fat utilization is very slightly impaired and is accompanied by a decreased body weight. Doses of 20 mgms. per kilo per day divided into two injections show a fairly distinct detrimental in- fluence upon both protein and fat utilization and for the first time a slight negative balance wasin order. Body weight was markedly diminished under this dosage. The water excretion of Dog. 50 was quite distinctly diminished under cocaine when compared with that of the fore-period. This finding does not hold true for Dog 51. The difference may be explained perhaps by the fact that Dog 50 was apparently much more sensitive in its reaction to cocaine with respect to the tem- perature raising influence than was Dog 51. Assuming this to be true more water was probably eliminated by the lungs in the first case than in the second which would account for lessened water elimination by the kidney. THE INFLUENCE OF COCAINE UPON THE ELIMINATION OF LACTIC ACID IN THE URINE. The presence of lactic acid in the urine in appreciable quanti- ties has been a subject of much investigation and discussion result- ing in a multiplicity of conflicting theories with respect to its sig- nificance. Out of the enormous literature’ relative to lactic acid only a few references that have a bearing upon the present paper may be cited. Thus, Araki’ has demonstrated that lactic acid appears in the urine in the absence of a sufficient supply of oxygen induced by various types of toxic compounds and epileptic seizures. The older work of Spiro® indicating that increased muscular activity leads to lactic acid excretion finds confirmation in the recent investiga- 7 Ryffel: Quarteriy Journ. of Med., iii, p. 413, 1909-10. ’ Araki: loc. cit. 9 Spiro: Zeitschr. f. physiol. Chem., i, p. 111, 1877. Frank P. Underhill and Clarence L. Black 245 tions of Ryffel!® and Feidman and Hill. According to the latter authors the appearance of lactic acid in the urine may be greatly diminished by breathing oxygen before and after exertion. They conclude that the increased production of lactic acid by the muscles is due to oxygen want, a view that was earlier denied by Ryffel.” Viewed from the standpoint of ultimate origin, it is possible that lactic acid is intimately associated with the carbohydrate store of the body; for Araki found, under the experimental conditions, less lactic acid in the urine of starving animals than could be dem- onstrated in the urine of those well fed. On the other hand, phosphorus, which leads to a disappearance of the carbohydrate store, causes a large output of lactic acid which may be accompan- ied by an increased elimination of ammonia." It is presumed that the increase of the latter urinary constituent is for the purpose of neutralizing the lactic acid produced. In the experiments to be recorded the rabbits were kept upon a diet consisting of 300 grams of carrots and 20 grams oats which experience had demonstrated would usually be entirely eaten each day. Experiment 3. Rabbit B. During each day of the fore period this animal left small portions of the carrots uneaten. After the subcutaneous cocaine injections no food was ever left. For the first two days of the cocaine period no evidences of ab- normal symptoms were observed. On the third day, however, there was considerable dilatation of the pupil. Beginning with November 9, the tenth day of administration, irritability and restlessness were noticeable. The appetite remained good, all food being eaten shortly after the daily cocaine administration. About 10 minutes after cocaine injection on November 11 the animal was seized with convulsions and respiration almost ceased, but recovery was complete three-quarters of an hour later. On the succeeding two days convulsions were in evidence shortly after cocaine administration, but in each instance recovery was complete. The animal died jn a convul- sion on November 14. The liver which was immediately excised contained 8 per cert of glycogen. From the data in Table 3 it will be observed that the injections of cocaine were progressively increased from approximately 15 mgms. per kilo to 20 10 Ryffel: Journ. of Physiol., xxxix, p. xxix, 1909. 11 Feldman and Hill: Jowrn. of Physiol., xiii. p. 4389, 1911. 12 Ryffel: Journ. of Physiol., xxxix, p. xxix, 1909. 13 Mandel and Lusk: Amer. Journ. of Physiol., xvi, p. 129, 1906. 246 | | DAILY | BODY | DOSE OF WEIGHT DATE | COCAINE | | TABLE 3. Rabbit B. Fore Period. | xy | Specific | Volume | Gravity Influence of Cocaine upon Metabolism URINE Total | Nitrogen Ammonia Nitrogen 1910 October 26 27 | (0.23) 28 2.34 120 1.024 0.75 1.8 (0.23) 29 2.32 105 1.025 0.93 1.8 (0.19) 30 2.32 125 1.025 0.96 1.4 | (0.15) - | Average per day | 2.34 132 | 1.023 0.88 1s : | (0.21) 3l November 1 a 3 34:5 | 2.32 4 34.5 2.32 5 34.5 2.32 6 34.5 2.32 f; 46 2.30 8 57.6 2.30 | | | | 215 | 1.018 215 1.019 250 1.015 210 1.016 185. | 210 | Lactic Acid Frank P. Underhill and Clarence L. Black 247 TABLE 3—Continued Cocaine Period—Continued | URINE DAILY BODY DATE DOSE OF cosas | EM | Volume specie | Tota | Ammonis | Late 1910 ios | November 9 69 2.26 | 235 1.015 | 0.61 L8 26 | (0.30) 10 89 | 2.22 195 Leo 0.61. | 6.3 | 25 G0). | il 101 Cae 180 1.020 | 0.63 18 33 | (0.28) 12 101 2 9 Vall 490 ieoz4eren 62 | 1.1 39 (0.13) 13 Average per day 1.018) 80.75 | 2.2. | 20 * Figures in brackets indicate percentages of total nitrogen. mgms. on November 7, to 25 mgms. per kilo on November 8, to 30 mgms. on November 9, to 40 mgms. on November 10, and finally to 45 mgms. per kilo on November 11. Frequent tests throughout the cocaine period failed to demonstrate an appreciable rise in rectal temperature. Experiment 4. Rabbit C. This animal behaved in a manner very similar to Rabbit B. A rise in rectal temperature of about 0.5° C. was the maximum increa e shown dur- ing the period of observation. The daily dose of cocaine given varied from approximately 10 mgms. per kilo on November 29 and 30, to 20 mgms. on December 1 to 6 inclusive, and from this time to the end of the experiment the animal received approximately 34 mgms. cocaine per kilo body weight. From the data in Tables 3 and 4 with rabbits and those in Tables 1 and 2 with dogs, it is evident that cocaine causes an appreciable increase in the elimination of lactic acid in the urine. In a general way the quantity of lactic acid thus excreted is in direct proportion to the amount of cocaine injected. The output of ammonia, however, does not appear to be significantly increased by the augmented elimination of lactic acid, an indication that in 248 Influence of Cocaine upon Metabolism TABLE 4. Rabbit C. Fore Period. : | es a i= Ee = | URINE DATE Sone ae BODY = COCAINE WHC ea) Wollime Specific a otal Ammonia | Lactic 4 a hes =| = L Gravity itrogen | Nitrogen Acid mgms. kilos. c.c. grams mgms. mgms. 1910 November 21 | | 2.26 290 1.014 | 0.80 1.0 | 24 | (0.12)* 22 | 2.24 295 1.014 0.85 1.0 20 | (0.12) 23 | 2.20 245 1.015 0.83 3.6 23 | | (0.43) 24 2.18 235 1.016 0.83 |. 4.5 22 (0.54) 25 2.20 230 1.016 0.80 3.6 hire (0.45) 26 2.22 190 1.018 0.82 3.6 20 (0.44) PA 2.24 200 1.019 0.81 4.5 2) (0.55) 28 2.26 230 1.018 0.83 23 Average per day 2.22 1.016 0.82 Cocaine Period. 29 (0.38) 30 4.5 23 (0.32) December 1 3.6 24 (0.27) 2 45 2.24 230 1.021 1.26 3.6 24 (0.27) 3 45 | 2.26 220 1022. 0:97 3.6 25 (0:37) 4 45 2.30 230 02 =" Se 4.5 26 (0.52) | Frank P. Underhill and Clarence L. Black 249 TABLE 4—Continued. DAILY c URINE DATE Deo ie ee z a ae og ee oe eee eer ee bia 5c: grams mgms. | mgms. 1910 December 5 45 2.30 215 1.022 0.93 4.5 30 | (0.49) 6 45 2.30 | 215 1.022 0.80 5.4 33 (0.67) | 7 75 2.26 | 205 1.022 0.74 6.3 33 | (0.85) | Shipckiy 25 2324.) 225 1.020 0.72 Scleap.e36 | | (1.12) | G sel in 75 2.26 | 190 1.024 0.83 9.0 40 | | (1.08) 10 75 2.30 | 230 1.020 0.95 9.0 42 | (0.94) | Average 2.26 | 226 1.022 | 0.99 | 5.5 30 per day | (0.55) * Figures in brackets indicate percentages of total nitrogen. this connection lactic acid:may be neutralized by some base other than ammonia. This is particularly true for dogs, but does not hold quite so well with rabbits, for with Rabbit C. the output of ammonia paralleled closely the elimination of lactic acid. The influence of diet upon lactic acid elimination under the experimental conditions may be indirectly inferred from the data of Table 5 obtained from Dog 52 during a period of inanition. Here it will be observed that in spite of largely increased doses of cocaine lactic acid output fell considerably. The larger quantities of lactic acid excreted during the first few days of the experiment may perhaps be explained on the assumption that the carbohy- drate store of the body during this interval had not been depleted. As soon as this condition had been reached a diminution in lactic acid output took place. These results are in harmony with the theories outlined by Araki, but are in opposition to the observa- tions reported for pernicious vomiting of pregnancy where lactic acid is eliminated in the urine probably as a result of the inanition 4 Underhill: This Journal, ii, p. 485, 1906-07; see also Underhill and Rand: Arch. of Int. Med., v, p. 61, 1911. 250 Influence of Cocaine upon Metabolism TABLE 5. Dog 52—Inanition. a — His sted bea DAILY BODY wind ne ATE DOSE OF se aed cocaine | WHONT | vume | Specie | Total | Ammonia | Tactic | mgms. kilos | cc. grams gram. mgms. 1910 | November 10 120 10:2) 460 1.050 6:57: |o 0931 41 | | (4.7)* 11 120 10.2 120 120587 |. ..04e23 5) 0289 38 | | (9.2) (ees |S 19.9 180 1.025 3:06 | 0.34 +) 739 | (11.1) 13 120 | 9.6 | 140 1.035 731; 0126 36 | | (9.5) . a | ae 200 | 1.040 | 4.92 | 0.31 | 32 120 (6.3) 16 2x 150 8.9 70 1.030 £280 (| 038 13 (7.3) 17 2x 150 8.8 160 1.030 3.60 | 0.25 21 (6.9) 18 2 x 150 8.6 100 1.018 0.48 | 0.03 5 (6.2) 19 2x 150 8.5 | 100 1.020 3.25 | 0.12 25 | | (3.6) a ! | eee Bere ’ * Figures in brackets indicate percentages of total nitrogen. which accompanies this pathological state. The observations notedabove are also opposed to the results obtained in phosphorous poisoning’ a condition in which carbohydrate is almost missing from the liver and blood. On the other hand, hydrazine® which behaves in a manner similar to phosphorus with respect to its influence upon the carbohydrate of the organism does not lead to the appearance of appreciable quantities of lactic acid in the urine. From these contradictory results it isapparent that lactic acid must have a diverse origin under the different conditions mentioned. The ammonia content of the urine voided by the dog in a state of inanition was not greatly influenced by the cocaine injections and did not bear a direct relationship to the elimination of lactic acid. 18 Frank and Isaac: Arch. f. exp. Path. u. Pharm., |xiv, p. 374, 1911. 16 Underhill: This Journal, x, p. 159, 1911. ; Frank P. Underhill and Clarence L. Black 251 From the observations here recorded the conclusion may be drawn that the appearance of lactic acid in increased quantity during cocaine poisoning is probably associated with the attendant increased muscular activity induced by the action of the drug upon the nervous system. What relation augmented lactic acid out- put bears to lack of oxygen as claimed by Araki is a problem dif- ficult of decision unless one accepts the view put forth by Feldman and Hill!? that increased muscular work results in a decreased amount of oxygen in the muscles, which in turn causes an increased production and subsequent excretion of lactic acid. It is also apparent that in cocaine poisoning greater quantities of lactic acid are eliminated by given doses of cocaine to well-fed animals than occurs under the same conditions during an interval of starvation. The average elimination of lactic acid during co- caine poisoning in a state of inanition was less than that of other animals maintained in a well-fed condition, but without cocaine administration. It seems probable, therefore, that during cocaine poisoning, carbohydrate material may be intimately associated with the production of lactic acid. CONCLUSIONS. In confirmation of previous investigation, it is found that co- caine introduced subcutaneously into dogs causes a temporary but significant increase in body temperature. With daily doses of 10 mgms. of cocaine hydrochloride per kilo of body weight for short periods of time no influence can be de- tected upon nitrogenous metabolism nor upon fat utilization. Fat utilization is slightly impaired and body weight is consider- ably decreased when daily injections of 15 mgms. cocaine are administered. When the dose of cocaine is increased to 20 mgms. per kilo body weight per day a distinct lowering of both nitrogen and fat utilization is noted. This may be accompanied by a slight nega- tive nitrogen balance. Lactic acid excretion in the urine is markedly increased in well- fed dogs and rabbits as a result of cocaine injection. In a starving 7 Feldman and Hill: loc. cit. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 3. 262 Influence of Cocaine upon Metabolism condition the dog eliminates less lactic acid after cocaine injections than is excreted by the normal well-fed animal. It is not unlikely that the increased lactic acid elimination after cocaine injections is associated with increased muscular activity induced by the drug. The ammonia output apparently bears little relation to lactic acid elimination under the experimental conditions. Lactic acid and carbohydrate metabolism are presumably inti- mately associated although there are indications that lactic acid may at times arise from more than a single antecedent. ON CREATINE IN THE URINE OF CHILDREN. By OTTO FOLIN anp W. DENIS. (From the Biochemical Laboratory of Harvard Medical School, Boston.) (Received for publication, February 29, 1912.) In a recent communication from Mendel’s laboratory! Rose showed by means of an extended series of analyses that the urine of children usually contains relatively large quantities of creatine. These observations are remarkable because they apply only to children and do not correspond to what is found in older people. In adults creatine is believed to be eliminated only when much creatine is taken with the food or when there is an unusual disin- tegration of tissue materials—a condition more or less the reverse of that prevailing in growing children. In view of the unexpected character of the results obtained by Rose we promptly repeated the work on three normal well nour- ished children belonging to one of us (Folin). We had intended to continue the investigation further before publishing anything, but in view of the criticism of Rose’s findings expressed by Wolf? and by McCrudden? and since our results completely verify and also extend Rose’s observations we have decided to publish them now. The subjects are Joanna, age 11, weight 38 K.; George, age 8 years, 8 months, weight 33.5 K; Teresa, age 3 years, 8 months, weight 16 K. If the appearance of creatine in the urine of children were essen- tially the result of carbohydrate starvation one would expect the night urine to be particularly rich in creatine. 1 This Journal, x, p. 265, 1911. 2 This Journal, x, p. 473, 1912. 3 Journ. of Exp. Med., xv, p. 110, 1912. 9 253 254 Creatine in Children’s Urine EXPERIMENT 1. The night urine obtained on the morning of October 17, 1911, yielded per 100 cc. of urine the figures recorded below. (The children had had some meat the preceding noon but none for supper. ) = T NAME CREATININE | CREATINE | mgm. mgm 1 iS Se . ..2 Se 52.0 | 19.0 Pi Gee 0: ee 89.0 | 22.0 a2 Ts .6e6.c ee 37.5 | 9.7 | EXPERIMENT 2. The afternoon urines obtained from the same children about a week later gave the following figures (per 100 cc. of urine). NAME CREATININE CREATINE mgm. za mgm. Telok saan. 2--| 76.5 46.5 ON: eee... | 66.0 15.0 SATE een | 16.3 Say In order to find out whether any hourly variation is present we next determined the creatinine and creatine for each hour from 9 in the morning till 3 in the afternoon in the urine obtained from one of the subjects, J. The results are recorded in Experiment 3. EXPERIMENT 3. Breakfast 9 a.m. (no meat); Dinner 1.15 p.m., including meat. T VOLUME CREATININE CREATI® | R INE cc. mgm. ae mgm. 47 34.8 | 3.7 32 28.8 | 7 30 25.2 | 9.0 40 28.0 | 8.0 80 30.4 | 3.6 98 24.1 | 12.0 The most striking point to be observed in the above results is the sharp and decisive rise in the creatine output, presumably as a result of eating meat at dinner. The breakfast also, although it included no meat is followed by a rise rather than by a fa!] in the creatine output. Otto Folin and W. Denis 255 We are inclined to believe that the creatine in children’s urine does not depend as Rose suggests on a peculiar carbohydrate metabolism but that it is due to an excessively high level of pro- tein consumption (in proportion to mass of muscles in the body). We know that the creatine output in response to creatine feeding depends very much on the level of protein metabolism maintained. Whether the creatine is taken with the food, as in the experiments of Folin and of Klercker, or whether it comes directly from the tissues, as in fevers (and possibly also in starvation) may be more or less immaterial in view of the fact ascertained by us‘ that creatinine and creatine like urea and amino acids® are promptly transported from the digestive tract to the blood and from the blood to the tissues. We hope later to prove (or disprove) ex- perimentally the validity of this point of view. In this paper we wish merely to corroborate Rose’s interesting findings as to the fact that creatine is nearly always found in the urine of children. EXPERIMENr 4. Twenty-four hour urine on a mixed diet including some meat at noon. VOLUME TOTAL NITROGEN CREATININE, | CREATINE cc. ap grams mgm. | mgm. 1340 10.1 643 258 1000 11.6 810 90 680 6.2 | 219 186 4 The absorption experiments with creatine and creatinine will be de- scribed in detail later. The creatinine we have traced from the intestine through the blood to the tissues by means of colorimetric creatinine estima- tions as well as by nitrogen determinations. 5 This Journal, xi, p. 87, 1912. 6 If the above hypothesis is correct it should be possible to reproduce in adults by forced feeding with protein which contains no creatine the condi- tion with reference to creatine found in children and it should also be pos- sible to obtain creatine free urine from children by reducing their protein consumption. 256 Creatine in Children’s Urine EXPERIMENT 5. Twenty-four hour urine from mixed diets containing no creatine. NAME DAY | VOLUME | Pisin ae CREATININE | CREATINE TEES +— te. pee + cc. grams mgm. | mgm. Re | First 1050: acl ees 720 160 dee .| Second 1085 7.8 399 140 tit tte | Third | 955 | 7.9 477 86 Ge ce First 790 8.6 620 100 Gad. Se Second 1330 | 10.6 385 119 Gaile toast Third 1075s 11.45 481 150 ~ First | 4.9 190 90 | 850 ca EXPERIMENT 5. Night urine from Dr. A. H. Wentworth’s children, Elizabeth, age 9 years, 6 months, and Charles, age 6, on a creatine-free diet. The figures are given for 100 cc. of urine. NAME DAY NITROGEN CREATININE ea haat | grams mgm. mgm. E.....,....| Second 0.86 34 3 Bee Third | 0.85 34 3 HS Sree Fourth ibaa l7/ 36 : 4 Che ace | Second 0.96 31 4 Crest ee Third 0.75 24 6 Coen eeeourth 0.96 30 4 EXPERIMENT 6. Through the kindness of Professor Wiener we are able to include in our determinations the morning urine obtained from his four healthy and unusually robust children, all of whom are vegetarians and have never eaten any food containing creatine.—Norbert, age 17 years, 3 months. Constance, age 13 years, 10 months, Bertha, age 9 years, 10 months, Fritz, age 6. CREATININE As before the figures are given for 100 cc. of urine. CREATINE A NEW METHOD FOR THE DETERMINATION OF HIPPURIC ACID IN URINE. By OTTO FOLIN anp FRED F. FLANDERS! (From the Biochemical Laboratory of Harvard Medical School, Boston.) (Received for publication, February 29, 1912.) Bunge and Schmiedeberg’s well known method for the deter- mination of hippuric acid in urine was published in 1876. That method is neither accurate nor convenient. It has survived evi- dently only because no one has succeeded in devising anything better. “The more recent methods which have been proposed from time to time have been only modifications of that method. They retain the tedious extraction by means of acetic ether and depend for their accuracy on the isolation of perfectly pure hippuric acid.” Benzoic acid is less soluble in water and much more soluble in organic solvents than is hippuric acid. The quantitative extrac- tion of benzoic acid and its determination by direct titration in the organic solvent (chloroform) is,a relatively simple, convenient and exact process for the determination of benzoic acid,’ in prod- ucts far more difficult to handle than urine. If hippuric acid could be conveniently hydrolyzed into benzoic acid and glycocoll the determination of hippuric acid in urine might be made almost as simple as the determination of benzoic acid. In our attempt to work out a method for determining hippuric acid according to this scheme we met with many unforseen difficulties and some sur- prises but the final outcome is, we believe, reasonably satisfactory. While it is generally recognized that it is possible to split hip- puric acid by either acids or alkalies, the former are in practice 1 Published with the approval of the committee as work done under a Bullard Fellowship, 1911-1912. 2 For the most recent modification see Dakin: This Journal, vii, p. 103, 1910. 3 Folin and Flanders: Journ. of the Amer. Chem. Soc., xxxiil, p. 161, 1911. 257 258 Hippuric Acid Determination uniformly preferred for that purpose. In fact hippuric acid is tacitly assumed to be more stable in weakly alkaline than in neu- tral or acid solutions, for in preparing hippuric acid from urine some alkali (caletum hydrate or sodium carbonate) is génerally added before the urine is concentrated. Definite data on the sub- ject we have not been able to find.4 It has not been our aim to furnish such data because our purpose was to accomplish the quan- titative hydrolysis of hippuric acid in urine under conditions that wouid permit a rapid and convenient extraction of the benzoic acid from the resulting mixture. In other words emulsion with the organic solvent used for the extraction was with us the most serious obstacle to be avoided and the hydrolysis had to be made with that end in view. Incidentally we have, however, ascertained a few specific facts as to the stability of hippuric acid which are worth recording. They are contained in the table below. WEIGHT | | mippurtc |voLuste| HYDROLYZING AGENT TEMPERATURE . | | HYDROLYZED ACID gram | ce. | gram per cent 0.2 | 50 | 0.01 gm. NaOH 16 hrs. on water) bath 0.0058 2.9 0.2 50 | 0.025 gm. NaOH. 16 hrs. on water | | bath 0.0066| 3.3 0.2 | 50 | 0.05 gm. NaOH 16 hrs. on water bath 0.0116 5.8 0.2 50 | 0.25 gm. NaOH 16 hrs. on water | | | bath 0.1966 | 98.2 0.2 | 50 | 0.2gm. Na,CO; 16 hrs. on water bath 0.0117 5.8 0.2 | 50 | 2.0 gm. Na,CO; | 16 hrs. on water | bath 0.0208 10.4 0.2 50 2 gms. urea 16 hrs. on water, bath None | None 0.2 | 50 0.5 gm. acetic acid | 16 hrs. on water | bath Trace | 0.2 150 | Excess of milk of lime | Boiled hr. Trace 0.2 100 | Excessofmilkoflime | Boiled3}hrs. | 0.0092} 4.6 0.03 30 4.5 gms. HCl Boiled 13 hrs. 0.027 | 90.0 0.15| 75 | 11.4gms. HCl | Boiled 14 hrs. 0.1323 | 88.2 0.15.| 70 18 gms. HNO; | Boiled 13 hrs. | 0.1242 | 82.8 «See, however, Dessaignes: Journ. Pr. Chem., (1) xxxvii, p. 244, 1846. Otto Folin and Fred F. Flanders 259 WEIGHT HXDROLYZING AGENT Sai eee gram per cent 18 gms. HNO; Boiled 3 hrs. 0.1280 | 85.3 18 gms. HNO; Boiled 8 hrs. | 0.1303 | 86.8 J 9 mgs. HNO; | +0.2gm.Cu(NO;)> | Boiled 14 hrs. 0.0620 | 41.3 J 9 gms. HNO; j + 0.2gm. Hg Boiled 14 hrs. 0.0240 | 16.0 9 gms. HNO; |} +35gms.NaNO; | Boiled 14 hrs. 0.620 | 41.3 if 9 gms. HNO; | S + 35 gms. NaNO; Boiled 3 hrs. 0.1395 | 93.0 18 gms. HNO; | 2 | +1gm.Cu(NO;). | Boiled 14 hrs. 0.1257 | 83.8 {18 gms. HNO; { 1 gm. Cu(NO3)> | (35 gms. NaNO; Boiled 14 hrs. 0.1411 | 94.1 f 18 gms. HNO; : } + .0.2 gm. Hg | Boiled 3 hrs. | 0.1227] 81.8 18 gms. HNO; } 35 gms. NaNO; | Boiled 3 hrs. PPO sie yg) Seiresil | 18 gms. HNO; } +0.2gm.Cu(NO;)2 | Boiled3 hrs. | 0.1441 | 96.0 18 gms. HNO; | | } +0.2gm.Cu(NOs;)2 | Boiled 4hrs. | 0.1500 | 100 Pees Cu(NOs)2 | Boiled 43 hrs. 0.0504 | 100.8 23 gms. HNO; ae Boiled4hrs. | 0.1000 | 100 23 gms. HNO; \ +0.2gm.Cu(NO.). | Boiled 43 hrs. 0.1487 | 99.1 J 23 gms. NHO; | | +0.2gm.Cu(NO;)2 | Boiled 43 hrs. 0.1973 | 98.7 The figures given in the above table show that while it is pos- sible to split hippuric acid quantitatively by boiling with mineral acids the treatment required for this purpose is rather heroic. The quantitative decomposition is much more easily accomplished by means of alkalies and there can hardly be any doubt but that hippuric acid is much less decomposed in the presence of dilute acids than in the presence of small amounts of alkali. In view of these findings it is clearly a mistake to render urine alkaline before 260 Hippuric Acid Determination concentrating it when preparing hippuric acid from urine and still more so when the hippuric acid is to be extracted for quantitative determinations. The lack of agreement among investigators on the transformation of benzoic acid to hippuric acid in the animal body is doubtless due in part at least, to losses of hippuric acid by its transformation back into benzoic acid in the urine, during the concentration of the latter. . In our method the hydrolysis of the hippuric acid is on the other hand an essential feature. In fact we lost much time in experi- menting with various acids, catalyzers, oxidizing reagents, etc., to bring about complete hydrolysis before we discovered that the greater part of the hippuric acid is split while the urine is being concentrated on the water bath. After having discovered such a convenient and effective method for splitting the hippuric acid it would seem that one should be able to merely acidify the urine and at once extract with chloroform. Ultimately a way will doubtless be found to do this but so far we have been unable to accomplish it satisfactorily. The extraction of the benzoic acid with chloroform is neat, clean, rapid and complete only when it is not complicated by emulsions. The best way which we have found to eliminate the emulsion and the coloring matters of the urine is to boil the urine for several hours with comparatively strong nitric acid. The method in detail, as finally adopted, is as follows: Measure 100 cc. of urine into a porcelain evaporating dish by means of a pipette. Add 10 cc. of 5 per cent NaOH and evaporate to dryness on the steam bath. (If the sample is placed on the bath at night it will be dry in the morning.) Transfer the residue to a 500 cc. Kjeldahl flash by means of 25 cc. of water, and 25 cc. of cone. HNO;. Add 0.2 gram copper nitrate, a couple of pebbles or glass pearls and boil very gently four and one-half hours over a microburner. . The necks of the flasks are fitted with Hopkins condensers, made from large test tubes which fit rather loosely. A good current of water flowing through the condensers prevents loss of benzoic acid or change in concentration of the nitric acid. The accom- panying photograph shows the arrangement of the apparatus. After cooling the condensers are rinsed down with 25 cc. of water, and the contents of the flask are transferred to a 500 cc. Otto Folin and Fred F. Flanders 261 separatory funnel by the use of 25 cc. more of water. The total volume of the solution is now 100 cc. Add to the solution sufficient ammonium sulphate to just saturate it (about 55 grams). Make four extractions with freely washed chloroform, using 50, 35, 25 and 25 ce. portions. The first two portions may be used to further rinse out the Kjeldahl flask. The separatory funnels may be shaken vigorously as there is practically no tendency to form an emulsion. The successive portions of chloroform are collected in another separatory funnel. Add to the combined extracts 100 cc. of sat- urated solution of pure sodium chloride, to each liter of which has been added 0.5 ce. of concentrated HCl. Shake well, draw the chloroform into a dry 500 cc. Erlenmeyer flask and titrate with qv sodium alcoholate, using four or five drops of phenolphthalein as indicator. The first distinct end point should be taken, al- though it may fade on standing a short time. 262 Hippuric Acid Determination The sodium ethylate is made by dissolving 2.3 grams of cleaned metallic sodium in one liter of absolute alcohol.5 It is advisable that it be slightly weaker rather than stronger than tenth-normal. It may be standardized against purified benzoic acid in washed chloroform, or with certain restrictions againsi tenth-normal hydro- chloric acid in aqueous solution. In a recent contribution, it was stated that the value found by titration in aqueous solution was slightly higher than that found by the chloroform. The cause of this variation has been traced to sodium carbonate, which is formed by the absorption of carbon dioxide. The point has an interesting theoretical, as well as practical side. A rather large quantity of the ethylate was gradually used with frequent opening over a period of three months. As it was nearly exhausted, quite a pre- cipitate was noticed in the bottom of the bottle. At this juncture titrations were made against acid solutions of equivalent normality in order to test the standard. The results appeared as follows: Ten cubic centimeters 35 oxalic acid in chloroform required 5.55, 5.6 and 5.6 ce. of the ethylate. Ten cubic centimeters 75 hydrochloric acid in aqueous solution required 5.4, 5.4, 5.4 cc. of theethylate. The ethylate was filtered, after which it was not quite transparent, but freed from nearly all the precipitated carbonate. The titrations were repeated with results as follows: Ten cubic centimeters 5 oxalic acid in chloroform required 5.6 and 5.65 ec. ethylate. Ten cubic centimeters 75 aqueous oxalic acid required 5.52, 5.55 and 5.58 ec. sodium ethylate. Ten cubic centimeters 75 aqueous hydrochloric required 5.55 and 5.55 ce. ethylate. To further emphasize the point, the same quantities of <> oxalic solution in chloroform were titrated after adding 0.1 gram of dry sodium carbonate to each: 10 ce. z oxalic acid required 5.5 and 5.6 ce. of the ethylate. From this it is plain that the sodium car- bonate does not influence the titration in chloroform, but of course does materially affect the aqueous titrations. The following results may be cited as showing the agreement in duplicates obtainable by this method. We believe that they are > This Journal, vil, p. 423. § Journ. Amer. Chem. Soc.,xxxili, p. 1625, 1911. al Otto Folin and Fred F. Flanders 263 more nearly correct than are the figures obtainable by any other method. atm: 5 eae VOLUME OF URINE TOTAL ERE SES TTS: | cc. cc. 1 5.30 800 | 0.764 5.45 0.786 9 8.55 1,040 1.602 8.65 | SE 3 4.65 1,180 0.990 4.70 1.000 4 5.30 1,200 sobs 5.35 1.158 5 7.00 610 0.770 6.90 0.760 6 5.30 870 0.832 ae25 0.825 7 8.60 700 1.086 8 30 : 1.049 8 14.6 730 1.920 | 14.4 [ 1.880 | ai aaa tbo apaaetan®, wits 08. gh geRMMEIRS oly tite el pai inte iGo ® 15 memitanivey Sh wks Mae lee Ts | Hee Tt? Reais, gas. tenthelicic Ley revels guts er Ayes) Oa m ape Niketa ans. evihe eee vk a petee ge voy eae | ie ote oY 24a dy ¢ 1 Ulirodras itn i Lea Siew ey. 54) Pe eee 7 — ~Laie ' 7 é Ajlta 257 (i ~ er (i. pe : a a ~ G ‘ = a 4 ; Ayr: 4 i hi LA Sipe =) Cpa te 4k Vege eer aa scent ay fi ue Se ar of ee a = =) ee z ve ‘ : ce -_- B: a a‘. fa) a ee TE AA TEN Case aan tik 2S Talis Vode ce ere Souinty 4 a ee : . ef oua’y av anima : re -_ “—_ vias rive Lei 4 a aed at ~ ta Pa ; vee ee nn: te Theta So a es yy wri nee ee Fe hrit <1 Merahiotis +). psi Alpe = iit 0). cea ’. ly ai iain ys hae pias agy Hiptica er. Se eh Sat nth: * fe vier’ piace - : _ i a * aan co ‘~S > he oy ut tae Fi... a] 7 ¥ - . : ~~ ms Wragr ine hy ert, te; Gas TO F aa: - * My , ar hd ? ‘ = ; Oi DS 3 % \ 2 Ke 7 _ 2 Sc tes = ee, ar: . rs ss eae ~~ ‘2 ma * i) ; | ae mi Ae ar ; 1) * — E os —_ Aa eee a ‘ i , : of) a J , 4 i > ary sea Bs ON THE BLUE COLOR REACTION OF PHOSPHOTUNG- STIC ACID (?) WITH URIC ACID AND OTHER SUBSTANCES. (PRELIMINARY PAPER.) By OTTO FOLIN anp A. B. MACALLUM. (From the Biochemical Laboratory of Harvard Medical School, Boston, Mass.) (Received for publication, March 19, 1912.) The beautiful blue color which is produced when phosphotung- stic acid and an alkali are added to uric acid lends itself unusu- ally well to quantitative work. After several months spent in trying to devise a direct method for a colorimetric determination of uric acid in urine on the basis of this reaction, we have reluc- tantly come to the conclusion that this is not feasible, because of the presence in urine of substances other than uric acid which give the same reaction. Jn uric acid solutions the colorimetric values obtained are sharply proportionate to the amount of uric acid present, and the color fades so slowly when the conditions are right that we do not hesitate to pronounce the reaction eminently suitable for the determination of small quantities of uric acid. The reaction is almost instantaneous and the color remains practically unchanged for almost ten minutes, so that no difficulty is experienced in making the necessary quantitative comparisons by means of a colorimeter. To secure the maximum color of the desired stability the strong alkalies usually employed in making the reaction can not be used; a saturated solution of sodium carbonate is very much better. In the course of our further studies we have discovered that the color in question is given not only by uric acid but is charac- teristic of phenols, and that in the case of more complex aromatic compounds it is particularly, if not exclusively, those containing a hydroxal group in the para position which give the color. This discovery has of course given a new turn to our investigations. 265 266 Color Reactions with Phosphotungstic Acid We believe that the reaction will be found fully as useful as Millon’s for the detection of certain aromatic groups in protein substances, and that it has the advantage of being particularly suitable for quantitative work. Among the substances which give the reaotion may be mentioned phenol, tyrosine, tannic acid, thymol, orcin, resorein, vanillin, and phloroglucin, besides a number of less definite protein materials. A more detailed study of this interesting reaction and its application for the detection and determination of such aromatic products will be undertaken as soon as we get through with the uric acid work. The best procedure which we have yet found for the determination of uric acid in urine is to precipitate the uric acid by means of silver sulphate and magnesia mixture, cen- trifuge, and make the color reaction on the precipitate in the presence of formaldehyde. (The latter is added to reduce the silver.) It would be useless to describe the method in detail at the present time, for we have found that different samples of phos- photungstic acid (and phosphomolybdie acid) do not produce the same intensity of color. In fact the material which produces the blue color with uric acid and with phenols is probably not phosphotungstic acid. Whether it is a tungsten product at all, or some other substance present as impurity, we have not yet been able to determine for lack of material. We have learned how to concentrate the active agent and to separate it from the greater part of the phosphotungstic acid, but more material and more work will be required before we shall know what it is and how to get it free from the useless, as well as expensive, phos- photungstic acid. STUDIES IN THE ACTION OF TRYPSIN. I.. ON THE HYDROLYSIS OF CASEIN BY TRYPSIN. By E. H. WALTERS. (From the Rudolph Spreckels Physiological Laboratory of the University of California.) (Received for publication, February 28, 1912.) I. INTRODUCTION. (a) Objects of the investigation. It is a well known fact that the digestion of protein is essentially a process of hydrolysis. This reaction may be represented by the following schematic equation: HXXOH + H.0 = HXOH + HXOH Protein + water = amino acid + amino acid. The mode of operation of this reaction is entirety beyond the category of our present knowledge. The researches of Emil Fischer! on the hydrolysis of the synthetic polypeptides will, it is hoped, throw some light on this question, for many of these synthetic products are known to be hydrolyzed by trypsin. It has been found that this reaction proceeds at all tempera- tures in neutral watery solutions free from a ferment or any other catalyser except, possibly, the ions of the water itself. Taylor? working with sterile solutions of casein, protamine sulphate, and nucleoprotein in pure water and Robertson’ working with casein have demonstrated directly that these bodies are hydrolyzed at 1Emil Fischer: Numerous papers in the Berichte der deutsch. chem. Gesellsch., and in the Zeitschr. f. physiol. Chem. during the past decade. ?Taylor: On Fermentation, Univ. Calif. Pub. Pathol., 1, p. 97, 1907. 3T. Brailsford Robertson: The Proteins, Univ. Calif. Pub. Physiol., iu, p. 174, 1909 (see footnote). 267 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO 4 268 Action of Trypsin ordinary temperatures. Robertson found that comparatively rapid hydrolysis occurs when neutral solutions of the caseinates in pure sterile water are kept at 36° C. For a 2.8 per cent solu- tion of neutral sodium caseinate the velocity constant (common logarithms being employed and time expressed in hours) was found after twenty days at 36° C. to be 0.000518. These observations are enough to show that a protein is not in equi- librium with its products. Kihne’* pointed out that the portion of the albumin molecule which is readily dissociated by the action of trypsin contains a high percentage of tyrosine and tryptophane, while the portion which was not acted upon by trypsin is characterized by the presence of glycocoll and a carbohydrate radical. The high tyrosine and tryptophane contents and the absence of glycocoll and a carbohydrate radical renders casein especially readily digestible. The action of trypsin on caseinates in neutral or faintly alkaline solution is, therefore, one of the best examples among protein reactions of the action of an enzyme in accelerat- ing an already progressing chemical transformation. Observations already made on the mode of the above reactions when accelerated by a proteolytic ferment and to which refer- ences will be made 7n loco indicate that the process of the hydroly- sis of proteins by trypsin or pepsin obeys the monomolecular reaction formula during the initial stages of the reaction when the influence of the products of hydrolysis is practically nil. This research was undertaken to determine, under more exactly defined conditions than hitherto, the influence of certain factors upon the hydrolysis of proteins by trypsin. The objects of the investigation are three-fold; firstly, to ascertain the relation be- tween the time of hydrolysis and the amount of protein hydro- lyzed; secondly, to ascertain the relation between the ferment con- centration and the velocity of-hydrolysis for different concentra- tions of protein; and thirdly, to determine the relation between the nature of the base combined with a protein and the velocity with which it is hydrolyzed. Rigidly comparable solutions of the caseinates can very readily 4Kihne: quoted after Mann, Chemistry of the Proteens, Macmillan Co., London, 1906, p. 148. E. H. Walters 269 be prepared,® and in view of the useful properties of casein in studying the process of the hydrolysis of a protein by an enzyme, the ease with which it is attacked by trypsin, its accurate quanti- tative estimation, and since the action of trypsin on casein has already been studied somewhat extensively, I have held to this system in the present investigation. (b) Influence of antiseptics. It is obvious that in experiments of this character the solutions under investigation must be kept sterile by the action of an anti- septic which does not alter the course of the reaction or impede the action of the ferment. It has been found by different observ- ers that toluol satisfies the conditions very satisfactorily. Weis® has found that toluol does not sensibly affect the action of tryp- sin or pepsin when used in just the quantities necessary to main- tain sterility. On the contrary thymol, chloroform, formol, benzoic acid and salicylic acid slightly retarded the enzyme action. Kaufmann’ found that toluol, chloroform, thymo!, and sodium fluoride destroyed the action of trypsin after a considerably long period of time; after twenty-four hours’ action, the most concen- trated solutions of the ferment were rendered inactive. This effect, however, is not entirely due to the action of the antiseptic as it is well known that trypsin is very readily rendered inactive on standing in pure water, although the destruction of the ferment may be accelerated in the presence of an antiseptic. Bayliss* found that toluol slightly accelerated the action of trypsin during the course of an experiment and that chloroform had no perceptible influence. I used toluol in all of my experiments in the propor- tion of 0.2 cc. to 100 cc. of casein solution as it was found that 100 cc. of a 2 per cent solution of “basic”? sodium caseinate contain- ing this amount of antiseptic were still sterile after the lapse of three weeks. No systematic study was made to determine the influence of the antiseptic but according to Weis the quantity used would not alter the reaction or the activity of the enzyme. 5'T. Brailsford Robertson: Journ of Physical Chem., xiv, p. 377, 1910. 6 Weis: Compt. rend. des trav. du lab. de Carlsberg, v. p. 133, 1900. 7Kaufmann: Zeitschr. f. physiol. Chem., xxxix, p. 434, 1903. 8 Bayliss: The Kinetics of Tryptic Action, Arch. sci. biol. (St. Peters- burg, 1904), 11 Suppl., p. 261; reprinted in the Collected Papers of the Physio- logical Laboratory, University College, London, xiii, 1903-5. 270 Action of Trypsin (c) The influence of alkalies and acids upon digestion by trypsin. It is now generally recognized that trypsin acts most energet- ically in faintly alkaline solutions although the evidence upon which this fact rests is extraordinarily at variance. According to Heidenhain® the action of definite percentages of NasCO; varies with the quantity of the ferment, but for moderate con- centrations the activity is most pronounced in solutions of 0.9 to 1.2 per cent Na,CO3. He also pointed out that the addition of.0.1 per cent HCl to an aqueous extract of the pancreas entirely stops its action. Kihne!® found that trypsin acts in HCl solu- tions up to 0.05 per cent, above whichits activity ceased altogether. , In his experiments, however, its action was most pronounced in solutions of 0.3 per cent NazCO3. Mays" and Ewald” also found that trypsin could digest fibrin in the presence of 0.3 per cent HCl only when large amounts of the protein were present, and Mays confirmed the statement of Kiihne that trypsin is destroyed by pepsin ard HCl. On the other hand, Engesser’s® experi- ments show that pancreatic juice did not lose its digestive power by two hours warming with a gastric juice containing 0.5 per cent HCl. Langley“ conducted a series of experiments which show that a glycerine extract of the pancreas when warmed for two and a half hours in solutions of 0.05 per cent HCl decreases considerably in its digestive power, a result diametrically opposed to that of Engesser’s. Lindberger!® found that fibrin was very slowly dissolved by trypsin in the presence of 0.012 per cent HCl and that the action of the ferment entirely ceased in the presence of 0.1 per cent HCl. He also observed that weaker acids, as acetic and lactic, had a much less retarding effect than the stronger HCl and that tryptic digestion was rapid and in some ®Heidenhain: Pfliiger’s Archiv, x, p. 557, 1875. 10Kuhne: Verh. Naturhist. med. Vereins zu Heidelberg, 1877, p. 193; quoted after Chittenden and Cummins, Amer. Chem. Journ., vii, p. 36, 1885. 11 Mays: Untersuch. a.d. physiol. Institut in Heidelberg, iii, p. 378, 1880; quoted from Maly’s Jahresbericht, x, p. 299, 1880. 12 Ewald: Zeitschr. f. klin. Med., i, p. 615; quoted from Maly’s Jahres- bericht, x, p. 297, 1880. 13 Engesser: Zeitschr.f. klin. Med.,ii, p. 192; quoted from Maly’s Jahres- bericht, x, p. 297, 1880. 14 Langley: Journ. of Physiol., iii, p. 246, 1880. 16 Lindberger: Maly’s Jahresbericht, xiii, p. 280, 1883. E. H. Walters 271 cases even more energetic in the presence of small quantities of these acids than in neutral solutions. Chittenden and Cummins" found that the addition of Na,CO; to 0.2 per cent increased the tryptic digestion of fibrin and be- tween 0.2 and 0.5 per cent NazCO3 the action was about the same and above 0.5 per cent the action was greatly retarded. These authors also found that very small amounts of hydrochloric and salicylic acids greatly retard the action, its proteolytic action being retarded to a minimum before any free acid is present. Three- tenths per cent combined HCl has a great retarding effect, and the same amount of combined salicylic acid plus 0.1 per cent free salicylic acid produces similar results. Much smaller quantities of combined salicylic acid (0.06 per cent) have the same effect. Combined hydrochloric acid has a greater hindering action than salicylic acid. Vernon!’ observed that trypsin is very rapidly destroyed in 0.4 per cent solutions of Na,CO3. His results indicate that a series of ‘‘trypsins’”’ might exist as different preparations were unequally affected by constant amounts of sodium carbonate, the least sensitive ones being more resistant to the action of the alkali. The preparations were also more resistant to the action of sodium carbonate in the presence of large amounts of protein. This statement was subsequently confirmed by Bayliss and Starling.!® Schierbeck!® states that carbonic acid augments the action of trypsin in alkaline solutions since it diminishes the alkalinity of the solution. According to Bayliss,2° Kanitz,24. Taylor,” Robertson and Schmidt,” and Kudo the influence of alkalies and acids is due to 16 Chittenden and Cummins: Amer. Chem. Journ., vii, p. 36, 1885. 17 Vernon: Journ. of Physiol., xxvi, p. 427, 1900. 18 Bayliss and Starling: Journ. of Physiol., xxxil, p. 129, 1905. 19 Schierbeck: Skand. Arch. f. Physiol., iii, p. 344, 1892. 20 Bayliss: The Kinetics of Tryptic action, Arch. sci. biol. (St. Peters- burg, 1904) 11 Suppl., p. 261; reprinted in the Collected Papers of the Univer- silty College Physiological Laboratory, London, xiii, 1903-5. 21 Kanitz: Zeitschr. f. physiol. Chem., xxxvii, p. 75, 1902. 22 Taylor: On Fermentation, Univ. of Calif. Pub. Pathol., i, p. 251, 1907. °23'T. Brailsford Robertson and C. L. A. Schmidt: This Journal, v, p. 3) 1908. 24 Kudo: Biochem. Zeitschr., xv, p. 473, 1909. 272 Action of Trypsin the OH and H ions since it is recognized that in most cases the acids and alkalies act in proportion to their degree of dissociation. Thus Kanitz has observed from the results of Dietz” that the action of the hydroxides of Ca, Sr, and Ba upon tryptic digestion is a function of their degree of dissociation and that the optimum OH ion concentration lies between 7's and zoo normal. This is an observation almost identical with that of Vernon’s® although his result is expressed in terms of sodium carbonate as per cent. In experiments on the digestion of protamine by trypsin Taylor learned by the aid of the gas cell that the most favorable initial concentration of alkali is that which is sufficient to neutralize about zoo0 acid solution after neutralization of the products of the hydrolysis, which are slightly acid. Kudo has found that trypsin acts best in neutral solutions and is inhibited by acids and alkalies in proportion to their degree of dissociation into H and OH ions. Robertson and Schmidt made an investigation to de- termine the part played by the alkali in digestion by trypsin Sodium caseinate and protamine sulphate were used as substrate and the alkalinities of the digests were followed throughout the digestion and determined by means of the gas-chain. It was found that the change in OH~ concentration with time obeyed the monomolecular formula for all alkalinities above 10-* normal and for concentrations less than this the velocity of the reaction diminished and the bimolecular formula held good. Moreover, this value, 10-§ n OH-, at which the order of the reaction changes is independent of the nature of the protein or the initial concen- tration of the alkali. It was concluded, therefore, that all alka- linities between yooo000 and about zsoo are equally favorable for tryptic action. The latter value yso0, as we have seen, is the one determined by Taylor. On the other hand the results of Berg and Gies?’-do not support this view although it was recognized that the H and OH ions were the favorable acid and alkali factors. No regular results were obtained in equivalent solutions of different bases and it appeared *% Dietz: Einfluss von Baryumoxyhydrat, Calciumoxyhydrat, Strontium- oxyhydrat auf die tryptische Verdauung, Inaug. Dissertation, Leipzig, 1900. *6 Loc. cit. For the calculation of this value consult Shield: Zeitschr. f phystk. Chem., xii, p. 167, 1893. 27 Berg and Gies: This Journal, ii, p. 489, 1906. E. H. Walters 295 that the cations or molecules (or both) exercised deterrent in- , fluences. Loeb?® explains the accelerating action of alkalies in tryptic digestion by assuming that the enzyme is a weak acid and upon the addition of alkali a salt is formed which is more strongly dis- sociated than the acid itself. This latter is based upon the fact that salts of weak bases and acids are more highly dissociated than the free bases and acids themselves. If the enzyme action is, therefore, due to the enzyme ion, its acting mass will be greater in the presence of enzyme salts. In my experiments the initial OH~ concentrations were under rigid control. The solutions were made so that the proportion of base to casein = 80 X 10 equivalents per gram. These solutions are neutral to phenolphthalein, 7.e., faintly alkaline, the OH- concentration being 107° n.”° (d) Method of measurement. Most of the conflicting data on the hydrolysis of proteins has resulted largely from the inaccuracy of the measurements. In most of the investigations the mode of measurement has not been adequate to estimate the actual transformation since in nearly all cases the errors in the methods have ranged from 10 to 30 pel cent. Inthe majority of the studies some alteration in the physi- cal properties of the solutions such as electrical conductivity, viscosity, osmotic pressure, optical activity, etc., have been used as a means of determining the degree of hydrolysis. Moreover, the chemical methods that have been employed have been based upon more or less empirical estimates in which the substances | measured bore no direct chemical relation to the amount of sub- stance undergoing transformation. ‘These methods will be alluded to throughout the paper in connection with the experiments in which they were employed. The most reliable of the properties of the proteins for obtaining constants characteristic of them is undoubtedly the number representing the quantity of nitrogen bound up in the molecule. 28 Jacques Loeb: Biochem. Zeitschr., xix, p. 534, 1909. 29See T. Brailsford Robertson: Journ. of Phys. Chem., xiv, p. 528, 1910. 274 Action of Trypsin Casein can be very accurately determined by making use of this reliable property and this method has been used in this investiga- tion. The numbers representing the quantity of nitrogen, ex- pressed in per cent, in purified casein from cow’s milk obtained by different investigators are practically identical as may be observed from the following figures.*° Percentage of nitrogen in Observer purified casein from cow’s milk EVA ATS TEE cts 5. < aoa eek ae tee 15.65 ChittendenandPainter......0. 402255... 15.91 Lehniannang'tempel .; 2-2 ee. ee 15.60 Elienberpeteseree. 22.02 eS Re eee 16.64 PQS ENO: cna 6 5 REE tenet A ETS aN A Ne 15.70 From the mean of six determinations on the anhydrous puri- fied casein used in the experiments described below I have ob- tained the number 15.81. To calculate the equivalent amount of casein from the nitrogen I have used the factor 6.4 which means that 1 cc. 7 alkali is equivalent to 9 mgs. of casein. The method actually employed was as follows: The casein was precipitated from 100 ce. of the solutions under investi- gation by a slight excess of 7y acetic acid (made up approximately by diluting 10 cc. of Kahlbaum’s glacial acetic acid to 1750 cc.). The quantity of acetic acid varied from 15 to 30 ce. according to the quantity of casein in solution. Not less than 15 cc. was added even in the most dilute solutions of casein as a slight excess did not appear to affect the accuracy of the deter- mination. Furthermore, the nature of the filtrates is an index to the quan- tity necessary for complete precipitation. The persistence of a cloudy filtrate after refiltering several times indicates incomplete precipitation which necessitates the addition of more acetic acid. A cloudy filtrate which cannot be removed by repeated filtering very often results by directly precipitating casein in neutral or faintly alkaline solutions by acetic acid. This cloudiness can be overcome by redissolving the precipitate in a slight excess of alkali and immediately re-precipitating with acetic acid, or by adding just enough ¥ KOH (usually about 1 cc.) to remove the opalescence in solutions of this type and then immediately add- ing acetic acid in slight excess to completely precipitate the casein. The latter method was adopted as it is more convenient and a clear filtrate always resulted when Schleicher and Schiill’s No. 590 ‘‘white band”’ filters were used. A finely divided precipitate which often occurs when casein is 50 Quoted after Mann: Chemistry of the Proteids, p. 397, MacMillan Co. E. H. Walters 275 precipitated in dilute solutions and especially in solutions of the hydroxides of the alkaline earths and which is difficult and often impossible to filter can be avoided if the acetic acid is added slowly (a few drops at a time) and the solutions vigorously shaken during precipitation and then allowed tostand for about an hour before filtering. The precipitate was thoroughly washed by decantation and on the filter with distilled water as free of CO. and NH; as could be achieved by boiling and the filter paper containing the precipitated casein transferred to a Kjeldahl digestion flask and the total nitrogen determined according to the Kjeldahl method.*! The following are some of the figures obtained by this method with purified anhydrous casein: WEIGHED AMOUNT OF CASEIN SOLVENT P In 100 cc. ESTIMATED CASEIN PIT mgs. 3 mgs. AYEK O18 lo gels ee Or 86 85 NE(QIEL Goo $ bold eee 238 235 Inj Ele. 2 Pl) eee 854 845 MOMs Res tee 560 556 Ca(OH) eee scoess eek: 324 319 CH(OH ee ous ha os: | 93 89 Ba(Oliete seca woe. 738 728 JB. 0i5 De ie 219 212 In obtaining the above figures it was necessary to work under conditions which would eliminate as far as possible the error due to hydrolysis. If the solution in which the casein is dissolved is too alkaline the high concentration of the OH ions causes rapid hydrolysis before the casein is completely dissolved. On the other hand, considerable amounts of casein will be hydrolyzed in very dilute solutions due to the long period of time which has elapsed before the state of complete solution is reached. Robert- son” observed this fact in his studies in the electrochemistry of the proteins while measuring the conductivity of solutions of potassium caseinate in solutions of varying OH ion concentra- tions, and it was found that the proportion, 10 cc. of 7; KOH to 1 gram of casein, gave the most satisfactory results. 31 The Kjedahl method as described on page 5 of Bulletin No. 107 (Re- vised) of the Bureau of Chemistry, U. S. Department of Agriculture, was strictly followed in making the nitrogen determinations. 32 T. Brailsford Robertson: Journ. of Physical Chem., xiv, p. 528, 1910. 276 Action of Trypsin I have used this same proportion for NaOH and KOH but as casein dissolves more slowly in solutions of Ca(OH). and Ba(OH). a large proportion of alkali to casein (15 cc. to 1 gram) was re- quired. In preparing the casein solutions, therefore, from which the above results were obtained casein was dissolved in solutions of NaOH and KOH in the proportion of 10 ce. of 4 alkali to 1 gram of casein and in solutions of Ca(OH): and Ba(OH), in the proportion of 15 cc. of * alkali to 1 gram casein. Under these conditions hydrolysis would occur to a slight degree but I think the extent was reduced to a minimum. The casein was estimated in 100 ce. of the‘respective solutions immediately upon complete solution by the method outlined above. Il. EXPERIMENTAL. (a) General procedure. The casein employed in all of the experiments was Eimer and Amend’s C. P. Casein ‘‘nach Hammarsten,” further purified ac- cording to the method of Robertson®* which is as follows: Half a pound of the casein was triturated with about 12 liters of distilled water, the water being added in six successive portions. On each addition of water the casein was well stirred up in it in a porcelain mortar and then allowed to settle, then the supernatant water was poured off and fresh water was added. It was then washed in asimilar manner in 5 kilos of Kahlbaum’s C. P. alcohol, 99.8 per cent, and then in 5 kilos of Kahlbaum’s C. P. ether, distilled over sodium. The mortar containing the casein drained as free from superfluous ether as possible,*4 was then placed.in an incubator over sulphuric acid at 40 to 50° C., the flame was turned out under the incubator and it was allowed to cool for about twenty-four hours. The casein is now found, if these operations have been conducted carefully, to be in the form of a dry, pure white powder, still containing, however, a considerable quan- tity of ether. The casein was now spread out, within the incubator, in a layer not over 1 cm. deep, the flame under the incubator was lighted, fresh sulphuric acid was introduced if necessary, and it was allowed to stand for twenty-four hours at 40 to 50° C. The casein is then found to be free from appreciable water or ether. 33 Thid. 34 At this point it is necessary to avoid exposing the mortar to the moist air of the room a minute longer than is necessary, otherwise the evaporat- ing ether causes condensation of sufficient moisture to spoil the product unless it is again treated with alcohol and ether. EH. Walters 277 Robertson finds that casein prepared in the above manner gives every indication of being a pure product. In one instance the same author® finds that casein thus prepared loses 5.8 per cent of its weight when dried for five hours at 70 to 80° C. I find that casein prepared as above loses 3.8 per cent of its weight after five hours’ heating at 70° C. or 4.19 per cent: after five hours heating at 100°C. At 100°C., however, Lacqueur and Sackur*® find that casein is decomposed. The commercial trypsin prepared by Griibler of Leipzig was used in all of the experiments. This preparation contains a small insoluble residue and as more concordant results could be ob- tained with filtered solutions than with suspensions, filtered solu- tions were employed throughout, although they did not appear to be as active. Taylor*’ also observed the same conditions in his experiments on the hydrolysis of protamine by trypsin. The incubator used was the double walled type employed by bacteriologists. It was provided with two doors, the inner a glass door and an outer double walled one. It would easily hold seventy-two Erlenmeyer flasks of 200 cc. capacity; its inside dimensions being 36 cm. deep, 45 cm. wide, and 72 cm. high. A temperature, constant within 0.5°C., could be maintained through- out the course of an experiment. ; Schleicher and Schiill’s ‘‘white band” quantitative filters No. 590 (11 cm.) were used throughout, as these were found to give an inappreciable blank in the nitrogen determinat‘ons and they held the precipitated casein especially well and filtration was comparatively rapid. (b) Relation between the time of hydrolysis and the amount of protein hydrolysed. Henri and Larguier de Bancels** have studied the digestion of gelatine and casein by trypsin using the electrical conductivity 35 T. Brailsford Robertson: This Journal, ii, p. 326, 1907. 36 Lacquer and Sackur: Beitrage z. chem. Physiol. und Pathol., iii, p. 193, 1902. 37 Taylor: On the Hydrolysis of Protamine with Especial Reference’ to the Action of Trypsin, Univ. Calif. Pub. Pathol., i, p. 7, 1904. 38 Henri and de Bancels: Compt. rend. acad. sci., cxxxvi, pp. 1088 and 1581, 1903. 278 Action of Trypsin method as measurement. They followed the curve only forty minutes, however, but during this brief interval it was found that the process of tryptic digestion follows the law for monomolecular reactions. The constants were also in fair agreement in series with two different substrate concentrations. Furthermore, their results confirm the hypothesis that the action of trypsin is not a pure catalytic reaction and that an intermediate compound is formed between the trypsin and substrate. Bayliss®® has investigated extensively the progress of the action of trypsin on casein and gelatine by the electrical conductivity method. He prepared an eight per cent solution of sodium case- inate and to 6 cc. of this solution were added 2 ec. of { ammonia, 2 ce. of a 2 per cent solution of trypsin, and a few drops of toluol. The conductivity was measured at different times and its in- crease at 39° C. plotted in a curve which tends to become asymp- totic to the base line, indicating that the velocity of the reaction approaches zero, and that an equilibrium point is reached before the reaction is completed. A mathematical analysis of his re- sults shows that the velocity constant calculated from the mono- molecular equation diminishes somewhat rapidly during the course of the reaction. The following are some of the values obtained when the equation K = - log. — in which ¢ is the time which has elapsed since the beginning of the reaction, ais the initial concentration of the substrate, and x is the amount of products formed during the time ¢, so that a—z is the sub- strate concentration at the end of the time f, is applied to the rate of hydrolysis. MirstGeEneminutes). << ct sects ee es ee nO) Seconditenvminutess sss. ee ee eee 1K = (D) ABhindstensminiites.<: : oc.. cece Shee Were eases ce er Reece ae K=0 Bourthscensmimnuutes.,; -. sas sete eee K = 0.0022 Minthetenimalmuces:.. -- nee eae ee eae Ls Sat eee I 0 Seventhatensminutess: <2 eta ee ne ope ony eee Ke —0 Nintistensmimites: 2 oc. & sap amt ete nee Nn tes De iKe—20 39 Bayliss: The Kinetics of Tryptic Action, Arch. des sct. biol., 11 Suppl., p. 261, 1904; reprinted in the Collected Papers of the Physiological Laboratory, University College, London, xiii. E. H. Walters 279 This phenomenon was due either to the retarding action of the products of hydrolysis or to the destruction of the trypsin or to both factors simultaneously. Bayliss worked with very alkaline solutions and this high degree of alkalinity must have caused a very rapid destruction of the trypsin. In a very extensive investigation on the hydrolysis of casein by trypsin Robertson*® by a different method found that the action of trypsin on calcium caseinate obeys the monomolecular formula during the first stages of the reaction. The amount of casein digested was estimated by a volumetric method based upon the fact that whenever a solution containing casein is neutral to phenolphthalein the proportion of base to casein = 80 X 107° equivalents per gram, 7.e., using phenolphathlein as indicator, 1 gram of casein is almost exactly equivalent to 8 ce. 4} alkali solu- tion. Briefly, the method consisted in dissolving the casein, which was precipitated by acetic acid, in a slight excess of a standardized Ca(OH). solution and subsequently titrating the uncombined alkali with a standard solution of HCl. From the results obtained the method appears to be a very accurate one. This method has been used by Hart*! in estimating the quantity of casein in milk. With the view of throwing further light on this question, the following experiment was undertaken. Seven liters of a 0.4 per cent solution of ‘‘basic’’ sodium caseinate were made by dissolving 28 grams of purified casein in 224 cc. of 7; NaOH and diluting to 7 liters with distilled water free from carbon dioxide and 100 cc. placed in Erlenmeyer flasks of 200 cc. capacity provided with tightly fitting rubber stoppers. Sixty-six flasks all of the same kind, lightly stoppered,* were placed in the incubator and after the solution in each flask had reached the temperature of the incubator, as indicated by a thermometer immersed in the liquid, 0.2 cc. toluol and 1 cc. of a 0.2 per cent filtered solution of tryp- sin were added to each flask which was tightly closed and returned to the incubator and digested at 37.5° C. + 0.5°.. Three samples to which no tryp- sin had been added were taken out and the casein determined in the usual way and is considered as the initial amount of casein present. Three sam- ples were taken out and the casein determined in each after every fifteen 40'T. Brailsford Robertson: This Journal, ui, p. 317, 1907. 41 Hart: This Journal, vi, p. 445, 1909. 42 The flasks should not be closed tightly while being warmed, otherwise the increasing pressure may cause them to break. 280 Action of Trypsin minutes for the first three hours; after every half hour for the next three hours; and after every hour for the next three hours so that the reaction was followed for nine hours after 82.29 per cent of the casein had been com- pletely hydrolyzed. To reduce the error of handling such a large number of samples to a minimum I proceededin thefollowingway. Toluol was first added to each flask after its contents had arrived at the temperature of the incubator, tightly closed, and replaced in the incubator. One sample at a time was then taken out, the trypsin solution added by means of a warmed pipette, and the time accurately noted. This sample was replaced in the incubator and allowed to digest for 9 hours. The second and third sam- ples were withdrawn in like manner, the trypsin solution added, and the flask replaced and allowed todigestfornine hours. The next three samples treated exactly the same were allowed to digest for eight and one-half hours. This process was continued so that the last three samples were only allowed to digest for fifteen minutes. I repeated this operation three times on a small scale before the actual experiment was inaugurated and found that I could handle it very conveniently with the smallest possible error. If this reaction obeys the law of mass action, the rate of change at any moment will be proportional to the concentration of the casein at that moment according to the equation where a is the initial amount of casein present, z the amount of it hydrolyzed in time ¢, and K the velocity constant. Integrat- ing this expression, we get: —ln(a—2x) = Kt+ constant At the beginning of the reaction, f = 0, x = 0, and we have:. — ln a= constant, Ina—In(a—2)=Kt Now from the values of a — zg obtained it is possible to cal- culate the velocity constants at different times. Instead of using natural logarithms I have used common logarithms throughout which is 0.4343 times the natural. The following were the re- sults obtained when equation (2) is applied to the rate of hydrol- ysis: E. H. Walters 281 TABLE I. TIME IN CASEIN DIGESTED, | Serer e CASE- LOG 10 a K MINUTES I.E., Z IN, Les, G-—2 a-x Atle mgs. mgs. 0 0 367 15 22 345 0.02685 1S. 10-4 30 38 329 0.04747 16 x 107 45 54* 313 0.06913 15 << 10-4 60. 65 302 0.08466 14 xX 1074 75 75 292 0.09929 13° x 107 90 84 283 0.13288 12°51, 1074 105 94 273 0.12851 12> bc 10-4 120 104 263 0.14471 12 >< 104 135 115 252 0.16327 12=> xr10-4 150 133 234 0.19545 13 10s 165 142 225 0.21249 1 ams be | 180 150 217 0.22821 12.5 xX 107+ 210 174 193 0.27911 13) ><.10=' 240 183 184 0.29985 12.5 X 1074 270 194 173 0.32662 12 107 300 221 146 0.40032 13.5 X 1074 330 234 133 0.44082 $3 SOO 360 254 113 0.51159 1A 10-4 420 278 89 0.61528 14.5 xX 1074 480 292 75 0.68961 | 14 x 1074 540 302 65 0.75176 |t4 x 107 *This number is the mean of two determinations. With one exception the figures in the second column are each the mean of three determinations. These results suffice to show very clearly that the velocity of hydrolysis at any moment is proportional to concentration of the casein at that moment. The slight lagging of the constant at the beginning of the experiment is in all probability due to some uncontrollable error in handling the experiment during such short intervals with the method employed and which would gradually be eliminated during the progress of the experiment. The results indicate, also, that the products of hydrolysis have very little, if any, influence upon the velocity of the reaction. Fig. I, in which the ordinates represent the amount of casein hydrolyzed in milligrams and the abscissae the time in minutes, shows the point in these experiments more lucidly. 282 Action of Trypsin (c) The relation between the concentration of the trypsin and the velocity of hydrolysis for different concentrations of protein. The literature on the relation between ferment mass and the velocity with which a protein is hydrolyzed is strewn with many > ee) | ae e SELAH Se EEE (- Ame eee MGS.— CASEIN TIME MINUTES FIG.1. contradictory statements. This confusion resulted largely from the use of the oft-described method of Mette* in determining the rate of hydrolysis. This method consists in subjecting short “Mette: quoted after Samojloff. Arch. des sci. biol. de St. Petersburg, ii p. 707, 1893. ; E. H. Walters 283 capillary glass tubes containing a solid protein such as fibrin or egg-albumin to the action of a ferment and measuring the quantity of protein dissolved. This method has been criticised by Taylor“ who has shown that it possesses only a qualitative value for work of a physico-chemical nature. Moreover, the very limited knowl- edge concerning the nature and mode of action of ferments led to many questionable experimental conditions with respect to acidity, alkalinity, or salt content of the digests or the influence of various external factors which have recently been brought to light and shown to have a very considerable influence on the ve- locity with which a protein is hydrolyzed by an enzyme. With the accumulation of facts pertinent to the chemical na- ture of the proteins, methods have sprung up whereby purer and much simpler substrates admitting of a more accurate measure- ment can be obtained. The more recent observers, therefore, concur in stating that the velocity of protein hydrolysis by an enzyme within a short range of temperature, and for certain con- centrations of substrate, is directly proportional to the concen- tration of the ferment provided the latter is not rapidly destroyed by high concentrations of acids or alkalies. As early as 1859 Briicke® studied the digestion of fibrin by pepsin and observed that the digestive power increased slowly as the quantity of pepsin was increased up to a certain limit above which additional quantities of pepsin had little or no effect. This early observation by Briicke was confirmed by the subse- quent investigation of ae 46 Mayer,*7 Ellenberger and Hof- meister,*® and Klug.*® Borissoff,®5° in 1891, by the aid of Mette’s method, found that “4 Taylor: On the Hydrolysis of Protamine with Especial Reference to the Action of Trypsin, Univ. of Calif. Pub. Pathol., i, p. 7, 1904. 4 Brucke: Sitzungsber. Wien. Akad., xxxvii, p. 131, 1859; quoted after Samojloff, Arch. des. sci. biol. de St. Petersburg, ii, p. 701, 1893. 46 Maly: Hermann’s Handbuch der Physiol., v, (2) p. 73, 1881. 47 Mayer: Zeitschr. f. Biol., xvii, p. 351, 1881. 48 Ellenberger and Hofmeister: Arch. f. Wiss. u. pract. Thierheilk., ix, p. 185, 1883; quoted after Carl Oppenheimer, Ferments and their Action, Eng. trans. by Mitchell, London, 1901, p. 95. 49 Klug: Pfliger’s Archiv, lx, p. 43, 1895. 59 Borissoff: La substance zymogene de la pepsine et sa transformation en pepsine active, (Thesis in Russian) St. Petersburg, 1891; quoted after Samojloff, Arch. des sci. biol. de St. Petersburg, ii, p. 705, 1893. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 284 Action of Trypsin the rate of digestion of egg-albumin by trypsin was proportional to the square root of the quantity of trypsin present, and Samoj- loff®' using the same method found the same rule to hold good only in dilute solutions. E. Schiitz*® showed that the velocity of the hydrolysis of fibrin by pepsin was proportional to the square root of the quantity of the ferment when the quantity varied as 1 to 64, and Jul. Schiitz®* finds that this rule holds for the peptic digestion of egg-albumin, and Walter and Vernon® confirm it for digestion by trypsin. In 1895 Sjoqvist®* introduced a new method for determining the velocity of hydrolysis of proteins by ferments. -He noted the changes in the electrical conductivity in solutions of egg-albumin when acted upon by pepsin and hydrochloric acid. In his experi- ments four different concentrations of pepsin were allowed to act upon constant concentrations of egg-albumin at 37° C. and the conductivity of the different solutions measured at given inter- vals. He found that the rate of hydrolysis during the first stages of the reaction was proportional to the square root of the mass of pepsin. On the other hand Schiitz.and Huppert,®” Pollok,®*? and Saw- jalow®® by the aid of Mette’s method were unable to confirm the rule of Schiitz but found that the quantity of digested protein was proportional to the quantity of ferment. Loehiein®® found the rate of hydrolysis proportional to the square root of the quantity of pepsin and directly proportional for trypsin. He employed the acidimetric method of Volhard® which consists in precipitat- ing the undigested casein by sodium sulphate from solutions of 51 Samojloff: Arch. des sci. biol. de Si. Petersbuyg, i, p. 699, 1893. 52 Emil Schiitz: Zeitschr. f. physiol. Chem., ix, p. 577, 1895. 33 Julius Schiitz: Zeitschr. f. physiol. Chem., xxx, p. 1, 1900. 54 Walter: Arch. des sct. biol. de St. Petersburg, vii, p. 1, 1899. 55 Vernon: Journ. of Physiol., xxvi, p. 405, 1900. 8 John Sjéqvist: Skand. Arch. f. Physiol., v, (part iii), p. 354, 1895. 57K. Schiitz and Huppert: Pfliiger’s Archiv, xxx, p. 470, 1900. 58 Pollok: Bettr. z. chem. Physiol. u. Pathol., vi, p. 95, 1904. 59 Sawjalow: Zetischr. f. physiol. Chem., xlvi, p. 307, 1905. 69 Loehlein: Beitr. z. chem. Physiol. u. Pathol., vii, p. 120; quoted after Taylor, Univ. Calif. Pub. Pathol., i, p. 243, 1907. 61 Volhard: Miinch. med. Wochenschr., Nos. 49 and 50, 1903; quoted after T. Brailsford Robertson, This Journal, ii, p. 328, 1907. E. H. Walters 285 casein hydrochloride, and subsequently titrating the quantity of free acid in the filtrate. Faubel® by the aid of the same method found that digestion by trypsin was proportional to the concentration of the ferment. Fuld® and Gross™ found that the time required for digestion by trypsin is inversely proportional to the quantity of ferment. Gross followed the reaction by noting the moment at which the digesting solution ceased to give a precipitate with 1 per cent acetic acid. Palladin® has recently studied digestion by trypsin and found that when a protein is in a state of solution the amount hydrolyzed is directly proportional to the concentration of the ferment. The author made use of the following method. A solid protein (fibrin) was dyed with “spirit blue, blue shade” and immersed in a solution of trypsin. It was found that as hydroly- sis proceeds the solution becomes colored proportionately to the amount of protein dissolved. By making colorimetric compari- sons with solutions of known concentrations the amount of pro- tein digested was estimated. The method was also employed to determine the quantity of trypsin in solutions of unknown concentrations. Weis® carried out some experiments on the digestion of the protein from wheat by means of malt extract. In some of his experiments the concentration of protein was varied as well as the concentration of the ferment (trypsin or pepsin). It was observed that within a certain limit the amount of protein digested was proportional to the quantity of acting ferment above which it varied as the concentration of the substrate. Arrhenius®’ and Euler®® have calculated the constants from the results ob- tained by Weis and Arrhenius concludes that the quantity of protein hydrolyzed is inversely proportional to the square root of the substrate concentration. The somewhat complicated sys- 82 Faubel: Beitr. zur chem. Physiol. u. Pathol., x, p. 35, 1907. 3 Fuld: Arch. f. exp. Pathol. u. Pharm., lviii, p. 468, 1908. 6 Gross: Arch. f. exp. Pathol. u. Pharm., lviii, p. 157, 1908. 6° Palladin: Pfliger’s Archiv, cxxxiv, p. 337, 1910. 86 Weis: Compt. rend. des trav. du lab. d. Carlsberg, v, p. 133, 1900-03. 87 Arrhenius: Immunochemistry, p. 84, MacMillan Co., New York, 1907. 68 Kuler: Allgemeine Chemie der Enzyme, p. 130, Wiesbaden, Verlag von J. F. Bergmann. 286 Action of Trypsin tem (substrate and ferment) used in these experiments renders the results not very convincing. Henri and Larguier des Bancels (loc. cit.) and Bayliss. (loc. cit.) by the electrical conductivity method confirmed the rule of direct proportionality for the tryptic digestion of gelatin and casein, and Taylor (loc. cit.) showed that the tryptic digestion of protamine obeys the same law. Hedin®® studied the digestion of casein, serum-albumin, and white of egg by trypsin. He found that the rate of digestion was directly proportional to the quantity of ferment acting when all other conditions were kept constant, and also the rate of digestion is directly proportional to the time under other- wise constant conditions. In other words, the velocity of hy- a drolysis obeys the formula, log 10 ee Kft, where a is the initial concentration of substrate, x the amount of it transformed in time t, f the concentration of ferment, and K the velocity constant. Adirect proportionality between the substrate concen- tration and the rate ot change was observed, although the con- stants were not concordant for different substrate concentrations and the effect per unit of casein increases as the total amount of casein diminishes, and finally becomes constant. The activity of the ferment diminished during the course of the reaction and the products of hydrolysis appeared to exert a considerable depressant influence. It was also noted that the total effect was not affected by dilution with water, in other words, if the ratio between the ferment and trypsin be kept constant, the effect for equal volumes is proportional to the concentration. Hedin measured the reaction by estimating the quantity of nitrogen that escaped precipitation by tannic acid. This method has been criticised by Taylor.7? Later experiments on the digestion of casein by trypsin yielded similar results when the rate of change was measured by estimating the quantity of phos- phorus split off during the course of the reaction. Quite recently Robertson’! made an investigation of this subject using a more refined method of measurement. Calcium 69 Hedin: Journ. of Physiol., xxxii, p. 468, 1904; xxxiv, p. 370, 1906. 70Taylor: On Fermentation, Univ. Calif. Pub. Pathol., i, p. 236, 1907. 7°T. Brailsford Robertson: This Journal, ii, p. 317, 1907. E. H. Walters 287 and barium caseinates were used as substrates. It was found that the velocity of hydrolysis of calcium caseinate is directly proportional to the amount of trypsin for small concentrations of substrate. For concentrations, above z¢5 Ca(OH). saturated with casein, the velocity constant calculated from the mono- molecular formula increased with increasing quantities of fer- ment. In this case the author states that the constant for the the velocity of hydrolysis is equal to af + bf?, where f is the concentration of the ferment and a and 6 are constants. By representing the ratio log jo to the number of cubic centi- a-—2z meters of trypsin solution by y it was found that y =41 4 11 f. This phenomenon was not improbably due to the destruction of the ferment by the uncertain OH~ concentrations of the solu- tions employed as my observations are not in accord with this theory. In most of the experiments quoted above the concentrations of the substrate were almost always constant. It appeared important, therefore, to repeat some of these experiments with various concentrations of protein. Basic sodium caseinate was used as substrate in all of the experiments and observations were made on the relation between the concentration of the trypsin for 0.125, 0.25, 0.5, 1, 2, and 4 per cent solutions. EXPERIMENT 1. Eight grams of purified casein were dissolved in 64 ce. 75 NaOH and diluted to 6400 cc. with distilled water free from carbon diox- ide. Sixty-two Erlenmeyer flasks of 200 cc. capacity containing 100 cc. of the above solution were lightly stoppered with clean rubber stoppers and placed in the incubator. As soon as the solutions had reached the tempera- ture of the incubator, 0.2 cc. of toluol were added to each digest and the flasks tightly stoppered and replaced’in-the incubator. After the tempera- ture of the solutions was readjusted, two samples were removed and the casein determined in each which gave the initial amount of casein present. As quickly as possible 0.5 cc. of a 0.025 per cent solution of trypsin were added to 6 flasks (set A), one flask at a time being removed from the incubat- ing oven, and allowed to digest at 37.5° C. + 0.25°. To six flasks of another set (B) was added 1 cc. of trypsin solution, to six flasks of another set (C) 2 ce., to six flasks of another set (D) 3 cc. and so forth up to 9 cc. of trypsin solution, thus making ten sets of six flasks each representing ten different concentrations of ferment to a constant concentration of substrate. Two flasks from each set were withdrawn after one, two, and three hours and the undigested casein determined in each in the usual way. Insets A, B, and C 288 Action of Trypsin containing 0.5, 1, and 2 cc. of trypsin respectively, the casein digested dur- - ing the first two hours was too slight to be measured accurately and so the tabulated results for these sets give only the change that occurred during the first three hours. Now if the velocity of transformation is directly proportional to the con- centration of the trypsin, formula (2) becomes Log iy 2 Raph NF, Oe ee (3) a—-waZz where k; is a constant and f is the concentration of the ferment. The fol- lowing tables contain the observed results as well as those calculated by means of the above formula. The initial amount of casein in each digest was found to be 108 mgs. TABLE II. Sertes A, B, and C. a = 108: ¢ = 3 hours. cc. TRYPSIN 5s soos pide = fe Ki + = 0.5 104.85 0.01285 8.5 XK 10 1 98.70 0.03910 13. KOs 2 90.90 0.07486 12.5 X 103 Series D. f = 3ce. t a HOURS es ie ares Ka Series E. f = Acc. Ki 0.04175 102 DO 83.50 0.11173 QromcnlOm E. H. Walters 289 TABLE IJ—Continued. Serves F. f = 5cee | iat a—2Z LOGio ae : Ki 0 108 ] 97 .20* 0.04575 9 X 10 2 85.28 0.10257 105 105¢ 3 78.63 0.13783 4) <1 *This number is based upon one determination. Series G. f = 6cc. t a es a-zZz LOG1o a Ki 0 108 1 93.15 0.06424 10.5 X 107° 2 80.43 0.12800 10.5 X 10° 3 72.90 0.17069 S25 10m Series H. f = 7ce. t a HOURS a-—-z LOG10 ee | Ki 0 108 if 88.43 0.08682 12.5 < 107° 2 79 .29 0.13420 9.5 X 1073 3 | 68 .04 0.21066 LOR al Ome Series I. f = 8ce. t a oa a—Zz | LOGi0 Ga Ki 0 108 1 87.53 0.09126 TRS SX LOm 2 75.6 0.15438 9° 5)x< 10m? 33 64.53 | 0.22366 Dea Ge LO 290 Action of Trypsin TABLE II—Continued. Series J. f = 9ec. ws — : aes oS LOGio0 perioe | Ki es — ! 0 108 | 1 85.95 | 0.09917 1 xe 10m 2 73.80 | 0.16536 9 x 1073 3 60.75 0.24987 | oa Ome The numbers in the column under a — x are each the mean of two determinations. The figures in the last column are very satisfying and show that at this substrate concentration the velocity of hydrolysis is proportional to the concentration of the ferment. The figure on the following page brings out this fact more clearly. ExPERIMENT 2: Four grams of purified casein were dissolved in 32 ec.7> NaOH and diluted to 1600 ec. with distilled water free from carbon dioxide, and 100 cc. measured out in each of 12 Erlenmeyer flasks of 200 cc. capacity and 0.2 ce. toluol and 1, 3, 5, 7, and 9 cc. of a 0.065 per cent solution of trypsin was added to the digests. The experiments were carried out in duplicate. The initial quantity of casein was found to be 217.35 migs., this being the mean of two determinations. The solutions were allowed to digest at 38.5° C. + 0.5° C. for three hours. The results obtained are tabulated in the following table: TABLE III. Q@=2(,05¢6— > hours: CUBIC CENTIMETERS a—t Log 10 Ki TRYPSIN a 1 170.10 0.10646 sje 10m The experiment was done in duplicate and the figures in the second column are each the mean of two determinations. EXPERIMENT4. Exactly asin the two previous experiments except that a 1 per cent solution of casein, 16 grams of pure casein dissolved in 128 cc. +, NaOH and diluted to 1600 cc., was used in place of a 0.5 per cent solution. The concentrations of the trypsin used were 1, 2, 4, 6, and 8 cc. of a 0.2 percentsolution. The digestion was carried out at 39.5° C. + 0.5° for three hours. TABLE V. @) = Shion OUrs: P CUBIC CENTIMETERS | a-—2Zz LOGio0 Ky TRYPSIN ad é 1 651.75 0.11931 40 x 1073 ve 486.54 0.24628 41 x 10° 4 293.90 0.46519 39 X 10 6 161.23 0.72594 40 <10s 8 90.12 0.97857 41 10 rs EXPERIMENT 5. In this experiment a 2 per cent solution of ‘‘basic’’ sodium caseinate, made by dissolving 32 grams of casein in 256 cc. 7 NaOH and diluting to 1600 cc., was used as substrate. 1, 2,4, 6, and 8cc. of a0.4 per cent solution of trypsin were employed. The digestion was carried. out for three hours at 39° C. + 0.5°. The initial quantity of caseinin 100 cc. was found to be 1618 mgs. E. H. Walters 293 TABLE VI. a= 1618 Sie— 3S hours: J, CUBIC CENTIMETERS os TRYPSIN 1 | 1239 0.11591 38.5 X 1073 2 965 0.22400 37.5 X 1073 4 573 0.45083 37.5 X 1073 6 332 0.68784 38 x 1073 8 223 0.86068 so4-><10- EXPERIMENT 6. Precisely as in the previous experiments except that a 4 per cent solution of ‘‘basic sodium’’ caseinate, made by dissolving 64 grams of purified casein in 512 cc. 77 NaOH and diluting to 1600 cc., was employed as substrate. ec. was found to be 3123 mgs. three hours at 39° C. + 0.5°. The concentrations of trypsin used were 1, 3, 5, 7, -and 9 ce. of a 0.8 per cent solution. The initial quantity of casein in 100 The solutions were allowed to digest for The following results were obtained. TABLE VII. i — ol2a tsa Ours: f | : CUBIC CENTIMETERS a—2Zz LOGi0 Ki TRYPSIN 3 a 1 2420 0.11075 37. x 1073 3 1437 0.33711 37.5 X 1073 5 846 | 0.56720 38 x 107 7 544 0.75897 36 xX 1073 9 407 | 0.88498 33 x 10-3 In the foregoing experiments it will be observed that the pro- portion of ferment to protein was practically the same and that the velocity constants in the entire series approach the same value, the non-concordance being due to the decay of the ferment through physical disturbances. It is a well known fact that trypsin preparations gradually lose their digestive power by the physical disturbances of the laboratory.” EXPERIMENT 7. In this experiment five different concentrations of ‘‘basic’’ sodium caseinate (0.2, 0.4, 1, 2, and 4 per cent solutions) were 72 See Taylor: On the Hydrolysis of Protamine with Especial Reference to the action of Trypsin, Univ. Calif. Pub. Pathol., i, p. 7; zbid., On Fer- mentation, p. 249, 1907. 294 Action of Trypsin digested simultaneously with varying amounts of the same ferment solu- tion. tion was carried out for three hours at 36° C. + 0.5°. The figures in the column under a — z are each the mean of two determinations. were obtained. A 0.2 per cent solution of trypsin was used throughout and the diges- The following results TABLE VIII. (A) a = 177 mgs. bes: ; Per ann) «| eee STED | CNDISEOREP Al eae rere x a—zZz St. MO i ts | [ mgs. mgs. 119 58 | 0.161513 28.5 0.154291 : 27.5 A VCLAR Ch eI. '.\=' oie ait seletpt nce wag arene Meee A aoe | 28 (B) = 360 mgs 1 i) aig Py oraeseo 26.5 2 233 | 127 | 0.0754160 27.0 an 280 | 80 | 0.0725786 26 Oneal 4 394 | 46 | 0.0744616 27.0 | aie INSURER SOE: ORR Span Go oleic Gl aeeOr eee tay ea 26.6 (C) a = 882 mgs. _ 1 Vt. 38) 0.026700 23.5 2 262 620 0.025513 22.5 3 380 | 502 | 0.027196 24.0 4 468 | 414 | 0.027373 24.0 IASViCT ALE: Ae en eI ak oe SOR eR ae TT ee a ee eee 23.5 (D) a = 1733 mgs Z 337 | 1396 0.015651 27 3 463 | 1270 0.015000 26 4 528 | 1205 0.013151 23 10 1055 | 678 0.013585 | 23.5 SA VETALES Ne SAEED cee ous is Se eat se a ee ee 24.9 E. H. Walters 295 TABLE VIII—Continued. (£) a = 3326 mgs. 2 320 | 3006 0.007322 24.5 4 518 2808 0.006126 20.5 6 750° | 2576 0.006165 20.5 10 | 1136 | 2190 0.006049 20.0 as pots ted | 2) eee LAER EOE SOS ei ee rie enn nS ereiegea 21.4 _ The results of the above experiments show quite conclusively that the velocity with which casein is hydrolyzed by trypsin is directly proportional to the concentration of the ferment. They also indicate a general proportionality between the velocity of hydrolysis and the concentration of the substrate. The slight variations that occur in the constants in the different experi- ments are due to slight differences in temperature and to small variations in the intensity of the ferment. The last experiment brings out the notable fact that the constants decrease as the concentration of the substrate increases indicating a slight tendency to depart from the rule of direct proportionality. This fall in the velocity constant however is not so great in my experiments as was observed by Taylor™ in the tryptic digestion of protamine. In fact it was so slight that it escaped unobserved in the previous experiments. It is a fact to be noted that by increasing the concentration of the ferment the turbidity of the casein solutions is increased. This fact is striking par- ticularly in the case of strong solutions, from 0.5 to 4 per cent. The turbidity of these solutions gradually increased by the addi- tion of increasing quantities of ferment so that the solution containing the greatest amount of trypsin (10 cc.) resembled a solution of calcium caseinate of the same concentration which is, normally, very much more turbid than solutions of sodium caseinate. (d) The relation between the nature of the base combined with a protein and the velocity with which it is hydrolyzed. There has been little said concerning the relation between the nature of the base combined with a protein and the velocity 3 Taylor: On Fermentation, Univ. Calif. Pub. Pathol., i, p. 236, 1907. 296 Action of Trypsin with which it is hydrolyzed. Robertson (loc. cit.) made some experiments using calcium and barium caseinates assubstrates and found that the velocity constant for the hydrolysis of barium case- inate is only about two-thirds as great as for calcium caseinate. These experiments, however, were not carried out simultaneously and this difference may possibly be due to differences in the digestive power of the ferments used. Also, there may possibly have been a slight difference in the degree of alkalinity of the two solutions employed as the solutions were made by shaking up solutions of joo alkali with casein. A later investigation revealed the fact that Ca(OH). dissolves casein more rapidly than Ba(OH).:,% and also that casein combines with bases in equivalent molecular proportions.” From a consideration of the réle of inorganic substances in nutrition as well as for theoretical reasons, it appeared of some import to carry out some experiments with the various caseinates of the alkalies and alkaline earths. Experiments were made with 0.4 and 2 per cent solutions of basic Li, Na, K, NHs, Ca, Sr, and Ba caseinates. Solutions were made in the usual way so that the proportion of base to casein = 80 X 10 equivalents per gram. The experimental procedure was precisely the same as in those described above and the experiments were done in du- plicate and simultaneously for each substrate concentration so that the acting mass and intensity of the ferment would be the same in each series. In series A of 0.4 per cent solutions of casein, 1 cc. of a 0.5 per cent solution of trypsin and 0.2 cc. of toluol were added to 100 cc. of the hydrolyte and the digestion carried out at 37° C. = 0.5° for three hours. To each digest of 100 ce. of the 2 per cent solutions in series B were added 1 ce. of a 3 per cent solution of trypsin and 0.2 cc. toluol and the diges- tion allowed to continue at 38°C. = 0.5° for three hours. The following results were obtained: : 74T. Brailsford Robertson: Journ. of Physical Chem., xiv, p. 377, 1910. 7° T. Brailsford Robertson: Ibid, xv, p. 179, 1911. E. H. Walters 297 SERIES A, 0.4 PER CENT SOLUTIONS. (Temperature 37° + 0.5°.) TABLE iX. ~ 0.4 per cent basic lithium caseinate. HOURS uraes Hietenl aa sd 0 377 1 232 0.21085 lee Om 2, 125 0.47943 2A Om 3 73 I 0.71302 24 X 107 a ee ee ee TABLE X. 0.4 per cent basic sodium caseinate. a-—z Gye K HOURS MILLIGRAMS LOGI ar 0 367 1 225 0.21249 Pal << LOE 2 133 0.44082 22 Ome 3 68 0.73216 24 <10m2 0.67325 = NS TABLE XII. » a-—-z MILLIGRAMS 0.4 per cent basic ammonium caseinate 376 230 LOGio = K a-—z 0.21346 PANES << UNE 128 0.46798 230 10m 69 | 0.73684 24.5 X 107? 298 Action of Trypsin TABLE XIII. 0.4 per cent basic calcium caseinate. — a Us z = ] LOGI9 K a-— MILLIGRAMS 0.20412 0.40353 0.70769 20'-" Xo 10s 20... KOR 23.5 K 10°? TABLE XIV. 0.4 per cent basic strontium caseinate. t a—T = HOURS MILLIGRAM PS) POGIS K 351 | 222 0.19896 20)... 2% 10m 130 0.43137 21.0%, 10pe 67 0.71924 24 xX 107? Wn re Oo TABLE XV. 0.4 per cent basic barium ariumecaseinate. t cs HOURS MILLIGRAMS BOS kK 0 | 357 1 | 222 0220632, |", 20.5 10s 2 144 0.39431 20 210m 3 75 0.67761 22.5 X 10°? SERIES B, 2 PER CENT SOLUTION. (Temperature 38° + 0.5°) TABLE XVI. 2 per cent basic lithium caseinate. = a ims siceroeee UAL ier 0.26350 0.48535 0.75197 26 X 10°? 24 X 10°? 25 X 107? wn = © E. H. Walters 299 TABLE XVil. 2 per cent basic sodium caseinate. aaa fae ate aa SEIT) = = K 0.26649 29.61 1052 563 0.47454 24. —X<-1072 0.75670 25 xX 10°? TABLE XVIII. 2 per cent basic potassium caseinate. = a t ome | LOG10 ——— MILLIGRAMS a— z 0 1 j Di De AO 2 509 25.5 X 10? 3 289 20) 1052 t TABLE XIX. 2 BP per centbasic ammonit cent basic ammonium caseinate. caseinate. a@a~-z ks a K MILLIGRAMS ae a-—-7Zz 0.27271 yf (ae a ae 0.48824 24.5 X 107 0.74787 25° 10-4 TABLE XX. 2 per cent basic calcium caseinate. marion ares scien BENS) K 0 1728 1 990 0.27190 2456 1057 2 660 0.41800 21 < 10; 3 358 0.68366 230 105- THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 300 Action of Trypsin TABLE XXI. 2 per cent basic strontium caseinate. E : al mio 4 — be air MILLIGRAMS La K a) ee ee : b a ee 0 1670 | 1 950 0.24500 24.5 X 10~ 2 600 | 0.44457 | 22 X10 3 340 0.69124 Zo ye sliaee TABLE XXII. 2 per cent basic barium caseinate. | | ae coi eieee ie heh Si 4 = | K 0 1701 | 1 980 0.23947 24 X 10°? 2 650 0.41779 21 X 10°? 3 350 0.68663 23 X 10~? scene OES 4 4 From an examination of the above tables it will be noticed that the constants are quite concordant, especially in the weaker solutions. For high substrate concentrations, however, there is a slight tendency downwards in the case of the salts of the alkaline earths which are hydrolyzed with a slightly lower ve- locity than the corresponding salts of the alkalies. It is concluded that the nature of the base combined with a protein has little or no influence in the process of hydrolysis. Ii is also interesting to note in this connection that the ve- locity of hydrolysis is uninfluenced by the extent to which the “‘basic’’ caseinates of the alkalies and alkaline earths are dis- sociated into their respective ions. Robertson” has calculated the dissociation-constants of the ‘“‘basic’’ caseinates of the alkalies and alkaline earths on the assumption that the “basic” caseinates (lissociate into two protein ions each possessed of twice as many valencies as there are molecules of base bound up in the molecule. of the caseinate. It was found that at 0.01 N concentration (0.005 N concentration of the neutralized alkali or 0.01 N con- centration of the neutralized alkaline earth) the caseinates of the alkalies are almost completely dissociated while the casein- ates of the alkaline earths are only about 50 per cent dissociated. 76 T. Brailsford Robertson: Jbid, xiv, p. 60, 1910. E. H. Walters 301 III. DISCUSSION AND RESUME OF RESULTS. The mode of action of trypsin in the above experiments cor- responds to case II outlined by Armstrong”’ in explaining the action of sucroclastic enzymes. This is a case in which the con- centration of the ferment is relatively large and practically unaffected by the products of change. The active mass is a function of the total mass from the beginning of the experiment and the change is expressed as a logarithmic function of the time. Now as Taylor” has pointed out, the fermentation of a pro- tein belongs to a class of mediated catalysis in which the trans- formation proceeds in many stages according to the following: Protein + water — product; + product; : Product; + water — po + po: etc. and finally Pn + H2O = end product + end product, each stage never being completed en bloc. As we have seen the process of hydrolysis obeys the monomolecular formula, which demands as a first postulate that the reacting substance exists at each moment in the form of unchanged substrate or end product. We must conclude, therefore, that we only follow the course of the first reaction and whatever the velocities or modes of transformation of the intermediary reactions are they do not perceptibly alter the course of the reaction measured. The well known monomolecular formula does not anticipate a reversion of the reaction and according to the foregoing experi- ments there is no indication of such. The reaction proceeds according to the law to the point of equilibrium which appears to be near the point of complete hydrolysis. The synthesis of proteins by a reversion of the above reaction through the action of enzymes is now a matter of fact. Taylor”? was able to effect the synthesis of protamine through the action of trypsin and Robertson®® has synthesized paranuclein through 7 Armstrong: Proc. of the Royal Soc. of London, \xxiii, p. 500, 1904. 78 Taylor: On Fermentation, Univ. Calif. Pub. Pathol., i, p. 125, 1907. 7 Taylor: On the Synthesis of Protein through the action of Trypsin, Univ. Calif. Pub. Pathol., i, p. 343, 1907: This Journal, iii, p. 87, 1907. 80 T. Brailsford Robertson: This Journal, v, p. 493, 1908; Ibid, iii, p. 95, 1907; Univ. Calif. Pub. Physiol., iii, p. 59, 1907. 302 Action of Trypsin the agency of pepsin. In these experiments the protein synthesis was brought about by allowing large quantities of the respective ferments to act upon concentrated solutions of the products of the complete digestion of the respective proteins. These exper- iments suffice to show the action of ferments in accelerating the reverse of the reaction alluded to above. Now if there is no indication of reversion in the process of hydrolysis how is this phenomenon to be explained? For the interpretation of this Robertson®! has advanced an hypothesis of Reciprocal Catalysis which appears to be a rational explanation of the above facts. This hypothesis assumes a combination between the enzyme and protein,® the enzyme being assumed to carry water into the protein molecule and parting with the water to recoup itself from the medium, while the protein molecule subsequently splits up into the products of its hydrolysis, and the enzyme-product finally reacting with the water regenerating the ferment. Consid- ering both the protein and enzyme to be amphoteric electrolytes, which is a reasonable conclusion from experimental data, the various steps in the process of hydrolysis are represented by the following schematic equations: HF — COH-N — + HFFOH = —COH-N................. (1) | HOF Protein + Ferment = Protein-ferment compound. HF —COH-N = —COOH + HoaN— +FF..............(1ID FOH Protein-ferment compound = Product + Product + Anhydrous Ferment. FEE HO = AP EOH@S 622-070 ey ne ae (IIT) Anhydrous Ferment + Water = Hydrated Ferment. 81 T. Brailsford Robertson: loc. cit.; Die physikalische Chemie der Pro- teine, Dresden, Verlag von Theodor Steinkopff, p. 404. 82 Biological chemists have pretty generally accepted this to be a fact, see literature quoted elsewhere in this paper under Vernon, Bayliss, Robert- son, Henri, and Hedin. E. H. Walters 303 while the synthesis is the reverse of these reactions; thus assum- ing the existence of two varieties of the same ferment, one accelerating hydrolysis, and the other accelerating synthesis, each operating under very different conditions. The existence of two such forms of an enzyme has been observed by Euler.*® It is supposed that during the process of hydrolysis the station of equilibrium in reaction III is far to the right and is reached with great velocity compared with that of either of the reactions I and I] measured in the direction from right to left and at any moment the concentration of the anhydrous (synthesis-acceler- ating) form of the ferment FF is negligible compared with that of the hydrated (hydrolysis-accelerating) form HFFOH. Under these conditions then it is obvious that the velocity of hydrolysis would be directly proportional to the concentration of the fer- ment, which is experimentally the case. And similarly the mono- molecular formula would hold good if reaction I proceeded at a very great velocity compared with reaction II, which is a logical deduction from experimental and theoretical evidence. It has been pointed out that the velocity constant gradually decreases as the concentration of the substrate increases. The above hypothesis also offers a reasonable explanation of this fact. According to the above scheme for the process of hydroly- sis the enzyme must cause a greater or smaller shifting in the station of equilibrium between the protein and its products. At high ferment concentrations the enzyme accelerates hydrolysis of the protein more than its synthesis because the hydrolysis- accelerating form of the enzyme is initially present in consider- able excess of the synthesis-accelerating form; but at high con- centrations of substrate the protein accelerates the dehydration of the erzyme more than its hydration because it is initially present in great excess of its products of bydrolysis. This latter condition will, therefore, cause a slowing of the velocity of hydroly- sis which is experimentally found to occur. The question of synthesis is quite a different problem and one that does not concern us here. The reaction of synthesis being a reaction of the second or a higher order a decided shift in the 83 Kuler: Zeitschr. f. physiol. Chem., lii, p. 146, 1907. 84 See T. Brailsford Robertson: Die physikalische Chemie der Proteine, p. 408, Dresden, 1912, Verlag von Theodor Steinkopff. 304 Action of Trypsin station of equilibrium between the Products — Protein in the direction of synthesis by the addition of large quantities of en- zyme, and also shifting the position of equilibrium between the hydrated and the anhydrous form of the enzyme in the direction of the latter by the presence of large amounts of the products of hydrolysis or by increasing the temperature, must be brought about before the reaction will proceed in the direction of synthe- sis. The foregoing experiments may be briefly summarized as fol- lows: 1. The method of estimating the velocity with which a protein (casein) is hydrolyzed by determining the nitrogen in the undi- gested portion after precipitation with acetic acid yields results admitting of an accurate physico-chemical interpretation. 2. Upon the addition of a slight excess of alkali to neutral or faintly alkaline solutions of casein immediately before precipi- tation with acetic acid, precipitation is hastened and a clear fil- trate is assured. 3. The relation between the time of hydrolysis and the amount of “‘basic” sodium caseinate hydrolyzed, is, for all stages of the reaction, what would be expected from the monomolecular = = KE a—z 4. The velocity with which “basic” sodium caseinate is hydro- lyzed by trypsin is directly proportional to the concentration of the ferment. 5. There is a general proportionality between the concentration of the substrate and the velocity of hydrolysis, although the velocity constant decreases slightly as the concentration of the substrate increases. 6. The nature of the base combined with casein has little or no influence in the process of hydrolysis. ‘Basic’ caseinates of Li, Na, K, NH; Ca, Sr, and Ba hydrolyze with approximately equal velocities for all concentrations of substrate between 0.4 and 2 per cent. 7. There is no relation between the degree of dissociation and the rate with which “‘basic”’ caseinates are hydrolyzed by tryp- sin, as in the solutions employed the caseinates of the alkalies are almost completely dissociated while the caseinates of the alka- formula Logio E. H. Walters 305 line earths are on!y about 50 per cent dissociated, yet both series of ‘salts’ are hydrolyzed by trypsin at approximately the same velocity. 8. There is evidence of rapid autohydrolysis in solutions of “neutral” and “basic” caseinates of the alkalies: and alkaline earths. Finally, I wish to express my appreciation of the valuable advice and incessant interest of Dr. T. Brailsford Robertson at whose suggestion this work was undertaken and with whose aid it has been accomplished. F Arar fi 7 7 j an 4 a = is ‘ - . a j — wee re F A Peay 4 Qe ~ ! . a tue vielrret ? or cerns & Oy y' ei wee oe ie | ate ee oe eet | a re “3 Ar’ Sea = +e oO Scar tte ‘ »¢@ - ah ie vr Leta, SoS c ~ --°2 Seen "art be nem u n), (£8 Af leas Son {hee ame oie. ae tee wt a. ‘up oer: ‘9 er ). a TEP Pea. Tay resin ce ee 439 Bae Hoe ie i er os eat “Ei i} iy vil sie ‘ Le *: bidet Sar inves sa eet ite vi &padds Sea ul ah sri. WF 7 = Pay as } niarpracse re - t ci tan mpilaad iy elo pil cians eet ee Ree P ; iu a} ics xs Brito «4 ag hey es { vr a a e° st ee ie al ane Y i * tae Ane‘ & y=! Toa ewe Y ‘yp os hi yen. % gers) Ton a i TAY Vw Lol" ax ari ptr? 4b-t24 4D8 Lenehan ~— < ‘; ‘4 a < —- a i Yt ipwrit a a ieee g OC eb Ton ee ree appease (amst ates Oe ae te chee the's weal ulti ls 2:who- th ON THE REFRACTIVE INDICES OF SOLUTIONS OF CERTAIN PROTEINS: VII. SALMINE. By T. BRAILSFORD ROBERTSON. (From the Rudolph Spreckels Physiological Laboratory of the University of California.) (Received for publication, March 2, 1912.) A. THE PREPARATION OF SALMINE SULPHATE. A member of the protamine group, namely salmine! was prepared by the method of Kossel? as follows: The ripe testicles of the Pacific salmon were minced and the macerated mass which was thus obtained was shaken up in tall glass cylinders with five or six times its volume of distilled water. The thick suspension of sperm which was thus obtained was syphoned off from the subnatant con- nective tissue and curdled by the addition of 80 cc. per liter of 7 acetic acid. The curdled mass of sperm was then washed in ten times its volume of 95 per cent alcohol and this washing was repeated twice; it was then washed once in the same volume of absolute alcohol and then in the same volume of ether. The powder, wet with ether, which was thus obtained, was spread out upon filter paper to dry in the air in a warm dry place. Each 15 grams of the dried sperm was then stirred up in 350 cc. of 1 per cent by volume H,SO,, for about six hours. This mixture was then filtered through hardened filter paper and the filtrate obtained from the extraction of 15 grams of sperm was placed in a tall glass cylinder of about 4000 cc. capacity which was then filled with absolute alcohol. After allowing the precipitate to settle, the supernatant fluid was syphoned off, the precip:tate contained in two cylinders was collected in one and this was filled with alco- hol again.* The entire precipitate, suspended in alcohol, from the extract 1 According to Taylor (Univ. of Calif. Publ. Pathol.,i, p. 7, 1904) the pro- tamine which is contained in the sperm of the Pacific salmon is identical with the salmine found in the sperm of the European salmon. 2 A. Kossel: Zeitschr. f. physiol. Chem., xxv, p. 165, 1898. 3 It is necessary to avoid washing with alcohol too frequently, as on sus- pending the protamine sulphate in alcohol for a third or fourth time a very stable suspension is produced from which protamine is only deposited very slowly. 3°97 308 Refractive Indices of Salmine Solutions of 300 grams of sperm, was dissolved by the addition of about 4 liters of hot water (about 80° C.), the least soluble portion was filtered off, and the re- mainder reprecipitated by the addition of 10 volumes of alcohol. This precipitate was washed once in the same volume of alcohol as that employed in precipitation and then in a like volumeof ether. The finalsuspension of protamine sulphate in ether, obtained afte: syphoning off the supernatant ether, was collected in a hardened filter, dried over sulphuric acid at-40° for two days and then pulverized and sifted. The product is a friable white powder. The yield from 300 grams of sperm was 14.6 grams. The empirical formula of the substance which is thus obtained is according to Kossel (loc. cit.) and Taylor (loc. cit.), C3oHs7Ni7O¢. 2H.SO,. It readily dissolves in water up to about 2 per cent at 20°. It diffuses through parchment paper (Taylor). Its solu- tions are very faintly acid in reaction. According to Taylor the acidity of a 0.5 per cent solution, measured by the gas-chain, is 300 H+; I found that one gram of my preparation in 0.25 per cent solution required the addition of .9.6 cc. of # KOH to render the solution just alkaline to rosolic acid, corresponding, in 0.5 per cent solution to an acidity of less than s¢y; since an acidity determined by titration in protein solutions must obviously be considerably in excess of the true acidity, it may be inferred that the acidities of solutions of my preparation were not appreciably in excess of Tay- lor’s estimate cited above. Since a 1 per cent solution of protamine sulphate contains 0.0424 equivalents of H,SO, per liter, it is evident that protamine sulphate does not, in aqueous solution, undergo hydrolytic dissociation to any very appreciable extent. Accord- ing to Taylor, a perfectly neutral preparation of protamine sulphate may be obtained by a special: and lengthy process of preparation and purification. KB. THE PREPARATION OF SALMINE CHLORIDE. Several attempts were made to prepare salmine carbonate accord- ing to the method recommended by Taylor (loc. cit.), in order to prepare the chloride from this substance. Many difficulties were found to attend this procedure, however. If excess of Ba(OH). be added to a dilute solution of protamine sulphate great difficulty is encountered in removing this excess by means of CO, even at 50° C. After several hours’ passage of CO, clear filtrates can be ob- tained which contain barium, a fact which is probably attribu- a2 T. Brailsford Robertson 309 table to the formation of the barium salt of a carbamino derivative of the protamine.‘ Moreover, as Taylor points out, great difficulty is experienced in obtaining clear filtrates; indeed I have found the only successful method to consist in filtration under pressure through a Chamber- land filter, a process which is attended by considerable loss of the protamine, since it is, to some extent, retained by the filter.5 Accordingly, salmine chloride was prepared directly from the sulphate in the following manner: To a carefully weighed amount (1.48 grams) of protamine sulphate dis- solved in 100 cc. of water was added an exactly sufficient weight of carefully chosen barium chloride crystals, dissolved in about 20 cc. of water, to pre- cipitate the H,SO, in the protamine sulphate. This mixture was then set aside in a tall glass cylinder at 50° for twenty-four hours at the end of which time a compact precipitate of barium sulphate had settled to the bottom of the cylinder from which the clear supernatant fluid could readily be de- canted. This fluid was filtered through a hardened filter and the protamine chloride precipitated by the addition of 5 to 6 volumes of absolute alcohol. After allowing the precipitate to settle the supernatant fluid was syphoned off and the precipitate was washed in 1 liter of absolute alcohol and twice in 1 liter of ether (iiber Natrium destilliert) and was finally collected on a hardened filter and dried over H.SO, at 36° for twenty-four hours. It was then pulverized and sifted and dried for another twenty-four hours. The yield was only about a third of a gram, which is attributable to the fact that the precipitate, after washing in alcohol only settled very incompletely, a phenomenon which appears to be characteristic of very anhydrous (or, as Taylor believes, very highly purified) preparations of salmine. The empirical formula of this substance is, according to Kossel Cszo Hs7Ni70¢.4HCl. It dissolves readily in water, yielding very faintly acid solutions. C. THE DETERMINATIONS OF REFRACTIVITY. Portions of a 2 per cent solution of salmine sulphate were diluted to 1.5 per cent and to 1 per cent and the refractive indices of these solutions and of water were measured at 22° C. in a Pulfrich refrac- tometer, using a sodium flame as the source of light. On another 4M. Siegfried: Ergeb. d. Physiol., ix, p. 334, 1910. 5 This is true also of protamine chloride. A1 per cent solution of pro- tamine chloride, after filtration through a porcelain filter under pressure, was found to be reduced in concentration to about 0.5 per cent. 310 Refractive Indices of Salmine Solutions occasion a 1 per cent solution was diluted to 0.5 per cent and the refractive indices of these solutions were measured at the same tem- perature. The following were the results obtained: The values headed a are calculated from the formula n — n; =‘a X c where n is the refractive index of the solution, , that of the solvent, and c is the percentage of salmine sulphate or of salmine in the solution. TABLE 1. | = CTIVE! In REFRACTIVE) 7 For SALMINE SUL- SOLUTION pak PHATE a FOR SALMINE Distilled water......... 1.33410 H.SO, = 1 per cent salmine sulphate= 42.4cec. 7; H2SO; per HOOVEEe eee e.g... | 1.33450 0.5 per cent salmine sul- phate.. ae .| 1.33497 1.0 per ent erieine She phates s.2-2- =. .: | (1) 1.33584 1.0 per cent salmine eal phate.. ree .| (2) 1.33584 | > 0.00174+0.00007| + 0.00173 +0 .00009 1.5 per ccat mulirine wale | phate.. MIPS 5h 1.33671 2.0 per aaa: palnnizie sul- phate.cn-qaesne nes 1.33759 The calculation of a for salmine sulphate is performed by adding together all observed values of n — m; and dividing this sum by the sum of the percentages of salmine sulphate employed. That for salmine is calculated upon the assumption that 1 gram of salmine sulphate contains 0.792 grams of salmine and that when the refrac- tivity of the sulphuric acid in the compound be subtracted from the observed values of n — 7, the remainders represent the re- fractivity of salmine. The details of these calculations follow: T. Brailsford Robertson 311 TABLE 2. POSSIBLE ERROR IN DETERMINATION OF n—n) PERCENTAGE OF SALMINE SULPHATE PERCENTAGE OF SALMINE m—m71 FOR SAL- |MINE SULPHATE n—n, FOR SALMINE 2.0 0.00349 0.00008 16 1.188 | 0.00261 0.00008 1.0 0.792 | 0.00174 0.00008 1.0 | 0.00174 0.00008 0.5 0.00087 It will be observed that the numbers enumerated in the third column of the above table, divided by those in the first column, yield a constant quotient. We may therefore conclude that the change in the refractive index of water which is brought about by dis- solving salmine sulphate therein is directly proportional to the concen- tration of the salmine sulphate. A 0.5 per cent solution of salmine chloride was prepared and its refractive index and that of water were determined at 18°C. The following was the result obtained: M=REFRACTIVE INDEX OF SOLUTION SOLUTION aT 18°C. a FOR SALMINE CHLORIDE Distilled water.......... 1.33356 One-half per cent chlo TAG S 1.33442 0.00172+0.00016 A1 per cent solution of salmine chloride contains 0.0446 n hydro- chloric acid and 0.837 grams of salmine. Hence in a 1 per cent solu- tion of salmine chloride the refractivity of the hydrochloric acid (calculated by interpolation from the refractivity of a 7 solution of HCl) is 0.00032. From this and the above value of a for salmine chloride it follows that the change in the refractive index of water due to the intro- duction of 0.837 per cent of salmine in the form of its chloride is: 0.00172 + 0.00016 — 0.00032 = 0.00140 = 0.00016 312 Refractive Indices of Salmine Solutions Hence the change in the refractive index of water due to the introduction of one per cent of salmine in the form of its chloride is: 0.00167 + 0.00019 Thus, within the experimental error, the refractivity of salmine in solutions of salmine chloride is identical with its refractivity in solutions of salmine sulphate. CONCLUSIONS. 1. The change in the refractive index of water which is brought about by dissolving salminesulphate therein is directly proportional to the concentration of the dissolved salmine sulphate. 2. The value of a (= change in the refractive index of the sol- vent due to the introduction of 1 per cent of the protein) forsalmine sulphate is 0.00174 + 0.00007. 3. The value of a for salmine chloride is 0.00172 + 0.00016. 4. The value of a for the base salmine when combined to form salmine sulphate is 0.00172 += 0.00009. 5. The value of a for salmine in the form of salmine chloride is, within the above experimental error, identical with its value for salmine in the form of salmine sulphate. STUDIES ON THE EFFECT OF LECITHIN UPON THE FERMENTATION OF SUGAR BY BACTERIA. By ALBERT A. EPSTEIN anp H. OLSAN. (From the Pathological Laboratory of Mt. Sinai Hospital, New York.) (Received for publication, March 5, 1912.) This work was undertaken with a view to studying the effect of lecithin upon the process of sugar fermentation in vitro. Many functions have been ascribed to lecithin, and considerable dis- cussion has arisen of late concerning its réle in metabolism. It has been assumed by a number of investigators, that lecithin exerts an inhibitory action upon the oxidative processes in the animal body. Diabetes was at one time believed to be due to the inhibitory action of lecithin on the oxidation of sugar. This hypothesis has been chiefly upheld by Liithje, who found that the sugar output in diabetics was usually increased by the administration of egg yolk, a substance rich in lecithin. Bang, as is known, suggested that lecithin and dextrose combined in the blood, forming a sub- stance called jecorin. Some have even asserted that the alleged decrease in the intrinsic or fundamental metabolism, and the decrease in the oxygen con- sumption occurring in adiposis, are attributable to the presence of a larger amount of lecithin in the body fluids than that nor- mally present. For example, Kempner and Schepilewsky! found that white mice usually increased in weight after receiving injec- tions of lecithin. Acting upon this belief, Russian and French investigators have suggested the use of lecithin therapeutically in cases of emaciation and wasting diseases. In support of their claims, the authors quote a number of experiments made upon animals, the results of which do not appear as convincing as the authors believe them to be. ‘Kempner and Schepilewsky: Zeitschr. f. Hyg., xxvii, p. 213, 1898. 313 314 Effect of Lecithin upon Fermentation Lately Yoshimoto? made a number of animal experiments with lecithin, and instead of finding a diminished output of nitrogen in the urine, he found it to be increased—the increase being the exact equivalent of the nitrogen present in the lecithin admin- istered to the animals. This investigator, therefore, came to the conclusion that lecithin does not exert an inhibitory action on metabolism. The attempt to ascertain any such function of lecithin in vivo is naturally associated with many difficulties. Even were we to find a diminution in the nitrogen output after feeding lecithin to animals, the. conclusion that lecithin produced this result by virtue of its action upon nitrogenous metabolism, would not be justifiable. Such a result might be due to an indirect action; for example, we could readily conceive of an increased mobilization of fats produced by lecithin; or possibly an increased oxidation of sugars, the result of which would be a sparing of protein material, leading consequently to a diminished output of nitrogen in the urine. This effect would manifestly be, not the result of inhib- itory action of lecithin upon nitrogenous metabolism, but the indirect result of an increased combustion of fats and sugars. The difficulties which arise in the study of a problem of this character in vivo, are chiefly due to the fact that it is almost impossible to dissociate a single function or chemical process from every other in the body. It seemed, therefore, desirable to approach the subject in a somewhat simpler way. Vallet and Rimbaud,’ as well as Ren- shaw and Atkins,‘ have recently attempted to solve the problem of the réle of lecithin in biological processes by studying its effect upon the growth of bacteria. Their results show that lecithin does not materially influence the growth of bacteria. However, the method pursued by the above investigators, when taken in conjunction with the object sought after, is not free from criti- cism. Cellular growth and cellular function should not be con- fused; the one need not be an index of the activity of the other. The protoplasm of many cell and tissue-forms is endowed with some of the functions present in the living cells. This is true of * Yoshimoto: Zeitschr. f. physiol. Chem., lxiv, p. 464, 1910. $ Vallet et Rimbaud: Compt. rend. soc. biol., Ixviii, p. 302, 1910. ‘Renshaw and Atkins: Journ. Amer. Chem. Soc , pp. xevii-xeviii, 1910. Albert A. Epstein and H. Olsan 315 many enzyme or ferment-bearing cells. The extract of the yeast cell, for example, can ferment sugar as does the living cell itself. Here we have a function that is resident in the cell material and is active even in the absence of cell life. A number of experiments have been made by Kuettner,® and also by O. Schwarz,® with lecithin and the different digestive ferments. Schwartz, for example, ascribes the inhibitory action of blood serum upon tryp- sin, to lipoids, presumably lecithin. We have therefore, deemed it necessary to take up the study of lecithin in connection with the fermentation of sugars by bac- teria. In so doing we deal with a single comparatively simple process. The character of the agents used in the tests to produce the fermentation, is also relatively simple. For our purposes we used three types of bacteria; namely, Bacil- lus coli communis, Bacillus mucosus capsulatus, and Bacillus acidi lactici. Each of these organisms ferments certain sugars. Our measure of bacterial activity, therefore, was the production of gas and acid. The amount of lecithin employed in our media at no time ex- ceeded 0.4 per cent. This, of course, is an amount of the sub- stance in excess of that found in biological fluids. If, therefore, lecithin could modify oxidative processes in respect to sugar fer- mentation, then it would manifest itself by an increase or decrease in the production of gas or acid, or both. We tested the fermen- tative action of the above bacteria on twelve different types of sugars. This was done to ascertain whether or not the chemical constitution of sugar played any rdéle in the rate and character of its decomposition by bacteria. The list of sugars includes the alcohol, aldehyde, and ketone type of the different saccharides (mono-, di-, and polysaccharide). A hexavalent alcohol and an aldehyde pentose are also represented in the series. A 1 per cent solution of each sugar in nutrient bouillon (neutral to phenophthalein) was used. The media were distributed into fermentation and straight test tubes; 10cc. of each medium being used as the unit. To one set of tubes, a 4 per cent emulsion of ’ Kuettner: Zeitschr. f. physiol. Chem., 1, p. 472, 1906-07. 6 Schwarz: Wien. klin. Wochenschr., 1909. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 316 Effect of Lecithin upon Fermentation lecithin was added, allowing 1 cc. for each tube. To the other set, sterile salt solution was added in like amount. Both sets of tubes were inoculated with a loopful of an emulsion of each bacterium and incubated at 37.5°C. All the tests were _ made duplicate. A parallel series of uninnoculated tubes were incubated and used as controls. The amount of gas produced in the fermentation tubes was recorded in cubic centimeters at the end of twenty-four and forty- eight hours incubation. The following table shows the results obtained in tests on gas production: 317 Albert A. Epstein and H. Olsan GsOIN dd GZqdAHAC1V OF |Of |O8 |O'E |O'E | °° sanog gp | UIYWT00] yyTM 03 |02@ |0'% |0'S {OT |°°*‘sanog ay UINI Poul UIB] J 0's 0'% OF 0's 0% “** *gInoy QF 0% (02 |0'F 10'S (0% |'''‘saunoy bo bedtd Suaccatal t :1O1}08] Iprow snyrowg $% /O€ | O'S | O'% | °* sanoy gp | UIyQIO9] YALA 0% /0% |o% oT |**'sinoq wn! pow ULB] 9% |0°% /o0'% | 9% |-°-‘sinoy gp o% [oz foe | oz |----simog yg ¢ TMPOU MPA :snjeynsdeo snsoonur snyjloeg Of |0% | OE | O'% |:°°*sanoy gp | UIYAT097 YIM 0% |OT |0'S | OT |°°‘‘sanog 72 { UINI pou UB] J O€ (08 | OS | O'T |°°*Sanoy gF 0% |OT [OT | OT |---sinoy ye / uINIpeul ulelg : STUNUIUIOD 1109 sny]IOVg 3s013x0q] TIOHOOTV INGIVAINL G@aI4VHOOVS |AGIHVHOOVS SAqCINVHOOVSIA SA d1UVHOOVBONOW “s4ajamyuas 919qnd Ur UoYoNpod svB Buimoyy T TGV 318 Effect of Lecithin upon Fermentation SUMMARY. Bacillus coli communis. INCREASED GAS DECREASED GAS Dextrose Levulose Galactose Maltose Lactose Inulin Saccharose Dextrin Glycerine Mannit Arabinose Bacillus mucosus capsulatus. Galactose Dextrose Maltose Levulose Lactose Saccharose Mannit Arabinose Raffinose Inulin Dextrin Glycerine Bacillus acidi lactict. Maltose Galactose Dextrose Saccharose Lactose Inulin Dextrin Levulose Glycerine Arabinose Raffinose In the above table we note that with Bacillus coli communis lecithin favors an increase in gas formation in the monosaccharides dextrose and galactose, and the disaccharides lactose and sac- charose; while it inhibits gas formation in the trisaccharide raffi- nose, the remaining sugars are unaffected. Lecithin aids gas production by Baciilus mucosus capsulatus in the monosaccharides galactose and levulose, while it arrests gas production with dextrose. Lactose and arabinose are influ- enced in like manner. With Bacillus acidi lactici lecithin also aids the formation .of gas in the disaccharides maltose and saccharose, while it checks gas fermentation in all the monosaccharides, excepting dextrose, in the dissacharide lactose, and in raffinose and in mannit. The remaining media are unaffected. 319 Albert A. Epstein and H. Olsan esoulquly aASOINGd aqdsHaUaTY AN ~~ qjuUueyy ING -VAVXSH IOHOD1V ai aife \e'@) 6 14n9).6 eye) 6.8 peonpoid prov JON ait ae qoR] plow “g “pour ULB] Whee o.14) 9) ua sie ce «8 peonpoid prow JON Be eo sdeo ‘onul "g “pour ule [ Pe ee poeonpoid prov JON ‘1JOO “¢ wunipeut UIBId cece eeenee ]019u09- WIN pour UlBe[_d il Q 5 2 Fs 5 B 3 ae ee B 8 IOHOOTY gaily HOOVea adluvHoov INATV AIL -410d “TL SACINVHOOVSIA ‘HOBN 2f Jo srazouNIyU99 dIqNd UT ‘pipau svbns juasaf{ip ay) U2 uoyonpo.d prop buinoyy (8) 2 ATAVE Effect of Lecithin upon Fermentation 320 ye] prow “g -4]ND ‘peonpoid prov yan a 4 0 6 e)0 level * aerate sdvo ‘onu ‘g 4ing ee) Jer ies Hee Pe ‘peonpoid pioe 9N “***** STuNUTUIOD 100 *g aIn4[ND ae ce [013U09 UNI peur UIYyID07 ; yo ~— : GL cr 9 ceo | 8 PF 8°6 LT | D Peak Peta ios ce'9 | O82 | SEZ See eam a) 2 0L°9 | €°¢ 8'6 OCI |; SOL ¢ SicZn | arGule bao 68 | S010 SOs) POs Gb.0 oo=|) 270 69| GG) Ss F Ler 6°0 c6°0 \¢8’0 ¢°0 9% Se°s| SEs | 179 UGS | ¢9'9 6 0- : “I 6'0 PF | S| OF| SLs T'¢ cc’ TO al (a | €°9 cg8'Z| 8¢| GF 0@ ¢6 0 €°0 a) £°0 Cheol Shel Oe toZien ee bee Piceane se |e le be Piel eel alg 3 : 5B 3 Pee eae Sie corauny aah tata ae ' eee SaaravHoovera SUdIUY HOOVSONON -VAVXGH u1y}199] 7ua0 4ad 7°09 fo aauasaid ay) Ur DipaU wwbns ay) UO UOYINposd prop Bbuimoyy “(9)% ATaVL Albert A. Epstein and H. Olsan 321 As the above tables indicate, the results of acid production are more uniform than those obtained with gas production. Bacillus coli communis causes an increased acid production in the presence of lecithin with all the sugars, excepting dextrose and raffinose. With the latter two sugars this bacillus produces less acid in the presence of lecithin than otherwise. Bacillus mucosus capsulatus produces more acid in the presence of lecithin than otherwise, with all the sugars excepting glycerine and raffinose; upon the latter lecithin has noeffect. Bacillus acidi lactici in the presence of lecithin produces more acid in all the sugar media. It is significant that in all the tests for acidity, lecithin has a distinct tendency to increase rather than to decrease the acid production; and if we take acid-production as an index of oxida- tive processes, then we must conclude that lecithin aids oxidation of the sugars. Lecithin itself contains certain radicles which on decomposition yield acid and gas. Its one constituent, cholin, can (as Hase- broek’? has shown) on anaérobic putrefaction be split into ¢arbon dioxide, methane, ammonia and methylamine. Although the con- ditions existing in our tests and those which occur in anaérobic putrefaction are totally different, nevertheless the question might be asked, whether or not, a decomposition of lecithin in the presence of the bacteria takes place, whieh may be responsible, at least in part, for the results recorded. Tests were therefore instituted to answer this particular question. Cultures were made of the organ-. isms in sugar-free bouillon, with and without lecithin; and, on comparing the results obtained, it was found that the bacteria employed do not cause any acid or gas production from lecithin, even after seventy-two hours’ incubation. It is necessary to call attention to the fact that in all our tests, lecithin is presumably present in a free state; and although we have reason to believe from the work done by one of the authors (E.) in another connection, that lecithin enters into combination with peptone bodies, such as are present in our culture media, we must for the present infer that the lecithin present is in a free state; and the conclusions to be drawn must apply to the action of leci- thin present in this state. ™Hasebroeck: Zeitschr. f. physiol. Chem., xii, p. 148, 1888. 322 Effect of Lecithin upon Fermentation To summarize briefly, our conclusions, therefore, are: (1) Free lecithin may modify the bacterial fermentation of different sugars; hence, oxidative processes. (2) The action of lecithin increases the fermentation of some sugars and lessens that of others. There is apparently no definite relationship between the action of leci- thin upon sugars and their chemical composition. To sum up: the tendency of lecithin is to increase rather than to decrease fermentation. THE BALANCE OF ACID-FORMING AND BASE-FORMING ELEMENTS IN FOODS, AND ITS RELATION TO AMMONIA METABOLISM. By H. C. SHERMAN anp A. O. GETTLER. (Contribution from the Havemeyer Laboratories of Columbia University, ' No. 205.) (Received for publication, March 8, 1912.) In recent years the ash constituents of foods have come to hold an increasingly prominent place in considerations of food ‘values. It is now generally recognized not only that the food as a whole should supply adequate amounts of each of the chem- ical elements which is essential to the body structure, but also that these elements should stand in normal quantitative relations to each other. ; Conspicuous among the quantitative relations is that between the acid-forming and the base-forming elements of the food. It has long been known that certain foods contain a surplus of base- forming over acid-forming elements as evidenced by the fact that on burning they yield a strongly alkaline ash, whereas other foods lose acid-forming elements in ashing and yet yield a neutral ash showing that acid-forming elements must have predominated in the food. It was, however, not possible to make any useful quantitative comparisons on the basis of the data which had been obtained by the usual methods and recorded in the accepted tables of ash analyses, because these data represented only the compcsition of the material which remained after ignition, regardless of the fact that in many cases a large part of the acid-forming elements exist in the food as constituents of the organic matter and pass off during the ignition. This is particularly true of sulphur, which, so far as known, exists in foods chiefly as a constituent of protein and often is expelled almost entirely during the burning, 323 324 eanutsieaeeeeern «0. Peas idried*r....-:.- Peas idrieditzc.cn.c-.:.- - Potataes ep... ss: Potatoes eseeee..- -.:. Potatoes eee... Raisinsia-cere see Raspberry juicef....... Rice, eas cee -...| ; Purmipse seer... 2.68 6.86 | TPurnips ene « @ 6.80tT 9.41 Wheat, entire*......... 9.66 Wheat, entiref......... 12.39 Wheat, flour*2235..-2:.. | *Data determined in this laboratory. + Data partly compiled, partly determined in this laboratory. 1 Data published by Katz. ** Data insufficient to permit calculation of acid to calorie basis. tt Possible loss of sulphur compounds in drying previous to analysis. H. C. Sherman and A. O. Gettler 331 calorie basis or on the basis of dry matter. Milk shows a slight predominance of bases. In vegetables and fruits the predomi- nance of bases is usually much greater, a 100 calorie portion of potato for example furnishing enough bases to almost exactly neutralize the excess of acids from a 100 calorie portion of lean beef. The few nuts so far examined yield different results, the peanuts showing an excess of acid-forming elements while the base-forming elements predominate in almonds and chestnuts. It will be of interest to study other edible nuts and todetermine whether the partial or complete substitution of nuts for meat produces a marked effect’ upon the balance of acid-forming and base-forming elements in the diet as a whole. METABOLISM EXPERIMENTS. In addition to the determination of the balance of acid-forming and base-forming elements in a variety of foods our investiga- tion was planned to include a study of the extent to which the acid arising from oxidation of an “‘acid-forming” food is neu- tralized by ammonia when such a food is metabolized in the hu- man body. It was desired to study this point upon a healthy man with ordinary articles of food avoiding any extremes of diet or any unusual condition which might interfere with the normal working of the neutralizing mechanism. First ExpERIMENT:—A healthy man (A. O. G.) twenty-seven years old, 5 feet 7 inches (1.70 meters) high, weighing 142 pounds (64.5 kilograms) took for ten days (November 27 to December 7, 1910) a diet which was uniform throughout except that during the first four and the last two days it contained 340 grams of potato while from the fifth to the eighth day inclusive the potato - was replaced by amount of rice (80 grams weighed dry) sufficient to furnish approximately the same energy value (about 300 cal- ories). Since 340 grams of the potatoes here used furnished an excess of base-forming over acid-forming elements equivalent to 15 cc. of normal base while 80 grams of the rice contained an excess of acid-forming elements equivalent to 6.7 cc. of normal acid, the change in diet corresponded to the production in the body of 21.7 cc. of normal acid per day. Beginning with the third day of the experiment the urine was collected in 24 hour THE JOURNAL OF BIOLOGICAL CHEMISTRY. VOL. XI, NO. 4 332 These washings were carried out inside an incubator over sulphuric acid in order to avoid the deposition of atmospheric moisture upon the filter. T. Brailsford Robertson 343 almost milky. This opalescence continued to be very marked down to a dilution of one-sixty-fourth. The eggs of onefemale Strongylocentrotus purpuratus were divided into three portions. One portion was not sensitized at all, another was sensitized by four minutes’ immersion in ¥ SrCl:, and a third by 4 minutes’ immersion in } CaCl. Two drops of thick egg- suspension were added to 2 cc. of each of the dilutions of the solu- tion just described. The following were the results: Eggs Sensitized with SrClz. Dilution of the solution of the fertilizing agent Effect of immersing the eggs in this solution. LL a | ere Immediate agglutination. In fifteen minutes irregular crinkled and col- lapsed membranes surround the ma- jority of the eggs. lpartin 400...........) Immediate agglutination. A dense Veware wn S00... ........ | precipitate surrounds the aggluti- DP pari 1,600........... ( nated masses of the eggs so that the Bparhi. 3,200) 62... 25s J individual eggs cannot be discerned. 1 part in 6,400............. Immediate agglutination. A dense precipitate surrounds each egg. After two hours in all those cases in which the egg itself can be observed there is a distinct membrane. [Sets SN a Agglutinatedinoneminute. Blisters onseveral of theeggsinfourminutes. No change after seventy-five min- utes. DE BI OO. oe oice sec soins Slight agglutination in oneminute. No further change in seventy-five minutes. [iii cay Migs Uv. | | a No agglutination. From this dilu- tion down to that of 1 part in 409,- 600 the substance had no action upon the eggs. 6 This agglutination is not to be confused with the phenomenon of ‘‘sticki- ness’’ which is exhibited by all eggs which have been treated with strontium chloride. When eggs which have been treated with strontium chloride are dropped into sea water they soon sink to the bottom of the vessel and adhere to it in a thin layer; even if shaken up before they sink, they do not adhere in clumps, at most one or two sticking together very loosely so that they can readily be shaken apart again. Eggs which ‘‘agglutinate,’’ in the sense in which the word is used above, ‘‘clot’’ or form large clumps resembling coag- ula almost the instant they are dropped into the agglutinating mixture. 344 lipartiin:200...2 28: 1 partum 4002... 1 =art in 800...... 1 part in 1,600.... 1 part in 200...... 1 part in 400...... I part. m'iSO0r... 1 part in 1,600.... Lpart inj332005. 2: 5.45.0 Odcytase Unsensitized Eggs. nS cist oe Irregular membranes and _ blisters and agglutination within one min- ute. In four minutes 10 per cent of the eggs are cytolyzed. In five minutes crinkled and collapsed membranes upon all of theeggs. In thirty minutes 50 per cent cyto- lyzed or converted into “‘shadows.”’ x. See Irregular blisters on 20 per cent of the eggs in five minutes. Agglutina- tion occurred at once. Irregular crinkled membranes in about 50 per cent in fifteen minutes. fet PISS 7, Agglutination occurred at once. In fifteen minutes 100 per cent have perfect spherical membranes. ce eee eee No agglutination. From this dilu- tion down to that of 1 part in 409,600 the substance had no action upon the eggs. Eggs Sensitized with CaCl. re See Blisters and irregular crinkled and collapsed membranes and _ pro- nounced agglutination within one minute. In fifteen minutes 10 per cent cytolyzed. In twohours 80 per cent cytolyzed. Be tek 3 2 Pronounced agglutination in one min- ute. Intwo hours blisters upon all of the eggs but no complete mem- branes. Rete eed Agglutinated in one minute. In five minutes large and perfectly spheri- cal membranes on 100 per cent, each membrane having a fine precipitate imbedded in it here and there upon the periphery. ...........Agglutinated in one minute. In two hours a few have membranes but the number cannot be clearly made out owing to the floccuient pre- cipitate which surrounds the eggs. .....No agglutination. From this dilu- tion down to that of 1 part in 6,400 the substance had no action upon the eggs. T. Brailsford Robertson 345 From the results it is clear that a substance precipitable from serum by BaCl: or by acetone and soluble in dilute acids and salt solutions is capable of bringing about membrane formation, partial or complete, at certain dilutions. The eggs can also be sensitized to the action of this substance by previous immersion in solutions of SrCl. and CaCl, and the sensitizing action of these substances is clearly seen to reside in their power of forming an insoluble com- pound with the fertilizing agent and precipitating it upon the eggs. So dense is this precipitate when SrCl. is employed as the sensitiz- ing agent that if membranes are formed in solutions of dilutions of from 1 part in 400 to 1 part in 3200 they cannot be observed because the precipitate completely envelops the eggs and hides them from view. The relative sensitizing powers of CaCl, and SrCl, can readily be compared from the following summary of results enumerated above: Unsensitized eggs...............- PS Loe) ty ee lpartin 800 Eggs sensitized with CaCl;......................... lpartin 1,600 PPOReABILIZCG WItn OFCls ............02--20-2-2 5s 1 part in 25,600 It will be recollected (cf. above) that the power of serum to agglutinate sea-urchin eggs runs parallel with its power to fertilize them and that SrCl. and CaCl, sensitize the eggs to both processes. It would appear highly probable, therefore, that the substance which can be isolated from active sera by the process outlined above is the substance which is responsible for the fertilizing, eytolyzing and agglutinating action of these sera upon sea-urchin eggs. Prior to the experiment reported in detail above I made several impure preparations of the substance, contaminated with BaSO,, proteins, etc., all of which fertilized and agglutinated sea-urchin eggs after sensitization with SrCle. The small amounts of this substance which I have as yet been able to obtain have not sufficed to carry out any extensive inves- tigations upon its chemical properties. The preparations which I have made, however, yield Millon’s test. From data which I have obtained and which will be reported subsequently, it appears probable that the fertilizing agent is not present as such in circulating blood, but is derived from the 346 Odcytase breaking down of corpuscles in shed blood. The fertilizing agent is also thermostabie, resisting an exposure of nineteen hours to a temperature of 58°. It consequently appears to be analogous to the eytases or cell-liquefying substances observed by Metchni- koff? and others to be derivable from white corpuscles. I there- fore suggest that this substance be termed oécytase. 7 E. Metchnikoff: L’7mmunite dans les maladies infectiouses, Paris, 1902. ON THE COMBINED ACTION OF MUSCLE PLASMA AND PANCREAS EXTRACT ON SOME MONO- AND DISACCHARIDES. By P. A. LEVENE ann G. M. MEYER. (From the Rockefeller Institute for Medical Research, New York.) (Received for publication, Mareh 14, 1912.) Through the experiments reported in a previous communication! the present writers have demonstrated that by the combined action of muscle plasma and pancreatic extract d-glucose was con- verted into a disaccharide. On the other hand, under the same conditions of experiment maltose was cleaved into glucose. Nat- urally it became important to make clear whether the action was applicable also to other sugars. Of the previous writers only Hall? allowed the muscle plasma to act on other sugars than glu- cose. This investigator extended his experiments to d-levulose, l-arabinose and d-xylose, and was led to the conclusion that the glycolytic action of his enzyme mixture was limited to glucose only. In the course of our previous work the optimal conditions for the action of muscle plasma and pancreatie extract were determined with greater certainty, and this made it urgent to repeat and to extend the experiments of Hall; all the more since the action of the enzyme mixture is viewed at present in a different light. Of the hexoses levulose, mannose and galactose were employed. The mannose was obtained through the courtesy of Dr. Hudson of Washington, and we wish to express our appreciation for his kind- ness. Of the pentoses /-arabinose, d-xylose and d-ribose were used for the experiments. Lactose was the disaccharide tested. In regard to hexoses it was found that mannose remained un- changed under the conditions of our experiments, but d-levulose showed under the influence of muscle plasma and of pancreatic 1 Levene and Meyer: this Journal, ix, pp. 97-107, 1911. 2 Hall: Amer. Journ. of Physiol., xviii, pp. 283-294, 1907. 347 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 348 Action of Muscle Plasma and Pancreas Extract extract a diminution of reducing power which could be restored to nearly the original power’by means of hydrolysis with dilute min- eral acid. In this respect our conclusions differ from those of Hall. The analysis of tables published by Hall, however, shows some disappear- ance of levulose through the action of the enzyme mixture, but the writer is inclined to explain these changes by faults in technique, namely by bacterial contamination. On the other hand, there is no record in his report of the concentration of the sugar employed in his experiments, and the importance of this factor has been emphasized in our previous publication. All our experiments were controlled by bacteriological examination, and only those experiments were taken into consideration which proved free from any bacterial growth. Hence the disagreement between our re- sults and those of Hall in regard to d-levulose is probably caused by the difference in the sugar concentration employed in the experi- ments of Hall and in ours. Regarding the pentoses our experiments are in full accord with those of Hall, and there was never observed a diminution in the reducing power of the pentose even when the concentration of the sugar solutions employed in the experiments was very ¢onsiderable. Also in regard to lactose the results of our experiment harmonize with those of Hall. On the basis of all this experience one is led to the conclusion that the muscle plasma combined with pancreatic extract possesses the power to cause condensation of only two closely related hexoses, namely of d-glucose and of d-levulose, and that it remains without action on mannose, xylose, ribose, and lactose. The same enzyme mixture also has the power to bring about the hydrolysis of mal- tose, but not of lactose. EXPERIMENTAL PART. The enzyme mixtures were prepared in the manner described in the previous communication. All the details of sugar analysis were the same as there described. The condensed levulose was hydrolyzed by heating on a water bath for two hours with 5 per cent hydrochloric acid. The results of the analysis are given in the following tables. P. A. Levene and G. M. Meyer 349 d-Levulose. | ee a = i} x s aa is) : S ; = g ee & a ga | 26/258 leer] zoe] AB) oem (Ome |PSCE ane! se 28 2am Sigs a >aSe gp a og BA aan sei gerre i a ae a. At beginning of experiment...... 0.5 After thirty-six hours........... 0.5 b. At beginning of experiment...... | 0.5 After thirty-six hours............; 0.5 After hydrolysis. . ih ..CRRS c. At beginning of eaearieut seeaare 1.0 After thirty-six hours........... 1.0 d. At beginning of experiment...... 2.0 After thirty-six hours........... 2.0 Galactose. 2 Pe z eS ie 2h | -goe 2 | f0n |oBS Beas og 3) Aonlesog gi | gah 2a seed] 2 A emailer |S oes a. At beginning of experiment............| 0.5 | 23.55 Alter onirty-six HOUPS..........ssedch- 0.5 | 23.60 0 b. At beginning of experiment 0.5} 19.0 After thirty-six hours................../. 0.5] 19.1 0 c. At beginning of experiment 1.0 | 22.8 Attemoninoy=six NOUNS:,.......-.-..«....| -L20 22.6 0 d. At beginning of experiment............| 2.0} 22.9 ARP Mine y-niX HOUES! 5... ccs. 2c. os of 20) 22-9 0 l-Arabinose. B Fe z ~~ 2 as oa -& && Zo | kar | nha lenge Sa | 880/288 [2288 Qa | eh | Ses Besa] » 5 = A =Z 2 RO le ono 5 oO o < y a. At beginning of experiment........... water thirty-six hours................. 0 b. At beginning of experiment............ Atter thirty-six hours................. 0 350 Action of Muscle Plasma and Pancreas Extract d-X ylose. 8801 o o e807 oO SUaLaWILNGD SUqLaWIINGO | SGLGN1ING)D dIGN9 OT O1MN9 COT : o1GNd OT : udd snvuo udd Ssnvup N N 00 00 “add SAVUD ‘ Us) _ ‘a8OTAX ‘aSOL0V1 ‘THOMIU-D | UaALAWILNGO UGLAWILN GO o~wo UALAWILN 2d HN oad uma olgno wad QE CN Oe, o1fn0 uaa x ~ SNO’HN SNOYHN | Oot SNO’HN SNO’HN SNO°HN oman SNOYHN A= SuaLaw SuaLaw Oo) COTS IS SuaLaw a -IIN@O O1809 “LINGO O1N0 | ot os ot st -ILNGO O1€N9 N aago suaian Rat eo aaga suaLan SSS aaso sugnaw os -IINGO 91400 | CO Omnm “LINGO ola#n9 | CO oO SH H -IINGO organo | © © sie. pk eit Ss = ~ : 2) eee ae 3 ee aes 3S 5 od haat Q Sak sei: ] — as li. Tints oO Pe aes 3 ie $e RES oder meet ee asa aca: Sa eaahek esa Or Ole: fae = aca = le A= ro = lari ae Roe wo Rope w =f ee R525 E wo KO RBeRS o£ o4 0.4 oa o 4 a 3 : * SHOE ogo oa wr wet i=") wo 1 “bs a> am a Pap Sen ah at ehas cos ; ; 4 d am Aig Ais Aid dca AS) fe | ce hp * “i + 8 o ® om oO | 252 5 225 ‘= RE ore b= 4 ow oO a a Qo a44< aaa ae St bec 8 os qq P. A. Levene and G. M. Meyer a. At beginning of experiment After thirty-six hours c. At beginning of experiment After thirty-six hours b. At beginning of experiment After thirty-six hours 35! Mannose. | a | | Q ia + | of | & Z| zw aye of lgees og Sas = SE6E eb [gam |Oak lzso8) @z |) 8az|mas|Zo2d) & iz) ° Zz a pe) eet vi cee ean | 0.5 | 24.4 | 48.8 | 15.05 3 ere Aw 0.5 | 24.2 | 48.4) 15.05) 0 bill Eee ae 1.25 | 31.8 | 25.44 7 ote rca Ie Ghole” | -Zo.4 0 ee ee = Fan TZOvibsGeo) | oOLo | 21225 ee er ie 1ObieeGeoco.o | 11.25) 0 <= In calculating the grams sugar for 100 cc. solution the following values for 1 cc. 4; NH:CNS equivalent to milligrams sugar were used. (COLOICE LN GA Ree ee Mgs. 5 3.58 Arabimnosescessns. -. csc 1s. 2.00 Ss ee. Gp lc Re a 2.56 5.78 IRADOSO see oe ine sw disk wie 2.56 5.76 Tig tOSeree te eieis.cies se 200 1.89 * ere a a ee Afr aR, Bere: ae as tule ron ne ~ 2 oe. ; 7 7 SI of in é ’ fo j es «il re 7 ‘ i: so y f { } Be a oie. Ai Wy aa Set saad as 7 mage ae) Ste + ptopal, ; a) t ee ‘ Y 1 o ODF or iaqine ania - seeiielane— ROE Bs wai - BR prow il cree a) ON THE ACTION OF VARIOUS TISSUES AND TISSUE JUICES ON GLUCOSE. By P. A. LEVENE anv G. M. MEYER. (From the Rockefeller Institute for Medical Research, New York.) (Received for publication, March 14, 1912.) The literature on the glycolytic action of various animal organs contains most contradictory and confusing statements. While some writers claimed the presence of sugar-destroying enzymes in all organs and tissues, other observers detected such enzymes only in few organs, and only under very definite conditions, namely, in the presence of some auxiliary substance. Thus, in recent years, Arnheim and Rosenbaum! and Stoklasa? and his co-workers claimed a general distribution of sugar-destroying enzymes in all animal tissues. Rapoport® observed gylcolytic action only in blood and fibrin and obtained negative results from his experiments with other organs. Finally, R. Hirsch‘ and O. Cohnheim!' observed that the glycolytic action is brought about by the combined action of pancreas extract and of the liver, or by pancreas extract and mus- cle plasma. It is possible that the observations reported by every one of the writers are correct, and the apparent contradictions were brought about by the different conditions of the respective experi- ments. It became evident from our work on the combined action of muscle plasma and pancreatic extract that alone variations in the sugar concentration may change the results of the experiments to such an extent that a marked disappearance of glucose will 1 Zeitschr. f. physiol. Chem., xl, p. 220, 1903-1904. 2 Pfliiger’s Archiv, ci, p. 311, 1904. 3 Zeitschr. f. klin. Med., \vii, p. 208, 1905. 4 Hofmeister’s Beitrdge, iv, p. 535, 19038. 5 Zeitschr. f. physiol. Chem., xxxix, p. 336, 1903; xlii, p. 401, 1904; xliii, p. 547, 1904-1905; xlvii, p. 253, 1906. 353 354 Action of Tissues on Glucose take place in one instance, and no change in the sugar content in the other. An analogous observation was made in the work on leucocytes, which will be reported subsequently. Another factor determining the result of the experiment lies in the degree of alkalinity or acidity of the solutions used in the experi- ment. The importance of this factor was pointed out first by the work of Hall, and was corroborated by our own experiments. In agreement with Hail we found that Henderson’s phosphate mixture offers the best medium for the study of the so-called glycolytic process. Still another cause for divergence in the characteristics of the results one may find in the difference in the species of the animal whose organs were employed in the experiment. Recent work on the enzymes of animal tissues has brought to light the great differences in the enzyme content of analogous organs of animals belonging to different species. Of course some of the dis- crepancies in the results of individual writers may have been caused by the different degrees of antiseptic precaution exercised by them. Only rarely were experiments controlled by bacteriological exami- nations. Further, the analytical methods employed by different workers varied greatly in their accuracy. And yet another source for possible error may be found in the fact that rarely was there made an attempt to search for the products of the disappearing sugar. The knowledge of the products formed in course of the experi- ment is important not only for theoretical reasons but as a means of detecting bacterial contamination. Thus in all our experiments of the last two years carbon dioxide was only rarely detected in the reaction mixture, and its presence always indicated bacterial con- tamination; therefore, we are inclined to believe that in the experi- ments of other writers, where there was reported the formation of carbon dioxide from sugar, this resulted from bacterial activity, and not through the actions of tissue enzymes. All these considerations led us to subject to a revision all the older observations on the presence in animal organs of either “‘slycolytic’’ enzymes or of activators of the enzymes, all the more since it became evident that the so-called glycolysis under the combined action of muscle plasma and pancreas consisted in a con- densation and not in a destruction of glucose. In course of this P. A. Levene and G. M. Meyer 355 work the organs of the rabbit and of the dog were employed. In one series of experiments the enzymotic effect of the tissue pulp was tested and in another of the tissue juices. A still different series of experiments aimed to investigate the presence in vari- ous organs of enzyme activators. For this purpose the action of various organs aided by extracts of other organs was tested. Every experiment was controlled by bacteriological examinations, made by Dr. Bronfenbrenner; only experiments that proved free from any bacterial growth were taken into consideration. The results of the experiments were as follows: A. In the experiments with the rabbit, without the aid of acti- vators only the action of liver and of muscle tissues were tested. The results in both experiments were negative. With the aid of pancreas extract also only the same two tissues were tested. Posi- — tive results were obtained only with the muscle plasma. B. In the experiments with the dog without the aid of activators the following tissue juices were employed: muscle, lung, intestine, kidney, pancreas and spleen. All experiments were negative. As activators the extracts of the pancreas and of the spleen were employed. With each activator were employed the same tissue juices as in the first series. The results were the following. The addition of pancreas extract did not alter the action of the tissue juices; on the other hand, after the addition of spleen extract as activator there was observed a fall of the reducing power of the sugar solution in the experiments with muscle, lung, liver and pan- creas. The action was of very moderate intensity. In the experiments with tissue pulp the following organs were used: muscle, spleen, liver, lung and pancreas. The results in all experiments were negative, excepting the liver. The experiments with liver tissue and additional glucose showed at the end of the experiments no change in the original reducing power; on the other hand, in the control experiments with liver tissue alone there was’ observed an increase in the reducing power. Hence it follows that in the experiment with additional glucose there was a com- pensation of phenomena so that the rise of glucose formation was observed by a simultaneous disappearance of glucose. With the addition of pancreas extract as activator the same organs were used. The results were the same as in previous series. 356 Action of Tissues on Glucose With the addition of spleen extract as activator the presence of enzymotic action was observed in the experiments with muscle, lung, liver and pancreas tissues. Thus in the dog the spleen and not pancreas is the organ contain- ing the activator for the enzyme causing the condensation of glucose. ; The general conclusion from the present experiments is that under the conditions here reported, namely, in the presence of antiseptics and under the conditions where access of oxygen is not totally excluded, animal tissues or their juices, aided or unaided by auxiliary substance fail to bring about a destruction of glucose. Wherever a fall in the reducing power of a sugar solution was brought about by the combined action of tissue and activator this was due to a condensation of the glucose. However, the writers realize that under some other conditions an actual gly- colysis may take place and it is hoped the exact conditions will be determined at some future date. EXPERIMENTAL PART. ‘The animals used in the experiments were killed by bleeding from the jugular vein. The organs were removed under aseptic conditions and immediately employed for the preparation of either plasma or. tissue pulp. Tissue Plasma. For the preparation of this, the organs were hashed and extracted for several hours with the Henderson’s phos- phate mixture, and then strained through cheese cloth; the residue was ground with sand and pressed in a Buchner press at 300 atmos- pheres. All the liquids were combined and employed for the experiments. Tissue Pulp. Five grams of the freshly prepared organ pulp were added to a flask containing 45 cc. of a solution of glucose in Henderson’s phosphate solution. Intestinal Extract was prepared in the following way. The intes- tines were washed and the mucous membrane scraped off witha knife. The substance was taken up in a Henderson’s phosphate solution containing glucose and well agitated. Activators. Ten grams of the hashed organ were treated in iden- tically the same manner as previously described for the pancreas. P. A. Levene and G. M. Meyer 357 The final solution was made up to 10 cc. and 1 cc. added to each flask of 50 ce. Sugar Estimations. The reducing power was determined in all cases on the strained mixture. Ten cubic centimeters of this liquid were coagulated by boiling and the addition of acetic acid; and made up to 100 ce. without filtering. This solution was fil- tered through a dry filter and five, ten or more cubic centimeters used for each reduction. The reduced copper was determined by the Volhard method. Organs of a rabbit. A rabbit was starved for three days and placed in a room below freezing temperature for three hours prior to its execution. o eh g28 & gE Spa Z % Soe Q r os na oo 8 ° ° = Zoe goad ro es > Z Aon a= a 5 > 2) Ome OnZ Ps 2 S x wpe Sag 2 = § eo oz Pao 2 B < Z Zi o a Before:---. ae } ; 5.32 After. iar ees,| Sone (20.0 40.0 Beforesy.).. ) 24.0) 48.0 | 17.18 After. _$| Muscle [Pancreas ia 42.0 | 15.08 12.5 Hydrolyzed....| | 93.6| 47.2 | 16.90 Before. 33; 2 é 23.5 47.0 16.82 After are aka | tis | 47.0 14 a5 De ble: cad et 45.8 | 16.39 AST Lan | 23.1} 46.2 | 16.53 1 ce. liver plasma—no reduction. 1 ce. liver plasma hydrolyzed—no reduction. 358 Hydrolyzed. Before?. sid; Hydrolyzed.. Before:...-.2-7 Afiterss: ee ATter. nee | 4 ‘ \ i Action of Tissues on Glucose Tissue plasma —_,— Muscle Muscle Muscle } | Muscle Muscle Lung | Lung Lung Lung | Intestine Intestine Intestine Intestine Intestine Kidney Pancreas Spleen | ACTIVATOR None | Spleen | Pancreas | Spleen | Pancreas | None Spleen | | Pancreas | Spleen None Spleen | Pancreas Pancreas of dog. NH.CNS ° ° bo Be NOOO FN eB eH OC —————— ee oN Ww Ww bd —_—_S Oooo oon eos — eea—e—=*$8 By te be Nee i be : ¢ SONNNWSOSOO DR ¢ bo G9 Ot 0 OD OO De OF fF ORDOANMMNMOKY ODEN OOONOUDAWOORD _ bd bt by bo Co & bo CWSWwh hyd t ww NONCDAANW WwW OS comin aut ceri eatin! gute ane axe ama NH,CNS per CUBIC CENTI- METER 41.2 | he lop) | | | } SSSSSSESSSR SSR 4 Be ie Pe oe ai Gre Ss ‘ ¢ Ko ie lt UOT OU aT NOON DWNNODWDAANAGDDTDODONOCANKOTDOROTOONDDWO PODD > a oD HD © © « | | Z52 id gBE|SE | ai=a aSa| ace | a5 Rog. SHz) BRE | SRG 203) S la 12.05, 2.25 13.65 | ae | } i | | 20.4 | 19.69 0.71 | 16.08 16.30) 14.75| 14.89) 14.60 { | | LOSS PER CENT 4.8 6.4 15.7 3.5 P. A. Levene and G. M. Meyer 359 Organ pulp of dog, without activator. a Z23 ao & aE eps | e& z mn ne on8 ee 2 Z Zoe aez i 8 S Seevimeee eee = z = Gee oses | see |g B Zz Zz, 3 gs | 8 Before...... \ 16.9 33.8 After....... oe eee 34.0 Before..... } oe 93.7 |, 47.4 After... ee 23.7| 47.4 | Before...... ——— 19.2 38.4 hee | \19.0| 38.0 0.14 | 1.0 Before...... \ eee | { 25.4] 50.8 | After....... | \25.5] 51.0 | Before...... (2327 47.4 | After....... \ 26.) AT2 Pao EE: a 5 Bz EO8 Sk 8 ° Dox ase 208 4 4 Za 2=5 Ao Q e See | she | gfs 2 g Bo haere lesa 5 Before... 40.4| 14.46 ferai Se oecen (ee 14.03 2.9 Before... | > Spl $440) (1231 After.... {| “U8 Le cee 33.8] 12.10 | 0.21 1:7 Before... Li ks 35.4 1- 12.70 ieee a oe ea 12.70 | 0.00 Mees aa eicen [34.2 12.24 After... ail (aes) tom) 0.14 | 1.1 Before...\ | 5, nas 32.0} 11.45 | After... | = ean gue mea 31.0 | 11.09 0.36 St Organ pulp of dog, without glucose. SE PO ES Ob | rr ro ee No reduction PUMPER TIDY 5. ios we ins» Se ins aS Rie ase > aloe Simic No reduction PEIPMEEMOTOAS PULP (os osc 5 oi 3's Ss ve dslcle's gece No reduction 2 LD Eon eee of No reduction 5 grams Before.......10—9.3 —0.7 NH,CNS = 0.25 per cent liver pulp |Hydrolyzed .10 —7.9 — 2.1 NH,CNS = 0.75 per cent 4 > » aera , ary . 4 a ee , Wage” be THE ACTION OF LEUCOCYTES ON GLUCOSE. By P. A. LEVENE anp G. M. MEYER. (From the Rockefeller Institute for Medical Research, New York.) (Received for publication, March 14, 1912.) The final products of sugar combustion in the animal organism are carbonic acid and water. This information is undisputable. The knowledge of every other phase in the complex process of glycolysis has not advanced beyond the state of conjecture. The paucity of undisputed knowledge regarding this very important biological process is due to the fact that intravital reactions are made visible only with great difficulty, and secondly to the great variety of theoretically possible processes which may lead to the breaking of the links between the individual carbon atoms of the sugar molecule. There are three principal types of sugar degradations. One begins directly with the oxidation of the end carbon atoms, the second consists in a gradual dissociation of formaldehyde. It represents a reversion of the synthesis of sugar from formaldehyde. The first step of this reaction leads to the formation of pentose from glucose. The reaction of the third type leads to a splitting of the six-carbon chains of glucose into two three-carbon chains—with the formation of either dioxyacetone or of lactic acid as the first phase of.the reaction. CH,.0OH CH.OH COOH CO; : ~ PS (CBOn),; — > (CHOH), => | (CHO); a 2(COOH),. a CHO COOH COOH CO, II. CH,OH(CHOH);CHOH CHO = CH:0H(CHOH);CHO + HCHO 361 362 Action of Leucocytes on Glucose CHO CHO CHO CHO COOH | | | | CHOH ,, COH CO CO CH (OH) ee Was | CHOH Ge = (CH) (GH: CH; IIIa. | | | CHOH CHOH CHOH CHO CHO CHO | | | | | | CHOH CHOH CHOH CHOH-—! qo) co | OH | | | | CH.0H CH:OH CH:0H CH,0H CH: — CH; b. C5H120¢ = 2 CH;CHOH— COOH Cc. CeHi20¢ 74 CH:0H—CO—CH,0H Each one of the three types of reactions was taken into consid- eration for the explanation of the process of sugar combustion in the body and attempts were made to interpret the process on the basis of every one of the three reactions. The method employed for the purpose of bringing to light the mechanism of biological cleavage of sugar were the following: 1. The search in the tissues for substances that may originate from carbohydrates. 2. Experiments on the action of tissues and tissue extracts on sugar. 3. Perfusion of organs with carbohydrates and their cleavage products. 4, Feeding healthy and diabetic animals on carbohydrates and their oxidation product. “ None of the methods has furnished thus far convincing evidence which would permit the acceptance, or would force the rejection of either one of the possible interpretations of sugar combustion in the animal. For the reasons which will become evident later, we shall briefly review the work which aimed to analyze the possi- bility of the intermediate cleavage of hexose into two molecules each containing a chain of three carbons. Fou substances had been named in connection with the biological sugar combustion: lactic acid, glyceric aldehyde, methylglyoxal and dioxyacetone. In connection with sugar combustion in the animal organism only lactic acid received serious consideration. The formation of lactic acid in surviving tissues at the expense of disappearing gly- cogen was accepted by older writers. In those experiments the P. A. Levene and G. M. Meyer 363 . influence of bacteria was not excluded, and the subject received a revision in recent years The views of the writers who were actively engaged in the investigation of lactic acid in the animal organism are equally divided, some regarding protein and others carbohydrates as the source of the substance. The writers who in recent years contributed evidence in support of the carbohy- drate origin of lactic acid were Spiro,! Hoppe-Seyler? and his co- workers, Araki? and Zillessen,t Embden,* and particularly Graham Lusk and A. R. Mandel. The most emphatic partisan of the protein origin of lactic acid was Minkowski’ and his views are supported by evidence adduced by Asher and Jackson,’ and by Neuberg and Langstein.°® The views of Hoppe-Seyler and his co-workers were based on observations on animals placed in conditions which brought about insufficient oxygenation of the tissues. In all such conditions there was observed: the élimination of lactic acid through the urine. Inouye and Saiki made similar observations in epilepsy. Emb- den reached his conclusions on the basis of perfusion experiments. The perfusion of livers rich in glycogen resulted in lactic acid for- mation. The same result was observed when blood containing sugar was perfused through a liver poor in glycogen. On the other hand, perfusion of a liver freed from glycogen with blood of a very low sugar content failed to bring about lactic acid formation. Lusk and Mandel based their view on experiments on dogs which received combined phloridzin and phosphorus treatment. It is known that phosphorus injected into normal animals causes the elimina- tion of lactic acid. On the other hand, the injection of phloridzin caused the removal, through the urine, of glucose from the body tissues. Ifthe phosphorus injection was preceded by a phloridzin injection it failed to give rise to an elimination of lactic acid. The 1 Zeitschr. f. physiol. Chem., i, p. 111, 1877. 2 Festschr. zu Virchow’s Jubilaum. 3 Zeitschr. f. physiol. Chem., xv, pp. 335 and 546, 1891; xvi, pp. 201 and 453, 1892; xvii, p. 311, 1893; xix, p. 422, 1894. * Zeitschr. f. physiol. Chem., xv, p. 387, 1891. 5 Centralblatt f. Physiol., xviii, p. 832, 1905. 5 Amer. Journ. of Physiol., xvi, p. 129, 1906. 7 Arch. f. exp. Path. u. Pharm., xxi, p. 67, 1886; xxxi, p. 214, 1893. 8 Zeitschr. f. Biol., xli, p. 393, 1901. ® Arch. f. Physiol., Suppl. 1903, p. 514. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4 364 Action of Leucocytes on Glucose interpretation given by Lusk and Mandel to these observations was that normally under the influence of phosphorus lactic acid is formed at the expense of glucose; and that phloridzin in remov- ing glucose also removes the mother substance of lactie acid. On the other hand, Minkowski observed that birds, after re- moval of the liver, eliminate through their urine considerable quantities of lactic acid, and that its value is influenced by the pro- tein intake and not by that of carbohydrates. Jackson and Asher in perfusion experiments failed to detect any influence of carbo- hydrate on the lactic acid formation and Neuberg and Langstein demonstrated the appearance of lactic acid in the urine after ad- ministration of alanine. Thus the evidence in support of either view was mostly indirect and not sufficiently convincing. These considerations impelled us a year ago to undertake the study of the products formed in course of glycolysis by means of tissue extracts. The results were all negative. For various reasons it was deemed advisable to test the influence on glucose of living leucocytes. Lepine!® and his co-workers and Mayer" had already advanced the view that leucocytes were concerned in the process of glycolysis. Their evidence was indirect and the products of leucocytic glycolysis remained unexplained by them. In the present investigation all the experiments were performed under absolutely aseptic conditions. The solutions were always tested for aérobic and anaérobic microédrganisms by Dr. Bronfen- brenner to whom we are greatly indebted for that part of the work. The leucocytes were suspended in a sugar solution containing 15 per cent of the Henderson phosphate mixture. The results of the experiments were the following: 1. Under the influence of leucocytes a sugar solution loses part of its reducing power. 2. The reducing power cannot be restored to the original by boiling with mineral acids. 3. The rate of glycolysis is in inverse proportion to the sugar concentration. (The last two points are interesting in connection with the influence of muscle plasma and pancreatic extract of glucose. There the fall in the reducing power was in direct pro- 10 Le diabete sucré, Paris, 1909. 11 Arch. de Physiol. 2. P. A. Levene and G. M. Meyer 365 portion to the sugar concentration, and the original reducing power could be restored by hydrolysis with mineral acids.) 4. If distilled water is employed in place of the phosphate mix- ture the leucocytes fail to exert any influence on glucose. 5. If toluol is added to the phosphate mixture the leucocytes do not demonstrate any action on glucose. 6. As product of the action of leucocytes on glucose, paralactic acid was discovered. It was identified as the zinc salt. Volatile . acids were not detected. 7. The quantity of lactic acid found was lower than that of the disappeared glucose. Whether the missing sugar underwent decomposition into other substances than lactic acid, or was used for synthetic purposes by the leucocytes remains to be established. EXPERIMENTAL. Leucocytes. Medium sized dogs were given two injections of 1.5 cc. turpentine into the pleural cavity at an interval of three days. Ether narcosis was used at the first injection. Eighteen hours after the second injection the liquid which had formed in the pleural cavity was withdrawn by aspiration. This exudate, which contained the greater portion of the turpentine, was dis- carded. The following day the aspiration was repeated. This fluid was received into sterile bottles. ‘The color of the exudate was a straw yellow occasionally tinged red. The quantity of exu- date obtained from each dog varied greatly. Three or more dogs were injected simultaneously so as to insure an ample supply of leucocytes. The combined exudate was centrifugalized and the leucocytes washed twice with sterile physiological saline. The centrifugal flasks contained glass beads to facilitate the breaking up of the agglutinated mass and thus aid proper washing and mixing. Solutions. In the first three experiments the leucocytes were well shaken with sterilized 1 per cent Henderson phosphate solu- tion and this suspension added to flasks containing the desired quantity of glucose. Merck’s “highest purity” glucose was used in all experiments. Just enough water was added previous to sterilizing, so that it remained liquid after cooling. In later experiments, where only one concentration of glucose was desired, 366 Action of Leucocytes on Glucose the sterilized glucose syrup was dissolved in the phosphate solution and this then added to the leucocytes and well shaken. The quan- tity of glucose-phosphate solution which was added to the leuco- cytes depended to a certain extent upon the quantity of leucocytes at-hand, although no attempt was made to count the cells. The mixture was usually made up to a volume nearly one-half of that of the total exudate. As the exudate contains approximately 10 per cent leucocytes, the glucose was acted upon by a 20 per cent leucocyte suspension. ‘The mixture was kept for thirty-six hours , at 37°C. The flasks were stoppered with cotton and well covered and sealed with tin foil to prevent evaporation. Methods of Analysis. Immediately after mixing the leucocytes with glucose and after thirty-six hours samples were withdrawn for analysis. The leucocytes were allowed to settle and the clear supernatant liquid only used for the sugar determinations. The liquid was freed of protein by boiling and acetic acid. The reduced copper was estimated by the Volhard method. The details of the sugar determination are identical with those already described in a previous communication. Carbon Dioxide. A measured Talitne of the leucocyte mixture was used for this determination according to the method of Fre- senius and Classen. The liquid was freed as far as possible from protein by heat while still alkaline. Phosphoric acid was used to acidify the mixture. ; Lactic Acid. The filtered residue from the carbon dioxide deter- mination was evaporated nearly to dryness. Anhydrous sodium sulphate was then added and all carefully ground in a mortar to an impalpable dry powder. . This was then extracted with anhydrous ether, until the extract gave no further test for lactic acid. All extracts were combined and freed of ether. The residue was taken up ina little water and boiled with zine carbonate. The filtered. aqueous solution was evaporated to dryness and the total weight obtained. This residue was usually more or less colored. It was redissolved in water and clarified by boiling with animal char- coal. After evaporating to a small volume zinc lactate soon crystal- lized. The crystals were used for further analysis and identifi- cation. Volatile Acid. The uncoagulated leucocyte mixture was dis- tilled with steam into #4 barium hydrate. Glacial’ phosphoric acid was added through a separatory funnel in such a manner as ~ P. A. Levene and G. M. Meyer 367 to prevent any loss of CO, and volatile acids. The adapter from the condenser dipped into the barium hydrate which was contained in a tall stoppered cylinder, connected with a series of wash bottle likewise containing barium hydrate and finally a soda lime tube to guard against absorption of carbon-dioxide through accidental back pressure. A. Experiments showing the relation between rate of glycolysis and concen- tration of sugar solution. EXPERIMENTI. 600 cc. exudate was obtained fromtwodogs. Four flasks with glucose at different concentrations were prepared. { a ' ee 2 Eom z : 3 E = a2 [86 |2e2| 8 (e828) ¢ e& |obm|Oek| SS |*ae, ae | BR | Be2)mes) zo | 25s] #8 oO o FZ om ~ a | a. At beginning of experiment...... BS Ors | 19.0 | 38.0 After thirty-six hours........... | 0.5 | 19.0 | 38.0 ) 5. At beginning of experiment ..... | 1.0] 30.0 | 30.0 After thirty-six hours........... 1.0 | 30.0 | c. At beginning of experiment......| 2.0 | 45.0 After thirty-six hours........... 2.0 | 39.5 | After hydrolysis... re es d. At beginning of lavseiteaee 2.0 | 36.6 After thirty-six hours........... ui 2.0 | 33.0 | After hydrolysis.................| 2.0 | 33.0 EXPERIMENT II. 500 cc. exudate. Four flasks at two different concen- trations. The remainder 100 cc. leucocyte suspension used as control | a g eB | E 2| g6| Seb | 2 26 | gat | Cee | sox 3 a g /gae (mee /i58/ ¢ | 35 I. At beginning of experiment....| 1.0 | 24.4 | 24.4] 8.72 After thirty-six hours......... \ 10°) 22-5 | 2208.07 | 0.65 | 7.5 II. At beginning of experiment....| 1.0 | 23.9 | 23.9 | 8.55 After thirty-six hours.........| 1.0] 22.0} 22.0} 7.87| 0.68} 7.9 III. At beginning of experiment....| 1.0 | 18.5] 18.5 | 6.62 After thirty-six hours......... .O| 16.8 | 16.8} 5.91 10.7 IV. At beginning of experiment...., 1.0 | 18.3 | 18.3 | 6.55 After thirty-six hours.......... .O | 16.5} 16.5 | 5.90 10.0 368 Action of Leucocytes on Glucose B. Experimenis aimed to test the formation of volatile acids during the process of glycolysis. ExPERmMENT Ja. 100 cc. control leucocytes (from A experiment II) acidified and distilled with steam into7; Ba(OH)2. The Ba(OH), was fil- tered from the carbonate and titrated; phenolphthalein used as indicator. 20 ce. f¢ Ba(OH) required 17.6 cc. 7g HCl = 2.4 ce. EXPERIMENT Is. 99 cc. of the leucocyte-glucose mixture was treated in identically the same manner. The 99 cc. consisted of— BUGAR BEFORE BUGAR AFTER LOss grams grams grams i cece. 2.61 2.42 0.19 Eh! 27 cere 2.30 2.12 0.18 TE 25'ce ees Say: 1.65 1.47 | 0.18 EVA Zeek. ee 1.13 1.00 | 0.13 7.69 7.01 | 0.68 | | | | 37.3 ec. 7p Ba(OH): after distillation and filtering from the carbonate required 34.7 cc. 74 acid = 2.6 ce. 75 neutralized. C. Experiments showing the development of lactic acid in course of glycolysis. EXPERIMENT Ia. The residue from B. Experiment Ia. was extracted in a Schwartz extractor with ether for lactic acid. There was only the merest trace of residue obtained from the ether extract, which was neutral tolitmus. EXPERIMENT Ip. The residue from B. Experiment Ib. was extracted for lactic acid. Yield of crude zinc lactate dried at 100° = 0.3184 gram. 0.0961 gram recrystallized salt, air dried after drying to constant weight lost 0.0106 gram H.O = 12.3 per cent H.O. Calculated for two molecules H.O = 12.5 per cent. 0.1362 gram recrystallized salt on ignition yielded 0.0458 ZnO = 33.62 per cent. Calculated = 33.4 percent. EXPERIMENT I]. Contents of flasks I and II (A. Experiment II) were coagulated and extracted for lactic acid. No lactic acid was obtained from II. The yield from I = 0.305 gram crude zinc lactate. 0.2616 gram recrystallized and air dried salt on heating to constant weight at 100° lost 0.032 gram = 12.35 per cent H.O. Calculated for two molecules H.O = 12.5 percent. 0.1135 recrystallized and dried at 110° after ignition gave 0.0383 ZnO = 33.7 per cent ZnO. Calculated = 33.4 per cent. EXPERIMENT IIIa. Contents of flask I Experiment E (170 cc.) was coagu- lated and filtered, the filtrate neutralized and evaporated nearly to dryness. The syrup was acidified with a small quantity of glacial phosphoric acid and ~ Po A. LeveneiandiG: M:: Meyer 369 well ground with highest purity anhydrous sodium sulphate. This powder was then repeatedly extracted with hot ether. The ether extracts were combined and treated as previously mentioned to obtain zinclactate. Yield of crude zinc lactate dried = 0.4798 gram. 0.2868 gram recrystallized at 100° dried salt dissolved in 2.84 cc. of water gave arotation in the polari- scope of — 0.22°. 0.3274 gram of the recrystallized salt after drying at 110° weighed 0.2868 gram. ‘Loss = 0.0406 gram = 12.31 percent H.O. Calculated for two molecules H,O = 12.5 percent. 0.0999 gram of dried zinc salt gave 0.0337 gram ZnO = 33.8 per cent ZnO. Calculated = 33.4 per cent. EXPERIMENT IIIB. The controls, flasks II and III Experiment E (170 cc.) were subjected separately to the identical treatment as flask I. No zine lactate was obtained. D. Experiments showing effect of distilled water and dilution; 700 cc. exudate. I. Three flasks with 1.86 grams glucose, mixed with leucocytes and phos- phate solution. II. Two flasks with 3.75 grams glucose with leucocytes and distilled water. III. 50 ce. leucocytes and phosphate solution set aside for control. | gh |gb |Qezlsoe| , | 88 [oe pene ojo gpa | ig SS I. At beginning of experiment....| 4.0 | 35.0 8.75) 3.12 After thirty-six hours......... | 4.0) 32.4| 8.1] 2.90] 0.22} 7.8 At beginning of experiment..... 4.0 32.8 8.2| 2.94, After thirty-six hours......... | 4.0] 30.0] 7.5 | 2.68|0.26| 8.8 At beginning of experiment..... 4.0 | 32.6 | 8.15) 2.92 | After thirty-six hours......... 4.0| 29.6| 7.40 2.65 | 0.27) 9.2 II. At beginning of experiment..... 2.0 | 39.3 | 19.65) 7.04 | After thirty-six hours......... | 2.0 | 39.4 | 19.70. 7.05; 0 0 At beginning of experiment...., 2.0 | 35.4 | 17.7 | 6.32 | After thirty-six hours......... 2.0 | 35.6 | 17.8|6.33| 0 0 IIT. 25 cc. of leucocytes and phos- phate solution gave no appre- | ciable reduction of Fehling’s| solution. There was like- wise no reduction after two hours hydrolysis with 2 per cent HCl. 370 Action of Leucocytes on Glucose E. Effect of adding toluol; control of action of phosphate solution on glucose; 800 cc. exudate. I. One flask of 200 cc. glucose, leucocytes and phosphate solution. II. One flask of 200 ce. as I, with addition of toluol. III. Glucose and phosphate solution. a a | eS |e m| B6/ Boe 3 fa eee ieee less) | ¢ gQ Sons & Qo | a mn ef | eft | Qe2| Sok oe a | Baz lied | ess] 8 | = o Oo | Zz 5 a Pa I. At beginning of experiment.....| 2.0 | 34. After thirty-six hours..........| 2.0 | 31. 0.45 | 7.4 II. At beginning of experiment... .. 2.0 | 34.3 After thirty-six hours.......... 2.0 | 34.3 0 0 III. At beginning of experiment....| 2.0 | 35.2 | After thirty-six hours.......... 2.0 | 35.2 0 | 0 ON THE ACTION OF TISSUE EXTRACTS CONTAINING NUCLEOSIDASE ON a AND 8 METHYLPENTOSIDES. By P. A. LEVENE, W. A. JACOBS anp F. MEDIGRECEANU. (From the Rockefeller Institute for Medical Research, New York.) (Received for publication, March 14, 1912.) Through the work of Levene and Jacobs it was established that the purine bases enter the molecule of the plant nucleic acid and of some animal nucleic acids in the form of a d-riboside. The struc- ture of these may be represented by that of guanosine: N=—=C NH, H H OC GN: CH.—C——_C——C——CH:0H pbs oiy |i Eiet s ORAS ley | ome agin = N= C=-N Regarding two points of their structure there exists at present no definite information. The first concerns the place of the union between the two molecules. The formula given here assumes a union in position 7 of the purine base, but the experimental evi- dence admits with the same degree of probability also the position 8. The second pertains to the two possible stereoisomeric forms of the pentosides. The assumption of the lactonic structure of gly- cosides admits of the existence of two isomeric forms of each gly- coside conditioned by the asymmetric nature of the end-carbon. Hence theoretically there are possible a and 6 forms of the nucleo- sides in the same manner as there exist a and 6 forms of any other pentoside, and it therefore remains to be established whether the natural nucleosides belong to the a or the 8 series. Three methods are available for the solution of the last problem. The first was introduced by Fischer! and is based on the specific 1 Zeitschr. f. physiol. Chem., xxvi, p. 61, 1898. 371 372 Action of Tissue Extracts on Methylpentosides power of certain enzymes to cause the cleavage of glycosides of one order leaving intact the other stereoisomeric form. Thus emulsin is capable of hydrolyzing 8 glycosides but not the a forms. On the contrary maltose has no capacity for disrupting the B forms, but possesses one for the a forms. This method was not available in the present investigation, for the reason that neither emulsin nor maltose, in the form as we were able to procure them, had the capacity for cleaving the nucleosides. The second method was introduced by Armstrong? and is — on the observation of the mutarotation of the sugar liberated from the glycoside. If the mutarotation is analogous to the rotation which characterizes the transformation of the a-isomer into the stable form, the glycoside is regarded as the a-glycoside and vice versa. This method is available when there exists an enzyme capable of hydrolyzing the glycoside with a sufficiently high degree of intensity, and besides, when the transformation of the isomeric sugars into their stable forms proceeds at a low rate of velocity. Unfortunately the sugar liberated from the nucleosides possesses a low rotation and is very rapidly transformed into the stable form, so that it shows only a low degree of mutarotation, which becomes evident only under very definite conditions. On the other hand, the cleavage of the nucleosides by the nucleases pro- ceeds very slowly, so that in a moderately short interval only little ribose is liberated. Another difficulty was encountered in the fact that the hypoxanthine formed in course of the cleavage (inosin was used in the experiments) combined with some of the unchanged nucleoside, giving rise to a precipitate of carnin. All these diffi- culties made the method of Armstrong of little value for the pur- poses of the present investigation. The third method was introduced by Hudson,* and is based on certain numerical values of the rotations of sugars and their gly- cosides. This method permits of establishing the nature of a glycoside when only one formis known, but when simultaneously there exists information regarding the specific rotation of at least one form of the sugar. According to Hudson the difference in the molecular specific rotation of two sugars has a constant value of 2 Journ. Chem. Soc., 1xxxiii, p. 13805, 1903. 3 Journ. Amer. Chem. Soc., xxxi, p. 66, 1909. P. A. Levene, W. A. Jacobs and F. Medigreceanu 373 16200. Hence the knowledge of the value of the specific rotation of one form permits of obtaining the value of the other form. Further, the sum of the molecular specific rotations of the two forms remains constant for every sugar and all its glycosides. Thus, if one possesses the knowledge of the sum of the specific rotations of the two forms of a sugar, he is also in possession of the information regarding the sum of the specific rotations of the glycosides. Hence it is possible to calculate by a simple arithmetical process the specific rotation for the second isomeric glycoside when that of the first is known. Thus accepting the difference in specific molecular rotation expressed by the formula — a + 6 = 453° = 108, and accepting for one isomer [a], = — 14.65, the rotation of the other will equal — B = — a — 108 = — 122.65°. When the rotations of the two forms are given Hudson suggested the following rule for naming the a and 8 forms: ‘‘The names should be selected that for all sugars which are genetically related to d-glucose the subtrac- tion of the rotation of the 6-form from the a-form gives a positive difference and for all sugars which are genetically related to l-glucose an equal negative difference.”” According to this rule the unknown form of d-ribose is to be named the 6-form. The information obtained in this manner furnishes also the value for the sum of the rotations of the two isomers of d-ribose, — 137.30°, which is alsothe sum of the rotations of all glycosides of the same sugars. Applying this rule for inosin, of which theknown form has the rotation of —49.2°, one is led to the conclusion that the other form has the rotation of —87.80°, and is therefore f-d- riboside. Thus this process of reasoning leads to the conclusion that the natural nucleosides belong to the a series of glycosides. This view may be correct, but in the absence of all other evidence, one would hesitate to declare this deduction perfectly conclusive. Hence it seemed desirable to search for additional data that would give more force to the above conclusion or would compel its rejection. With this aim in view it was attempted to obtain more informa- tion regarding the action of the nucleosidases present in the animal tissues. It has been mentioned already that glycosidases of plant origin possess a selective hydrolytic aptitude for only one form of 374 Action of Tissue Extracts on Methylpentosides glucosides and Fischer and Nobel* have demonstrated that the glycosidases of the animal tissues were capable of hydrolyzing only one form of the glucosides, namely the 6-form. Hence the @ and 8 forms of methylxylose and methylarabinose were prepared and added to a solution containing the active nucleosidases. Methylribose could not be obtained in crystalline form, and therefore the a and the 8 forms could not be separated one from another. The efficiency of the enzyme was always tested on nucleosides. To our surprise all tissue extracts failed to act on any one of the pentosides, the ribosides included. The action of the enzyme was tested by the optical method and by the reduc- ing power of the solution for Fehling’s solution. Thus the present experiments failed to contribute to the knowl- edge of the structure of the nucleosides, but have furnished new information regarding the nature of nucleosidases showing that they possessed a greater degree of specificity than is known to be the property of many glycosidases. EXPERIMENTAL PART. Organ plasma was prepared in the manner described in a pre- vious communication of Levene and Medigreceanu.® All other details of the experiments were the same as there described. Q-METHYLARABINOSIDE EXPERIMENTS. In neutral phosphate solution (1 per cent). EXPERIMENT WITH EXTRACT OF INTESTINAL Mucosa. I 16; 711: Enzyme solution, 1 ce. a-Methylarabinoside solution, 3 cc. Control: Enzyme solution, 1 cc. Phosphate solution, 3 cc. 10 min. 24 hrs. 48 hrs. 96 hrs. Experiment: +4.87 +4.86 +4.86 +4.86 Control: 0.00 0.00 0.00 0.00 4 Sitzungsberichte Berliner Akad., v, p. 73, 1896. 5 This Journal, ix, p. 65, 1911. P. A. Levene, W. A. Jacobs and F. Medigreceanu 375 EXPERIMENT WITH PANCREAS PLASMA. Tle a: Enzyme solution, 1 cc. a-Methylarabinoside solution, 3 cc. Control: Enzyme solution, 1 ce. Phosphate solution, 3 ce. 10 min. 24 hrs. 96 hrs. Experiment: +5.00 +4.97 +4.96 Control: +0.06 +0.06 +0.06 Q-METHYLXYLOSIDE EXPERIMENTS. In neutral phosphate solution. EXPERIMENT WITH Extract oF INTESTINAL Mucosa. Fore ads Enzyme solution, 1 ee. a-Methylxyloside solution, 3 ce. Enzyme and phosphate solution, Control: see a-methylarabinoside experiment. 10 min. 24 hrs. 96 hrs. +2.90 +2.90 +2.89 EXPERIMENT WITH PANCREAS PLASMA. bg Urea Enzyme solution, 1 cc. a-Methylxyloside solution, 3 ce. Enzyme and phosphate solution, Control: see a-methylarabinoside experiment. 10 min. 24 hrs. 96 hrs. +3.00 +2.98 +2.98 Q@-METHYLGLUCOSIDE EXPERIMENTS. In neutral phosphate solution (1 per cent). EXPERIMENT WITH Extract OF INTESTINAL Mucosa. P93 Lt Enzyme solution, 1 ce. Glucoside, 5 per cent, 3 cc. Enzyme and phosphate solution, Control: see maltose experiment. 10 min. 18 hrs. 120 hrs. +3.08 +3.06 +3.05 376 Action of Tissue Extracts on Methylpentosides ETS OS at; Control: EE oy i Control: EXPERIMENT WITH PANCREAS PLASMA. Enzyme solution, 1 ce. Glucoside, 5 per cent, 3 cc. Enzyme and phosphate solution, see maltose experiment. 10 min. 18 bre. 120 hrs. +3.15 +3.16 +3.16 EXPERIMENT WITH KIDNEY PLASMA. Enzyme solution, 0.5 cc. Glucoside, 5 per cent, 3.0 cc. Phosphate solution, 0.5 cc. Enzyme and phosphate solution, see maltose experiment. 10 min. 18 hrs. 120 hrs. +3 .00 +2.98 +3.02 AMYGDALIN EXPERIMENTS. In neutral phosphate solution (1 per cent). EXPERIMENTS WITH Extract oF INTESTINAL Mucosa. Experiment I. Control (1): Control (2): Experiment: Control (1): Control (2): Experiment II. Control (1): II, 17, ’11. Enzyme solution, 1 cc. Amygdalin, 7 per cent, 3 cc. Enzyme solution, 1 ce. Phosphate solution, 3 cc. Amygdalin solution, 7 per cent. 10 min. 24 bra. 96 hrs. 144 hrs. —1.10 —1.24 —1.29 —1.32 0.00 0.00 0.00 0.00 —1.48 —1.70 —1.90 —2.00 II, 20, 11. Enzyme solution, 1 ce. Amygdalin, 10 per cent, 3 ce. Enzyme and phosphate solution see preceding experiment. P. A. Levene, W. A. Jacobsand F. Medigreceanu 377 Control (2): Amygdalin solution, see preceding experiment. 10 min. 24 hrs. 96 hrs. 300 hrs. —1.47 —1.62 —1.65 —2.05 Reduced Fehling’s solution. EXPERIMENTS WITH PANCREAS PLASMA. Experiment I. II, 17, 711. Enzyme solution, 1 ce. Amygdalin, 7 per cent, 3 cc. Control (1): Enzyme solution, 1 cc. Phosphate solution, 3 cc. Control (2): Amygdalin solution, see experiment with extract of intestinal mucosa. 10 min. 24 hrs. 48 his. 72 brs. 200 hrs. Experiment: —1.05 —1.10 —1.25 —1.35 —1.43 Control (1): +0.06 +0.06 +0.06 +0.06 +0.06 Experiment II. II, 20, ’11. Enzyme solution, 1 cc. Amygdalin, 10 per cent, 3 cc. Control (1): Enzyme and phosphate solution, see preceding experiment. Control (2): Amygdalin solution, see experiment with extract of intestinal mucosa. 10 min. 24 hrs. 96 hrs. 300 hrs. Experiment: —1.45 —1.62 —1.85 —2.12 EXPERIMENT WITH KIDNEY PLASMA. Experiment I. II, 17,’11. Enzyme solution, 0.5 cc. Amygdalin, 7 per cent, 3 cc. Phosphate solution, 0.5 ce. Control (1): Enzyme solution, 0.5 ce. Phosphate solution, 3.5 ce. Control (2): Amygdalin solution, see experiment with extract of intestinal mucosa. . 10 min. 24 brs. 72 brs. 200 hrs. Experiment: —0.90 cloudy —1.14 —1.20 Control (1): —0.04 cloudy —0.03 —0.03 Experiment II. II, 20,’11. Enzyme solution, 0.05 ce. Amygdalin, 10 per cent, 3 cc. Phosphate solution, 0.5 ce. 378 Action of Tissue Extracts on Methylpentosides Control (1): Enzyme and phosphate solution, see preceding experiment. Control (2): Amygdalin solution see experiment with extract of intestinal mucosa. 10 min. 24 brs. 72 hrs. 120 hrs. 300 hrs. —1.30 cloudy —1.60 —1.80 —2.00 Did not reduce Fehling’s solution. B-METHYLXYLOSIDE EXPERIMENTS. In neutral phosphate solution (1 per cent). EXPERIMENT WITH Extract oF INTESTINAL Mucosa. I, 16,."ht Enzyme solution, | ce. 8-Methylxyloside solution, 3 cc. Control: Enzyme and phosphate solution. see a-methyl- arabinoside experiment. 10 min. 24 brs. 96 hrs. —2.60 —2.60 —2.56 EXPERIMENT WITH PANCREAS PLASMA. FS 16: EF. Enzyme solution, 1 ce. 6-Methylxyloside solution, 3 ee. Control: Enzyme and phosphate solution, see a-methyl- arabinoside experiment. 10 min. . 24 hrs. 96 hrs. —2.00 —2.00 —1.98 B-METHYLARABINOSIDE EXPERIMENTS. In neutral phosphate solution (1 per cent). EXPERIMENT WITH Extract OF INTESTINAL Mucosa. ib 16.71 Enzyme solution, 1 ce. 8-Methylarabinoside solution, 3 cc. Control: Enzyme and phosphate solution, see a-methyl- arabinoside experiment. 10 min. 24 hrs. 96 hrs. +0.72 +0.68 +0.68 P. A. Levene, W. A. Jacobs and F. Medigreceanu 379 EXPERIMENT WITH PANCREAS PLASMA. ¥, 16, "FH. Enzyme solution, I ce. 8-Methylarabinoside solution. 3 cc. Control: Enzyme and. phosphate solution, see a-methyl- arabinoside experiment. 10 min. 24 hrs. 96 hrs. +0.68 +0.64 +0.63 METHYLRIBOSIDE EXPERIMENTS.. In neutral phosphate solution (1 per cent). EXPERIMENT WITH ExtTracr or INTESTINAL Mucosa. IE, 2, tt Enzyme solution, lL ce. Methylriboside solution, I ce. Phosphate solution,,3-ce: Control: Enzyme solution, | ce. Phosphate solution, 3 ce.. 10 min. 24 brs. 48 hrs. 96 hrs. Experiment:: —0.39 cloudy —0.40 —0.40 Control :: 0.00 cloudy 0.00 0.00 EXPERIMENT WITH KIDNEY PLASMA. fl oles i Enzyme solution, 0.5 ce. Methylriboside solution, 1.5 ec. Phosphate solution, 2.5 cc. Control: Enzyme solution, 0.5 cc. Phosphate solution, 4. Occ. 10 min. 48 hrs. 144 hrs. Experiment: —0.50 —0.54 —0.55 Control: —0.06 —0.06 —0.06 EXPERIMENT WITH HEART MuscLeE PLASMA. Ft..2: 741, Enzyme solution, 0.5 ce. Riboside solution, 1 ce. Phosphate solution, 2.5 cc. Control: Enzyme solution, 0.5 ce. Phosphate solution, 3.5 ce. 10 min. 24 brs. 144 hrs. Experiment: —0.34 —0.37 —0.38 Control: —0.02 —0.02 —0.02 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 380 Action of Tissue Extracts on Methylpentosides At the end of the methylriboside experiments, none of the solutions reduced Fehling’s solution. However, after the solution had been boiled with H.SO, the sugar test was positive. d-RIBOSE MUTAROTATION EXPERIMENT. DVe26 ste Ribose solution, 10 per cent., in 2 dm. long observation tube. Temperature 0°C. 15min. 18 min. 25min. 30min. 40 min. 45 min. 60 min. Observer I : —2.73 —2.80 —2.98 Observer II: —2.72 —2.85 —3.01 —3.05 18 hrs. 20 hrs. 24 hrs. Observer II: —3.44 —3.44 —3.86 (Room temperature) STUDIES ON THE ABSORPTION OF METALLIC SALTS BY FISH IN THEIR NATURAL HABITAT. I. ABSORPTION OF COPPER BY FUNDULUS HETEROCLITUS. By GEORGE F. WHITE anp ADRIAN THOMAS. (From the Woods Hole Laboratories of the U. S. Bureau of Fisheries.) (Received for publication, March 15, 1912.) While it is well known that various fish may take up certain substances dissolved in their surrounding medium, the rate and amount of absorption has not been established very definitely for the most diverse metallic salts. Itis our purpose to carry on a complete and comprehensive study of this problem utilizing a large number of salts. The influence of dilution will be noted as well as varying the anion in the different salts studied. Consid- erable work has been done on the concentration of metallic salts and the time necessary to cause death in certain fish, but our method will be to expose the fish to the salts under conditions approximating as closely as possible those in nature, and to test for the presence of the added substance in the fish taken from the medium while still living. A few isolated experiments have been performed on this line, but we believe that the method of our work is sufficiently improved over these, that repetition if any, is not only desirable but necessary, since a thoroughly systematic view of the field is lacking. The work of Sollmann! on ‘The Effect of a Series of Poisons on Adult and Embryonic Funduli’” may be referred to for the rate of poisoning and degree of toxicity of various substances, the pois- ons nicotine and digitaline being found to act most rapidly. Also the study carried on by Loeb? and his co-workers of the inhibition by one salt of the poisonous effect of another on Fundulus, both 1 Amer. Journ. of Physiol., xvi, p. 1, 1906. 2 Biochem. Zeitschr., xxxi, p. 450, 1911; xxxiii, p. 480, 1911. 381 382 Absorption of Copper by Fundulus salts being normally present in sea water, is interesting in connec- tion with our study of Fundulus, although of not direct bearing on our problem. Further, Sumner® has carried out some careful and complete investigations of the osmotic relations between Fun- dulus and other fish and their surrounding medium. The relative toxicity of various poisons in different media was studied, and such substances as cupric chloride were found to be less fatal in sea water of high salt content, or fresh water to which sugar had been added, than water of low salt or sugar content. These results seem to conclusively prove that it was the increase of the osmotic pressure in the surrounding medium by the addition of these substances, which prevented the toxic action of the salt. The fish selected for our experiments was Fundulus heteroclitus, for many reasons. It is small and abounds in the shallow waters around Woods Hole, and is a hardy fish, living practically indefi- nitely in tanks through which sea water is kept running. Of course this latter arrangement would be impossible for experiments of the character described in this paper, but it was found that the following method was successful. Large glass vessels, of about 10 liters capacity, were filled with . sea water containing the desired amount of salt—copper sulphate —and during the whole period while the fundulus were kept in these, a constant stream of air was blown in, furnishing sufficient oxygen for the fish to live for days. The aération was conven- iently accomplished by drawing off the excess air through a tube from a bottle attached to the water outlet of a Richards suction pump, the water being siphoned out of the bottle through another tube. Care was taken not to cause “air-sickness”’ by the use of too strong a blast. By this process, the only disturbing influence, outside of the added poison, was the accumulation of the excre- tory products from the fish, which difficulty was obviated or les- sened by changing the medium about every twelve hours. Fifteen to twenty Fundulus were placed in each of several ves- sels, and after being subjected to the action of the poison for the desired length of time, were taken out while still active and pre- pared for analysis. They were thoroughly washed with fresh water and a stream of water was also passed through the alimen- $ Biol. Bull., x, p. 298, 1906; Amer. Journ. of Physiol., xix, p. 61, 1907. George F. White and Adrian Thomas 383 4 tary tract to remove all traces of copper sulphate not actually absorbed in the body of the fish. They were then cut up in small pieces and analyzed for copper as follows: The fish were dried to constant weight at 110° to 120°. and about 10 grams of the dried flesh taken for analysis. Each sample was placed in a Kjeldahl flask with 80 te 100 cc. conc. sulphuric acid and 5 grams of potassium sul- phate. A small amount of paraffin was added to prevent excessive froth- ing. The solutions were digested until they became clear, this process requiring between ten and eighteen hours. The conditions of digestion were therefore similar to the Gunning method for the analvsis of proteins for nitrogen, in consequence of the presence of the absorbed copper. Oxida- tion of the organic matter by treatment with concentrated nitric acid was previously tried, but found unserviceable due to uncontrollable frothing. The solutions were diluted to 200 cc., a few drops of phenolphthalein solu- tion added, and nearly neutralized with 50 per cent sodium hydroxide solution. Thesolution, measuring about 250cc., was then electrolyzed with a current of 0.5 to 0.75 ampere and a potential difference of 2.5 volts. The current was obtained from two storage batteries set up in series. The electrolysis was allowed to proceed for five hours at which time it was com- plete, trial having shown this to be sufficient. The cathodes were then dried and weighed in the usual manner. In order to prove that there was no other deposit on the electrode than copper, blank tests were made on normal Fundulus, that is Fundulus not having been placed in a copper sulphate solution. Several of these tests were made during the course of the experiments, and absolutely no deposit was found on the cathode in any case. Moisture determinations were made on the Fundulus flesh so that the per- centage of copper absorbed could be calculated either for the original flesh or the dried material, this latter being the most desirable. The results of six such analyses of normal Fundulus are as follows: 77.30; 77.33; 77.49; 76.77; 78.19; 77.12. Average, 77.43 per cent. The results of the poison experiments are given in Table I, the strength of the solutions being referred to normal, and the amount of copper absorbed being expressed in percentage by weight of metallic copper in the dried flesh. A summary of the results is presented in Table II, percentage of copper being calculated for the undried and the dried flesh. It may be seen from the data that very appreciable amounts of copper were taken up by the Fundulus, even although the most concentrated solutions were practically of low copper content. Scaly fish may then absorb poison to a degree which is of the same 384 Absorption of Copper by Fundulus TABLE I. Copper absorbed by Fundulus. EE EEE eel Te | priep Fiesn || NORMALITY ce anes r hours per cent per cen ae 1 0.0210 ( 4 0.0200 | 0.0160 is 0.0190 | 0.0110 ees 6 0.0083 — 2 0.0212 { 0.0086 0.0100 2 0.0080 3 0.0213 0.0070 0.0190 3 0.0114 0.0110 1 | 0.0100 7050 A 0.0080 p =, S200 | 0.0170 2 | 0.0060 ¥ 8 0.0060 am 0.0650 =o 5: | gee 0.0070 | 24 0.00500 | 0.0200 is | 0.00427 4 0.0360 : 0 0100 4000 1 0.00600 48 0.0070 1 0 0040 | 0.0060 0.0030 24 0.00299 2 0.0060 0.00310 T0080 0.0070 0.00400 0.0088 | 0.00300 0.0193 | | 0.00400 0.0080 q 0.00400 order of magnitude as oysters, which Bothe’ has shown have taken up under natural conditions from 0.017 per cent to 0.050 per cent of their body weight in copper, such flesh containing at the same time more water thanthe Fundulus. The greatest absorption with the Fundulus takes place in the first part of the period of exposure to the poison, and there is a gradual increase with length of time until enough has been accumulated to seriously affect the life of the fish. As much absorption may occur in the dilute solutions as in the concentrated if sufficient time is allowed; thus there is as great a percentage of copper in the flesh after an experiment of four hours duration in ros solution as after three hours in a 3, 4 Amer. Food Journ.. vi, p. 2, 1911. George F. White and Adrian Thomas 385 solution, and a larger percentage after four hours in a stn solution. In the extremely dilute solution of scss normality, ninety-six hours is required to produce an accumulation of 0.0040 per cent of copper. TABLE II. Average results of absorption experiments: | TIME CuIN DRIED FLESH | Cu IN UNDRIED FLESH NORMALITY | per cent | 1 0.0160 | 0.00361 ahs 2 0.0156 0.00352 3 0.0201 0.00454 1 0.0100 0.00226 5o0 3 0.0164 0.00360 4 0.0230 0.00529 [ 1 0.0035 0.00079 - 2 0.0065 0.00147 PETS 3 0.0103 0.00232 | 4 0.0195 0.00440 / 2 0.0075 0.00169 2000 3 0.0120 0.00269 | 4 0.0125 0.00280 : {| 24 0.00509 0.00115 pe 48 0.00650 0.00147 24 0.00306 0.00069 3000 48 0.0035 0.00079 | 96 0.00400 0.00090 Since fish take up the considerable amount of copper shown by our experiments, it may be asked in what manner this takes place. To answer this tentatively, a brief study of the distribution of the absorbed salt in the fish was made. Since the Fundulus is so small that it is difficult to dissect and separate its organs, the larger tautog (Tautoga onitis) was selected. This was placed in trios copper sulphate solution for two hours under the same conditions as the Fundulus, and analysis made for copper in the blood system— heart, gills, blood vessels—the alimentary tract—stomach, intes- tines, etc.—and the flesh. The results are given in Table III. The amount of copper taken up by the fish, 0.007 per cent of its total body weight (dry), is practically identical with the result for Fundulus, 0.0065 per cent, obtained in the same dilution and for 386 Absorption of Copper by Fundulus TABLE ITI. Distribution of copper absorbed by Tautog in 1,, solution. | PER CENT COPPER IN DRY MATERIAL MATERIAL fs — - I > a Whole fish. 3 eetecacy =: 0.008 Blood system! =>: - 0.010 Alimentary tract.......... 0.003 Flésh:<2heeee:. . «| 0.009 Residue---eee ee... | the same time period. Therefore we may assume that the two species do not act materially differently towards the poison. The ~ results are quite positive in their character. The largest percen- tage of copper was found in the blood system and it is therefore reasonable to conclude that it is through the gills, where the sur- rounding medium comes in most intimate contact with the blood, that the absorption is the greatest. A final and absolute state- ment of this we do not put forward, since this one experiment is not sufficient to firmly establish any theory. But nevertheless it is very suggestive. Taken in connection with the work of Scott and White® on the permeability to salts of the gill membranes of a fish, it receives some support. It is interesting to note that we found visible evidence of the copper in the fish, the tautog especially showing the green color caused by the reaction between the copper sulphate and the pro- tein substance. Microscopic sections’ of the whole Fundulus (cross-section), and even of the brain and spinal cord, treated with potassium ferrocyanide, were colored markedly brown by the formation of copper ferrocyanide, whereas normal Fundulus evidenced no such change. Work on the problem of absorption of salts will be continued. ’ Science, xxxil, p. 768, 1910. ® The tissues, preserved in alcohol, were embedded in paraffin, cut with the microtome, the paraffin dissolved in turpentine which latter was removed by alcohol. After washing out the alcohol with water to prevent precipi- tation of K,Fe(CN)., the sections were washed in K,Fe(CN),., washed with water and alcohol, and mounted in balsam. THE DETERMINATION OF ALUMINUM IN FECES. By CARL L. A. SCHMIDT anv D. R. HOAGLAND. (From the Laboratories of the Referee Board of Consulting Scientific Experis at the University of Pennsylvania.) (Received for publication, March 16, 1912.) On looking about for a satisfactory method for determining aluminum in the feces of experimental subjects who were given either alum or the aluminum-containing residue from baking powder, we encountered a number of difficulties. The determina- tion of aluminum as the hydroxide was impossible on account of the presence of phosphates. The determination of aluminum by precipitation both of iron and aluminum as phosphates and then determining the iron and the phosphorus, determining the alumi- num by difference, is too long and besides, the value for aluminum so obtained may include the combined errors of the other two determinations. In attempting to use the Peter’s method,’ in which iron is reduced by ammonium thiosulphate and the alumi- num determined as the phosphate, we encountered a number of dif- ficulties; yet this method gives a direct determination of aluminum, so the attempt was made to find out the conditions under which it could be used. Substances present in the feces which affect the determination of aluminum by this method are: Organic matter, silica, tin (from canned foods), iron, calcium and phosphates. Organic matter can be removed by mixing with the feces several cubic centimeters of concentrated sulphuric acid and ashing in a silica dish. All of the aluminum in the ash is not soluble in hydro- chloric acid. It is necessary to dissolve out the acid-soluble part and then fuse the insoluble residue with sodium carbonate.? The silica is dehydrated and the solution added to the main portion. Tin can be removed by precipitation as the sulphide in an acid 1 Circular 26, U. S. Bureau of Standards. 2 Ibid., p. 6. 387 388 The Determination of Aluminum in Feces solution. Complete removal of the iron and probably calcium also, is not accomplished in the first precipitation of the aluminum. On redissolving the precipitate in hydrochloric acid and again precipitating the aluminum as the phosphate, only a negligible trace of iron remains in the precipitate. A large excess of am- monium phosphate should be avoided since it is difficult to wash out and the time of heating the precipitate to constant weight is thereby greatly increased. The latter process necessitates a very high temperature. A small precipitate may be brought to a constant weight by heating over a blast lamp for an hour, but for precipitates of 150 mg. or more, a higher temperature or a greatly prolonged heating is neces- sary. We have found that on heating precipitates in a Méker muffle furnace at a dull white heat for an hour to an hour and a half, precipitates as high as 300 mg. can be brought to constant weight. Porcelain crucibles cannot be used. Platinum, while re- maining constant in weight, soon crystallizes and is rendered worth- less. A glazed silica crucible will remain constant in weight for several determinations, but on prolonged heating will lose weight. A transparent silica crucible will remain constant somewhat longer. The empty crucible should be cleaned and weighed after being used, to make sure that there has been no loss in weight. The details for carrying out a determination of aluminum in feces are as follows: Five to ten grams of feces are treate: with several cubic centimeters of concentrated sulphuric acid and ashed in a silica dish. The soluble aluminum is dissolved out by warm- ing with dilute hydrochloric acid. The residue on the filter paper is washed and then ignited and fused with sodium carbonate in a platinum crucible. The melt is dissolved out with dilute hydro- chloric acid, the silica dehydrated, and the whole added to the main portion containing the aluminum. The volume at this point should be about 300 ce., and contain about 2.5 ec. of con- centrated hydrochloric acid. Tin is precipitated from the hot solution by hydrogen sulphide and filtered off. Di-ammonium hydrogen phosphate is added to the solution—0.5 gm. for each 100 mg. of aluminum phosphate present. The solution is heated, and while hot 5 grams of ammonium thiosulphate (in solution) and after several minutes 6 to 8 grams of ammonium acetate (in solution) and 4 cc. strong acetic acid are added. Heating is Carl L. A. Schmidt and D. R. Hoagland 389 continued for about half an hour to expell SO, the precipitate allowed to settle, filtered and washed once by decantation. The precipitate is redissolved in 2 to 2.5 cc. of concentrated hydrochloric acid, the solution diluted to about 300 cc., 0.5 gram of ammonium phosphate added for each 100 mg. of aluminum phosphate present and the aluminum again precipitated as described above. The precipitate is filtered and washed several times with hot water to remove chlorides and ignited in a transparent silica crucible until constant weight is reached to remove excess of P,O;. The precipitate obtained in this manner is easily filtered and washed. The free sulphur present serves to make the precipitate more flocculent and more easily filtered. By using ammonium salts throughout as precipitating reagents the necessity for very thorough washing of the precipitate is eliminated, since small amounts of such salts remaining in the precipitate are volatilized on ignition. Washing the precipitate with ammonium nitrate has been recommended, but we have found such procedure unneces- sary. The minimum amount of hydrochloric acid necessary to keep the aluminum in solution before precipitation, should be used, thus making it easier to reduce its concentration in the aluminum precipitation. The method carried out as above gives good check results. Using this method with various amounts of aluminum in feces and in solutions containing known amounts of aluminum we ob- tained results as follows: (1) Solution of pure Al Cl; (single precipitation). | FOUND CALCULATED aia — —— AIPO, Equivalent Al:O3 Theoretical AlsOs ( 0.1721 0.0720 0.0720 (a).. SaaS Fa 0.1723 | 0.0721 (| 0.1718 0.0719 ie { 0.0106 0.0044 : 0.0106 0.0044 (2) A known volume of AICI; was added to 50 cc. of a mixture containing per liter the following salts: 390 The Determination of Aluminum in Feces Grams Sinditim mhornbates <2) 5 +... - << tne lee ee ee 25 PGTANBTITIGHIGTIOG: 5 5 6 o.6.<.+0.6% << cet See eee eee eee 15 NGERICTCINOTIO Eco. co oss soc 3 nue.s DbletE ee en ee eee 5 ORIENT CHIOLICC* Ss] cot... oe ee eee eee 11 Maanramnpchloride... ..)..0) 05 OBR OR CU 11 ; a ae [oe FOUND CALCULATED AIPO, Equivalent Al.O: | Theoretical AlsOs 0.1677 | 0.0702 0.1684 0.0705 0.0707 0.1686 0.0706 | (3) Duplicate samples of dried feces of men who were given alum gave the following results. AIPO, Equiva- AIPO, lent Al. (Aieks stents: {0.1799 0.0399 f 0.0915 * sia | 0.18 0. 0405 ') Heneneie a 1) 0.1942 0.0431 (f) 0.0333 awe 0s So ee } 0.0348 (ele ee 0.1848 0.0410 (g) 0.2143 } 0,1846 0.0410 SAT US TGS ae \ 0.2143 (Ap 2iiiil? egtene: 0.2141 0.0475 \.0.2120 0.0471 Equiva- lent Al. 0.0203 0.0200 0.0074 0.0077 0.0476 0.0476 (4) Determinations of aluminum in feces of different subjects on a con- stant diet, but who were not given any aluminum salt, gave the following results, as total amounts of aluminum excreted in a period of two weeks. Subject A Gram Al PiTSt CWORVECKS os oo ee te ee a see 0.040 SeCOnGeGwOMWCeKs.: sos ciscicen. eee SRR kate ean See 0.053 ACHITONU WOLWEERS Sc. cos alte Sets oR As Oe, ee Es 0.063 HNourihsiwOaweekss =. )4. Uber eee Ce ee eee 0.054 Bibb twonweeksac... } 1 Wi iis Ure > ul , a Mary : wilfe uri os i { a | ‘sri he woh a : | pe ' CHTaL at he hoa ey 5 or a aig tn) ‘2 Jud | Oat -43 Lo JTieie a re i ~~. ot t \ ———— : a - % Px =. i y i » i r a, r Te 4 p a a a t D sr nes, ee er a ‘ ilies ee. cee bem ser Tiere A ¢ : => ; DORR Ga Pat heatde =" a alt Fh le a iavouia" Tos ri ry f im R Os Aes) 8 OA eS, ) 7 : =P pe ee ees +¥ RESEARCHES ON PURINES. ON 2,8-DIOXY-6,9-DIMETHYLPURINE AND 2,8-DIOXY-1-METHYL- PURINE. SIXTH PAPER.! By CARL O. JOHNS. (From the Sheffield Laboratory of Yale University.) (Received for publication, March 19, 1912.) Although the dioxy-dimethyl-purines are of considerable interest owing to the fact that they are isomeric with theobromine,? 2,6-dioxy-3,7-dimethylpurine (IX), yet very few of the many possible isomers have been described. Of the nine isomerides of 2,8-dioxy-dimethylpurine only one member has been described, namely, 2,8-dioxy-3,7-dimethylpurine (XII). This compound was obtained by Emil Fischer* who chlorinated 3,7-dimethyluric acid (X) and reduced the resulting chloride (XI) with hydriodic acid. In a previous contribution* from this laboratory it was shown that orthodiaminopyrimidines, in which a hydrogen atom of an amino group has been substituted by an alkyl group, condense readily with formic acid or urea to form purines. This method has now been applied in the synthesis of 2,8-dioxy-6,9-dimethyl- purine (IV). 2-Ethylmercapto-4-methyl-6-chlorpyrimidine® (I) was heated in a sealed tube with methylamine and the result was a quantita- tive yield of 2-ethylmercapto-4-methyl-6-methylaminopyrimidine ‘(II). This, in turn, was converted to 2-oxy-4-methyl-6-methyl- aminopyrimidine (III), which, when nitrated, gave 2-oxy-4-methyl- 1 This Journal, xi, p. 73, 1912. 2 Beilstein’s Handb., iii, p. 954. 3 Ber. d. deutsch. chem. Gesellsch., xxviii, p. 2487, 1895; xxx, p. 1851, 1897; XXXli, p. 474, 1899. 4 Johns: This Journal, ix, p. 161, 1911. 5 Amer. Chem. Journ., xl, p. 351, 1908. 393 394 Researches on Purines 5-nitro-6-methylaminopyrimidine (VI). The yields were satisfac- tory. When the nitro-compound was reduced with freshly pre- cipitated ferrous hydroxide an 83 per cent yield of 2-oxy-4-methyl- 5-amino-6-methylaminopyrimidine (V) was obtained. By heating with urea, the diamino-compound was easily converted to 2,8-dioxy— 6,9-dimethylpurine (IV). This paper also contains a description of the synthesis of 2,8- dioxy-1-methylpurine (VIII), which compound was prepared by heating urea with 2-oxy-3-methyl-5,6-diaminopyrimidine® (VII). As three of the isomers of 2,8-dioxy-monomethylpurine have been described previously,’ the only member of this series which is still unknown is 2,8-dioxy-7-methylpurine. These researches will be continued. N=CCl N=CNHCH; N==CNHCH; | oa . — one a — > 0c oa | | | N—C-CH; N —C-CH HN —C:CHs I II Il | N= C-CH3 vy = CNHCH3 N= CNHCH3 | ee es oc. C—-NH +— OC CNH. <— OC -ENO; cae | | | | Co | Pei ro HN C——Ni- CH: HN —C.-CH3 HN —C-CH3 IV V VI N=CNH> CH;-N—CH HN—CO | | | | | OC foo —_——> OC THON OC i ee I | oat | | > | co CH wrt hi | aioe CH3:-N —CH' N ~=C—NH CH;3-N—C—N VII VIll IX § Johns: This Journal, xi, p. 77, 1912. 7 Fischer and Ach: Ber. d. deutsch. chem. Gesellsch., xxxii, p. 2736, 1899; Johns: This Journal, ix, p. 63, 1909; Amer. Chem. Journ., xli, p. 63, 1909. Carl O. Johns 395 HN—CO N+>CCl N ==CH & loi pe 4G NCH, '—— OC: C—NGHe => 0G. C-—-N°-CHs ee) co co | 4 fe Woe whe CH;-N—C—NH CH;-N—C—NH CH,-N—C—NH x xI XII EXPERIMENTAL PART. 2-Ethylmercapto-4-methyl-6-methylaminopyrimidine. N==CNHCH; C:H;SC .CH | N=—C-Cn. Fifteen grams of 2-ethylmercapto-4-methyl-6-chlorpyrimidine® were mixed with 21 cc. of a 33 per cent aqueous solution of methyl- amine and 30 cc. of water and this mixture was heated in a sealed tube at 100°C. over night. A heavy transparent oil formed and this solidified to a white crystalline mass on cooling. The reaction product, thus obtained, was easily soluble in cold ether, benzene or alcohol but it was almost insoluble in hot water. It dissolved readily in cold concentrated hydrochloric acid. When crystallized from dilute alcohol it formed beautiful, flat, anhydrous prisms that melted to an oil at 87°C. The yield was quantitative. Calculated for 7 CsHisNsS: Found: Ni] beer RARER e eres 22.95 23.03 2-Oxy-4-methyl-6-methylaminopyrimidine. N= CNHCH: eee OC CH | i HN — C-CHs Twenty grams of 2-ethylmercapto-4-methyl-6-methylamino- pyrimidine were dissolved in 200 cc. of concentrated hydrochloric 8 Johns: Loe. cit. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4. 396 Researches on Purines acid and the solution was evaporated to dryness on the steam- bath. The residue was then found free from sulphur. It was dissolved in hot water and the solution was made slightly alkaline with ammonia whereupon crystals formed rapidly. These were easily soluble in cold acetic acid and slightly soluble in hot alcohol. They were moderately soluble in hot water from which solvent they separated in compact, biconcave, anhydrous blocks. These turned brown at about 290°C. and decomposed slowly above that temperature. The yield was 83 per cent of the calculated. Calculated for CsHsONs: Found: IN EEN SS on. n ic oie oto CREE: SPER Cee 30.21 30.00 2-Oxy-4-methyl-5-nitro-6-methylaminopyrimidine. N=CNHCH; bea OC CNOz Lavaeal HN — C-CH; Twelve grams of 2-oxy-4-methyl-6-methylaminopyrimidine were dissolved in 30 cc. of concentrated sulphuric acid. While this solution was kept cool, 4.8 cc. of nitric acid (density 1.5) were added gradually. After five minutes the resulting solution was poured on crushed ice and the acids were neutralized with ammonia. A crystalline precipitate separated at once. The yield was quanti- tative. The nitro-compound was practically insoluble in hot alcohol and but slightly soluble in hot water. It dissolved moder- ately in glacial acetic acid. By cooling this solution slowly small, stout prisms were obtained,while by cooling rapidly slender, pointed crystals were formed. The crystals turned dark at about 250°C. and decomposed slowly when heated above that temperature. Calculated for CeHsO3Na: Found: Nae EPI eee isi» » a steiels, GRE nS een 30.43 30.12 Carl O. Johns 397 2-Oxy-4-methyl-5-amino-6-methylaminopyrimidine. N= CNHCH; ee OC CNH, | HN — C-CH; Fourteen grams of 2-oxy-4-methyl-5-nitro-6-methylaminopyri- midine were dissolved in a mixture of 200 cc. of concentrated ammonia and an equal volume of water. A concentrated aqueous solution containing 174 grams of crystallized ferrous sulphate was added. Ferric oxide was precipitated rapidly, the reaction being exothermic. The sulphate was precipitated by the addition of a solution of 203 grams of crystallized barium hydroxide and the excess of baryta was removed by means of ammonium carbonate. After shaking a few times, the mixture was set aside over night, after which it was heated to about 80°C. and filtered by suction, the precipitate being washed with hot water. On evaporating the filtrate to dryness a crystalline crust was left. This was dissolved in hot water and the color was removed with blood coal and, on cooling, a bulky mass of colorless, anhydrous needles was obtained. These charred rapidly above 270°C. They were soluble in less than ten parts of boiling water, easily soluble in cold glacial acetic acid and slightly soluble in hot alcohol. The yield was 83 per cent of the calculated. Calculated for CeHiONa: Found: IN PONT looc coclore & erere oq nis @ pare eie ene 36.36 36.53 2,8-Dioxry-6 ,9-dimethylpurine. N= C- CH3 gene Ore C—NE [ths SECO HN —G—NCH; Two grams or urea and an equal weight of 2-oxy-4-methyl-5- amino-6-methylaminopyrimidine were ground together and the mixture was heated in an oil-bath at 180 to 190°C. for an hour. The mass first melted but soon solidified. The reaction-product 398 Researches on Purines was dissolved in dilute ammonia and decolorized with blood coal. After boiling off most of the ammonia, the solution was acidified with acetic acid. Compact prisms were formed by cooling the solution slowly, but by cooling rapidly a bulky mass of needles was obtained. The yield was quantitative. The purine dissolved in about sixty parts of boiling water but was difficultly soluble in cold water. It was almost insoluble in boiling alcohol. It dis- solved readily in dilute alkalies. It possessed the usual stability of dioxypurines and did not melt at 320°C. The crystals con- tained 2 molecules of water. 1.193 grams of substance lost 0.1965 gram at 135 C. Calculated for C7HsO2Ni.2H20: Found: 1S INO ee (tan eet SON Te OO ek SPCR, NAA Merde 16.66 16.47 0.2204 gram of anhydrous substance gave 0.0891 gram of H2O and 0.3772 gram of CQ. Calculated for C7Hs02Ni: Found: Corer es: ot bal Aas Saree are besos 46.66 46.66 13M 2S 2 0 EE Dn che hy ie ae ae tatinlrr mtntton 4.44 4.49 ING oss rete ir. 2s lean Se ELE Pen ease eee 31:11 31.14 This purine did not form a picrate. It could be crystallized from an aqueous solution of picric acid and the crystals formed were those of the free base. When crystallized from 20 per cent hydro- chloric acid it gave a hydrochloride that was easily soluble in hydrochloric acid and that hydrolyzed readily in water. When warmed with 30 per cent nitric acid it was oxidized and on evapor- ating the solution a red residue was obtained. This turned a brilliant purple when moistened with ammonia. 2,8-Dioxy-1-methylpurine. CH3-N— CH OC, C—_-NEH joo N=C—NH Carl O. Johns 399 A mixture of pulverized urea and an equal weight of 2-oxy-3- methyl-5,6-diaminopyrimidine? was heated in an oil-bath at 170 to 180°C. for an hour. The reaction product was dissolved in hot dilute ammonia and the solution was decolorized with blood coal. On acidifying with acetic acid the purine crystallized from the hot solution in the form of small, anhydrous plates. These did not melt at 320°C. The yield was almost quantitative. The purine was difficultly soluble in cold water andalmostinsoluble in boiling alcohol. One part of the purine dissolved in about 200 parts of boiling water. It was soluble in concentrated acids but the salts dissociated in water. 0.1782 gram of substance gave 0.2852 gram of CO: and 0.0610 gram of H;O. Calculated for 6HeO2Na: Found: Obs e oyctogiies a ce 00% oii tt ee Seas 43.37 43.64 1p II. MIE ri SSC tle ce ctc eo oaee ene 33.73 33.77 33.60 ® Johns: Loc. cit. Se va ae oe Latah = i 4 am’ eve hi anit * "a Se ~ cee.) ae Faye. || Gy ‘saa te [OR c, x tr y 4 igen 7 a We at is i ae ke aa aa bavi SOME “(ie ‘china timeacahe ar hai .! i anmes Hipp wore Tve aa aa ! ae: % “rdbpiniolt Kaaatt age ee Mipligg Seer) | pele OF Sduleed vi > BAO Gea Iq Seer s0OP 49h hy: ued iy Loe an + tole uctind f : in a settee 7 . i ¢ 7 Pais BR AG Mii aN ab KL) ne, CT arity Coes Oa 1 Or Sa % z ae - = he poe bichdbce gave a. > » v na 5a 2 eae r \. a T 4 ¢ i 1 . : : : ‘ ha a % TRC 8, POPS I, " es Ny ri Seb. a 3 a 4 re Bern "= ON GA POTTS CR TS oa , Ve Fy. E é ty ba dy? frtets brit wey 7S a ee nee ti 2ir “wil Witte © Fak wee ‘eee THE CHEMICAL ANALYSIS OF THE ASH OF SMOOTH MUSCLE. By EDWARD B. MEIGS anp L. A. RYAN. (From the Wistar Institute of Anatomy and Biology and the Hare Chemical Laboratory of the University of Pennsylvania.) (Received for publication, March 22, 1912.) The salts dissolved in the tissue fluids of animals and plants play an important physiological réle, and a large amount of work has been done in the last twenty years with the object of discove:- ing the nature of these salts and the manner of their combination in different tissues. The usual method of dealing with these prob- lems has been to make ash analyses; the blood, lymph, and striated muscle of many animals have been analyzed in this manner. Very few analyses, however, have been made of the ash of smooth muscle. Kiihne states that it is richer in sodium than in potas- sium,! and this statement has been quoted in later text-books and in scientific articles. But Kiihne gives only the bare statement quoted above with no reference and no account of the work on which it is based. Neumeister, on the other hand, says, “Die glatten Muskeln zeigen in ihrem chemischen Verhalten von den ‘quergestreiften kaum Abweichungen,” and the context makes it appear that this statement refers, among other things, to ash analy- sis.2— But here also there is no reference and no further account of work on which the statement might be based. Halliburton makes a similar though somewhat vaguer unsupported statement.’ Finally, Macallum in a recently published general article says, “Analysen des Natriums und Kaliums im glatten Muskel zeigen, 1 Kiihne: Lehrbuch der physiologischen Chemie, Leipzig, p. 333, 1868. 2 Neumeister: Lehrbuch der physiologischen Chemie, 2 Aufl., Jena, p. 442, 1897. 3 Halliburton: Tezt-book of Chemical Physiology and Pathology, London, p. 398, 1891. 401 402 Ash of Smooth Muscle dass letzteres reichlicher als ersteres vorhanden ist, obgleich das Ueberwiegen nicht so gross ist, wie in der quergestreiften Faser.”’ In a footnote he says that this statement is based on unpublished data of his own.‘ We are familiar with two accounts of more or less complete analyses of the ash of smooth muscle. The first of these is by Saiki® who worked on the stomach and bladder muscle of the pig and found the ash of these tissues widely different from that of the striated muscle of the same animal. Saiki finds considerably more sodium than potassium in the pig’s smooth muscle, and much less phosphorus and sulphur and more chlorine than earlier investi- gators have found in the striated muscle of the same animal. According to Saiki, therefore, the ash of smooth muscle is a good deal more like that of the blood plasma than is that of striated muscle. But Costantino® has recently analyzed the smooth muscle of various animals for sodium, potassium, and chlorine, and finds in general that the potassium is much higher than the sodium, though the difference is usually less marked than it is in striated muscle. Costantino has analyzed the retractor penis of the pig for chlorine and gets a figure somewhat lower than Saiki’s. Saiki’s figures are compared with those of Costantino in Table I, and Katz’s’ figures for the striated muscle of the pig and ox are given in the same table. There are various possible explanations for the differences between Saiki’s results, on the one hand, and those of Costantino and Macallum on the other. According to the figures given by Katz the pig is a rather éxceptional animal in regard to the sodium and potassium content of its striated muscle. Katz analyzed the ash of striated.muscle in the human being, pig, steer, calf, stag, rabbit, dog, cat, chicken, frog, haddock, eel and pike. In the pig there is less than twice as much potassium as sodium, while in the other animals there is from three to fourteen times as much. It may be that pig’s smooth muscle is even more different from that of other animals in this respect than is its striated muscle. It *Macallum: Ergeb. d. Physiol., p. 642, 1911. 5 Saiki: This Journal, iv, p. 483, 1908. 6 Costantino: Biochem. Zeitschr., xxxvii, p. 52, 1911. 7 Katz: Arch. f. d. ges. Physiol., lxiii, p. 1, 1896. 403 08°SZ L8T'0 06° FG 86°62 IT 22 $02 0 8400 Z90°0 9°62 | ¥02 180'0 Ik oF UGdLvy mM sdari1os Edward B. Meigs and L. A. Ryan oR ale Se Seah oe ele ‘ajasnus yjyooug T 2Ta&VL 800°0 680°0 €Fo 0 zai a L960 c9¢ 0 290°0 = ouly © eu pelat 9, ere srued . 3 -ue4soy IOPIBIJAI /XO ‘*YoRuloys :XQ ecw eet eee tere eee aid au) jO Jappe]q pus youul04s ay} IO} Sodvs9aeV 8 1yIRS BN Ds | | ‘anssy, ysaif ay) fo sabpjuaosad sp waarb ajasnut paznr4js puy yjoows fo syuanqysuoy 404 Ash of Smooth Muscle seems strange, however, that the smooth and striated muscle of the pig should be so much more markedly different in sodium and potassium content than are the two kinds of muscle in other ani- mals, and that the magnesium, phosphorus, and sulphur content of the pig’s smooth muscle should be so low. Katz did not find the pig’s striated muscle exceptional in respect to its content of these three elements. If one studies the sodium and potassium figures which Saiki gives for his individual analyses one finds a rather wide variation; the sodium varies from 0.2 per cent to a little over 0.3 per cent and the potassium all the way from 0.039 per cent to 0.081 per cent. In the text of Saiki’s article (p. 492) is the statement, ‘‘A comparison of the composition of fresh pig’s muscle of different types with blood serum (resembling lymph) of the same species, indicates that the assumption of an admixture of lymph may explain the higher content of sodium and chlorine and the lower percentages of potassium, magnesium and phos- phorus.” This statement taken in connection with the rather wide variation in Saiki’s figures for sodium and potassium brings up the question how much of the tissue analyzed by him was really smooth muscle. Another fact that must be taken into account is that Saiki extracted his tissue with ether before ashing it, while Costantino did not. It is quite possible that preliminary extraction with ether would lower the phosphorus content in the ash of a tissue.® We have made analyses of the ash of frog’s stomach muscle. We have chosen the frog as the object in our investigation, partly because the tissues of this animal have become the standard in all sorts of physiological experimentation, and partly because it is easy to obtain from the frog’s stomach a tissue which is undoubt- edly 90 per cent irritable smooth muscle at the time the chemical investigation is begun. We have used the large American bull- frog (Rana Catesbiana); the stomach of a single individual of this species sometimes yields more than 2 grams of smooth muscle. Our muscle was prepared as follows: From six to eighteen frogs were killed and the stomachs were dissected out. The stomachs were never left in the dead frogs for more than three hours, and usually for a much shorter period. The muscle was always still irritable at the time the chemical examination 8 See Katz: Loc. cit. pp. 9 and 10. Edward B. Meigs and L. A. Ryan 405 was begun. The stomachs were laid on a glass plate and the muscular coats were cut open along the line of the lesser curvature, while the: mucous mem- brane was left as a still unopened tube. The mucous tube was then torn loose from the muscle, and in this way contamination of the muscle with stomach contents was avoided. After the separation of the mucous and muscular layers of the stomach, the small amount of submucous connective tissue which usually adheres to the inner surface of the muscle was stripped off. The sheets of muscle so obtained were pressed several times against hardened filter paper, weighed, and then either fused, or dried and inciner- ated, according to the nature of the analysis that was to be carried out. If such sheets of muscle as we used for analysis are fixed and examined his- tologically, it will be found that they contain from 90 to 95 per cent of smooth muscle and from 5 to 10 per cent of serous connective tissue. The same relations are found if one examines slices of the fresh tissue, and it is probable, therefore, that the ‘‘smooth muscle”’ which we used for analysis contained less connective tissue than the ‘‘striated muscle’’ used by Katz and the other investigators of the inorganic constituents of striated muscle. Saiki suggests (p. 492) that smooth muscle may contain larger lymph spaces than striated muscle. We have made a careful investigation of this question. Samples of muscle were fixed in various ways and embedded in parafin. From these, thin transverse sections were cut and stained by methods which are supposed to be specific for smooth muscle. From other samples cross sections were sliced off free hand with a sharp.razor and examined microscopically in Ringer’s solution. Such sections were often stimulated by an electric current beneath the microscope, and the muscle in them usually contracted, showing that they were still alive. In both fixed and living preparations it was found that the muscle fibers occupied from 80 to 90 per cent of the total volume of the preparation; and the interstitial spaces, from 10 to 20 per cent. The lymph spaces between the fibers are therefore smallerin the smooth muscle of the frog’s stomach than they usually are instriated muscle. We are prepared to assert that at least 75 per cent of the total volume of our tissue was occupied by the muscle fibers, and we should judge that the average was nearer to 85 per cent. We did not extract our dried tissue with ether before ashing it for various reasons. It is possible that the ash of tissue extracted with perfectly dry ether more nearly represents the inorganic salt content of the tissue than in the case where the ether extraction is omitted. It would be very interesting to determine the differences between the ashes of extracted and unextracted tissue and we hope to take up this question lateron. We think it is impor- tant, however, as a preliminary step to have total ash determinations for smooth muscle, which should be comparable, as far as possible, to the ash determinations which have already been made for the striated muscle of the same animals. In making our determinations we used in general the methods employed 406 Ash of Smooth Muscle by Katz,® and we carried along with each pair of determinations on smooth muscle a similar pair on striated muscle. Our methods were exactly similar to those of Katz in the determinations of potassium, sodium, iron, calcium, and magnesium. We determined the phosphorus in both striated and smooth muscle as Katz did in three portions; extracting our tissue first with boiling water, and then with 95 per cent alcohol in the Soxhlet apparatus. The phosphorus was determined in the water extract, in the alcoholic extract, and in the residue which had been extracted with both water and alcohol. In the stri- ated muscle we found, as Katz did, that a little over 80 per cent of the phos- phorus appeared in the water extract. .In the case of smooth muscle, on the other hand, the phosphorus which appeared in the water extract was a little under 70 per cent of the total: our figures for smooth muscle are given in the experimental protocols. We have not put these figures in our tables, because we think (as, indeed, Katz admits) that they are far from correct quantitatively. Our water extracts from smooth muscle were quite opales- cent, and left, on evaporation, considerable quantities of greasy material which might well have been lipoid. The results indicate that smooth muscle contains a good deal more non-diffusible phosphorus than striated muscle, but they require further elucidation. - In analyzing for phosphorus we fused our residues with sodium hydroxide and potassium nitrate in silver dishes instead of ashing them in platinum crucibles as Katz did. ; In the case of chlorine a preliminary determination was made by the method employed by Katz; in this determination the chlorine was found to amount to 0.0988 per cent of the fresh tissue. We were not entirely satisfied with this determination and made two others, in which the fresh tissue was fused with sodium hydroxide and potassium nitrate and analyzed for chlo- rine by the Volhard-Arnold method described by Hawk.!° In these two experiments the chlorine was found to be 0.1200 per cent and 0.1191 per cent of the weight of the fresh tissue respectively. The sulphur was determined by the method described by Hawk on pp. 381-383 of the work just quoted. The tissue was fused with sodium hydrox- ide and potassium nitrate in a silver dish over an alcohol flame, and the sulphates were subsequently precipitated by means of barium chloride. Striated muscle analyzed by these methods for phosphorus, chlorine, and sulphur gave results not far from those obtained by Katz (see tables). Our results are given in Table II. ® Katz: Arch. f. d. ges. Physiol., |xiii , p. 1, 1896. 10 Hawk: Practical Physiological Chemistry, 3d Edition, Philadelphia pp. 390 and 391, 1910. 407 award B. Meigs and L. A. Ryan “(punrgsajn DuDy fo ajosnue yoouo}s) aasnue Ypoowss Z9 18 | 3 ‘ ; ; ; . ; osoe 0 | cose pQugNosy VUBY Joy sosvi9aAy §,248o | geco'0 | cose 0 | sosVIIAY | BUBIqsSeze,) goso'0 | siss'0 | Il; Buey 10) ZcolOe|esve 0 |. = = I sein3sy INQ ‘(pjuanasy Duby PUD puvigsayDg DUDY) aJ9SnU PI}DUiS' Sac ieee Pore, Feceht rr Fea! ‘anssyy ysaif oy? fo sabpyuaoiad sp waarb ajasnu s bosf fo sjuanjysuo,) ‘Tl AIAVL 408 Ash of Smooth Muscle THE RELATION OF OUR RESULTS TO THOSE OF SAIKI, MACALLUM AND COSTANTINO. We agree with Macallum and Costantino and differ from Saiki in finding the sodium and potassium content of smooth muscle not widely different from that of striated muscle. We find in general that the ash of smooth muscle is much more nearly like that of striated muscle than Saiki’s figures would indicate; our figures for potassium, magnesium, phosphorus, and sulphur are much higher; and our figures for sodium, calcium, and chlorine, much lower than his. Saiki lays some emphasis on the large amount of calcium which he found in smooth muscle, and suggests (pp. 492 and 493) that this element may have some connection with the well-known power of smooth muscle to remain for a long cime in a state of marked tonic contraction. Our figure for cal- cium is only about one-eighth of that of Saiki, and only about one- fourth of what Katz found for the striated muscle of the frog. We are not inclined to attribute very much significance to the percent- age of calcium found in either striated or smooth muscle. In Katz’s results this element is by far the most variable of all, ranging from 0.002 per cent in the steer to 0.018 per cent in the rabbit among mammals; and rising to nearly 0.04 per cent in the eel and pike. Saiki’s figures seem to show that the amount of calcium varies widely even in the smooth muscle of the same animal; they range from 0.022 per cent to 0.042 per cent. Our figure for iron is very low, which indicates that our tissue contained very little blood. We were careful to dissect off the larger blood vessels from the outside of the stomach, and our method of pressing our sheets of muscle against filter paper proba- bly freed them from blood quite effectually. The gross and micro- scopic appearances of our tissue agreed in indicating that there was very little blood in it. Our results show that smooth muscle contains about twice as much chlorine as striated muscle, and in this we are in general agreement with both Saiki and Costantino. And we agree with Saiki, Costantino, and Macallum in finding more sodium in smooth muscle than in the striated muscle of the same animal, though we do not think the difference is so marked as Saiki’s figures would indicate. Histological examination has convinced us that our Edward B. Meigs and L. A. Ryan 409 smooth muscle contained a larger proportion of muscle fiber than do most samples of striated muscle; we therefore think that all the work done so far on smooth muscle shows that the fibers of this tissue contain considerably more sodium and chlorine than do those of striated muscle. It is interesting to note that the amounts of sodium and chlorine which we find in smooth muscle are about such as unite to form sodium chloride. GENERAL DISCUSSION. There is reason to believe that a large proportion of the sodium and chlorine of the blood plasma and lymph, and of the potassium and phosphorus of the striated muscle fibers exists as diffusible salt; and it is rather generally supposed that semi-permeable membranes surround the muscle fibers and prevent the inter-diffu- sion of the sodium chloride and potassium phosphate. Overton" has collected a good deal of evidence which tends to prove that the muscle fibers are surrounded by such semi-permeable membranes; he believes that such surrounding membranes constitute a very general peculiarity of both animal and vegetable cells. On the hypothesis that such membranes exist there have already been founded theories of stimulation and of anesthesia, and it is alto- gether probable that many of the nutritional processes are con- trolled by the nature of the bounding surfaces between cells and the surrounding lymph. Overton’s work makes it seem probable that the fibers of stri- ated muscle are surrounded by membranes permeable to water and to fat solvents, and impermeable to sugars and inorganic salts. But the hypothesis that such membranes exist is, as Overton acknowledges, far from explaining all the known facts. The question arises, how growing muscle fibers get their supply of potassium phosphate. The concentration of potassium and phos- - phorus in the blood plasma and lymph is very low, and salts do not diffuse from a region of lower to one of higher concentration. The difficulty can be overcome only by supposing that the potas- sium and phosphorus enter the muscle fibers not as potassium phos- phate but in some other probably organic combination or com- binations. 1 Overton: Arch. f. ges. Physiol., xcii, pp. 115 and 346, 1902; ev, p. 207, 1904. 410 Ash of Smooth Muscle It is highly improbable that this process occurs only in growing muscle fibers. The existence of a membrane absolutely imper- meable to salts is hardly conceivable, and it is not difficult to find experimental evidence for the view that potassium phosphate escapes from the muscle fibers quite readily. The work of Urano! and of Fahr™ shows that the slightest injury to an excised muscle causes a large escape of potassium and phosphorus from its fibers into a surrounding isotonic sugar solution. Considerable quanti- ties of these elements may be lost without impairing the muscle’s irritability. Further, Klug and Olsavsky™“ have shown that mus- cular work causes an increase in the phosphorus excreted in the urine. It is, therefore, not a tenable hypothesis that the “semi-perme- able membranes”’ surrounding living cells are absolutely imper- meable to inorganic salts. The impermeability must be regarded as relative, and the solution of the most interesting physiological problems will depend on our knowledge of the degree and varia- bility of this relative impermeability. In the case of striated muscle, there can be little doubt that under normal circumstances there is a frequent, if not continual, loss of potassium phosphate by the fibers, and this loss must be compensated by their taking these elements from the lymph in some other combination. It is generally believed that the striated muscle fibers contain a cer- tain amount of phosphorus in organic combination with lecithin and nuclein, and it has been shown by Katz" that the phosphorus from these sources appears in the ash of the tissue. Overton lays a good deal of stress on the fact, that the intake of water by striated muscle from hypotonic solutions is smaller than it should be on the supposition that all of the water within the muscle fibers acts as a solvent for the muscle salts. To explain this he has introduced the hypothesis that a part of the water within the muscle fibers is held in a sort of chemical combination by the muscle colloids and thus prevented from acting as a solvent for~ the salts. This hypothetical organically combined water he calls “Quellungswasser;” the term will be translated in this article by the phrase organic water. Urano: Zeitschr. f. Biol., 1, p. 212, 1907; li, p. 483, 1908. '3 Fahr: Ibid, lii, p. 72, 1908. 14 Klug and Olsavsky: Arch. f. d. ges. Physiol. liv, p. 21, 1893. 1870. ett. Edward B. Meigs and L. A. Ryan 4Il The evidence which Overton has accumulated makes the exist- ence of semi-permeable membranes very probable in the case of striated muscle; but we wish to point out that the fact that the ash of a tissue is different from that of the lymph is not of itself sufficient to show that the cells or fibers of that tissue are sur- rounded by semi-permeable membranes. To explain the condi- tions existing in smooth muscle, for instance, without invoking the aid of semi-permeable membranes, it is only necessary to sup- pose that the potassium, phosphorus, magnesium, and sulphur of the ash exist in the living tissue in a non-diffusible form; and to extend somewhat the conception of organic water which Overton hinself has introduced. For, if a little more than half of the water of the tissue existed as organic water, the percentages of sodium and chlorine given in the tables would be sufficient to make the concentration of these elements in the remaining inorganic water the same as they have in the lymph. There are reasons for believing that the fibers of smooth muscle are not surrounded by semi-permeable membranes. For example, a great many facts indicate that these fibers lose fluid when they contract,'* and it is difficult to see how this could occur if they were separated from their surroundings by membranes impermeable to inorganic salts. Further, the changes of weight undergone by smooth muscle in various solutions of sugars and salts are so differ- ent from those undergone by striated muscle in the same solutions that it is difficult to believe that the two sets of phenomena have anything in common. Some of these peculiarities in the behavior of smooth muscle have been already reported;!7 others have been experimentally determined by one of us and may be briefly de- scribed here. Both smooth and striated muscle gain in weight if they are im- mersed in a half strength Ringer solution. If the rates of gain be determined at short (five to ten minutes) successive intervals and the results plotted as curves, it will be found that the curves in the two cases have entirely different characters. Striated 16 Meigs: Amer. Journ. of Physiol., xxii, p. 477, 1908; xxix, p. 317, 1912. 17 Meigs: Ibid, xxvii, p. xvii, 1911. 18 By a half-strength Ringer solution is meant a solution with the follow- ing formula; NaCl, 0.32 gram; KCl, 0.01 gram; CaCh, 0.012 gram; NaHCO;, 0.01 gram; H,O, 100 grams. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 4, A412 Ash of Smooth Muscle muscle gains most rapidly in the first five minutes and less rapidly in each succeeding period; the curve of gain in weight is, at least in the early stages, concave to the abscissa and of such a character as to suggest that the water intake may be the result of osmosis. Smooth muscle, on the other hand, gains no more rapidly inthe first five minutes than in the next four or five succeeding equal periods. Indeed, the gain may be slower at first so that the curve becomes slightly convex to the abscissa; but it tends in general to take the form of a straight line. Striated muscle maintains its original weight for many hours in a 7.5 per cent cane sugar solution, while smooth muscle may nearly double its weight in this solution in the course of an hour. But, if nine parts of the sugar solution are mixed with one part of Ringer’s solution, the smooth muscle shows no marked tendency to gain weight in the mixture. These facts are opposed to the view that the smooth muscle fibers are surrounded by semi-permeable membranes. The mixture of sugar solution with Ringer solution has nearly the same osmotic pressure as the sugar solution itself, yet the muscle gains weight rapidly in this and fails to gain in the mixture. The behavior of the muscle in these solutions, therefore, bears no relation to their osmotie pressure. It must be added that both striated and smooth muscle live for twenty-four hours or more in half strength Ringer, in 7.5 per cent cane sugar solution, and in mixtures of the sugar solution with Ringer at room temperature. The changes of weight under- gone by the tissues in these media must therefore be regarded as the reactions of living tissue. We have some very incomplete evidence to show that the potas- sium and phosphorus of smooth muscle is present in the living tissue in a non-diffusible form, and that smooth muscle contains more lipoid than striated muscle. We have cut pieces of smooth muscle across the fibers, kept them for several hours in isotonic saccharose solution, and then compared their potassium content with that of fresh muscle used as a control. We find that very little (less than 5 per cent) of the tissue’s potassium diffuses out into the sugar solution under these conditions. The boiling water extracts from smooth muscle contain a con- siderably less proportion of the tissue’s total phosphorus than do similar extracts from striated muscle (see ante, p. 406, and experi- Edward B. Meigs and L. A. Ryan 413 mental protocols for phosphorus), though thesmooth muscle extracts contain a larger proportion of solid matter. Both of these facts are evidence for the view that a large proportion of the smooth muscle phosphorus comes from the lipoid of the tissue, and is there- fore non-diffusible under normal conditions. The facts which are known at present, then, point to the follow- ing conclusions in regard to smooth muscle. 1. The fibers of this tissue are not surrounded by semi-permeable membranes. 2. Most of the water of the smooth muscle fibers is held by the colloids of the living tissue as organic water. 3. Most of the potassium, phosphorus, sulphur, and magnesium, which appear in the ash of smooth muscle, are present in the living tissue in a non-diffusible form. PROTOCOLS OF THE EXPERIMENTS ON SMOOTH MUSCLE. Potassium and Sodium: I. 6.787 grams fresh smooth muscle yielded 0.0508 gram combined KCl and NaCl and 0.1299 gram K2PtCle. 0.1299 gram K2PtCl, is equivalent to 0.0397 gram KCI. 0.0508 gram— 0.0397 gram = 0.0111 gram, quantity of NaCl. 0.0397 gram KCl contains 0.0208 gram K. 0.0111 gram NaCl contains 0.0044 gram Na. 6.787 gram muscle. therefore, contains 0.0208 gram or 0.3063 per cent K and 0.0044 gram or 0.0648 per cent Na. II. 6.3422 gram fresh smooth muscle yielded 0.0546 gram combined KCl and NaCl, and 0.1364 gram K2PtClg. 0.1364 gram K2PtCl, is equivalent to 0.0417 gram KCl. 0.0546 gram — 0.0417 gram = 0.0129 gram, quantity of NaCl. 0.0417 gram KCl contains 0.0218 gram K. 0.0129 gram NaCl contains 0.0051 gram N a. 6.3422 gram muscle contains, therefore, 0.0218 gram or 0.3437 per cent K and 0.00 51 gram or 0.0804 per cent Na. Iron: I. 15.2106 grams fresh smooth muscle yielded 0.0003 gram Fe:(POx)2 = 0.0001 gram Fe = 0.0007 per cent Fe. II. 15.1614 grams fresh smooth muscle yielded 0.0003 gram Fe2(PO,)2 = 0.0001 gram Fe = 0.0007 per cent Fe. Calcium: I. 15.2106 gram fresh smooth muscle yielded 0.0009 grain CaO = 0.0006 gram Ca = 0.0042 per cent Ca. II. 15.1614 grams fresh smooth muscle yielded 0.0009 gram CaO = 0.0006 gram Ca = 0.0042 per cent Ca. : Magnesium: I. 15.2106 grams fresh smooth muscle yielded 0.0093 gram Mg:P20;7 = 0:0020 gram Mg = 0.0132 per cent Mg. 414 Ash of Smooth Muscle II. 15.1614 grams fresh smooth muscle yielded 0.0089 gram Mg2P.0; = 0.0019 gram Mg = 0.0126 per cent Mg. Phosphorus: I. The water extract from 10.8756 grams fresh smooth muscle yielded 0.0374 gram Mg2P20; = 0.0104 gram P = 0.0958 parts P per 100 parts muscle. The alcohol extract from 10.8756 gram smooth muscle yielded 0.0139 gram Mg:P;:07 = 0.0039 gram P = 0.0356 parts P per 100 parts muscle. The residue from 10.8756 grams smooth muscle which had been extracted with water and alcohol yielded 0.0057 gram Mg:P:20; = 0.0016 gram P = 0.0146 parts P per 100 parts muscle. This sample of smooth muscle, therefore, contained in all 0.0958 + 0.0356 + 0.0146 or 0.1460 per cent P. II. The water extract from 10.3912 grams freshsmooth muscle yielded 0.0343 gram Mg2P,0; = 0.0096 gram P = 0.0919 parts P per 100 parts muscle. The alcohol extract from 10.3912 gram smooth muscle yielded 0.0097 gram Mg:P.20; = 0.0027 gram P = 0.0260 parts P per 100 parts muscle. The residue from 10.3912 grams smooth muscle, which had been extracted with water and alcohol, yielded 0.0039 gram Mg2P.0; = 0.0011 gram P = 0.0105 parts P per 100 parts muscle. This sample of smooth musele, therefore, contained in all 0.0919 + 0.0260 + 0.0105 or 0.1284 per cent P. Chlorine: I. 3.5270 grams fresh smooth muscle yielded 0.0169 gram AgCl = 0.0042 gram Cl = 0.1191 per cent Cl. II. 3.5008 grams fresh smooth muscle yielded 0.0169 gram AgCl = 0.0042 gram Cl = 0.1200 per cent Cl. Sulphur: I. 6.3489 gram fresh smooth muscle yielded 0.0797 gram BaSO, = 0.0109 gram S = 0.1724 per cent S. II. 6.0210 grams fresh smooth muscle yielded 0.0658 gram BaSO, = 0.0090 gram S = 0.1501 per cent S. The experiments described in these protocols are those of which the results are given in Table II. We have analyzed other samples of smooth muscle for potassium, sodium, phosphorus, and chlorine; but have not used the results of these analyses in calculating our averages, chiefly because they were preliminary single analyses without accompanying parallels. The results were, however, not far from those which appear in the table and may be given here: Potassium, 0.3458 per cent; sodium, 0.0506 per cent; phos- phorus, 0.1494 per cent; chlorine, 0.0988 per cent. The percentages of water and total solids in our samples of tissue were determined by drying them at between 100° and 110°C. until they reached constant weight. This usually required between twenty-four and seventy- two hours, though the loss of weight after twenty-four hours was very slight. The figures in the table represent the widest differences found in six deter- minations, and the average is calculated from all six. In the four deter- minations not given in the table the total solids were found to be 17.49 per cent, 17.60 per cent, 17.82 per cent, and 17.93 per cent respectively. THE TOXICITY OF SUGAR SOLUTIONS UPON FUN- DULUS AND THE APPARENT ANTAGONISM BETWEEN SALTS AND SUGAR. By JACQUES LOEB. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, March 26, 1912.) In 1901 the writer pointed out the fundamental difference in the behavior of electrolytes and non-conductors in regard to antag- onistic action on the eggs of Fundulus.! While it was to a large extent possible to antagonize the toxic action of one electrolyte by another it was impossible to antagonize the toxic action of an electrolyte by a non-electrolyte. Gies and I found an apparent exception to this rule in the case of the salts of heavy metals, e.g., ZnSO; which could be antagonized by an excess of cane sugar,? and Sumner found afterwards that the toxic action of CuSO, could also be deferred through the addition of cane sugar. We were inclined to explain this apparent exception on the assump- tion of the formation of saccharates with a diminution in the con- centration of the free metal ions. In the case of the antagoniza- tion of one salt by another we are dealing with a common action of both salts upon one or several colloids on the surface of the organism.* The writer recently made experiments on the toxic action of sugars on Fundulus, the results of which, at first sight, seemed to speak in favor of the possibility gf an antagonization of the toxic action of sugars by salts. A more thorough analysis, however, showed that in this case the real antagonism was between two elec- 1 Amer. Journ. of Physiol., vi, p. 411, 1902. 2 Pfliiger’s Archiv, xciii, p. 246, 1902. 3 Amer. Journ. of Physiol., xix, p. 61, 1907. ‘ Science, xxxiv, p. 653, 1911. 415 416 Toxicity of Sugar Solutions trolytes. The method employed was identical with that used in the previous experiments by Mr. Wasteneys and the writer. A series of sugar solutions was prepared, 500 cc. of each, and six Fundulus were put into each of these solutions, after having been washed repeatedly in fresh and distilled water. The number of fish which had survived the treatment was ascertained daily. Tables I and II give the records of two simultaneous experi- ments, one with various concentrations of cane sugar alone, the other a the same concentrations of cane sugar solutions made up in a § solution of NaCl + KCl + CaCl (in the wane propor- tion) instead of in water. A comparison of the two tables gives the typical ae of an antagonistic action; while in the pure cane sugar solutions all the fish were dead on the seventh day, in the solutions of cane sugar TABLE I. Number of surviving fish in various concentrations of cane sugar in water. CONCENTRATION OF CANE SUGAR AFTER DAYS 1 ss isi 8 ft 2 1 6 4 5 3 6 3 0 5 0 0 a TABLE II. Number of surviving fish in various concentrations of cane sugarin NaCl, KCl, CaCl; CONCENTRATION OF CANE SUGAR AFTER DAYS Cm OI OW eH Jacques Loeb 417 made up in 7 NaCl + KCl + CaCl, practically all the fish were alive in all the sugar solutions below =. Nevertheless the salt did not antagonize the toxic action of the sugar solutions in this case, but the action of a fermentation product from the sugar, namely an acid. In previous publications by Mr. Wasteneys and the writer it was shown, that the toxic effects of acid on Fundulus could be antagonized by salts, and that the concentration of acid the toxie action of which can be antagonized, is not inconsiderable.® Table III gives the maximum concentration of acid which is antag- onized by a mixture of NaCl + KCl + CaCl, of various concen- tration. TABLE III. CONCENTRATION OF THE MIXTURE OF NaCl, MAXIMAL QUANTITY OF x. HCl wHicu-THE KCl anv CaCh FISH WILL RESIST IN THESE SOLUTIONS 0 0.1-0.2 a5 0.2 ts 0.5 3 1.0 ¥ 122=1-4 M z 0.8 8 t 0.6 M z 0.3 M 5 02 The sugar solutions with and without salt soon become turbid and later almost opaque and an examination showed that the solu- tion was teaming with bacteria. A titration of the sugar solutions with NaOH gave a rather high degree of acidity as was to be expected. Twenty-five cubic centimeters of the 35 cane sugar solution in ¥ NaCl + KCl + CaCl, required on the ninth day 7.5 ec. 730 NaOH for neutralization. The toxic action of the sugar solution on the fish was therefore due in part not to the sugar but to a fermentation product, namely an acid. That the fish died of acid poisoning was also indicated by their externa! appearance. As I pointed out in a previous paper the epidermis 5 Biochem. Zeitschr., xxxili, p. 489, 1911; xxxix, p. 167, 1912. 418 Toxicity of Sugar Solutions becomes white in the case of acid poisoning, and the fish which died in the sugar solutions became white before death occurred. Similar results were obtained in experiments with dextrose, as Tables IV and V will indicate. The experiments represented in these two tables were made simultaneously. TABLE IV. Number of surviving fish in various concentrations of dextrose in water. CONCENTRATION OF DEXTROSE 6 6) 6.cemGen | Ash sel £0 feiO la Oniteaat | TABLE V. Number of surviving fish in various concentrations of dextrose in * NaCl + KCl + CaCh. CONCENTRATION OF DEXTROSE AFTER = = Br DAYS M | M : : =e ; 7 = I The dextrose solutions were less toxic when they were made up in ; Ringer solutions than when made up in H,O. As was to be expected the dextrose solutions soon became turbid and opaque and contained a considerable amount of free acid. Again the inference was unavoidable that the antagonism in this case existed between the free acid and the salt and not between the sugar and the salt, and this surmise was supported by the fact that the epi- dermis of the fish showed the whiteness characteristic for the effect of acid. This idea that the acid and not the sugar solution killed the fish could be put to a further test by a comparison of the toxic action of sugar solutions which were allowed to ferment and sugar Jacques Loeb 419 solutions which were renewed sufficiently often to prevent a high concentration of acid through the action of bacteria. Into each of two dishes with 500 cc. ¥ dextrose six Fundulus were put. The one dish remained unaltered; the fish of the other dish were trans- ferred into a fresh % dextrose solution every twenty-four hours. In these latter solutions fermentation began also and some acid was formed, but this solution remained clear and the amount of acid was too small to do much harm. Care was taken that the new sugar solutions always had the same temperature as the old ones. In the dextrose solution which was not changed all the fish were dead after four days, as in the previous experiments, and they died with the symptoms of acid poisoning. The six fish which were transferred into a fresh ¥ dextrose solution every day are still alive and apparently normal today, on the twenty-sixth day of the experiment. This experiment was repeated with the same result. A similar experiment was started with an ¥% solution of cane sugar. Four dishes, each with 500 cc. 3 cane sugar, were prepared and six Fundulus put into each dish. In two dishes the solution was not renewed, the fish from the other two solutions were trans- ferred every day into a fresh | solutions of cane sugar. In the dishes in which the cane sugar solution was not renewed the fish were all dead after three and five days respectively. The fish which were transferred every day are partly alive today after fifteen days, in one dish five out of the six put there originally are still alive and two in the other dish. These experiments leave no doubt that at least part of the toxic effects of the sugar solutions is due to the formation of acid and that the antagonism expressed in Tables I and II and IV and V is the antagonism between acid and salts. It became a matter of interest to find out to what extent a pure sugar solution may be called toxic for these fish. For this purpose six fish each were put into a 7's, 5, + and ¥ solution of cane sugar and the solution renewed each day. Table VI gives the results. The rapid death in the } and } cane sugar solution cannot be ascribed to a product of fermentation, e.g., acid; here we are pos- sibly dealing with a direct action of the sugar. The js and ¥ solutions, however, behave almost like an indifferent salt-free solution. 420 Toxicity of Sugar Solutions TABLE VI. Number of surviving fish in various concentrations of cane sugar. CONCENTRATION OF CANE SUGAR AFTER DAYS |; =o | | ee | Mm M f M | 18 8 4 2 ee 1 6 6 6 6 2 6 6 5 4 3 5 6 5 1 4 5 6 4 0 Ul 4 5 1 9 4 Ae wih 0 Fula sob Ten es ea | The writer is not in a position to judge whether or not these results can be applied to the interpretation of the symptoms of patients with glycaemia. During the experiments the fish were not fed and therefore did not take up any salts. SUMMARY OF RESULTS. 1. Fundulus live longer in solutions of cane sugar and dextrose made up in a ™ solution of NaCl + KCl + CaCl, than in the same sugar solution without salts. 2. It is shown that in these solutions (through bacterial action) a considerable amount of acid is formed and that the apparent antagonism between sugar and salt is in reality a case of antagonism between acid and salt. 3. This conclusion is corroborated by the observation that % solutions of dextrose or cane sugar, which kill the fish in a few days become almost harmless if the solution is renewed every twenty- four hours, whereby the multiplication of bacteria and the forma- tion of acid is considerably diminished. 4. These experiments also show that ¥ or weaker solutions are in themselves nearly harmless and that their apparent toxicity upon Fundulus is due to the formation of acid or other products under the influence of bacteria. Concentrations of sugar equal to or greater than * are, however, harmful independently of the acid formation by bacteria. CARBOHYDRATE ESTERS OF THE HIGHER FATTY ACIDS. III. MANNITE ESTERS OF LAURIC ACID. By W. R. BLOOR. (From the Laboratories of Biological Chemistry, Washington University, St. ; Louis, Mo.) (Received for publication, March 27, 1912.) In the earlier papers of this series the preparation and chemical and physiological properties of some mannite esters of stearic acid have been described. These compounds, because of their high melting point and low digestibility were not well adapted to - physiological investigations, and it was decided to prepare similar compounds of a fatty acid lower in the series, in the hope that they would prove more suitable. Lauric acid was chosen as being of sufficiently lower carbon content to give a decided difference in properties, and because its wide distribution promised a ready supply of material. To prepare laurie acid, laurel oil was saponified and the fatty acids fractionally distilled at low pressure according to the directions of Krafft.'. The fraction so obtained, which boiled at 225° under 100 mm. pressure, was found to have too low a melting point. It was, therefore, recrystallized from ice cold alcohol until a product was obtained which melted at 42°C. and distilled at 225° under 100 mm. pressure. This sample was used in the preparation of the mannite esters described below. MANNITE DILAURATE. Mannitan dilaurate was prepared from the lauric acid as fol- lows: Ten grams of mannite (Kahlbaum) were dissolved in 200 ce. warm (38°) concentrated sulphuric acid and 23 grams lauric acid stirred in. When all 1 Krafft: Ber. d. deutsch. chem. Gesellsch., \xiii, p. 4344, 1903. 421 422 Mannite Esters of Lauric Acid had dissolved the mixture was digested over night in an incubator at 38°C. To separate the mannite esters from the acid mother liquor, the solution was poured with constant stirring into iced, saturated ammonium sulphate, let stand for a short time, transferred to a filter, let drain as completely as pos- sible, then washed once with saturated ammonium sulphate. After again draining, the mass was transferred to boiling alcohol in which it separated into two layers, a lower, watery layer containing most of the ammonium sulphate and acid, and an upper alcoholic layer which contained the mannite ester. The lower layer was siphoned off and the alcoholic solution allowed to cool. The ester was filtered off and recrystallized from alcohol until the melting point was constant. The yield from the second recrystallization from alcohol was 13 grams, corresponding to about 45 per cent of the lauric acid used—a very- much lower yield than was obtained from stearic acid. Griin in his work on the synthetic fat has also found that the sulphuric acid synthesis becomes less effectiv the smaller the molecular weight of the fatty acid.* The product obtained greatly resembles the homologous stearic acid compound. It separates from alcohol in microscopic needles which when dried are snow white. It is practically insoluble in cold alcohol, and only slightly soluble in cold ether, benzol or chloro- form. It dissolves in these solvents when heated to boiling and separates on cooling in crystalline form. It is heavier than water. Melting point (uncorrected), 122°C. It is slightly dextrorotatory. Because of its slight solubility there was the same difficulty in making the polarimetric reading as-with the stearic acid ester. One gram of mannitan dilaurate dissolved in 25 ec. chloroform at 50°C. and read in a1 dm. tube gave a reading of +0.34°. [a], = +8.5° AnaLysis. A fatty acid determination was made in the regular way. The ester dissolved in hot alcohol was saponified by alcoholic alkali, most of the alcohol driven off and the residue taken up with water. The fatty acid set free by the addition of sulphuric acid was filtered, washed with hot water until the wash water showed no trace of sulphate, then dried and weighed. (1) 0.5938 gram mannitan dilaurate yielded 0.7218 gram laurie acid = 75.68 per cent. (2) 1.0449 gram mannitan dilaurate yielded 0.7920 gram lauric acid = 5.79 per cent. Calculated for mannitan dilaurate, CsHio0O3(Ci:H2;COO). = 75.75 per cent. ‘ Griin: Ber. d. deutsch. chem. Gesellsch., xl, p. 1778. > W.R. Bloor 423 ComBusTIon. (1) 0.1512 gram yielded 0.3773 gram CO: and 0.1424 gram H,0. (2) 0.1496 gram yielded 0.3727 gram CO: and 0.1423 gram H,0. (3) 0.1387 gram yielded 0.3452 gram COb. Calculated: Found: 1 2 3 OC tea 68.18 68.06 67.95 67.88 |: (os tye Bee 2 ea eben be 10.61 10.47 10.57 Mannitan dilaurate, because of its high melting point, did not seem promising, and no feeding experiments were made with it. ISOMANNID DILAURATE. Isomannid dilaurate was prepared from the mannitan dilau- rate by a short heating to 200°C. in the same way as isomannid distearate was prepared from mannitan distearate.2 The product was freed from small portions of the unchanged mannitan ester by repeated treatment with cold ether, filtering and evaporating to dryness, until it dissolved clear in a small portion of the cold sol- vent. It was then taken up with ether, titrated with alcoholic alkali and phenolphthalein to remove any free lauric acid, filtered, treated with bone black to remove color, again filtered and evapo- rated to dryness. The pure white product was further purified by several recrystallizations from ice cold alcohol. The product is snow-white and when melted and cooled appears crystalline. It is lighter than water and emulsified readily with warm soap solution. When heated above 100°C. it slowly volatilizes with decompo- sition. It is readily soluble in cold ether, benzol or chloroform and is quite soluble in cold alcohol. Its melting point is 37.5°C. (uncorrected). Optical activity. Like the corresponding stearic acid ester it is strongly dextrorotatory. Determinations were made in ether and benzol solution. . ETHER. 0.343 gram in 50 cc. ether in 2 dm. tube gives reading +1.73; = -++-125.5°. 0.600 gram in 16 cc. ether in 1 dm. tube gives reading +4.69; = +-125.1°: 3 Bloor: This Journal, xi, p. 141, 1912. 424 Mannite Esters of Lauric Acid Benzou. 0.494 gram in 50 ec. benzol in2dm. tube gives reading +4.95; = +125.0. Refraction. Abbe-Zeiss refractometer. n4o = 1.4570 Nags = 1.4555 nso = 1.4535 Neo = 1.4500 Its low melting point and ready saponification rendered isoman- nid dilaurate a promising substance from a physiological point of view, and experiments were conducted to determine its availability for the animal organism. Because of the difficulty of obtaining pure laurie acid in large quantity, and since for physiological pur- poses a pure compound was not required, a mixture of isomannid esters of lauric, myristic, etc., acids was used. The acidswere obtained from cocoanut oil, which consists mainly of esters of capric (20 per cent), lauric (40 per cent) and myristic (24 per cent) acids together with small amounts of palmitic (10 per cent) and oleic (5 per cent) acids. The palmitic and oleic acids and most of the capric acid were removed as follows. Commercial cocoanut oil was saponified with alcoholic potash. The soaps were dissolved in water, boiled, skimmed free from unsaponified material, and the fatty acids set free with sulphuricacid. After washing several times with hot water, then cooling, the cake of fatty acids wasdried and dissolved in just sufficient alcohol to avoid any-separation when cold. Enough saturated magnesium acetate solution was added to precipitate the palmitic acid, the mixture set in the ice chest over night, then filtered ona Buchner filter. The precipitate contained the palmitic acid. For the separation of the oleic acid, recourse was had to the solubility of its lead soap inether. To the filtrate from the magnesium precipitate a hot satu- rated water solution of lead acetate was added as long as a precipitate was formed, and the mixture was set in the cold over night. The lead soaps were filtered off, pressed as dry as possible, then boiled out with water. They melted and sank to the bottom, and on cooling solidified. The water was poured off, the solid mass dried, melted to get rid of the remaining water, shaved fine and extracted several times with cold ether. The insoluble residue was boiled out with dilute hydrochloric acid sufficient to remove the lead, washed well with boiling water and finally cooled. The mixture of fatty acids so prepared, consisting largely of lauric acid, was used in the preparation of the mannite esters for the feeding experiments. The proportions of the fatty acids varied somewhat, depending on the sample of oil used and on the condi- tions of the separation, so that ester mixtures prepared from differ- ‘W. R. Bloor 425 ent samples differed to some extent in melting point and optical activity. Isomannid esters were prepared from these fatty acid mixtures by the method outlined above. Weighed quantities of the esters together with meat were fed to animals, and the amount of digestion determined by the optical activity of the ether extract of the feces. It was obviously neces- sary to know the optical activity of the ether extract of normal feces under parallel conditions. Some preliminary experiments were, therefore, carried out: in which the animal was fed lean meat and lard. PRELIMINARY (CONTROL) EXPERIMENTS. A cat, weighing 1.6 kilos, was starved for two days, then fed on each of three days, 75 grams of hashed lean beef, 7.5 grams lard and 3 grams bone ash (added in order that the resulting feces could be readily dried and powdered for extraction). The feces were collected, dried, broken up and extracted three to four hours with ether in a Soxhlet extractor. The ether extract was filtered, evaporated to 50 cc. and readings — in a polariscope using a 1 dm. tube. First day—reading of 50 cc. ether extract of feces in 1 dm. tube = +0.04°. Second day—reading of 50 cc. ether extract of feces in 1 dm. tube = +0.07°. Third day—reading of 50 ce. ether extract of fecesin 1 dm. tube = +0.03°. Average reading for one day’s feces on above diet = +0.05. Other blank experiments were carried out in the intervals of the ester experiments with results substantially the same. FEEDING EXPERIMENTS WITH THE ISOMANNID ESTERS. The mixture of esters used in these experiments had a melting puint of 25° and specific rotation of +87.0°. Between each of the experiments came two days of feeding of meat mixed with a little wood-charcoal, so that the feces from the ester-feeding days were sharply marked off. EXPERIMENT I. The cat, after a preliminary starvation period of two days, was fed 50 grams hashed lean meat, 3.5 grams isomannid esters and 3gramsboneash. The resulting feces were dried, broken up and extracted three to four hours in a Soxhlet extractor. The ether extract after filtering was evaporated to 50 cc. and readings taken with the polariscope using a 1 dm. tube. Average of readings = +0.05°. Subtracting reading of blank +0.05°—indicates complete absorption. 426 Mannite Esters of Lauric Acid EXPERIMENT II. The same animal was fed 50 grams hashed lean meat, 4 grams of the isomannid esters and 3 grams bone ash. The feces were col- lected and extracted as usual. Reading of the extract in 50 cc. ether in 1 dm. tube = +0.29.° Blank = +0.05°. Corrected reading = +0.24°. Corresponding to an ester content of +0.14 gram. Absorption = 97.15 per cent. EXPERIMENT III. The same anima’, after the usual preliminary feeding with meat and charcoal was fed 50 grams of hashed lean meat, 4 grams of isomannid esters and 3 grams of bone sash. The feces were collected and extracted as usual. Feces extract in 50 cc. ether in 1 dm. tube = +0.39°. Blank = +0.05°. Corrected reading = +0.34°. Corresponding to 0.19 gram of ester. Absorption=95.3 per cent. The results of the experiments indicate a practically complete utilization of the isomannid esters of the fatty acids used. The slightly better absorption in Experiment I may have beem due to the previous starvation. In the paper immediately following this one are reported experiments on dogs, in the course of which, after feeding isomannid esters, the contents of the whole intestinal tract were removed and examined for unabsorbed esters. The . results bear out the findings of the above experiments, and show clearly that mannite esters of the fatty acids, if of suitable melting point, are as well utilized by the animal organism as are ordinary fats. Because of the excellent utilization of the isomannid esters of the fatty acid. mixture used above (mainly lauric acid) it became of interest for purposes of comparison to determine the degree of utilization of the homologous isomannid ester of stearic acid.4 The same animal was used as in the other experiments (a cat weighing about 1.6 kilo). EXPERIMENTI. After two days’ starvation, the cat was fed 50 grams lean beef, 3 grams bone ash, and 4 grams isomannid distearate dissolved in 12 ce. of cotton seed oil. (The resulting mixture melted at 45°C°, but remained soft at body temperature. An unsuccessful attempt was made to obtain and feed a mixture melting at body temperature by using more cotton seed ‘Tsomannid distearate is described in the second paper of this series. Bloor: loc. cit. W. R. Bloor 427 oil. The use of so much oil caused a diarrhoea.) The feces were collected and extracted with ether as in the previous experiments. First passage of feces. Ether extract in 50 cc. ether in 1 dm. tube reading 0.84° Blank (for whole day), 0.05°. Corrected reading, 0.79°. Corresponding to a weight of ester of 0.43 gram. Second passage of feces. Ether extract in 50 cc. ether in 1 dm. tube read- ing 1.10. Corresponding to a weight of ester of 0.60 gram. Total ester recovered in feces, 1.03 grams. Per cent absorption, 73 per cent. ExpPERIMEN? II. The same animal! was fed 3.5 grams of isomannid di- stearate in 10 cc. of cotton oil with 50 grams lean beef and 3 grams bone ash. Feces collected in two portions and extracted with ether as usual. Portion1. Reading of 50 cc. ether extract in 1dm.tube = +0.87°. Blank (for whole day) = +0.05°. Corrected reading +0.82°. Corresponding to 0.44 gram ester. Portion 2. Reading of 50 cc. ether extract in 1 dm. tube = +1.03°. Corresponding to 0.55 gram of ester. Total ester recovered infeces = 0.99 grain. Per cent absorption = 72 per cent. The utilization of isomannid distearate by the animal body is quite comparable to that of a high melting fat, e.g., tristearin.® SUMMARY. Mannitan and isomannid di-esters of lauric acid have been pre- pared and described. The isomannid esters of lauric and closely related fatty acids have been shown to be as well utilized by the animal organism as ordinary fats. The work is now being extended to the prepara- tion of the esters of the higher fatty acids with the true carbo- hydrates. 5’ Arnschink: Zeiischr. f. Biol., xxvi, p. 434, 1890. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. XI, NO. 4 ia thed bominniet: tetera soe sfianesing ARS 1) Sar eee la hae Be obisa: ¥ Saf) ise*éIon c y me 2 eeians he gay? tiger priarmibi ce fit eS salt fiF ‘Hed i 3S rode ot is = ieipata Took! : IP War ister opm aig oS ei eee (sen daa snsedanlot is dileacniaon dee 6NI axl Longe 24 cigce. salar glee Gatoanrgecky ; os ode) ab Remp gadis | ni aedaieetie ites main Xo Speen te! z Li Ra +> « or ohh eyes + heer tis Sb. tide: as * ” te ree “a hand la u 1 ing ee 4" 0T.0 GalBoas. rf ne 2 defo acehes hiya eo ynibita 1a aes ah SRE Ee Sparro anil 2a98¢ Ys aneanng © m2" : ‘4 : : tino 1 2h iste) a ¥, (A iatte bf ped Ate a o iid sakting 3 Angie’ hpi g ars REEL LBM b Pi tovaraoeey moe 1 Aer Z fish iP EF ne cuttgpgll eer: - ol mavtdegsing katate og on ot sarangi. i) Esta UC | ‘tier lip TeAAs to he " 4 ate pes -o «he Metomtixe Ga ace roiiag creed ty eed ton Ode en. Js ithe Riker whe dy wo Bio gmbpak.” | ee ae pie SO ae ~ (ge ota sot ; te Ree cine, eae : OL eae 2 FOCUS areca ge Rate es tbh Ted: = Gs~ hin ele alti $s eta iia ae se i ' eetiaade ais aa Sbrchts t et 2 ee Bers EA: 30.9 R ROTH crcaeaanind 7 i : ~ iget 30 Ss oe ~ i a ree avait pete d URE RO O i ict ay.9 pvt ton teakeesh'bt nile nm soeancat 3 Sree gl = aE Lee : RPA ant aaa . ae al pc = sy 4)’ as ical il aieeiay Pig ee pf Yor bs) 30 i mes Spot * 7 *) ie ES : iat a3 mr ER rh Bhar pie he ae} ves fo it gran zs Fz + ae g? AAS 3S at. Seige tiga ora G ! sae = Suits he ry ates, cf tang ‘| bode a = AG Teeitgeeete = Werte ing vas bt = fhe tS ae. 2 Gee as wv VAST tes as v2 . mal ' 5. * 44 ? — pas 7 ie _ Ps er es oer ‘ae erie sia . pend : + eas ta ave " tebeatinie whotamoote 30 ON FAT ABSORPTION. By W. R. BLOOR. (From the Laboratories of Biological Chemistry, Washington University, St. Louis.) (Received for publication, March 27, 1912.) It is pretty generally believed that under normal intestinal conditions, most, if not all, of the food-fat is saponified in the intes- tine before absorption, and is absorbed as soaps. The question whether all fat must be split before absorption is still in doubt. It has been shown that fat-like substances such as the petroleum hydrocarbons! and wool fat,” although they emulsify well, are not absorbed (a fact which in regard to the petroleum hydrocarbons has recently been disputed). Frank has demonstrated that the ethyl esters of the fatty acids, all of which form good emulsions and are readily saponified by the intestinal lipases, do not pass into the chyle unsplit.4 On the other hand there is much evidence to show that food fats may be transferred to the fat depots of the animal body without apparent change. The research on the carbohydrate esters of the higher fatty acids’ was undertaken primarily in the hope of obtaining a fatty substance of such characteristic properties that it could be traced through the processes of absorption. With such a substance it was intended, among other things, to test further whether a fatty substance could pass into the chyle unchanged. The isomannid dilaurate described in the preceding paper seemed suitable for this purpose since it has great optical activity ([a]*,? = +125°) which is lost on saponification, it emulsifies and saponifies readily, and 1 Henriques and Hansen: Zentralbl. f. Physiol., xiv, p. 313, 1900. 2 Connstein: Arch. f. (Anat. u.) Physiol., p. 30, 1899. 3 Bradley: Proceedings Amer. Soc. Biol. Chem., Baltimore, Dec., 1911. 4 Frank: Zeitschr. f. Biol., xxxvi, p. 568, 1898. 5 Bloor: This Journal, vii, p. 427; xi, p. 141. 429 430 Fat Absorption has been shown to be well utilized by the animal organism. Ex- periments were accordingly undertaken to determine whether iso- mannid dilaurate, when submitted to the processes of digestion and absorption could be detected in the chyle. After a preliminary period of starvation the animals (dogs) were fed considerable amounts of the isomannid esters, described in the paper immediately preceding this, together with hashed lean beef, and after a sufficient time had elapsed for digestion to be well under way they were etherized, a cannula tied into the thoracic duct® and the chyle collected for a period of three or four hours. The animals were then killed and the contents of the gastro-intes- tinal canal removed and examined for unabsorbed esters. The fat of the chyle was extracted by shaking out several times with ether, and finally the chyle was evaporated to dryness and again extracted with ether. The ethereal extracts were examined for optical activity. No attempt was made to determine the exact nature of the fat. EXPERIMENT J. A young dog (female), weighing 9 kilos, after starving two days, was fed 160 grams of hashed lean beef and 19 grams of isomannid esters of the fatty acids (mixture of lauric, myristic, etc.) obtained as described in the preceding paper. The ester mixture had a melting point of 25° and a specific rotation of +88.4°. Time of feeding, 8:45a.m. At 1:30 p.m. the animal was etherized and at 2:05 p.m. the cannula wassecurely tied into the duct—five hours and twenty minutes after feeding. Collection of chyle. Portion 1. From 2:05 p.m. to 3:05 p.m., 33 cc. of very white milky chyle. Portion 2. From 3:05 p.m. to 4:05 p.m., 34 cc. chyle rapidly losing its white color. Portion 3. From 4:05 p.m. to 5:05 p.m., 35 ee. chyle of yellowish color. Portion 4. From 5:05 p.m. to 5:35 p.m., 14 ce. chyle of yellowish color. Animal killed. The blood pressure remained at a good height (130to 140 mm.) throughout. Ether extract of chyle. Portion 1, acidified with sulphuric acid and extracted four times with warm ether. Extract = 0.8 gram. Per cent of fat in the chyle = 2.4 per cent. Optical activity, 0.8 gram in 35 cc. ether in 2 dm. tube, none. 6 The operations were performed by Dr. D. E. Jackson of the department of Pharmacology of this school, to whom I take this opportunity of express- ing my thanks. W.R. Bloor 431 Portions 2, 3 and 4 combined and extracted similarly. Weight of total extract, 0.8 gram. Per cent of fat in chyle of last three portions = 0.97 percent. Optical activity, none. The fat was saved to be united with material from later experiments for further examination. Examination of intestinal tract. The whole gastro-intestinal tract was removed shortly after death, slit throughout its whole length, and the contents scraped out under alcohol. The stomach was found to be very much distended with gas and contained considerable undigested material; the small intestine was empty except for masses of intestinal worms. The large intestine contained old fecal material only. The alcoholic mixture was allowed to settle in the cold, the clear liquid poured off and evaporated to dryness. The alcohol-insoluble portion was dried separately. After drying, the two portions were united and extracted several times with hot ether. The extracts were united, washed with water, treated with a little bone black, filtered and evaporated to dryness. The dry residue was taken up with ether and the ether extract examined for optical activity. Ninety-nine cubic centimeters of ether extract in 100 mm. tube gave a rotation of +1.88, corresponding to a weight of ester of 1.9 gram. The ether extract was titrated with alcoholic alkali and phenolphthalein. Acidity = 6.7 cc. % alkali corresponding to 1.34 gram lauric acid. The extract was now acidified, washed with water, evaporated to dryness, taken up with ether, filtered, again evaporated to dryness, and the residue weighed. Weight, 4.0 grams. The ether-soluble material from the intestinal canal may then be listed as follows: Grams Pipes tannid esters......-...2. 2 50ST ee 1.9 Patrvememisras laurie acid: 2° 205..00.5 SP RRS 1.34 Oe eee ee SUS, | ee aaaeeme oo et, lke STs, 0.76 THe o-oo Seo og ee ene ee Sen Ge ite ooo Seer re el) Eight hours and fifty minutes (four hours under ether) after feeding, 90 per cent of the ester had been split and 83 per cent absorbed. EXPERIMENTII. Dog (female), weight, 6 kilos; after two days starvation was fed 150 grams hashed lean meat and 13.2 grams of the isomannid esters. This sample of ester melted at 20°C. and had a specific rotation of +100° Time of feeding, 11:10 a.m. It was decided to begin the operation earlier than in the first experiment since it was found that digestion was almost complete at the time of the operation (five to six hours after feeding). The animal therefore, was etherized at 1:20 p.m. The cannula was in the duct and collection of chyle begun at 1:50 p.m.—two hours and forty minutes 432 Fat Absorption after feeding. The chyle was received in a small amount of ammonium oxalate solution, in order to obviate the difficulty of working with clotted lymph. Collection of chyle. Portion 1. 1:50 to 2:50, 25 cc. white and milky. ~ Portion 2. 2:50 to 3:50, 35 cc. losing its dead white color. Portion 3. 3:50 to 4:50, 28 cc. yellowish. Portion 4. 4:50 to 5:20, 11 cc. almost clear. Animal killed. The respiration was labored throughout, and the blood pressure was low and irregular, so that the animal required constant watching. Ether extractions of chyle. Portionl. Weight of extract,0.5gram. Fat content of chyle, 2 per cent. Optical activity, 0.5 gram in 20 cc. ether in 200 mm. tube, none (0.5 gram of the isomannid ester would have given a rotation of +5.0°). Portion2. Weight of extract,0.4gram. Fat content of chyle, 1.2 percent. Optical activity, none. Portions 3 and 4. Weight of extract, 0.1 gram. The extracted lymph from 1, 2, 3 and 4 was neutralized and evaporated to dryness with the aid of alcohol, and the dry material again extracted. Weight of extracted substance, 0.4 gram. The chyle fat was united with that of Experiment I and saved for further examination. Examination of gastro-intestinal contents. The stomach and intestinal tract were treated as in Experiment I. The stomach was very much distended with gas, and contained a considerable amount of undigested food, while the small intestine contained oply mucus. The whole:tract was scraped out and the contents dried and extracted with ether, asin the first experiment. The ether extract was washed with water, treated with bone black, filtered and examined for optical activity. iVolumerotsether extracte..- >. season eee eee 475 ce. Rotatronuneltzam: tube sgsckoe oie oe ee eee +1.0° Corresponding to a weight of ester of.................. 4.75 grams. Acidity. The ether extract was evaporated to dryness, the residue taken up with chloroform and titrated with alcoholic alkali and phenolphthalein. Titration. 11.84 cc. ¥ alkali = 2.36 grams calculated as lauric acid, cor- responding to 2.5 grams of the dilauric ester. The chloroform solution was acidified, washed with water and evaporated to dryness. Weight of dry residue, 10.0 grams. The constituents of the ether soluble portion of the intestinal contents were then W. R. Bloor 433 Grams MimenneCeSter....; . 2.2. eee eels os ie edd wek iss 4.75 eeI AS LAUTIC ACIG... 2 220 Meee ls oats eae ss 2.36 Unidentified........ PRE eRe er ee ee 2.89 LS FP ain ce SYR ad Be meet TE 2 crn a te ea 10.00 Unabsorbed ester and f atty ACI Coe eee eee fel Six hours and ten minutes (four hours under ether) after feeding, 64 per cent of the total ester had been split and 46 per cent absorbed. ‘The chyle fat. The combined ether extract of the chyle from the two experiments was now examined. It was purified by solution in ether, evaporation to dry- ness, re-solution, neutralization, treatment with bone black, and filtering. The clear, slightly yellowish filtrate was evaporated to small bulk and examined for optical activity. Twenty-five cubic centimeters of ether extract, containing 1.8 gram of the purified fat wasexamined ina polariscope using a 100 mm. tube: Reading 0.08°, showing clearly that no unsplit ester had passed into the chyle (1.8 gram of the ester in 25 cc. of ether would have shown a rotation of +7.2°). Melting point, 32°C. Refraction, Abbe refractometer—n‘* = 1.456. (Trilaurin nq®°= 1.440.) The fat was now saponified and the fatty acids separated. Melting point of mixed acids, 30°C. Mean molecular weight. The mean molecular weight of the fatty acids was determined by titration. 0.357 gram in 100 cc. chloroform with phenolphthalein = 1.69 cc. ¥ alkali, from which the mean molecular weight was calculated to be 211. Iodine number (Hiibl method): 0.26 gram of fatty acid absorbed 0.043 gram I. Iodine number, 16.5. The chyle fat therefore consists probably of trilaurin containing some triolein. SUMMARY OF THE EXPERIMENTS. I. Weight of dog, 9 kilos. Time of operation, five hours and twenty minutes after feeding. Duration of experiment, four hours. Weight of esters fed, 19 grams. Weight of esters absorbed, 15.8 grams. Esters in chyle, none. II. Weight of dog, 6 kilos. Time of operation, two hours and forty minutes after feeding. Duration of operation, four hours and thirty minutes. Weight of esters fed, 13.2 grams. Weight of esters absorbed, 6 grams. Esters in chyle, none. 434 Fat Absorption CONCLUSIONS. The results of the experiments show quite conclusively that none of the isomannid esters had passed unchanged into the chyle, although considerable quantities had been digested and absorbed. This result is in entire agreement with the findings of Frank’ with the ethyl esters of the fatty acids and emphasizes the proba- bility that readily saponifiable fatty acid esters do not escape saponification under the favorable conditions in the normal intes- tine (excess of lipase, rapid removal of the products). Whether fatty substances of any kind may pass into the chyle unchanged remains to be proven. * Frank: Loc. cit. ECHINOCHROME, A RED SUBSTANCE IN SEA URCHINS. By J. F. McCLENDON. (From the Embryological Laboratory of Cornell University Medical College, New York City, and the U. S. Bureau of Fisheries, Wocds Hole, Mass.) (Received for publication, March 30, 1912.) INTRODUCTION. My interest in echinochrome arose from studies in permeability. In the same way that haemolytic agents cause haemoglobin to leave the red blood corpuscles, so do cytolytic agents cause echino- chrome to leave the cells containing it. R. Lillie is of the opinion that this is due to the action of the cytolytic agent in increasing the permeability of the cell surface. In the elaeocytes,- wandering cells of the body fluid of Arbacia punctulata, the cytoplasm is crowded with spherical chromatophores. Some of these may be colorless, but usually they are colored bright red with echinochrome. Similar chromatophores, though not so close together, occur in the eggs. In the unfertilized egg they are evenly distributed throughout the cytoplasm. But after fertili- zation, the chromatophores all migrate to the surface within half anhour. During cleavage of the egg, they are massed in the cleav- agefurrows. The pigment occurs also in the test of this sea urchin, and gives the animal the characteristic color, which varies from a bright red (especially in young individuals) to a dark red, and may be almost black in old specimens. In reference to the fact that the pigment may be caused to leave the chromatophores and pass into the cytoplasm and thence into the medium, the following questions may be asked: (1) How is the pigment held in the chromatophores? (2) What is its func- tion? (3) What is its chemical nature? The present paper is eoncerned with these questions. 435 436 Echinochrome HISTORICAL. Echinochrome was studied spectroscopically by McMunn,! who found it in the elaeocytes of the sea urchins, Strongylocentrotus lividus, Amphidotus cordatus, Echinus esculentus? and E. sphaera. The spectrum showed faint absorption bands, which ¥aried with different solvents and different reac- tions of the same solvent. McMunn thought that he noticed changes in the spectrum on the addition of powerful reducing agents, such as stannous chloride, and concluded that echinochrome functioned as an oxygen carrier. However, the absorption bands in its spectrum are difficult to make out except in absolute alcohol (or glycerine) and in this solvent I observed that stannous chloride caused a precipitation of the pigment, which interfered with the examination. A. B. Griffiths? attempted an elementary analysis of the substance. He dried the elaeocytes and extracted them with chloroform, benzol or carbon bisulphide. On evaporation of the solvent he analyzed the substance with- out further purification, although evidently it contained many impurities. From four analyses, he deduced the formula Cyo2HssNj2FeS.Oi2, which would make C = 67.8 per cent H = 5.5 per cent, and N = 9.3 per cent. He states that on boiling with mineral acids it is transformed into haemato- porphyrin, haemochromogen and sulphuric acid (E + acid = 2C3,H3sN,Os + C3sHxNFeO; + H.SO,). Griffiths agrees with McMunn that echino- chrome is an oxygen carrier, and states that the oygen is held rather firmly, and in nature is removed only by the reducing action of the cell containing the pigment. EXPERIMENTAL. The pigment in the elaeocytes, eggs and tests of Arbacia, shows no absorption bands, but after extraction it shows very similar bands in its spectrum to those described for echinochrome by McMunn. He published drawingsof the spectra and measured the wave lengths corresponding to the edges of the bands. It is well known that bands become broader as the solution is more concen- trated, and for that reason I measured the wave length of aline of the spectrum corresponding as nearly as could be determined to the center of each band. By taking the mean between the wave lengths of the edges of the band in McMunn’s data I have com- pared his with mine. The discrepancies may be accounted for, *MeMunn: Quart. Journ. Micro. Sci. (2), xxv, p. 469, 1885; xxx, p. 51, 1889. ? Griffiths: Compt. rend. soc. biol., exv, p. 419, 1892; Proc. Roy. Soc. Edin- burg, xix, p. 117, 1892: Physiology of the Invertebrata, New York, 1892; Respira- tory Proteids, London, 1897. J. F. McClendon 437 first by the fact that the mean is not the exact center of the band in a prism spectrum, and secondly there is a personal equation in observation. I found the pigment extracted from elaeocytes, eggs or tests to give about the same spectra, though a few isolated obser- vations seemed to vary. These might have been due todecom- position products with different spectra. My data... | 5296 | 4844 | 5504 | 5302 4844 | | 5296 | 4844 a 4844 sae) | 4844 | 5154 | 4844 McMunn | 5512 [5128 4848 | 5370 4998 5205 | 4848 Neither McMunn nor Griffiths succeeded in crystallizing echino- chrome. Dr. A. P. Mathews had observed that on the addition of iodine in potassium iodide (KI3) crystals form easily. In 1910 I obtained quantities of these crystals, but did not succeed in recrys- tallizing them without great loss by the formation of amorphous masses. The iodine compound in absolute alcohol showed an additional, but very dim band in the spectrum (wave length 5628 or 5696). It crystallized in red or orange needle-like crystals, triangular in cross section, sometimes rhombic in side view and often forming rosettes. They were but slightly soluble in water ‘unless hot or containing acid, soluble in absolute alcohol (the rhombic crystalsseeming more soluble than the needles) and slightly soluble in ether. If a solution in water is shaken with ether the latter is not colored. If an alkali is added to the KI; solution no crystals are formed (due to combination of the base with the echinochrome) but HCl does not prevent their formation. Some of this iodine compound which was kept for several months in a dry state became more soluble in ether and crystallized in flat thin, red or orange rhombic plates. Perhaps the substance had decomposed with the liberation of iodine, for I succeeded in crys- tallizing the mother substance and obtained the same plates, in addition to red or orange needles, never triangular in cross section, but sometimes forming rosettes. I extracted echinochrome from the tests with strong, slightly acidulated alcohol and purified it by repeated precipitation with 438 Echinochrome alkali and solution in acid alcohol, and filtration. Finally I dissolved the precipitate in water plus HCl, filtered and shook the solution with ether. The ether did not remove all of the echino- chrome and the formation of haptogen membranes caused much loss of material. The ether was evaporated at room temperature, as heat seemed to decompose the substance. Occasionally a few crystals formed at the edges of the solution but the main mass of the residue was amorphous. The next season (1911) I tried to purify echinochrome without the use of acids or alkalies. The body fluid of the sea urchins was allowed to clot and the elaeocytes thus obtained were placed directly into acetone, which extracted the pigment. The extract was filtered and evaporated at room temperature. The residue was washed with carbon tetrachloride (which does not easily dis- solve echinochrome) to remove fats, and again dissolved in the smallest quantity of acetone and filtered to free it from traces of lecithin. This solution was evaporated, dissolved in absolute ether and filtered to remove salts, evaporated to constant weight and analyzed by Dennstedt’s method. A mean of two analyses gave: C = 51 per cent, H = 7.7 per cent. The echinochrome purified by precipitating with alkali gave C = 53.3 per cent, H = 4.4 per cent, N = 1.5 per cent. The nitrogen was deter- mined by Kjeldahl’s method and therefore may not be reliable, — since the constitution of the molecule is unknown. Traces of sulphur and phosphorus, possibly due to impurities were found, but no iron. ‘The ether-soluble crystals from spontaneous decomposi- tion of the iodine compound gave C = 57.9 per cent, H = 6.5 per cent. It was stated above that echinochrome is precipitated by alkalies in alcohol. I precipitated echinochrome with NaOH in 95 per cent alcohol and washed in the same alcohol to remove the excess of NaOH. From the amount of NaOH that was neutralized by the pigment I concluded that it combined with from 18 to 25 per cent of Na. Analysis gave C = 31.5 per cent, H = 6 per cent, Na = 19.5 percent. Therefore we may say that the echinochrome behaves as an acid, or else is amphoteric. The former view is 3 Alkali does not precipitate it in water; the particular base was immate- rial, ammonia was added but the presence of sea salts allowed the libera- tion of other bases. J. F. McClendon 439 supported by the fact that on passing an electric current through the aqueous (colloidal) solution, the echinochrome shows a nega- tive charge (is anodic) and again, if histological sections are placed in such a solution the acidophile portions are stained more strongly than the remainder. In fact its behavior is very similar to that of a weak solution of eosin, except that it is very easily washed out by alcohol. However the substance is probably amphoteric (the acid charac- ter being stronger than the basic) since its aqueous solution is precipitated by phosphomolybdic and phosphotungstic acids but not by tannin. From the analyses given above it would seem that no one has succeeded in obtaining echinochrome in a reasonably pure state. It is very unstable and probably breaks up into a host of decompo- sition products all having practically the same spectrum. [If it is kept in the dry state for a great length of time, or is evaporated on a bath not over 50° for a shorter time, part of it becomes insol- uble in ether but not in alcohol. When heated in the combustion tube it first stiches then boils and sublimes as crystals on the top of the tube, then very soon turns brown and chars. After being crystallized from a solution in ether the crystals often become smaller and irregular in outline. Perhaps the crystals evaporate or lose water of crystallization, but I think that both these possibilities are improbable. The crys- tals may decompose into an amorphous substance. On first obtaining crystals, I feared that they were crystals of some other substance merely colored by echinochrome, but this: seems impossible from later observations. In extractions made for the purpose of studying the lipoids in Arbacia eggs, red or brown substances (echinochrome or its decom- position products) appear in every fraction, rendering analysis difficult and indicating the instability and wide solubility of the substance. In order to test the statement that it is an oxygen carrier I separated the cells from 50 cc. of body fluid by the centrifuge, and mixed them with sea water to make 50 cc. This suspension, and 50 cc. of sea water as a control, were placed in two similar graduated tubes. The air was pumped out for six hours (until the water boiled), air was then admitted and the tubes sealed. They 440 Echinochrome were shaked one-half hour and the volume of air measured at atmospheric pressure. The suspension had lost 1.25 ec., the con- trol only 0.8 cc. In another experiment the suspension lost 0.95 ee. and the sea water 0.8 cc. It was thought that in the absence of oxygen the cells would take the oxygen from the echinochrome. However no color change could be observed with the naked eye or the spectroscope, and the greater absorption of air by the sus- pension may have been entirely due to oxidation in the cells. In similar experiments, with an aqueous solution of the pigment, and distilled water for a control, and using pure oxygen, the two tubes gave the same absorption, as shown by two examples: Oxygenvabsorbed: by HsOt2eis sicirrt atin eer des A Oxygen absorbed by echinochrome...................... The question, how echinochrome is held in the chromatophores, cannot be fully answered. The chromatophores when free from pigment are highly refractive and stain strongly with the intra- vitam stain, neutral red, and when fixed they stain strongly with Delafield’s haematoxylin, indicating a lipoid nature. The pig- ment may be in solution in the lipoid. The fact that the spectrum is different (shows no bands) in life from the spectrum of the extract may indicate chemical com- bination of the pigment with the chromatophores. The fact that echinochrome stains acidophile tissue may show a possible mode of such combination, if it be found that the chromatophores contain bases. However I do not think we can rely on the spectroscopic evidence, for the absorption bands are very faint in aqueous solu- tion unless it be alkaline, and the cells containing the pigment intertere with the passage of light and make the observation diffi- cult. I have never seen absorption bands in echinochrome ex- tracted from the fresh cells with distilled water. The same state- ment is made by McMunn. If the substance is held by chemical combination why does it come out so easily? The same argument may be made against the possibility that the echinochrome is held in the chromatophores because it is more soluble in them than in water. When the cell is stimulated me- chanically or chemically the pigment comes out of the chromato- J. F. McClendon 441 phores with explosive rapidity. The cell need not be killed to accomplish this. The mere act of normal fertilization causes some of the chromatophores in the egg to lose their pigment. The only alternative hypothesis I know of is, that the pigment is manufactured in the chromatophore, and cannot normally get out ~ because the surface of this body is impermeable to it. An increase in permeability of the chromatophore allows the pigment to escape. Such an increase in permeability might be due to an aggregation change in the colloids of the limiting membrane or surface film. Echinochrome is held in the chromatophores of the sea urchin’s cells probably in the same way that chlorophyll is held in the chro- matophores of the green plant cell. ‘ re r ee CRRA | . : aie SLAMS Sg ; 2 Seta avi gO wi dain men's Dre ak Ne Ps. ed Satie SS we Ho tha Arabs. sf "hel ee ‘ | Li rans gus aon tae & . vat e w dactantas Apis SL NA, BP? sah eth kos ii, ee alae porn " ij : ut MAB, 2: i Seah peti a to al Iie bisa, re et aSHe AGF * t vidiec diab Por a mye tha ‘ a} wendy at me hi THE PHYSIOLOGICAL ACTION OF SOME PYRIMIDINE COMPOUNDS OF THE BARBITURIC ACID SERIES. By ISRAEL S. KLEINER. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven, Connecticut.) (Received for publication, March 27, 1912.) Aside from the fact that. certain pyrimidine compounds are con- stituents of the nucleic acid molecule, their possible biochemical importance is attested by a structural relation to the purines, creatine, creatinine, allantoin and other compounds of physiologi- eal interest. Moreover a few pyrimidines are known to have a marked pharmacological action. The first substance of this type used in physiological experiments was alloxantine, which is formed by the reduction of alloxan. Beet HN—CO 06——_NH | bey | | 2 e . +H oc CcH—o—C—OH CO Re oa | HN—CO HN—CO oc-——-NB Wohler and Frerichs* fed 5 to 6. grams of this tomen but could not recover any in the urine; nor was alloxan found. The urine was rich in urea, and a breaking down of alloxantine to urea and other products was believed to be probable. Ne mention is made of any toxic effects, although if any had been experienced they would undoubtedly have been described because the subjects were human beings. Koehne? fed alloxan and alloxantine in 8-gram doses to dogs. Each caused a mild diarrhea: without othersymptoms.. No alloxan or alloxantine was excreted in the urine; but small amounts of oxalic and parabanic acids were found. Working independently of Koehne with the same compounds Lusini’ obtained results different in some respects at least. In his experi- 1 Wohler and Frerichs: Ann. d. Chem. u. Pharm., |xv, pp. 335-349, 1848. 2 Koehne: Inaugural Dissertation, Rostock, 1894, 39 pp. 3 Lusini: Ann. di chim. e di farmacol., xxi, pp. 145-160, 1895; pp. 41-257; and xxii, pp. 341-351, 1895; pp. 385-394: from Chem. Centralbl., 1895, i, p. 1074; ii, p. 838. 443 THE JOURNAL OF BIOLOGICAL CHEMISTRY, XI, NO. 5 444 Action of Certain Pyrimidines ments he found that both of these substances attacked the skin of frogs, dogs and rabbits. Both acted upon the cerebro-spinal centers, this action being divided into two periods, (a) hyper-reflex-excitability followed by rigidity, and (b) hypo-reflex-excitability and paralysis. This second stage had previously been noted by Curci,‘ who because of the use of a larger dosage had overlooked the first stage. These and other minor effects varied slightly with the compound used, alloxan being in general more toxic than alloxantine. Alloxantine strongly reduced the hemoglobin of the blood both in vitro’ and in vivo. Among other phenomena produced in frogs by alloxan, mydriasis is noteworthy. Alloxan also had a powerful influence on the heart; the contractions were diminished in vigor, diastolic pauses lengthened, and finally the heart stopped in diastole. According to Lusini, alloxan was non-toxic when given per os, 0.5 gram being easily borne. It did not reappear in the urine but parabanic acid and alloxantine were found. When Lusini fed alloxantine he recovered only slight traces in the urine. A small amount of dialuric acid was found together with parabanic acid and murexide in larger quantities. Lusini HN | reached the conclusion that the group, OC is able to stimulate and then : | HN HN—CO inhibit the nerve centers, and that the grouping, O = has no such power. It is, according to Lusini, the ketone-like CO which seems to have the stimu- lating property, and the abundance of these groups increases the toxicity of alloxan. More recently Steudel® has attempted to ascertain whether pyrimidines may be built up to purines in the animal body. The compounds used in- cluded those which Behrend and Roosen? had described as intermediate products in the synthesis of uric acid in the chemical laboratory. At the outset it may be stated that a purine synthesis in vivo was not established. Steudel fed the substances to a bitch weighing 6.2 kg. in doses of 1 gram per day with meat and attempted to isolate them or their purine deriva- tives in the urine. 4-Methyluracil and 5-nitrouracil were found unchanged in the urine. 5-Nitrouracil-4-carboxylic acid, however, did not reappear in the urine. Steudel believes that it underwent a complete decomposition in the organism, although he does not consider the possibility of the non- absorpti-n from the alimentary tract and does not report any analyses of 4Curci: Cited by Lusini: Ann. di chim. e di farmacol., xxi, pp. 145-160, 1895; from Chem. Centralbl., 1895, i, p. 1074. 5 This property was described by Kowalewsky: Centralbl. f. d. med. Wis- sensch., xxv, pp. 1-3, 17-18, 1887. 6 Steudel: Zeitschr. f. physiol. Chem., xxxii, pp. 285-290, 1901. 7 Behrend and Roosen: Ann. d. Chem., ccli, pp. 235-256, 1889. Israel S. Kleiner 445 the feces. Of the following pyrimidines none was recovered in the urine after feeding, nor was any difficultly soluble condensation product detected : isobarbituric acid, isodialuric acid, thymine and uracil. This author points out the striking difference in behavior between thymine (5-methy]-2,6- dioxypyrimidine) and 4-methyluracil. Structurally they differ only in the position of the methyl group but the former is broken down in the body while the latter is not. If, however, a nitro group is substituted for the methyl in thymine the physiological character of the pyrimidine is reversed; for it now passes unchanged through the kidney. Although a purine synthesis could not be demonstrated, Steudel deter- mined to extend his experiments with other pyrimidines, namely, 2,4-di- HN—CO pal amino-6-oxypyrimidine, H3AV—C CH and 2,4,5-triamino-6-oxypy- HN—CO | | rimidine, NG eval: which Traube® had obtained as intermediate iM —N ile products in his synthesis of guanine. Both were administered as the sul- phates in 1-gram dose in the manner above described. Both were reported to be toxic, which was surprising inasmuch as none of the other compounds had been accompanied by any untoward symptoms. Feeding of the 2, 4-diaminopyrimidine was followed by vomiting and the triamino compound provoked equally serious disturbances. About an hour after the substance was taken, there occurred attempts at vomiting without any vomitus being ejected. The animal had no appetite and lay on one side almost all day. The urine contained protein, hyaline cylindroids and the unchanged tri- amino compound. The last was recovered as the sulphate and identified by tHe violet color produced by saturating it with ammonia. By subcu- taneous injections the lethal dose for rats was determined as 0.2 gram for 2,4-diamino-6-oxypyrimidine sulphate and 0.1 gram for 2,4,5-triamino-6- oxypyrimidine sulphate. Autopsy of the rats poisoned with the diamino substance revealed nothing characteristic; but the kidneys of the animals which had received the triamino compound contained numerous concre- tions and resembled microscopically the kidneys of dogs poisoned with adenine.? From these results, Steudel concluded that the attachment of amino groups to the pyrimidine ring transforms harmless, indifferent substances into poisonous ones. The toxicity of adenine, 6-aminopurine, he regards as an analogous phenomenon in the purine series. He believes that an examina- tion of other amino derivatives of the pyrimidine and purine compounds will prove the universality of this law. No analyses of the two amino- 8 Traube: Ber. d. deutsch. chem. Geselisch., xxxili, pp. 1371-1383, 1900. ® Minkowski: Arch f. exp. Path. u. Pharm., xli, pp. 375-420, 1898. 446 Action of Certain Pyrimidines pyrimidines fed are presented by Steudel, nor are any data as to their solu- bility given. In a later contribution Steudel!® reported the investigation of other mem- bers of this series. Pseudo-uric acid and isourie acid did not result in a purine synthesis although both have been transformed into uric acid in vitro. A similar result was obtained when hydrouracil was fed. 2-Thio-4-methyl- uracil, like 4-methyluracil described above, was quickly excreted in the urine. 2-Amino-4-methyluracil, which differs from thiomethyluracil only in the substitution of an amino group for sulphur in the 2-position did not appear in the urine, nor was any other characteristic product found. Steu- del concludes that none of the pyrimidines used by him are adapted to a synthesis of purine compounds in the dog. The pharmacological action of some pyrimidines was studied by Fischer and von Mering"™ with interesting results. They discovered that certain alkyl derivatives possess an action similar to that of sulphonal. The latter, diethylsulphone-dimethylmethane, is rich in alkyl groups; and it was the idea of these authors to experiment with other alkyl organic compounds, many of which Fischer had synthesized, in the hope of ascertaining the essential or most effective groupings for hypnotic action. Of especial inter- est are the cyclic compounds employed, which are derivatives of barbituric acid and of malonyl guanidine. It was found that the 5-monoalkyl deriva- tives of barbituric acid have no hypnotic action, nor has the 5, 5-dimethyl derivative; but when both hydrogen atoms in the 5 position are replaced by alkyl groups, at least one being higher than methyl, the compound acquires sleep-producing powers. This reaches its maximum in 5, 5-dipro- pylbarbituric acid. Some of the compounds studied proved toxic; for example, substitution of the H in the 1 position by CH; or of the O in the 2 position by S transformed 5,5-diethylbarbituric acid into a toxic com- pound. However, 5,5-dipropylmalonyl guanidine, HN—CO as | | AGH : HN=C og | ENCgEI, HN—CO well as diethylmalonuric acid, HN COOH had no pharmacological | ols OG. <¢ | | \osHs HN—CO effect. 5, 5-Diethylbarbituric acid, HN—-CO has been used widely in | | ACeHs OC oe | | ‘CaHs HN—CO 10 Steudel: Zeitschr. f. physiol. Chem., xxxix, pp. 136-142, 1903. 11 Fischer and von Mering: Therapie der Gegenwart, N. F., v, pp. 97-101, 1903. Israel S. Kleiner 447 medicine as an hypnotic under the name ‘‘veronal,’’ its sodium salt as ‘‘medinal,’”’ and the dipropyl compound to a less extent as “‘proponal.’’ Fischer and von Mering™ have found that most of the veronal is excreted from the body unchanged. Recently P. Fischer and Hoppe, Bachem," Groéber,’® and Jacobj'* have added many new facts to the literature of veronal. Wolf? injected uracil, thymine, and cytosine in 10 to 50 mgs. doses into the circulation of cats, but observed no effect upon arterial pressure, intesti- nal volume, respiration, or rate of blood-clotting. Sweet and Levene!’ fed thymine to a dog with an Eck’s fistula (on the basis of Steudel’s contention that thymine is destroyed by the normal dog). A marked diuresis resulted and thymine was found in the urine in considerable amount. This is inter- esting in view of the close relationship between this methylated pyrimidine and the methyl substituted xanthines: theophylline, theobromine, and caffeine which are also diuretics. Mendel and Myers! have however recently shown that thymine is not completely destroyed normaliy by the dog, nor is uracil norcytosine. Diure- sis was not observed by them after the administration of thymine. The output of purines, creatinine and urea + ammonia was not influenced by administering any of these to rabbits, dogs ormen. None of the compounds had any marked pharmacological effects. This is especially interesting because cytosine is an amino-pyrimidine, closely related to the compound alleged by Steudel to be toxic. EXPERIMENTAL PART.?° Dogs, rabbits, and guinea pigs were used in the physiological studies. The dogs were not catheterized, as the time relations were not of interest; but in the case of rabbits the urine was removed by artificial means in some experiments. The dogs’ food always was mixed with bone so that the feces became firm and did not contaminate the urine. For the same reason the rabbits and guinea pigs were given some grain in addition to carrots. The compounds were administered. subcutaneously, intraperitoneally, or by mouth. 12 Fischer and von Mering: Therapie der Gegenwart, April, 1904. 18 P. Fischer and Hoppe: Miinchener med. Wochenschr., 1909, p. 1429. 144 Bachem: Arch. f. exp. Path. u. Pharm., \xiii, p. 228, 1910. 16 Grober: Biochem. Zeitschr., xxxi, p. 1, 1911. 16 Jacobj: Arch. f. exp. Path. u. Pharm., xvi, p. 241, 1911. 17 Wolf: Journ. of Physiol., xxxii, pp. 171-174, 1905. 18 Sweet and Levene: Journ. of Exp. Med., ix, pp. 229-239. 1907. 19 Mendel and Myers: Amer. Journ. of Phystol., xxvi, pp. 77-105, 1910. 20 The experimental data in this paper are taken from the writer’s dis- sertation for the degree of Doctor of Philosophy, Yale University, 1909. 448 Action of Certain Pyrimidines The substances used were barbituric acid and its amino-deriv- atives. HN—CO HN—CO HN-—-CO ket frend fecal OC CHe HN=C CH: ' HN=C -CHNHs ba HN—CO HN—-CO HN—CO Barbituric acid Malony! guanidine 5-Aminomalony] (Malonyl] urea) guanidine OO) HN—CO | haga i H:sNC CH H2aNC CNHe | Il ll N—CNHe N—CNHo2 2, 4-Diamino-6-oxy- 2, 4, 5-Triamino-6-oxy- pyrimidine pyrimidine Barbituric Acid. Barbituric acid was made essentially according to Michael’s?! method, the principle of which consists in condensing urea with diethylmalonate in the presence of sodium ethylate. The yield represented 71 per cent of the theoretical and the acid obtained gave the following results on analysis (Kjeldahl-Gunning method). Caleulated for CiHaN203: Found: IN ee ook. ea ene ee 21.87 21.87 21.657 21°53— Barbituric acid crystallizes in two forms, the anhydrous as needles, and the hydrated as rhombic prisms. It is slightly soluble in water. A rough solubility determination showed that a 2.68 per cent solution can be prepared at 40 to 43°. Efforts to obtain a method for estimating barbituric acid quanti- tatively in urine were unsuccessful, although a qualitative color test afforded a means of getting rough values colorimetrically. . The difficulty lies in the fact that many of the properties of bar- bituric acid are possessed also by hippuric acid. As the latter 21 Michael: Journ. f. prakt. Chem., (2), xxxv, pp. 449-459, 1887. 22 These two analyses were made several months later. Israel S. Kleiner 449 occurs constantly in all ordinary urines, and in considerable amount in the urine of herbivora, it proved an effective bar. Both com- pounds are precipitated by mercuric sulphate and by silver nitrate; both are soluble in ethyl acetate and amyl alcohol, and insoluble in ligroin, petroleum ether and benzene. Thecolorreaction referred to is based on Baeyer’s* observation that nitroso-barbituric acid, in the presence of ferrous acetate, yields a deep prussian blue color. The directions for this test are as follows: 3 cc. of urine are treated with three drops of 2 per cent sodium nitrite solution; about 0.5 cc. of 10 per cent sulphuric acid is added and the solution is now made alkaline with sodium carbonate solution; on addition of one or two drops of strong ferrous sulphate solution a beautiful blue appears in the presence of barbituricacid. When the expression ‘‘NaNO-FeSO, reaction” is employed hereafter it will be understood that this test is meant. Other members of this series give this reaction but thy- mine, cytosine and uracil do not.** Since urine frequently assumes a deeper color when subjected to this treatment, a direct colori- metric estimation was not attempted but a crude method was worked out, in which the greatest dilution allowing a positive test was considered the standard. It was thus found that a 0.0023 per cent solution of barbituric acid in water is the limit for this test, and hence the standard for comparison. Barbituric acid is precipitated by mercuric sulphate solution. It gives Jaffé’s reaction as applied to creatinine. A red color results when ferric chloride solution is added to barbituric acid. The sodium salt was made by dissolving the acid in the amount of NaOH calculated to form the disodium salt, concentrating and allowing the salt to crystallize. Needle crystals were obtained; but that they were probably a mixture of the mono- and disodium salts is evident from the nitrogen determination. 23 Baeyer: Ann. d. Chem. u. Pharm., cxxvii, pp. 199-236, 1863. 24 None of the compounds of the barbituric acid series give the character- istic reactions of uracil, thymine or cytosine. For example, if thymine in substance be treated with diazobenzol-sulphonic acid a reddish purple color . results; tested in the same way barbituric acid gives a red, malonyl guani- dine and cyanacetylguanidine a deep orange and the others a yellow or green color. When uracil or cytosihe is dissolved in about 5 cc. of water, bromine water added in slight excess and the solution boiled, a deep purple precip- itate results on the addition of baryta water. None of the barbituric acid series studied gives this test. 450 Action of Certain Pyrimidines Calculated for Calculated for CiH2N203Na2: CaHsN203Na: Found: DY See Sn a ee ee . et AT no ee 18.66 17.58 As illustrations of the general method employed in the animal experiments two typical protocols will first be given. EXPERIMENT 1. A rabbit weighing 2 kg. was given 0.519 gram barbituric acid in about 25 cc. of water at 40°, hypodermically. The urines of the next two days were precipitated with mercuric sulphate, the precipitate decom- posed with hydrogen sulphide and, after removing the mercuric sulphide, the colorimetric determination made. The amount excreted was estimated at 0.026 gram. No hypnotic or toxic action was exerted by the compound. Its acidic character, however, made it harmful to the tissues at the point of injection; this caused an opening in the body wall which led to the death of the animal seven days after the injection. EXPERIMENT 17. 0.64 gram of sodium barbiturate in 45 cc. water contain- ing 0.1 cc. 7; NaOH at 38° were injected intraperitoneally into a rabbit weighing 1.6 kg. Diarrhea resulted in about two hours and this condition persisted for five days. The urines of the first two days gave positive NaNO,-FeSO, tests and these corresponded to 0.04 gram of barbituric acid. These as well as other experiments with barbituric acid are tabulated on the opposite page. From this table it is seen that the fatal termination of Experi- ments 1 and 3 must be ascribed to the acidic properties of barbi- turic acid; for when larger amounts of the sodium salt were given as in Experiments 12 and 17 no toxic effects resulted. The only physiological effect, which may be ascribed to its structure, is its diarrheal action; but a greater number of experiments need to be done to settle this point. In this connection it is interesting to recall the fact, noted above, that Koehne* observed a mild diar- rhea after feeding alloxan and alloxantine. Again, the fact that barbituric acid has no hypnotic action harmonizes with Fischer and von Mering’s* experiments on substituted barbituric acids, in which, as detailed above, they found that the lower the substi- tuted alkyl groups, the less hypnotic the influence possessed by * the complex. In barbituric acid, the lowest degree is reached and no hypnotic action is observed. 26 Koehne; Inaugural Dissertation, Rostock, 1894. 26 Fischer and von Mering: Therapie der Gegenwart, v, pp. 97-101, 1903. Israel S. Kleiner 451 TABLE I. Animal Experiments. Barbiiuric Acid (Malonyl urea). AMOUNT GIVEN MODE OF ADMINISTRATION HSS ES ROT SES | Subecutaneously | Not toxic except for necrosis at point of injection, which caused death seven days later. Some excreted. (2) Rabbit...| 2.1 0.32) 0.15 | Subcutaneously | Not toxic. Exeret- | ed about one-third (?) (3) Rabbit...| 2.2 | 0.53) 0.24 Intraperitoneally} Death in three days. Diarrhea at first. Diminished flow of urine (28 cc. in two days) containing 0.09 gram (?).Au- topsy revealed fib- rinous adhesions in peritoneal cavity. 0.10 | Intraperitoneally | Recrystallized prep- aration used. Not toxic. Under ob- servation fifty-six (1) Rabbit...| 2.0 (4) Rabbit...| 1.9] 0.2 days. (6) Rabbit...} 1.9} 0.6/0.3 | Per os.. Marked diarrhea. Excreted about 3) in urine. (12) Guinea 0.5; 0.3 Subcutaneously | Na salt used. Not pig.... ° toxic. Excreted 0.01 gram. (?) Intraperitoneally | Na salt used. Ex- creted about 0.04 gram (?) Diar- rhea for five days. otherwise not tox- ic. Under obser- vation thirty-one days. } a3 (17) Rabbit..| 1.6 | 0.64 452 Action of Certain Pyrimidines Malonyl Guanidine. In synthesizing malonyl guanidine Michael’s?? procedure was essentially followed. The pyrimidine was obtained in the form of its sodium salt which was dissolved in water and dilute NaOH, and the free pyrimidine precipitated with acetic acid. Malonyl guani- dine crystallized in fine white needles which, after drying in a desiccator, were analyzed for nitrogen. Calculated for CsHsN202+H20: CsHsN202: Found: INE ec ree. os os eee es 28 .96 33.07 32.05 31.91 The low nitrogen values are probably due to incomplete removal of the water of crystallization by simple desiccation. Inasmuch as the analysis was fairly close and the preparation was pure white no further purification was attempted. It was only slightly solu- ble in water. At 40 to 43° a 0.049 per cent solution was the strong- est obtainable. This, of course, renders malonyl guanidine itself unsuitable for injection experiments and the sodium salt was accordingly used. In preparing this, the pyrimidine was dissolved in NaOH, as little in excess of the calculated amount as would bring about solution being used. On concentration, fine pale pink needles crystallized out. From the analyses, which follow, this salt must contain four molecules of water of crystallization which are lost in the desiccator very slowly. Calculated for Found: CsHsNaN3024+4H20: = (air-dry) ING ee ete cl. 2: le eedee. eee 19.00 18.77 Calculated for Found: , CsH«NaN30:2: (desiccated) PRU Nic Ste, A Sr iri mee Be EC ol oto 28.19 25.31 A method for recovering malényl guanidine from urine is at once suggested by the slight solubility of the free substance. However, if urine is acidified and allowed to stand, uric acid and, if concentrated sufficiently, hippuric acid will also crystallize out. The NaNO,-FeSO, reaction described above for barbituric acid is also applicable to malonyl guanidine. The limit for this test in urine is 0.004 per cent. Another mode of estimation by means of this color reaction was tried as follows: 0.002 gram in 27 Michael: Journ. f. prakt. Chem., xlix, pp. 26-43, 1894. Israel S. Kleiner 453 3 cc. water was converted to the prussian blue compound and dilu- tions made until the blue was no longer distinctly discernible in a 100 cc. cylinder. The concentration just above this was con- sidered the standard. By sucha rough method it was found that a distinct blue can be seen when there is an amount corresponding to 0.0004 per cent present. Sodium malonyl guanidine is not precipitated by ammoniacal silver nitrate solution, but is precipitated quantitatively by mer- curic sulphate solution. With picric acid and alkali a red color is formed as in Jaffé’s test for creatinine. From the animal experiments (see Table II) it is seen that mal- onyl guanidine is non-toxic, at least in the doses for the animals used. The failure to detect the substance or a related compound in the urine of Experiment 5 may be due to the small amount injected. TABLE II. Animal Experiments. Malonyl Guanidine. AMOUNT GIVEN NUMBER AN eee x REMARKS AND RESULTS = MODE OF | ADMINISTRATION (5) Rabbit. . .| | 0.09 | 0.04 | Subcutaneously | Sodium salt used. | | | No effects. Not | detected in urine. | 1.9 | 0.41 | 0.21 | Subcutaneously | Sodium salt used. tected in urine. (9) Rabbit... 12.1 |-0.22 | 0.10 | Subcutaneously | Sodium salt used. | Noeffects. De- (26) Rabbit. | Mild diarrhea; no | other effects. All (?) excreted in | | * urine. (24) Dog.....| 10. | ZL | 0:21 | Per as Free malonyl guani- | dine used. Some absorbed and ex- j . . | ' ¢reted in urine. | | No toxic effects. 454 Action of Certain Pyrimidines 5-Aminomalonyl Guanidine. This compound is quite difficult to obtain in good yield as it decomposes very easily. The most advantageous method was found to be a modification of one described by Traube?® in which the sulphate of this compound can be prepared directly from mal- onyl guanidine. The directions of Traube were followed as far as the formation of 5-aminomalonyl guanidine sulphate by reduc- tion with H2S, but instead of extracting this salt with hot water, the sulphur was removed by means of CS, and the sulphate con- verted into the hydrochloride by treatment with BaCl,. Traube’s suggestion of adding alcohol to induce crystallization was not found to be advantageous since the crystals, when finally obtained, had a pink tinge. Consequently, the fluid was concentrated under diminished pressure and allowed to crystallize. Light yellow ros- ettes of needles formed very slowly. Analysis of this preparation (A) by the Kjeldahl-Gunning method showed that, in spite of the tinge of yellow color, the salt was quite pure. Another preparation (B) made by the same method gave a higher nitrogen percentage. Calculated for (CsHeN.O2) HC1+H:20: Found: A desiccated 28.55 Nee ee ee is. 28 .57 f desiccated 28.08 B air-dry 29.41 Its solution, which is acid to litmus, very quickly turns red, owing undoubtedly to a slight oxidation. It stains the tissues red and has a faint disagreeable odor. Boiling with NH,OH yields a solu- tion colored like potassium permanganate and this changes to dark blue on addition of KOH. It will give the NaNO.-FeSO, reaction, but not readily or brilliantly. The Jaffé color reaction for creatin- ine is not given by this salt. It is precipitated both by ammoni- acal silver nitrate and mercuric sulphate, but as very little can be injected into an animal and as it was found to be toxic no attempt was made to isolate it from urine. The toxicity of this compound is shown in the following illus- trative protocols and the accompanying table (Table III) which summarizes all the experiments. The hydrochloride was used in each case. 28 Traube: Ber. d. deutsch. chem. Gesellsch., xxvi, pp. 2551-2558, 1893. Israel S. Kleiner 455 EXPERIMENT 7. November 30. 3:10 p.m. A rabbit weighing 1.4 kg. was given subcutaneously 0.37 gram in 35 cc. water. 3:20 p.m. Has defecated very soft stools. Moves around restlessly. 4:15 p.m., 4:50 p.m. Apparently well. December 1. 8:45 a.m. Rabbit found dead. Autopsy: kidneys are very light colored; intestines intensely reddened; liver, light brown; large amount of bloody fluid in peritoneal cavity. EXPERIMENT 23. March 11. 11:00a.m. A guinea pig weighing 450 grams was given 0.036 gram of the salt in about 10 ce. water subcutaneously. March 12. 2:15 p.m. Has eaten 60 grams carrots and 9 grams oats. Urine, 64 cc., alkaline; specific gravity, 1.017; albumin present, but no casts. March 18. 2:40 p.m. Has eaten 70 grams carrots and 3 grams oats. Urine, 43 cc.; alkaline; specific gravity, 1.024; large amount of albumin; granular, granular partly hyaline, and cellular casts found; NaNO,-FeSO, test negative. Marchi4. 2:35p.m. Has eaten 85 grams carrots and 4 grams oats. Urine 30 ce.; alkaline; albumin present; casts. Marchi5. 2:50p.m. Has eaten 93 grams carrots and 3 grams oats. Urine 57 ec.; alkaline; specific gravity, 1.021; albumin; casts. March 16. 8:40 a.m. Animal appears well. Weight 390 grams. The animal daily ate more food until March 20, when the usual amount (150 grams carrots and 15 grams oats) was entirely consumed. On March 18, its weight had dropped to 360 gram but then rose to 440 grams on March 24. The urine still contained a trace of albumin. On Apri! 3—twenty-three days after the injection—the animal was still living and apparently well. EXPERIMENT 22. March 11. 4:00 p.m. A female rabbit weighing 2.44 kg. was given a subcutaneous injection of 0.19 gram in about 40 cc. water. March 12. 9:00a.m. Stools partly diarrheal. 2:15 p.m. No urine. Has eaten 145 grams carrots but no oats. March 18. 2:40 p.m. No urine. Has eaten 130 grams carrots but no oats. - March 14. 11:00 a.m. No urine, no feces. Has eaten 85 grams carrots and 8 grams oats. March 15. 2:50 p.m. Has eaten 46 grams carrots but no oats. Urine, 163 ce.; alkaline; specific gravity, 1.018; NaNO.-FeSO, test negative; albu- min and granular casts present; slight reduction of alkaline copper solu- tion (after removing albumin). March 16. 8:40 a.m. Has eaten 35 grams carrots. Animal is very weak, breathes slowly and can not hold its head up. 10:05 a.m. Breathes more quickly but head is on floor of the cage. 11:50 a.m. Still breathing; extremely weak. 2:10 p.m. Found dead. Urine, 52 cc.; alkaline; specific gravity, 1.010; albumin and casts present; reduction positive. Autopsy. Weight 2.26 kg. All viscera hyperemic; blood of liver does not clot readily; kidneys edematous; bladder empty; animal is quite fat. Sections of tissues preserved. 456 Animal Experiments. NUMBER AND ANIMAL (7) Rabbit... (22) Rabbit. . (8) Rabbit...) (10) Guinea pig (27) Guinea pig.... (25) Guinea pigs (23) Guinea Action of Certain Pyrimidines TABLE III. T ; | AMOUNT GIVEN | —_—_—_—_——+ lwerona) Per apa Total kg. gram : 1.4 | 0.37 | 0.26 | Subcutaneously 2.4} 0.19 | 0.08 | Subcutaneously 2.6 0.11 | 0.04 | Subcutaneously 0.41 | 0.05 | 0.12 | Subcutaneously | 0.54 | 0.061) 0.11 | Subcutaneously 0.54 | 0.048 0.09 | Subcutaneously ig....| 0.45 p 0.036 0.08 | Subcutaneously §-Aminomalonyl guanidine. | REMARKS AND RESULTS Fatal in less than eighteen hours. Albuminiuria; casts; glycosuria. Death in five days. Fifty-three cubic centimeters urine in first forty-eight hours. Albumi- nuria until fourth day. No glyco- _ suria. Recovery. Albuminuria. Death in four days. Au- ; topsy: organs ap- | pear normal. Blood | does not clot read- | ily. | Not fed on day of | injection. Albu- minuria; mucus | eylindroid seen. | Fatal in less than two days. Autop- sy: one fetus pres- ent; large amount of bloody subcu- taneous effusion. | Kidneys seem con- | tracted. | Albuminuria for at least seven days; casts and leucocy- tes In urine; recov- ery; under obser- vation twelve days. Albuminuria; casts; recovery. Israel S. Kleiner 457 TABLE 111—Continued. == — = a ne ee ae ae eT | AMOUNT GIVEN NUMBER AND | erro MODE OF REMARKS AND RESULTS ANIMAL Pre | Per ADMINISTRATION ABKS MI | Total | kilo- | | gram No symptoms; no albuminuria. Un- | der _ observation eighteen days. No symptoms; no .| 0.52 | 0.04 | 0.08 | Per os albuminuria. Un- der observation thirty days. PL a eS es oe From these results it appears that a lethal subcutaneous dose for rabbits is 0.08 gram per kg. and for guinea pigs 0.11 gram per kg. It is also evident that when the compound is fed it is not toxic. In Experiment 20, the urine was repeatedly examined for substances giving the NaNO,-FeSO, test but with negative results. The feces, however, in both Experiments 20 and 21 were tinged with pink at times. Probably not enough of the compound is absorbed from the alimentary tract at one time to prove toxic; it may be mentioned, however, that the hydrochloride is fairly solu- ble. That the compound acts mainly on the kidneys is evident from the protocols and the table, but substantiating evidence is - given by the histological examination, made by Professor H. Gideon Wells to whom I am greatly indebted for the following report. EXPERIMENT 22—Rabbit. Kidney. Shows extensive necrosis of the con- voluted tubules, perhaps one-fourth of the tubules seen in section showing total necrosis of the epithelium. The necrotic epithelium desquamates into the lumen of the tubule which it fills up, and all stages of transition from masses of necrotic epithelium to granular and hyaline casts which pack the collecting tubules can readily be made out. These casts, being very abundant and staining intensely with eosin, give the sections a striking appearance. The tubular epithelium where not necrotic is strikingly little altered, some tendency to vacuolization of the cytoplasm being the chief abnormality noted. Glomerules congested, swollen, and in some a little granular material and occasional red corpuscles free in the space outside the tuft; in general the glomerules show relatively little change. There is an occasional small area of interstitial hemorrhage. To summarize, the poison has caused a marked necrosis of the epithelium of the convoluted 458 Action of Certain Pyrimidines tubules, but without affecting other renal structures to any considerable degree. Liver. No definite changes except the accumulation of masses of yellow- ish brown pigment in many of the stellate cells. Spleen. Some of the endothelial cells of the splenic sinuses contain brownish pigment, otherwise no change. The pigmentation of the liver and spleen suggests a hemolytic action by the potson. EXPERIMENT 27—Guinea Pig. Kidney. Shows the same necrosis of the secretory epithelium of the tubules and the same formation of casts as described in Rabbit 22, but very much less marked, only occasional tubules showing the lesion. Liver. No pigmentation or other distinct changes. Spleen. Much more pigment than in Rabbit 22. No other changes. Adrenal. No changes. EXPERIMENT 10—Guinea Pig. Kidney. Granular and hyaline casts are very abundant and conspicuous, although there are fewer tubules showing necrosis than in either of the other specimens. When found it is typical, exactly the same in appearance as in 22 and 27. The casts much more often show desquamated epithelial cells within them. Marked congestion, but no other changes. The constancy of the finding of necrotic tubular epithe- lium in all three kidneys is conclusive evidence that this is a specific effect of the poison given. Liver. No distinet alterations. 2,4-Diamino-6-oxypyrimidine. Both 2,4-diamino-6-oxypyrimidine and its precursor, cyana- cetylguanidine, were used in the experiments on animals. They were made by Traube’s?® method with some modifications. Guani- dine hydrochloride, according to this procedure, is condensed with | eyanethylacetate forming, in part, the pyrimidine; but mainly cyanacetylguanidine, which is easily converted into the pyrimi- dine by alkali. H2N COOC2H; HN—CO HN—CO baw al Lecl HN =C a CHe Sa HNC CHe a me sa fae ee fee od | HN CN HsN CN N—CNH The yield of cyanacetylguanidine was 35.8 per cent of the theo- retical, if this were the sole end-product. The mother-liquor was of a dark red color and on concentration yielded a large amount 29 Traube: Ber. d. deutsch. chem. Gesellsch., xxxili, pp. 1371-1383, 1900. Israel S. Kleiner ‘£56 of material which was used in the preparation of the pyrimidine. The first crop was recrystallized from hot water, pulverized and desiccated. To determine whether the substance obtained was cyanacetylguanidine or the pyrimidine, advantage was taken of the fact that the latter crystallizes with one molecule of water of crystallization while the former is water-free. Calculated for (CaHsN.O)+H20: CsHeNiO: Found: 18 EO) oot Si 12.5 0.0 1.2 This preparation was consequently cyanacetylguanidine with very little, if any, pyrimidine admixture. Nitrogen determina- tions by the Kjeldahl-Gunning method gave low figures, perhaps because some HCN may have been formed and lost or because of a very slight admixture of the pyrimidine. Calculated for 4HsN,O: Found: Ifo 2 2 babe, Selec See eyes 44.44 41.99 41.96 A much better yield is obtained by using guanidine sulphocy- anide in place of the hydrochloride, as the mother liquor in this case is not as dark colored and may be evaporated to dryness without much loss of material. In this modification, when used as a step in the preparation of the pyrimidine, it is not necessary to remove the NaSCN formed until the 2,4-diamino-6-oxypy- rimidine is precipitated as the sulphate, since the latter can be washed free from inorganic salts with water. Cyanacetylguanidine is quite soluble in water—a 2.5 per cent solution being easily maintained at 40°—and is suitable for injec- tion. Cyanacetylguanidine forms a rose red isonitroso compound (or is converted into the isonitroso derivative of 2,4-diamino-6- oxypyrimidine) on adding NaNO, and H,SO, to its solution; as this is quite insoluble it may be isolated from the urine. Accord- ing to Traube the isonitroso compound has an intense yellow or yellowish green color; however, with our preparation the brilliant red compound formed first and did not become yellow until addi- tional acid was used. The color test with NaNO, and FeSO, as described above is also positive for cyanacetylguanidine. For the transformation of cyanacetylguanidine into its isomer, 2,4-diamino-6-oxypyrimidine, it was put into boiling 2-5 per cent THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 5. 460 Action of Certain Pyrimidines NaOH, animal charcoal added, boiled a few minutes, and filtered into a beaker placed in an ice-bath. As some NH; is split off by boiling with alkali in this way, the operation must necessarily be performed quickly. The solution was now made weakly acid with H.SO, and white or yellowish needle-like crystals of the sulphate of the pyrimidine appeared. When recrystallized from hot water large silky, grayish needles were obtained and these were again recrystallized from water in the presence of dilute H,SO, and some charcoal; the crystals resulting were of a light yellow, almost white color. . According to Traube the sulphate, when recrystallized from water, contains one molecule of water of crystallization which is not driven off at 100°. Its composition is (CsHgN,O.). HsSO.+ H,.O. Our preparation agreed in its nitrogen content with this formula, as the following analyses indicate. Calculated for (C4HeN.sO)2°H2SO;.+- H20: Found: IN ee eis: 5: = 5 Sc sera eee 30.43 30.42 4 30.27 This salt is sparingly soluble in water. A rough solubility de- termination showed that at 43° a greater concentration than 0.49 per cent could not be maintained and at a slightly lower tempera- ture much of the substance instantly crystallized out. An aqueous solution gives a positive NaNC.-FeSO, test. Because of the poor solubility no injection experiments were performed. However, Steudel’s*® experiment, in which he reports this compound toxic when fed to a dog, was repeated in exactly the same manner and with the same relative dosage. EXPERIMENT 15. February 16. 10:20 a.m. Bitch weighing 9.6 kg. fed 180 grams chopped meat with bone meal, to which was added 1.55 grams of the sulphate of the pyrimidine. 10:20 to 11:40 a.m. Under observation almost continually. The animal, which has always been playful, shows no unusual behavior, but is appar- ently normal. 2:00 p.m. Animal still lively. 2:15 p.m. Ate.some meat and drank water. No nausea observed. 5:10 to 5:20 p.m. Animal well. February 16. 9:00 a.m. Fed meat, cracker and bone meal. Urine, 170 ce.; specific gravity, 1.055; acid; no albumin. On adding NaNO, and H.SO, 30 Steudel: Zeitschr. f. physiol. Chem., xxxii, pp. 285-290, 1901. Israel S. Kleiner 461 a rose-colored precipitate appeared which was filtered off and washed with hot water, alcohol andether. For the total volume of urine this would have amounted to 1.17 grams. It was dissolved in KOH, reprecipitated by HCl, filtered etc. and analyzed by the Kjeldahl-Gunning method (for nitrates). Calculated for isonitroso derivative of 2,4-diamino-6-oxy- monamino-dioxy- pyrimidine pyrimidine ° (=CaHsNs02): (=C.HiN,O3): Found: Sloosstt. eee 45.16 35.89 39.25 39.56 Feeding suspensions of the salt to a guinea pig and to a rabbit gave simi- lar non-toxic results (see Table IV). The sulphate, as in Steudel’s investi- gation, was used in every case. TABLE IV. Animal Experiments: 2,4-Diamino-6-orypyrimidine. | ANIMAL een pec (15) Dog..... Per os No toxic effect. Large proportion excret- (18) Guinea ed; deaminized (?) pig....| 0.16 | 0.14 | 0.87 | Per os Fed in saccharose suspension from pipette. Nosymp- toms. Under ob- | servation seven days. (19) Rabbit. .| 1.96 | 0.51 | 0.26 | Per os Given in suspen- | | sion in water. No albuminuria. Na- NO, — FeSO, test positive. Unable to obtain an isoni*ro- so compound as in Experiment 15. | — — It is therefore evident that his pyrimidine is not toxic when given per os as the sulphate. Doses larger than those reported toxic by Steudel were without effect upon the rabbit and guinea pig as Experiments 18 and 19 indicate. 462 Action of Certain Pyrimidines 2,4,6-Triamino-6-oxypyrimidine. This pyrimidine was prepared according to Traube’s* directions and was isolated as the sulphate. When crystallized quickly the salt appears as small rods or rectangular prisms but if allowed to crystallize slowly large needles are formed. After desiccation, an analysis gave the following results. Calculated for (CsH7NsO)H280.+H:0: Found: Nee eta: « Sara Seno oeat 27.24 27.87 The solubility of this salt is about the same as that of the diamino compound, 2.e., it was found to be possible to obtain a 0.49 per cent solution at 43°. In this case, however, the fluid became dark dur- ing the manipulation and, after drying, the residue was dark brown in color. It is thus evident that some chemical change—a decom- position or oxidation—occurred and hence the determination can only be regarded as an evidence of the very slight solubility of the substance at low temperatures and of its instability, when in solution, at a high temperature. According to Traube, if an ammoniacal solution of the sulphate be shaken so as to afford contact with the air the fluid assumes an intense violet color resembling permanganate solution. This re- action, according to our experience, is better performed and with more uniform success, if a few milligrams of the substance are placed on a porcelain surface together with one or two drops of NH,OH and evaporated to dryness on a water-bath; the violet tinge is here seen against the white surface. In trying to dissolve some of the salt in 50 per cent alcohol it was discovered that although very little went into solution the latter became colored with this same violet tint. This pyrimidine also resembles uric acid in two reactions, namely, the murexide and Schiff’s tests; the murexide test is given very brilliantly indeed. Addition of bro- mine water to an aqueous solution was found to produce a deep reddish-purple color which vanished, leaving a yellow solution, when the bromine was in excess. The NaNO.-FeSO, reaction is positive if the triamino pyrimidine be first dissolved in boiling water; this is probably due to a trace of the diamino being formed by the action of the water as, theoretically, if the 5 position is occupied by an amino group no isonitroso derivative can be formed. 31 Traube: Ber. d. deutsch. chem. Gesellsch., xxxiii, pp. 1371-1383, 1900. Israel S. Kleiner 463 In using this salt in physiological experiments we again obtained results quite different from those reported by Steudel. EXPERIMENT 14. February 11. 2:45 p.m. Bitch weighing 9.8 kg. (same animal as in Experiment 15) was given 1.58 grams of the sulphate of this pyrimidine mixed with about 180 grams of chopped meat and some bone meal. No unusual symptoms were noticed by 4:00 p.m. 4:30; 5:00; 5:30; 7:40; 9:15 p.m.; animal observed and was well and play- ful. February 12. 8:50 a.m. Animal well. Urine, 134 cc.; dark orange-red in color; specific gravity, 1.049; acid; no albumin. Some of the urine was made acid with H.SO, and was concentrated to small volume. HgSO, solution was added and the precipitate filtered off; a few crystals were found and, as they gave a violet color on treatment with NH,OH and evaporation, were probably some triamino-sulphate which had crystallized before adding the HgSO,. The mercury precipitate was unfortunately lost through an accident. 4:00 p.m. Fed meat and bone. February 13. a.m. Urine light yellow in color. Relatively larger doses were fed in suspension to a rabbit and a guinea pig with similarly negative results; these are summed up in the following table (Table V). TABLE V. - Animal Experiments: 2, 4, 5-triamino-6-oxypyrimidine. AMOUNT GIVEN | a \WEIGHT ieee oatrecee fe on | BEMARES AND RESULTS Total | kilo- i | | gram | kg. gram | gram | | (14) Dog.....| 9.8 | 1.58 ; 0.16 | Per os No toxic effect. Some excreted (?). Urine | red. (13) Rabbit..| 2.5 | 0.2 | 0.08] Per os Notoxiceffect. Sec- | Ovo"2|.0520 ond dose four days after first. Urine red after second dose. (10) Guinea Fed, suspended in pig saccharose solu- (young).| 0.13 | 0.13 | 1.0 | Per os tion, from a pip- ette. No toxic | effects. Urine col- | ored dark red. 464 Action of Certain Pyrimidines It is accordingly evident that even the triamino compound, which Steudel claims is the more toxic of the two, has no harmful influence upon the organism when administered by way of the mouth. Cyanacetylguanidine. HN—CO aN=6 Ott = CN The preparation and properties of cyanacetylguanidine are described above in the description of the process of making 2,4- diamino-6-oxypyrimidine. This compound was used because it is a precursor of the diamino and triamino pyrimidines just described and might readily be pres- ent as an impurity if these compounds were carelessly prepared. In- asmuch as from the following experiments it is seen to be toxic after injection, a reason for the difference between our results and Steudel’s is thus suggested. EXPERIMENT 28. March, 30. 12:30 m. Injected subcutaneously, into guinea pig weighing 680 grams, 0.38 gram cyanacetylguanidine in-15 ce. water. . 4:00 p.m. Animal shows hyperexcitability. 6:20 p.m. Still very excitable. March 31. 1:15 p.m. Apparently well except for continued hyperexcit- able state, which is not as great as on the previous day. 4:00 p.m. Has eaten 15 grams oats and 95 grams carrots during twenty- four hours. Weight 668 grams. Urine, 43 cc.; alkaline; specific gravity, 1.031; no albumin present; strong NaNO:-FeSO, test; upon addition of NaNO, in substance, and H.SO, a pink isonitroso derivative was obtained which amounted to 0.161 gram, if computed to total volume. This was analyzed with the following results. Calculated for C4HsNsO2(=isonitroso derivative of eyanacetylguanidine): Found: IN PROUD eo & ee 3o) ERI asap SE Cage a 45.16 36.47 5:10 p.m. Apparently well. April 1. 4:00 p.m. Has eaten 15 grams oats and 105 grams carrots. Weight, 664 grams. Urine, 46 cc.; alkaline; specific gravity, 1.030; no albu- min; NaNO:-FeSO, test positive. Israel S. Kleiner 465 April 2. 4:00 p.m. Weight, 667 grams, Urine, 38 cc.; NaNO.-FeSO, test negative. * EXPERIMENT 31. A pril 2. 11:45 a.m. Young guinea pig, weighing 191 grams, given 0.4 gram cyanacetylguanidine in 17 cc. water by subcutan- eous injection. 12:15 m. Apparently well. 2:10 p.m. Animal found in violent spasms, especially the posterior parts .of the body. There is hyperexcitability. 2:20 p.m: Head raised alittle more and pig runs around some, pawing at its chin at intervals. Twitchings continue. 4:13 p.m. Violent convulsion; lies on its side and moves its limbs rapidly. 4:18 p.m. Animal gradually rights itself and grips the side of the wire cage with its teeth. Waves of convulsions, starting at the posterior part and running forward, occur. 4:22 p.m. Dies in the same position; body quickly in rigor. The urine excreted, 2 cc., was found to contain no albumin but on addition of NaNO; and H2SO, a pink precipitate appeared which after dissolving in Na2COs and adding FeSO, produced the deep prussian blue color. EXPERIMENT 30. March 31. 2:55 p.m. A dog weighing 5.8 kg. was fed 100 grams chopped meat containing 0.94 gram cyanacetylguanidine. 3:10 to 3:20; 4:25 to 4:30 p.m. Apparently no effects. 4:50 p.m. Drank water; no nausea. April1. 9:15a.m. Dogapparently well. 3:00p.m. Fed meat, lard, bone and cracker meal. Urine, 226 cc.; acid; specific gravity, 1.025; no albumin; strong NaNO,-FeSO, test. To an aliquot portion was added solid NaNO: and H.SO, and the reddish brown precipitate which amounted to 0.389 gram analyzed. Calculated for C4sHsNs02 (=isonitroso derivative of cyanacetylguanidine): Found: POPE ERE eros borer siors ie diace ts vist alaleg » erate 45.16 34.43 April 2. 3:00 p.m. Dog well. Urine, 138 ce.; specific gravity, 1.032; acid; no albumin; strong NaNO2-FeSO, test. No loss of appetite or other unfavorable symptoms. April 3.. 9:30 a.m. Dog well. Weight, 5.6 kg. Urine gives uncertain NaNO2-FeSO, test. These and one other experiment are summarized in Table VI. The low nitrogen values found in Experiments 28 and 30 suggest the possibility of a deaminization of cyanacetylguanidine in the body. The substitution of O for NH in its isonitroso derivative would result in a compound containing 35.90 per cent of nitrogen; the figures found, 36.47 per cent and 34.43 per cent, correspond with this percentage. 466 Action of Certain Pyrimidines TABLE VI. Animal Experiments: Cyanacetylguanidine. | AMOUNT GIVEN aha Sate \WEIGHT | Per byte tenia oe REMARES AND RESULTS Total kilo- gram kg. gram | gram (28) Guinea Hyperexcitability. pig...., 0.68 | 0.38 | 0.56 | Subcutaneously Excreted consider- able as a deamin- | ized (?) substance. (31) Guinea Hyperexcitability. pig....| 0.19 | 0.40 | 2.1 Subcutaneously Violent convul- sions. Fatal in four and three | quarters hours. (29) Dog..... 8.4 | 0.70 | 0.08 |-Per os No symptoms. Urine | gave positive Na- | NO,-FeSO, test. | | No albuminuria. (30) Dog..... 5.8 | 0.94 | 0.16 | Per os No harmful effect. Considerable ex- creted as adeamin- ized(?) substance. This toxic action agrees with the results of some unpublished trials by Mr. J. J. Costello, who observed similar effects in Pro- fessor Mendel’s laboratory when the sulphate of this compound was subcutaneouly injected. A few of his figures follow. of guinea pig Results Oe ee Bere. a oa tn Gee ae ee Hyperexcitability—recovery. a el Sa ee Hyperexcitability for two days-recovery. AE ates OSS Ss rs Ree ae Death in sixteen hours. gD ea ie a ee WO By IR Death in fourteen hours. OOM tee 4 Sa 5 ee Death in three and one-half hours. DISCUSSION. In considering the physiological and pharmacological behavior of the members of this series the most striking fact is the toxicity of 5-aminomalonyl guanidine with its chief effect upon the epithe- lium of the convoluted tubules. Its harmlessness when adminis- Israel S. Kleiner 467 tered per os may be due either to an absorption so slow as to allow of elimination before a toxic concentration is reached, or to a trans- formation—perhaps by deaminization—into a non-toxic compound in the intestinal wall. The toxicity after subcutaneous adminis- tration may possibly be attributable to some hydrolytic or oxida- tion product formed during solution inasmuch as the solution quickly assumes a red color. The absence of hypnotic powers in barbituric acid and malonyl- guanidine is in harmony with the ineffectiveness of the lower alkyl barbituric acid derivatives and of 5,5-dipropylmalonylguanidine.” The diarrheal action of barbituric acid is noteworthy because of a similar action ascribed to alloxan.® Steudel’s* claim that 2,4-diamino-6-oxypyrimidine and 2,4,5- triamino-6-oxypyrimidine are toxic, cannot be substantiated. In duplicating his experiments in which he fed these compounds to a dog, no similar results could be obtained; the animal used was a very playful one as was Steudel’s but it did not become less lively after ingesting these substances, nor was vomiting or albuminuria observed or any other of the effects noted by that author. The lethal doses for rats he gives as 0.2 gram and 0.1 gram for the sul- phates of the diamino and triamino compounds, respectively, when injected subcutaneoulsy. The smallest volumes which can possi- bly contain these amounts at 43° are 40 cc. and 20 cc. respectively. Moreover, it was shown above that such concentrations are not suitable for injection and this leads us to believe that Steudel used products which were more soluble than these aminopyrimidines. Moreover he published no analyses of his compounds. Cyanace- tylguanidine, however, is a precursor of both pyrimidines; it is quite soluble as is also its sulphate; and finally, when injected sub- cutaneously it is toxic. These properties would indicate that this compound was an admixture of Steudel’s preparations and would account for their toxic action. However, when fed to dogs, cyana- cetylguanidine is not toxic although his preparations were; and the only apparent explanation for this is that still another com- taminating substance was responsible in this case. That cyana- cetylguanidine is toxic is not surprising since, from its structure, ™ Fischer and von Mering: Therapie der Gegenwart, v, pp. 97-101, 1903. % Koehne: Inaugural Dissertation, Rostock, 1894, 40 pp. 4 Steudel: Zeitschr. f. physiol. Chem., xxxii, pp. 285-290, 1901. 468 Action of Certain Pyrimidines HN—CO HNC CH: H2N CN it might possess the properties of guanidine or of nitriles. Guani- dine, the toxicity of which has long been known, causes* peculiar shaking movements of the head and ears, paralysis of the hind limbs, clonic muscular contractions and muscular twitchings of the entire body. Different nitriles have different effects but the typi- cal phenomena are described* as vomiting, dyspnoea, tetanic con- vulsions and opisthotonus. Hence, probably cyanacetylguanidine embraces some of the toxic effects of both of these poisons (see Experiments 28 and 31). The behavior of 2,4-diamino-6-oxypyrimidine and cyanacetyl- guanidine in the body affords suggestions for further work upon the intermediary metabolism of these substances, as the few experi- ments indicate that a deaminization may occur in vivo. The dif- ferences between the theoretical percentage of N for the compounds administered and ‘those recovered from the urine are too great (6 to 11 per cent) to be ascribed to the method of analysis or to faulty technique. Moreover, an analysis of the pure isonitroso derivative of the diamino pyrimidine by the same method gave a satisfactory nitrogen value. The possibilities for the transforma- tion of this pyrimidine are shown by the following scheme. HN—CO HN—CO HN—CO Gl re HaNC CH +H:20=NH3+ OC CHe or HNC CH2 ll lee eos N—CNH2 HN—CNH HN—-Ce With cyanacetylguanidine a somewhat similar problem is pre- sented as deaminization can result in one of three compounds: HN—CO HN—CO HN—CO HN—CO | HNC CH2+H.,0 =NH;+ OC CH or OC CH: or HN—C CH, | Laas lang HaN CN ae H.2N CN HN—CNH HN—CO 36 Gergens and Baumann: Arch. f. d. ges. Physiol., xii, pp. 205-214, 1876; Pommerenig: Beitr. z. chem. Physiol. u. Path., i, pp. 561-566, 1901. 36 Kobert: Lehrbuch der Intoxikationen, Stuttgart, ii, p. 862, 1906. Israel S. Kleiner 469 If either of the last two complexes result it is of great interest as no precisely similar transformation of an acyclic into a cyclic compound is known in physiology. HN— fr Lusini’s conclusion that the grouping OC has first a stimulat- | HN ing and then an inhibiting action on the nerve centers and that HN—CO the grouping Ate | has no such power can not be substan- ; | tiated inasmuch as barbituric acid, which is non-toxic, contains this urea grouping and differs very little in structure from alloxan which Lusini found to be toxic. SUMMARY. The administration of barbituric acid per os is followed by no marked physiological effects except diarrhea; when given subcu- taneously the free pyrimidine has a local action on the tissues due to its acid properties. The sodium salt has no local action. Malony! guanidine when fed, or when injected subcutaneously as the sodium salt, provokes no noteworthy symptoms. 5-Amino- malonylguanidine hydrochloride, 2,4-diamino-6-oxypyrimidine sul- phate and 2,4,5-triamino-6-oxypyrimidine sulphate, when fed, are also without marked action. Subcutaneous injection of 5-aminomalonylguanidine hydro- chloride leads to grave changes in the tubular epithelium of the kidney; casts and albumin abound in the urine; and death fre- quently results. 2,4-Diamino-6-oxypyrimidine sulphate and 2,4,5-triamino-6-oxy- pyrimidine sulphate, which Steudel reported as toxic, are too insol- uble to inject in appreciable quantity. Inasmuch as cyanacetyl- guanidine, a precursor of both of these, is quite soluble, and was found to be toxic when injected subcutaneously, doubt is expressed as to the purity of the diamino and triamino pyrimidines used by Steudel, especially as this author also observed nausea, etc., after feeding them to dogs, whereas no symptoms whatever occurred in the present investigation under similar conditions. 470 Action of Certain Pyrimidines A color reaction is described which is common to all of this series, although 2,4,5-triamino-6-oxypyrimidine and 5-aminomalonylgua- nidine do not react well. By aid of this reaction and in other ways, evidence was gained that, after administration of a compound of this series there was excreted in the urine the compound used (or a derivative) in every case except with 5-aminomalonylguanidine, and perhaps 2,4,5-triamino-6-oxypyrimidine. Evidence is presented to indicate that 2,4-diamino-6-oxypy- rimidine and cyanacetylguanidine may be deaminized in the body. My thanks are due Prof. Lafayette B. Mendel who directed the physiological investigations and Prof. Treat B. Johnson, who aided and advised in the syntheses of the compounds employed as well as in the questions of organic chemistry: involved. PHYTIN AND PHOSPHORIC ACID ESTERS OF INOSITE. By R. J. ANDERSON: (From the Chemical Laboratory of the New York Agriculiural Experiment Station, Geneva, N. Y.) (Received for publication, April 2, 1912.) In continuation of the physiological investigation concerning the metabolism of the organic-phosphorus compound known as phytin, which has been and is being carried out at this institution by Dr. Jordan, a closer study of the chemical properties of this substance, phytin, became necessary. Much work has already been done and reported on this subject by various investigators. Definite information, however, concerning different kinds of salts formed by the free phytic acid or inosite phosphoric acid is seldom met with in the literature. Frequently impure salts have been analyzed. Posternak, who first successfully prepared phytin in pure form,! also studied its chemical properties. Among the salts mentioned? is one, calcium-magnesium, as well as one crystalline, calcium- sodium, double salt, for which he gives the formula, 2C,;HsP.2O Na. +C,H,P.Ca. + 8H:O. Winterstein® describes a calcium-magne- sium compound which, after removing the calcium with oxalic acid and precipitating with alcohol, contained 42.24 per cent P.O; and 12.97 per cent MgO. Patten and Hart,‘ working in this laboratory, isolated from wheat bran an impure magnesium- calcium-potassium compound. Levene® describes a semi-crystal- line barium salt which corresponds to a tetra-barium phytate. Vorbrodt® mentions a crystalline barium salt obtained by partially ¥ 1 Rev. gén. de bot,, xii, p. 5; Compt. rend. acad. des sci., cxxxvii, p. 202. 2 Compt. rend. acad. des sci., ¢Xxxvii, pp. 337 and 439. 3 Ber. d. d. chem. Gesellsch., xxx, p. 2299. 4 Amer. Chem. Journ., xxxi, p. 566. 5 Biochem. Zeitschr., xvi, p. 399. 6 Anzeiger Akad. Wiss. Krakau, 1910, Series A, p. 414. 471 472 Phytin and Esters of Inosite neutralizing phytic acid with barium hydroxide and evaporating in vacuum, to which he assigns the formula, CyH».0.Ba7P1). Although crystalline, this compound was undoubtedly impure. By neutralizing the mother-liquor from the above with barium hydroxide he obtained an amorphous precipitate of the composi- tion C, 5.75; H, 0.77; Ba, 52.97; P, 14.60 per cent. This corre- sponds approximately with a hexa-barium phytate. Of the several salts mentioned in this paper some were obtained from commercial phytin and from an organic-phosphorus-mag- nesium compound by precipitating with barium chloride and barium hydroxide; others were prepared from previously purified phytic acid. These products will be more fully described in the experimental part. The tri-barium phytate, CsH:20.[(PO3H)2 Bals, is obtained pure as an amorphous white powder by repeatedly precipitating barium phytate in 0.5 per cent hydrochloric acid with a like volume of alcohol. It may also be obtained in crystalline form by dissolving the amorphous salt in a 10 per cent solution of phytic acid in which it is very soluble and from which it again slowly crystallizes out on standing at ordinary temperature. A penta-barium phytate, CsHi,O27P.Ba;, is obtained when a solution of the tri-barium phytate in 0.5 per cent hydrochloric acid is neutralized with barium hydroxide and then made faintly acid with acetic acid. The penta-barium ammonium phytate, CsH 2027PsBas(NHs)s, is obtained when the above menticned amorphous tri-barium salt is digested with dilute ammonia. The penta-magnesium ammonium phytate,CsHi2027PsMg;(N H,)o, is thrown down as a white amorphous precipitate when excess of magnesia mixture is added to an aqueous solution of phytic acid, or when ammonium phytate is precipitated with magnesia mixture. A tetra-cupric di-calcium phytate, CsH)20.7P,CuyCaz, in nearly pure form is obtained when a slightly acid solution of calcium ammonium phytate is precipitated with excess of copper acetate. If the magnesium ammonium phytate is precipitated under the same conditions an impure compound is obtained which contains about 1 per cent Mg, 0.6 per cent N, 34 per cent Cu and 15.6 per cent P. No effort was made to obtain these salts pure. It was R. J. Anderson 473 only desired to find out to what extent other bases were removed when precipitating with copper acetate. Starkenstein’ claims that commercial phytin always contains free inosite together with inorganic phosphates and that merely drying the substance at 100°C. causes nearly complete decomposi- tion into inorganic phosphate and free inosite. That phytin is so easily decomposed seemed very improbable as several months’ work on the substance has shown that it is relatively stable when pure and when no mineral acids are present. Moreover Contardi® reports that when phytin is heated in an auto- clave with pure water for several hours to a temperature of 200°C. ‘only very small quantities of inosite could be isolated. In order to determine if inosite is present in determinable quantity 100 grams of commercial phytin in the form of the acid calcium salt, imported from Europe-and which had been kept in the laboratory for several years, was shaken up with 1 liter of water, filtered at once and washed with water. The filtrate was precipitated with barium hydroxide, again filtered and the excess of barium precipitated with carbon dioxide and the filtrate from the latter evaporated on the water-bath. In the very slight residue which remained, consisting mostly of barium carbonate with a trace of barium chloride, no trace of inosite could be detected by the most painstaking method of isolation. Of the same phytin, 100 grams were dried to constant weight at 115°C. and was then treated in the same manner. Even here no trace of inosite could be obtained. Subjecting to the same treatment 50 grams of the same phytin, after previously mixing with 0.5 gram inosite, resulted in the recovery of 0.4 gram inosite. This proves that phytin is by no means so easily split as Starken- stein claims. The results in his case may have been due to other causes besides mere drying at 100°C. The same author (loc. cit.) also states that when phytic acid is precipitated with ammoniacal magnesia mixture it is not the mag- nesium ammonium compound which is formed but only the diffi- cultly soluble magnesium phytate. Thisisanerror. Under these conditions the previously mentioned penta-magnesium ammonium phytate, CgH12027P6Meg;(N Hy), is formed. For the free phytic acid Posternak* proposed the empirical for- mula, C2HgO9P2, which he considered to have the following con- stitution: 1 Biochem. Zettschr.. xxx, p. 59. 8 Atti R. Accad. dei Lincei Roma (5), xviii, 1, p. 64. ® Compt. rend. acad. des sci., ¢xxxvii, p. 4389. 474 Phytin and Esters of Inosite H | Pers PO (OH): cHo-Po (OH), H and which finds expression in the name “anhydro-oxymethylen di-phosphoric acid. As is well known the free acid, as well as its salts, is easily split under the influence of dilute mineral acids into inosite and ortho- phosphoric acid. This fact and the discovery by Neuberg?® that both inosite and phytin yield furfurol when distilled with phos- phorus pentoxide and phosphoric acid, respectively, lead him to believe that the inosite ring exists already formed in phytin. In accordance with this view he proposed the following structural formula for the acid: ae (OH)3P P(OH)s | | O O CH—CH (OH); | | (OH)3 ne O—CH CH—O—P oN | | »o (OH); P—O—CH—CH—O—P (OH)3 This is just treble the molecular weight of the anhydro-oxymeth- ylen di-phosphoric acid of Posternak. Suzuki and Yoshimura" considered that phytic acid was the hexa-phosphoric acid ester of inosite. Starkenstein” believes that phytin represents a complex pyro- phosphoric acid compound with inosite and he proposes the follow- ing constitutional formula: 19 Biochem. Zeitschr., ix, pp. 551 and 557. 1 Bull. Coll. of Agric. Tokyo, vii, p. 495. 12 Biochem. Zeitschr., xxx, p. 56. R. J. Anderson A75 (OH)2 (OH) P =0-HO-HC—CH-OH-0=P \ pO P=0-HO-HC CH-OH-O=P (OH)2 (OH)e (OH). P = O-HO-HC—CH-OH-O= P (OH): a i a 4 NY Vorbrodt (loc. cit.) proposes still another formula. It is impossible at the present time to decide definitely between any of the above constitutional formulas, as the substance has not yet been synthesized in the laboratory. ; As represented by the empirical formula, CsH2O27P.s, phytic acid corresponds to a hexa-phosphoric acid ester of inosite plus 3H.0, CsH,O; [PO(OH).|, + 3H20. At present it is impossible to say whether the compound repre- sents a pyrophosphate or if the water is linked in some other way. That the acid contains twelve acid (OH) groups as expressed in the formula of Starkenstein, which wouldalso be the case if it were a, hexa-phosphoric acid ester of inosite, and not eighteen (OH) groups as in the formula of Neuberg, seems certain, for in no case have we been able to prepare any salt in which more than twelve H- valences were replaced by bases. As observed by Starkenstein only one-half of the twelve (OH) groups are particularly reactive. This finds expression in the fact that the barium-salt obtained from acid solutions contains only 3Ba to 6P. As suggested by the above author, it is probable that these reactive hydroxyls are adjacent but linked to different phosphoric-acid residues. The salts with binary bases would then be represented by the following: THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 5 476 Phytin and Esters of Inosite — OH —-P=0 m -P=0 — OH A further confirmation of this is found in the fact that the tri- barium phytate as well as other similar salts of phytic acid with binary bases are strongly acid in reaction. The presence of only eight acid (OH) groups, however, can be shown by titrating an aqueous solution of the acid with decinor- mal sodium hydroxide using phenolphthalein as indicator. Patten and Hart (loc. cit.) who titrated with decinormal barium hydroxide using phenolphthalein as indicator obtained results agreeing with a hexa-barium salt. Of special interest in connection with the constitution of phytin are the phosphoric acid esters of inosite. Neuberg and Kretschner!? report obtaining a poly-phosphoric acid ester of inosite by their method of preparing phosphoric acid esters of the carbohydrates and glycerine, that is, by the action of phosphorus oxychloride. The product however, could not be obtained pure as it was found impossible to separate it from the inorganic phosphates. Contardi' claims to have prepared the hexa-phosphoric acid ester of inosite by heating inosite with an excess of phosphoric acid in a stream of carbon dioxide to 160° to 165°C. The product was purified as the barium salt and after decomposing the latter with sulphuric acid the free ester was obtained, which he describes as identical with phytic acid. The same author claims to have prepared poly-phosphoric acid esters of mannite, quercite and glucose by the same method. Carré,’° however, repeating these experrments found that the products described by Contardi were merely mixtures of free phos- phoric acid and the polyhydric alcohols in question together with 1S Biochem. Zeitschr., xxxvi, p. 5. M4 Atti. R. Accad. det Lincei Roma, (5), xix, 1, p. 23. 1S Voids spe ceo: 1© Bull. soc. chim. de France. (4), ix, p. 195. R. J. Anderson 477 their decomposition products mixed with some monobarium phos- phate. Many fruitless efforts have been made in this laboratory to syn- thesize phytic acid and the hexa-phosphoric acid ester of inosite. All experiments in this direction lead only to the tetra-phosphoric acid ester of inosite, CsHs(OH).04 [PO(OH)el.. The method of Contardi was modified to the extent that inosite, either dry or with water of crystallization, was heated with phos- phorie acid, previously dried at 100°C. to constant weight, in vacuum to a temperature of 140° to 160°C. for about two hours. The same product, viz., the tetraphosphoric ester was obtained whether the phosphoric acid was present in large or small excesss above six molecules of H;PO, to one molecule of inosite. When it was present in less quantity than this, however, for instance one molecule of inosite to three molecules of H3PO,, then a mixture of esters was formed. It was found impossible to separate these prod- ucts completely owing to the fact that they possess about the same solubility. The tetraphosphoric ester is most conveniently isolated by means of its barium salt. The separation of the ester from the excess of the phosphoric acid or barium phosphate succeeded because its barium salt is much less soluble in dilute alcohol acidified. with hydrochloric acid than is barium phosphate. The new ester is a well characterized compound, very similar in appearance and reactions to phytic acid. By heating with acids, inosite and phosphoric acid are regenerated. It gives a white precipitate with the ordinary molybdate solution, and with excess of silver nitrate a white precipitate is also produced. These reac- tions are identical with those of phytic acid. The inosite used in these experiments was prepared from the crude magnesium compound previously mentioned and carefully purified by recrystallization. The reason why phytic acid could not be obtained by the action of phosphoric acid on inosite is no doubt to be found in that it is not a simple ester but a complex compound as suggested by Starkenstein. It is, however, difficult to understand why the hexa-phosphoric ester was not obtained by this method. The only explanation that can be offered is that under the conditions of these experiments it is not stable. 478 Phytin and Esters of Inosite One reason alleged by Starkenstein for considering phytin a pyrophosphate is based upon its giving a white precipitate with silver nitrate. This is certainly a characteristic reaction of pyro- phosphates. Yet the tetraphosphoric ester gives a pure white precipitate with the same reagent. As the ester cannot be in the form of a pyrophosphate the fact that phytic acid gives the same colored silver compound is not necessarily an indication that it represents a pyrophosphate compound. The phytic acid used in these experiments was prepared from products obtained from two different sources. The starting mate- rial in one case was a calcium phytate imported from Europe; the other was a crude natural magnesium organic phosphorus compound extracted in this country and kindly supplied us by Dr. Carl S. Miner of Chicago. As shown by the analyses,of the carefully purified salts and of the free acid, these two preparations were identical and they were also identical with the product described as phytic acid by Poster- nak and other investigators. EXPERIMENTAL PART. Tri-barium phytate. The commercial phytin was purified for analysis by means of the barium salt. Thirty grams calcium phytate were dissolved in a small quantity of 0.5 per cent hydrochloric acid, diluted to about 2 liters with water and a concentrated solution of 30 grams barium chloride was added. The precipitate was dissolved without fil- tering by the addition of just sufficient dilute hydrochloric acid. It was then precipitated by adding barium hydroxide to faintly alkaline reaction. The mixture was then acidified with acetic acid and after standing over night was filtered and well washed in water. It was re-precipitated in the same manner three times. After finally filtering and washing in water the substance was dis- sulved in about 1 liter of 0.5 per cent hydrochloric acid, filtered and the filtrate precipitated by adding a like volume of alcohol. After repeating this operation the substance was filtered, washed free of chlorides with 50 per cent alcohol and finally washed in alcohol and ether and dried in vacuum over sulphuric acid. R. J. Anderson 479 The product so obtained was a light, perfectly white semi-crys- talline or amorphous powder. Placed on moist litmus paper, it showed a strong acid reaction. It is very slightly soluble in water, slightly soluble in acetic acid and readily soluble in mineral acids. For analysis the substance was dried at 130°C. 0.2728 gram substance gave 0.0352 gram H;O and 0.0643 gram COs. 0.2763 gram substance gave 0.1749 gram BaSO, and 0.1675 gram Mg2P20;. 0.1909 gram substance gave 0.1206 gram BaSO, and 0.1154 gram Mg2P2O7. For CsH120.{ (PO;H).Ba]s; =) 1120: Calculated: C = 6.42; H = 1.60; P 6.60; Ba = 36.78 per cent. Found: C642; H =\1.44:. P P 1 16.89; Ba = 37.25 per cent. 16.85; Ba = 37.17 per cent. The barium salt prepared in the same manner from a natural crude magnesium organic phosphorus compound gave the following result on analysis: 0.2057 gram substance gave 0.0273 gram H.2O and 0.0480 gram CO». 0.1422 gram substance gave 0.0886 gram BaSQO, and 0.0841 gram. Mg»P>0O7. Found: C = 6.36; H = 1.48; P = 16.48; Ba = 36.66 per cent. The two salts are therefore identical. Crystallized tri-barium phytate. One gram purified phytic acid was dissolved in 10 cc. water and’ 4 grams of the above mentioned tri-barium phytate added. It was filtered from traces of undissolved particles and allowed to stand for two days at room temperature. The substance has then sepa- rated as a heavy crystalline powder of irregular form. From less concentrated solutions the substance separates in small, needle- shaped crystals. The substance was filtered, washed well in water and finally in alcohol and ether and dried in the air. For analysis it was dried at 120°C. 4 0.1972 gram substance lost 0.0153 gram H,O. 0.2028 gram substance gave 0.1251 gram BaSO, and 0.1216 gram Mg»P.0;. Found: P = 16.71; Ba = 36.30 per cent. Calculated for 5 H.O: 7.44; Found: 7.75 per cent. 480 Phytin and Esters of Inosite Penta-barium phytate. This salt is obtained on neutralizing a solution of the tri-barium phytate in 0.5 per cent hydrochloric acid with barium hydroxide and then acidifying with acetic acid. The precipitate was filtered, washed thoroughly in water, alcohol and ether and dried in vacuum over sulphuric acid. The product was a white amorphous powder. For analysis the substance was dried at 130°C. 0.2970 gram substance gave 0.0307 gram H;0 and 0.0500 gram CO». 0.2507 gram substance gave 0.2080 gram BaSO, and 0.1207 gram MgP207. 0.1856 gram substance gave 0.1543 gram BaSO, and 0.0899 gram Mg»P.0;. For C5H14007P Bas = 1391. Calculated: C = 5.17; H = 1.00; P = 13.37; Ba = 49.37 per cent. Found: C = 4.59; H = 1.15; P = 13.42; Ba = 48.82 per cent. P = 13.50; Ba = 48.92 per cent. Penta-barium ammonium phytate. When the tri-barium phytate is digested in dilute ammonia it is transformed into the penta-barium ammonium salt and ammo- nium phytate. The latter product, however, was found to contain some barium. Two grams of the analyzed tri-barium phytate were digested for two hours in 25 ce. of 2.5 per cent ammonia, filtered and washed in dilute ammonia and finally in alcohol and dried in vacuum over sulphuric acid. The product was a heavy white amorphous pow- der. On moist litmus paper it showed a neutral reaction. For analysis the substance was dried at 130°C. 0.1509 gram substance gave 0.1205 gram BaSO, and 0.0762 gram Mg»P-O;. 0.1747 gram substance gave 0.0026 gram N (Kjeldahl).” For C>5H12027P Bas (NHg4)2 = 1425. Caleulated: P = 13.05;Ba = 48.19; N = 1.96 percent. Found: P = 14.07;Ba = 46.99; N = 1.48 per cent. By evaporating the filtrate from the above to dryness on the water-bath an amber-colored mass remained which after drying at 130°C. gave the following result on analysis: Found: P = 20.51; Ba = 6.65; N = 10.48 per cent. 17 This and subsequent nitrogen determinations were made by Mr. M. P. Sweeney. R. J. Anderson 481 Penta-magnesium ammonium phytate. Two grams phytic acid were dissolved in 400 cc. water and then precipitated by adding excess of magnesia mixture slowly and under . constant shaking. After the precipitate had settled the super- natant liquid was decanted, the residue filtered and washed with water until free from chlorides and finally washed in alcohol and ether and dried in vacuum over sulphuric acid. The product was a fine white amorphous powder and weighed 2.7 grams. It reacts neutral on moist litmus paper. For analysis it was dried at 130°C. 0.1089 gram substance gave 0.0832 gram Mg,P;0; for P. 0.1089 gram substance gave 0.0705 gram Mg»P.O; for Mg. 0.1248 gram substance gave 0.0039 gram N | Kjeldahl 0.0893 gram substance gave 0.0028 gram N { (Kjeldahl). For CeH12027PsMegs (NHs3)2 = 859_5. Calculated: P = 21.64; Mg = 14.13; N = 3.25 per cent. Found: P = 21.29; Mg = 14.13; N = 3.12, — 3.13 per cent. If the phytic acid is first neutralized with ammonia and then precipitated with magnesia mixture the same product is obtained. Two grams phytic acid in 400 ce. water were neutralized with ammonia, precipitated with excess of magnesia mixture, filtered, washed free of chlorides with dilute ammonia and then in alcohol and dried in vacuum over sulphuric acid. For analysis the sub- stance was dried at 130°C. Found: P = 21.49; Mg = 13.96; N = 3.47; 3.48 per cent. Tetra-cupric di-calcium phytate. To a sulution of 2 grams phytic acid in 200 cc. water excess of calcium chloride was added and the solution then neutralized with ammonia. The precipitate was just dissolved in dilute hydro- chloric acid and the solution precipitated with copper acetate. The bluish-green colored copper compound was filtered off, washed with water until free from chlorides and then in alcohol and dried in vacuum over sulphuric acid. The dry substance was a light-blue amorphous powder. It is very slightly soluble in water or in very dilute acids, readily soluble 482 Phytin and Esters of Inosite in the ordinary dilute mineral acids. It is readily soluble in 2.5 per cent ammonia with a deep-blue color. In this solution con- centrated ammonia or alcohol produces a light-blue colored pre- cipitate. The compound represents a nearly pure tetra-cupric di-calcium phytate. It contained 0.17 per cent N. For CeHi205 (PO;Cu),4.(PO3;Ca)> = 1036. Calculated: Cu = 24.51; Ca = 7.72; P = 17.95 per cent. Found: Cu = 25.58; Ca = 7.69; P = 16.85-per cent. If a slightly acid solution of magnesium ammonium phytate is precipitated with copper acetate a light blue colored copper com- pound is obtained. After washing and drying it gave the follow- ing result on analysis: Mg = 1.11; Cu = 34.27; N = 0.64 and 0.52; P = 15.66 per cent. This compound is exceedingly soluble in dilute and concentrated ammonia. By the careful addition of alcohol to the ammoniacal solution a substance separates in light biue colored crystals on standing. This is evidently a complex copper-ammonium salt but it was not further examined. Phytie acid. This was prepared after the method of Patten and Hart (loc. cit.). The analyzed tri-barium salt was decomposed with the cal- culated quantity of decinormal sulphuric acid. After removing the barium sulphate, the solution was precipitated with copper acetate. The copper compound was decomposed with hydrogen sulphide, the copper sulphide filtered off, the filtrate concentrated in vacuum and finally dried in vacuum over sulphuric acid. The products obtained from both the calcium phytate and the magne- sium compound were light amber colored, very thick liquids and corresponded in all respects with the body described by other inves- tigators as phytic acid. For analysis the substance was dried at 130°C. a. From calcium phytate. 0.3193 gram substance gave 0.0917 gram H,O and 0.1288 gram COs. 0.1505 gram substance gave 0.1424 gram Mg,P2O;. R. J. Anderson 483 b. From the magnesium compound. 0.2789: gram substance gave 0.0804 gram H.O and 0.1101 gram COs. 0.1236 gram substance gave 0.1160 gram Mg,P20>. -For CeHoy O27P5 = 714: Calculated: Found: Found: 54 II MO RRee etre ON LLY 10.08 10.57 10.76 =) 3.36 8.21 3.22 IPS 2 A Ce 26.05 26.37 26.16 Titrated against decinormal sodium hydroxide using phenolphthalein as indicator the following results were obtained: 0.2648 gram acid required 30.7 ec. 4} NaOH. Calculated for 8NaOH: 29.65 ce. 0.1593 gram acid required 18.60 cc. 3) NaOH. Calculated for 8NaOH: 17.60 cc. Inosite from the crude magnesium compound. Twenty-five grams of the air-dried substance, containing 20 per cent of moisture, was heated with 100 ce. of 30 per cent sul- phuric acid in a sealed tube for about three hours at a temperature of 140°C. Two tubes equally charged were heated at the same time. After cooling the reaction mixture was of dark brown color and a considerable quantity of magnesium salts had crystallized out. The contents was washed into a beaker, filtered and diluted with water to about 1500 cc. The sulphuric and phosphoric acids and the magnesium were then precipitated by barium hydroxide, filtered and well washed in hot water. The filtrate was evaporated to about 350 cc. and the excess of barium removed by carbon dioxide, filtered, the filtrate decolorized with animal charcoal and then evap- orated on the water bath toa syrupy consistency. This wastaken up in a small quantity of hot water, filtered and alcohol added to the filtrate until a cloudiness was produced. By scratching with a glass rod crystallization began; more alcohol was then added and the mixture placed in the ice-chest over night. After filtering and washing in alcohol and ether and drying in the air the product weighed from 5.1 to 5.4 grams. From the mother liquor a further quantity of crystals from 0.4 to 0.6 gram could be obtained on the addition of ether and allowing to stand for twenty-four hours in the cold. 484 Phytin and Esters of Inosite For purification the raw product was dissolved in six parts of water and again brought to crystallization by the addition of alco- hol as before. It was then obtained in large, thin, colorless plates. It gave the reaction of Scherer. The dried substance melted at 220°C. (uncorrected). Dried at 100°C., 0.4136 gram substance lost 0.0669 gram H.O and 0.1600 gram lost 0.0258 gram H,0. The dried substance was analyzed. 0.1342 gram substance gave 0.0791 gram H.O and 0.1981 gram CQ. For CsHe (OH). = 180. Calculated: ‘C = 40.00; H = 6.66; 2H,O = 16.66.per cent. Found: C = 40.26; H =,6.59; 2H.O = 16.17 — 16.12 per cent. This substance was used in subsequent experiments with phos- phoric acid. Some 40 grams of inosite were prepared in this way. Tetra-phosphoric acid ester of inosite. Crystallized inosite (4.32 grams, 2 molecules) was powdered and mixed in a distillation flask with 24 grams phosphoric acid (about 24 molecules or double the quantity required to form the hexa- phosphoric ester). The acid had been previously dried at 100°C. to constant weight. The flask was connected with the vacuum pump and heated in an oil bath to 140° to 160°C. for about two hours. By 120° water began to come over and the reaction was practically complete at the end of one hour. After cooling the reaction mixture was a thick, reddish-brown colored, nearly solid mass. This was dissolved in about 1 liter of water and a solution of 40 grams of barium chloride in 400 cc. of water was added. The barium salt of the ester was then precipitated by the addition of about 1 liter of alcohol. A solution containing phosphoric acid and barium chloride in the same dilution as above remains perfectly soluble on the addi- tion of a like volume of alcohol. The voluminous flaky precipitate was filtered off at once and thoroughly washed in 334 per cent alcohol. For purification the substance was dissolved in 700 ce. of 0.5 per cent hydrochloric acid, filtered from slight insoluble residue, the filtrate diluted with 500 cc. of water, some barium chloride added and then precipitated by the addition of a like volume of alcohol. This was repeated a second time. The substance was R. J. Anderson 485 then dissolved in 500 cc. of 0.5 per cent hydrochloric acid, precipi- tated by adding barium hydroxide to slightly alkaline reaction, then acidifying with hydrochloric acid and adding 500 cc. alcohol. After filtering and washing as before the substance was again twice precipitated from 0.5 per cent hydrochloric acid solution with alcohol and finally washed in 50 per cent alcohol, alcohol and ether and dried in vacuum oversulphuricacid. Theproduct weighed 8.9 grams. It was a white voluminous amorphous powder. On moist litmus paper it showed a strong acid reaction. The solu- bility of the product was practically the same as forthe tri-barium phytate. For analysis it was dried at 100° and 130°C. 0.3252 gram substance lost 0.0281 gram H.0O. 0.2697 gram substance gave 0.0442 gram H.O and 0.0878 gram COs. 0.2038 gram substance gave 0.0300 gram H;O and 0.0685 gram COs. 0.2482 gram substance gave 0.1505 gram BaSO, and 0. 1434 gram Mg>P.0;. 0.1833 gram substance gave 0.1108 gram BaSO, and 0.1075 gram Mg,P20;. 0.1776 gram substance gave 0.1074 gram BaSO, and 0.1038 gram Mg>P,0;. For CsHs(OH)20« (PO3H)2 Ba], = 770.7. Calculated: C = 9.34; H = 1.55; P = 16.08; Ba = 35.64 per cent. Found: C = 8.87; H = 1.83; P = 16.10; Ba = 35.68 per cent. C = 9.16; H = 1.64; P = 16.34; Ba = 35.57 per cent. P = 16.29; Ba = 35.58 per cent. Calculated for 4 HO: 8.55; Found: 8.64 per cent. ll Another lot prepared by heating 1.80 grams dry inosite (1 mole- cule) with 7.9 grams dry phosphoric acid (about 8 molecules) and isolated in the same manner gave the following results on analysis: 0.2879 gram substance lost 0.0240 gram H20. The dried substance was analyzed. 0.2639 gram substance gave 0.0452 gram H.O and 0.0936 gram CO>. 0.1480 gram substance gave 0.0866 gram BaSO, and 0.0846 gram MgP2O;. 0.1632 gram substance gave 0.0959 gram BaSO, and 0.0933 gram Mg>P:0;. Found: C = 9.67; H = 1.91; P = 15.93; Ba = 34.48 per cent. H.0 = 8:33; P = 15.93; Ba = 34.58 per cent. A third lot prepared by heating 1.80 grams dry inosite (1 mole- cule) with 5.88 grams dry phosphoric acid (6 molecules) and isolat- ing in the same manner as before gave the following: C = 9.69; H = 1.75; P = 16.06; Ba = 36.33 per cent. It is apparent therefore that in each of the above experiments the same compound was produced. 486 Phytin and Esters of Inosite The free tetra-phosphoric ester. About 5 grams of the purified barium salt was decomposed by digesting it with the calculated quantity of decinormal sulphuric acid. After removing the barium sulphate the solution was pre- cipitated with excess of copper acetate. ‘The copper precipitate was filtered, thoroughly washed with water, suspended in water and decomposed with hydrogen sulphide. The copper sulphide was removed by filtration, the filtrate concentrated in vacuum and finally dried in vacuum over sulphuric acid until it was of a thick, syrupy consistency. For analysis the substance was dried at 130°C. 0.3020 gram substance gave 0.0933 gram H2O and 0.1577 gram CO». 0.1605 gram substance gave 0.01387 gram MgeP20;. For CsHe(OH): O,[PO(OH)2]4 = 500. Calculated: C = 14.40; H = 3.20; P = 24.80 per cent. Found: C = 14.24; H = 3.45; P = 24.09 per cent. 0.1663 gram substance required 16.5 cc. decinormal sodium hydroxide using phenolphthalein as indicator. This corresponds to five acid (OH) groups. Calculated for 5NaOH: 16.63 cc. ft ‘4 Properties of the free ester. The concentrated aqueous solution of the ester is very similar to phytic acid. It is a very thick amber-colored liquid of sharp acid, slightly astringent taste and strong acid reaction. On longer keeping in the desiccator over sulphuric acid it becomes hard and brittle and may be powdered. It is then very hygroscopic. The dry substance is slowly but completely soluble in alcohol, readily soluble in water. The concentrated aqueous solution gives a white precipitate with silver nitrate in excess which dissolves on largely diluting with water. The precipitate is readily solublé in ammonia, dilute nitric, sulphurie and acetic acids, insoluble in glacial acetic acid. With ferric chloride it gives a white or faintly yellowish precipi- tate which is very sparingly soluble in acids. With lead acetate a white precipitate is produced, readily soluble in dilute nitric acid but sparingly soluble in acetic acid. With barium chloride it gives a white precipitate slightly soluble in acetic acid but readily soluble in hydrochloric and nitric acids. R. J. Anderson 487 Calcium chloride does not give a precipitate but on heating the calcium salt is thrown down as a white precipitate which redissolves on cooling. Magnesium salts do not cause a precipitate and on heating the solution merely turns cloudy; on cooling it clears up again. With the ordinary molybdate solution it gives in the cold a white voluminous flaky precipitate which slowly turns yellowish in color. Phytie acid under the same conditions gives a white precipitate which remains unchanged in the cold. On drying at 110° or 130° the substance turns very dark in color. The ester, like phytic acid, fails to give directly the Scherer reaction for inosite. Inosite from the tetra-phosphoric ester. Ten grams of the purified barium salt was heated with 25 ce. 30 per cent sulphuric acid in a sealed tube to about 150°C. for three hours. After precipitating the sulphuric and phosphoric acids with barium hydroxide the inosite was isolated by the usual method and recrystallized from hot dilute alcohol. It was filtered and washed in alcohol and ether and dried in the air. Yield, 1.52 grams. It was obtained in the form of small colorless six-sided plates, free from water of crystallization. The air-dried, water-free substance melted at .221°C. (uncor- rected.) 0.2094 gram substance gave 0.1259 gram H.O and 0.3033 gram COs. 0.1360 gram substance gave 0.0827 gram H.O and 0.1991 gram COs. For CeHi205 = 180. Calculated: C = 40.00; H = 6.66 per cent. Found: C = 39.50; H = 6.72 per cent. C = 39.93; H = 6.80 per cent. As already mentioned, if a mixture of inosite and phosphoric acid is heated when less than six molecules H;PO, are present to one molecule inosite, a mixture of esters is obtained. It was found impossible to separate these bodies as barium salts and obtain pure compounds since their solubilities are apparently nearly alike. Dry inosite (3.60 grams, 2 molecules) and 5.88 grams dry phos- phoric acid (6 molecules) was heated in a distillation flask as before to 180° to 190° for about two hours, until water ceased coming over. The reaction mixture was in the form of a very bulky, thin 488 Phytin and Esters of Inosite flaky mass, very brittle and of yellowish-brown color, mixed with some very dark-colored substance. It was broken up with a glass rod and removed from the flask and treated with water in which the dark-colored portion was readily soluble, but the lighter-colored substance was insoluble in this medium. It was powdered in a mortar and thoroughly washed in water and alcohol and dried in vacuum over sulphuric acid. The substance was apparently insoluble in boiling water, in boiling dilute acids and in glacial acetic acid; also insoluble in alco- hol, ether and other organic solvents. After drying at 130° the substance was analyzed. 0.2500 gram substance gave 0.0838 gram H,O and 0.2085 gram CO). 0.1500 gram substance gave 0.1145 gram Mg»P.20O;. 0.1542 gram substance gave 0.1178 gram MgeP.0;. Found: C = 22.74; H = 3.75; P = 21.28; 21.29 per cent. This agrees approximately with a mono-pyro-phosphoric ester of inosite but the phosphorus is too high. It was decided to purify it by means of the barium salt. The substance was dissolved by boiling in dilute sodium hydroxide in which it gave a dark amber colored solution. After filtering, it was precipitated with barium chloride, the barium precipitate filtered and washed free of alkali. It was then dissolved in 500 ce. 0.5 per cent hydrochloric acid and precipitated by barium hydrox- ide. After filtering and washing it was repeatedly precipitated with alcohol from 0.5 per cent hydrochloric acid solution until finally a small amount of a white amorphous powder was obtained. After drying at 130° this was analyzed. 0.2028 gram substance gave 0.0412 gram H.O and 0.0979 gram CO:. 0.2207 gram substance gave 0.0413 gram H.O and 0.1042 gram CO:. 0.1982 gram substance gave 0.0996 gram BaSO, and 0.1103 gram Mg>P,0;. Found: C = 13.16; H = 2.27; P = 15.51; Ba = 29.57 per cent. C = 12.88; H = 2.09. In this compound the relation between the carbon and phos- phorus is nearly 6C to 3P which would indicate a tri-phosphoric ester. The substance was, however, far from pure and lack of material prevented any further investigation of this body, which is apparently a mixture of various esters. ON THE PRESENCE OF ACTIVE PRINCIPLES IN THE THYROID AND SUPRARENAL GLANDS BEFORE AND AFTER BIRTH. By FREDERIC FENGER. (From the Chemical Research Laboratory in Organotherapeutics of Armour and Company, Chicago.) (Received for publication, April 4, 1912.) From a chemical standpoint the two best known of the ductless glands are the thyroid and the suprarenals. The active principle of the suprarenals has been separated in pure crystalline form and we have well defined methods for itsidentification and quantitative estimation. The activity of the thyroid gland is measured by its iodine con- tent. Sajous, in his recent work,! calls attention to the fact that it is absolutely established that an iodine compound is the active agent of the thyroparathyroid secretions. Reid Hunt states? that the active principle of this gland is associated with iodine and that the therapeutic activity of the various preparations from this gland is proportional to the amount of iodine in thyroid com- bination present therein, and that consequently the iodine may be used as a basis for standardization of such preparation. Beebe’ confirms this statement. It has been stated that the thyroid gland of new-born animals‘ does not contain any iodine. To the writer it appeared unlikely that either the thyroid or suprarenal glands should be free from their active principles up to the time of birth. We know that the ductless glands inject their secretions into the circulatory and lymphatic system.’ If, there- 1 The Internal Secretions and the Principle of Medicine, i, p. 156, 1911. 2 Journ. Amer. Med. Assoc., Oct. 24, 1908, p. 1386. 3 Ibid., lvi, p. 658, March, 1911. 4 Tbid., lv, p. 1983, Dec. 3, 1910. 5 Ott: Internal Secretions, 1910, p. 93. 489 490 Thyroid and Suprarenal Glands fore, the secretions of these glands are necessary, not merely for the maintenance of life and healthy metabolism, but also to govern the growth of the young animal, we might reasonably expect to find these glands active not merely at time of birth but also inthe fetus, especially as these glands only produce internal secretions which, as far as we know, do not enter the alimentary tract. These considerations led the writer to conduct the experiments described below. The lack of available material, z.e., normal healthy glands, would prevent a thorough and practical investigation of this subject as far as the human body is concerned. Of the domestic animals, cattle are best adapted for such experiments. For this series of experiments, which were carried out during March, 1912, thyroids as well as suprarenals were used and in the case of cattle, four stages of age were selected, namely, the fetus about three months old, the fetus about eight months old, young suckling calves six to eight weeks old, and full-grown cattle. In general the suprarenal glands from beef, hog and sheep seem to be of fairly uniform and proportional size and color. The thy- roid glands on the other hand, varied enormously both in size and color. This is especially true of beef and sheep. An investiga- tion is now being carried on in order to look into this matter more thoroughly, and the results will be reported in a later paper. In this paper only normal-sized healthy glands are considered. It should be borne in mind that the period of gestation for cows is nine to nine and one-half months; for sheep five months; and for hogs four months. ‘ The method of preparation was briefly as follows: The fresh glands were trimmed and weighed, minced, dried at 35° to 50°C. to constant weight, and freed from fat by extraction with petroleum ether. All determinations were made in duplicate on composite samples of the number of glands specified in the tabulation. The thyroid and suprarenal glands were obtained from the same animals in case of all sheep and hog fetus and three months old beef fetus. The thyroid glands from eight months old beef fetus as well as those from sucking calves and all the grown animals were not out of the same animals as the suprarenal glands of corresponding age. Frederic Fenger 491 The iodine determinations were made according to Hunter’s excellent method.® The active principle of the suprarenal glands was determined colorimetrically according to the iodic acid method suggested by Hale and Seidell’ with the exception that samples of desiccated beef, hog and sheep suprarenals of known physiological strength were used for comparison instead of the proposed permanent standards. The results are given in the table on page 492. Those obtained on the sheep and hog glands are somewhat incomplete partly due to the fact that the glands in the fetus are very small, and also because suckling lambs and pigs are not commonly used for human food, and consequently obtainable only with great difficulty. The results obtained above indicate definitely that the thyroid gland of these animals contains iodine, not merely at time of birth but long before. Since the amount of iodine in the thyroid is an indication of the relative activity of this gland there is evidently a gradual rise in activity of the gland in the fetus, and this activity is increased rapidly shortly after birth, reaching its maximum in the young growing animal. The iodine content of the glands from the full-grown animals is very low. This is, however, not unusual as the iodine content varies considerably. The glands were collected during the same period as the glands’ from the various fetus and the analyses are given here for comparative purposes only. ; The active principle of the suprarenals is also present in the fetus long before maturity, and in comparatively higher quantities than in the full-grown animal. As time permits and opportunities present themselves, it is the writer’s intention to confirm these results, and to carry on further and more extended investigations along these lines, in the hope that the data so obtained may throw further light on the activity of the ductless glands. In conclusion it may be stated that in all his experience withthe thyroid glands from beef, hog and sheep, the writer has never found a sample of known origin that did not contain iodine. 6 This Journal, vii, p. 321, 1910. 7 Amer. Journ. of Pharm., Dec., 1911, p. 551. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 5 492 Thyroid and Suprarenal Glands SHEEP THYROIDS BEEF THYROIDS —_ - n *> SUCTION WASHBOTTLE Coe Fic. 2. ApparATus FOR USE WITH SUCTION. obtained is enough to titrate with a very fair degree of accuracy by the help of 34 acid and 5 alkali using alizarin red as indicator. The process is in every way similar to the method described on the preceding pages, except that the ammonia is collected in an ordinary small Florence flask (instead of in a measuring flask or Otto Folin and Chester J. Farmer 501 test tube) containing 10 cc. of 74 acid and about 40 cc. of water. The solution is titrated in the ordinary manner, and the end point is sufficiently sharp to give very satisfactory results. Results obtained in this way are recorded below. Those who are color blind as well as those whose ability to match colors is rather poor can use the above miniature Kjeldahl process to good advantage. Had the problem been purely a problem of total nitrogen deter- minations it is doubtful whether it would have been worth all the time that it has cost to develop the colorimetric procedure after it once had become clear that the color reaction seemingly could not be applied directly to the digestion mixture (see p. 493). As will be seen from the other analytical methods now pub- lished (see pp. 507-536) the total nitrogen determination was only one part of a general colorimetric scheme of analysis. The determinations recorded below are cited to show the accu- racy of our new method for the determination of nitrogen in urine. The middle column represents figures obtained by titrating the ammonia as described above. The figures represent grams of nitrogen per liter of urine. NEW METHOD KJELDAHL’S METHOD i 7 = | 1 7.9 8.1 8.0 2 10.0 10.2 9.9 3 3.7 4.1 at 4 10.5 10.0 10.2 5 3.8 4.1 3.9 6 9.4 9.3 9.2 7 7:5 7.3 7.3 8 9.2 9.3 9.2 9 9.0 9.1 9.0 10 9.3 9.1 9.2 11 8.5 8.3 12 9.1 9.3 13 9.1 9.4 14 Sez 5.3 15 3.7 33.0 16 7.5 (atl Diabetic urine. 17 7.5 7.6 18 8.4 8.4 Diabetic urine. 19 13.1 13.1 Nephritic¢ urine. 20 10.0 10.2 Nephritic urine. a ’ ‘Cae 5 ae" a ‘A . ‘ sh aR TROATE 10° 7 weAigtiey Gb . oR Ww oe te! inidditosatg Pe i ee oF ort sy iS hot ei BON 7 +5 i Tit » 9 ten rer a ee pi iy L, RMSE i Any 2 ed TM e it a oa tig i m } yBeu! fi PREY. NY a et? ody ot a . ‘ : a ‘ | Oleg hi Toe nieiee FS Ae Maan’ pater lleblet zh aingtuic: syodeiond : 4 _ t \ Ane : yok “by ote}! Oi rats wig 2 Wigvind tis fiesta. = im ~*~ ‘ troeay “ths oid Jeddah omg agit bed pets ¢ = pes tee Melee es: i as: VT > | Semetere we ‘ aT : oT AT PT SE re > Me : ne, tadto ssienis mae 39 ‘Ui . ve tan ‘ L4.@prt nny he eT, OE) S| MY a} oy ae {ete %! : aa a my Be by A RSs Ws tiahios dave: ast oD fast toate e ke Le PEGI Le | ts = eS un AEG banish ; Y tJ ti SEONG: th seor a iriSI ol atti anise Fis .¥ Fri rT} eS on be Sa . * ~ ' Fi tie? foryqys bad i i De Ad i F is = 4 + ‘| ¢ S es 3 Sat eet Ak 4" 9e4 a ’ te 44 ss a teacher a be = a * - Tae a . auivesd ea: tee“ 4 “ 7) 2 eae 2 ees ; ae mA Re ; rz ae : 1S eee ¥ * a a : on { Mi > ae > ei 4 s 63.7 a > me , ae a Ae — | tam <: c 7 ae — Le hot Oe re4 : cc. = ee a £.¢ > prey > ' bG Lips = x a i ia * | » s e- 3 oe 4 rae . =e rid nye at oe ee Sah mails Si kexd ‘ ~ Z Tee tt 2 ines hs we sy Wisi 2 ita 4 \ ; ae { is salina te $B LUDIAT VG 7% yay PRA ut i Aes Sid vey ‘a ge eet Sars ee Oe Pas 8 RN TE Se AN APPARATUS FOR THE ABSORPTION OF FUMES. By OTTO FOLIN anp W. DENIS. (From the Biochemical Laboratory of Harvard Medical School, Boston.) (Received for publication, April 12, 1912.) The new microchemical method for the determination of total nitrogen in urine described in the preceding paper requires very little in the way of laboratory equipment except a hood to carry off the sulphuric acid fumes. The decomposition of even a trace of urine with boiling sulphuric acid does produce more irritating fumes than would be tolerated in any small laboratory though the amount is insignificant compared with that produced in an ordinary Kjeldahl digestion. To overcome this difficulty we have devised an inexpensive little apparatus which has proved surprisingly effective for the removal of such fumes. As originally made it consisted merely of a broken pipette resting on top of the test tube in which the fumes were generated and drawing off the fumes by means of a water pump (an ordinary cheap one made entirely of glass). For the sake of safety we led the fumes through a large bottle containing a 10 per cent solution of sodic hydrate. We still use this bottle as an accessory although we now know that very little acid comes off and believe that it would probably be perfectly safe to let the fumes run directly through the pump and into the pipes that carry off the water from the latter. This arrangement had one drawback. There was always more or less condensation of acid and water in the body of the pipette and this acid solution would drain back into the test tube and thus delay the process of decomposition and in addition would drip on the table top when the absorber was removed from the test tube. To overcome this deficiency we sealed up the lower end of the pipette and while still hot and soft invaginated it by pushing the bottom upwards by means of a pointed stick of wood. 593 504 Fume Absorber A small hole was then made with a long wire nail in the tip of the invagination and we thus secured a capacious pocket large enough to hold the condensed water and acid obtained from a dozen digestions (see illustration). The fume absorber thus made is abundantly capable of taking care of all the fumes made in the ordinary Kjeldahl, or Neumann digestion as well, and we now do not use the hood at all for such purposes. With the additional help of funnels cut off half an inch above the stem and ground smooth on a wet grindstone, the apparatus, we find, can even be used for carrying off fumes from beakers and evaporating dishes. The accompanying photograph shows a somewhat more elaborate apparatus and how it may be used for carrying off fumes from test tubes, flasks, beakers and evaporating dishes. This apparatus is made for four exhaust tubes run by a single pump. To accomplish this a large bore glass tube carrying four side tubes connects on the one hand with the pump (or rather with the bottle containing the alkali) and on the other with the individual absorbers. The only important point about its con- struction is that the side tubes shall be of such a diameter that Otto Folin and W. Denis 505 the bent stems of the absorbers can just slip in for a distance of several.inches. The joint thus made, particularly when wet, is quite tight enough and no rubber connection is needed. One or two or all the exhaust bulbs can be used without changing anything and when any one exhaust tube has been used a number of times and is nearly full of condensed water and acid it is simply withdrawn, emptied and rinsed, and is again ready for use. We believe that the single absorber at each student’s desk might prove a valuable accessory in class room laboratories where hoods so often are inadequate and ineffective.' 1 Eimer and Amend now make the apparatus for us and it will be listed in their next catalogue. Atlas pats it ah | 15 7 | res nish Bpar %f ws ‘ ec * i ee Se - ‘ a= 6g str & rs labor’: hombetisal hin erihh datiu aul uw » canis: iti , % E : jour 16 a Pe~ wil cA? i a ON THE DETERMINATION OF UREA IN URINE. By OTTO FOLIN. WITH THE ASSISTANCE OF C. J. V. PETTIBONE. (From the Department of Biological Chemistry of Harvard Medical School.) (Received for publication, April 12, 1912.) I. Observations on my magnesium chloride method and on Ben- MUERTE Se ial 8 os oa cide Me RP ere reed SA gs ae eS 507 II. A new method in which the urea is decomposed with phos- Peeeiemree Peiemet foe oe ot), 320) OD TRA tas Pe Cle 512 III. A new method in which the urea is decomposed in boiling po- SPMMIRBEACCEALC SOLUGION «.. «.<.-)<5)« fa coco ajauths vn me nn Ane Bees a ae 513 IV. (With W. Denis) The determination of urea in the pres- TE ESL 2. Ba ea ae er pene cs, (2 Si oA cs 520 The apparent simplicity of the magnesium chloride method for the determination of urea in urine has proved rather deceptive and although this method probably has been used and is still used more frequently than any other in connection with metabo- lism work the literature which has grown up around it during the past ten years is not lacking in unfavorable criticisms. The chief source of error in the determination is due to incom- plete decomposition of the urea. Earlier investigators (Hugoun- enq, Kossel) had shown that urea is quantitatively decomposed when heated for a short time in sealed tubes or in the autoclave and had published methods for the determination of urea in urine based on that principle. In the magnesium chloride method the boiling point of the mixture containing the urine or urea solu- tions is raised by the addition of the salt and the rather high temperature needed for the speedy decomposition of the urea is thus secured in a most convenient manner. The essential point in the process is of course that a temperature of not less than 150° shall be maintained in the urea solution for the prescribed period. Lack of experience in how to obtain and to maintain this tem- 507 THE JOURNAL OF BIOLOGICAL CHEMISTRY, XI, NO. 5 508 Determination of Urea perature is the chief cause of the failures to get accurate results. One important factor in this decomposition of urea into ammonia and carbonic acid has, however, not been adequately recognized, namely this, that the time necessary for the complete decompo- sition of urea under uniform conditions of volume and temperature depends very materially on the amount of urea to be decomposed. That the time of decomposition of urea depends among other factors upon the amount taken is of course ‘an elementary, self- evident proposition but that the small amount of urea repre- sented by the difference between 5 cc. of dilute and 5 cc. of con- centrated urine may require almost a whole extra hour’s heating (at 150°C.) is anything but self-evident. In this fact is to be found the explanation why the time of heating in the method has been gradually increased from thirty minutes to an hour and a haif. It also indicates that the heating time can again be reduced to thirty or forty minutes by limiting the amount of urea taken for an analysis to a maximum of 60 or 70 mgms. In 1908 Kober? called attention to the fact that the long dura- tion of the distillation of the ammonia in the determination of urea is due not to the formation and subsequent decomposition of cyanuric acid derivatives, as I had suggested in my first paper on the subject, but that it is due to the presence of the large amounts of magnesium chloride. That such is the case I had found a couple of years before the appearance of Kober’s paper when I tried to remove the ammonia by distillation from a large batch (several pounds) of magnesium chloride. The hypothesis that condensation products similar to those obtained when urea is heated in a dry condition may be formed was advanced on the basis of the assumption that free water is practically absent from the mixture,? an assumption recently revived by S. R. Benedict? as explaining why the method is more accurate than the autoclave methods. Kober implied that it is practically impossible to distil off am- monia from solutions containing magnesium or calcium salts and by inference that my method for determining urea is hopelessly unsuitable. Without publishing any experiments on the subject 1 Journ. Amer. Chem. Soc., xxx, p. 1279, 1908. 2 Zeitschr. f. physiol. Chem., xxxvi, p. 336, 1902. 3 This Journal, viii, p. 415, 1910. Otto Folin 509 Kober leaves it to be understood as ‘‘quite obvious” from the mass law and the reversible reaction MgCl, + 2 NH,OH @ Mg(OH), + 2 NH.Cl that the ammonia cannot be obtained by distillations from solu- tions containing 15 to 20 grams of calcium or magnesium salts as in my method for determining urea. In a recent article on the determination of urea Henriques‘ and Gammeltoft have reproduced Kober’s deductive argument against my method in a rather more specific and positive form. Experiments on the subject are, however, still missing. But the mass law is only generalization. The reversible reac- tion quoted above conveys no information as to the final outcome of the distillation of ammonia in the presence of magnesium salts. The analogous reaction can equally well be presented for the ammonia distillation in the Kjeldahl process yet we know from experience that distillation yields satisfactory results. The question how completely ammonia can be obtained from ammonium salts under the conditions of distillation prevailing in the urea determination is of course easily determined experi- mentally. The results cited below were obtained in November 1908 in the order given from a standard ammonium sulphate solution (25 ee. of which contained 25.5 ec. 4; NH3) when dis- tilled for one hour with 15 grams of magnesium chloride,® 700 to 800 cc. water and 20 cc. 7.5 per cent sodium hydrate solution. (1) 25.5 (7) 25.4 (12) 25.5 (18) 25.45 (2) 25.5 (8) 25.4 (13) 25.3 (19) 25.35 (3) 25.25 (9) 25.5 (14) 25.4 (20) 25.4 (4) 25.25 (10) 25.5 (15) 25.6 (21) 25.35 (5) 25.5 (11) 25.5 (16) 25.5 (22) 25.4 (6) 25.5 (12) 25.3 (17) 25.5 (23) 25.4 When the amount of ammonia in the solution was doubled so that it contained 51 cc. 4 NHsz the results of the distillation 4 Skand. Arch. f. Physiol., xxv, p. 154, 1911. 5 The magnesuim chloride used was free from ammonia and as it was a fused salt (in sticks) 15 grams were used instead of 20 grams for these dis- tillations just as in the urea determinations. 510 Determination of Urea were less satisfactory. After one hour’s distillation the following figures were obtained: (1) 50.3 (5) 49.9 (9) 50.1 (2) 50.3 (6) 49.8 (10) 50.25 (3) 50.0 (7) 49.9 (11) 49.9 (4) 50.3 (8) 50.0 When distilling according to the directions given for urea deter- minations, 7.e., until the distillates failed to give an alkaline reac- tion with litmus paper, the figures cited below were obtained (in about one hour and twenty minutes): (1) 50.8 (3) 50.65 (5) 50.65 (7) 50.85 (2) 50.7 (4) 50.5 (6) 50.6 (8) 50.7 The above results are in harmoay with the fact that so many different investigators not only in my laboratory but in many other laboratories have obtained satisfactory results for the nitro- gen of pure urea solutions. The difficulties in so far as there have been any have come when urines of various concentrations were substituted for the urea solutions and the chief cause of these difficulties as mentioned above has been the incomplete decomposition of the urea. The chief criticisms raised against the magnesium chloride method for the determination of urea is not that it is less accu- rate than any other method but that too much skill, experience and time is necessary in order to obtain reliable results.® In his last paper on the estimation of urea S. R. Benedict (loc. cit.) describes a new method which he believes to be very accurate, giving figures slightly lower than those obtained by means of the magnesium chloride method, though “the agreement between the two methods is often as close as two duplicate determinations by the same method.” In working on pure products, creatinine, uric acid, and allan- toin, Benedict finds that whatever difference there is between the two methods is rather in favor of his new one. He clearly recognizes, however, that the ammonia obtained from those pro- * It is interesting to note that this criticism comes chiefly from Amer- ican laboratories where metabolism experiments for the past few years have been conducted on a wholesale, factory-like basis. Otto Folin 511 ducts in urine work is negligible and accordingly recommends the use of sodium hydroxide instead of sodium carbonate for distilling off the ammonia. As a matter of fact the method which he thus recommends would yield according to his own experiments fully as much ammonia from creatinine + uric acid as does the ‘‘ Folin” method; and he did not try this method with allantoin. As the method he did try with allantoin decomposed more than 50 per cent of 30 mgms. it is reasonably certain that it would decompose quantitatively such small traces of allantoin as may be present in 5 cc. of urine. The following urea determinations in urines made by Benedict’s method (using sodium carbonate as alkali) and by the magnesium chloride method show that the two do indeed yield substantially the same results. The figures represent grams of nitrogen per liter of urine. NEW PHOSPHORIC ACID * BE y NEDICT’S METHOD es FOLIN'’S METHOD 5.1 5.2 5.2 6.6 6.6 6.5 2.9 2.8 2.8 2.7 2.7 2.7 3.3 3.4 3.3 4.4 4.2 14.1 5.9 5.8 5.7 7.5 7.4 17.4 2.8 2.8 12.7 6.5 6.5 2.0 2.0 2.0 8.8 8.7 8.7 8.6 8.5 8.6 7.4 7.2 7.3 8.9 8.8 8.8 a 9.4 9.4 * See p. 512. The magnesium chloride method as used for the above determi- nations has been somewhat simplified in that the decomposition is carried on in a Kjeldahl flask (capacity 500 cc.) by the help of a small so-called micro-burner. A large test tube filled with cold water and suspended in thé neck of the flask by means of a cork or copper wire is used as a condenser. Only one-half of 512 Determination of Urea the test tube should be inside the Kjeldahl flask. After adding about 350 ee. water (hot) and alkali the ammonia can be dis- tilled off directly from the Kjeldahl flask in about an hour not- withstanding the higher initial concentration of the magnesium chloride. II. On p. 508 I indicated that since the decomposition time depends in a large measure on the actual amount of urea to be decom- posed that time could be very materially reduced by diluting the urine so that not over 60 or 75 mgms. of urea is used for each determination. At best, however, the determination would prob- ably require over two hours by the magnesium chloride method. The task which I have endeavored to accomplish was to evolve a method for the determination of urea which should at least, approximate in speed and convenience the method for total nitro- gen described in the preceding paper. The problem was to decompose one or a few milligrams of urea and either titrate the ammonia with very dilute acid and alkali or to determine it colorimetrically by means of Nessler’s solution. The magnesium chloride was found not to be suitable as a means of producing the necessary temperature on such a small seale. The procedure described below accomplishes the purpose fairly well with 1 ee. of undiluted urine. Measure the urine (1 ec.) with an Ostwald pipette into a Jena test tube. Add three good sized drops of pure phosphoric acid, one drop of indicator (alizarin red), a few grains of talcum powder and boil the mixture over a free flame until about one-half of the water has escaped. This requires only two to three minutes. Now place the test tube in a bath (paraffin, oil, or sulphuric acid), previously heated to 175° to 180°C. for fifteen minutes. The urea is completely decomposed in that time. The content of the tube is then dissolved by the addition of water (1 to 2 ec.) and a little heat. After adding 0.5 to 1 cc. of 50 per cent caustic potash? the ammonia is removed by a strong air current in ten minutes. It is collected in 25 ce. of sy hydrochloric acid and the excess of the acid is titrated with =, sodium hydroxide using 7 KOH is better than NaOH because of the greater solubility of potas- sium phosphate. Otto Folin 513 alizarin red as indicator. With the paraffin bath in order this determination can be finished in about ‘half an hour. Results obtained in this manner and calculated in grams per liter are cited on p. 511. No one would hesitate to consider those figures satis- factory. They were obtained by Mr. Pettibone only after several months’ fruitless endeavor in other directions. Iil. The method just described while representing a great saving of time when compared with any other reliable method was not considered entirely satisfactory. Like Benedict’s new method it depends on a bath, kept at a certain temperature, for the heat that is to decompose the urea. While this may not be much of a drawback, particularly if one has to make a large number of determinations at the same time, still it is a drawback that it seemed worth while to endeavor to get rid of. To solve the prob- lem I have returned to the principle used in the magnesium chloride method, 7.e., the use of a salt to obtain the high boiling point necessary for the speedy decomposition of urea. The salt finally adopted for this purpose is potassium acetate. By means of this salt any temperature up to 158° to 160° can be obtained. Potassium acetate is unfortunately somewhat hygro- ~ scopic though less so than magnesium chloride. The hygroscopic quality is, however, more objectionable in the new method about to be described because one of the advantages striven for is to get around the preliminary boiling off of water called for in the phosphoric acid method just described as well as in my earlier magnesium chloride method and in Benedict’s method. With any dry salt and a definite amount of water any given tempera- ture which that salt is capable of giving might be obtained at once without any preliminary concentration provided enough of the salt is taken. Being rather hygroscopic, the different brands of potassium acetate on the market differ markedly in the amount of water they contain. The best German brands are sufficiently dry for the purpose here involved while the American brands, as at present sold, contain very much more water and should be dried before being used. The salt loses its water very readily and we dry it, about a pound at a time, by having it in a large 514 Determination of Urea porcelain dish standing on a warm plate (at about 115°) for about twenty-four hours. The plate must not be too hot as the acetate decomposes rather easily. The method described below is based on the use of such dry salt. An important accessory in this new procedure for the deter- mination of urea is a temperature indicator. This indicator was originally devised for use in connection with the magnesium chloride method but it has proved less useful there than in con- nection with the new method because in the old method so much coloring matter is formed as to obscure the indicator. This tem- perature indicator consists of powdered chloride-iodide of mercury (HgICl) inclosed in a sealed glass bulb not over 1 mm. in diameter. This salt is bright red at ordinary temperatures. At 118°C. it turns lemon yellow and melts to a clear dark red liquid at 155°C. It solidifies again at about 148°C. and resumes its red color grad- ually only in the course of about twenty-four hours. The melting point temperature 153°C. is fortunately a temperature very read- ily obtained and maintained by means of potassium acetate and as the acetate begins to cake and solidify at 160° to 161° there is no danger in this combination of having either too high or too low a temperature without its being unmistakably apparent. The salt in question, HgICl, is prepared by heating in a dry state intimately mixed mercuric chloride and mercuric iodide in molecular proportions at 150° to 160°C. for six to eight hours. At the end of the heating the product should be powdered and used as it is for it cannot be purified by the use of solvents. It should be kept dry until sealed up as indicated.® Since the urea according to this method is decomposed in a practically saturated potassium acetate solution, the acid to be used for retaining the liberated ammonia must of course be acetic acid. Acetic acid in the presence of so much acetate is an ex- tremely weak acid. In fact it is barely capable of holding the ammonia under the conditions of the determination so that for a time it was thought that the low results which were constantly obtained were due to the escape of ammonia. The decompo- 8 Kohler: Ber. d. d. chem. Gesellsch., xii, p. 1187, 1879: The indicator properly sealed up in bulbs as well as the other special appliances needed in this determination can, however, be obtained from Eimer and Amend, New York. Otto Folin 515 sition of the urea in this method may therefore be said to be accom- plished in an almost neutral medium. As indicated by alizarin red, the medium is neutral or alkaline, certainly not acid. The method is as follows: The urine is diluted so that 1 cc. contains 0.75 to 1.5 mgms. of urea nitrogen. Dilutions of 1 in 20, 1 in 10 or rarely 1 in 5 are usually adequate for this purpose. One cubic centimeter of the diluted urine is then transferred by means of an Ostwald pipette to a large Jena test tube (200 mm. by 20 mm.) previously charged with 7 grams of dry potassium acetate (free from lumps), 1 ce. of 50 per cent acetic acid, a small sand pebble, or better, a little powdered zinc (not zinc dust) to prevent bumping during the boiling, and a temperature indicator. The test tube is then closed by meansof a rubber stopper carry- ing an empty narrow “calcium chloride tube” (without bulb) as a condenser (size of calcium chloride tube, 25 cm. by 1.5 cm.). The test tube and condenser are then suspended by a burette clamp or similar device so that it can easily be raised or lowered with reference to the small flame of the micro-burner. As soon as the acetate is dissolved and the mixture begins to boil, which usually occurs in about two minutes, the indicator begins to melt showing that the desired temperature (153° to 160°C.) has been reached. -The boiling is continued in a gentle, even manner for ten minutes at the end of which time the decomposition of the urea is already completed. The apparatus is removed from the flame and the contents are diluted by the addition of 5 ce. of water. The water is added by means of a pipette through the calcium chloride tube so as to rinse the sides of the tube and the bottom of the rubber stopper from traces of ammonium acetate which may be there. Not more than 5 ce. of water should be used for this purpose. An excessof alkali,2 cc. of saturated sodium hydrate or potassium carbonate solution, is added and the liberated am- monia is driven off by means of a strong air current into a 100 ec. measuring flask containing about 35 cc. of water and about 2 cc. of 4, acid. The time required for this will of course depend on the strength of the air current. In this laboratory ten minutes is allowed and is abundant. The ammonia thus set free is deter- mined colorimetrically against 1 mgm. of nitrogen in the form of ammonium sulphate exactly as in the total nitrogen determi- nation described in the preceding paper. 516 Determination of Urea In execution the determination of urea described above is about as simple and free from complications requiring unusual skill or experience as it is possible to make a quantitative method. While in the process of development, however, this was not the case and it sometimes appeared as though it would not be possible to find the conditions which could be depended on to yield theo- retical results. For a long time the results were almost invariably too low al- though an occasional theoretical figure showed that such was not necessarily the case. The deficiency in the ammonia found was supposed to be due to the inability of acetic acid to prevent its escape and numerous futile efforts were made to detect the loss and to prevent it. In time the losses were found to be due to the formation of condensation products which do not give up their ammonia to the air current and it was further found that the acetic acid concentration or the absence of water was the factor which determined this formation. Because of the weak- ness of acetic acid in concentrated acetate solution, glacial acetic acid rather than dilute acid was used to retain the ammonia. This was wrong. With glacial acetic acid and dry acetate, whether two or three drops or any larger quantity is used, the results were almost invariably too low. And the greater the amount of acid taken the greater was the loss of nitrogen. This fact suggested that probably acetamide was formed. But when ammonium sulphate was substituted for urea there was no loss. Later it was found that when 1 ce. of ammonium sulphate solution containing 5 mgms. of nitrogen or over was used with glacial acetic acid, all ofthe ammonia could not be recovered by means of the air current, though it could be obtained by distillation, thus showing that acetamide was probably formed. But since urea behaved similarly when only 1 mgm. of urea nitrogen was present it was clear that the amide formation could not be the cause of the failure to recover it all. The acetamide theory furnished, however, the solution of the problem from the analytical standpoint. By substituting 50 per cent acetic acid for the anhydrous acid the difficulty dis- appeared. Urea corresponding to as much as 5 mgms. of urea nitrogen will be completely decomposed by ten minutes’ boiling with 7 grams of potassium acetate and 1 cc. of 50 per cent acetic Otto Folin 517 acid and the ammonia will be recovered quantitatively by means of the air current. Five or six milligrams of nitrogen represents, however, the upper limit under the conditions described. If 10 mgms. of nitrogen are taken, whether inthe form of ureaor of ammo- nium sulphate, 50 per cent acetic acid does not entirely prevent the formation of more or less stable condensation products. When as much as 10 mgms. of ammonia are present there is also danger of losing some mechanically for in the upper half of the test tube there is then an abundance of ammonia as well as of acetic acid vapors during the boiling. It is of course desirable that this ammonia be kept down as near the boiling liquid as is practicable. Consequently it is desirable first to avoid bumping and secondly to keep the steam pretty well confined. To keep the steam down the amount of water present in the system must be kept low. It is possible to get a temperature of 153°C. and over with only 3 grams of potassium acetate and 2 ce. of water by boiling the mixture so hard that the surplus water is constantly kept circulating in the upper part of the test tube. The whole test tube and the lower half of the condenser as well, will then be very hot from contact with the steam. A similar result is of course obtained by using 7 grams or even more of acetate which is not dry. By using 7 grams of reasonably dry acetate, however, one obtains with 2 ce. of water a mixture which can be gently boiled above 153°C. with the evolution of so little steam that the upper part of the test tube remains quite cool. The flame from the micro-burner necessary to maintain boiling in such a solution need not be over 0.5 ce. long and, at that, the bottom of the test tube must be some distance above it.’ _ If too much heat is applied the acétate cakes at the bottom of the mix- ture, if too little it cakes at the top. With the small flame froma micro-burner and a windshield it is, however, very easy to keep solution boiling without caking.?° A few additional points should be mentioned in connection with this new method for determining urea. 2 Bottomless beakers make excellent windshields for such small flames and wind-shields of some sort are indispensable in most laboratories. 10 Such a boiling solution was once left over night and was found in the same condition, 7.e., boiling and clear, in the morning. 518 Determination of Urea 1. When the urea is decomposed in boiling acetate solution at 150° to 160° that solution as already indicated is only faintly acid. The solution does not retain quantitatively either acetic acid or ammonia and a certain amount of each is present in the vapors above the boiling mixture. In a pure water solutionof ammonia and acetic acid, on the other hand, the ammonia does not escape with the vapors when boiled provided the amount of ammonia present is not too large. The acetic acid vapors above the acetate solution therefore probably help very much to keep the ammonia from escaping. 2. Bumping in a boiling test tube is always disagreeable. In this case no bumping whatever is wanted, first because the vapors inside are charged with more or less ammonia and secondly be- cause in a bumping solution the acetate will suddenly cake at the bottom. Bumping is easily prevented by the presence of a rough piece of gravel the size of an ordinary glass bead. A small pinch of powdered zinc is even better than the pebble for this purpose. The acidity of the solution is so weak that the action on the zine in spite of the high temperature is very slight. 3. Curiously enough the presence of the zinc appears to some- what modify the hydrolytic power of the hot acetate mixture. It reduces to a marked degree the decomposition of allantoin yet does not interfere with the decomposition of urea. In the presence of zinc not over one half of half a milligram of allantom-N can be recovered. 4, The essentially neutral acetate mixture used in this method represents probably the mildest direct hydrolysis yet applied for the purpose of determining urea in urine. Other urinary constitu- ents, except of course the ammonia, contribute probably very little indeed to the result. Neither creatinine nor hippuric acid gives even a trace of ammonia. Uric acid sometimes seems to give enough to make the qualitative test positive, at other times the qualitative test is negative and, at all events, the test (whether much or little uric acid is taken) is quantitatively imperceptible when made in the presence of standard urea solutions (see next page). Allantoin, as already indicated, may give off about one half of its nitrogen in the presence of zinc, otherwise it behaves like urea provided its quantity does not exceed 0.5 mgms. of allantoin-N. The darkening of urine, conspicuous in the mag- Otto Folin 519 nesium chloride method, is almost entirely absent in the acetate mixture even when undiluted urine is subjected to the treatment. In this respect the marked charring effects obtained in the phos- phoric acid method described above, as well as in Benedict’s acid sulphate method, is rather disconcerting, though apparently harmless." 5. As with the total nitrogen determination described in the preceding paper the best air current is compressed air since the ammonia can then most conveniently be collected directly in the 100 cc. measuring flask. Suction with a good water pump can be used, however, and as we have satisfied ourselves repeatedly, will also take out all the ammonia in ten minutes. The most rapid stream of which the pump is capable should always be used for there is no danger of losing any ammonia (see p. 523). 6. For preparation of the standard ammonium sulphate solu- tion, the Nessler solution, and for the details of the color com- parison, etc., see the preceding paper. 7. This method has been designed primarily for a colorimetric reading of the ammonia. As the method is perfectly reliable for larger quantities of urea up to 5 mgms. of nitrogen titrations with 35 solutions may be applied. With a colorimeter at hand the colorimetric method is, however, equally convenient and rather more accurate. 8. The test tubes in which the urea is to be decomposed should be dry. The amount of water present in a freshly rinsed test tube is considerable, relative to the total amount present in the reagents. The parallel urea determinations recorded below were made for the purpose of determining whether this new method gives essentially the same values as the magnesium chloride method. I expected rather lower results with the new method but this 1 The home dried potassium acetate used in these determinations was as a matter of fact not entirely free from ammonia when tested qualita- tively for it. The trace found was sometimes a trifle increased, sometimes not, after uric acid had been heated in the mixture. In no case, however, did these traces appreciably affect the color of the ammonia correspond- ing to 1 mgm. of nitrogen. Potassium acetate containing less then 1 per cent of moisture and free from ammonia is now made for us by J. T. Baker Chemical Company, Phillipsburg, N. J. 520 Determination of Urea expectation was not realized. The differences are immaterial. The figures represent grams of urea-N per liter of urine. COLORIMETRIC METHOD MAGNESIUM CHLORIDE METHOD 14.2 14.1 8.2 8.2 8.4 8.5 6.7 6.8 10.5 10.6 8.6 8.7 5 3.4 | ww oo IV. The determination of urea in urines containing sugar has been recognized as a special problem ever since the publication in 1903 of Morner’s illuminating paper on the different methods then available for the determination of urea in human urine. Mo6rner’s own procedure” for the preliminary removal of the sugar has remained a tedious but indispensable prerequisite for the deter- mination of urea in such urines. None of the methods published since that time represent any improvement in this respect and the determination of urea in diabetic urines is still a comparatively long and laborious operation. The colorimetric potassium ace- tate method described above appeared at first to be no more suitable for sugar urines than any other. In the presence of sugar the results obtained were invariably from 20 to 50 per cent too low. A more systematic investigation of the subject has, how- ever, shown that it is possible by means of this method to meet the unusual conditions which must be fulfilled if urea is to be quantitatively converted into ammonia in the presence of sugar. The reason why sugars interfere with the decomposition of urea was formerly ascribed to the formation of nitrogenous ‘‘melanins”’ but the loss of nitrogen is in all probability due to the formation of definite, stable ureids."’ The difficulty involved is therefore analogous to the difficulty encountered in the use of acetic acid (see p. 516). The disturbing effects of acetic acid were overcome % Morner: Skand. Arch. f. Physiol., xiv, p. 319. 1 Folin:» Amer. Journ of Physiol., xiii, p. 46, 1905. Otto Folin | 521 by reducing its concentration below the point at which it begins to give condensation products with urea. The remedy against the ureid-forming tendency of the sugars is the same. When the amount of sugar present is sufficiently small the combination with urea does not take place and the results obtained are quantitative. The dilution necessary to prevent the fatal ureid formation in the case of dextrose is, however, very great, so great in fact, as to be entirely out of reach in the titration methods for deter- mining urea. The presence of 10 mgms. of dextrose in the acetate mixture used to decompose urea (about 2 mgms.) is accompanied by a loss of 40 to 50 per cent of urea nitrogen. With 5 mgms. of dextrose present the loss of urea nitrogen sinks to about 20 per cent and this loss remains about the same whether the urea nitrogen present is 1 mgm. or is reduced to one half or even to one-tenth of that amount. The losses due to sugar depend therefore chiefly on the amount of sugar present and only to a much smaller extent upon the amount of urea to be decomposed. One milligram of urea-nitro- gen is, however too large an amount for a determination in the presence of sugar. With this amount a loss of about 5 per cent is encountered in the presence of 1 mgm. of dextrose. - With the colorimetric method for determining ammonia, a method which until now has been used only in water analysis, it is of course possible to determine much smaller amounts of urea than those corresponding to 1 mgm. of nitrogen. In fact it was only by virtue of the special adaptation of the method worked out in this laboratory that it became possible to work with as much as 1 mgm. or more of nitrogen. One tenth of a miligram of nitrogen can be determined with a very satisfactory degree of accuracy by diluting the Nesslerized ammonia to only 10 cc. (instead of 100 cc.) before reading the color. But experiments have shown that 0.1 to 0.3 mgm. of urea nitrogen can be deter- mined in the presence of as much as 2 mgms. of dextrose. It is therefore possible by simply diluting diabetic urine until 1 ce. contains about 0.1 mg. of urea nitrogen to determine the urea without any preliminary removal of the sugar when the dextrose- nitrogen ratio (D:N) is as high as 20:1. The determination is made as follows: 1 ce. of urine previously 522 Determination of Urea diluted from 20 to 100 times is decomposed in the usual manner with the potassium acetate and acetic acid. The ammonia is then driven into a second test tube containing about 2 ec. of water and 0.5 ec. of 4 hydrochloric acid. To the contents of this test tube are then added first, a couple of cubic centimeters of water, then 3 ee. of diluted (1:5) Nessler’s solution. The colored solution obtained is then rinsed and washed into a 10 cc. measuring flask and the volume made up to the 10 cc. mark. The whole is trans- ferred to a dry cylinder of a Duboseq colorimeter andthedepth of the color is determined in the usual way against the standard containing 1 mg. of nitrogen per 100 cc. of solution. The following determinations may be cited to show the extent to which the figures for the urea come up with the dilution of urines containing sugar. VOLUME OF URINE DILUTION per cent a ov G2 09 00 09 He de Cosa a The following figures were obtained after adding 10 per cent dextrose to normal urines of known urea content. 1 = - TRUE UREA NITROGEN DILUTION | UREA NITROGEN FOUND PER 100 cc. 1:10 Ve 0.62 | 0.97 i} 1:100 0.98 0.97 1:10 0245 0.69 fl \} 1:10 0.69 0.69 (3) fp b240 0.32 0.50 ue 1:50 0.50 0.50 ON THE DETERMINATION OF AMMONIA IN URINE. By OTTO FOLIN anp A. B. MACALLUM. (From the Biochemical Laboratory of Harvard Medical School.) (Received for publication, April 12, 1912.) In Folin’s air current method for the determination of ammonia 20 or 25 ce. of urine is used and from this volume of liquid all the ammonia can be removed in from one to three or four hours, the time depending on the rapidity of the air current. The accu- racy of that method has never been questioned. But a number of investigators have abandoned the attempt to make use of it because they did not have a strong enough air current to work with and others, as indicated above, have had to run their air currents several hours in order to obtain all the ammonia when an hour to an hour and a half should be enough. They have assumed that the water pressure in their laboratories has not been sufficient to produce the required air current. This is a mistake. A pressure of 40 to 45 pounds per square inch is probably available in most laboratories and such a pressure is sufficient to produce a very effective air current provided the water pump used is a good one.! In the paper on the air current method for determining ammonia attention was called to the fact that the rapidity with which a given air current removes ammonia from solutions depends very much on the volume of the solution. To shorten the time of the determination of ammonia in urine it is therefore only necessary to reduce the volume of urine used. In the two preceding papers 1The water pump listed in the Kny-Scheerer Company’s Catalogue (List 120), No. 2458, p. 272, produces, when properly adjusted, an entirely adequate air current with such a water pressure. Its only drawback is that its attachment nut does not fit any American made thread and its attach- ment therefore requires the help of a mechanic. It should not be bought without the vacuum gauge because the latter makes the adjustment to the point of maximum efficiency very much easier. 523 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 95 524 Determination of Ammonia (on the determinations of total nitrogen and of urea) it was shown that ammonia could be removed quantitatively from 10 cc. of solution in ten minutes or less. The application of the technique described in those papers to the determination of ammonia is more or less self-evident.2._ The ammonia determination according to this method is carried out as follows: Into a test tube measure by means of Ostwald pipettes 1 to 5 ec. of urine. (The volume taken should give 0.75 to 1.5 mgms. ammonia-nitrogen. With normal urines 2 cc. will most often give the desired amount. With very dilute urines 5 cc. may be required, while with diabetic urines rich in ammonium salts even - 1 ce. may give too much and the urine must be diluted.) Add to the urine a few drops of a solution containing 10 per cent of potas- sium carbonate and 15 per cent of potassium oxalate, and a few drops of kerosene or heavy, crude machine oil (to prevent foaming). Pass the strong air current through the mixture for ten minutes (or as long as is necessary to drive off all the ammonia) and collect the ammonia in a 100 cc. measuring flask containing ‘about 20 ce. of water and 2 cc. of #4 acid. Nesslerize as described in the paper on the total nitrogen determination, p. 495, and compare with 1 mgm. of nitrogen obtained from a standard ammonium sulphate solution and similarly Nesslerized. Results obtained by this rapid methed and parallel results ob- tained by the original air current method are given in the table on the opposite page. It is clear that the figures obtained by the new rapid process are practically identical with the figures obtained by the old air current method. No absolutely sharp end-point is obtainable when a rapid air current is passed through urine. A trace of something capable of giving a color with Nessler’s solutions continues to come long after all the ammonia has been removed. This is a point of dis- tinction between urine and ammonium salt solutions. What this 2 The determination of ammonia as described in this paper is not so far as the development is concerned a mere application of the principles de- scribed in the preceding papers. On the contrary the investigation of this method was started simultaneously with the other two and was finished first. Otto Folin and A. B. Macallum Grams ammonia-nitrogen per liter of urine. 525 FOLIN’S METHOD bi baie | FOLIN'S METHOD pgs ares 0.55 0.58 0.80 0.80 0.52 0.54 0.60 0.60 0.44 0.50 0.62 0.62 0.45 0.48 0.26 0.27 0.40 0.45 1.35 1.38 0.43 0.44 0.51 0.54 0.47 0.47 0.44 0.48 0.42 0.42 110 1.09 0.37 0.39 1.99 2.00 0.48 0.50 xl substance is we do not know, though we have devoted considerable time to its investigation. The effect of this substance in actual ammonia determinations is so small as to be hardly, if at all, perceptible. ete \ . : mt) i i 1 j ys 8 Wet: fi bie ge IO. . 1% ay ae Te rhe Wat ‘ de ob As ai cts ‘ Ms Pon deed tat ; ,a uh *) OVARY it 4 eb id Pare iv act oe eo) TER 3S NEW METHODS FOR THE DETERMINATION OF TOTAL NON-PROTEIN NITROGEN, UREA AND AMMONIA IN BLOOD. By OTTO FOLIN anp W. DENIS. (From the Biochemical Laboratory of Harvard Medical School, Boston.) (Received for publication, April 12, 1912.) evtennndmomdrawing blood: ....... 60.0 cusken ters hewte sides» oh ok wu ee 527 II. Isolation of non-protein nitrogenous constituents................ 528 III. Determination of the total non-protein nitrogen................. 529 IV. Determination of the urea.................. Preto AST GER ot See 531 VeeDetermmationyof the-ammonia. 2...) oe.) soe): Sees. chases ow se. 532 The analytical technique described in the preceding three papers lends itself peculiarly well to the determination of total unco- agulabie nitrogen, urea and ammonia in blood, milk, eggs and other liquids where we are dealing with minute amounts of these different constituents. In two earlier papers results obtained by adaptations of these methods to blood analysis were published. ! The procedures by means of which these results were obtained are described in this paper. I. METHOD FOR DRAWING BLOOD. Before going into the details of the: chemical work it would seem worth while to describe our method of drawing blood because so far as we have been able to learn it is somewhat different from the procedures employed by physiologists and because we believe it to be expeditious, neat and exact and therefore particularly suitable for quantitative work. We use neither cannulae nor syringes but simply hypodermic needles and pipettes. The needles are about 1 mm. in diameter and about 25 mm. long. They are immersed in a dilute solution 1 This Journal, xi, p. 87; Ibid, p. 161, 1912. 527 528 New Methods for Blood Analysis of vaseline in ether and then allowed to drain and dry on a clean paper for a few minutes before being used. (This does not apply of course to the drawing of human blood when the needles must be thoroughly sterilized.) An adequate supply of these needles is kept on hand so that we do not need to use any needle more than once in any one experiment. The needle is attached to the tip of a 2 or 5 cc. pipette by means of a short piece of narrow pure gum tubing. A small pinch of powdered potassium oxalate is introduced into the upper end of the pipette (which must be clean and perfectly dry) and is allowed to run down into the tip and the needle. The other end of the pipette is connected with a rubber tube which in turn connects with a mouth piece consisting of a short tapering glass tube. Close to the pipette the rubber tube carries a pinchcock. To draw the blood one of us inserts the needle in the vein or artery and the other regulates the flow of the blood by means of the pinchcock and by suction. The exact quantity of blood desired is thus obtained without any waste and without clotting. II. ISOLATION OF NON-PROTEIN NITROGENOUS CONSTITUENTS. - To separate the non-protein nitrogenous constituents from the protein materials we make use of pure (acetone-free) methyl aleohol and an alcoholic solution of zine chloride. Ordinary methyl] alcohol cannot be used because the impurities in it, par- ticularly the acetone, combine with more or less of the urea so that it escapes. decomposition in the subsequent treatment and is not quantitatively recovered. We have satisfied ourselves by means of determinations on pure urea solutions that the presence of acetone results in a loss of urea. As soon as the blood is drawn it is transferred into measuring flasks half filled with methyl alcohol and the flasks are then filled up to the mark with methyl] alcohol and vigorously shaken. Two cubic centimeters of blood we dilute to 25, while for 5 of blood we use 50 ce. flasks. At the end of two hours, or as soon after that as is convenient, the contents of the flasks are filtered through dry filters. To the filtrate are then added two or three drops of a saturated alcoholic solution of zine chloride and after stand- ing for a few minutes the mixture is again filtered through a dry Otto Folin and W. Denis 529 paper. The zinc chloride brings down an appreciable precipitate and the last traces of coloring matters so that when the second filtration is made a perfectly colorless filtrate is obtained. Five cubie centimeters of these filtrates, corresponding to 0.4 or to 0.5 ec. of blood, depending on whether 2 or 5cec. of blood were drawn, are taken for each determination. The precipitation procedure described above is the one which we ordinarily use. There are objections to it. We are not certain that traces of protein-like materials may not escape precipitation by this as by every other method and we do know that the filtrate does not contain all of the non-protein materials. When relatively large quantities (equi- valent to 100 mgm. of nitrogen per 100 cc. of blood) of creatine, or asparagine are added to blood and treated as described above there is invariably an appreciable loss of material. To overcome this loss we have tried to triturate and wash the first alcoholic precipitate with methyl alcohol, and with some substances, as for example, with glycocoll, urea and acetamide, we are thus able to get practically quantitative results while with others, such as creatine, asparagine, and tyrosine, we still do not get quite all. Moreover, such trituration and washing does leach out a small amount of the coloring matters of the blood so that except for special experiments with less soluble substances we consider the simpler procedure rather more satisfactory. In the case of muscle analysis, on the other hand, we thoroughly tri- turate and wash with the alcohol. Incidentally it should be said that, muscles as soon as cut out, while still twitching, are cut with a pair of sharp scissors and immediately immersed in methyl alcohol (about 50 cc. in an Erlenmeyer flask). After being allowed tostand for afew hours the coagu- lated muscle is thoroughly ground up and then extracted overnight with a fresh portion of alcohol. The various extracts and washings are then combined, filtered into a 100 cc. volumetric flask and after the addition of a few drops of alcoholic zinc choride solution, made up to volume with methyl alcohol and again filtered. We invariably start with 5 grams of muscle and use 10 cc. of the filtrate for each determination of total nitrogen as well as of urea. Ill. DETERMINATION ©F THE TOTAL NON-PROTEIN NITROGEN. ‘To determine the total non-protein nitrogen of the blood 5 cc. of the alcoholic filtrate is transferred to a large Jena test tube of the same kind as is used ip urine analysis (see p. 494). One drop 530 New Methods for Blood Analysis of sulphuric acid, one of kerosene and a pebble are added and the methyl alcohol is driven off by immersing the test tube in a beaker of boiling water for five to ten minutes. When the alcohol is removed 1 ec. of concentrated sulphuric acid, a gram of potassium sulphate, and a drop of copper sulphate solution are added and the mixture is boiled, cooled and diluted as in the analysis of urine (see p. 494). From this digestion mixture the ammonia is removed in the usual manner. It is, however, not collected directly in a measur- ing flask (as in urine analysis) but in a second large test tube previously charged with 1 ce. of 75 acid and 2 to 3 ce. of water. The reason for this variation is that 0.4 to 0.5 ec. of blood contains only 0.1 to 0.2 mgm. of non-protein nitrogen. The final Ness- lerized solution cannot be diluted to 100 ec. and smaller volu- metric flasks cannot be used as receivers during the air current —— treatment because of spattering. Large test tubes are efore used as receivers and the ammonia is Nesslerized in these before the liquids are transferred to measuring flasks. Ordinarily the colored solutions obtained from cat’s blood are transferred to 25 ec. flasks and are then found to have a depth of color which permits of a sure and accurate reading in the colorimeter. In some of our absorption experiments the total non-protein nitrogen runs up to very high figures and then the solutions are diluted to 50, sometimes even to 100 cc., before being read in the color- imeter. Human blood contains scarcely more than one half as much non-protein nitrogen as cat’s blood. In the case of human blood we therefore never draw less than 5 ec. and we take 10 ce. of the filtrate for each determination. In all other respects we use the same procedure for human blood as for cat’s blood. In all ordi- nary cases 7 to 8 cc. of diluted Nessler’s reagent (dilution 1:5) are added for the production of the color. If much ammonia is present so that the resulting colored solution must be diluted to 50 or 100 ce. correspondingly larger amounts of Nessler’s reagent are added. The calculation of the analytical results to milligrams of nitro- gen per 100 ce. of blood is not difficult but the formulae given below may prove useful. In these formulae the standard solu- tion contains 1 mgm. of nitrogen (as ammonium sulphate) Ness- Otto Folin and W. Denis 531 lerized in a 100 ce. flask and the colorimeter prism of the standard is set at 20 milimeters. = x D in which RF stands for the reading of the unknown and D represents the volume to which its ammonia has been diluted gives the desired figure. The reason for the figures is that we are here working with 0.4 cc. of blood. When 5 ce. of blood is taken and it is diluted to 50 the formula 40 = D. becomes RP P< When working with human blood and taking 10 cc. of the fil- trate obtained from 5 cc. of blood diluted to 50 the formula is 20 R <_D, It may be thought that we are using unnecessarily small amounts of blood in these analyses. We are, however, by no meanssure that: working with larger amounts would yield more accurate re- sults and we have satisfied ourselves by scores of duplicate analyses that the method as outlined gives trustworthy figures. Further, the smaller the quantity of blood which can be made to give reliable results the greater becomes the usefulness of the method. The work which we have already done on cats could not have been done on such a small animal except by means of these micro- chemical methods. Finally, small amounts of blood must be used for the urea determinations because of the disturbing effects of the sugar present (see p. 520). IV. DETERMINATION OF UREA. Having described in some detail the preliminary treatment of the blood for the removal of the proteins and also the procedure for determining the total non-protein nitrogen, the urea deter- mination in blood can be described very briefly. Five cubic centimeters of the alcoholic filtrate from cat’s blood (or 10 ce. from human blood) are taken for each determination. This amount is measured into one of the large Jena test tubes in which the decomposition is to be made. A drop of dilute acetic acid and two or three of kerosene are added and the test tube is then closed by a two-hole rubber stopper. Through one of the holes in the stopper passes a glass tube drawn out to a capillary 532 New Methods for Blood Analysis several inches long. The capillary end reaches nearly to the bottom of the test tube. Through the other hole passes a short bent glass tube which is connected with a good water pump (see p. 523). The test tube is placed in warm water and the vacuum pump is started. In ten to thirty minutes the combined action of the gentle heat, the air current (through the capillary) and the vacuum removes all the alcohol. The rubber stopper is then removed and the capillary tube is broken off by bending it against the sides of the test tube and is left there. Two cubic centimeters of 25 per cent acetic acid, a temperature indicator, a pebble and 7 grams of dry potassium acetate are added and the decomposition of the urea is accomplished by heating it to 153 to 158°C. for about eight to ten minutes exactly as in the urea determination described for urine (see p. 515). The ammonia set free by the subsequent air current treatment is collected in a large test tube, there Nesslerized veo only 3 ec. of the diluted reagent), is made up te-volume in a 10 ce. volumetric flask and the color comparison is made as in the case of the total non-protein nitrogen against the same standard solu- tion of ammonium sulphate. We usually Nesslerize the total nitrogen, and the urea, and the standard, all at the same time. Since only 10 ce. is available of the solution corresponding to the urea, all of it must be poured into the Duboseq colorimeter cylinder for the making of the color comparison. Dry cylinders must therefore be used. If only one cylinder is available the urea should be read first. ‘We find it extremely convenient, however, to have several extra cylinders for the colorimeter and are thus able to read a series of urea determinations without stopping to rinse and wipe the inside of the cylinder for each determination. V. DETERMINATION OF THE AMMONIA. The accurate determination of the ammonia in blood is beset with far greater difficulties than any of the earlier inventors of methods for its estimation have realized. The blood decomposes spontaneously (and particularly in the presence of alkalies capable of setting free the ammonia) at all temperatures even when kept on ice. The ammonia thus produced by decomposition in the course of a few hours is much greater than the preformed ammonia Otto Folin and W. Denis 533 . present in the strictly fresh blood and when distillation methods are applied, whether in the vacuum or otherwise, the determination becomes little else than a measure of the decomposition. The decomposition in tissues such as the liver is even greater than in the blood and for this reason (among others) we are of the opinion that there is not a single experiment on record proving that macerated liver tissue splits off by hydrolysis the NH» groups from ordinary amino-acids when the latter are added to such tissue. . In view of the instability of blood or of certain components of blood the determination of its ammonia can be accomplished with a reasonable degree of accuracy only by the speediest kind of a process. Having oncethoroughly realized this fact the problem of determining this ammonia became with us a problem of learn- ing to work with the smallest possible amount of material—a serious problem in view of the minute quantities of ammonia present in normal blood. The Nesslerization process lends itself as does none other to the quantitative estimation of small amounts of ammonia but instead of working with milligrams, as in urine, or with tenths of a milligram, as with blood in the estimation of total nitrogen and urea, it became a question of working with hundreths of a milligram. The quality of the color produced by Nessler’s reagent with ammonium salt depends greatly on the amount of ammonia present, the tint is yellow or yellowish green when the amount of ammonia is very small (see p. 496) and such faintly colored solu- tions can not be read in a Duboscq colorimeter as ordinarily used. It would of course have been possible to fall back on the pro- cedure as it is used and has been used for a long time in water analysis, but we felt sure that this old process is not as reliable as the ammonia determinations we made by the help of a high grade colorimeter. By means of two important modifications of the Duboseq, colori- meter we have succeeded in meeting all the necessary conditions. The chief reason why a dilute Nesslerized solution cannot be read against a much stronger one is that the light is absorbed in passing through a deep layer of the solution. Two such fields cannot therefore be made to look alike. After having unsuccess- fully tried various kinds of screens for reducing the amount. of 534 New Methods for Blood Analysis light passing through the thin layers of concentrated solutions . we finally attained the desired result by the help of an iris dia- phragm attached to one side of the colorimeter. By means of this diaphragm we are able to make use of 0.5 mgm. of nitrogen as a standard and against it read a solution containing only a few hundredths of a milligram of ammonia nitrogen. The second modification consists in the use of a 100 mm. polari- scope tube as container for our unknown ammonia solutions instead of the cylinders which go with the Duboseq colorimeter. These cylinders are so large in diameter that the solutions would have to be made impracticably dilute in order to furnish a reason- ably high column. Ten cubic centimeters, for example, will reach to a height of only about 30 mm. in the Duboseq cylinders yet these are about the smallest colorimeter cylinders in the-mar=——— ket. With 10 cc. we can, however, comfortably fill a 100 mm. polariscope tube and, as it happens, such tubes just fit the Duboseq colorimeter when the solid movable glass-prism is removed. In the determination of the traces of ammonia here under discussion two precautions, not needed in any of the other methods described in the preceding three papers, are necessary. The first is that too much Nessler reagent must be avoided. The greenish tint observed in very dilute ammonia solutions when Nesslerized is almost wholly due to an excess of the reagent (see p. 497). The second precaution is the necessity of using only water that is strictly free from ammonia for diluting the unknown. The amount of ammonia in ordinary distilled water is sufficient to introduce a considerable error in this determination, while in those pre- viously described it does not matter, partly because the ammonia is so small as to be negligible in view of the fact that the standard and the unknown are diluted to about the same extent with the same water. In this case where we read through 100 mm. of the unknown solution against about 10 mm. of the standard the case is different and the ammonia of the water must be eliminated.’ ? Ammonia-free water is easily obtained from ordinary distilled water by the addition of a little saturated bromine water and a few drops of con- centrated caustic soda. See Claessen’s Tezxt-book, li, p. 116. Such water containing hypobromite and alkali cannot of course be used for the absorp- tion of the ammonia but only for diluting the reagent and for the final dilution to a definite volume. Otto Folin and W. Denis 535 We do not use such ammonia-free water for the small amount, 2 to 3 cc., employed for the absorption of the amimonia, but only for the water subsequently added in Nesslerizing and making up to a volume. The method for the determination of the ammonia is as follows: Ten cubic centimeters of systemic blood or 5 ce. of portal or mesenteric blood are drawn in the usual manner (described above) by means of a pipette and transferred directly to one of the large Jena test tubes so extensively used in this work. To it are added 2 to 3 ce. of the oxalate-carbonate solution described on p. 524 (15 per cent potassium oxalate and 10 per cent sodium carbonate) and about 5 cc. of toluol. The air current is then started and is run as fast as the apparatus can stand for 20 to 30 minutes. The liberated ammonia is collected, as previously described, in another large test tube charged with 5 to 6 drops of tenth-normal acid and 1 ce. of water. On account of the strong air current available in this laboratory, and also because of the relatively long period during which the process is carried out, we have found it desirable to cover the top of the test tube receiver with a small funnel from which the stem has been removed, thus obviating any loss which might be caused by spattering. At the end of the time indicated the contents of the receiver is Nesslerized in the usual manner but more cau- tiously, adding in all not over 1 cc. of the previously diluted re- agent (dilution 1:5). The solution is then carefully transferred to a 10 ec. volumetric flask, diluted to the 10 ec. mark, mixed, and with this solution the 100 mm. polariscope tube is filled and closed as for ordinary polariscope work. Two standard solutions, one containing 0.5 mgm. the other 1 mgm. of nitrogen, are Nesslerized simultaneously with the unknown solution made up to. volume (100 ec.) and one or the other is used as a standard. In this case, of course, the unknown remains stationary and the standard solution must be adjusted until the two colors match. In making this comparison it is necessary to keep moving both the diaphragm and the colorimeter prism in the standard solution until the right position of each is secured. The colorimeter, as thus used, represents, we believe, a new departure in colorimetry and we are taking steps to secure the 536 New Methods for Blood Analysis making of such instruments. So far we have used an ordinary diaphragm taken from a microscope and have fastened it by means of two screw clamps on top of the colorimeter platform on which stands the cylinder. A new zero point has of course to be estab- lished to allow for the altered position of the cylinder. Now we are compelled to use one instrument exclusively for such ammonia determinations but we hope later to see such instruments properly made by some manufacturer. In view of the fact that we have already published? a number of ammonia determinations, made as described above, it seems unnecessary to insert more figures here. We do not assert that even those figures may not ultimately be found to be too high but we do believe that they represent the nearest approach to the true values that have yet been published.— yee We believe that the methods described in ctl paper oa be found more serviceable than any hitherto available for the study of many important problems which can be solved only on the basis of blood and tissue analysis. We have so far published two papers (loc. cit.) and shall soon publish another more extensive one on the fate of the amino-acids absorbed from the digestive tract (and the gradual formation of urea). We hereby expressly revoke our earlier reservation (loc. cit.) of the field of research referred to in those papers by means of these methods. We would like to reserve for a while, however, the use of the methods for clinical investigations. We wish particularly toinvestigate nephritiec cases and fevers, and for this purpose are now gathering data as to the variations in the composition of normal blood. The retention of 3 to 4 grams of non-protein nitrogen in a person of average size should be easily demonstrable by means of these methods unless the normal variations are greater than we have yet found them. 3 This Journal, xi, p. 161, 1912. ON UROCANIC ACID. By ANDREW HUNTER. (From the Physiological Laboratories of the Universities of Edinburgh and Leeds, and the Department of Physiology and Biochemistry, Cornell Uni- versity, Ithaca, N. Y.)} (Received for publication, April 16, 1912.) In 1908 the writer isolated from a long-continued pancreatic digest of casein a crystalline substance, which he was able shortly afterwards to identify as urocanic acid (“‘ Urocaninsaéure’’). Up to that time the substance had been known only as an occasional - constituent of the urine of dogs. Two cases of its occurrence had been reported; the first in 1874 by its discoverer Jaffé,? and the second in 1898 by Siegfried. It would seem that the two animals concerned presented a rare anomaly of metabolism, not attrib- utable to any definite cause. Jaffé examined the urine of other dogs, and also of men, without again encountering the condition. Siegfried could find no urocanic acid in 110 liters of human urine. The origin and constitution of urocanic acid have till now re- mained uncertain. Siegfried conjectured a relation to the purines. The appearance of the substance among the products of casein digestion pointed at once in another direction. In preliminary communications? I indicated the probability that the mother sub- stance of urocanic acid is histidine. Observations made soon afterwards suggested a more definite conclusion regarding its struc- ture; but the hypothesis formed could not at the time be deci- sively tested without a fresh supply of material. Efforts to procure this have consumed a great deal of time. During-the last three 1 The substance discussed in the present communication was isolated at Edinburgh, and identified as urocanic acid at Leeds. The remainder of the investigation was carried out at Cornell. 2 Jaffé: Ber. d. chem. Ges., vii, p. 1668, 1874; and vii, p. 811, 1875. 2 Siegfried: Zeitschr. f. physiol. Chem., xxiv, p. 399, 1898. 4 Hunter: Journ. of Physiol., xxxvii, Proc. Physiol. Soc., p. xxxvii, 1908; and this Journal, vi, Proc. Soc. Biol. Chem., p. xliii, 1908-9. 537 538 Urocanic Acid years large quantities of casein have been subjected for six or seven month periods to tryptic digestion, and in the product urocanic acid: bas been sought by the method which originally led to its iso- lation.® Disappointment has been the uniform result; the original experience has not once been duplicated. Attempts to discover a dog which excreted urocanic acid have met with no greater success. The statement of Swain® that a possibly related substance may occur in coyote urine led to an examination of that source also; but the animal investigated produced neither Swain’s substance nor urocanic acid.’ It is therefore fortunate that recent work by others has afforded the means of deciding immediately the question at issue, and of settling the problem of constitution. The circumstance re- moves any reason that may have hitherte-existed for withholding the details now communicated. Se eee The digest from which urocanic acid was obtained had been made with the object of preparing a supply of the ‘polypeptide of pancreatic digestion” described by Fischer and Abderhalden.*® In 5 liters of water, containing 10 ec. of concentrated Liquor ammoniae, there were dissolved 500 grams of ‘‘ Plasmon” and 10 grams of ‘‘pankreatin absolut.,: Rhenania.’”’ The mixture was maintained at 40° in the presence of abundant toluene and chloro- form. Two days later 10 ec. of ammonia, and sixteen days later 10 grams of pancreatin were added. Digestion was continued for seven months, at the end of which time the biuret reaction was but feebly positive. A loose jelly-like clot (plastein?), impregnated with tyrosine crystals, was filtered off, and the filtrate was concen- trated in vacuo at 40° to 50°. The second crop of tyrosine crystals having been removed, the liquid was diluted to about 6 liters, and treated with phosphotungstic acid. The washed precipitate was decomposed in the cold with baryta, and excess of the latter removed by sulphuric acid. The product was concentrated in vacuo and dried over sulphuric acid. The yield of crude “poly- peptide” was 56 grams. This was dissolved in 1 liter of 5 per cent 5 In the somewhat laborious operations involved I had during the Summer Session of 1911 the assistance of Miss Ruth Wheeler, to whom I here See fully acknowledge my indebtedness. 6 Swain: Amer. Journ. of Physiol., xiii, p. 30, 1905. 7 Hunter and Givens: This Journal, viii, p. 449, 1910. 8 Fischer and Abderhalden: Ze7tschr. f. physiol. Chem., xxxix. p. 81. 1903. Andrew Hunter 539 sulphuric acid, and the precipitation with phosphotungstic acid was repeated. From the precipitate were finally obtained 42 grams of brownish-yellow, extremely hygroscopic material, which dis- solved in water with strongly alkaline reaction, and gave no biuret reaction whatever. From this product it was decided to separate, if possible, arginine and histidine. One would naturally have expected the material to contain these bases in considerable amounts. It did give the intense ‘‘diazo reaction”? shown by histidine. But it was found that towards silver nitrate and fixed alkali it did not react in the way expected of an arginine solution. As a matter of fact subse- quent application of the silver-baryta method showed that neither arginine, nor yet histidine, was present in quantities that could be isolated. On the other hand there was produced by silver nitrate alone a quite considerable precipitate, which dissolved at once in the slightest excess of either acid or alkali. Attention was there-. upon directed to the separation of the substance so reacting. To this end the whole material was brought into aqueous solution (600 to 700 cc.), very nearly neutralized with nitric acid, and treated with 10 per cent silver nitrate as long as a precipitate resulted. Six or seven grams of the nitrate were required. The light brown gelatinous precipitate was collected on a filter, and thoroughly washed. It was then suspended in water, and dissolved by the aid of a little dilute sulphuric acid. The solution was freed from silver by hydrogen sulphide, from the latter by a stream of air, and from sulphuric acid by baryta. It reacted now acid, and on con- centration deposited 1.45 grams of crude crystalline material. This was purified by boiling with charcoal and by several crystalliza- tions. The final yield was 0.92 gram. The amount originally present in the digest must have been considerably greater. The substance thus obtained was sparingly soluble in cold, readily soluble in hot, water. Its solubility in alcohol was very slight, while in ether, acetone, ethyl acetate and carbon disulphide it was almost, if not quite, insoluble. It was dissolved with ease by glacial acetic acid, and by aqueous ammonia or sodium hydrox- ide. Its aqueous solution reacted acid to litmus. When rapidly crystallized from hot water, it formed branching groups of slender, beautifully iridescent, doubly refracting needles, sometimes nearly a centimeter long; on more gradual separation it appeared as well formed tetragonal prisms of the first and second orders. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL, XI, NO. 5. 540 Urocanic Acid The water-free substance melted with decomposition at 224° (corrected). ANALYSIS AND MouecuLar WEIGHT DETERMINATION. 0.1369 gram _air- dried substance lost 0.0281 gram at 110°. ; 0.2208 gram, dried at 110°, yielded 0.4182 gram CO, and 0.0879 gram H,O. 0.1193 gram gave 21.1 cc. N at 17° and 746 mm. 0.1081 gram, dissolved in 10.48 grams glacial acetic acid, depressed the freezing point of the solvent 0.324°. Calculated for CesHeO2N2.2H20: Found: 1S XO) <5 Ree gee 20.7 20.5 RO rae. FCR es Se, SA 52.1 GS? 1812 Gi: Bee SNE eS ote 6 0 4.4 4.5 PeeMee ei. yn) LR 20.3 20.5 Molecular weight........ 138.0 124.0 In crystalline form, solubility, melting point, and elementary composition the substance agreed exactly with the descriptions of urocanic acid. The two following reactions removed any un- certainty as to its identity therewith. (1) If a small quantity of the hydrated substance is treated with a drop of glacial acetic acid the crystals at first dissolve; but almost immediately thereafter, especially if the solution is shaken or rubbed, they separate again as a thick mass of small, opaque, white needles. This behavior is described by Siegfried as characteristic of urocanic acid. The opaque needles dissolve readily on addition of water or alcohol, or an excess of acetic acid. (2) When an aqueous solution of the substance is treated with an equal volume of 50 per cent nitric acid, a heavy microcrystalline precipitate of the nitrate is very rapidly deposited. The crystals of urocanic acid nitrate, produced in a similar manner, have a highly characteristic appearance. They are described by Jaffé as “small plates, bent in the form of a sickle, with the ends apparently frayed or eaten away; frequently several such plates are united to cross- or rosette-shaped aggre- gates.’’ A reference to the photographs reproduced? will demon- strate the quotation to be an exact description of the crystals yielded by the substance from casein. Concerning the identity of the latter there was therefore no rea- sonable room for doubt. But the molecular weight determina- tion above reported led to the formula CsHgO2Ne. This was in 9 The negatives were very kindly made for me by Dr. R. Cattley, Univer- sity of Leeds, whom I take this occasion of thanking for the service. Andrew Hunter 541 disagreement with the view of Jaffé (concurred in by Siegfried), which assigned to the acid the double formula Cy2.H)204Ns.'° Jaffé’s choice was determined by the single circumstance, that uro- eanic acid, when heated, yields by loss of carbon dioxide and water a base—“urocanine’’—to which apparently must be ascribed the formula C,,HipON,;. The evidence for the chemical individuality of this substance is not altogether convincing. Neither the base itself nor its salts with mineral acids could be obtained in crys- Various Forms oF UrocaNic Actp NITRATE. talline form. Its formula rests entirely upon analyses of an exceedingly hygroscopic chloroplatinate. The nature of urocanine 10 Jt was at first suspected that the determination itself might be in error. It was made with an apparatus, the only one then at my disposal, which did not exclude moisture. Experiment proved that it was nevertheless easy to obtain with the same instrument and the same sample of acetic acid satis- factory approximations to the calculated molecular weight of other organic substances. 542 Urocanic Acid decidedly calls for further investigation. Whatever may be the mechanism of its formation, the sequel will show with sufficient clearness that urocanic acid at any rate does possess the simpler formula indicated by its cryoscopic effect. To the older descriptions of urocanic acid I am able to add the following points. Its solutions are optically inactive. They are precipitated by silver nitrate; the precipitate increases in bulk upon exact neutralization with ammonia, but dissolves instantly in the slightest excess of either ammonia or nitric acid. Urocanic acid is precipitated also by mercuric chloride, and by phosphotungstic and picrolonic acids. The phosphotungstate dissolves in hot water, from which it crystallizes in minute cubes or short rectangular prisms. The picrolonate, gelatinous when formed by bringing together aqueous solutions but granular if precipitated in alcohol, is dissolved with great difficulty in boiling absolute alcohol, with less difficulty in boiling water, and with comparative ease in boiling dilute alcohol. {[t can be recrystallized from water as bright yel- low sheaves of long filamentous needles, from 75 per cent alcohol as dense clumps of yellow plates, which singly take the form of elongated rhombs. It decomposes about 268° (corrected), after gradual discoloration from about 230°. On bringing together saturated aqueous solutions of urocanie acid and picric acid there is no immediate precipitate; but there gracually separate yellow iridescent macroscopic prisms of the picrate, which melt at 224° to 225° (corrected). Urocanic acid in dilute sodium carbonate solution gives a very intense red reaction with diazobenzenesulphonic acid. It does not evolve any nitrogen on treatment with nitrous acid. It instantly reduces a cold alkaline permanganate solution with immediate liberation of manganese dioxide. A substance giving the diazo reaction, and obtained from pro- tein in the way described, could hardly be other than an imidazole derivative, standing in some relation to histidine. Its precipita- tion reactions were in harmony with this conclusion. The immedi- ate reduction of cold alkaline permanganate, which the imidazole ring itself will not bring about, pointed to the possession of an unsaturated side chain. These considerations, taken im con- junction with the empirical formula CyHsQ.Ne, suggested the probability that urocanic acid is an imidazole-acrylic acid, related Andrew Hunter 543 therefore to histidine in the same way as cinnamic acid is to phenylalanine." CH—N# SEES I ae pou ae pou om on tes . NH» ia ane ds Histidine Imidazole-acrylic or Urocanic acid Such a constitution would account for the acid reaction to indi- cators with simultaneous possession of basic characters, the want of optical activity, and the failure to react with nitrous acid. But before it could be held to be fully established, confirmatory evi- dence of a stricter nature was essential. This can now be sup- plied. A substance known independently to have the structure represented above has been recently described by Barger and Ewins.” They obtained it in two ways: (1) from ergothioneine (the betaine of thiohistidine) in the manner illustrated by the scheme below: - CHiN CH—NH Ch— Ni |! Sos | Ye.su | Scu cn 7 C——N 7 NTS. | mori; ',: | HNO; | CHe ———— CH —— nag et | | CH—N (CH3)3 : CH CH | co-—0O COOH COOH 1! This view of the constitution of urocanic acid occurred independently, as I learn from a private communication, to Professor Treat B. Johnson of Yale University, whose studies on thiohydantoins have led him to an interest in the derivatives of histidine (see Johnson and Guest: Amer. Chem. Journ., xlvii, p. 242, 1912). It was Professor Johnson who drew my attention to the ’ paper of Barger and Ewins mentioned below, and he therefore who furnished the stimulus that occasioned the immediate publication of my results. It is a pleasure to record my appreciation of the friendly spirit that has charac- terized Professor Johnson’s side of the correspondence. 2 Barger and Ewins: Journ. Chem. Soc., xeix, p. 2336, 1911. 544 Urocanic Acid (2) by the action of trimethylamine on a-chloro-6-imidazole-pro- pionic acid. Of this product sufficient has been placed in my pos- session® to enable me to say with certainty that it is identical with urocanic acid. It crystallizes in precisely the same forms; it behaves in the same way with glacial acetic acid; and its nitrate has the peculiar and characteristic shape of urocanic acid nitrate. Specimens of the two products heated side by side melted together at 231° to 232° (corrected) ; a mixture of both in equal proportions melted simultaneously with a sample of the compound from casein. Barger and Ewins describe a phosphotungstate crystallizing in smallrectangular plates, and a picrate forming golden yellow prisms. The comparison places it beyond reasonable doubt that urocanic acid is B-imidazole-4(or 5)-acrylic acid. That such a substance should make its appearance in a pancreatic digest is somewhat astonishing. Its origin certainly cannot be ascribed to the action of trypsin. What the responsible factor was, whether the particular ferment preparation employed contained a deaminizing enzyme of peculiar nature, or whether the responsibil- ity lay with some accidental circumstance in the manipulation of the product; it has not been possible to determine. It is doubtless of significance that from the digestion mixture arginine, as well as histidine, had disappeared. The attempt to duplicate the occur- rence has not been abandoned, and an explanation may yet be found. One naturally thinks of bacterial action. The incubated mixture at no time exhibited evidence of organismal growth; yet in the absence of bacteriological control that source of decomposi- tion cannot be by any means excluded. With this in mind I have grown some of the commoner organisms in casein and histidine solutions. The result hoped for has not so far been attained. Experiments in this direction also are being continued, although, so far as Iam aware, no analogous case of the conversion by bac- teria of an amino- into an unsaturated acid has been reported. 13 To Drs. George Barger and Arthur J. Ewins I take this opportunity of expressing my grateful recognition of the courtesy with which they at once acceded to my request for a specimen. 14] had previously found for urocanic acid the melting point 224°, while Barger and Ewins report for their substance 235° to 236°. The fact is, as Siegfried also noticed, that the value found varies widely with the manner of heating. This is probably equally true for the picrate, which according to Barger and Ewins melts at 213° to 214°, according to my own determina- tion at 224° to 225°. Andrew Hunter 545 The appearance of imidazole-acrylic acid under the circumstances described in this paper is not more remarkable than its occasional occurrence as an excretory product in the dog. In this character it almost certainly represents an intermediate step in the catabo- lism of histidine. The type of amino-acid transformation which would thus be illustrated apparently occurs in plants—witness the formation of cinnamic and p-cumaric acids—but has not hitherto been met in animals. The production of cinnamoyl-glycocoll observed by Dakin™ to follow administration of phenylpropionic acid to cats presents perhaps the nearest analogy. Other origins than the one assumed are of course not impossible. But the formation of the unsaturated acid is not the only problem offered. The phenomenon of its excretion is equally puzzling. It is known that moderate doses of cinnamic acid are easily and completely oxidized within the animal organism.!® A case where the analo- gous imidazole derivative cannot be similarly disposed of is almost certainly a metabolic anomaly. The elucidation of the structure of urocanic acid adds therefore a fresh interest to the search, still being prosecuted, for an animal which regularly excretes that substance. It would be of interest to determine whether even normal dogs do not excrete small quantities of urocanic acid in response to enteral or parenteral administration of histidine or its derivatives. Experiments to decide the point are in contempla- tion. The metabolic fate of histidine has been the subject of studies by Abderhalden and Einbeck,!? Abderhalden, Einbeck and Schmid,!* Kowalevsky,!® and Dakin.”° But in none of the experi- ments reported was urocanic acid specifically sought. 15 Dakin: This Journal, v, pp. 173 and 303, 1908; also vi, p. 203, 1909. 16 Cohn: Zeitschr. f. physiol. Chem., xvii, p. 274, 1893; and Dakin: this Journal, v, p. 413, 1909. 17 Abderhalden and Einbeck: Zeitschr. f. physiol. Chem., \xti, p. 322, 1909. 18 Abderhalden, Einbeck, and Schmid: Jbid., lxviii, p. 395, 1910. 19 Kowalevsky: Biochem. Zeitschr., xxiii, p. 1, 1910. 20 Dakin: This Journal, x, p. 499, 1912. ; 1 tse | ' ie ine 7 MRT stirygecs Pasa «" ON SPHINGOSINE.! By P. A. LEVENE anp W. A. JACOBS. (From the Laboratories of the Rockefeller Institute for Medical Research’ New York.) (Received for publication, May 2, 1912.) Sphingosine was discovered by Thudichum? on hydrolysis of a cerebroside, phrenosine. Discussing the chemical properties of the substance, its behavior towards bases and acids, the author took into consideration the possibility of the substance having the structure of an amino-acid or of an alkaloidal base. In his final conclusion he expressed preference to the view of the basic nature of the substance. In later years Thierfelder* repeated the work of Thudichum, and in the main substantiated his views. The work of Thierfelder, however, was directed principally to the study of the properties of 1A report on the results of the present investigation has appeared in the Proceedings of the Meeting of the American Society of Biological Chemists, held December 28th to December 30th, 1911, published in the March number of this Journal. In Heft 6, vol. lxxvii of Hoppe-Seyler’s Zeit- schrift fiir physiologische Chemie, published on April 9, there appeared two articles by Thierfelder, Riesser and Thomas in which the authors arrived at the same conclusions as reported by us. The appearance of the two articles was caused undoubtedly by the publication of our report, since Heft 5 of vol. Ixxvii of Hoppe-Seyler’s Zeitschrift contained no mention of Thierfelder’s name among the authors of twenty-six articles received for publication. In a footnote to one of the articles Professor Thierfelder claims the sole privilege for work on sphingosine and related substances, for the reason that Thudichum’s work had been half forgotten at the time when Thierfelder directed his attention to cerebrosides. We do not feel that this justifies the request made by Thierfelder, that the work which had been in progress in our laboratory for more than a year should be abandoned before it is completed. 2 Die chemische Konstitution des Gehirns, Tiibingen, 1901. 3 Zeitschr. f. physiol. Chem., xliv, p. 366, 1905; Kitagawa: Jbid., xlix, p. 286, 1906. 547 548 On Sphingosine the cerebroside, which he named ‘‘cerebron.”” The author ex- pressed no definite view regarding the chemical structure of the base. The results of the present investigation have made certain that sphingosine is an unsaturated monoaminodihydroxyalcohol. This conclusion is based on the following data: 1. The substance contains all! its nitrogen in form of primary amino nitrogen. 2. The presence of a double binding in the molecule is demon- strated by the readiness with which sphingosine absorbs hydrogen when treated according to the method of Paal. A substance is thus formed which has the composition of dihydrosphingosine. It was analyzed in the form of a sulphate and a triacetylderivative. 3. The presence of two hydroxyl groups in the molecule is evi- dent from the fact that sphingosine forms a triacetylderivative which no longer contains the original primary amino group. The substance forms a dimethylether, and finally it can be reduced to an amine, sphingamine. As yet it is not certain whether or not the carbon atoms are linked in a normal chain. Attempts were made to reduce the dihydrosphingosine to the corresponding amine, but instead of the heptadecylamine there was always obtained the unsaturated sphingamine. Efforts to obtain the saturated amine are now in progress. Also work is in progress on the respective position of the double bond, and of the hydroxyl] groups. EXPERIMENTAL PART. Sphingosine. ‘The base was obtained on hydrolysis of ‘‘cerebrin” prepared by a slight modification of the process described by Parcus.4 The conditions of hydrolysis were similar, but not identical with those described by Thierfelder. The base was prepared in crude form as the sulphate, which was then transformed into the free base and into the acetate. The discoverer of the base mentioned that it could be made to crystallize out of ether. It was found in course 4 Journ. f. prakt. Chem., xxiv, p. 310, 1881 P. A. Levene and W. A. Jacobs 549 of this work that crystallization proceeded much more readily out of petrolic ether. The sulphate was obtained in the form of a white crystalline powder. It melted with decomposition at 233° to 234°C. (uncor- rected). A great many samples were analyzed. The analysis of one of these gave the following results: 0.0996 gram of the substance dried in chloroform-vacuum bath over phos- phorus pentoxide gave on combustion 0.0940 gram of H.O and 0.2280 gram of COs. 0.2400 gram of the substance, employed for a Kjeldahl nitrogen estima- tion, required for neutralization 6.9 cc. of 74 acid. Calculated for (CizHasN O2)2H2SQ.: Found: Oe os a gene 58 rr 61.08 61.05 H oy nt id ee, bongs Ole eee 10.78 10.60 Ifo. ade s ol: De eee 4.19 4.06 The optical activity of the substance was the following: 0.5304 gram of the sulphate was dissolved in a mixture of 5 cc.of chloro- form and 1 ce. of glacial acetic acid. The total weight of solution was 8.7514 grams. The rotation in pure D-light was —1.50°, hence [a]? = — 13.12° (+0.00). Diacetate. Dissolved in glacial acetic acid and petrolic ether, the substance crystallized in form of very long needles of the following composition. 0.1402 gram of the substance gave on combustion 0.1290 gram H.O and 0.3206 gram of COs. Calculated for C17HasNOxz. (C2Hs02)2: Found: (OR Sehr ok Oe 5 a 62.22 62.36 LE 2 ome beasts ee 10.61 10.22 Amino nitrogen estimation. A solution of 0.300 gram of the sul- phate in 10.0 cc. of glacial acetic acid was employed for an amino nitrogen estimation according to the method of Van Slyke. Five cc. were used for each experiment. All nitrogen was given off in thirty minutes. In each experiment 11 cc. of nitrogen were formed at ¢ = 21°C. and p = 760 mm. Calculated for (CizH2302.N H2)2H2S0« Found: ES ee ee 4.19 4.17 550 On Sphingosine Dihydrosphingosine. The hydrogen absorption value of sphingosine is obtained most conveniently when the free base is dissolved in ether and shaken with aqueous colloidal palladium prepared according to Paal. 0.100 gram of palladium to about 0.500 gram of the base dissolved in about 150 ce. of absolutely pure ether gave the most satisfac- — tory results. The absorption was completed in about thirty min- utes. The velocity of the operation was greatly increased by the addition of 1 cc. of glacial acetic acid to the ethereal solution. 0.500 gram of the substance absorbed 50 cc. of hydrogen (without correc- tion fort and p). Theory requires 45 cc. of H. 0.6486 gram of the substance absorbed 59 cc. of H; theory requires 56 ce. The ethereal solution of dihydrosphingosine was evaporated to dryness and the substance converted into the sulphate and into the triacetylderivative. The sulphate was obtained in form of a white crystalline powder. Its melting point was only slightly different from the unsaturated compound, being 235°C. Calcylated for (C17H37 NO2)2H2SOs: Found: Cre oe i eee 60.61 60.90 1 8 (by pele RRR ate 9 ets ok oA ia eS zc 11.38 Lei The optical activity of the substance was difficult to determine for the lack of a sufficiently satisfactory solvent. Approximately it was as follows: 0.0776 gram of the substance dissolved in about 3 cc. of aleohol containing sulphuric acid. and weighing 2.8640 grams gave a rotation of —0.29° ina 2 dm. tube. [a]? = — 10.67°. 0.1214 gram of the substance gave on combustion 0.1153 gram of H.O and 0.2948 gram of COs. Calculated for CirHai NO2z. (CH3CO)s: Found: CA a co. ee 66.76 66.76 1 In ee od Be ee AeA ed 10.50 10.60 P. A. Levene and W. A. Jacobs 551 Acetylderivatives. On treatment of the free base with acetic anhydride di- or tri- acetylsphingosine can be obtained. The first is obtained by dis- solving the base in boiling acetic anhydride and evaporating the solution under diminished pressure. The triacetylderivative is prepared by allowing the base to digest with acetic anhydride in a boiling water bath with return condenser for one hour and only then evaporating the solution to dryness. The further treatment in both instances is identical. The residue obtained on evaporat- ing the solution under diminished pressure is taken up in chloro- form and again evaporated under diminished pressure. This residue is taken up in hot acetone and the substance allowed to crystallize. For analysis the substances were dried in a vacuum- chloroform bath over phosphorus pentoxide. Diacetylderivative. 0.1317 gram of the substance gave on combustion 0.1256 gram of H2O and 0.3283 gram of COs. 0.2720 gr-m of the substance was dissolved in 10 ce. of glacial acetic acid and used for amino nitrogen estimation according to Van Slyke. Five cubic centimeters of the solution were employed for each experiment. There was formed 9.3 cc. nitrogen at 24°C. and 758 mm. pressure. The substance was allowed to react one hour, although the reaction was practically com- pleted in twenty minutes. Calculated for (Ci;H3iO2. N H2).(CHsCO)e: Found: No, Sg Be Obs oor ae ee 68.1 68.16 18. OR Bn besiege Dea 10.8 10.59 IS (oto bid hamlets nae ee 3.4 3.78 The physical constants and saponification value of this substance were not determined. Triacetylderivative. 0.1194 gram of the substance gave on combustion 0.1064 gram of H.O and 0.2950 gram of CO. 0.2500 gram of the substance was dissolved in 10 ce. of glacial acetic acid and employed for anamino nitrogen estimation. No formation of nitrogen took place. Calculated for Ci7H32NO2.(CH3CO)s: Found: (Gy ccc 2:6 coke ee O7e15 67.38 isl. (042 2G See eee 9.98 9.98 The substance melted sharply at 102° to 103° C. (uncorr.). 0.3383 gram of the substance was dissolved in 60 cc. of methy! alcohol, containing 10 cc. of a ¥ solution of sodium hydrate in methyl alcohol. The solution was heated on boiling water bath for two hours, allowed to stand over night and titrated. It required 24.90 cc. of 7) alkali to neutralize the acetic acid formed on saponification. The theory required 24.75 cc. 552 On Sphingosine Dimethylsphingosine. This substance is formed in course of hydrolysis of cerebrosides by means of methylalcohol and mineral acid. Thierfelder, who was the first to have the substance in his hands. erroneously re- garded it as a.new base. Since our first communication, Thier- felder and Riesser® substantiated our view on the substance. The methylderivative was obtained in form of a sulphate on concen- trating the mother liquors from the crude sphingosine sulphate. The sulphate was then transformed into the free base and this again transformed into the hydrochloride. The hydrochloride crystallizes out of alcohol in- the form of large glittering plates. The substance was identified by the fact that, similarly to sphin- gosine, it contained all its nitrogen in form of primary amino nitro- gen; it contained one unsaturated bond and on boiling with hydro- iodic acid formed the required amount of methyliodide. 0.1615 gram of the substance gave on combustion 0.1654 gram HiO and 0.3926 gram of COs. Calculated for CisH2sNO2. HCl: Found: CORR HS: Te ee 65.18 66.65 Eee FeO, 25 2 Ae POE RS ae 11.5 11.24 Hydrogen absorption value. One gram of the substance dissolved in ether containing 2 cc. of glacial acetic acid. On treatment with palladium accord- ing to Paal it absorbed 67 cc. of hydrogen. Theory requires 72.6 cc. Amino nitrogen estimation. 0.3500 gram of the hydrochloride dissolved in 10 ce. of glacial acetic acid. Five ce. of this solution used for amino nitro- gen estimation according to Van Slyke. There formed 12.4 cc. of nitrogen at t = 24° and p = 758mm. Calculated for Ci9H3902.N H2HCl: Found: FR I oes ak 4.01 4.00 Methyl estimation. 0.1206 gram of the hydrochloride boiled with hydro- iodic acid of specific gravity = 1.71 in the apparatus of Zeisel and Fanto. There was obtained 0.1286 gram of silver iodide. Calculated for C1tzHaz NO2.(CHs)2.HCl: Found: LONG bh au doch cc vce ON 8.45 7.00 The physical properties of the substance did not permit a more accurate estimation. In a control experiment with sphingosine ® Zeitschr. f. physiol. Chem., \xxvii, p. 508, 1912. P. A. Levene and W. A. Jacobs 553 crystallized out of petrolic ether no silver iodide was formed. On the other hand, the base obtained directly after removing the sul- phuric acid from apparently pure sphingosine sulphate still caused the formation of some silver iodide. The highest value obtained in this manner was equivalent to CH; = 2.32 per cent. Sphingamine. Attempts were made to reduce dihydrosphingosine to the corre- sponding amine. The normal heptyldecylamine has been obtained synthetically. Hence a comparison of the two bodies should have determined the fact whether or not the substances were identical. In several experiments the reduction was attempted by means of hydroiodic acid, and in one experiment the dihydrosphingosine was transformed into the dihydrodichlorsphingosine, which was then reduced by means of metallic sodium and alcohol. However, under all conditions the unsaturated substance was formed. Reduction with hydroiodic acid was carried out in sealed tubes at 125°C. The reaction-product was dissolved in ether. The ethereal solution was dried with anhydrous sodium sulphate, and then diluted with one-third of its volume of 98 per cent alcohol, and the solution treated with metallic sodium. The substance obtained from the solution was transformed into the sulphate. The analysis of the substance obtained from three different experiments follows: I. 0.1042 gram of the substance gave 0.1100 gram H.O and 0.2562 gram of COs.. II. 0.1014 gram of the substance gave 0.0998 gram H,O and 0.2494 gram of CO:. ILI. 0.1153 gram of the substance gave 0.1233 gram H.O and 0.2800 gram of CO». 0.1480 gram of sample I was used for Kjeldahl nitrogen estimation. It required for neutralization 4.1 ce. of 75 acid. ’ Caleulated for (CizHasN )2H2SOu: Found: { IL Ill (Ci op 52 oe 67 46 67 67.2 67.85 2 82 11.05 11.97 44 554 On Sphingosine Reduction of dihydrodichlorsphingosine. The chlorderivative was obtained by digesting dihydrosphingosine with thionylchloride in a water bath at 50°. The crude substance without purification was dissolved in a mixture consisting of two parts of ether and one of 98 per cent alcohol and reduced with metallic sodium. The sub- stance obtained in this manner was transformed into the sulphate and analyzed. 0.1251 gram of the substance gave on combustion 0.1310 gram of H,O and 0.3097 gram of COs. Calculated for (Ci7HasN )2H2S0Os: Found: Cs oe ER a ee ees 67 .46 67.51 ee. ee ee nee 12.02 11.72 ERRATUM. On page 217 of this volume, No. 3, tenth line from the top, for strict read generic. INDEX TO VOLUME XI. Absorption, of copper by Fundulus heteroclitus, 381; of fat, 429; of fumes, apparatus for, 503; of metallic salts by fish, 381. Acid-forming elements in foods, 323. Alanine, action of potassium thiocy- anate upon, 97. Alcohol, recovery of from animal tissues, 61. Aluminum, determination of in feces, 387. Ammonia, determination of in blood, 527; determination of in urine, 523; in portal blood, origin and significance of, 161; —— me- tabolism, relation of balance of acid-forming and base-forming elements of foods to, 323. Analysis, of ash of smooth muscle, 401; of blood and tissues, pro- tein metabolism from the stand- point of, 87, 161; of urine, phos- photungstic acid as a clarify- ing agent in, 81. ANDERSON, R. J.: Phytin and phos- phoric acid esters of inosite, 471. Animal tissues, recovery of alcohol from, 61. Antagonism between salts and sugar, 415. Apparatus for the absorption of fumes, 503. Autolysis, influence of upon chol- esterol, 37. Bacteria, effect of lecithin upon fer- mentation of sugar by, 313. Balance of acid-forming and base- forming elements in foods, 323. Barbituric acid series, physiological action of some pyrimidine com- pounds of, 443. Base-forming elements in foods, 323. Bean (Phaseolus), haemagglutinat- ing and precipitating properties of, 47. BENNETT, C. B.: The purines of mus- cle, 221. Benzoic acid in urine, quantitative determinations of, 201. Buack, CuaRENCE L.: see Under- hill and Black, 235. Blood, and tissue analysis, protein metabolism from the stand- point of, 87, 161; new methods for determination of total non- protein nitrogen, urea and am- monia in, 527; portal ——,, ori- gin and significance of ammonia in, 161; sera, mammalian, isolation of odcytase from, 339; serum, optical method for determining the concentrations of various proteins in, 179. Buoor, W. R.: Carbohydrate esters of the higher fatty acids. II. Mannite esters of stearic acid, 141; Carbohydrate esters of the higher fatty acids. III. Man- nite esters of lauric acid, 421; On fat absorption, 429. Carbohydrate esters of higher fatty acids, 141, 421. Casein, hydrolysis of by trypsin, 267. Chemical analysis of ash of smooth muscle, 401. Chemistry of dog’s spleen, 27. 555 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XI, NO. 5. 556 Children, creatine in the urine of, 253. Cholesterol, influence of autolysis upon, 37; quantitative determi- nation of, 37. Cocaine, influence of upon metab- olism, 235. Couuison, R. C.: A brief investiga- tion on the estimation of leci- thin, 217. Concentrations of various proteins in blood serum, optical method for determining, 179. Copper, absorption of by Fundulus heteroclitus, 381. Corprer, Harry J.: Chemistry of dog’s spleen, 27; Errors in the quantitative determination of cholesterol by Ritter’s method: the influence of autolysis upon cholesterol, 37. Cotton seed, utilization of proteins Om iE Creatine in the urine of children, 253. Denis, W.: see Folin and Denis, 87, 161, 253, 503, 527. Determination, of aluminum in feces, 387; of ammonia in urine, 523; of benzoic, hippuric and phenaceturic acids in urine, 201; of cholesterol, 37; of con- centrations of various proteins in blood serum, optical method for, 179; of hippuric acid in urine, 257; of total nitrogen in urine, 493; of total non-protein nitrogen, urea and ammonia in blood, new methods for, 527; of urea in urine, 507. 2,8-Dioxy-l-methylpurine, 393. 2,8-Dioxy-6,9-dimethylpurine, 393. Disaccharides, combined action of muscle plasma and pancreas ex- tract on, 347. Dog’s spleen. chemistry of, 27. Index DunHAM, Epwarp K.: see Mandel and Dunham, 85. Echinochrome, a red substance in sea urchins, 435. Elimination of lactic acid during cocaine poisoning, 235. Ewuiotr, J. H. and H. S. Raper: ‘ Note on a case of pentosuria presenting unusual features, 211. Epstein, ALBERT A. and H. Osan: Studies on the effect of lecithin upon fermentation of sugar by bacteria, 313. Esters, carbohydrate, of higher fatty acids, 141, 421; mannite, of lauric acid, 421; mannite, of stearic acid, 141; phosphoric acid, of inosite, 471. Estimation of lecithin, 217. Extractive-free meat powder, util- ization of proteins of, 5. FARMER, CHESTER J.: see Folin and «Farmer, 493. Fast of one hundred and seventeen days, nitrogen distribution dur- ing, 103. Fasting, putrefaction processes in the intestine of a man during, 169; studies, 103, 129, 169. Fat absorption, 429. Fatty acids, carbohydrate esters of, 141, 421. Fecal nitrogen, origin of, 5. Feces, determination of aluminum in, 387; hydrogen ion concentra- tion of, 129. FENGER, FREDERIC: On the presence of active principles in the thy- roid and _ suprarenal glands before and after birth, 489. Fermentation of sugar by bacteria, effect of lecithin upon, 313. Fine, Morris 8.: see Mendel and Fine, 1, 5. Index FLANDERS, FRED F.: see Folin and Flanders, 257. Fotin, Orro: On the determination of urea in urine, 507; —— and W. Denis: An apparatus for the absorption of fumes, 503; New methods for the determina- tion of total non-protein nitro- gen, urea and ammonia in blood, 527; On creatine in the urine of children, 253; Protein met- abolism from the standpoint of blood and tissue analysis (first paper), 87; Protein metabolism from the standpoint of blood and tissue analysis (second paper). The origin and sig- nificance of the ammonia in the portal blood, 161; ——— and CuesteR J. Farmer: A new method for the determination of total nitrogen in urine, 493; —— and Frep F. FuanpErs: A new method for the determina- tion of hippuric acid in urine, 257; —— and A. B. Macauuum: On the blue color reaction of phosphotungstic acid (?) with uric acid and other substances, 265; On the determination of ammonia in urine, 523. Foods, balance of acid-forming and base-forming elements in, 323. Fumes, apparatus for the absorption of, 503. Fundulus heteroclitus, absorption of copper by, 381. Fundulus, toxicity of sugar solu- tions upon, 415. GETTLER, A. O.: see Sherman and Gettler, 323. Glands, thyroid and suprarenal, presence of active principles in before and after birth, 489. Glucose, action of leucocytes on, 361; action of tissues and tissue juices on, 353. 557 Haemagglutinating properties of the bean (Phaseolus), 47. _- Hanzuik, Pau J.: On the recovery of alcohol from animal tissues, 61. Hawk, P. B.: see Howe and Hawk, 129; Howe, Mattill and Hawk, 103; Sherwin and Hawk, 169. Hippuric acid in urine, new method for the determination of, 257; quantitative determinations of, 201. HoaGuanp, D. R.: see Schmidt and Hoagland, 387. Hows, Paut E. and P. B. Hawk: Studies in water drinking: XIII. (Fasting studies: VIII.) Hydro- gen ion concentration of feces, 129; ——, H. A. Martin. and P. B. Hawk: Fasting studies: VI. Distribution of nitrogen during a fast of one hundred and seventeen days, 103. Hunter, ANDREW: Onurocanic acid, bar: Hydantoins, 97. Hydrogen ion concentration of feces, 129. Hydrolysis of casein by trypsin, 267. Inosite, phosphoric acid esters of, 471. Jacoss, W. A.: see Levene and Jacobs, 547; Levene; Jacobs and Medigreceanu, 371. Jouns, Cart O.: Researches on pu- rines. On 2-oxy-l-methylpurine. 73; Researches on purines. On 2-oxypurine and 2-oxy-8-meth- ylpurine, 67; Researches on purines. On _ 2,8-dioxy-6,9-di- methylpurine and 2,8-dioxy-l- methylpurine, 393. Jounson, Treat B.: Hydantoins: the action of potassium thio- cyanate on alanine, 97. 558 KLEINER, IsRAEL S.: The physiol- ogical action of some pyrimidine compounds of the barbituric acid series, 443. Lactic acid, elimination of during cocaine poisoning, 235. Lauric acid, mannite esters of, 421. Lecithin, effect of upon fermenta- tion of sugar by bacteria, 313; estimation of, 217. Leucocytes, action of on glucose, 361. LEVENE, P. A. and W. A. JacosBs: On sphingosine, 547; ——, W.A. JacoBs and F. MEDIGRECEANU: On the action of tissue extracts containing nucleosidase on a- and £-methylpentosides, 371; and G. M. Meyer: On the action of various tissues and tissue juices on glucose, 353; on the combined action of mus- cle plasma and pancreas extract on some.mono-.and di-saccha- rides, 347; The action of leuco- cytes on glucose, 361. Logs, JacquEes: The toxicity of sugar solutions upon Fundulus and the apparent antagonism between salts and sugar, 415. Macauuium, A. B.: see Folin and Macallum, 265, 523. Mammalian blood sera, isolation of oé6cytase from, 339. Manpet, JoHN A. and EpWwarp K. Dunuam: Preliminary note on a purine-hexose compound, 85. Mannite esters, of lauric acid, 421; of stearic acid, 141. Martiuu, H. A.: see Howe, Mattill and Hawk, 103. May,.CLaRENCE E.: Concerning the use of phosphotungstic acid as a clarifying agent in urine analy- sis, 81. Index McCurenpon, J. F.: Echinochrome; a red substance in sea urchins 435. Meat powder, extractive-free, util- ization of proteins of, 5. MEDIGRECEANU, F.: see Levene, Jacobs and Medigreceanu, 371. Meies, Epwarp B. and L. A. Ryan: The chemical analysis of the ash of smooth muscle, 401. MENDEL, LAFAYETTE B. and Morris S. Frye: Studies in nutrition. V. The utilization of the pro- teins of cotton-seed, 1; Studies innutrition. VI. The utiliza- tion of the proteins of extrac- tive-free meat powder; and the origin of fecal nitrogen, 5. Metabolism, ammonia, relation of balance of acid-forming and base-forming elements in foods, to, 323; influence of cocaine up- on, 235; protein, from the stand- point of blood and tissue analy- sis, 87, 161. Metallic salts, absorption of by fish, 381. Method, for the determination of hippuric acid in urine, 257; for the determination of total nitrogen in urine, 493; optical, for determining concentrations of varicus proteins in blood serum, 179; Ritter’s, for the quantitative determination of cholesterol, errors in, 37. Methods for the determination of total non-protein nitrogen, urea and ammonia in blood, 527. Methylpentosides, a- and B-, action of tissue extracts containing nu- cleosidase on, 371. Meyer, G. M.: see Levene and Meyer, 347, 353, 361. Monosaccharides, combined action of muscle plasma and pancreas extract on 347. Index Muscle plasma and pancreas extract, combined action of on mono- and di-saccharides, 347. Muscle, purines of, 221; smooth, chemical analysis of the ash of, 401. Nitrogen, distribution of during a fast of one hundred and seven- teen days, 103; fecal, origin of, 5; total, in urine, new method for the determination of, 493; total non-protein, in blood, new method for determination of, 527. Nucleosidase, tissue extracts con- taining, action of on e- and p- methylpentosides, 371. Nutrition, studies in, 1, 5. Osan, H.: see Epstein and Olsan, 313. Odcytase, isolation of, 339. Optical method for determining the concentrations of proteins in blood serum, 179. Ox-serum, refractive indices of solu- tions of the proteins of, 179. 2-Oxy-l-methylpurine, 73. ~ 2-Oxy-8-methylpurine, 67. 2-Oxypurine, 67. Pancreas extract and muscle plas- ma, combined action of on mono- and di-saccharides, 347. Pentosuria, a case of, presenting unusual features, 211. PETTIBONE, C. J. V.: see Otto Folin, 507. Phaseolus, haemagglutinating and precipitating properties of, 47. Phenaceturic acid in urine, quan- titative determination of, 201. Phosphoric acid esters of inosite, 471. Phosphotungstic acid, as a clarify- ing agent in urine analysis, 81; blue color reaction of, with uric acid and other substances, 265. 559 Physiological action of pyrimidine compounds of the barbituric acid series, 443. Phytin, 471. Portal blood, origin and significance of ammonia in, 161. Potassium thiocyanate, action of on alanine, 97. Precipitating properties of the bean (Phaseolus), 47. PROCEEDINGS OF THE AMERICAN SOCI- ETY OF BIoLOGICAL CHEMISTRY, vii. Protein ingestion, low and _ high, putrefaction processes in the intestine of a man during a period of, 169; —— metabolism, from the standpoint of blood and tissue analysis, 87, 161. Proteins, in blood serum, optical method for determining the con- centrations of, 179; of cotton seed, utilization of, 1; of extrac- tive-free meat powder, utiliza- tion of, 5; of ox-serum, refrac- tive indices of ,179; refractive indices of solutions of, 179, 307. Purine-hexose compound, prelimi- nary note on, 85. Purines, of muscle, 221; researches on, 67, 73, 393. Putrefaction processes in man dur- ing fasting and during subse- quent low and high protein ingestion, 169. Pyrimidine compounds of the bar- biturie acid series, physiologi- cal action of, 448. Raper, H.S.: see Elliott and Raper, 2A: Recovery of alcohol from animal tissues, 61. Refractive indices, of solutions of proteins of ox-serum, 179; of solutions of salmine, 307. 560 Ritter’s method for the determina- tion of cholesterol, errors in, 37. Rospertson, JT. BrarL~srorp: On the isolation of odcytase, the fertilizing and cytolyzing sub- stance inmammalian blood sera, 339; On the refractive indices of solutions of certain proteins. VI. The proteins of ox-serum: a new optical method of determining the concentrations of the various proteins contained in blood sera, 179; On the refractive indices of solutions of certain proteins. VII. Salmine, 307. Ryan, L. A.:see Meigs and Ryan,401. Salmine, refractive indices of solu- tions of, 307. Salts, apparent antagonism between sugar and, 415; metallic, ab- sorption of by fish, 381. Scumipt, Cart L. A. and D. R. HoaGLanpD: The determination of aluminum in feces, 387. SCHNEIDER, Epwarp C.: The hae- magglutinating and precipitat- ing properties of the bean(Phase- olus), 47. Sea urchins, echinochrome from, 435. Sera, mammalian blood, isolation of oécytase from, 339. Serum, blood, optical method for determining the concentrations of various proteins in, 179. SHERMAN, H.C. and A. O. GETTLER: The balance of acid-forming and base-forming elements in foods, and its relation to ammonia metabolism, 323. SHERWIN, C. P. and P. B. Hawk: Fasting Studies. VII. The pu- trefaction processes in the intes- tine of aman during fasting and during subsequent periods of low and high protein ingestion, 169. Index Smooth muscle, chemical analysis of the ash of, 401. Sphingosine, 547. Spleen, dog’s, chemistry of, 27. Stearic acid, mannite esters of, 141, STEENBOCK, H.: Quantitative deter- minations of benzoic, hippuric and phenaceturic acids in urine, 201. Sugar, apparent antagonism be- tween salts and, 415; effect of lecithin upon fermentation of by bacteria, 313; —— solutions, toxicity of upon Fundulus, 415. Suprarenal gland, presence of active principle in before and after birth, 489. THoMaAs, ADRIAN: see White and Thomas, 381. Tissue, and blood analysis, pro- tein metabolism from the stand- point of, 87, 161; —— extracts containing nucleosidase, action of on a- and B-methylpentosides, 371; ——— juices, action of on glucose, 353- Tissues, action of on glucose, 353; animal, recovery of alcohol from, 61. Thyroid gland, presence of active principles in before and after birth, 489. Toxicity of sugar solutions upon Fundulus, 415. Trypsin, hydrolysis of casein by, 267 ; studies in the action of, 267. UNDERHILL, FRANK P. and CLAR- ENCE L. Buack: The influenee of cocaine upon metabolism with special reference to the elimina- tion of lactic acid, 235. Urea, determination of in urine, 507; new method for determination of in blood, 527. Index Uric acid, blue color reaction of phosphotungstic acid with, 265. Urine analysis, use of phosphotung- stic acid as clarifying agent in, 81. Urine, determination of ammonia in, 523; Determination of urea in, 507; new method for the determination of total nitrogen in, 493; of children, creatine in, 253; new method for the determination of uric acid in, 257. Urocanic acid, 537. 561 Utilization of proteins of cotton- seed, 1; of proteins of extrac- tive-free meat powder, 5. Watters, E. H.: Studies in the action of trypsin. I. On the hydrolysis of casein by trypsin, 267. Water drinking, studies on, 129. Wuitrt, Grorce F. and ADRIAN Tuomas: Studies on the ab- sorption of metallic salts by fish in their natural habitat. I. Absorption of copper by Fun- dulus heteroclitus, 381. at ae wh i wreted cries 7 SS Mita iM Cie . U4] sore 6) [ < ely a! a » + 72s S v if ‘ io a * hi ae > - ‘Le Owes y: ~~ ~ A ne dt) « A= 2s. 2 f _ e = Vi aa ay if cs = <> Fe rr 24 j Gj : + Ps % ; ns Se! ae eo + Je hi = Gapad _ ry id % —* ae = ’ aa ‘4 A) ae? ar, = cm ih i 4s) xy 29 tape id ie, + @ i od 2 ips ties nie a 7 es ; +> ° a ie 14 é = piel - 1 ~ = _ 7 es a: . at a aca lite ee : “t i it he or “Gio ael1g io eran Sey imonritaetn y .acpyblifies + “Ot Avon sie OR bel Lees “3! oar ie Vi abcd bt Y a tap eid A ’ TY 2a A S| : ! mio ¢ ' : Ae a or a8 r is - JO? ah ’ «i te = hts i a? Be - 2 i ao). ay awe Tee an a I + Gi eshte aie Sanaa rnd >) 1h a as en @ PYS reas iy | 7" as i. oy ’ Tt (owt. Ve. = Ts Tie 4s eo) : 4 os Oe ob tia - _ seo ke > 4 ead enya ae ’ 7) cies wie ‘a af: } ; Wi . i AA ey Meese a teak) Wa { a i mut itd I eG, ye hal WC ie aT hel rs \ — i; i 1 { ; Lj fi ve \ Py i Ca a Vil} J Pyros '\ QP The Journal of biological 501 chemistry Biolezical & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY vy mh whee: va Se tshiey 3%}. et