ere Sa eear aN ake Try SY See SERN eh Secse ; THE JOURNAL | OF BIOLOGICAL CHEMISTRY EO UNDED BY CHRISTIAN A. HERTER AND SUSTAINED IN PART BY THE CHRISTIAN A. HERTER MEMORIAL FUND EDITED BY H. D. DAKIN, New York, N. Y. A. N. RICHARDS, Philadelphia, Pa. £. K. DUNHAM, New York, N. Y. DONALD D. VAN SLYKE, New York, N. Y. LAFAYETTE B. MENDEL, NewHaven,Conn. CLARENCE J. WEST, New York, N. Y. WITH THE COLLABORATION OF J. J. ABEL, Baltimore, Md. P. A. LEVENE, New York, N. Y. STANLEY R. BENEDICT, New York, N. Y. JACQUES LOEB, New York, N. Y. R. H. CHITTENDEN, New Haven, Conn. A. S. LOEVENHART, Madison, Wis. t OTTO FOLIN, Boston, Mass. GRAHAM LUSK, New York, N. Y. WILLIAM J. GIES, New York, N. Y. E. V. McCOLLUM, Baltimore, Md. L. J. HENDERSON, Cambridge, Mass. A. B. MACALLUM, Toronto, Canada. REID HUNT, Boston, Mass. J. J. R. MACLEOD, Cleveland, Ohio. W. A. JACOBS, New York, N. Y. JOHN A. MANDEL, New York, N. Y,. WALTER JONES, Baltimore, Md. A. P. MATHEWS, Chicago, Ill. J. B. LEATHES, Sheffield, England. 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. Vv. C. VAUGHAN, Ann Arbor, Mich. ALFRED J. WAKEMAN, New Haven, Conn. | m ~~ 2 5 | l, | 4 |e VOLUME XxXxXI BALTIMORE 1917 CopyRIGHT 1917 BY THE JOURNAL OF BIOLOGICAL CHEMISTRY _ PUBLISHED BY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH FOR JOURNAL OF BIOLOGICAL CHEMISTRY, INC. r _ WAVERLY PRESS a Tue Wivtiiams & Witkins Company __Bartimorg, U.S. A. CONTENTS OF VOLUME XxXxXI. GRrEZNWALD, Isipor, and Weiss, Morris L. The fate of inosite ad- BMMIETeTEONCOROUPS rte 22s. lf. eu ears ts meeue 2) ee ees Curriz, James N. The citric acid fermentation of Aspergillus niger. vere spleen Ce Se Ss oe cc gious as ew tae Lene eee Jones, WALTER, and Reap, B..E. Uracil-cytosine dinucleotide..... Reap, B. E. Guanine mononucleotide (guanylic acid) and its prep- MEAMIOHE EOIN, Yeast, MUCICIC ACIG... 026. 62 ey es ce ee Harpine, Vicror Jonn, and Mason, Epwarp H. The estimation of Loree Tas arhingl seve bys 0G ee ere ee a Witper, Russert M. Intravenous injections of #B-hydroxybutyric AMUMACELOSACCHICHACIOS! = 22. .6<2 <5 002 5.gwns cee uncer ckelses ee ae Hoaauanp, Rautpw. The quantitative estimation of dextrose in meanmsyerallave: {HUSEYDOE ses eee eee oe ee eee note Bioor, W. R., and MacPuerson, D. J. The blood lipoids in anemia. . Fark, K. Grorce. Studies on enzyme action. XIV. Further experi- MieMissOM IPO TIC HCWOUS. 0.50 2. 6d esc vse es hmm ie KurtyamMa, SaHicgeNnosu, and Menpen, Larayerte B. The physio- fovicaivbehayror OL TANINOSE 5... 5.5 os. cw coe ade tee ope crammes OsBoRNE, THomAs B., and Mrenpet, Larayette B. The réle of vitamines in the CLIC eee ets, oe ee ee ee Bogert, L. Jean. A note on modifications of the colorimetric desea mination of uric acid in urine and in blood.................... Griitinc, E.M.K. The nutritive value of the diamino-acids occurring in proteins for the maintenance of adult mice.................. FRANKEL, Epwarp M. Studies on enzyme action. XV. Factors influencing the proteolytic activity of papain................ HENDERSON, YANDELL, and Morriss, W. H. Applications of gas analysis. I. The determination of CO, in alveolar air and blood, and the CO, combining power of plasma, and of whole McCo.tvm, E. V., and Prrz, W. The ‘‘vitamine’’ hypothesis and de- ficiency diseases. A study of experimental scurvy............ Dupin, Harry. The influence of bile on phenol production........ ‘ Kinessoury, F. B., and Sepewick, J. P. The uric acid content of the DOMORGE ME W=DOLRSS) s+ ...>,. este ee ee ons eee Pee eee 2 McCuenpon, J. F., SHepiov, A., and Tuomson, W. The hydro- gen ion concentration of the ileum content................-. Goss, B. C. Light production at low temperatures by catalysis with Mepvandemerimve oxide hydrosols:..::.:...:........<5-+52<%: Z Warner, D. E., and Epmonp, H. D. Blood fat in domestic fowls in re- AO MMONeS ITO UOUCtION: 62. ot chyessjce eg ee ee ee oe aye 125 . 149 165 173 201 iv Contents Reap, B. E., and TorrineHam, W. E. Tritico nucleio acid......... 295 Morsr, Max. The proteoclastic tissue enzymes of the spleen...... 303 SHERWIN, Cart P. Comparative metabolism of certain aromatic acids. 307 Harvpy, E. Newron. Studies on bioluminescence. VIII. The mechanism of the production of light during the oxidation of pyrogallol. 2.0.34. 58-¥-.. 22 ae ee soa 311 Jones, WALTER, and Reap, B. E. The structure of the purine mon- onuUCcleOtides sc. iio. Fagen oe dae ioe c tels Oe SR eee 337 Lorn, Jacques. The similarity of the action of salts upon the swell- . ing of animal membranes and of powdered colloids........... 343 Lewis, Howarp B. The metabolism of sulfur. II. The influence of small amounts of cystine on the balance of nitrogen in dogs main- tained ona low protein diet....... ..4... . 4-2: /4eeeeee eee 435 Givens, Maurice H. Studies in calcium and magnesium metabolism. III. The effect of fat and fatty acid derivatives................ 44] Harr, E. B., and Humpurey, G. C. The relation of the quality of protems) to milk production. UNI 2 2 2226s. ones eee 445 STEHLE, Raymonp L. A study of the effect of hydrochloric acid on the mineral excretion of dogs........:«.: do. 4:24. ©. 229 461 Peters, Joun P., Jr., and Gryetin, H. Rawue. The relation of adrenalin hyperglycemia to decreased alkali reserve of the blood... 2. ofc dt cid Medios ake ato Gved seller 471 Foster, G. L. A modification of the McLean-Van Slyke method for the determination of chlorides in blood.................... 483 HoacGuanp, Raupn, and MansFretp, C.M. The function of muscular tissue in urea formation. |... ...05..+..: sess ste 7) eee 487. HoaGuanp, Rauru, and Mansrietp, C. M. Glycolytic properties of MUSCUIATSUISSUC...¢ 350. ad ctu Ss ses ek ee ee ee 501, Contents Vv McCienpon, J. F., SHepLov, A., and THomson, W. Tables for finding the alkaline reserve of blood serum, in health and in acidosis, from the total CO». or the alveolar CO, or the pH at E Domh CLOR Rares (0) 0s eee a Be an Lee 519 Sure, Barnett, and Harr, E. B. The effect of temperature on the Beariions ar lyaine. with. DItTOUS ACI... 2... ss.cc.u: <1 ewlemeien ks o6) DOT McGuiean, Huan, and Ross, E. L. Methods for the determination of blood sugar in reference to its condition in the blood........... 533 Baumann, L., and Hines, H. M. The origin of creatine. II........ 549 Denis, W., and Minot, A.S. The production of creatinuria in normal Gr dlivliisys 5 Baws £0.73 Skee, ae ee ea RI ree OS As 561 Rosertson, T. Brartsrorp, and Derxtprat, M. Experimental studies on growth. IX. The influence of tethelin upon the early growth: of the white mouse... -..../1....5....... deco BLogaayeek. Dhe blood lipoids in nephritis... ..-...........3.%5 575 OstERHOUT, W.J.V. The dynamics of the process of death.......... 585 LEvENE,.P. A. The structure of yeast nucleic acid.................. 591 Levene, P. A., and Meyer, G. M. The removal of nitric acid from BOlonssOL orzanicucompounds 42-0, 22-22: 22: seat -a.. Soe 599 Crank. ob. Che preparation of lyxoses....:.<.1...0.0. 6056012 sdesa 605 Levene, P. A. Chondrosamine and its synthesis.................. 609 Levenr, P. A., and Meyer, G. M. The relation between the con- figuration and rotation of epimeric monocarboxyli¢e sugar acids. Pit he piiemwlavaraziGes: . . oo h.0o. o= 0c svelas oe eats oe ae = 623 Levens, P. A., and Mseyer,G.M. Cerebrosides. III. Conditions for HyenGl ysis. Glcere brosidess— cis. fi in oss ee eee Soh os nee Pee 627 LEVENE, P. A., and West, C. J. Cerebrosides. IV. Cerasin........ 635 Levens, P. A., and West, C. J. Cerebrosides. V. Cerebrosides of pieckdneya liver wand egg yolk... 4.552252. . o och oe Sb ees oe ace 649 lhayoless @, \(Wellithato 3.0.0.4 ee ee ee es ee 655 CORRECTIONS. On page 538, Vol. XXXI, No. 3, September, 1917, the 5th line from the bottom, for This, however, does disprove Scott's contention read This, however, does not disprove Scott’s contention. J On page 613, for Lyxohexosamine (synthetic chondrosamine) t Lyxohexosaminic acid (synthetic) read Lyxohegosamine (synthetic chondrosamine) Lyxohexosaminic acid (synthetic) On page 619, 20th line, for 60 gm. read 6.0 gm. On page 645, over the heading of the second column of the tabulation for C9 HoosNOpe read Co9H 105N Ore. THE FATE OF INOSITE ADMINISTERED TO DOGS. By ISIDOR GREENWALD anp MORRIS L. WEISS. (From the Harriman Research Laboratory, Roosevelt Hospital, New York.) (Received for publication, May 7, 1917.) In a recent publication Anderson! has reviewed the literature on the utilization of mosite by animals and has reported the re- sults of his own experiments. All investigators who have worked with inosite have found that it is not readily utilized by animals. After administration by mouth, a considerable amount of in- osite may disappear but, after subcutaneous administration, most of it is found in the urine. The destruction of inosite given by mouth has generally been ascribed to the action of the in- - testinal flora. In man, Anderson found that of 0.5 gm. of inosite per kilo of body weight given by mouth, only 9 per cent was found in the urine and none in the feces. In the dog, however, with doses of 2 gm. per kilo, apparently very little was absorbed from the intestine. A large part of the inosite administered _ could be recovered from the feces but only a small amount was _ found in the urine. - Ingestion of inosite did not raise the res- piratory quotient and Anderson concluded that “‘inosite is not utilized to any extent by the dog.” At the time Anderson’s experiments were published we had already been engaged with an investigation of the same subject. Our first experiments were planned to ascertain whether or not Inosite was, physiologically, related to the carbohydrates. Pre- ‘vious work had indicated that it was not. Kiilz? and Mayer had failed to observe a formation of glycogen from inosite. It ds true that Mayer had found a small amount of lactic acid in the urines of rabbits receiving inosite but, since rabbits readily 1 Anderson, R. J., J. Biol. Chem., 1916, xxv, 391. 2 Kiilz, E., Sitzungsber. Ges. Beférd. ges. Naturwiss., Marburg, 1876, No. 4, given in Maly’s Jahresber. Thierchem., 1876, vi, 45. 3 Mayer, P., Biochem. Z., 1907, ii, 393; 1908, ix, 533. 1 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. l 2 Inosite excrete lactic acid under a variety of conditions, this could not be regarded as a strong indication of a relation between inosite and carbohydrates. In our own experiments we used phlorhizinized dogs. In these animals, although there is a normal amount of glucose in the blood, it apparently cannot be utilized. Moreover, the or- ganism seems to form glucose from every available source. The ratio of glucose to nitrogen in the urine of fasting phlor- hizinized dogs remains for days between 3.0 and 3.6 (the varia- tion is much smaller in any one experiment) a value scarcely equalled and, in all probability, never exceeded even in the severest forms of diabetes mellitus. This ratio is assumed to represent the maximum formation of glucose from protein. If, therefore, this ratio rises after the administration of any sub- stance, that substance may be regarded as having been con- verted into glucose in the organism. There are certain excep- tions, such as narcotics and other toxic substances, which lber- ate the small amounts of glycogen which such phlorhizinized dogs still retain. In such cases the amount of glucose excreted in excess of the usual amount (“extra glucose’) bears no con- stant relation to the amount of material administered and the effect is obtained ‘only with the first few doses and is absent there- after, for all the available glycogen has been excreted as glucose.‘ At first the usual technique was employed. After the ratio of glucose to nitrogen in the urine (G:N ratio) had reached a constant level, the inosite, in aqueous solution, was injected subcutaneously. The G:N ratio rose very little. However, it was noticed that it did rise, even though very slightly, in every experiment. Unchanged inosite was also found in the urine (Tables IX and X). It was then determined to test the capa- city of the normal dog to oxidize inosite. It was found that about one-half of the inosite administered was excreted un- changed (Table I). It was believed that this poor utilization might be due to the fact that the inosite was rapidly absorbed from the aqueous solution and as rapidly excreted, before the tissues had much opportunity to change it. In order to secure a more gradual absorption of the inosite, it was then given in ‘Sansum, W. D., and Woodyatt, R. T., J. Biol. Chem., 1915, xxi, 1. I. Greenwald and M. L. Weiss 3 suspension in cottonseed oil at intervals of 4 to 6 hours. Deter- minations of the ratio of total carbon to nitrogen in the urine indicated that a fairly regular excretion of inosite could be se- cured in this manner (Table II). An experiment with glucose and a phlorhizinized dog showed that a fairly uniform excretion of the added glucose could be secured. Also, the length of the experimental period was increased from 12 hours to at least 48 hours, in which period the inosite was administered at in- tervals of 4 to 6 hours. In the normal fasting dog it was found that a considerable portion of the inosite administered could not be accounted for in the urine, either as such or by calculation from the ratio of carbon to nitrogen. Excretion into the intestine was improb- able for there were usually no feces. In one experiment (Table VII) the dog was fed a constant mixture of hashed boiled beef, cracker meal, Crisco (a hydrogenated fat), bone ash, and water. The periods were marked off with carmine. Determinations of the total carbon and nitrogen in the feces were made and an attempt was made to isolate inosite from the feces of the experi- mental period. This was unsuccessful and the ratios of carbon to nitrogen in the feces gave no reason for believing that inosite, or any derivative thereof, was present in the feces. A simple retention was also unlikely for in only two experi- ments (Tables IV and VI) was inosite found in the urine more than 12 hours after the close of the experimental period. There was a discrepancy between the amount of inosite calculated from the “extra carbon” and that actually isolated that was somewhat greater than was to be expected from the results ob- tained in the recovery of inosite added to urine. Among the possible derivatives of inosite were phenols and oxalic and lac- tic acids. The excretion of none of these was found to be in- creased (Tables I to IV).2 It was thought possible that some of the inosite might, be excreted, with or without previous change, in combination with sulfuric or glycuronic acids. No increase in the excretion of combined or ethereal sulfuric acid was observed. 5 Dubin, H. (J. Biol. Chem., 1916-17, xxviii, 429), has also found that the excretion of phenols in dogs is unaffected by the administration of inosite. As regards the excretion of inosite, his results are in accord with those of Anderson.! 4 Inosite There was, however, a slight, though unmistakable, increase in the excretion of glycuronic acid (Tables Viand VII). This was deter- mined by the method of Tollens,®’ in which the glycuronic acids are first precipitated with basic lead acetate and ammonium hydrox- ide, the precipitate is filtered off and washed, and then treated with boiling 12 per cent hydrochloric acid, distilling off the fur- furol formed. This is then precipitated as the phloroglucide.? There is very little evidence that the substance responsible for this increase is really glycuronic acid. The results are reported in terms of this only as a matter of convenience. That the in- crease was not due to the presence of inosite or of a reducing substance in the urine or to the administration of cottonseed oil was shown by suitable control experiments (Table VIII). It is very likely that the substance, whatever it may be, is identical with the dextrorotatory non-reducing substance observed by Mayer in the urine of some of the rabbits to which he adminis- tered inosite. In the experiments with phlorhizinized dogs (Tables XI to XV) it was found that the administration of imosite in cotton- seed oil was followed by a slight though unmistakable increase in the glucose: nitrogen ratio. Inosite is not a toxic or narcotic substance and the ‘extra glucose’ can hardly be regarded as being derived from the glycogen or other carbohydrate of the body. Moreover, the amount of “extra glucose’? increased as the experiment was continued, leaving very little doubt that the glucose was actually derived from the inosite administered. That it was not due to the cottonseed oil was shown by two ex- periments. In one of these (Table XIV) the dog received in- jections of cottonseed oil alone for four 12 hour periods and then, after a 24 hour interval, the inosite was administered in the usual manner. The oil alone produced only the slightest, ifany, rise in the glucose: nitrogen ratio. In the inosite period, how- ever, the usual rise was observed. In the other control experi- ment (Table XV) the dog received 15 ce. of cottonseed oil every 6 hours during the fore- and after-periods. This was at least 6 Tollens, C., Z. physiol. Chem., 1909, 1xi, 95: 7 In these experiments the paper was not included in the mixture sub- jected to distillation, thus avoiding the large correction employed by Tollens. I. Greenwald and M. L. Weiss 5 50 per cent more oil than was used for the suspension of the inosite. Nevertheless the glucose: nitrogen ratio rose in the usual manner with the administration of the inosite and fell thereafter. The excretion of acetone and of 6-hydroxybutyric acid was generally diminished in the experimental periods. Determinations were made of the total carbon in the urine. From this amount were deducted the amounts of carbon present as glucose, as acetone, and as 6-hydroxybutyric acid. From the residual value (“‘rest C’’), a carbon: nitrogen ratio was calculated. From this,. following the method introduced by Lusk for the calculation of the “extra glucose,” the ‘‘extra carbon” and the inosite equivalent thereto were calculated. The amount of inosite actually found corresponded, as well as could be expected, with the amount calculated in this manner. In only one ex- periment (Table XIV) was this not true. In this experiment, particularly in one period, the amount of “‘extra carbon” was very high. It is possible that the excretion of some unknown constituent was very irregular, giving rise to the results obtained. If the amount of “extra glucose” be added to the amount of inosite calculated from the “extra carbon,” the sum, except in the experiment already alluded to, is almost exactly equal to the amount of inosite administered. In view of the error in- herent in these determinations and calculations, the correspond- ence is surprisingly close. Apparently, therefore, inosite, though slowly and incompletely, is converted molecule for molecule into glucose. EXPERIMENTAL. The general plan of the experiments and the analytical methods were those generally employed in this laboratory.® Total carbon was determined by oxidation with sulfuric acid and potassium dichromate, passing the gaseous products of oxidation through a heated combustion tube containing copper oxide and lead chromate, drying with calcium chloride, and finally absorbing the carbon dioxide in soda-lime. Inosite was prepared from “steep water’? by the method of 8 Greenwald, I., J. Biol. Chem., 1914, xviii, 115; 1916, xxv, 81. 6 Inosite Griffin and Nelson.’ The isolation from the urine was accom- plished by Mayer’s method, crystallizing the inosite from alcohol and from acetic acid. In the experiments with phlorhizinized dogs this crystallization was repeated several times before the product was considered pure enough to weigh. The method of Meillére and Fleury? gave a purer product but the yield was much lower and extremely variable. Oxalic acid was determined by Dakin’s method" and phenols by the method of Folin and Denis.!? TABLE I. Excretion of Inosite after Administration in Aqueous Solution. Normal Fasting Dogs. 12 Hour Periods. “Extra pee Carbon.| C:N. Hees Inosite. Oxeks Remarks. gm. gm gm. gm gm 0.852) 0.589} 0.692 0.929} 2.114) 2.276) 1.475] 3.71 Weight 4.15 kilos. 5.81 gm. ino- site in 100 cc. H,0 in 6 hourly doses in first half of period. U.901| 0.648) 0.719} 0.028) 0.07 0.962) 0.657] 0.683 0.803} 0.700) 0.873 0.008 0.996) 0.859) 0.864 0.020 1.100, 2.950) 2.683} 2.00 | 5.06 | 0.010} Weight 5 kilos. 8.05 gm. inosite in 100 ec. H.O in 5 hourly doses ‘ in first half of period. 1.156) 0.976} 0.844 0.010 The column headed inosite gives the amount of inosite calculated from the “extra carbon.’’ The isolation of inosite from the urine was unsatis-_ factory because of mechanical losses, etc. 9 Griffin, E. G., and Nelson, J. M., J. Am. Chem. Soc., 1915, xxxvil, 1552. We are indebted to the Corn Products Refining Company for the supply of steep water. 10 Meillére, G., and Fleury, P., J. pharm. et chim., 1910, series if i, 348. 11 Dakin, H. D., J. Biol. Clem. 1907, iii, 77. 12 Folin, O., wad Denis, W., J. Biok Chem., 1915, xxii, 305. TABLE IT. Hourly Excretion of Inosite after Administration in Suspension in Cotton- seed Oil. Dog 11. gee ae Carbon.} C:N. ear Inosite. yer Remarks. hrs. gm. gm. gm. gm. gm. 12 0.963} 0.873} 0.906 0.011 12 0.826] 0.728) 0.882 2 0.175) 0.259) 1.49 | 0.102] 0.256 Weight 8 kilos. 4 gm. inosite in 10 cc. cottonseed oil. 2 0.183} 0.376) 2.06 | 0.212) 0.533 2 0.190} 0.658) 3.46 | 0.487] 1.223 4 gm. inosite in 10 cc. cottonseed oil. y 0.192) 0.791} 4.03 | 0.600; 1.508 2 0.175} 0.547} 3.14 | 0.390] 0.981 2 0.144) 0.316) 2.19 | 0.186) 0.468 2 0.169] 0.261] 1.54 | 0.108) 0.272 Potala nec... 1.208) 3.208 2.085] 5.241 Composite...| 1.228] 3.267 2.172) 5.461] 0.005 10 1.180} 1.201) 1.017) 0.138) 0.347| 0.008 12 1.347|) 1.207) 0.896 0.009 A composite of the 2 hour urines was prepared and analyzed, with the results given above. the amount of “extra carbon.” composite urine. TABLE III. Excretion of Inosite after Administration in Oil The values for inosite are those calculated from 4.4 gm. of inosite were isolated from the for Four 12 Hour Periods. Dog 17, Normal, Fasting, Weight 9 Kilos. Nitrogen. Carbon. C:N. Aeetaeeh poate gm. gm. gm. gm. 2.096 1571 0.750 0.0045 1.591 ISB yA 0.861 0.0030 1.415 1.579 1.116 0.306 0.0023 1.639 2.874 1.754 1.400 0.0034 1.939 2.944 1.519 e200 a OR0020 2.450 3.710 1.514 1.505 0.0017 2.205 2.210 1.000 0.221 0.0045 1.575 1.550 0.987 0.137 | 0.0023 1.621 1.506 0.929 1.309 1.230 0.940 evil, 205) 2 2a | Ore ae ee Tnosite. caer Found. ice gm gm, gm, 0.769 0.50 4.85 3.520 2.70 6.00 3.017 3.05 6.00 3.780 | 3.10 | 6.00 0.350 0.00 11.99 9.69 22.85 8 Inosite TABLE IV. Excretion of Inosite after Administration in Oil for Four 12 Hour Periods. Dog 21, Normal, Fasting, Weight 12 Kilos. Inosite. Nitrogen. Carbon. GaN. i xtra | Phenols. |= are Gator | round. | Admitit gm. gm. gm. gm. gm. gm. qm. 1.156 0.981 0.848 0.045 1.034 0.965 0.933 ' 0.049 0.995 2.289 2.301 1.39 0.047 3.50 2.56 9.2 0.932 3.600 3.864 2.76 0.058 6.94 Seog 8.0 1.448 3.926 PCA 2762 0.066 6.59 5.40 9.0 1.422 4.023 2.829 2.74 0.066 6.90 5.10 8.0 1.311 1.802 Lays 0.62 0.054 1.56 1.30 1.259 1.225 0.973 0.09 0.057 0.23 0.35 1.280 1.354 1.058 io 0.068 hs 0.13 1.144 1.023 0.895 0.067 BRO GAL rll ee sychatei ca kengdrat eal Mas wena ||| Nee DAS) || PAN Al7/ 34.2 *The urine was contaminated with blood from a wound. The C:N ratio is probably too high and the ‘“‘extra carbon’’ and inosite have there- fore not been calculated. TABLE V. Excretion of Inosite after Administration in Oil for Four 12 Hour Periods. Dog 25, Normal, Fasting, Weight 20 Kilos. Inosite. Nitrogen. Carbon. C:N “Extra, ate aa Calculated. Found. A duiieie: gm gm. gm gm gm gm 2.360 2.113 0.895 2.565 2.240 0.873 2.888 3.100 1.073 0.965 2.43 1.70 8.46 3.637 3.947 1.085 1.258 3.16 2.50 8.00 5.057 5.103 1.002 1.330 3.34 Seat 8.00 5.642 7.062 1.252 2.894 7.28 6.77 8.00 4.970 3.802 0.765 4.431 3.103 0.700 2.528* 1.902 0752 | Poth <5 «| a eee. ee 16.21 | 14.34 | 32.46 *10 hour period. ; The “extra carbon” and inosite have been calculated upon the basis of a basal C : N ratio of 0.739, the average of the last three periods. This gives the maximal values for “extra carbon’’ and inosite. ; TABLE VI. Excretion of Inosite after Administration in Oil for Four Successive 12 Hour Periods. Dog 25, Weight 19 Kilos. e Tnosite. + Sulfur. 3 : = , 3 ee) 3 Ss 8 : 2 | = rs) = 5 2 ere 8 | 38 ¢ a eae a (e) 1) 5 ie) ics < 4 12) oO gm. gm. gm gm. gm gm. gm gm. gm. 2.052 | 1.502) 0.732 0.084; 0.016} 0.125 2.336 | 1.781) 0.763 0.097) 0.018} 0.148 2.622 | 3.676} 1.402) 1.72 | 4.32 | 3.84 | 12.4 | 0.091] 0.017) 0.189 2.622 | 4.528} 1.727) 2.57 | 6.46 | 5.99 | 10.0 | 0.092) 0.016) 0.176 3.926 | 5.829) 1.487) 3.29 | 8.28 | 6.94 | 12.0 | 0.131) 0.016} 0.250 4.075 | 6.451] 1.583] 3.81 | 9.58 | 7.67 | 10.0 | 0.118] 0.012} 0.232 4.550 | 3.857| 0.848] 0.91 | 2.29 | 1.7 0.164) 0.015) 0.211 4 U4 | 3.401) 0.712) 0.31 | 0.77 | 0.31 0.163) 0.014! 0.162 3.474 | 2.248) 0.647 0.165 4.301 ; 2.789) 0.649 Motalcsl er ss silacs «os sx. 12.61 [31.70 26.49 | 44.4 Period. hrs. 10.5 12 12 12 12 12 12 13 10 14.5 The basal C : N ratio has been taken as 0.747 (the average of the ratios in the fore-periods) for the first two experimental periods and as 0.648 (the average of the ratios in the last two periods) for the others. TABLE VII. Excretion of Inosite after Administration in Oil for Four 24 Hour Periods. Dog 25, Weight 19 Kilos. Fed Daily. Food: gm. Baleumbeeteee sir te tee eee ees TR AR Ce 190 L DRGIREP GRE ge Sd ogee ner Spree ole Saree ne oe oir neiear 76 \CHIGOD. 2-2 6.0 oklG BAS Be eee a cd fn rd 5 ea eee 57 Evian emsls DEP aa we ce oe vest rea ie mys Soceeae See Asie oka treet Geer aahs 19 WWD. 05 3.4 ccs ob BO eo eC Se oie na eee ree 760 ce a Inosite. | Sulfur. 3 Feces Z ree te g , 5 i, | teen serie te 5. na = ee foie | So |e be Saem | 2S | Bot ate ee A 16) 'e) : 16) ey il ce EP ITS) oO 'e) Zz 2) gm, gm. gm. gm. gm. gm. gm. gm, gm. gm. gm. 11.10 | 7.18/0.647 0.452/0.076)0.338 9.75 | 6.30)0.646 0.358/0.078/0.320} 20.7] 1.61)12.9 9.22 | 9.92)1.076) 3.93) 9.87] 8.32} 20.9/0.323/0.061|0.489 9.87 |12.23)1.239) 5.81/14.61]/13.89} 22.0/0.286)0.051\0.464 10.51 |12.68]1.207|) 5.85]14.72/13.87| 22.0/0.379|0.060)/0.412 11.87 |11.79/0.993] 4.07|10.24| 8.01) 22.0/0.443/0.045/0.522) 34.4) 3.38/10.2 10.49 | 6.87|0.655 0.412/0.049)0.280 10.47 | 6.82/0.652 0.441/0.049/0.309} 16.6) 1.73) 9.6 Aion) |. || a rr 49.44/44 09) 86.9 9 10 Inosite TABLE VIII. Effect of Subcutaneous Administration of Cottonseed Oil on the Excretion of Glycuronic Acid. Dog 25, Weight 19 Kilos. Fed Daily. Periods 24 Hours Each. Food: gm. Boiled beetle 3i ss fe eciae se atte, Oe eee ere 190 Cracker me@all..)4.0. 00-8 ted, oe sees ee Oe ee 76 COTISCOK: Sustissseees we bic tere Ses tones a ee oie Ae ee ee 57 Bone ashi. jis csnceeeiten le heya ake ynin sav aie eee eee ee 19 WALOT S o..ch0 cleo 5 tsciera cSo lela g She, Cyhte mtteccaia: SI RIoI SST a aie tele nee ap teen 760 cc. Nitrogen. Glycuronic lactone. Cottonseed oil. gm. gm. ce. 11.56 0.300 10.97 0.348 10.25 0.325 4X 20 10.57 0.349 4 X 20 9.90 0.308 3 X 20 10.31 0.349 4 X 20 Lost. 9.97 0.355 9.01 0.308 TABLE IX. Inosite Administered in Aqueous Solution to a Phlorhizinized Dog, Weight 9.5 Kilos. Periods 12 Hours Each. f Glucose. a) Carbon. Tnosite. - ol 2 E oO . g B A g 3 ez o q r= ~ ° "eo c S - ® seh a) S 8 8 © ate = s = ; CH! ap 3 Te s a Cs : 25 : S c= ise) nm Spee |e] 2 led a eae eee ~ = oo oO ~~ = a pt o o) je ° 1s ° fox} ed ica] 3 io) eet a mo | A | Die] s < Q a : io) z oO | & {A gm gm.| gm gm.| gm gm gm. | gm.| gm. | gm.| gm.| gm. | gm. 10.08 |32.5/80.4/3.22 0.091|0.441/19.39/6.21|0.617 8.95 |29.1/28.1)3.25 0.091|0.573/17 .02|5.10|0.572 6.92 |24.9/24.8/3.61)1.94/0.065/0.482)19. 64/9. 46)1.365/5.04/12.7| 8.6)14.4 7.18 |24.2|23.6/3.37 0.077|0.448/14.76|4.87|0.678 5.83 |20.4 3.49 0.094/0 .629/12.44/4.00|0. 686 I. Greenwald and M. L. Weiss iat TABLE X. - Inosite Administered in Aqueous Solution to a Phlorhizinized Dog, Weight 12 Kilos. Periods 12 Hours Each. Glucose. 2 Carbon. Inosite. ah > F 2 3 =} ai E = on g 3 a 12 Bele bes. |.) |e ee | cease eles Pees | eee ae | 2 |) E24 |e te eke B MPa fect eipeedmaieas | cx | 842°) Oey. 1 EF se Re le gm. gm. | gm gm.| gm. | gm gm gm gm. | gm. | gm. | gm 5 .483* |18 .57/18 .6|3.39 0.039/0.120/11.15} 3.68/0.671 6.205 |19.60)19.6/3.16 0.060/0.234)12.52) 4.58/0.738 ; 5.915 |22.36/22.8/3.78/2.49/0.049/0.215/18.01) 8.98]1.519] 4.62 }11.6| 8 [12.7 5.817 |19.94/19.9/3.43 0.139/0.793)12.67| 4.29)/0.738 6.355 |22.01/21.8/3.46 0.099)}0.430)13.65) 4.64/0.730 *11 hour period. TABLE XI. Inosite Administered in Suspension in Oil to a Phlorhizinized Dog, Weight 20.6 Kilos. Periods 12 Hours Each. Glucose. 2 Carbon. Tnosite. = c PR © « 3 : 3 2 z ; A 3 R as! 5 = ~ ® =) i=} — fo} 1) iS - = ~ n oO J (5) oO FA = - a -_ 30 cs: 2 g 5 |S j So) Mbt stl etal ecepe || > ° 5 FI Z oc} 2 om 3 R Zz +7] 5 & g = OM pce Vers | Ce ulivermiimeersoligl |e tS elas Z fea aw = a |S a : .@) : oo erste |b gm. gm. | gm. gm.| gm. | gm.| gm. | gm. gm. | gm.| gm. | gm. 4.635* |16.34)15.9 7.026 — |25.62/24.5 3 3. 6.852 |27.10/26.3)/3. 19.75) 7.78/1.136)2.34/5.88/5.12)10.5 3 3 3 i 7.183 |26.09|/25.4 418.13} 6.10/0.849/0.39/0.98)0.58 2 7.150 {25.66 4.576** |14.69 IND c | satel Seis eens eee ee a ete | 2.8 ee , *8 hour period. **6 hour period. The “rest”’ carbon and the inosite have been calculated upon the basis of a basal C : N ratio of 0.795, which gives the maximal values for “‘extra’”’ carbon and inosite. Inosite Administered in Suspension in Oil to a Phlorhizinized Dog, for Four Successive 12 Hour Periods. Glucose. = Carbon. Tnosite. > E 3 z a ; g 3 2 3 § g, = 8 = g 5 : 4 ae ; "3 3 = 4 a fQ Ay oO 5 < Q a : ie) = 6) os So gm gm. gm. gm gm gm gm gm qm gm gm gm 7.104/24.57/24.7/3.46 7.068) 26.00/26 .4/3.68 0.451/4.26]16.74/4.15|0.587 7 .244)24 .65/24.7/3.40 0.525/4.57|17.9215. 68/0. 784 : 7 .278|28.70|29.1|3.77| 0.5110. 184]4. 72/22 24/8 .53/1.172| 2.71] 6.81] 4.15)10.45 6 937/27 . 62/28. 613.98] 1.94/0.169]2. 66/20. 22/7. 90/1.139| 2.35} 5.88] 3.81/10.45 6 352/25 .39/24.2/4.00} 3.68/0.079/0.82|18.37|7.84)1.234) 2.76] 6.93] 4.54/12.45 5820/24 .92/24.7/4.28] 3.84/0.060/0.51/18. 52/8. 23]1.414] 3.57) 8.99} 6.76/10.45 5 205/22. 18/21.9]4.36] 0.85/0.013/0.07/14.75|5.89]1.132) 1.73] 4.34) 2.10 3.854/14.89|14.9|3.87 0.007/0.03) 9.22/3.18|0.825 3.226]11.77|11.2|3.74 6.98/2.28]0.707 1.651| 6.18 3.75 Potala ses esi 10.82 | 13.12/32.95|21 .36/43.80 The caleulations have been made upon the basis of basal ratios of 3.70 and 0.800 for G:N and C: N, respectively. It is believed that these ratios give minimal values for ‘extra glucose’? and maximal values for inosite: TABLE XIII. Inosite Administered in Suspension in Oil to a Phlorhizinized Dog, Weight 15 Kilos, for Four Successive 12 Hour Periods. Glucose. 2 Carbon. Tnosite. = a B 3 a pe 3 Ec < 5 Mes eye [iat | Beas 2 2) aoa 1a | ele) € | 2 | eel a eee eee = Cie ae |e F< S ra} 6 ce 3 cs a pea les: Gi -Q Ay 1) : = Ron a 3 () 3 6) & < gm. gm. | gm gm gm gm gm gm. gm gm gm. | g 3.851* |12.81]12.3/3 .33 0.044/0.167) 7.11) 1.91)0.495 7.658 |24.40/23 .3/3.19 0.120/0.463)14.01| 4.02)0.525 8.395 |27.54/27.3|/3.28] 0.92/0.128/0.829|19.13] 7.72/0.920) 1.01) 2.53) 2.62/10.2 8.160 |28.52/27.213.49| 2.6110. 133/0.544/21.11] 9.44]1.157] 2.10) 5.27] 6.63)12.4 7.788 |27.46|26.5|3.53] 2.80/0.087|0.339|21.31|10.18]/1.307| 3.95) 9.93) 7.55)12.1 7.359 |25.54/26.3/3.47| 2.21/0.102/0.308/18.96] 8.60|1.169| 2.72) 6.83] 6.53) 9.0 7.349 |25.23/24.9)3.43] 1.91/0.092/0.293|17.04| 6.66|0.906] 0.78] 1-96] 2.18 7.148 |22.61|22.7|3.16 0.097|0.292|/15.43] 6.24|0.873] 0.52] 1.31] 0.00 6.315 |19.95|20.5/3.16 0.051)|0.222)12.94) 4.87/0.772 5.708 |19.03 3.33 0.054|0.156)12.56} 4.89)0.857 Totalicle ssc geen |e 10.45 11.08|27 .83)25.51|43.7 *6 hour period. The “extra glucose”? has been calculated upon the basis of a dominant ratio of 3.17. The wide variations in the value of the C:N ratio made the selection of « dominant ratio difficult.’ The value finally selected and used was 0.80. . 12 I. Greenwald and M. L. Weiss 3 TABLE XIV. Effect of Administration of Cottonseed Oil Alone, and of Inosite in Oil, to a Phlorhizinized Dog, Weight 11.6 Kilos. Periods 12 Hours Each. Glucose. i Carbon. 3. ; rt 2 Cottonseed oil. ct elated uae 5 aye iee| = Z, Seniesa ime ks | oils qm gm. | gm gm gm. | gm gm ce. 9.559 |32.36 3.39 4.812* |16.74 3.48 4.931* |17.29 3.0 10.02 |36.31 3.62 40 9.561 |35.04 307, 60 9.504 |33.48 3.52 40 8.857 {30.19 3.41 60 Tnosite. Ad- Calculated. |Found| minis- 8.478 |29.72 3.51 as 3.950* |13.44} 13.0) 3.40 8 .68|3 .34/0.844 gm. gm. | gm. 4.197* |14.22) 14.5) 3.30 8 .92)3 .26/0.778 7.774 |30.30| 29.1) 3.90} 6.10/20.17/8.12/1.045) 2.00 5.02 3.18] 7.88 7.723 |28.65| 28.7) 3.71] 4.09/20.49/9.10]1.178] 3.01 etsy 3.36)12.0 6.656 |26.85) 26.4) 4.04) 5.72/18.36/7.68/1.154) 2.43 6.12 ae PANO) 5.793 |21.98) 21.1) 3.80} 3.59]16.92)8.18/1.412] 3.62) 9.09(?)**| 2.83)12.0 5.124 |18.63) 17.6) 3.64) 2.35]12.72/5.31]1.061| 1.39 3.52 0.18 DI TAS* |) S42 3.07 5.51|2.16/0.784 on 128" | 9.21 2.94 5.99)2.33/0.745 5.946 17.33 2.91 ‘TRONS | SERS alee ee 21.85 12.45) 31.32 12.67/41 .88 *6 hour period. **Note the unusual discrepancy between amount of inosite calculated from ‘‘extra’’ carbon and that actually found. 14 Inosite TABLE XV. Inosite, Suspended in Cottonseed Oil, Administered to a Phlorhizinized Dog Receiving Oil throughout the Experiment. Dog 29, Weight 22 Kilos. Glucose. Carbon. ; Inosite. g : g ; 2 : a 2 z g 5 3 3 = : elma a = 3 Pc : oe eee me eee | eo: s it io 5S ~ a ire} % a Ky o B ees PP SV Ws Jeo) = |e | gm. gm. gm gm gm gm. gm gm. | gm. | g 6.010} 19.39 6.222] 19.53) 19.1 12.06 | 35.86} 35.3 11.36 | 36.74] 36.7 11.09 | 38.00) 37.9 10.85 | 38.59} 38.1 10.35 | 37.95} 37.6 23.14) 8.80)0.736 1.64) 24.18) 9.56/0.842) 1.10) 2.77] 1.80) 9.7 26.39}11.29]1.002| 2.85) 7.17) 6.43] 10.0 26.42)11.07|1.020) 2.98] 7.50} 6.54! 10.0 5.98) 25.73)10.64/1.028) 2.93) 7.36} 5.51) 10.0 9.65 | 33.77| 32.8 20.75) 7.29)0.755 9.57 | 30.65) 27.8 19.78] 7.59/0.793 Ropar etre |\osetereee ales Seve 20230 9 .86)24.80)20.28) 39.7 © or Ny oO e bo “IO ON eb uy or fo) ~J wwwwwwnw wow Shioy ee oo ~I LW) bo on SS oo Ne) or THE CITRIC ACID FERMENTATION OF ASPERGILLUS NIGER.* By JAMES N. CURRIE. (From the Research Laboratories, Dairy Division, United States Department of Agriculture, Washington.) PLATES 1 AND 2. (Received for publication, April 20, 1917.) INTRODUCTION. The citric acid fermentation induced by certain fungi has been elaborately studied, especially by Wehmer.! His work on this subject and also on the oxalic acid fermentation is familiar to ail students of fungi. Wehmer? believed that the production of citric acid in more than mere traces was characteristic of the group of fungi to which he gave the generic name C7tromyces and that the oxalic acid fermentation was characteristic of Asper- gillus niger. This seems to have been accepted by all the other workers who have investigated either the citric or oxalic acid fermentation. Martin,® in a very recent study of the citric acid fermentation, discarded all cultures of Aspergzlli with the assump- tion that their fermentative action was well known and that they did not produce citric acid. It has been noted‘ that many cultures of Aspergillus niger produced citric acid. Although the literature on the chemical activity of Aspergillus niger is voluminous, only one reference has been found relative to citric acid production by this group * Published with the permission of the Secretary of Agriculture. 1 Wehmer, C., Beitr. zur Kenntnis einheimscher Pilze, Hannover, 1893, No. 1. * Wehmer, in Lafar, F., Handb. technischen Mykologie, Jena, 2nd edition, 1905-07, iv, 242. 3 Martin, J. A., Am. J. Pharm., 1916, Ixxxviii, 337. ‘Thom, C., and Currie, J. N., J. Agric. Research, 1916, vii, 1. 15 16 Citric Acid Fermentation of fungi. In 1913 Zahorski® was granted a patent in the United States on a method for producing citric acid by fermenting sugar solutions with Sterigmatocystis nigra. This is one of the many names that ae been used to designate fungi of the black Asper- gillus group.4| Zahorski, however, states that Sterigmatocystis differs distinetly from Aspergillus. The writer at first supposed that Zahorski had worked with some very unusual culture of Aspergillus niger. This impression was probably wrong, for any one of about twenty cultures studied under certain conditions produced citric acid in abundance. In fact almost any culture of Aspergillus niger upon concentrated sugar solutions will produce much more citric acid than oxalic acid. For conducting the citric acid fermentation a well selected culture of Aspergillus niger is far superior to any culture resembling Wehmer’s Citromyces with which the writer has ever worked. In the beginning it was hoped that Aspergillus niger cultures could be divided into two general groups, one of which produced citric acid and the other oxalic acid. This would lend some aid to the problem of classifying this puzzling group of black Asper- gilli. In this respect the data are disappointing. No cultures produced citric acid only under all conditions or oxalic acid only. under all conditions. Many of the workers who have studied the citric acid fermen- tation performed only a few experiments without being guided by a fundamental knowledge of the metabolism of fungi or of the conditions favorable to the reaction with which they were concerned. Experiments conducted in this way are not likely to make a very definite contribution to any problem. In this paper three fundamental factors with regard to Aspergillus niger have been considered: (1) The inorganic salt requirements; (2) the general equation of metabolism; and (3) the reaction of the medium. Few concise statements can be made concerning the metabolism of an organism capable of producing such a variety of chemical transformations as Aspergillus niger. What is true for one set of conditions may not be true for another set of conditions differ- ing ever so little. Much of the preliminary work which served ©. no other useful end than to inform the experimenter will not be ’ Zahorski, B., U. S. Patent No. 1,066,358, July 1, 19138. NS Currie 1 described in detail, although general conclusions from such pre- liminary work may be related. Results were sometimes obtained which could not be duplicated. Such results have not been included without comment to this effect. Methods Employed. The cultures were selected from those used in the previous study and are designated by the same numbers, with the omission of the first two digits. ; The chemical reactions must proceed in the mycelium. This floats on the surface of the substrate but wrinkles in such a manner that it presents an enormous surface of contact. These wrinkled structures often project 5 to 6 cm. into the substrate. This pecu- liarity of growth enables the mold to exhaust a deep substrate much more rapidly than if it depended on diffusion alone. Never- theless the ratio of the surface to the volume of the media must be uniform in order to obtain results that have comparative values. A volume of 50 cc. of media contained in a 200 cc. Erlen-_ meyer flask was used in nearly all of the experiments here reported. The cultures were grown at 28°C. To determine the quantitative relations between the products formed and the sugar consumed, the medium was drained off and the mycelium repeatedly washed. The medium and washings were combined and made up to a definite volume. Separate portions of this solution were then taken for the estimation of oxalic acid, citric acid, and sugar. Oxalic acid was in all cases estimated by double precipitation as calcium oxalate and titration with standard permanganate. The sugar was estimated by reduction with Fehling’s solution and calculated from the reducing factor of invert sugar. The citric acid was estimated either by the method of Pratt® or the method of Kunz.?7. While either method will give fairly satis- factory results if used with discretion and patience, a really con- venient and accurate method for estimating citric acid is still wanting. Carbon dioxide was determined by drawing a gentle current of 6 Pratt, D. S., U. S. Dept. Agric., Bureau of Chemistry, Circ. 88, 1912. 7 Kunz, R., Arch. chem. Mikros., 1914, vii, 285. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 18 Citric Acid Fermentation carbon-dioxide-free air through the flasks containing the grow- ing mold and absorbing the carbon dioxide produced in caustic potash. Diligent search was made for other organic acids, especially malic and tartaric. These were never found and in all probability are never formed at all. The total acidity is nearly exactly ac- counted for by the sum of the oxalic and citric acids regardless of the proportion in which they occur. Both citric and oxalic acids have repeatedly been isolated and identified. The oxalic acid can be recovered directly from the fermented liquors by evaporation and crystallization. Citric acid because of its very great solubility has been recovered only through the calcium salt. Several pounds of calcium citrate have been prepared which by analysis corresponds to the formula Ca3(CsH;O07)2.4H2O and in no wise differs from calcium citrate prepared from the citrus fruits. The citric acid prepared from this calcium citrate by decomposition with sulfuric acid has some- times crystallized out anhydrous and sometimes with one mole- cule of water. There is some discussion in the literature on the isomerism of citric acid. Witter? claimed that an anhydrous isomeric citric acid was obtained by recrystallizing a sample of ordinary citric acid which had been dehydrated at 130°C. His conclusion was based on chemical data. On the other hand, Meyer® concluded from physicochemical data that the acids were identical. The anhydrous acid was the more stable form at high temperatures and the hydrated acid the more stable form at low temperatures. The conclusions of Meyer are in all likelihood correct. Mineral Requirements of Fungt. Earlier workers with fungi used very complex culture media. Raulin’s!” fluid which is still much used by mycologists is a good illustration of this type of medium. The composition as given by Raulin is as follows: 8 Witter, H., Ber. chem. Ges., 1892, xxv, 1159. 9 Meyer, J., Ber. chem. Ges., 1903, xxxvi, 3599. 10 Raulin, J., Ann. Sc. nat. (sér. 5, Botanique), 1869, ii, 224. : gm. DMeviillles! TAG, cole Mee Oe ee ee Sar 1,500 ce. Cane sugar..... EAN eno ns fine ithe tok og RE 70 Tanai, iol. Belo. ate NS. Ieee ea ae aint 4 AEM TINE ADE GIeUUGe ne siete eae steraie sie ove. arse: be ckonctdievs avs a. Mbeya ATMO MII MOS PATCH. acts tie scsi a\s nig ciees «ale eels 2 ee weiter ASS UIT IG A) OVATE eee ees eee a lok hs orrs <0, aoes ova 0,0 Bua lav ahoy a: Shas 31 cee MaomesimMECanOOMaAbees ricki oat. s ¢ ss tes satya dieveys,acd«,~) araeys TIMIMOMIITMPS UL ebm eyes cy eitts Ocisc + ss Ste. neat. edaccer PNG SUPA bes. se ata cetera CAEL PPA PI AS cio ASE HETERO USS Ula te see at an lenis s lace . oc 50 TRUE I TPO} oo oA en oS Se le es oo Te oe 1 MegSO:;.7H.O etm cre ote ts, ‘sicalts\aic\a eNeleMeeru As. cifaliey aes ele, 63 0 . 25 IKCll.w seaen bose ses 5 oe ee Ne eee aap) 0.25 Ttas(Oas7elhO). oo8 6.) 5 alate eRe eee IRD a> cee 0.01 IN GRIN(O piois ip: cu NEO I oe eee ae Varying quantity. The cultures were examined on the 7th day. The results in Table IV are in accord with all the other data on this particular question, and show clearly that the restriction of the supply of 30 Citrie Acid Fermentation nitrogen tends to increase the amount of citric acid at the expense of the end-products of metabolism. TABLE IV. Effect of Varying Quantities of Nitrogen on the Acid Fermentation of Asper- gillus niger 142. Results Expressed in n/10 Cc. per 50 Cc. of Medium. NaNOs per Acidity by titra- Weight of my- 1,000 cc. font Oxalie acid. : Citric acid. Sakata, qm. 1.2 166.9 108.1 58.3 0.398 1.4 169.3 130.6 . 39.7 0.500 1.6 162.2 137.5 21.2 0.548 1.8 161.0 141.0 24.4 0.634 2.0 152.4 143.5 14.5 0.606 2.2 158.2 150.9 14.1 0.616 2.4 158.0 152.5 yar 0.643 2.6 152.9 154.9 Guo 0.637 2.8 149.2 1532 7.8 0.644 3.0 114.8 124.1 0.0 0.695 In order to determine the quantity of potassium the medium should contain, four cultures were grown upon 50 ec. of medium of the following composition: gm. WEAR ET cco on. 5 iare S00 ois Site ee eS cn) Nin 1,000 ce. : Sacchanoseé::,. 40642. 5. 02.5 aoe ee ote eae 50 AS Fis 42 6 A eR ee ys. otk tk alee eee BTEC 2.0 MgSO, Sete Oe OLA CRO ERROR eric oO .0.0 0D O440.0 OD OIRO Cura as 0 aD TRG ae ete asco Sos « +. SR eee ee cd Varying quantity. The cultures were examined when 8 days old. For tabulation of results see Table V. The results indicate that the potassium requirements are very low, in most cases, 0.2 gm. per liter being sufficient. In order to determine the effect of varying quantities of magnesium sulfate on the acid fermentation of Aspergillus niger, Culture 142 was grown upon a medium of the following composition: FeSO..7H,0 ives Owes lhe ae ee a TABLE V. sje) oifers) 6) aila\te aie, alié|e),ere, ee) 0) \a)'s\\p sigs > 0) (= jo 0. 0\.ome 6 2 5 ¢ 0 ee #8 08 oo . Varying quantity. gm. 1,000 ec. 50 0.01 Effect of Varying Quantities of Potassium on the Acid Fermentation of As- pergillus niger. Culture. 142 69.4 28.7 The cultures were examined when KCl per 1,000 ce. (=p Ih (>) a> & bo Acidity by titration. 115. 136. 127. H He bo 110.4 118.8 97 .6 143.2 120.4 119.2 159.4 124.8 151.2 Oxalie acid. for) NS bo om OC 6 23.8 5 days old. Citric acid. 94.3 101.5 63.0 ek 98.2 93.8 133. 81 109. No & OO Results Expressed in N/10 Cc. per 50 Ce. of Medium. Weight of my- celium, 0.492 From the re- sults it appears that the most favorable quantity of MgSO,.7H.O for a medium of the above composition lies between 0.1 and 0.2 gm. per liter (Table VI). TABLE VI. Effect of Varying Quantities of MgSOs.7H:0 on the Acid Fermentation of Aspergillus niger 142. Results Expressed in N/10 Ce. per 50 Cc. of Medium. MgSO: per 1,000 ee. Beidity by titra- Oxalie acid. Sila 31.0 72.0 75.0 103.5 109.8 112.8 103.8 93.3 83.3 94.3 86.8 89.0 78.0 82.5 69.5 Citric acid. Weight of my- celium. gm, 0.123 0.394 0.521 0.551 0.547 0.460 0.456 0.436 ee Ne ea le ee 31 32 Citric Acid Fermentation In order to determine the effect of iron on the acid fermentation of Aspergillus niger, Culture 142 was grown upon a medium of the following composition: | gm. \\ CA) RE Bi OL Se. 5 Clowibaiicroi ts. - SeeecLaNcn saat ae eee 1,000 ee DACCHATORE 6. ro acsc te a ono SS roe Bete er erie. ae 50 POR soe ceils ee ee. Te eee ee ere iL WNaNQaniincsce see eee eee ae Rs aR dee ee 3 MegS0i7 AiO. cee oe VPA a en Bits ree eS 0.25 1.4 @} RSet cs poem Noi, here AS CO Saye erat 0.25 HeSO ROR Ss, .:: Per ee inc eee Varying quantity. The cultures were examined on the 7th day. No determina- tions were made other than total acidity. The results are shown below (50 cc. of medium were employed). FeSO,;.7H20 per 1,000 cc., gm... .. I/O GCIOICY MCC Rees. a - Heese 2 0.0 89.5 9.004 133.8 0.01 135.0 0.02 98.0 0.05 0.1 98.5 | 110.5 Other studies have been made on the effect of iron. In general the results were in agreement with those shown above. The addition of about 0.01 gm. of ferrous sulfate per liter to media containing nitrates stimulates the growth of mycelium and in- creases the rate of metabolism, especially in the earlier days of the fermentation. If the fermentation be continued for periods longer than 6 or 7 days, the unstimulated cultures tend to over- take and even to surpass the stimulated ones in total acidity. In order to determine whether the mixture of inorganic salts, contained in Czapek’s media, were not more complex than really necessary five cultures of Aspergillus niger were grown upon media of the composition shown below: qm. MESH T=) ee os 5 en a A | aco RR SRR Ce ACR G 1,000 ec. Sacehanraserere teed ac oo eee are, eRe eae ocr eae 50 NaNO Senter tete cee se die sche, ate OE TORE on oe Nae tee eta tas Once 2 BGS P10 5 or a ee nth A EE S Nene aoc e cine oar ek MgSO Bp Or rite ines, Sees Eee mane rOete Sie coasts ects 0.5 No “ . 1. The above salts. De ee a ESO) IONE eae pn id « +0.5 “™ + 0.01 FeSO;.7H26- “ce 4 “ <4 ce =. 0 5 NaCl. ‘ J. N. Currie 33 The results in Table VII show that the duplication of the potas- ~~ sium radicle and the introduction of chlorides and iron are not only unnecessary but in most cases even unfavorable to the acid fermentation of Aspergillus niger. TABLE VIL. Results Expressed in Ce. n/10 per 50 Cc. of Medium. Culture. Be Oty Acidity. Oxalie acid. | Citric acid. Caeleeee Mier a gm. 142 i 148.8 91.0 61.8 152.8 0.506 2, 142.8 103.6 44.0 147.6 0.469 3 152.0 118.6 40.8 157.4 0.495 4 137.6 88.6 60.4 149.0 0.486 69.4 il 35.6 41.4 94.8 136.2 0.462 2 4 38.0 83.5 1A) 0.472 3 79.6 42.4 42.9 85.3 0.454 4 103.0 2.0 Views) IIL G55 0.461 28.7 1 140.0 67.0 66.1 133.1 Or 2 82.8 43.2 OD. 78.9 0.521 3 78.6 32.6 47 9 80.5 0.570 4 92.8 37 .0 ley) 88.9 0.532 ° 74 1 115.4 72.0 47.3 THO 3: 0.438 2, 39.2 27.0 10.4 37.4 0.458 ! 3 45.4 32.0 19.0 41.0 0.493 4 44.4 34.8 12.4 3f 2 0.4388 96 1 O12": 44.4 45.9 90.1 0.597 2 78.8 34.0 44.3 78.3 0.549 3 TAL? 34.4 42.1 16.5 0.617 4 81.4 36.8 43.0 9.8 0.591 From the experience gained in all of the foregoing studies on the mineral nutrition, the reaction of media, and the general equation of metabolism of Aspergillus niger, and from many experiments not reported in this paper it was concluded that the most suitable medium for conducting the citric acid fermentation with Aspergillus niger should have about the following composi- tion per 1,000 ce. ! THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 34 Citrie Acid Ferment ation gm. SAO GOATORE icc oo. Se eh ne ee ET So ecieis eee 125 -150 INCELSIN OS ai0508 oc cos sh evs SE eee von hee 2.0 -2.5 1A I 2: 6 Pee rer aS eso ail inne a oN Oe a 0.75-1.0 IIo SO 67 gO ons Por ee ee ere I epee ah 0.20-0.25 HCl to pH 3.4 — 3.5 (5 — 4 ec. N/5). The medium ‘proposed by Wehmer*! contained ‘‘a small quan- tity of mineral salts (NHyNO;, KH,PO,, and MgSO,).”” Zahorski® used “‘small amounts of nutrient salts ‘such as ammonium, nitrate, potassium phosphate, and magnesium sulfate.” The addition of hydrochloric acid to the mixture of inorganic salts proposed above has several advantages. It raises the hydro- gen ion concentration to a point that makes complete steriliza- tion possible at a single heating in steam at atmospheric pressure for 30 minutes. As previously pointed out, it greatly reduces the dangers of infection of the lquors with organisms which might interfere with the citric acid fermentation without inhibiting the growth of the mold with which the liquors are inoculated. Many fermentations have been conducted in Erlenmeyer flasks on this medium. The general course of the fermentation is quite similar. There is little development of acid during the first 2 or 3 days. When a vigorous mycelial felt has developed the rise in acidity is very rapid, about 2 per cent in 24 hours, until the 7th or 8th day. After remaining nearly constant for 2 or 3 days the acidity begins to decline. When the fermentation proceeds properly the mold does not spore but remains white as shown in Figs. 3 and 4. The course of the fermentation under varying conditions is shown by the curves in Text-fig. 2. Curve I shows the course of the development of acid in a 12.5 per cent cane sugar solution. 100 cc. of the medium were contained in a 300 ec. Erlenmeyer flask. 5 ee. were removed at intervals and the acidity was de- termined by titration. Curve II shows the development of acid in a 15 per cent cane sugar solution. Curve III shows the course of the fermentation in a shallow pan (Fig. 5) containing 1 liter of a 15 per cent cane sugar solution. The mycelium had been developed in a previous fermentation and the rise in acidity began at once. 21 Wehmer, C., U.S. Patent No. 515,033, Feb. 20, 1894. - OY a ee |) a 3 | de N. Currie 35 With Culture 28.7 the acidity generally reaches about 10 per cent on the 8th day. Not infrequently the acidity is as high as 12 per cent although sometimes it fails to rise above 8 per cent. Along with the citric acid there are generally traces of oxalic acid and sometimes this latter acid may account for 3 or 4 per cent of the total acidity. The removal of this oxalic acid by partial neutralization of the fermented liquors with calcium car- bonate would probably present no serious difficulties. The variability of the fermentation under what appears to be identical conditions is a difficulty that has not been entirely over- cone. Wehmer? stated that the greatest difficulty he had to 1@. 12), TAPS Mes 20 22.126 DAYS Text-Fiac. 2. The course of the citric acid fermentation. contend with, aside from the infection of his hquors with undesira- ble organisms, was the varying fermentative power of his cul- tures (Variabilitat des Garvermégens). Throughout this study it has been emphasized that a complex biological reaction is involved which results in a number of products and which cannot be ex- pressed by a simple equation, representing the oxidation of a sugar to citric acid. All of the conditions that may influence the course of the fermentation are not within the control of the experimenter. Good results can be had only by the adoption of conditions that prove successful and can be duplicated with a high degree of uniformity. 36 Citrie Acid Fermentation While most of the experiments have been conducted in Erlen- meyer flasks of various sizes, the fermentation can be conducted with equal success in shallow pans. 1 liter of a 15 per cent sac- charose solution was fermented in each of three pans (Fig. 4). The fermentation was continued for 8 days. About 800 ec. of liquor could be recovered from each pan by pressing out the myce- lium in a hand filter press. Analytical data on the liquid from each pan are shown in the table below: | Pan. peat agg : Sy nae Calcium citrate. raniniceess Bi in thee precipitation. calcium citrate. liquor. per cent gm. per cent per cent per cent 1 11.54 3.8789 15,752 10.35 3.95 2 11.51 3.8209 15.28 10.19 4.69 3 11.00 3.7673 15.07 10.05 5.93 Already many substances of great technical value such as ethyl alcohol, acetic acid, butyric acid, and lactic acid are prepared by biochemical processes. In a recent paper on the chemical activ- ities of yeasts, molds, and bacteria, Ehrlich”? concluded with the prophecy that in time we would have a great chemical fermenta- tion industry in which many substances would be prepared which are now manufactured by expensive synthetic methods. Many such substances are known to occur as metabolic products of microorganisms. The painstaking investigation of all the condi- tions favoring the production of such substances will lay the only sure foundations for the development of a chemical fermentation industry. It is the hope of the writer that the work here recorded may prove a definite contribution to this much neglected but promising field of scientific endeavor. EXPLANATION OF PLATES. Puate 1. Fie. 1. Ten strains of Aspergillus niger. Showing sporification on a medium containing no iron. Fig. 2. Ten strains of Aspergillus niger. medium containing iron. Showing failure to spore on a 22 Ehrlich, F., Z. angew. Chem., 1914, xxvii, 48. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL XXXII. PLATE 1: 2 87 Sal o7 50 694 RrGzele (Currie: Citric Acid Fermentation.) ” yo we ANH As We ’ ‘7 ‘ee 7 % . 7 wy f THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXl. PLATE 2. (Currie: Citric Acid Fermentation.) dee Currie 4 PLATE 2. Fic. 3. Culture 28.7. I and III show the appearance of the culture when the citric acid fermentation proceeds properly. II shows the same culture on a medium which favors sporification. Fie. 4. Culture 28.7 fermenting 1 liter of liquid in a shallow pan. Fie. 5. Culture 28.7. Showing the appearance of the mycelium at the beginning of the third fermentation. Note the increased thickness and the deeply wrinkled structure of the mycelium. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 URACIL-CYTOSINE DINUCLEOTIDE. By WALTER JONES anp B. E. READ. (From the Laboratory of Physiological Chemistry, Johns Hopkins Medical School, Baltimore.) (Received for publication, May 7, 1917.) It has been shown that yeast nucleic acid is composed of the groups of four mononucleotides joined to one another through their carbohydrate groups.! When the nucleic acid is heated with ammonia it produces adenine-uracil dinucleotide,? evidently by hydrolytic rupture of its central nucleotide linkage as is indicated in the formula below (upper rectangle). We will now show that when the nucleic acid is heated with mineral acid its central nucleotide lnkage is not disturbed but the two terminal nucleotide linkages are broken and _ uracil- cytosine dinucleotide is formed, as is indicated in the formula | below (lower rectangle). The difference in the behavior of yeast nucleic acid toward ammo- _ nia on the one hand and toward mineral acid on the other hand . shows a remarkable specific difference between hydroxyl ions and hydrogen ions as hydrolytic agents; especially as the initial sub- stance and all of the hydrolytic products are easily soluble. Uracil-cytosine dinucleotide produces both uracil and cytosine but neither guanine nor adenine. It likewise forms both pyrimi- dine nucleosides but neither of the two purine nucleosides, and yields no easily split phosphoric acid. When an aqueous solution of the dinucleotide is treated with an alcoholic solution of brucine, a brucine salt is formed which crystallizes from hot water in macroscopic individual needles having the composition required for the formula CisH2sN5P2O0\.- 1 Jones, W., and Read, B. E., J. Biol. Chem., 1917, xxix, 123. 2 Jones and Read, J. Biol. Chem., 1917, xxix, 111. 39 40 Uracil-Cytosine Dinucleotide 4(Co3HagsN2O4).14H2O. As the brucine salt contains four equiva- lents of brucine, the mononucleotide groups that compose the dinucleotide must be united to one another through their carbo- hydrate groups.2. We formerly inferred this mode of nucleotide linkage! from a study of the rate at which phosphoric acid is liberated from yeast nucleic acid by hydrolysis with mineral acid. HO. O=P = O . C;H,;0>, . CsH4Ns HO” O H : HO. | i. /OH Oe = O + C;H,.O " C,H;N.02 HOT | : nH % O HO : HO. HO | OH : $ O H HOV O=P — O.C;H,0.2 . C;H.N;O HO” Formula for yeast nucleic acid showing how the two dinucleotides are formed. EXPERIMENTAL. Preparation and Properties of Uracil-Cytosine Dinucleotide. Commercial yeast nucleic acid in portions of 50 gm. was boiled with 250 ec. of 5 per cent sulfuric acid for 13 hours under an in- verted condenser. These are the conditions which our previously reported experiments show to be most favorable for the complete liberation of the purine-phosphoric acid with a minimum destruc- tion of pyrimidine nucleotides.? The product was treated with freshly precipitated silver oxide in such amount that a drop of the fluid formed a brown precipitate with sodium hydroxide, and after standing until perfectly cold the purine-silver compounds were filtered off. The solution was treated with warm saturated barium hydroxide until faintly alkaline to litmus and the pre- 3 Jones, J. Biol. Chem., 1916, xxiv, p. ili. W. Jones and B. E. Read 4] cipitated silver compound of the dinucleotide was suspended in hot water and decomposed with sulfuretted hydrogen. After treatment of the filtrate from silver sulfide with sulfuric acid for the removal of a trace of barium, the fluid was evaporated to a syrup at 50° under diminished pressure and the dinucleotide was precipitated with absolute aleohol. The precipitate easily hardens with absolute alcohol to a granular white powder. The dinucleotide is easily soluble in cold water and dextro- rotatory to polarized light. 1.5 gm. (7 per cent moisture) dissolved in 15 ec. of water gave a reading of + 2.82°ina2dm. tube. [a], = +15.0°. When boiled with twenty parts of 5 per cent sulfuric acid, the dinucleotide does not produce a trace of either guanine or adenine, and loses its phosphoric acid very slowly. 1.0362 gm. dried at 105° was boiled for 3 hours with 20 ce. of 5 per cent sulfuric acid. The cooled product was made alkaline with ammonia but no guanine was deposited even after standing for several hours in ice water. The warmed solution was then treated with magnesia mixture and after standing over night the precipitated magnesium ammonium phosphate was filtered off and weighed. mg WD imu cleobidesUseds casssa. a. ce Saeed eos oRotacmads wie dee 1,036.2 MetweOsGi>O obtained: . 2225/0. ..62 erties. sents 64.7 Calecnlatedsamount per gm:—hr.. 5.0.6... 1 css se le ook 20.8 In our reported studies of the rate at which phosphoric acid is liberated from yeast nucleic acid* we ascribed 20 mg. of magnesium ammonium phosphate per gm.-hour to the pyrimidine nucleotides. The filtrate from magnesium ammonium phosphate gave no purine precipitate with ammoniacal silver nitrate. The crude dinucleotide dried at 105° gave the following analytical results. The calculated values are for the formula CisHosN5P2016. I. 0.2147 gm. gave 0.2805 gm. CO, and 0.0812 gm. H,0O. II. 0.2804 “ required 8.47 ec. H2SO; (1 cc. = 0.0037 N). EET0.3267 -“ oe O81“ fe (Ga 0.0037 “ IV. 0.3587 “ gave 0.1081 gm. MgeP2O;. VerORaiog) 1. (0.1264 ~ * se 42 Uracil-Cytosine Dinucleotide Cc H | N Ve GU AtEGeT ; 2% cadlall, aauenn ace 34.34 3.97 11-13 9.86 Found. DOES «2, oe cpdcans & Rate Lae ae 35.63 4.20 1 EP Pe OS ee, Sec ook Wal al/ MD eo cra ic log ecg Oe Ee ee eal Ve. eee 8.42 Meo De ier TS A es ee 8.49 Preparation of the Uridine from Uracil-Cytosine Dinucleotide. 25 gm. of dinucleotide were heated in an autoclave for 2 hours at 140° with 140 cc. of 2 per cent ammonia. No guanosine was deposited even after the cooled product had stood several hours in ice water, and picric acid produced no precipitate of adenosine picrate. The fluid was transferred to a vacuum distilling appa- ratus, evaporated at 50° under diminished pressure to about 30 ec., and after the addition of 300 ec. of absolute alcohol, the solu- tion was saturated with dry hydrochloric acid gas. Uridine was then isolated by the method of Levene and La Forge.* After crystallization from hot alcohol the substance was obtained in large snow-white crystals which melted at 158-159° (corrected). I. 0.3451 gm. required 10.61 cc. H2SO, (1 ce. = 0.0037 gm. N). II. 0.3866 “ S 10.41 “ ieee (a 050037" fhe N Calculated for CoHi2N20¢ DHOdd domo UO sao DON Oooo on ogo so5 11.48 Found. De ict xt Saye ha er opahe BS oes Me This sav STU eT 11.38 10 Se anes AME rey CR corer iS clog o 0.5 ¢ 11.44 From 25 gm. of dinucleotide 5.2 gm. of pure uridine were . obtained. The mother liquor from uridine after evaporation of the alcohol gave qualitative tests for cytidine; 7.e., produced a crys- talline nitrate with nitric acid and a picrate with picric acid erystallizable from alcohol. 4 Levene, P. A., and La Forge, F. B., Ber. chem. Ges., 1912, xlv, 613. W. Jones and B. E. Read 43 Preparation of Uracil and Cytosine from Uracil-Cytosine Dinucleotide. 20 gm. of dinucleotide were heated with 100 cc. of 25 per cent sulfuric acid in an autoclave for 3 hours at 140°. After cooling, the black pigment was filtered off and the diluted red solution was treated with barium hydrate for the removal of sulfuric and phosphoric acids. The fluid was evaporated at 50° under dimin- ished pressure and treated with hot saturated picric acid solution as long as a drop of the fluid formed a precipitate with cold picric acid. A copious precipitate of cytosine picrate was formed (probably 6 or 8 gm.). A portion of the picrate upon recrystal- lization from hot water melted at 265-270° (uncorrected). Another portion of cytosine picrate was dissolved in hot water, acidified to Congo red with sulfuric acid, and after cooling somewhat was shaken out with ether for the complete removal of picric acid. The solution then was treated for cytosine by the silver-barium method. The final solution of cytosine sulfate was acidified to Congo red with sulfuric acid and evaporated to a small volume on the water bath. After cooling, it was strongly acidified with sulfuric acid and treated with two volumes of absolute alcohol. Cytosine sulfate was almost immediately thrown down in glistening crystal grains which appeared under the microscope as colorless transparent tables. This is the best method of preparing cytosine sulfate (C4H;N20)2- _ HSO0;.2H.0. A possible trace of uracil is eliminated, the col- oring matter remains in the alcoholic mother liquor, and the formation of basic cytosine sulfate is avoided. I. 0.2591 gm. lost 0.0264 gm. at 120°, gave 0.1686 gm. BaSOs, and required 16.54 cc. of H2SO, (1 ee. = 0.0037 gm. N). II. 0.2193 gm. required 13.92 ce. H,SOs. H:0 H2S04 N Calculated for (CsH;N3;0)2H.SO,.2H.O. .. 10.11 27.24 23 . 60 Found. Lehto isd oa eects et 10.19 Ale 23 .62 JUGS 28 8 See R OLE eee 23.49 From the pure sulfate, the pure picrate was prepared in trans- parent yellow needles which decomposed and melted sharply at 260-261° (uncorrected). 44 Uracil-Cytosine Dinucleotide The original filtrate from the crude cytosine picrate was treated with sulfuric acid and ether for the removal of picrie acid, and uracil was isolated by the silver-barium method in characteristic needle clusters. 0.2651 gm. required 17.87 ec. of H2SO, (1 ec. = 0.0037 gm. N). N Calculated tfor CgHiNsOnacaeen i + ck Co cet ec ieee ieee 25.00 10) 0 b« V6 EAeaRan or Sa eines ae Sebi ae Unchanged. Slightly turbid. After standing in the ice box for 23 hours they were neutralized and tested; 50 to 60 ec., equivalent amounts. 22 hours, glyceryl tri- ACCtALC HEN ce eee Asia s Get etnsiarie’s 0.76 0.50 0.32 0.29 These experiments show that greater amounts of alkali were needed to produce the corresponding inactivation with the lipase preparation than with the esterase preparation. They also show that the effect, 50 per cent loss in activity, caused by the H+ = 10-" is produced at once, standing 18 hours at room tem- perature causing no further inactivation. In the treatment of the esterase preparation with dilute acid or alkali there was a marked buffer action apparent, tending to bring the solutions to the neutral point. Alkali and acid were produced in the solutions so that when the mixtures were neu- tralized again, smaller quantities of the alkali and acid were required than were calculated from the quantities of acid and alkali originally added. These changes were not observed with | the lipase preparations. In some of these experiments the mixtures without and with treatment with acids and alkalis were titrated by the formol method. No increase in the formol titration was observed after > such treatments, so that the production of amino-carboxyl groupings. by the breaking up of peptide linkings is negligible with the concentrations of acid and alkali used. (c) Inactivation by Neutral Salts——An extended study of the effect produced by a number of different salts at various con- centrations on the activity of the castor bean preparations, mainly toward ethyl butyrate, showed that some of the salts retarded the actions very markedly, others to a less extent, while some acceler- ated the action of the enzyme.’ While the studies were incomplete in that the determination of the hydrogen ion concentrations was not included and only a limited number of experiments were made with esters other than ethyl butyrate, still regularities with’ 5 Falk, K. G., J. Am. Chem. Soc., 1913, xxxv, 601. Falk, K.G., and Sug- iura, K., ibid., 1915, xxxvii, 217. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 102 Enzyme Action. XIV different series of cations and anions could be observed, but there was no general underlying principle apparent governing the two actions, acceleration and retardation. These seemingly opposite actions may well be different phases of the same phenomenon. (d) Inactivation by Alcohols and by Acetone.-—Dilute solutions of methyl and ethyl alcohols and of acetone retarded the actions of both preparations, the amount of retardation increasing with the concentration of the aleohol or acetone. Solid preparations made by precipitation and washing with alcohol were always inactive. Solid esterase preparations precipitated and washed with acetone were active in a number of cases, but the activity was very much smaller than that of the corresponding solutions from which they were prepared. Similar solid lipase preparations were always inactive. The addition of glycerol, however, up to a concentration of 25 per cent had no effect upon the activity of the original castor bean preparation. (e) Inactivation by Esters’—The actions of the esters paralleled those of the aleohols. Methyl acetate inhibited the action more than ethyl butyrate, while glyceryl triacetate had very little, if any, such action. It was pointed out previously that these actions serve to explain part of the selective actions of the lipases which have been described in the past and that the action of the substrate on the enzyme must in all cases be taken into account when considering the properties of enzymes. . (f) Inactivation by Heat.—The original castor bean preparation containing the enzymes was inactivated by heating with water for a few minutes at 100°. The same preparation heated dry at 100— 110° lost 50 to 80 per cent of its activity. The same loss in weight brought about in a vacuum desiccator over phosphorus pentoxide was not accompanied by loss of activity. Drying first and then heating (the latter causing only 0.1 to 0.2 per cent increased loss in weight) caused 50 to 80 per cent loss in activity.’ The esterase preparation, C, = 10-7, on being boiled vigor- ously for 5 minutes and then cooled, showed C, = 107°, and a somewhat cloudy appearance. After being neutralized, no action 6 Falk, J. Am. Chem. Soc., 1913, xxxv, 616 and 1904. Falk and Sugiura, tbid., 1915, xxxvli, 217. 7 Valk, J. Am. Chem. Soc., 1913, xxxv, 616. ’ Falk and Sugiura, J. Am. Chem. Soc., 1915, xxxvil, 217. re} Kk. George Falk 103 was observed toward ethyl butyrate. The lipase preparation when boiled showed no change in hydrogen ion concentration while the activity was lost. It was shown previously that the esterase preparation on standing lost its activity very much more rapidly than did the lipase preparation, which, perhaps because of its physical condition, retained its activity very nearly unchanged for several days. Theoretical Considerations. The different ways in which the esterase and lipase preparations may be inactivated make it appear at first sight as if different reactions occurred in the inactivations. If, however, a definite chemical group is responsible for a definite enzyme action, it might perhaps be more reasonable to assume that inactivation followed a definite reaction. The preparations were essentially protein in character. There is no evidence that a dehydration, or loss of the elements of water, caused inactivation. Some of the reactions indicated that a possible hydrolysis may be a cause of inactivation. With proteins, hydrolysis is generally taken to occur with the —CO—NH-— group which goes over into the —COOH NH, —groups. Experiments with all the inactivations in no case showed an increase in the formol titration such as would be expected in this reaction, and therefore makes the assumption of such a hydrolysis improbable. Coagulation of the material accompanied some of the inactivations. The physical change alone does not appear satisfactory as an explanation, some change in chemical structure unquestionably accompanying or producing the physical phenomenon. Furthermore, the lipase material in suspension in water showed the same activity as when dissolved in 1.5 normal sodium chloride solution when tested immediately. The explanations of the chemical changes accompanying in- activation so far suggested are not satisfactory. The reagents~ used are simple. It is difficult to conceive of a very deep-seated chemical reaction taking place under so many different conditions, | none of a complex nature. The only chemical change which appears probable under these conditions is that involving a simple rearrangement within the molecule, such as a tautomeric change involving the change in position of the hydrogen atom. 104 Enzyme Action. XIV In considering the structure of proteins it is evident that such a rearrangement is possible in the peptide linking. The hypothesis to be suggested is that the active grouping of the esterase and lipase preparations is of the enol-lactim structure, —C(OH)=N-, the specificities being dependent in part upon the groups combined with the carbon and nitrogen, and that in- activation consists primarily in a rearrangement to the keto- lactam group —CO—NH-. The further work to be described here will show in brief that in the presence of simple peptides, esters are hydrolyzed under conditions which favor the production in the former of the enol- lactim grouping; that a substance, ethyl imidobenzoate, having the enol-lactim structure, possesses marked ester-hydrolyzing action as well as certain properties strikingly analogous to those of the naturally occurring lipolytic enzymes; and finally, that under conditions under which the occurrence or formation of the enol-lactim structure might be expected (action of alkali) ester- hydrolyzing substances are produced from proteins. Action of Dipeptides and of Amino-Acids. . ° In previous papers of this series® the ester-hydrolyzing actions of a number of amino-acids and peptides were studied. The work dealt more’ particularly with the specific actions of these sub- stances on a number of esters, and their behavior as a whole was studied without attempting to analyze more closely the groups responsible for the actions. In this paper the actions of some dipeptides will be described from the point of view of their struc- tural configuration, and in this connection a similar study of the amino-acids was necessary. According to the hypothesis suggested in the last section, ester- hydrolyzing action may be caused by the enol-lactim grouping —C—(OH)=N-— which loses this property in rearranging to the tautomeric keto-lactam grouping —CO—NH—. The action of alkali is generally considered to produce the former grouping in such organic compounds, perhaps with the simultaneous replace- ment of the hydroxyl hydrogen by metal. The best studied ° Falk, K. G., and Nelson, J. M., J. Am. Chem. Soc., 1912, xxxiv, 828. Hamlin, M. L., ibid., 1913, xxxv, 624 and 1897. \ K. George Falk 105 example of this appears to be isatin, but there is considerable evidence, much of it indirect it is true, that with compounds containing such groups, an equilibrium between the two tautomeric forms exists, the enol-lactim form predominating in alkaline solution, the keto-lactam form in neutral or acid solution. The hydrolytic actions of some of the simpler dipeptides chiefly on ethyl butyrate and glyceryl triacetate, since these were the esters used in the study of the naturally occurring lipases, as well as on a few other esters in neutral and shghtly alkaline solutions, were measured. The results found are given in Table I. The titrations were carried out by the formol method, and the results all corrected for blanks done under the same conditions. The hydrogen ion concentrations given in the headings refer to the solutions as made up at the beginning of the experiments by the addition of hydrochloric acid or sodium hydroxide; in a number of experiments the hydrogen ion concentrations were determined also at the end of the reactions and these are indicated by the numbers in parentheses in the body of the table.!° The results are given in terms of the amount of acid in tenths of millimols formed at 38° in the lengths of time indicated from 1.0 ce. of ethyl butyrate, 0.5 cc. of glyceryl triacetate, ete. Certain points appear clearly in the results as given. The actions are very much more marked in the alkaline solutions than in the neutral. At the concentration C, = 10~*°, 0.00001 normal hydroxyl ions, considerable action was obtained, espe- cially toward the acetates. There isin these cases a decrease in the hydroxyl ion concentration in the course of the experiments, approaching neutrality or going beyond in some cases. It is difficult to judge how far this influences the results, as it un- questionably does. The objection may be raised that the alka- linity of the solutions alone caused the hydrolysis, and that the peptides act only as buffer mixtures to keep the hydroxyl ion concentration predominant as compared to the water ester blanks, which, with the acetates, became neutral or slightly acid in re- action more rapidly. This objection is disposed of by compari- son with the results obtained with the amino-acid solutions later 10 The hydrogen ion concentrations in all the tables are indicated by the Sérensen symbol, pH, for the negative exponent of ten. XIV . Enzyme Action. 106 ee eee ———— FFF (0'8)0¢°0 6I OT GS (Q'8)0¢°0 61 OT GG (3° L)¢r°0 (¢°9)€T'0 (0°9)0 96 OT GG FS 0 Ig 0 €1'0 0 WW OT OP ‘ayer dyng [AYA (¢°¢)€F'0 GG OT GS (¢°¢)0F 0 GG G0 GS (0'9)F€ 0 GG ZO GG ¢°9)Z6°0 GG 10 GG (0° 1)Z2'0 GG ¢0°0 GG (¢°¢)18°0 SP ¢c0 GS (0°9)Z8°0 ¥G G0 GG (0°8)19'0 9 ¢c’0 GG (¢°¢)8r'0 61 ¢'0 BSG (0°9)¢6'0 oI ¢g'0 GS (G°L)ZL'T (8°¢)0€°0 61 G0]. & (0°8)9S'€ 61 ¢'0 GG (Z L)Sv € (0° 9)SF 0 (¢°¢)61'0 9G c'0 GS ‘ayRyooRtay [A004] Sy °29 90 0°6 0°8 OL 09 ; tid feta "10ST “OUIN[OA *Ppeatosqo suolpy ‘suapsy Uo sapydadig fo suoyoy arjfijosph yy ‘T AIAVL ooooo 1d N S88 oO oo COC 2S Se IDNA Nw a Scococooscoocococs S16 tS) O10 “auTDATS[ADAT) “mb ‘g0uBySqng 107 (02) S21 (0°S)SF'S (0'8)SF'Z (0°8)08"& (0°8)9¢° K. George Falk (0°8)F8'0 (0°8)8F'0 (0°8)#z'0 (0°8)98'0 (0°8)08°0 (GL)IF'T (0°8)61°0 (0°8)€Z'0 (0° 2)20°T (¢°2)60°T (0° 2)9F'0 (0' 9)0z'0 61 ” ” ¢'0 GG 61 ”» ” G0 GS 61 ” ” ¢'0 GS 61 ” ” G0 GS 6I ‘ayeyaority [AIOIATO GO GZ 6I ” » OT GS 61 ” ” On GG 61 ” ” OT GS 61 ”? ”? OT GS 61 ‘averdgng [Aq O'T CZ ZS ‘[I0 paosu0yyoy O'T CZ GG ‘THO 9ATIO O'T GG ‘o}vOZUOq 2 [Aueyd (uD) gO} 8 BG ‘operdyng [Ay G0 | 8 a ” yO} GG "oywozued [AYO ¢'0 GS CZ ‘ayeyoovtay TATOOATH GO GZ mae 9 jsucyg G0 | 8 rea %9 Mua eo] & GG AQP G0] gz (0°)60°0 ee 9 [Auoyd ¢'0 | = GZ (0°$)¢0°0 GG ‘oyeqoor [AYP ¢°0 GS 6I OT GS 6I OT GS Ol’ 10) OL ) OL’ OT Ol’ OL OL OL’ Sore Ore ooococo aurone[AoneT eurona[[AoApy autoA[s [Aono] ou1oA[s[Auvly euloA]S[AOATD eulone [Aone] aurone][ AoA oul A]s[Aone'T ouloA]S[AuBly auIdA]S[AATD ¢¢0'0 $¢0°0 ¢g0°0 ¢g0°0 ¢¢0°0 ¢S0°0 ¢g0°0 s¢0°0 ¢¢0°0 $S0°0 ¢0°0 ¢0°0 c10 0 ¢20°0 108 Enzyme Action. XIV where the buffer action is presumably the same, but entirely dif- ferent actions were obtained, both absolutely and relatively. A comparative study of the hydrolytic actions of the dipeptides in themselves also answers the objection. It is interesting to note that the action toward glyceryl tri- ‘acetate of glycylglycine at C, = 10~*° appears to be proportional to the amount of glycylglycine; but if increasing quantities of ester are used with a definite but small amount of glycylglycine, there is only a small increase in action.!! Toward ethyl butyrate the action is too small to allow definite comparisons. The actions of equimolar quantities of the five dipeptides may be compared. The following table gives the actions X 10~ on ethyl butyrate (1.0 ec.) and glyceryl triacetate (0.5 cc.) of 1 gm. molecule of each of the peptides under the conditions indi- eated in Table I, as caleulated from these results. ; Ratio, glyceryl butyrate. | triagotate, | tmanetete Gilyeviglyeme 1. in0 -yee ds cease 4.0 47.0 11.8 Theweylolyeime a5 4. sce as cee ee 5.3 48 .2 OF Gly cyllenietne:: ./..: 2. 50h 1).ceer ee 4.5 46.1 10.2 ReueyWeweme: 2) 2S. sect cee eee 9.0 46.1 5:1 IME vomalahi(oue Meme ncondu.audbioee.>4 cor 8.3 44.7 5.4 These results are strictly comparable, except for the differences (molar) in concentration of the peptides in solution, since the hydrogen ion concentrations of all changed practically to the same extent in the course of the experiments. Owing to the small actions actually obtained with ethyl butyrate, the experi- mental error is relatively much greater for the results with it than for the results with glyceryl triacetate. The constancy of the actions toward glyceryl triacetate is striking, the mean value being 46.4. Expressed in slightly different terms, this means that if the relations hold for all concentrations, 1 gm. molecule of the peptide will hydrolyze 0.464 gm. equivalents of glyceryl triacetate in 19 hours. Toward ethyl butyrate the variation in the actions is greater, the mean being 6.2, or very much less than the action 11 This is similar to the action of the lipase preparation described by Falk and Sugiura, J. Am. Chem. Soc., 1915, xxxvii, 226. K. George Falk 109 toward glyceryl triacetate under the given conditions. This is also brought out by the ratio of the action toward the two esters shown in the last column, these varying from 5 to 12, mean 8.3. The comparative action of the glycylglycine toward different esters may also by summarized. The following list contains the actions X 10 of 0.055 gm. of .glycylglycine at C, = 10-%° on one equivalent of each ester under the conditions as stated. Phenyl acetate. 3.98 Ethyl acetate. 1.84 Ethyl benzoate. 0 Glyceryl triace- tate. 2.84 < butyrate. 0.58 Phenyl ie 0 Methyl acetate. 2.40 Methyl benzoate. 0.20 The general formula for the peptides may be written as follows: CHR-—CO—NH-CHR’ CHR—C (OH) = N—CHR’ | | or NH; CO.H NH; CO.H The groups which may be considered involved in the actions are the amino and carboxyl groups or the central -CO—NH— group or its tautomer. In attempting to separate the actions of these groups, two lines of experimentation were followed. In the first place the actions of the amino and carboxyl groups were masked by studying the behavior of the glycylglycine ester hydrochloride and hydrobromide, and secondly, the actions of the amino and carboxyl groups alone were followed by studying amino-acids, all under the same conditions as those under which the peptides were studied. While these methods permit the studying of the groups alone it must be remembered that they leave out of account the important factor of the influence on the tautomeric equilibrium of the molecule as a whole. The results on p. 110 were obtained with the glycylglycine ester hydrogen halides. There was only very slight decomposition of the peptide ester as shown by the blanks on immediate titration and after standing. On the other hand, there was a marked tendency for the solution to become neutral on standing. There was not sufficient de- composition of the peptide ester blanks to account for this, the hydrogen halide evidently being involved in the reaction. How- ever, this was apparently slow enough to permit of marked ac- tions being observed with C, = 10-*° and 10-°%° initially. 110 Enzyme Action. XIV + RE ee Le ME NS PT 6.0 7.0 8.0 9.0 Glyeylglycine ethyl ester hydro- bromide, ethyl butyrate......... 0 (6.0)}0 (6.0)|0.64(6.5)|0.10(7.8) Glycylglycine ethyl ester hydro- chloride, ethyl butyrate ........ 0.05(6.0)|0 (5.8)/0 (6.0)/0.02(6.5) Glycylglycine ethyl ester hydro- bromide, glyceryl triacetate...../0 (5.2)/0.12(5.8) |0.54 (6.2)/0.98(6.2) Glycylglycine ethyl ester hydro- chloride, glyceryl triacetate...../0 (5.0)|0.03(5.5) 10.39(5.5)|0.89(6.0) The lack of action toward ethyl butyrate is significant. The action X 10-° of 1 gm. molecule of the peptide ester hydrogen halide toward glyceryl triacetate (0.5 ce.) is found to be 19.7 for the hydrobromide and 14.6 for the hydrochloride as compared with the action of 46.4 for the pure peptides. This difference may be accounted for by the more rapid increase in acidity tending to cause a shift of the tautomeric equilibrium to the presumably inactive keto-lactam form, and possibly by the difference in the composition of the molecule. The second method of studying the influence of the different groups separately is to compare the actions of some amino-acids with the peptides under similar conditions. This may be il- lustrated by comparing the following formulas for amino-acids and peptides. A. B. CHR—CO—NH-—CHR’ CHR i BN NH: CO. NH, CO.H By comparing the actions of equivalent amounts of substances of Formulas A and B under comparable conditions, it should be possible to find the action due to the grouping —CO—NH— in A | with the possible reservation that this group and the amino and carboxyl groups may exert reciprocal influences upon each other although no direct evidence of such influence has been obtained. Table II gives the results obtained with some amino-acids. Since considerable work was published in the previous papers of this series on the hydrolytic actions of amino-acids and their specificity toward certain esters, only a brief discussion of the K. George Falk TABLE II. Hydrolytic Actions of Amino-Acids on Esters. Substance. gm. Glycine. 0.033 0.0525 0.082 0.150 0.033 0.0525 0.082 0.150 Alanine. 0.045 0.16 0.045 0.16 Leucine. 0.045 0.10 0.16 0.10 0.16 Phenylalanine. 0.045 0.10 0.045 0.10 Tyrosine. 0.045 0.045 8 | Volume. 0 sig if Ester. ooo Oo FR Fe ee ce. .0 Ethyl butyrate. 0 Ethyl butyrate. 0 “cc “e 5 cc “ .0 Ethyl butyrate. 0 “ce “ce 0 “ec “ 5 “ce “ .0 Ethyl butyrate. 0 “ “ce .5 Glyceryl triacetate. “ nS “c .0 Ethyl butyrate. .5 Glyceryl triacetate. 0 “ “ 0 “ “ 0 “ “ .5 Glyceryl triacetate. 5 “ce ae 5 “ee “ce 5 “ “ .5 Glyceryl triacetate. .5 Glyceryl triacetate. Time of action. => Ss Z) me Re eB bo our C1 HD o1 mH o1 gn bo bo fora to bp ori or bo bh bw bw or or Ou tN bo He He Actions observed. SSS =—) 6.0 0.36 7.0 | 8.0 Noreococrod7”cjo G2 © (oe) Sor ore (SS) SS) He > HH Or Ww NN wre oOo (e/) results of Table II, and only in so far as they relate to the present subject, will be given. Considering the amino-acids alone first, it will be noticed that as a rule the action is as marked toward ethyl Omitting glycine, this is seen to be true throughout, and further, that the action even at C, = 10-*° is comparatively small and apparently butyrate as it is toward glyceryl triacetate. independent of the concentration of the amino-acids. The ir- 112 Enzyme Action. XIV regularities of the results are due to the relatively large experi- mental errors. With regard to glycine,” the first point to be brought out is that the ratios of the actions toward glyceryl triacetate and ethy] butyrate ranged from 1.30 to 1.37, mean 1.33. For the other amino-acids the ratio is also not very far from unity. This con- trasts sharply with the ratios of the actions of the dipeptides which ranged from 5 to 12, mean 8.3 This proves that the actions are not due to the hydroxyl ion concentrations, but that the amino-acids and peptides are the important factor. The hydrolytic action of 1 gm. molecule of glycine may be calculated just as with the peptides. Toward ethyl butyrate for 26 hours’ action it is found to be 6.1 X 10°; for 45 to 46 hours 8.2 * 10? (mean); toward glyceryl triacetate it is found to be 8.4 & 10? for 26 hours’ action, and 10.8 10? for 45 to 46 hours. With dipeptides the mean actions found for 19 hours were 6.2 X 10? toward ethyl butyrate and 46.4 X 10? toward glyceryl triacetate. This indicates that the action of the dipeptides toward the ethyl butyrate is due mainly to the amino and carboxyl groups and confirms the results obtained with the glycylglycine ester hydro- gen halides. Subtraction of the amino-acid glyceryl triacetate value from the value for the peptide leaves an action of 35.8 10 to be accounted for by the group —CO—NH-— or —C(OH) = N—. The mean value found with the peptide ester hydrogen halide was 17.2 X 10? but the difference as already stated may well be due to the shift in the tautomeric equilibrium as the solu- tion became neutral. The action of glycine ethyl ester hydrochloride was also studied and the following results were obtained. 0.10 gm. of glycine ethyl ester hydrochloride, 25 cc. of water, 23 hours’ action, 1.0 cc. of ethyl butyrate, or 0.5 cc. of glyceryl triacetate. DELS eee an eae ssc isdsnee Meceee eee 6.0 7.0 8.0 9.0 Ethyl-butyrates. 32605 5 SA 0.04 (5.0)|0.04 (6.0)/0.09 (6.0)|0.11 (7.0) Glyceryl triacetate ............... 0 (5.0)|/0.05 (5.8)|0.10 (5.5)|0.20 (6.0) 12 The glycine solutions were made up according to Sérensen’s directions for buffer mixtures and contained sodium chloride. K. George Falk 113 Just as with the glycylglycine ester hydrogen halides the solu- tion became neutral fairly rapidly. The very slight action when the amino-acid carboxyl groups are masked is also significant. The actions of the simpler amino-acids toward different esters were treated in detail by Hamlin and it was shown by him that if the esters are arranged in a series according to the extent of their hydrolyses, different arrangements result with the different amino-acids for the same hydrogen ion concentration and different as well from an isohydric acid solution containing no amino-acid. Since the keto-lactam group is present in other substances besides peptides, some experiments were made to find whether these exerted any hydrolytic action on esters in slightly alkaline solutions. Urea, studied at hydrogen ion concentrations between 10-*-° and 10-1°-° gave no action whatsoever on ethyl butyrate or glyceryl triacetate. Hippuric acid showed practically no action on ethyl butyrate and very slight action on glyceryl triacetate at H+ = 10-%-°, but the solutions rapidly became neutral or slightly acid. The hippuric acid exerted very little buffer action. A series of experiments was therefore run in which Sdérensen’s glycine buffer mixture for C, = 10-*° was added to the solution. The results follow. = : ; Action. Panes Glycine. Volume. aimee ae Gye = butyrate. triacetate. gm gm ce. hrs. 0.10 0.033 25 26 0.44(6.0) 0.60(5.8) 0.10 — 25 26 0 (6.0) 0.04(5.5) = 0.033 25 26 0.27(6.5) 0375-5) 0.20 0.082 25 46 0.66(7.0) 1.14(6.0) 0.20 = 25 46 0 (6.0) 0.10(6.0) = 0.082 25 46 0.60(7.0) 0.80(5.5) The action of the hippuric acid alone calculated from these results is 0.17 and 0.06 on the ethyl butyrate and 0.19 and 0.24 on the glyceryl triacetate; small actions it is true, but distinct. These results indicate that the structure of the compound as a whole is of importance in determining the equilibrium between the tautomeric forms, if these be involved in the actions. This question will be taken up again later. 114 Enzyme Action. XIV Action of Imido Ester on Esters. In order to obtain further evidence with regard to the hydro- lytic action on esters of substances containing the enol-lactim erouping, the behavior of an imido ester was studied. Imido TABLE III. Hydrolytic Actions of Ethyl Imidobenzoate on Esters. | | Blanks. Actions. | &B i oe} | = g Substance.* 3 a mediante re titra- yee Sa || ‘& |_. titration. ome Hi. hs/8)| pHL-4 | 3) pk: }S| & | irect + formol.| Direkt | P™y ge} PO | og! P | & = | ee oe gm. ‘hrs, | 5 0.075 | 23 6.0) 2.39+0.19 | 0.12+2.27 (8.0)/0.15 (6.5)|0.71 (6.5) 0.075 237.0, 0.64+0.04 | 0 +0.47 (8.5)/0.20 (7.5)|/1.39 (6.5) 0.075 | 23) 8.0) 0.11+0.11 | 0 +0.17 (8.5)\0.05 (7.5)/1.51 (7.0) 0.075 | 23} 9.0—0.0440.15.| 0 +0.10 (8.5)\0.08 (7.5)/1.34 (7.0) 0.06 | 22! 4.0 2.87+0 6.33+2.66 (3.8)|0.21 (3.8)|0.28 (3.8) 0.06 226.0 1.61+0 0.02+1.60 (8.5)|0.10 (6.5)|0.62 (6.0) 0.06. | 22/ 8.0) 0.10+0 0 +0 (9.0)\0.24 (7.0)}1.56 (6.8) 0.06 22/10 .0|\—0.10-+0 0 +0.14(9.0)0 (7.0){1.12 (6.5) pe S| BAS es eee ee 0.10 29 7.0 1.04++-0 0 +0.31 (9.0)(0.61 (7.5)|1.89 (7:5) 0.05 4.7.0) 0.51+0 0.06+0.34 (7.8) 0.27-(7.2) 0.05 | 22) 7.0 0.51+0 0 +0.34 (9.5) 1.18 (7.0) 0.05 | 47 7.0 0.51+0 10 +0.36 (9.5) 1.53 (5.8) 0.05 gm. portions in solution heated in incubator (38°) for different lengths of time after being brought to pH = 7.0, then tested at once and also after neutralization. Heated 4hrs. “cc 4 “cc “ce 93 “ce “ 93 “ce “ee 4S “ “ 4S “ce bdo ww =) “J 2% bo Ww & Oo bo 0} 7. — bo =) 8 0.040. Boiled 5 min. “cc 5 “ce | 20) 9.0\—0.24-+-0.2 20| 7 0 0.05+0.30 0 +0.43 (8.5) 0 +0.40 (8.5)) 0.04-+0. 34 (8.0) 0.04+0.39 (7.8)| 0.04+0.28 (7.0)| 0.04+0.33 (7.8)10 0.04+0.46 (8.0)0 (6.0) |0.06 (5.5) (6-0)|0.27 (5.5) K. George Falk 15 TABLE I1I—Concluded. 0.055 gm. of ethyl imidobenzoate hydrochloride, 25 cc. of water, 20 hours, action, pH = 7.0 initially. Immediate titration 0.27 + 0; titration after standing 0+ 0.37 (pH = 8.0) for blanks. Ester. Action. Ester. Action. cc. : cc. 0.5 Methyl acetate. 0.94(7.5)| 0.5 Ethyl benzoate. 0.03(7.8) 0.5 Ethyl ‘ 0.70(7.0)| 0.5(gm.) Phenyl benzoate.|0 (8.5) 0.5 Phenyl os 0.62(7.0)| 0.5 Ethyl butyrate. 0.20(7.8) 0.5 Glyceryl triacetate. 1.39(7.0)| 1.0 Olive oil. 0.16(9.0) 0.5 Methyl benzoate. 0.09(8.0)| 1.0 Cottonseed oil. 0.04(9.0) * Weighed as ethyl imidobenzoate hydrochloride. / esters possess the general formula Rae} . One of the 4 NN R’ simplest members of this class, and the one most readily pre- pared (in the form of hydrochloride) is ethyl imidobenzoate, OC,Hs 4 CsH; — C < . The results obtained with this substance \NH will be presented first in tabular form (in Table IIT) and then their significance will be discussed. The volume of the solution in each experiment was 25 ce. Ethyl imidobenzoate decomposes in aqueous solution, the decomposition being accelerated by acids or alkalis.’ In this decomposition ethyl benzoate, benzamide, benzonitrile, and ammonia or ammonium chloride may be formed. The decom- position is shown by the titration values of the blanks in the table. The column headed ‘Immediate titration” gives the results for titration to the first pink color with phenolphthalein without (“Direct”) and with (“Formol”) addition of neutralized formaldehyde solution to the solutions as first prepared. It is seen that in every case during the time of the action the direct and formol titration values have become practically interchanged. Because of the possibility of the formation of different products in the reaction, it is impossible to state the substances present at the different times. This makes it difficult to determine the ester blanks to be used as corrections. The experiments with 13 Stieghtz, J.. Am. Chem. J., 1908, xxxix, 29 and 166. 116 Enzyme Action. XIV ethyl benzoate show that the latter was not hydrolyzed in the reactions. Series of experiments with benzamide and with benzonitrile starting with C, = 10~*° to 10-*° showed no hydro- lytic action whatsoever toward ethyl butyrate or glyceryl tri- acetate. A series of experiments with ammonium chloride gave the following results, 0.025 gm. of NHyCl, 25 ec. of H.O, 19 hours’ action. Oy: ee ee Se ES ST eck coc 6.0 7.0 | 8.0 9.0 Ethyl butyrate....................|0.08(5.5)/0.13(6.8)|0.15(6.0)|0.18(7.8) Glyceryl. triacetate... asemecta-..- 0.12(5.0)/0.17(5.0)|0.22(6.0)|0.46 (5.5) These values were used throughout Table III as the ester cor- rections. The choice may appear to be somewhat arbitrary, but these values are if anything greater than those actually occurring, since in the imido ester experiments the ammonium chleride was probably not all present from the beginning. At any rate, this question is of minor importance when compared with the magnitude of the actions themselves. The action toward ethyl butyrate was found to be small except in one experiment. The actions toward glyceryl triacetate were comparatively large, however. A maximum action was observ- able at the hydrogen ion concentration of Cy = 10-*°, compared with more acid or more alkaline solutions. This may well be due to the more rapid decomposition of the imido ester in the latter solutions. With different amounts of imido ester and for different periods of time, no simple proportionality with the amount of hydrolysis is noticeable. This may also have been due to the decomposition of the imido ester. The action was also lost when the imido ester in neutral aqueous solution was allowed to stand at 38° for 23 hours or was boiled for 5 minutes. These experiments showed that the alkalinity of the solutions was of secondary importance in the actions. The results obtained with the different esters may be calculated to correspond to the actions which would be obtained with 1 gm. equivalent of the ester under the same conditions. The follow- ing values X 10°? are obtained in this way. doit; K. George Falk 117 Glyceryl triacetate. 1.77 Ethyl acetate. 1.34 Ethyl benzoate. Trace. Phenyl acetate. 1.62 % butyrate. 0.53 Phenyl! “ s Methyl i 1.45 Methyl benzoate. 0.23 This order of decreasing actions for ethyl imidobenzoate at C, = 107° is the same as that for glycylglycine at C, = 10-*° except that the positions of the first two members, glyceryl tri- acetate and phenyl acetate, are interchanged. This shows a marked similarity in behavior, while the minor difference may be due to secondary difference in structure. The action of 1 gm. equivalent of the imidobenzoate en glyceryl triace- tate (0.5 cc.+), 22 hours, Cg = 1077:° may also be given (mean values are shown). Creer ne: 1054-9 L080 1Oim2<° 11Oms:0 TOs 22° 1 Ope Action...... 8.7 18.4 37.8 42.8 33.2 34.6 The results, especially for C, = 10-7° to 10-° are not far removed from the results for the dipeptides at C, = 10-*°. The decomposition of the imido ester must also be taken into account here. The maximum action at C, = 107*° is also evident. Since the action toward ethyl butyrate is small or negligible, it is evident that the actions observed confirm the view based upon the behavior of the dipeptides, that the action is due — to the grouping —C(OR)=N-. Without considering the possible mechanism or cause for the reactions in either case, the ester-hydrolyzing action of the imido ester is similar to that of the naturally occurring lipases in that a maximum action is obtained at a definite hydrogen ion con- centration, and that the activity of both is destroyed by the action of acids, of alkalis, standing in solution, and heating in solution. Action of Alkali on Proteins. In order to determine whether or not the conditions which are known to favor the enol-lactim grouping in simple substances will produce ester-hydrolyzing groups or substances from pro- teins, a number of experiments were carried out in which casein, gelatin, and castor bean globulin were treated with alkali of different strengths and after neutralization tested for ester- hydrolyzing action. A typical experiment may be described as follows. THE JOURNAL OF BIOLOGICAL CHEMISTRY, XXXI, NO. l 118 Enzyme Action. XIV Three 2 gm. portions of casein (Kahlbaum’s) were treated with 25 ce. of sodium hydroxide solution of the following concentrations: (a) 3.0 molar, ()) 1.5 molar, (c) 0.6 molar. They were thoroughly shaken, and allowed to stand for 24 hours at room temperature. The mixtures then had the appearance of homogeneous light yellowish brown solutions. They were diluted with 75 ec. of water each, neutralized with concentrated hydrochloric acid to about C,, = 10~*-°, and dialyzed in collodion bags for 19 hours against running water. : (a) After dialysis, the volume had increased to 310 cc., a small amount of solid was present, C, = 10778 It was brought to C, = 10-7, and 45 ec. portions were tested. (b) Volume, 395 cc. after dialysis, turbid, C,, =10~*°. It was brought to C,, = 10-7", and 50 cc. portions were tested. ' (c) Volume increased to 400 cc., slightly turbid, C, = 10-°°. It was brought to C, = 10-7-°, and 50 cc. portions were tested. Actions (corrected for blanks), 47 hours, (a) Ethyl butyrate 0.08; Glyceryl tri- acetate, 0.56; (b) Ethyl butyrate 0.11; Glyceryl triacetate 0.50; (c) Ethyl butyrate 0.10; Glyceryl triacetate 0.48. In some experiments the alkali was removed completely by dialysis; in others, it was neutralized by acid without dialysis; in others, removed in part first by dialysis and then neutralized, and vice versa. Different strengths of alkali were used for various periods of time. In about half of the experiments toluene was added throughout in every preparation; in the remainder no toluene at all was added. No difference in results was obtained in the two series, and no growth was obtained either in agar or in blood agar. Some of the results obtained are summarized in Table IV. In the titrations with the gelatin experiments it was necessary to use the formol method in order to obtain satisfactory end- points. With the casein and castor bean globulin titrations formaldehyde was not added. The effect of boiling the alkali- treated protein solution before testing the actions was tried several times. With the casein mixtures no difference in action was observed between the heated and unheated mixtures; with the gelatin, in one case two-thirds of the activity was lost; in two other attempts there was no change.'4 On boiling these solutions there was a small increase in alkalinity, greater with casein than 14 Reference may be made to the thermostabile lipase described by Kendal, A. I., Walker, A. W., and Day, A. A., J. Infect. Dis., 1914, xv, 455. K. George Falk 119 TABLE IV. Hydrolytic Actions of Alkali-Treated Proteins on Esters. Treatment. Action. NaOH = Pe cs 2 3 ei se 5 Method of neutralization. 2 E = 2 : & 3 =] 3 ah Seeesel ed (he | ag | 2 le o ones 2 o o Shae) rs! O65 mleneele ee o | £ | 22| = | 25 oe ie) < cal Fey Et eo) isa] 16) Casein. gm. N cc. hrs. hrs. gm. 2.0 | 0.11; 15] 24 | Diluted, dialyzed. @.5| 45, | 0:35) 0.03}-0.87 20) |2'-0 25 22 | Diluted, neutralized, dialyzed, neutral- ized. 120\" 46810. 25) 0.05) 0737 ZHOR SOF A25 |, +22 sf 7.0} 46 | 0.30) 0.08) 0.48 ZeOm Oko e ol, 22 e 7.0} 46 | 0.30) 0.06} 0.55 2.0 }+1.0 | .25 | 48 | Diluted, neutralized, to: 4.0} 25 | 0.50 0.15 Zen eslaO) |) 25.) AS S G20) 25190250 0.10 PAO lel 25.1 48 ss 8.0} 25 | 0.50 eli, PAOR lO 25. | 48 oi LORO |e 25740250 1.47 PROM OM S25.) 24 . 8.0} 22 | 0:40 1.68 PROM le ON e224! e CAO 2220" ORA0 0.91 PAOnI ABO! | 25 18 ee 8.0} 24 | 0:24 Ue il} Z On eO. |) 25 |, 24 - 8.0) 22 | 0253 2.00 Pam ele Onl s25:.||, 24 ss 8.0} 22 | 0.50 173 SEO OL.4) 30 4 of SLO = 185 SO 1.86 540) 0451 30 4 s 820) 218) 0825 0.74 Gelatin. PROM lO mea2o.)|) 45 ee 7.0} 46 | 0.40) 0.13) 0.96 ANOm ate O 50) I) 942 es 8.0} 42 | 0.40) 0.22} 1.80 AR Om Omlee sO) 2 ss @.0) ©422| 0.25 86 BAD || OLE OA 55 * 8.0} 45 | 0.40 0.63 QAO MPOeZe| 2508)" 45 BOS 253 ne 8.0! 45 | 0.40 2.38 2:0) PEO |) 50) | 45 ce 8.0) 45 | 0.40 4.28 Castor bean globulin. 1.0 | 2.5.| 25 | 24 | Diluted, dialyzed. 8.5! 48 | 0.20} 0.24) 0.73 1.0 | 2.5 | 25 | 24 | Neutralized, diluted, dialyzed. 7.0) 22 |-0.15! 0.09) 0.27 120 Enzyme Action. XIV with gelatin. Dialysis removed or destroyed the activity of the gelatin preparations but only decreased that of the casein prep- arations somewhat. Dissolving the gelatin in water, bringing the solutions to the hydrogen ion concentrations 10-®° to 10-*°, and testing their actions toward ethyl butyrate and glyceryl triacetate gave only a trace of hydrolysis. A parallelism was observed between the action of the -gelatin preparation toward the glyceryl triacetate and the amount of formol titration. From the evidence presented-with the peptides and the amino-acids, — the amount of amino carboxyl groups appears to have no direct connection with the amount of glyceryl triacetate hydrolysis, so that this does not mean that the action observed here was due to the presence of peptides. The main point brought out in these results is the very marked hydrolytic actions of proteins after treatment with alkali. Not enough work was done on the question to show the dependence of the action on the method and details of treatment, on a number of different esters, etc. Dialysis of the casein and castor bean globulin preparation indicates that simple substances are not responsible for the actions. A systematic study of these relations is now being carried on in this laboratory by Dr. Florence Hulton Frankel. DISCUSSION. The experimental evidence presented in the first part of this paper led to the view that the inactivation of lipase was due to a tautomeric change or rearrangement within the molecule. Con- versely, therefore, active lipase material should be formed by a tautomeric change in which the equilibrium between the tauto- meric substances would be such that the lipolytically active structure was capable of existing. The lipase materials pre- pared from castor beans and soy beans in the course of this work ~ were essentially protein jn character, and the most obvious group- ing having such possibilities of tautomerism is that involved in the peptide linking. The next step in the work was therefore to find whether such tautomeric groupings show hydrolytic actions toward ester and whether these actions were lost by con- version into the keto-lactam structures. The work described with the peptides, amino-acids, and imido ester showed that the K. George Falk eal grouping —C(OR) = N —, in which R may stand for hydrogen or an organic radical, has hydrolytic action on esters and that the tautomeric group —CO—NR-— does not. This evidence does not show anything with regard to the active lipolytic grouping in enzymes, but the work with the action of alkalis on protein to form ester-hydrolyzing substances under conditions favoring an enol-lactim structure may be considered to be evidence bearing on this point. It must be emphasized that no direct conclusive evidence is presented as to the actual chemical configuration of the active lipase grouping. The steps in the reasoning may be summarized as follows. Inactivation (and therefore also activation) is assumed to be due to a tautomeric rearrangement whose possible nature is indicated. Simple substances possessing such structures show the actions and some other properties of naturally occurring lipases present in protein materials. Inactive proteins treated in such a way as to produce the supposedly active grouping show ester-hydrolyzing properties. Whether it is possible to go much beyond this in the present state of the knowledge of the chemical nature of proteins and the changes they undergo with simple treatment is an open question. However, one possible line of development bearing upon the present problem may be indicated. The equilibrium in solution between the tautomeric forms of ethyl acetate depends to a great extent upon the solvent.* This suggests that with the enol-lactim keto-lactam tautomerism in proteins the colloidal properties of the protein material may well exert an influence on the grouping comparable to the effects of solvents, and that the decreased stability or increased rates of inactivation of enzyme preparations, when separated to a greater or less extent from colloidal and other matter not connected with the actions, may be placed in parallel with the actions of the solvents on the equilibria between the tautomeric forms of ethyl acetacetate, ete. Reference may be made to some preliminary tests in this connection. Precipitation and adsorption of the proteins of the esterase and lipase preparations by aluminium hydroxide (alumina 15 Meyer, K. H., and Willson, F. G., Ber. chem. Ges., 1914, xlvii, 832. 122 Enzyme Action. XIV cream) showed that under certain conditions the activities of these substances were retained somewhat longer than in the absence of aluminium hydroxide. Attempts to adsorb dipeptides in the tautomeric enol-lactim form on egg albumin or gum arabic and then to neutralize the solutions and still retain the former in the active ester-hydrolyzing form failed. In the development of the hypothesis regarding the active erouping in lipase actions, the experimental work and discussion were limited almost entirely to the peptide linking. It is evident, however, that such tautomeric structures, enol-lactim and keto- lactam, may be present in other groupings, and the results of this investigation in no way limit the lipolytic activity to the peptide linking. In view of the complexity of the protein molecule, it is highly probable that such similar tautomeric groups may be present in combination with other groups and that the specificities of the actions are in part dependent upon these. This is especi- ally true of the actions described with the esterase preparation, in whichthere was marked activity toward ethyl butyrate. This action may be due to the free amino carboxyl groups under favor- able conditions as shown by glycine, and also to the enol-lactim structure in suitable combination. It must be admitted that the treatment of proteins with alkah to form active substances is rather strenuous. Unquestionably, simpler methods, comparable to those taking place in nature, will be found to produce the same effects. The fact that dilute alkalis inactivate the castor bean globulin lipase, while a certain higher concentration of alkali produces an_ ester-hydrolyzing substance from the inactive globulin preparation, indicates that differently placed groups in the molecule are involved in these two changes. SUMMARY. The inactivation of esterase and lipase preparations by acids, bases, neutral salts, alcohols, acetone, esters, and heat, led to the hypothesis that the active enzyme grouping in these substances possessed the enol-lactim structure, -C(OH) = N—, which be- came inactive by tautomerization to the keto-lactam structure, —CO-—NH-. K. George Falk 123 This hypothesis was tested by studying the actions of such group- ings in dipeptides and an imido ester. The dependence of the actions on the grouping was discussed, and certain similarities were pointed out between the behavior of the imido ester and naturally occurring lipases. The production of ester-hydrolyzing substances by the action of alkali on proteins under conditions which might be expected to form the hypothetical active grouping was shown. The bearing of this work on the nature of the chemical group responsible for lipase actions was discussed. 1 THE PHYSIOLOGICAL BEHAVIOR OF RAFFINOSE. By SHIGENOBU KURIYAMA anp LAFAYETTE B. MENDEL. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven.) (Received for publication, May 23, 1917.) Though raffinase is frequently found in plants, fungi, bacteria, yeast, and invertebrate animals (1, 2, 3, 4), its presence in the body in higher animals has never been positively demonstrated. The blood serum, bile, extracts of the mucous membrane of the stomach and small intestine, pancreas, thyroid, and testicle have no raffinose-splitting power (5, 6, 7). After injecting 11 to 22 gm. of raffinose subcutaneously in man, F. Voit recovered 65 to 92 per cent of it in the urine (8). In Magnus-Levy’s rabbit experiments, however, subcutaneously injected raffinose was recovered quantitatively in the urine (9). After feeding 10 gm. of raffinose to starv- ing hens, Kiilz noted some glycogen formation in the liver (10). Discuss- ing Kiilz’s results, however, Pfliiger considered the glycogen formation after raffinose feeding very doubtful (11). When raffinose was fed to a diabetic dog, severe diarrhea followed. An increase of urine sugar was rather uncertain (12). Comparing the rapidity of absorption of various kinds of sugar administered into intestinal loops of rabbits, Hédon demon- strated that raffinose is by far more slowly absorbed than the other sugars (13). Haldsz injected 49 to 147 gm. of raffinose into the rectum, in men, ~and examined the feces, eliminated from 2 to 6} hours later, for raffinose and its cleavage products. 2.6 to 74.7 gm. of raffinose disappeared (14). We have attempted to obtain additional evidence concerning the physiological behavior of raffinose. Part of this sugar used. was obtained through the courtesy of Dr. C. 8. Hudson in Wash- ington; the rest was a Kahlbaum preparation. The Influence of the Hydrogen Ion Concentration of the Medium upon the Activity of Raffinase. As a preliminary to renewed search for raffinase in the animal body, it became essential to learn what medium is favorable for the action of this enzyme. 125 126 Physiological Behavior of Raffinose Methods.—The hydrogen ion concentration of the medium was deter- mined by the indicator method. The standard solutions employed were mono- and dipotassium phosphate mixture, acetic acid and sodium acetate mixture, and hydrochlorie acid and disodium citrate mixture. Phenol- phthalein, neutral red, methyl red, Congo red, and tropeolin 00 were used as indicators (15, 16, 17). Raffinase solution was prepared from brewers’ bottom yeast by the method suggested by Hudson for sucrase preparation (10 day autolysis and 3 day dialysis) (18). The clear enzyme solution, preserved with chloroform, was neutral and contained 42 mg. of nitrogen per 100 ce. It gave a positive biuret reaction. No melibiase was present. To test the enzyme action in media with a desired hydrogen ion con- centration, either the standard regulator mixtures or sulfuric acid, di- luted in various degrees, were used. The latter were prepared by diluting 0.2, 2, 4, 10, 20, 30, 40, and 50 cc. respectively of 0.01 N sulfuric acid to 100 ec. with distilled water. The test was completed as follows: 15 ec. of a regulator mixture or a diluted sulfuric acid and 3 ec. of a 10 per cent raffinose solution were heated in a stoppered bottle in a ther- mostat at 40°C. 2 cc. of the raffinase solution were then added. After 40 minutes 1 ce. of a 3 per cent mercuric chloride solution was added, in order to stop the enzyme action instantly. The bottle was cooled, and after 4 hours, after which any multirotation of the cleavage products might be avoided, the filtrate of the mixture was examined polarimetrically. The temperature of the thermostat during the experiment did not vary more than 0.5°C. Another portion of the same mixture of the standard solution or diluted sulfuric acid with raffinose and enzyme solution was used for determining the hydrogen ion concentration. The mixtures showed the same hydrogen ion concentration before or after heating. When a stand- ard solution was mixed with raffinose and enzyme, the original hydrogen ion concentration was unchanged. In case of diluted sulfuric acid, the hydrogen ion concentration was markedly decreased by the addition of raffinose solution. In calculating the extent of the inversion of raffinose, a decrease of rotation to half of the original was considered to show the complete inversion of raffinose into levulose and melibiose. In order to see if the acidity itself was adequate to invert raffinose, the enzyme prep- aration was previously boiled for 1 hour in some experiments. From Table I and Fig. 1 it will be seen that under the ex- perimental conditions employed, the optimal zone of hydrogen ion concentration of the medium for the raffinase activity 1s pH 3.8-5.4. The same hydrogen ion concentration, no matter how it is prepared, shows the same influence upon the raffinase activity. The most favorable hydrogen ion concentration for the activity of sucrase is reported to be pH 4.44.6 (15) or pH 5.25-3.67 (19). Raffinase, which was formerly considered to be identical with sucrase, has nearly the same relation to hydrogen ion concentration of the medium. TABLE I. The Influence of the Hydrogen Ion Concentration of the Medium upon the Activity of Raffinase.. Medium. cee ee See Experiment I. Ventzke° per cent Ehosphatermixture-.- nae. so: oes Sal 0 0 oe AEE Eye ees 8.0 0 0 Acetate ak eS ete he lia 6.7 1.05 24.3 re ee eaenene Ficus es 5.3 ets 40.5 ss Ae BS A A eas 4.3 1.90 43.9 ; eS Oe, CO ee 3.1 1.40 32.4 Citrate at De see Rape 2.2 0.10 2.3 ie , (with boiled LAAT) 508 Sela ee eee eee Oe ee 2.2 0 0 Experiment II. EwOsphate MIXtUre.. 2... ss =. 23 8.7 0 0 te La rey. Ss acy 8.0 0 0 ag * Ba a Re RE Sr 7.4 0 0 Acetate Ey Rae Sas tnd hg 6.7 1.05 24.3 ot Cs WES Siem, ee RE 6.0 2.00 46.2 wy ND ee aes ee ener 5.3 2.05 47.4 os Sag teats prone ror cane 4.3 2.00 46.2 i uae ee OE at: 3.4 1.85 42.8 Citrate Re Wat ie ees aeiiot 2ee 0 0 on sg (with boiled RREZINAANC Orrin a ious, 2 owe sachs wakes 2.2 0 0 Experiment ITI. SLOVENES C10 ee ea eae 6.8 0.90 20.8 3 A. PN ls es 2,9) os 6.5 1.45 33.5 s CO eho Geo eae 6.1 1.75 40.5 - 28 Se a eee 4.9 2.15 49.7 = ES et eae 4.0 2.05 47.4 3 SE ess ty career Ine ie 3.5 2.00 46.2 we See eee ee eer Aric, Bee Sa a 3.0 1.35 31.2 f SO ANSE, ts 6 Cee re 2.5 0.90 20.8 “ “(with boiled enzyme).. Jay 0 0 Experiment IV. pe OUUR FES Chalet) C10 [ley hoa ia ee a 6.8 0.65 15.0 re “23% Salk Se ee 6.5 1.19 27.5 “e Se So. d Se ae ee 6.1 1.59 36.8 ed Pd 5 Sth ene ee ee 4.9 1-91 44.2 ee ee ies ae i a 4.0 1.90 43.9 ae Th ae sc ions Se ea a 3.5 1.55 35.8 se Sayre ee Sinn ok YS Fo 3.0 1.20 rahe pe a eee A 2.9 0.55 1227 ee “(with boiled enzyme). 225 0 0 127 xX Legulekor mixlivie : 7 f “ ° re) Sue zéc acid — Sf dl Z 2) = = = io) = = on > = KB 8 we 6 5 + 3 2 pH of medium Fire. 1. Does Raffinase Occur in the Alimentary Tract? Pautz and Vogel (5) and Fischer and Niebel (6) failed to detect raffinase in the body of higher animals. We have examined (a) human saliva, (b) bladder bile of a rabbit, (c) water extract of dog pancreas, (d) water extracts of the liver of a dog and a rabbit, and (e) water extracts of the mucous membrane of the small and large intestine of a dog and a rabbit. Methods.—The organs were comminuted with double volumes of water and kept for 24 hours with toluene at room temperature. The mixture was then filtered with cloth. When the filtrate was strongly acid, sodium carbonate solution was carefully added so that the final reaction was very slightly acid. 5 ce. of the filtrate were mixed with 1 cc. of a 10 per cent raffinose solution and 0.2 ce. of toluene, and incubated for 24 hours at 38- 40°C. The mixture was then diluted with 19 ec. of water and clarified with 5 ee. of colloidal iron solution. The final water clear solution was examined for its rotation and reducing power. For control, another sample of the same mixture was examined without incubation. Saliva and bile were tested in the same manner, but here the examination was performed S. Kuriyama and L. B. Mendel 129 both at the original reaction of these fluids and in a condition slightly acidified with dilute aceticacid. The saliva and pancreas extract employed were strongly amylolytic, and the extract of the mucous membrane of the small intestine contained sucrase. The experiments all failed to show the presence of raffinase. It has been reported that inulin, sucrose, and erythrodextrin can be inverted in the stomach, not by corresponding enzymes, but by the free hydrochloric acid of the stomach juice (20, 21, 22, 23). To test the possibility of raffinose being split in this way, media containing 1 per cent raffinose and 0.04, 0.1, or 0.2 per cent hydrochloric acid were incubated for 53 hours at 38—40° C. and then examined polarimetrically at once. TABLE II. The Inversion of Raffinose in Media Containing 0.04, 0.1, or 0.2 per cent Hydrochloric Acid. te Oe Examined immediately. Steet Rotation. Reduction. Rotation. Reduction. per cent Vex V3: +5.90 = +4.30 +4+-+ 0.2 +5.90 - +4.70 SPaPsr +5.90 — +4.60 +++ $ +4.83 SPSRqr 0.1 +5.85 _ 4.4.83 ene +5.38 46 4r Ls = 0.04 +5.85 4.5.35 oe From Table II it will be seen that the inversion of raffinose in the stomach is possible under suitable conditions. The Activity of Raffinose in Rabbit Serum. From the results shown in Table I it will be seen that the nor- mal reaction of the blood is unfavorable for the activity of raffinase (dog blood pH 7.32-7.64, rabbit blood 7.18—-7.70) (24). To ascer- tain whether the parenteral administration of raffinose can call forth the corresponding enzyme in the blood, it was necessary to determine how raffinase acts in the serum. 130 Physiological Behavior of Raffinose Methods.—The blood was taken from a jugular vein and cooled 3 hours, after which the serum was separated by centrifuging. A mixture was pre- pared as follows: 8 cc. of serum (or 3 ce. of 0.9 per cent NaCl) + 1 ce. of 10 per cent raflinose + 1 ec. of raffinase solution (the original concentration is’ designated as “‘raffinose stronger,’’ a five times diluted solution as ‘‘raf- finose weaker’’?) + 1 ce. of water (or 1 ce. of 0.025 n CH;COOH—“‘acid weaker,’’ or 1 ec. of 0.075 n CH; COOH—“‘acid stronger,’’ or 1 ee. of 0.025 N NaOQH—“‘alkali’’) + 0.2 ec. of toluene. The mixture was incubated for 40 hours at 38-40°C., and then diluted with 19 cc. of water. After clari- fying with 5 cc. of colloidal iron solution, the water clear filtrate was examined for rotation (Ventzke degrees with a 2 dm. tube) and reducing power. ; TABLE III. The Activity of Raffinase in Rabbit Serwm. Raffincse plus Reaction Rota- | Reduc- Boro eeune Rathiiase. Acid ale or (litmus). tion. tion. Ve Serum. Raffinase Water. Alkaline. |+1.95) — (weaker). oh os Acid (weaker). § +1.91) = ee ss Alkali. ss +1.92) — Saline solution. 3 as Water. Neutral. |+1.02/++-+ et de’ & Acid (weaker).| Acid. +1.02/+++ sf of as Alkali. Alkaline. |+1.91) — Serum. Raffinase Water. Neutral. |+1.92| — (boiled). Saline solution. ie “ +1.91) — Serum. Raffinase Water. Alkaline.}-+1.67}) ++ (stronger. ) y a Acid (weaker). sf +1.05)+++ es a “ (stronger).| Neutral. |+1.05/++-+- Saline solution. oe “ (weaker). | Acid +1.90} = i ‘ es “ (stronger).| “ +1.90) + » “A]SNOUVABIYUL “OL vs ‘wd BT'P + “BI pajiog “00 cz) ¢__,, SIZ I ‘avy : OG LG 9 66 ”? ”? ” G69 9G ”? ge Sy OST “A][VOUO} Lod BAQUI “OL WS [ZF] ZZ “Qoy] “00 “OUIN[OA *(O1) ssouTyeI + (BijasBUuyyes 10 (O01) osOUYsBI oyeqd = aiqqey jo uorpBaysTUTMI py aug ay) fo unay ay) fo sanog buryjydg-asoujoy ay “spqqvy o7ut hjjosajwadog pajoaluy ‘asourfoy fo worynzyUyy” 24L ‘A ATAVL 133 S. Kuriyama and L. B. Mendel = 08° 6+ = G8°6+ GS'6+ os 08 6+ =5 08 6+ 08'6+ = 08 6+ 06° 6+ g ” *A[SNOUSA BI -ul ‘Ol ‘wd QY'g + “BI ‘09 GE] Z ‘Ady GG») “A[SNOUDABIYUL “OL ‘BY E'S ‘wus 22°§ + “BI poylog *99 Gg) IZ ,, JOU 0 ” €T 5; “A[SNOUDABIY -UT ‘Ol ‘US OEP + “BI “90 GZ ZI ABIN PIT GG» 8 » 9 ” “AJsnowusaAsrjur ,, ,, 22°! G AB THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. l 134 Physiological Behavior of Raffinose TABLE VI. The Inversion of Raffinose in the Urine by Sulfuric Acid. | Rabbit: ¢ 2.) cceteete + Sac. cheer Te re MW OEE Lie Urine sample. | MBC. . ck io ee save bee Mar. 5. Mar. 5. Mar. 12. Mar. 21. Before inver- | Rotation, V°...........| +15.30) +11.40] +9.00 | +-10.50 sion. Reductionas- see _ _ _ — After partial | Rotation, V°...........|.+ 7.50) + 5.30} +4.61 | + 5.00 inversion. | Reduction............. +++) +4+ )/ +44 )/ +44 After total) Rotation, V°........... + 3.60; + 2.30) +2.15 | + 2.10 inversion. Reduction... 4.0 eeeee +++ )/++4+ |4+4++4+ /)4++4++ to the circulation, are not eliminated quantitatively through the kidneys. Heilner ascribed the loss of part of the sucrose parenter- ally administered to the appearance of the corresponding enzyme in the serum (28). Abderhalden considered that sucrase which is called forth by parenteral injection of sucrose is probably mobilized from a special part of the body where this enzyme is physiologically produced; namely, from the mucous membrane of the small intestine (29). If these hypotheses are true, the extent of the recovery of sucrose injected parenterally may differ from those of other sugars, for which no corresponding enzymes exist in the organism. Inulase and raffinase have never been found in the animal body (20, 30, 31). Injecting 2.8 and 2.2 gm. of inulin intraperitoneally, Mendel and Mitchell recovered 2.2 and 1.43 gm. respectively of it in the urine (32). Being of a colloidal character, inulin may not be able to pass through the kidneys as easily as erystalloid sugars. From this viewpoint, the parenteral injection of raffinose, which is of a crystalloid char- acter and finds no corresponding enzyme in the animal body, is of interest. Though Weinland and Abderhalden and Kapfberger claimed to find some special enzymes in the serum after parenteral administration of sucrose and lactose, they failed to obtain a comparable phenomenon after introduction of inulin (83, 34). In a few experiments, therefore, we examined the utilization of raffinose parenterally administered and the raffinose-splitting power of the serum of the same animals. In Kuriyama’s pre- vious experiments, an injection of sucrase, followed a few minutes S. Kuriyama and L. B. Mendel 135 later by sucrose injection, markedly increased the utilization of that sugar (27). Experiments of the same kind were performed with raffinose. Methods.—Full grown rabbits were used. They were fed on oats and corn, greens being added from time to time. A 10 per cent raffinose solu- tion, sterilized by boiling, was injected either into the peritoneal cavity or an ear vein. When raffinase and raffinose were injected successively, raffinase was always injected first, the interval between the two injections being 10 minutes. In control experiments, boiled enzyme solution was used. The serum examination for raffinase was performed as described on pp. 130 and 131. To activate the enzyme, i cc. of 0.075 Nn CH;COOH was sometimes added to the raffinose-serum mixture. Raffinose in the urine was determined polarimetrically after removing the disturbing substances (25 ce. of urine + 10 cc. of saturated mercuric acetate solution). In order to ascertain the nature of the dextrorotatory and non-reducing substance in the urine, the specimens were heated with sulfuric acid. For partial inversion (levulose + melibiose) the urine, mixed with sulfuric acid so that it contained 3.3 per cent sulfuric acid, was heated for 30 minutes at 75°C. For total inversion (levulose + giucose + galactose) the urine, acidified so that the mixture contained 1.2 per cent sulfuric acid, was heated for 64 hours at 100°C. The results are shown in Tables V and VI. Of the raffinose, administered intravenously or intraperitone- ally into rabbits in doses of 1.96 to 2.95 gm. per kilo of body weight, 88.4 per cent (as an average of six injections of raffinose alone) was recovered in the urine. The urine was always acid and contained neither reducing substance nor protein. After partial inversion of raffinose in the urine, the dextrorotation decreased to about one-half of the original degree; after total inversion to about one-fifth. These figures coincide with the properties of raffinose (35) and show that raffinose existed un- changed ‘in the urine. When raffinase and raffinose were in- jected successively, no better utilization was called forth than when raffinose alone was injected. The difference between these experiments and the more favorable utilization of sucrose + sucrase (27) may be due to an insufficient amount of raffinase or the immediate disappearance of this enzyme from the cir- culation. The serum, taken from time to time, possessed no raffinose-splitting power, even in cases where the mixture was slightly acidified to facilitate the enzyme action. 2 oie a +++ = > ~ “ls = S = PROV, PRY LLANE ap ESy. “4398 mens) 4F98 “4F08 vL Sy 62S ISG L V9Or L916 9° 666 b COG - [8 GP 6°95 GL ‘wis ¢ ‘uleseo “wis “usd “us “us ¢ ‘aso1oNg]z ‘aso1ong) G ‘asotong) ¢ ‘ulesED LE 0€ 8§ &P 612 ogl g0g 84 x ‘XI INIA ‘IIA jo1juoy Lof Buripaag asosongy = {ipa Sacae BON il) SPIO JOAN -*(yno poysea) Aydurg) “90g “IOS oo I 76 0 aBLL 9°FL v 99 oer OY TL ws ¢ “ws ¢ ‘ulasvo =| ‘UIesBO “ud ¢ “us ¢ ‘osou ‘osou ou N Soeur mechs % SP GG GE 08% 002 94% ‘IA “A “AI ‘spwawmrad xy ‘poly, i.mjuawmyy oy) ur asourfoy fo avy ay, ‘TIA WTEViL uae “UOISIOA “ul 10usV woTjponpey = aa ‘UOISIOA |" “ur as0y0 aes : Pa -U104s JO 1} §Juse}u0H poy ‘poy fos WOTjovOY Pig “WJOS TUG ott souvivoddy “"guao wad) | 9081], | “OOBI], pEgEUeD sees But | uwas0o04]4) “TOATT g9 Tg "wB Yysta 2 a "US G “ud ¢ ‘osou ‘osou -ey “Ia fee ee Surysvy 19}JB poo] LP 9% a aleve isl ©, ofa: biel Chexehal(s A eReane wb ‘qsvy Avp ¢ 10978 WyS1aMm Apo jo sso'T 908 Iz do Vieng ‘qySiom Apog — “ger OFICM I yo fee + ON ‘Burpaay asourfoy iazfp savy ay} Ur woyDUog Uabooh)5 = +4 ++ - — | ttt | ++ | 444+ | +++ [+44 | uorszos -Ul 10}JV morenped *(oulin = - ++ -- - — a - ++ + — | wors10a ree — SAS a 0 2 ee Se. pee | es ee ean nag sPLOV SRO: SPIO aBIOV. RAO: OV; POV.) TRANON | “TEIMIN Sale apy = = UOT oRey ; $009] "sa00] “sa00] S000] "sa00] "sa00] pIBy] “Soooy ON pie PIVH{|'S99J ON |['sao0J ON] AAOZVAA|'Sod9f ON] Aroze AA] AtozwA\| °° °° ooUBIvdd a 7 = a: = LOH Slory stairs testa ch ate |e UOIsaBA, -UL 109SV uoTpoNpeyy = = = a ae at, el + Slectecta late te at = “WOISIOA “OUTJSO}UI SULSLOFO asie] jo bs ‘phoy ‘[BIININ | ‘[BANEN | ‘[BIyMeN |'[VyyneN | ‘ploy ‘ploy ‘ploy [Teen | [eayMeN foc uOTyOVEYy] szUe}UOD: 7 *(qno0 poysea) SEAS ag PLS) me ers) Hes SHOR PM Pa ‘PI | “PMY RE ee oounrvoddy = = = = = t+ [tte | +4 | +4 | 4+ | cores eu tera) . uoryonpeyy — = = = = = — —_ = — “UOISIOA -ut}soqut SS a a SSeS eae [[eus jo sju9qU0y ‘ploy ‘ploy | [Bayne ‘ploy | [eyyney | “ploy “ploy ‘ploy |[eaynen |peagnenN foo uMOTIOwEY “(4no peyse a) PINT PMA SUL Te BONY 3) PIM Pee Pe PIL ‘pin | °° eouvreaddy 138 Physiological Behavior of Raffinose Glycogen Formation in the Liver after Raffinose Feeding; the Fate of Raffinose in the Alimentary Tract. Judging from Kiilz’s and Sandmeyer’s reports (10, 12), the utilization of raffinose in the alimentary tract seems to be rather difficult. Haldsz’s experiments suggest that part of it can be inverted in the rectum (14). Some experiments in this direction were performed on white rats. Methods.—Full grown white rats fasted for 72 hours. This period was so arranged that it ended at9 p.m. A certain amount of selected food was then left in the cage over night. At 9 the next morning the animal was decapitated and the glycogen content of the liver was determined by Pfliiger’s method. The glycogen was hydrolyzed with hydrochloric acid and the reducing sugar produced was determined by Allihn’s gravimetric method. The contents of the stomach and small and large intestine were collected separately. Each portion was diluted with water to about 10 ce. The filtered contents were tested for reducing power both before and after hydrolysis with 3.3 per cent sulfuric acid for 30 minutes at 75°C. The urine and feces were collected together. They were unavoidably contami- nated with particles of the food. The food was given in the form of paste. As raffinose causes severe diarrhea, casein was mixed with it in some experi- ments, in the hope of decreasing the purgative action of the raffinose. For control experiments, sucrose and casein or eithér sucrose or casein alone were given. From Table VII it will be seen that after raffinose feeding, glycogen was not formed in the liver to any noticeable extent. Though raffinose was present in the stomach and small intestine, no reducing substance was found. It may be that the cleavage products of raffinose were very easily absorbed and could not be detected in the stomach and small intestine. Judging from the absence of glycogen formation in the liver and the failure to find raffinase in the intestinal mucous membrane of vertebrate animals, it is very likely that raffinose was not inverted, at least to any noteworthy amount, in the stomach and small intestine. In the large intestine, however, some reducing substances were found, probably owing to the bacterial action. This point will be dis- cussed later. Why did not the reducing substances, produced in the large intestine, become a source of glycogen in the liver? Was the amount of the cleavage products of raffinose insufficient or were the monosaccharides further destroyed in the intestinal canal, S. Kuriyama and L. B. Mendel 139 without entering into the circulation? Studying the fate of dextrose in the large intestine and in a feces-dextrose mixture in vitro, Bingel concluded that the sugar was absorbed very slowly in the large intestine and the amount of the sugar destroyed by bacteria was as great as that absorbed (36). Introducing a large amount of sugars (mono-, di-, and trisaccharides) into the large intestine, Haldsz found that a noteworthy amount of the sugar was absorbed in 5 to 6 hours, the amount destroyed by bacteria being rather negligible. For the absence of the sugar in the urine in his experiments, the slow absorption in the large intestine was considered to be one reason (14). In our experi- ments, diarrhea caused by raffinose seems also to have been a factor explaining absence of glycogen from the liver. In control experiments with sucrose feeding, the food was easily inverted and became a source of glycogen in the liver. When casein was added to raffinose, some glycogen formation was observed. In these cases, raffinose may not have passed so quickly through the alimentary tract as when this sugar was given alone, and consequently may by chance have been inverted. According to Bendix’ and Stookey’s experiments, casein alone can become a source of some glycogen in the liver (37, 38). Rat VII. in our experiments seems to confirm their results. It is there- fore probable that the liver glycogen obtained after raffinose- casein feeding is at least for the most part due to casein itself. The behavior of raffinose in the alimentary tract is somewhat ‘similar to that of inulin. The glycogen formation in the liver after inulin feeding to fasting rabbits is slight or uncertain (39, 40). After inulin feeding, Miura found a reducing substance in the stomach and large intestine, also sometimes in the small intestine. The Fate of Raffinose Administered into Intestinal Loops of Dogs. In the preceding rat experiments raffinose was inverted in the large gut, but.not in the small intestine. To examine this point further, we studied the fate of the sugar, introduced into intestinal loops of dogs. Methods.—Full grown dogs fasted for 48 hours before operation. The urine contained neither protein nor sugar. As anesthetics, urethane -So}UL jo doo] 1aMo7 -So}Ul jo doo, sodd “oury} |[Bus “our} [eurs Be oe oF ABA. Set eae eh tee gts Sys Serle auozs¢ £76 $99 " 9°%8 page a pala ddoon esoumye 68°T olen 69°T "UB He SPRY Gh Fa og ping woryestiy = = ae worjonpey 2 -[eaqnaN ‘[RANON yeanonp tect “-UoTjoROYy S 18Z 6FZ 80z *) folds! etieie ie) .0lfe) ie (ONO Fel viele (0! et © *299 ‘QuM]O A e 00°Z 00°Z 00°% oss squh DPaonpodzul ssouyper JO JUNO oo eG og 08 So naWfel¥ wo he:(she) o: s\hal eps wigs) o, ella © B) Ss) ye iele TA) ‘dooy] jo yysuey] iG as ee ee Pe ee 2 = = % Stee le ie eas celeriiels seer ae auOZes() i) ae 2 '¥6 zeae nae ee ‘poldAOIAL BSOULJEY, 2 is ie eee We eae es Ona t ‘pmp wows] 3 = = — uoTJONpayy] - * Ro ee = “[BINON -[B.1jNIN yoanany feet WOTJVOY fal Ch 06Z GGZ Se (eerie ‘s)\eliel sits ie. © \ople! si ehele: 99 ‘auINnyO A — ° by 00°Z 00°Z OOS2e ee ICR sss wb ‘psonpodyut osougyet jo JunoWy Sp Ch 87 z9 ages sic? ais G hs ees slo BOOT Ome eg] = — “royye ++ ‘rayyo + ‘royyo + ss (“WS OF) OUBY IOI |(“UIS Fg) OUBYJoI |("UIs C'9g) JURY IOI) “OTJOYASOUYy — — jae 26 FIT 66 | BY “VySIOM Apog 30d © Ul 6 IL BT eee ttre sbeceeee eer eee ees x08 put ‘ON = \ ‘sboq fo sdooT Jourysajuy ojur paonpowuy ‘asowfoy fo 20 2Y,], - : Re ‘WIA AT&L \ by 141 S. Kuriyama and L. B. Mendel “UOTPONPEL PUL UOT}IOL UIIMJoq BUIIIYIP oY} WOIJ po}VTNo[Bo SVM osoUTeA JO JUNOUTe oY "9S0JOV[VS PUB ‘ASO[NAY] ‘osoonys JO S}uNOW’ [eNbo Jo o1n4xIW B oq 0} poPBNo[B SBA OOULYSGNS Sutonpor Jo JUNOUI OIL], y — a + aigubaltalcantarial iehejeliegiakieie (elieiens hele) age) ai cclne™ citaielia (site lepcaitere? obeue QUOZBSC) £0°0 8.0 0 sorts quB TOTVEIOI WIOI] poyV[NoTBo ‘asougzery 0 0 worljponpad Pf ZC re \ OCONEE ONCE O ON OnG Mane ODGOr tec Lon Ord O tM Ge .0 osc W 5 -wb UOTPRIOI UOT 02°2\ f ‘Qsoon[S sv po}E[No[vo ‘oouvysqns Suronpoey “ploy ‘ploy ‘poy [wn CCC C0) cL Tye ce z9 6Z Cicy ORDO ONDAO. CHORE ORC OCA ONO licech crnich Aidlertcr Go Bor 99 ‘QUINOA 0 9F 019 Gay “uaa aa ‘pore 660 CGarl Te SP ECU i) -AOVOL OSOULJRYy. |. 88°0 1g°0 It '0 “ss quB ooueysqns Surnpoy,. DIOR US ee ans “poy ‘pLboy ‘ploy Sr ob @! 04 4 wvie (>, e's igus el eae 0) akele vata; UOTLOVOY, 812 SSS 1&zZ a of elipaliby @ (eo\e/ 0) 'e (ane! evel @renene “90 ‘QuIN]O A 00°% 00°% 00°% ceeeesese sess sah HaonposjUt ssouypes Jo JUNOUTY ZI CL eT CO Oro eta DEO CNEL OATS. ont WS Onn uciare Oe. eLr wma ‘dooy jo Yyysuey Ub, A “ouTIy) ‘OUT}S9} UI asiey, jo dooT 142 Physiological Behavior of Raffinose (applied subcutaneously) and ether were used. Two loops were established on the small intestine: the upper end of the first loop was at the beginning of the duodenum, leaving the opening of the bile and pancreatic duct with in the loop; the lower end of the second loop of the small intestine was just above the ileocecal valve. The loop on the large intestine began a few centimeters from the cecum. Glass cannulas were inserted into both ends of the loops. 20 ce. of a 10 per cent raffinose solution were intro- duced from the upper end of each loop and the intestine was gently mas- saged, so that the raffinose solution was distributed through the whole length of the loops. Every precaution was taken not to damage the in- testinal mucous membrane and the blood vessels providing the alimentary tract. After closing the abdominal cavity, the animal was kept warm and under a slight anesthesia for 2 hours. The contents of the loops were then washed out with a warmed physiological saline solution. The fluid discharging from the lower end of the loop was mixed thoroughly with toluene. After killing the animal by bleeding, all the loops were again washed out. The remaining parts of the intestinal canal (the middle part of the small intestine and the beginning and end parts of the large intestine) were also washed out. The urine was collected from the bladder. The proteins in the irrigation fluid were removed by colloidal iron solution (25 ec. of fluid + 5 or 10 ce. of colloidal iron solution). In the water clear filtrate the reducing substance was determined by Allihn’s gravimetric method. The fluids, obtained by washing the remaining parts of the intestine and the second washing of the loops, showed neither reducing power nor rotation. Before determining the rotation of the urine, it was clarified with mercuric acetate. As Table VIII clearly shows, raffinose was very slowly ab- sorbed by the small intestine, especially by the upper part. No evidence of its inversion was observed. In order to determine the nature of the dextrorotatory substance, obtained from the small intestine, the protein-free irrigation fluid was heated with sulfuric acid as described previously. After partial ‘inversion the rotation decreased to nearly half of the original degree, and after total inversion to nearly one-fifth (Table IX). This cor- responds to the properties of raffinose (85). The slow absorption of raffinose in the small intestine in our experiments is noticeable. There have been many investigations of the absorption of various kinds of sugar from the small intestine (41, 42, 13, 48). For full grown dogs Ré6hmann and Nagano reported that among the disaccharides sucrose was absorbed the most quickly, lactose the most slowly. No evidence of the in- version of lactose was observed, while sucrose and maltose were partly inverted before their absorption (42). Comparing the S. Kuriyama and L. B. Mendel 1438 TABLE IX. The Inversion of Raffinose, Recovered from the Intestinal Loops, by Sulfuric Acid. Doze) sew a tee: it ge II. Sample. — i Lop of mansion | Gun?) Ha | aa | Kane | pe [gue Before in-| Rotation, V°......./+3.85/+4.10/+3.22/+2.28/+4.00/+3.40 version. Reduction: ...4... — _ — — — _ After partial | Rotation, V°.......|+1.70/+1.80/+1.50/+1.19/+2.04/+1.69 inversion. | Reduction......... +++]+++|+++]+++])++4+]/4+++4+ After total | Rotation, V°......./+0.80/+0.80 inversion. | Reduction......... +++/++-+ rapidity of absorption of various kinds of sugar “(20 ec. of 25 per cent solution were introduced) in a loop 1 meter long of the small intestine of rabbits, Hédon found that raffinose disappeared most slowly. 10.2 per cent of the amount introduced was absorbed in 2 hours. The rapidity of absorption increased gradually in the following order: lactose, maltose, sucrose, levulose, ete. (13). The irrigation fluid from the loop of the large intestine contained a noteworthy amount of reducing substances. In the osazone test, yellow crystals appeared while heating, and increased mark- edly after cooling.. Microscopically, most of the crystals had the appearance of phenylglucosazone but some the shape of phenyl- lactosazone. This suggests the production of a small amount of melibiose (35) together with other cleavage products. As the exact nature of the reducing substances was not clear, the amount was calculated to be a mixture of equal portions of glucose, levu- lose, and galactose. The raffinose was calculated from the dif- ference between rotation and reduction. In the urine of Dog I, a large amount of glucose was found, contrary to the other two experiments. The amount of reducing substances produced in the large intestine was not enough to explain this glycosuria. The operation and anesthesia may have been the cause. In the urines of Dogs II and III, a small amount of a dextrorotatory and non-reducing substance was found. After acid hydrolysis, it showed reducing power. This substance was calculated to be raffinose. The fact that raffinose can pass 144 Physiological Behavior of Raffinose unchanged through the intestinal wall and reappear in the urine was also demonstrated by other investigators (12, 44). The Presence of Raffinase in the Feces. The inversion of raffinose in the large intestine seems to be due to bacteria. In reality, numerous .kinds of bacteria have the power to invert raffnose. Investigating the properties of seventy- seven strains of colon bacilli, isolated from polluted water, feces, urine, and the animal body, Kligler found that forty-one strains of them had the ability to attack raffinose (45). In the study of 350 strains of streptococci, isolated from human, equine, and bovine feces, Fuller and Armstrong found that in human feces none, in horse feces 12 per cent, and in cow feces 73 per cent of the strains examined attacked raffinose (46). It is probable that such bacteria act favorably to the host, by decomposing raffinose into utilizable forms. A comparable phenomenon has also been reported for cellulose and intestinal bacteria (47). Though raf- finose is usually first decomposed into levulose and melibiose, it can also be inverted into galactose and sucrose (48). It is yet unknown whether the latter mode of inversion can also occur in the intestinal canal. When the fluids, obtained from the loops of the large intestine and mixed thoroughly with plenty of toluene, were kept for 48 hours at room temperature, their rotatory power decreased mark- edly. This was not the case, however, with the fluids from the small intestine (Table X). TABLE X.’ The Examination of Rotation of the Irrigation Fluids Both Immediately and after Keeping 48 Hours at Room Temperature with Toluene Addition. l Fluid obtained from Examination. Dog II. Dog III. a VP. Vass Upper loop of small intes- | Immediately............. +3.28 | +4.00 tine. Atter48thrsigs).: . asses |) 23228 +4.00 Lower loop of small in-| Immediately............. +2 \28 +3.40 testine::.2...../.........| After 4eubrs:).-. 2. ......| -E2 SOR Meeseae Loop of large intestine......| Immediately............. +2.30 +1.70 ; | Aftent4S@hngte set. 7. 3 +0.40 | +0.60 ie —. S. Kuriyama and L. B. Mendel 145 Raffinose or its cleavage products in the irrigation fluid of the large intestine, therefore, must have been attacked either by active bacteria or isolated enzyme. Though Ury stated that toluene addition was enough to eliminate the bacterial action in diastase determination in the feces (49), our bacteriological examination showed that our materials, obtained from the large intestine and mixed thoroughly with toluene, were not sterile. To make a sterile fecal extract, therefore, fresh dog and rabbit feces, obtained from an animal room, were ground with about ten volumes of water and mixed with both toluene and chloro- form. 7 to 8 hours later the mixture was filtered through cloth. The slightly acid filtrate was used for the digestion test. The procedure was the same as described previously. Cultivating from these samples, no bacterial growth was observed on agar- agar media in 48 hours. For control, the feces extract was previously boiled or mixed with mercuric chloride. The results with these extracts of the feces of dogs and rabbits indicate a small amount of raffinase, probably originating from bacteria. The feces extract can be more properly sterilized with a porcelain filter (50, 51). Our investigation of raffinase in the feces is not yet concluded. SUMMARY. A hydrogen ion concentration of pH 3.8—5.4 is the most favor- able for the activity of raffinase (from yeast). ~ Saliva (human), bile (rabbit), pancreas (dog), liver (dog and rabbit), and mucous membrane of small and large intestine (dog and rabbit) do not contain raffinase. The gastric juice may invert raffinose under suitable conditions. Blood serum (rabbit) is not a favorable medium for the activity of yeast raffinase; yet a sufficient quantity of the enzyme can exert activity. When yeast raffinase is injected intravenously into rabbits, the activity of the enzyme can be maintaimed in the serum for a short time only. When raffinose was injected parenterally into rabbits in doses of about 2 to 3 gm. per kilo of body weight, 88 per cent of the amount administered was recovered in the urine. The serum of the same animals failed to show raffinose-splitting power. When raffinase and raffinose were injected successively into 146 Physiological Behavior of Raffinose the circulation, no better utilization of the sugar was called forth than when raffinose alone was injected. No noteworthy glycogen formation in the liver was found after feeding raffinose to fasting white rats. The sugar was scarcely inverted in the stomach and small intestine. It was, however, changed in the large intestine. When raffinose was administered directly into loops of the small intestine of dogs, most of it was recovered after 2 hours without evidence of its inversion. In a loop of the large intestine, however, raffinose was easily inverted. This was probably due to bacteria. Part of raffinose can pass unchanged through the intestinal wall and reappear in the urine. In our experiments, the animals (rats and dogs) were previously fasted for a few days. This might have been unfavorable for the growth of intestinal bacteria, especially for raffinose-attacking bacteria. Under ordinary conditions, therefore, the inversion of raffinose may occur more extensively and perhaps even in the small intestine. The sterilized feces of dogs and rabbits seems to contain a small amount of raffinase, probably of bacterial origin. Raffinose is devoid of food vaiue until after its inversion. It may be that raffinose-digesting bacteria occur more frequently in the large intestine of species which consume foods containing raffinose and thus render the physiological utilization more prob- able for them. : BIBLIOGRAPHY. 1. Bierry and Giaja, Compt. rend. Soc. biol., 1906, xi, 485. 2. Barthet, G., and Bierry, H., Compt. rend. Soc. biol., 1908, lxiv, 651. 3. Straus, J., Z. Biol., 1908-09, lii, 95. 4. Bierry, Biochem. Z., 1912, xliv, 426. 5. Pautz, W., and Vogel, J., Z. Biol., 1895, xxxii, 304. 6. Fischer, E., and Niebel, W., Sitzwngsber. preus. Akad. Wissensch., 1896, i, 73. 7. Abderhalden, E., and Brahm, C., Z. physiol. Chem., 1910, Ixiv, 429. 8. Voit, F., Deutsch. Arch. klin. Med., 1896-97, lviii, 523. 9. Magnus-Levy, A., Oppenheimer’s Handb. Biochem., Jena, 1911, iv 1, 315. 10. Kiilz, E., cited in Pfliiger (11). 11. Pfliiger, E., Arch. ges. Physiol., 1903, xevi, 201. 12. Sandmeyer, W., Z. Biol., 1895, xxxi, 12. 13. Hédon, M. E., Compt. rend. Soc. biol., 1900, lii, 41, 87. 14. Haldsz, A. v., Deutsch. Arch. klin. Med., 1909-10, xeviii, 433. S. Kuriyama and L. B. Mendel 147 . Sérensen, S. P. L., Biochem. Z., 1909, xxi, 201. . Henderson, L. J., and Palmer, W. W., J. Biol. Chem., 1912-13, xiii, 393. . Walpole, G. S., Biochem. J., 1914, viii, 628. . Hudson, C.S., J. Am. Chem. Soc., 1914, xxxvi, 1566. . Michaelis, L., and Davidsohn, H., Biochem. Z., 1911, xxxv, 386. . Chittenden, R. H., Am. J. Phystiol., 1898-99, ii, p. xvii. . Ferris, 8. J., and Lusk, G., Am. J. Physiol., 1898, i, 277. 22. . London, E. S., and Polowzowa, W. W., Z. physiol. Chem., 1908, lvi, 512. . Menten, M. L., and Crile, G. W., Am. J. Physiol., 1915, xxxviii, 225. . Mendel, L. B., and Kleiner, I. S., Am. J. Physiol., 1910, xxvi, 396. . Hogan, A. G., J. Biol.. Chem., 1914, xviii, 485. . Kuriyama, 8., J. Biol. Chem., 1916, xxv, 521. . Heilner, E.; Z. Biol., 1911, lvi, 75. . Abderhalden, E., Deutsch. med. Woch., 1914, xl, 268. . Bieri and Portier, Compt. rend. Soc. biol., 1900, lii, 423. . Richaud, A., Compt. rend. Soc. biol., 1900, lii, 416. . Mendel, L. B., and Mitchell, P. H., Am. J. Physiol., 1905, xiv, 239. . Weinland, E., Z. Biol., 1906, xlvii, 279. . Abderhalden, E., and Kapfberger, G., Z. physiol. Chem., 1910, |xix, 23. . Scheibler, C., and Mittelmeier, H., Ber. chem. Ges., 1889, xxii, 1678. . Bingel, A., Ther. Gegenw., 1905, xlvi, 436. . Bendix, E., Z. physiol. Chem., 1901, xxxii, 479. . Stookey, L. B., Am. J. Physiol., 1903, ix, 138. . Miura, K., Z. Biol., 1895, xxxii, 255. . Mendel, L. B., and Nakaseko, R., Am. J. Physiol., 1900-01, iv, 246. .. Nagano, J., Arch. ges. Physiol., 1902, xc, 389. . Réhmann, F., and Nagano, J., Arch. ges. Physiol., 1903, xev, 533. . Albertoni, P., Ergebn. Physiol., 1914, xiv, 431. . Neuberg, C., Der Harn, Berlin, 1911, i, 425. . Kligler, I. J., J. Infect. Dis., 1914, xv, 187. . Fuller, C. A., and Armstrong, V. A., J. Infect. Dis., 1913, xii, 442. . Luntz, N., Die Wissenschaften, 1913, 1, 7. . Neuberg, C., Biochem. Z., 1907, iii, 519. . Ury, H., Biochem. Z., 1910, xxiii, 153. . Brugsch, T., and Masuda, N., Z. exp. Path. u. Ther., 1910-11, viii, 617. . Hirayama, K., Z. exp. Path. wu. Ther., 1910-11, viii, 624. Lusk, G., Am. J. Physiol., 1903-04, x, p. xxi. — THE ROLE OF VITAMINES IN THE DIET.* By THOMAS B. OSBORNE anp LAFAYETTE B. MENDEL. WITH THE COOPERATION OF Epna L. Ferry AND ALFRED J. WAKEMAN. (From the Laboratory of the Connecticut Agricultural Experiment Station and the Sheffield Laboratory of Physiological Chemistry in Yale University, New Haven.) (Received for publication, May 9, 1917.) The feeding experiments first made by Hopkins! have led to widespread recognition of the importance of small quantities of hitherto unidentified substances—in addition to the protein, carbohydrate, salts, and fat—as essential components of a ration adequate for prolonged maintenance or growth. The develop- ment of our own views on this subject has been alluded to in a recent publication.2, Réhmann,? however, has taken vigorous exception to the vitamine hypothesis. He asserts that “accessory foodstuffs are not necessary for the continued maintenance of fully grown animals,’’? and believes that if the long familiar nutrients are suitable in quality and quantity, nothing further is essential in the ration. Thus he says: “The assumption that some unknown substances are indispensable for growth is a con- venient device for explaining experiments that result in failure— a device that becomes superfluous as soon as the experiment succeeds.’” * The expenses of this investigation were shared by the Connecticut Agricultural Experiment Station and the Carnegie Institution of Washing- ton, D. C. 1 Hopkins, F.G., J. Physiol., 1912, xliv, 425. 2 Osborne, T. B., and Mendel, L. B., Biochem. J., 1916, x, 534. 3 R6hmann, F., Ueber kiinstliche Ernghrung und Vitamine, Berlin, 1916. 4 «*Akzessorische’ Nahrungsstoffe sind mindestens zur dauernden Erhaltung ausgewachsener Tiere nicht notwendig”’ (p. 142). 5 “Tie Annahme von irgendwelchen unbekannten Stoffen, die fiir das Wachstum unentbehrlich sind, ist ein bequemes Mittel, um die fehlge- schlagénen Versuche zu erkliren, das iiberfliissig-wird, sobald der Versuch gelingt’’ (p. 42). 149 THE JOURNAL OF BIOL(C + MISTHY, VOL. XXX!, NO. 1 150 Vitamines As evidence against the indispensability of vitamines Réhmann cites a new series of experiments on the growth and maintenance of white mice fed with artificial mixtures of food supposedly free from these so called accessory substances. It will be unnecessary to review the successive steps in his extensive investigations, because yeast was used to impart a suitable texture to many of the food mixtures earlier employed by him. The efficiency of yeast as a source of vitamine, to which we can testify from our own experience, has long been recognized. To meet this criticism Réhmann used, in his latest experiments, starch digested by ‘‘purified’’ diastase, and leavened the mixtures with baking powder. We shall there- fore concentrate attention upon the crucial experiments in his newer series in which the diets had the following composition: 1. Hiihnereiweissnahrung.* 2. Kasein-Vitaminfreie Nahrung.** gm. gm. logy white: 25: .ot..0. 64540. 5 4a *Kalzose’?y poh fan ae hv ee 22 Bee white anon: 203%. sk oe eee 2 Casein=ironiy ltr JN see 2 Potato starch (predigested).... 20 Potato starch (predigested).... 20 Potato starch (raw). =... +a ce 25 Potato starch (raw). sce eee 25 Wheat starch): :..:.)..2°a% ster 90 Wheat.starch... --2:22. 2.005 90 DExtTOSeMreteys.< Ses acca neteee ae 5 Dextrose: ov. vat cee Pee ee 20: Marrarine ts. 25%). ohn dese ie 10 Lard: :!. Ae eee 24 Daliises Mee sete, ai wrens see 3 RIES. Ai ous eae eee 6 Bakinespowder:2a.c sor ee 5 Baking powder...2-4...4- cae 5 * Pages 27-28. ** Page 43. Réhmann states that with Diet 1 five young mice having an average initial weight of 6 gm. were successfully reared. In comparison with most data on the feeding of growing animals the protein (4.4 to 10 per cent) was surprisingly low. Addition of casein gave still more successful results, for a third generation was obtained. It should be noted that the egg white used in these experiments was not purified, but ‘‘albumen ovi siccum dried in vacuum,’’ of which nothing is known respecting its effect on growth. The use of diastase is also open to criticism, since McCollum® has shown that extracts of some of the cereal grains possess decided growth-promoting properties. In subsequent experiments with Diet 2 R6hmann met the possible objection to the presence of vitamine by the statement that “the food stuffs were either treated with alcohol without warming, or were heated 3 hours at 120-150°” (p. 42). The ‘‘Kalzose’’ used was a commercial preparation of calcium-casein admittedly still containing some milk sugar, and doubtless other pro- ducts present in milk. In this series 0.2 gm. of ‘“‘Merck’s purified diastase,’’ washed with alcohol, also was used to digest 20 gm. of starch. 6 McCollum, E. V., and Davis, M., J. Biol. Chem., 1915; xxiii, 181, 231. T. B. Osborne and L. B. Mendel LSt With Diet 2 Réhmann states that he has succeeded in rearing young mice and subsequently maintaining them. His protocols, however, show that in some cases some of the mice were consumed by the others, and suc- cess was frequently induced not only by added alcoholic extracts of yeast or by small quantities of milk, but particularly with ‘“‘Filirateiwetss,”’ a product from milk ‘‘principally composed of proteins which remain in solution after the precipitation of the casein’’ (p. 42). According to his account of the method of preparation this product may have contained some or all of the other constituents of milk among which are those proved to be especially efficient in promoting growth. Presumably relying on the earlier statements regarding the loss of anti- scorbutic power in various natural substances which have been heated, Réhmann considered the application of heat or extraction with cold alco- hol to be sufficient to exclude the presence of vitamines. The effect of heat has been discussed by Funk.? The destruction by heat of both the fat-soluble and water-soluble vitamines present in milk is improbable. Furthermore, considerable difference of opinion regarding the relation of heated milk to infantile scorbutus still exists. Aleohol was doubtless used by Rohmann in view of the state- ment of various investigators since Hopkins that this substance is a solvent for the vitamine of yeast, milk, wheat embryo, ete. Whether Ro6hmann used absolute alcohol or alcohol containing more or less water to remove vitamines is not apparent from his published descriptions. Our own experience discussed below has shown that absolute alcohol is by no means an adequate sol- vent for the effective nutrition-promoting accessory substances in yeast. We have not yet learned how much water the alcohol must contain in order to extract these substances. It is doubtful therefore whether either of the methods selected by Réhmann ean be depended upon to exclude all traces of vitamines. The thesis that a successful, 7.e., positive, experiment in nutri- tion is far more significant than a negative one is doubtless valid. On the other hand, in dealing with substances which, like the alleged vitamines, are potent in surprisingly small amounts, the burden of proof with respect to the complete absence of effective substances so widely distributed among the natural foodstuffs falls on those who deny the need for them. Without the use of some water-soluble accessory substances such as has been dem- 7 Funk, C., Ergebn. Physiol., 1913, xiii, 125. 8 Osborne and Mendel, J. Biol. Chem., 1915, xx, 381. MeCollum and Davis, ibid., 1915, xxiii, 247. 152 Vitamines onstrated to be present in milk,’ “protein-free milk,’’® yeast,’ the extract of embryos of certain seeds,® animal tissue extracts," or doubtless in nearly all the commonly used animal or vegetable foods in their natural state, we, in common with many other investigators, have failed to induce growth or even maintenance in such a large proportion of our trials that the importance of the vitamine hypothesis has been forced upon us. Nevertheless, occasionally, though very infrequently, an animal has for a time grown well or been maintained for exceptionally long periods on a ration of isolated food substances supposedly free from all but traces of water-soluble vitamine, upon which an overwhelming majority of the same species promptly fails: to thrive.” For example, Rat 3030 on a diet of lactalbumin, “artificial protein- free milk,” starch, lard, and butter fat, grew from 286 gm. to 372 gm. and showed no signs of nutritive failure after 392 days. This food contained purified lactose. It may still have con- tained a trace of vitamine, though we regard this as an unlikely explanation of the outcome of the preceding experiment, in view of our numerous failures with artificial foods containing lactose from the same stock. From these very infrequent successes it would seem as if such a diet were adequate for maintenance provided that an animal can be induced to consume enough of it. One of the immediate effects of the addition of ‘‘protein-free milk,” small quantities of yeast, etc., to such a ration is an improved appetite attended by increase of weight. This is well exemplified in Chart I in which the food intakes on the different diets are arranged for comparison. It is now generally believed that if a young rat is fed on a diet consisting of purified protein, carbohydrate, lard, a suitable mixture of inorganic salts, along with some water-soluble vitamine, ° Osborne and Mendel, Carnegie Institution of Washington, Publication No. 156, pt. ii, 1911; Z. physiol. Chem., 1912, 1xxx, 356; J. Biol. Chem., 1913, RVamo ld 10 Funk, J. Biol. Chem., 1916, xxvii, 1. Funk, C., and Macallum, A. B., ibid., 1915, xxiii, 413; 1916, xxvii, 51. Hopkins, J. Physiol., 1912, xliv, 425. 11 Eddy, W. H., J. Biol. Chem., 1916, xxvii, 113. 12 Osborne and Mendel, J. Biol. Chem., 1912-13, xiii, 233; 1913, xv 311. McCollum and Davis, ibid., 1913, xv, 167. T. B. Osborne and L. B. Mendel 153 it will fail to complete its growth. Replacing a part of the lard by butter fat,!4 egg yolk fat, beef fat,’ or cod liver oil!” renders the ration adequate for promoting normal growth and reproduc- tion provided that the protein of the ration is also suitable. Experience has therefore demonstrated that adequate dietaries require at least two formerly unappreciated components. Hopkins, who first thus added yeast to a diet of purified food stuffs, reported that very smal! quantities of a protein-free alco- holie extract or even of the ether-soluble fraction of the alcoholic extract of yeast markedly accelerated the growth of rats. Hopkins writes: “There is some indication from my experiments that the optimum supply of the substances which induce growth is soon reached; but any attempt to ascertain the nature of their action by noting the relation be- tween their concentration and their effects would call for extensive experi- mentation which it would seem better to leave until definite substances have been isolated’’ (p. 440). Recently Funk and Macallum!° have shown that yeast added to artificial diets for growing rats accelerates their growth. Funk!® states: ; “Our experiments show that the quantity of vitamines necessary for stimulating growth in rats is by no means small. If yeast is added to the diet to the extent of 1 per cent the rats grow for a short time, after which they begin to decline. Experiments which are not recorded in this paper have shown that at least 3 per cent of yeast is necessary to insure a satis- “factory growth in rats. Still further experiments have shown that yeast has more effect in promoting growth than an addition of a few cc. of milk, as used by Hopkins. Yeast can be regarded as a complete food by itself. It was therefore necessary to ascertain whether the good results obtained with this addition are not merely due to a correction of the nutritive value 13 Osborne and Mendel, Z. physiol. Chem., 1912, Ixxx, 356; J. Biol. Chem., 1913, xv, 311. McCollum and Davis, ibid., 1913, xiv, p. xl; 1913, xv, 167. Hopkins, J. Physiol., 1912, xliv, 425. 14 Osborne and Mendel, J. Biol. Chem., 1913-14, xvi, 423; McCollum and Davis, ibid., 1913, xv, 167. 15 MacArthur, C. G., and Luckett, C. L., J. Biol. Chem., 1915, xx, 161. Osborne and Mendel, zbid., 1914, xvii, 401; 1915, xx, 379. McCollum and Davis, ibid., 1913, xv, 167. 16 Osborne and Mendel, J. Biol. Chem., 1915, xx, 379. 17 Osborne and Mendel, J. Biol. Chem., 1914, xvii, 401. 154 Vitamines of the protein used (in this case, casein) or to the presence in it of nucleic acid. Consequently a diet was prepared in which the total casein nitrogen was substituted by yeast nitrogen. The results obtained were not as satisfactory as when yeast was used in smaller amounts for its vitamine content only, and not for nutritive value.”’ Yeast has frequently been ‘used to cure polyneuritis induced in birds by a diet of polished rice. Whether or not the anti- neuritic component is identical with the growth-promoting one is a question which as yet has received no definite answer, al- though Funk and Macallum think that their results ‘indicate that the growth-promoting substance is analogous to and possibly identical with the beri-beri vitamine” but that ‘considerably larger quantities of vitamines are necessary for stimulating growth than for curing beri-beri.’’!8 Seidell’® reports that Lloyd’s reagent completely removes from autolyzed brewers’ yeast a substance capable of preventing polyneuritis in pigeons fed on polished rice. On the other hand Gibson and Concepcién” found that puppies and pigs even while growing develop symptoms of peripheral nerve degeneration when milk, whether fresh or heated, forms their sole diet. Our early experiments with “artificial protein-free milk’ indicated its marked inferiority to our ‘‘natural protein-free milk” when fed to growing rats, although the product contained like proportions of everything known to be present in the natural product. The fact that in our earlier experiments rats occagion- ally grew well on the “artificial” diet led us at first to suspect that some essential inorganic element was present as an impurity in the chemicals which we then used. Our later experience has, however, inclined us to the conclusion that the inorganic content of all these salt mixtures was probably suitable, but that the dietary deficiency lay in the lack of some essential undetermined organic food factor. To render our “artificial” food mixtures as efficient for pro- 18 Funk and Macallum, J. Biol. Chem., 1916, xxvii, 63. 19 Seidell, A., Public Health Report No. 325, 1916. 20 Gibson, R. B., and Concepcién, I., Philippine J. Sc., B, 1916, xi, 119. 21 Osborne and Mendel, J. Biol. Chem., 1913, xv, 311. T. B. Osborne and L. B. Mendel 155 moting growth as the foods containing our ‘natural protein- free milk,”’ we have recently fed smal! quantities of dried brewers’ yeast either separately or incorporated in the artificial food mixtures. On a ration of purified casein, “artificial protein- free milk,” starch, lard, butter fat, and 1.5 per cent of dried yeast, rats of both sexes have grown from about 50 gm. body weight to maturity, and have even produced young. This is a smaller proportion of yeast than Funk considered necessary for normal growth. Adult rats have been maintained for more than 300 days. For some as yet unknown reason the majority of the rats grew normally when the protein used was casein, whereas they have usually failed when it was edestin, and almost invariably when lactalbumin, cotton seed globulin, cotton seed proteins, or squash seed globulin was fed. This result surprised us because all of these proteins had earlier led to normal growth when used in rations containing natural “protein-free milk.’ The failure to grow on the “artificial protein-free milk’’-yeast foods was especially unexpected in the case of lactalbumin (Chart II); for our former experience had demonstrated that even exception- ally small proportions of this protein promoted normal growth.” The fact that in all of our numerous experiments with this ~ lactalbumin-‘‘artificial protein-free milk” diet young rats failed to grow with an addition of 1.5 per cent of yeast, and all soon died unless a change was made in the diet, whereas mature rats were maintained over very long periods, suggests that the unknown ‘nitrogenous constituents of milk? may possibly supplement some hitherto unrecognized deficiency in this protein. Attempts to find an explanation for this unexpected result have not yet given us a clue. McCollum and Davis* consider that the nitrogen of ‘‘protein- free milk” has essentially the same nutritive value as that of the milk proteins; but this assumption does not help to explain our results obtained with lactalbumin, for in this case we are con- fronted by an apparent deficiency in the chemical make-up of this protein which may possibly be supplemented by the unknown 22 Osborne and Mendel, J. Biol. Chem., 1915, xx, 351; 1916, xxvi, 1. 23 Our ‘“‘protein-free milk’’ has been found to contain, on an average, 0.68 per cent N. 24 McCollum and Davis, J. Biol. Chem., 1915, xx, 641. 156 Vitamines . constituents of milk. More precise knowledge of these latter, as well as of the products of hydrolysis of lactalbumin, is needed before definite conclusions can be reached. In contrast to casein (and perhaps edestin) some of the other proteins which we have fed may need a supplement which is not found in yeast, yet is present in the so called “protein-free milk.’”’ Whatever the nutritive value of the unknown constituents of milk may be, our experience manifestly leads to the conclusion that the supple- mentary value of ‘protein-free milk” in the diet is, as a general rule, decidedly greater than that of yeast. Indeed the value of the latter would perhaps not readily have been discovered if other proteins than casein had formed the basis for the food mix- tures with which the earlier experiments were conducted. The objection may be raised that since some rats grow on the “artificial” diets alone, those animals which have done well on the yeast diets might have grown equally well without it. The improbability of this assumption is shown by the fact that whereas the great majority of the rats without yeast failed to thrive after a comparatively short period, nearly all of those receiving yeast with an appropriate protein as already indicated continued to grow normally for a long time. Moreover, the removal of the yeast was almost invariably followed by immediate cessation of growth and ultimate decline which could be promptly checked and converted into rapid recovery by the addition of a small amount of yeast. The numerous failures of growth in the experiments without yeast or other intentionally added sources of vitamines demon- strate that the quantities present in our other food ingredients must be far too small, at the best, to furnish enough of the growth- promoting material to satisfy the requirements of the majority of the animals tested. This experience does not support the suggestion made by some investigators, that such ingredients of our foods as casein and lactose, prepared from milk, and used in many of our food mixtures, are the carriers of a sufficient quan- tity of vitamines to vitiate experiments designed to test the effect of specially added accessory materials of this class. Our experiments indicate that the rapidity of growth is related 25 McCollum and Davis, J. Biol. Chem., 1915, xxiii, 181. and 231. Funk and Macallum, ibid., 1916, xxvii, 62. T. B. Osborne and L. B. Mendel 157 to the quantity of yeast fed. While there is a considerable variation among the individual rats with respect to the quantity of yeast needed, in general we have found 1.5 to 2 per cent of yeast in the food sufficient for promoting normal growth. Some rats have grown well with only 0.25 to 1 per cent of yeast in the diet, but almost invariably in these cases an increase in the amount of yeast given was followed by an increase in the rate of growth (Chart III). It might be assumed that when the rats grew on the smaller quantities of yeast their total food intake was greatly increased, so that the actual absolute intake of yeast was not very different from that of those animals receiving relatively larger percentages of yeast in their ration. A study of the food intakes of these rats, however, shows that this is not the case. The food consumption of the rats on the smaller quantities of yeast was less than that of those on the larger quantities, because their growth was slower and consequently they needed less food; and the change from a small quantity of yeast to a larger one was followed by growth with a resultant increase in the food intake. How does the yeast exert its beneficial effect? Does it merely add something which renders the food more palatable and so stimulates the animal to eat more liberally of it? Or does it | exert some favorable influence upon the metabolism of the rat, and thus improve its general condition so that more food is con- sumed? Satisfactory growth is associated with liberal eating; but whether the animal eats because it grows or grows because ‘it eats is a difficult question to settle. Hopkins! believes that “any effect of the addendum upon appetite must have been secondary to a more direct effect upon growth-processes;”’ and our experience leads us to the same belief, especially in those experiments in which the yeast was fed separately. In these cases the yeast could not have affected the inherent palatability of the ration, and the fact that the rats immediately increased their consumption of the ‘artificial’? food mixture points toward an improvement in the general condition of the animals which led them to raise their level of food consumption to keep pace with their more rapid growth and consequent need of more food. This experience fully confirms Hopkins’ results with feeding separately small quantities of milk.” 26 Cf. Hopkins for a full discussion (p. 441) of the relation of food intake to growth, which is confirmed by our own experience. oe Cyl CO Vitamines An attempt was made to concentrate the effective substance in the yeast by fractionation with alcohol. Moist yeast obtained from a brewery was filtered and the residue subjected to hydraulic pressure. The press cake thus obtained was heated to boiling in a large volume of distilled water made slightly acid with acetic acid, and filtered. The filtrate was evaporated and the resulting pasty mass thoroughly mixed with absolute aleohol and evaporated. After repeated evaporations with frequent additions of absolute alcohol the almost brittle mass was ground with absolute alcohol and centrifuged. The alcoholic solution, containing solids equal to 2.1 per cent of the dry yeast, was concentrated in a vacuum at about 65°C. The portion insoluble in absolute alcohol was pressed in the hydraulic press and dried in a vacuum over sul- furic acid. This was equal to 16.2 per cent of the dry yeast. Each of these fractions was incorporated with our “artificial protein-free milk’ and fed in suitable rations to rats. The fraction soluble in absolute alcohol exerted no beneficial in- fluence on their growth, whereas the addition of the residue solu- ble in water but insoluble in absolute alcohol in nearly every case led to a marked increase in their rate of growth. Here again the need of a sufficient quantity of the water-soluble frac- tion was demonstrated; for the use of 0.5 per cent of the yeast residue in the food was followed by a failure to make normal growth, whereas with 2 per cent of it in the food resumption of growth took place (Chart IV). That the above method of pre- paration has not tended to concentrate the effective substance in the yeast is shown by the fact that it required just as much of this residue, representing 16.2 per cent of the dry yeast, as it does of the whole yeast, to induce normal growth when fed with the “artificial protein-free milk’’ foods. The experiments just described confirm the presence in yeast of something comparable with the so called water-soluble vitamine. They offer no evidence regarding the presence or absence of the fat-soluble one, since a liberal supply of butter fat was used in all of the food mixtures. Funk and Macallum found that diets which contained yeast and butter showed only a slight superiority over those which contained yeast and lard. Our own experi- ments in which lard was the only fat component indicate that yeast contains only a very small amount, if any, of the fat- T. B. Osborne and L. B. Mendel 159 soluble factor. The lack of the butter fat may help to explain Funk and Macallum’s failures with the phosphotungstice acid fractions of yeast:'® for unless all the other necessary factors of the diet are adequately supplied, the presence or absence of the water-soluble vitamine cannot be demonstrated. Despite the success which has attended the use of yeast as an adjuvant to otherwise inadequate food mixtures, notably in the case where casein or edestin furnished the bulk of the protein, such yeast-containing ‘artificial’? food mixtures have not yet demonstrated a nutrient efficiency equivalent to that manifested through the use of “protein-free milk” or certain other naturally occurring food products like cotton seed meal. The refusal of some rats to eat an adequate amount of the yeast-containing foods has proved a stumbling block to exact. comparisons. Al- though some of the animals brought up on the yeast-containing foods have given birth to young, thus far none of the latter has been reared. The charts follow. 160 Vitamines | ErrEct of Yeast ano |PRoTEIN- FREE MILK: ane ano GRowTH| | heel Zaha | ; z Rare a DAYS BODY WEIGH 2 ee SE PACE EEC ee S = Rt —_GIMAMO DAYS Cuartr lI. Illustrative graphs showing the effects of small additions of yeast and of “‘protein-free milk’’ respectively on the rate of growth and: the intake of food (charted in gm. per week). The food mixtures had the following composition. Rat 16849. Rats 29610 and 2991c7. Benod pened oieg Peiodelie per cent|per cent|per cent per cent Casein es ke ci eer ears 18 18 18.0 18.0 ‘Artificial protein-free milk’? IV*.| 29 29.5 29.5 Natural ‘‘protein-free milk’’....... 28 Starchitee ta. fcc ae oes ae Soe 28 29 Didiad DA 5) Buttertate sector co ote circ 18 18 18.0 18.0 Lard eee eee: aniecice otros sien if if 7.0 7.0 Veastweers. tar. elec th. stoaare hte. 0.1 gm. per day. * The composition of “artificial protein-free milk’’ IV is given in J. Biol. Chem., 1913, xv, 317. TSB. Osborne, and i. B. | COMA PCawdmel east wo) | aaa | dase) —_| pee Tg Be LacTALGUMIN : DAYS Cuart II. if Heme aur cu | gf 1 Mendel 161 Illustrative graphs showing the ready growth on casein food and the failure to grow well on lactalbumin food and cotton seed globulin food when yeast was used as a course of water-soluble vitamine salong with following composition. “artificial protein-free milk.” Casein food. Cotton seed globulin food. LOLS See ee a ae “Artificial protein-free milk’’ IV... per cent 18.0 29.5 26 .0-25.5 18.0 TAD 1.5-2.0 per cent 18.0 29.5 18.0 18.0 15.0 1.5 The food mixtures had the Lactalbumin food. per cent 18.0 29.5 17.0-16.5 18.0 16.0 1.5-2.0 162 Vitamines 80 |-F Frdcr oF VARIOUS] PROPORTIONS] OF i aa, eee ee fami | || sortie DAYS Cuart III. Illustrative graphs showing comparative growth-promot- ing value of different proportions of yeast used as a source of water-soluble vitamine in the diet. The food mixtures had the following composition. per cent CASEI ered hice ee te re el se sorato caste 18.0 “Artrticial protein-tree mmlk’) VV 3. 2.0m.» 2 oe es 29.5 SG arenes nn eed ee ou URS cies Neue oernh ashot he 2 27.25-26.0 Buther bait crces ccc hsc ee eres ee ee oe eee 18.0 Tse ee Poe in Be a ee he Crean 720 VAS ees ae Le I a ee ST Ee ene Pa ciar 2A (0025=185 T. B. Osborne and L.-B. Mendel 163 COMPARISONLOF YEAST FRACTION ouRCEs oF VITAMIN PROMOTING GROWTH | Cuart IV. Showing a comparison of the efficiency of the alcohol- soluble fraction (a) and the water-soluble fraction (b) (described on p. 158) as a source of water-soluble yeast vitamine in promoting growth. The food mixtures had the following composition. per cent per cent (CRISTEA oA. Se areca ale eee ea 18 18.0 “Artificial protein-free milk’? IV............ 28 28.0 Siac MMPI ees tine ne eh og oe lee 29 28 .5-27.0 JEYORETP TENE eee NE gee et ea 18 SOR se IDG, 2:5 5 oa ee uf 7.0 Alcohol-soluble fraction (a)................. 0.044 Water-soluble fraction (b)................... 0.5-2.0 A NOTE ON MODIFICATIONS OF THE COLORIMETRIC DETERMINATION OF URIC ACID IN URINE AND IN BLOOD. By L. JEAN BOGERT. (From the Laboratory of Pathological Chemistry of the School of Medicine and the Sheffield Laboratory of Physiological Chemistry, Yale Uni- versity, Hew Haven.) (Received for publication, May 16, 1917.) For the estimation of uric acid in urine, a rapid micro method such as the procedure suggested in the Benedict-Hitchcock modi- fication! of the Folin-Macallum method has numerous advantages over the slower and more laborious Folin-Shaffer? method. A colorimetric method is especially useful where uric acid excretion is determined for short periods, involving the accurate determina- tion of small quantities in order to place emphasis on slight varia- tions in uric acid output. Loss by manipulation is reduced to a minimum. Previous attempts to use the Benedict method in these laboratories have, however, failed to yield accurate results, owing chiefly to rapid variations in the color of the unknown uric acid solutions. Benedict himself refers to this difficulty indirectly. In speaking of the deeper color produced by a given quantity of uric acid when potassium cyanide has been added to the solution, he says: “This effect seems to be due chiefly to a marked diminution in the rate of fading of the color from solutions containing the cyanide.’’ He adds: “The slower fading of the color under these conditions is a distinct advan- tage.’’ Later he states that, due to the fact that reoxidation of the colored compound is accelerated by filtration, an unfiltered clear standard will read higher than the same solution after filtration. Myers and Fine? make note of the fact that the colors must be matched ‘‘at once’’ as they 1 Benedict, S. R., and Hitchcock, E. H., J. Biol. Chem., 1915, xx, 619. 2 Folin, O.; and Shaffer, P. A., Z. physiol. Chem., 1901, xxxui, 552. 3 Myers, V. C., and Fine, M. S., The Chemical Composition of the Blood in Health and Disease, New York, 1915, p. 17. 165 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 166 Urie Acid ’ “fade at a fairly rapid rate.’’ Benedict directs the comparison of the unknown solution with a “‘simultaneously prepared’’ standard solution, assuming that the two solutions lose color at the same rate. From our experience we have reached the conclusion that the rate of fading of the color is quite different in the unknown and standard solutions. In this case, any delay in reading, such as may be caused by the development of turbidity in either solution, will result in a considerable error. We were led to a more detailed study of the rates of fading in the standard and unknown solutions by the following experiment. A sample of urine was treated according to the procedure suggested in the Benedict uric acid method and the resulting unknown solution com- pared with a standard solution set at 20 mm. in the colorimeter. The unknown solution read at 17.5 mm. The solutions, when matched in color, were left in the colorimeter 20 minutes. The unknown was then noticed to be much lighter than the standard and was adjusted to match the standard at 19.9mm. 10 minutes later the unknown was again found lighter than the standard and matched at 21 mm. This suggested that the standard solution was not fading at all or was fading at a much slower rate than the unknown solution. Both solutions were, therefore, left standing over night in closed flasks and were read against each other on the following day, 20 hours after they were made up. With the standard again set at 20 mm. in the colorimeter, the unknown matched at 44 mm. The old standard was then compared with a fresh standard containing the same amount of uric acid and set at 20mm. The reading given by the 20 hour standard was 48 mm. The color of the unknown was too faint to compare in the colorimeter with the freshly prepared standard but it was calculated (20: 44:: 48: X) that a depth of about 106 mm. would be neces- sary to match in color 20 mm. of the new standard. Further experiments led to the belief that, although the rate of fading of the standard solutions varies slightly, under the conditions used by us, an allowance of 5 per cent loss of color in the standard per hour for the first 2 or 3 hours after maximal color has developed is probably justifiable. Thus, when the standard solution in the colorimeter is first set at 20 mm., it is moved to 20.5 mm. before reading a second unknown against it + hour later. By making this small correction, it is possible to read several unknown solutions against the same standard and, with attention to other factors discussed later in this paper, duplicate analyses have been obtained which checked very closely. vs L. J. Bogert 167 Benedict recommends a delay of 4 minute after adding the reagents to the standard uric acid solution to allow maximum color development before diluting. We find that the develop- ment of color does not seem to be affected by dilution and that the maximum color is not attained for 15 minutes after the addition of the alkali to the standard uric acid solution. This is illustrated by the following experiments, in which different standards were read at intervals of 5 minutes against freshly prepared standards, diluted after standing 3 minute, and immedi- ately placed in the colorimeter at 20 mm. The readings were made as rapidly as possible, usually in less than 1 minute. The results are given below. Standards. Fresh. After mm min mm 20 5 19.9 20 10 17.6 20 10 17.5 20 15 15.0 20 15 1oRO 20 15 15.4 20 15 15.0 No further development of color occurred after 15 minutes. This led us to adopt the routine of allowing all standard solutions to stand 15 minutes before using. The development and fading of color in the unknown solutions were next investigated and were found to vary greatly in different unknowns, even in duplicate analyses on the same urme. The data from such an experiment are given below. In each case the unknown solutions were compared with fresh standards, diluted to the mark after standing } minute, and used zmmediately. Standards. Sample 1. Sample 2. mm. mm. mm. 20 Read at once. 9.5 9.0 20 After 5 min. Une 1.4 20 seed (aS 8.2 8.0 20 sera =< 8.5 8.0 9.5 8.0 20 ean 168 Urie Acid Since the lowest readings in duplicate analyses were found to check invariably within the limits of error in reading, we adopted the procedure of keeping each unknown solution in the colori- meter several minutes, making readings every 2 or 3 minutes, until the poimt of maximum color development had been reached. Although this point is reached in 5.minutes in each of the series of readings given above, samples have occasionally been found which required as long as 20 minutes for full development of the color. Difficulty was sometimes caused by turbidity in the solutions. When the modified uric acid reagent devised by Benedict was used, the solutions were seldom turbid. We found, however, that dilution with 20 to 30 ce. of water before adding the sodium carbonate solution was also of assistance in avoiding turbidity. In the very rare instances where the solutions did become turbid, a perfectly clear fluid was obtained by centrifuging for 1 or 2 minutes. This is more rapid than filtering as advised in the Benedict method and avoids the oxidation, with consequent loss of color, caused by filtration. We have finally used the method as follows. 2 cc. of urine are measured into a centrifuge tube, diluted to about 5 ec. with water, stirred, and treated with twenty drops of ammoniacal silver — magnesium solution.’ The contents of the tube are well mixed with a stirring rod and the latter is washed down with water. The tube is then centrifuged 2 to 4 minutes. The supernatant liquid is poured off as com- pletely as possible, the tube being inverted and the inside of the lip touched with a towel or piece of filter paper. Compressed air or suction is useful in removing the last traces of ammonia. The residue in the tube is then treated with two drops of 5 per cent potassium cyanide solution and the mixture thoroughly stirred with a slender stirring rod for half a minute. About 1 ec. of water is added and the solution is again stirred. 2 cc. of uric acid reagent are then added, the mixture is stirred, and washed into a 50 ce. flask with 20 to 30 ce. of water. After the addition of 10 ec. of 20 . per cent sodium carbonate solution, the solution is diluted to the mark and compared in the colorimeter with a solution obtained by treating 5 ce. of standard uric acid solution (containing 1 mg. of uric acid) with two 4 Neuwirth, I., J. Biol. Chem., 1917, xxix, 478, note 4. 5 The composition of all solutions used may be found in the paper by Benedict and Hitchcock,! with the exception of the modified uric acid re- agent described by Neuwirth. L. J. Bogert 169 drops of 5 per cent potassium cyanide solution, 2 cc. of uric acid reagent, 10 ce. of 20 per cent sodium carbonate solution, diluted to 50 ce., and allowed to stand 15 minutes before using. If any turbidity develops, it is removed by centrifuging. The unknown solution is placed in the colorimeter with as little delay as possible after it is made up and readings are taken every 2 or 3 minutes until maximum color development is reached, which usually occurs in 5 to 10 minutes. If the standard has stood more than 15 minutes, allowance is made for a loss of color of 5 per cent per hour. Standards are seldom used longer than 1 hour. By this procedure uric acid added to urine has been determined quantitatively and duplicate analyses have been obtained which agree uniformly within per 2 cent, as shown by the following figures. Volume of urine. Urie acid. Colorimeter readings. cc. mg. mm. 1,080 346* 15.6 348 15.5 1,005 283 de 285 16 910 286 15.9 286 15.9 1,350 307 22.0 304 22.2 * These figures were all obtained from a subject who was on a purine- free diet. In attempting to use the Benedict modification of the Folin- Denis method for determining uric acid in the blood,® even greater difficulty was experienced in obtaining consistent results. In addition to the errors caused by rapid changes of color due to development and fading of color in both the standard and un- known solutions, as discussed previously in this paper, the solu- tions were found to be tinged with yellow, instead of water clear, after adding colloidal iron and filtering. This yellow color could not be removed by adding salt solution or avoided by the addi- tion of less iron. It apparently resulted in loss of the uric acid 6 Benedict, S. R., J. Biol. Chem.,°1915, xx, 629. 170 Urie Acid by oxidation, as very faintly colored or colorless solutions were usually obtained from these analyses on adding uric acid reagent and sodium carbonate solution. The loss of uric acid was roughly proportional to the amount of iron which had been added. Ona few occasions only, in the course of many analyses, were results obtained in which there had probably been no loss of uric acid. Of known amounts of uric acid added to samples of blood, only 25 to 50 per cent were recovered. The omission of the precipitation by colloidal iron was sug- gested by Dr. Louis Baumann,’ who had experienced similar difficulties. By following this procedure, in which reliance is placed on the treatment with dilute acetic acid and heat for the removal of most of the protein, results have been obtained which agree within the limits of experimental error. 10 ec. of fresh oxalated blood are pipetted into a casserole containing 50 ee. of boiling 0.01 N acetic acid and heated to boiling until coagulation is complete. It is then filtered into a second casserole, washing the coagulum and the casserole in which coagulation took place with 200 ce. of boiling water. The filtrate, which should be almost if not entirely water clear, is evaporated to 50 cc. either on a water bath or over an asbestos mat with a small central hole for the free flame. This latter precaution is to avoid oxidation on the sides of the dish. When the volume-of the solution is about 50 cc. it is washed quantitatively with boiling water into a 100 ce. casserole, after filtering if there is any precipitate. This solution is then concentrated to about 10 ce., washed with hot water into a centrifuge tube, 2 cc. of silver magnesium mixture are added, and the whole is well stirred with a slender glass rod. It is advisable at this point to allow the mixture to stand at least 1 hour as the uric acid precipitate often forms quite slowly. After centrifuging, the supernatant liquid is carefully removed and the same procedure outlined under the determinations on urine is followed for preparing and reading the colored solution. A standard containing 0.5 mg. of uric acid, diluted to 50 ce. and set at 20 mm. in the colorimeter, gives about the right depth of color for matching against a normal blood, if a 10 cc. sample has been taken and the resu!ting colored solution made up to 25 ce. In the case of bloods containing larger amounts of uric acid, the unknown may be diluted to 50 ce. or a standard containing 1 mg. of uric acid in 50 ce. may be used. The results of a few analyses are given below to show that, 7 The writer desires to express gratitude to Dr. Louis Baumann, Towa State University, for a number of helpful suggestions based on work done by Mr. Thorsten Ingvaldsen and received by private communication. L. J. Bogert 171 after endeavoring to control all the variable factors outlined above, accurate and consistent results have been obtained. Some of the samples were from pathological cases. 10 cc. of blood. 0.93 mg. of uric acid. 10 “cc “ec “ce -+ 1 mg. of 1 94 “ec “ “ “ec uric acid. Volume of blood sample. Uric acid per 100 cc. ce. j mg. 10 2.4 10 2.5 10 3.0 10 2. 10 4.2 10 4.2 10 tke 5 7.5 5 3.1 10 3.0 —_—— THE NUTRITIVE VALUE OF THE DIAMINO-ACIDS OCCURRING IN PROTEINS FOR THE MAINTENANCE OF ADULT MICE.* By E. M. K. GEILING. (From the Department of Animal Husbandry, University of Illinois, Urbana.) (Received for publication, May 25, 1917.) Review of the Previous Work on the Nutritive Value of the Diamino- Acids Occurring in Proteins. The functions of the different amino-acids in the animal body have been studied mainly by means of three general methods. The first consists of feeding completely hydrolyzed proteins—hydrolysis may be effected by either acids or enzymes—from which the amino-acids whose functions are to be ascertained have been removed. This method has been developed and extensively used by Abderhalden and associates who have obtained some very valuable results. More recently Hopkins and coworkers have employed this method with considerable success. According to the second method rations made of pure proteins lacking or deficient in one or more amino-acids are fed to animals. The behavior of the experimental animals on such a ration will serve as a good index of the function of the missing or deficient amino-acids. Osborne and Mendel and associates (among others) have pursued this line of investigation and have secured results of great significance and value to the science of nutrition. The third method of attacking this problem has been used by Abderhalden and more recently by Mitchell;! it consists in supply- * The results presented in this paper formed part of a thesis submitted to the Graduate School of the University of Illinois in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Depart- ment of Animal Husbandry. 173 174 Nutritive Value of Diamino-Acids ing the nitrogenous portion of the ration with pure amino-acids. This method promises to be very successful, if the technique is well developed. The main difficulty seems to be that of securing a sufficient consumption of the food for considerable periods of time. At present, however, its application must be somewhat limited on account of the high cost of the amino-acids. Much of the earlier work on the indispensability of certain of the amino-acids to the animal organism is of little value, because the rations used were inadequate in the nutrients other than protein. For example, in the literature one finds rations used composed only of hydrolyzed protein and cane sugar. Obviously, the poor results obtained with such a food mixture may be due just as much to the absence of the accessory substances, mineral salts, or fat as to the deficiency in the protein make-up. Henriques and Hansen,’ using the first named method, removed the diamino-acids from an enzyme digest of a protein by precipitation with phosphotungstic acid; after removal of the excess of the reagent from the filtrate, the product was evaporated to dryness in vacuo. The resultant mixture centained presumably only monoamino-acids and formed the sole source of nitrogenous material for their feeding experiment. Only one rat was used in their experiment. The authors concluded from this single trial that the diamino-acids could be entirely dispensed with by the animal organism. Mitchell! aptly criticizes this work in the following words: “Unfortunately the particular protein used was not mentioned, and the lack of all details of the phosphotungstic acid precipitation precludes any attempt to judge of the completeness of separation of monoamino- from diamino-acids. The experiment lasted only 26 days. During the last 17 days only were positive nitrogen balances obtained and they were such that they could hardly be said to constitute convincing evidence of the nutritive adequacy of the ration fed, especially in view of the fact? that the determination of the total urinary nitrogen per day of rats, with the exercise of the utmost care as to collection and preservation, is sub- ject to errors of 10 per cent or more, due to incomplete collection. The body weight increased slightly during the last half of the experiment, but during the last 3 days it declined slightly but consistently. It is un- fortunate that the experiment was not continued further to determine whether this final decrease was significant or not. This criticism is especi- ally justified by the many experiments that may be quoted on the feeding of synthetic rations, in which entirely erroneous conclusions may be drawn if attention is confined to the first 20 or 30 days of observation.’’ Abderhalden‘’ in a recent extensive communication discusses the bio- K. M. K. Geiling 17D logical value of the amino-acids and presents some very interesting mate- rial. He has succeeded in obtaining a completely hydrolyzed mixture of amino-acids from meat by digestion first with enzymes and finally with 10 per cent sulfuric acid. No trace of any intact polypeptide compound could be discovered. On this mixture as the sole source of nitrogen a dog lived in good health and increased about 10 kilos in weight for 100 days. Lysine was removed from the hydrolyzed product and resulted in a negative nitrogen balance. When hydrolyzed gliadin formed the sole source of nitrogen, a negative balance was also obtained. Even with the addition of lysine to the hydrolyzed gliadin, a positive balance could not be secured, although an improvement was observed. The test periods lasted only 7 days. The results obtained with arginine were inconclusive, as the experiments were discontinued before an after-period with arginine was completed. It is suggested that ornithine can replace arginine, but the evidence for this is not very good, as Abderhalden admits. From the experiments with histidine and cystine no definite conclusions were drawn. Histidine was removed with HgCl, and the resultant amino-acid mixture gave a negative nitrogen balance. Addition of histidine had not the effect of restoring a positive balance. From this Abderhalden concludes that the removal of histidine must have brought about a change in the amino- acid mixture, probably destroying or rendering inactive some of the es- sential amino-acids. Cystine is regarded as probably being an essential amino-acid. Its removal was effected with acetic acid, but was not quan- titative. The excess acetic acid could not be evaporated without destroy- ing some of the amino-acids. Here again the amino-acid mixture, when supplemented with cystine, was inadequate. Considerable difficulty was experienced in some cases in getting the animals to eat the food. Other animals again vomited their food or contracted diarrhea. Prob- ably the chief criticism against this work is that the experimental periods were of too short duration, usually only 7 days with any one ration. Ackroyd and Hopkins*:* recently have made some interesting con- tributions to our knowledge concerning the function of some of the amino- acids in the animal body. Rats were used as experimental subjects. The nitrogen requirements were supplied by casein, completely hydrolyzed with 25 per cent sulfuric acid. The amino-acids to be studied were re- moved from this hydrolyzed product and the resultant mixture, supple- mented with cystine and tryptophane, formed the sole source of nitrogen in the rations used. Rats fed a ration from which the arginine and the histidine had been removed by the Kossel and Kutscher method ceased to grow and lost weight. When these amino-acids were replaced, growth was resumed at a normal rate. Further experiments showed that the animals grew when either arginine or histidine was present, thus indicating that these two amino-acids are interchangeable in nutrition. The close simi- larity between these two amino-acids is shown by writing their structure in the following manner: 176 Nutritive Value of Diamino-Acids CH.N CH.— NH T Der | >». NH, C-—NH CH, NH | CH, CH, | | CH . NH: _ CH.NH: | COOH COOH Histidine. Arginine. (Iminazole-amino propionic acid.) | (Guanidine-amino propionic acid.) “The essential molecular changes involved in passing from one form to the other—the opening or closing of a ring and the addition or removal of an amino group—may be inferred from our general knowledge of the chemical powers of the body, and may be regarded as essentially physio- logical processes.”’ In another series of experiments these investigators have shown the special part played by arginine and histidine in purine metabolism. A comparison of the structural formula of these two amino-acids with that of guanine, a typical purine base, will show their close relationship. HN— CH, HC—N HN — CO ye r\ fee! H,N—C CH, | CH H.N-C C—NH Be eee | Se ; ] > Zz N — C— NZ ee CHV. Cia.) | COOH CH (NH) COOH Arginine. Histidine. Guanine. Allantoin is the main end-product of purine metabolism in rats. Quanti- tative determinations of allantoin in the urine of rats will, therefore, reveal the operation of any factors which have influenced purine metabolism. In this connection the following experiments were conducted: Rats were first fed a complete amino-acid mixture to determine the normal allantoin excretion; then arginine and histidine were withdrawn in a second period, and were again replaced in a third period. A marked decline in the allan- toin excretion was noted during the period when arginine and histidine were withdrawn from the ration. To obviate a criticism that the falling off in allantoin is due to a general lowering of metabolism, owing to the removal of a factor essential to nutrition, two further experiments were performed to settle this point in the following manner: Rats were fed a ration from which the tryptophane, which is essential for maintenance, had ee EK. M. K. Geiling 177 been removed, and the allantoin was determined in the urine. The animals declined in weight, but the allantoin excretion remained unchanged, thus indicating that arginine and histidine function in a special way in purine metabolism, and that the decline in allantoin during the second period was due to a removal of these amino-acids, and not to a lowering of the general metabolism. A similar decline in weight with unchanged allantoin excretion was noted, when the vitamine supply of the ration was withdrawn. Previous to this work Abderhalden and Einbeck’ studied the rela- tionship between histidine and purine formation in the animal organism. They supplemented a complete ration with histidine and obtained no in- crease in uric acid or allantoin excretion. Later these experiments were repeated in conjunction with Schmid,* but again negative results were obtained. In these experiments only the total nitrogen and allantoin were determined in the urine. Kowalevsky® also studied the rédle of his- tidine in the metabolism of a dog. A dog was fed with milk, bread, and cane sugar, and later the ration was supplemented with histidine. No increase in uric acid was noted in this latter period. There was an in- crease in urea and ammonia, and the urine was acid. The disagreement of this earlier work with that of Ackroyd and Hopkins can probably be ac- counted for by a difference in experimental procedure. The former workers fed histidine in addition to a complete ration, whereas the latter authors removed histidine completely from the diet and then noted the effect. This method is certainly the better one, and results secured by employing it are more significant than those obtained when the amino-acid was practically fed in excess. Osborne and Mendel,!° using the second general method of experimenta- — tion, namely, feeding purified incomplete proteins and then supplementing them with the amino-acids deficient or lacking in the ration, have made some important contributions to our knowledge of the physiological action of certain amino-acids. Lysine was shown to be essential for growth in rats. This was done by feeding gliadin, the alcohol-soluble protein of the wheat kernel, as the sole protein of a ration. The animals maintained their body weight for long periods of time, but failed to grow normally. On the addition of lysine to the ration normal growth was resumed. Simi- lar results were obtained when other proteins low or lacking in lysine, like zein of maize, were used. Subsequently !! these investigators have shown that for the normal growth of rats the rations used by them must contain somewhat over 2 per cent of the protein as lysine. Buckner, Nollau, and Kastle,!2 using chickens as experimental subjects, fed them rations presumably high or Jow in lysine. The chicks were fed complex mixtures, such as wheat, wheat bran, sunflower seed, hemp seed, and skim milk, a ration taken as high in lysine, and a mixture of barley, rice, hominy, oats, and gluten flour, a ration taken as low in lysine. The animals receiving the first named ration grew normally, while those fed the latter remained stunted. From these results the authors conclude that the marked differences shown by these two lots of chicks in the rate of growth and development are due to ‘‘differences in the amino-acid content 178 Nutritive Value of Diamino-Acids of the two rations and in all probability to differences in the lysine con- tent.’’ This work is, however, open to serious criticism. It may be shown that their method for determining lysine in feeds is thoroughly unreliable. Furthermore, the grain mixtures fed the two lots of chickens differed so radically that the interpretation of their results is difficult. Osborne and Mendel'’ have lately obtained results similar to those of Buckner, Nollau, and Kastle with chicks, but instead of using mixtures of grains, corn gluten, containing about 1 per cent of lysine, and a mixture of equal parts of corn gluten and lactalbumin, yielding about 10 per cent lysine, were used to supply the protein requirements. The balance of the rations was made up with protein-free milk, starch, butter fat, and lard. The data show in a striking way the difference in the efficiency of the two rations used. After 55 days the chick receiving the corn gluten ration gained only 52 gm., while the other chick receiving the corn gluten and lactalbumin ration gained 283 gm. The stunted chick exhibited no signs of malnutrition other than failure to grow. However, McCollum, Simmonds, and Pitz,!4 in experiments on rats de- signed to ascertain the supplementary relationships among the naturally occurring foodstuffs, are forced to the conclusion that “in the protein mixture of the maize kernel and the oat kernel, lysine certainly is not the essential protein cleavage product which is present in amount so small that it is the limiting factor which determines the biological value of the pro- teins of these seeds.’’ Thus, zein, though lacking in tryptophane and lysine, supplements the proteins of the oat kernel in a surprisingly efficient manner. Also, gelatin with its high lysine content does not improve the proteins of the maize kernel. Rats fed a ration in which zein, supplemented with lysine and trypto- phane, was the sole source of protein, did not always grow as rapidly, nor was growth as prolonged as was expected by Osborne and Mendel. Hence, in a subsequent experiment arginine was also added to the same ration (zein being very low in arginine) and resulted in improved growth. His- tidine was added in one ease, in addition to the arginine, and a slight increase in the growth was noted, but Osborne and Mendel'® do not regard this as significant, inasmuch as another rat grew quite as rapidly in the absence of this amino-acid. This conclusion appears justified by the work of Ackroyd and Hopkins quoted above, in which it was shown that arginine and histidine seem to be interchangeable. Myers and Fine!® studied the effect of feeding rations low and high in arginine on the creatine content of rat muscle. The animals fed the “Osborne and Mendel’ ration, low in arginine, contained about 2.5 per cent less creatine than did those fed on a diet rich in arginine. The fact that both arginine and creatine contain the guanidine ring makes it prob- able that the latter may be derived from the former in the body. When these data are submitted to a statistical examination, the differences be- come much more striking, and hence more significance may be attached to the conclusions than is done by the authors themselves. Concerning the nutritive valve of cystine, Osborne and Mendel!’ re- E. M. K. Geiling 179 port results showing that when proteins low in cystine, such as casein, are supplemented with this amino-acid, a much smaller amount of the protein is required in the ration to produce normal growth. For example, rats grew normally on a ration containing 15 per cent of casein, but when the protein was reduced to 9 per cent, growth was promptly limited. The addition of isolated cystine to the ration with 9 per cent casein at once rendered the ration decidedly more adequate for growth. “Growth can be facilitated or repressed at will by the addition or with- drawal of the extra cystine from the diet containing 9 per cent of casein.”’ These results at least suggest strongly that cystine too may be made a limiting factor in growth. From the works just reviewed it will be seen that, with the exception of the work of Henriques and Hansen, al] the data secured on the nutritive value of the diamino-acids were obtained with growing animals. It appears that lysine and cystine may be considered as essential for the growth of animals and that either arginine or histidine must be present to . render the ration adequate for maintenance. Preparation of the Food Materials. In synthetic feeding experiments it is essential that the food materials used should be as pure as possible, for sometimes impurities present in even small amounts may have a very de- cided influence. Below are given, in brief, the methods used in the preparation of the different nutrients. Casein.—Prepared from a solution of skim milk powder in twelve parts of water by repeated precipitation with dilute hydrochloric acid and solu- tion in dilute sodium hydroxide. The product was finally extracted with ‘alcohol and ether, dried, pulverized, and put through a 40 mesh sieve. Dextrin.—According to the method of McCollum and Davis,!* high grade corn starch was moistened with 0.5 per cent citric acid solution and made into a stiff paste which was heated in an autoclave at 15 pounds’ pressure for 3 hours. The resultant gelatinous mass was cut into slices, washed with alcohol, dried, ground in a mill, and passed through a 40 mesh sieve. Butter Fat.—Best quality creamery butter was melted at a low tem- perature (40°C.) and then centrifuged for 1 hour. The clear liquid was syphoned off and allowed to harden. Protein-Free Milk.—In most oi the experiments reported in this paper, the protein-free milk was prepared according to the method of Osborne and Mendel.!® In some of the later experiments, a slight modification of this method was used in that a second filtration was introduced after neutralization of the filtrate from the lactalbumin. While this filtration removes some calcium phosphate, it also reduces the nitrogen content of the final product by about 0.1 per cent. In previous work done in this labora- 180 Nutritive Value of Diamino-Acids tory?" it was shown that this modified preparation is adequate to cover the mineral requirements of adult mice. Hydrolyzed Casein.—Two preparations of this product were used— one digested with enzyme for 2 months and the other for 34 months. 200 gm. of casein (either purified or commercial) and 20 gm. of pancreatin were nuxed with 2 liters of water made slightly alkaline with ammo- nia, and kept in an oven at 40°C. for the periods mentioned above. The mixture was stirred every day. Toluene was added to prevent bacterial erowth. At the end of each:‘month 10 gm. of pancreatin were added. At the conclusion of the digestion period, the mixture was boiled and filtered. The precipitate was washed with boiling water, and the volume of the filtrate reduced to about 2 liters by evaporating on a water bath at a low temperature. Nine and a half volumes of redistilled 95 per cent alcohol were added to precipitate all the peptones, proteoses, and complex peptides. How complete this precipitation with alcohol is, is not known. The solution was filtered and the alcohol distilled off at a low temperature. The precipitate was washed with boiling water to dissolve out the tyro- sine. The washings thus obtained were allowed to cool, and the tyrosine crystallized out. The tyrosine was washed with ice water, dried, and added to the material obtained from the alcoholic filtrate. The resultant product, When dried, was a brown powder, with a sharp tasté, ‘and did not give the biuret test (showing that only the very simplest peptides ° could be present). The amino nitrogen before and after hydrolysis of this product was determined according to the method of Van Slyke®! in order to arrive at an estimate of the amount of peptide nitrogen present. The product’ hydrolyzed for 2 months contained 36.2 per cent peptide nitro- gen, and the one hydrolyzed for 3} months, 24.25 per cent. Monoamino-Acid Mixture.—About 50 gm. of hydrolyzed casein, pre- pared as outlined above and containing 24.25 per cent peptide nitrogen, were dissolved in 5 liters of 5 per cent sulfuric acid. Phosphotungstic acid, dissolved in 5 per cent sulfuric acid, was added in excess to precipi- tate the diamino-acids, and any peptides which may have a diamino-acid component.22, The mixture was allowed to stand at room temperature for 48 hours, after which it was filtered. The excess phosphotungstic acid and sulfuric acid were removed from the filtrate with barium hydroxide. The solution was filtered and evaporated to dryness at a low temperature. The product thus obtained was of a light yellow color and had a fairly sharp taste. The peptide nitrogen in this preparation was 23.54 percent. To test whether this monoamino-acid mixture contained any diamino- acids in peptide form, the following experiment was carried out. A sample of the mixture was completely hydrolyzed with 20 per cent hydrochloric acid, the acid removed by evaporation on the water bath at a low tempera- ture, and the residue taken up with water. The solution was filtered to remove the melanin, of which there was very little, made up to volume, acidified with hydrochloric acid, and phosphotungstie acid was added, according to the directions as outlined by Van Slyke.?!. No precipitate was formed; only a shght scum floated on the surface, which was not a E. M. K. Geiling 181 phosphotungstate of the diamino-acids, since it was insoluble in boiling water. This shows conclusively that the diamino-acids were all removed from the monoamino-acid mixture in the first precipitation with phospho- tungstic acid, even those that may have been present in peptide form. Diamino-Acids.—The diamino-acids used were obtained from gelatin by hydrolyzing with 25 per cent sulfuric acid for 40 hours on an electric sand bath, then diluting the mixture with water to give a 5 per cent sul- furic acid solution. Phosphotungstic acid dissolved in 5 per cent sulfuric acid was added in excess, and the mixture was allowed to stand at room temperature for 48 hours, and filtered. The precipitate was washed with a 2 per cent sulfuric acid solution of phosphotungstic acid, decomposed with barium hydroxide, and the mixture was filtered. The excess alkali was removed with sulfuric acid, and the light colored solution was again filtered and evaporated to dryness at a low temperature. Other Amino-Acids.—Cystine was prepared from wool according to Denis’** modification of Folin’s method.*4 The Hopkins-Cole method was used in the preparation of tryptophane from casein. Arginine, histidine, and lysine were obtained from gelatin, blood meal, or casein according to the method of Kossel and Kutscher. EXPERIMENTAL. The experiments about to be reported were all conducted with adult mice for the purpose of ascertaining whether or not the diamino-acids, arginine, histidine, and lysine, which are precipi- tated with phosphotungstic acid in acid solution, are neces- sary for the maintenance of adult mice. In all the experiments with hydrolyzed protein products it was assumed that cystine is an indispensable amino-acid for the maintenance of mice. ‘Therefore such preparations, whether complete or containing only monoamino-acids, were supplemented with cystine in the large majority of cases. Some evidence will also be presented con- firming the correctness of this assumption. The method of attacking this problem was to feed hydro- lyzed casein, prepared as described above minus the diamino- acids precipitable by phosphotungstic acid in acid solution. This preparation—“‘monoamino-acid mixture’”’—generally supple- mented with tryptophane and cystine, was fed in an othcrwise adequate non-nitrogencus ration, except for the small am unt of nitrogen occurring in protein-free milk. In other words, the diamino-acids were made the limiting factor in an otherwise complete ration. If such a food mixture is sufficient to cover the maintenance requirements, the conclusion of Henriques and THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 1 182 Nutritive Value of Diamino-Acids Hansen is justified. The body then is able to synthesize the diamino-acids required. On the other hand, if this ration proves inadequate, it would tend to show that the material precipitated by the phosphotungstiec acid in acid solution is necessary for the maintenance of adult mice. In this investigation the problem was somewhat complicated by the fact that the casein was not completely hydrolyzed, from which it may be argued that some of the essential monoamino- ° acids may have been completely precipitated in peptide linkage by the phosphotungstic acid. Consequently the failure of main- tenance may be due just as much to the absence of these essential monoamino-acids, as to the lack of the diamino-acids. To ensure the presence of sufficient tryptophane, it was always added to the rations, except in one case. Should it, however, be shown that this ‘‘monoamino-acid mixture,”’ supplemented only with trypto- phane and the diamino-acids, is adequate for maintenance, then the above objection becomes void. The next step was to as- certain whether all or only some of the diamino-acids are neces- sary for maintenance of adult mice. Before the main problem of the investigation could be taken up, several preliminary points had to be settled. These will be discussed in their logical order. The details of the experimental procedure were similar to those followed by Mitchell, except that some of the rations used were made into a biscuit form. The biscuit was prepared by thoroughly mixing the dry ingredients — and adding enough warm water to form a stiff mush. This mush was dried in a thin layer on glass plates at a low temperature and subsequently the brittle mass was broken in pieces of convenient size. A small tin plate, about 3 inches in diameter, was placed below the porcelain crucible con- taining the food for the purpose of catching any of the ration which might be thrown out of the crucible by the mice. Considerable difficulty was experienced in getting some of the experimental animals to consume a> sufficient amount of a few of the rations fed. In these cases the mice scattered the food but most of it fell on the plate below, and could be recovered and fed again, if it was not soiled. In such instances the food intake mn casurement must be regarded as an approximation only. Experiment 1.—The first step was to compound a ration which would serve as an intermediate step between the grain and the experimental rations—a preliminary ration—and also ascertain the percentage of casein necessary for maintenance. For this purpose the following ration was made into a biscuit form. EK. M. K. Geiling 183 Ration 1. per cent Skim milk powder (6.01 per cent nitrogen, or 6.01 6.33=38.043 per cent protein) yielding 10.65 per cent protein in the ration. 28 JEAUT ENE TID eal Be i eis ec nt pen ee a mee A ele 5 ACTOS CMI ar ease ree eee one oe a oe heed 10 IDEXGEI nee eee eee he ti ae os ORE ee onn.2 See 57 All the ingredients were passed through a 40 mesh sieve before being mixed. The details concerning the weight of animals and duration of the experi- ment are appended: Days. No. 7la 2. | No. 72b 9. ]|-No. 71b 9. | No. 72 9. No. 62 ¢’. No. 62a 0’. gm. gm. gm. gm. gm. gm. 1 26.4* 18.7 26.7 25 .2 21.2 24.4 7 19.2 L725 23.5 23.7 19.5 23.9 14 Ge SSA 19.4 22.6 21.0 26.5 28 15.8 IIEISZE 19.8 22.0 19.5 26.0 35 16.8 15.5 22.6 22.8 19-7 26.3 42 eo 14.8 22.1 23.5 20.0 26.2 49 17.6 15.5 22.1 23.4 20.0 26.8 56 ile faa 1622: 21.8 23.7 19.5 28.5 63 17.2 16.3 22.1 23.4 19.0 28.5 70 17.6 16.6 22.4 23.1 18.2 28.5 77 NE 175 24.0 23.9 20.7 28.3 84 16.8 Ne 23.5 22.5 20.2 26.7 91 ilyfee3) 2S 24 .0** PA as OM fa Pia 98 19.0 18.7 105 18.8 19.0 112 18.0 19.2 119 19.7 18.7 126 18.2 18.5 * Gave birth to young but ate them. **Weight on 88th day. From these figures it will be seen that Mice 7la @ and 72b 9 maintained themselves satisfactorily for 126 days. The other four mice showed satisfactory maintenance at a higher weight for 84 to 88 days. In every case there was a drop in weight during the first 7 days. Such a drop in weight was observed in practically every experiment in this investigation following a change in ration. The food intake was quite satisfactory. Each 184 Nutritive Value of Diamino-Acids animal was fed about 2 gm. of ration per day. All the animals used in the present investigation were placed on this ration for several weeks before being used for an experiment. In later work this preliminary ration was somewhat altered, more butter fat and lard being added. Experiment 2.—It was thought advisable to test whether adult mice could be maintained on a lower protein diet. For this purpose the fol- lowing ration was made up: Ration 2. per cent DMextrim. ck he Fen ie. S Pe Ree Re se See ee TO 55 TiS CCOSCE Cee oe ek Bn ote Raclics tats BR cee URI, eM Pe Oe eee 15 SUCTOS@ Sees eta ec wate kb Rie De ee ee SS te footee ee 5 BUGTE share cschoe eres ee eee Ser eine ie nee 5 Skim milk powder—equivalent to 7.6 per cent protein in the MVGLOME Soi i ord ge ee Be ee a Ae re oct Sr 20 Four mice were fed this ration and took the food satisfactorily. Days No. 74¢ No. 74a o. No. 75 2 No. 75a 9 gm gm. gm gm 1 21.8 24.9 29.7 17.2 vi 17.9 20.0 26 .2* 19.2 14 16.2 18.2 20.2 19.7 21 15.0 17.2 20.8 20.7 28 15.2 17.5 20.7 18.5 35 15.3 18.7 20.6 19.3 42 15.2 IY fats) 18.0 18.1 * Gave birth to young but ate them. Since maintenance of weight on this ration did not appear to be as satisfactory as on Ration 1, it was decided not to reduce the protein of any of the rations below 10 per cent. These results are in agreement with those obtained by Wheeler? who found that growing mice require a higher per cent of protein than rats, which are able to maintain themselves on a very low per cent of protein. K. M. K. Geiling 185 Experiment 3.—A further test was made to ascertain whether a ration composed of 10 per cént purified casein was adequate for the maintenance of full grown mice. The ration used was made up as follows: Ration 3. per cent ZU CCE CASEI SENN Ar yack et ete kk) Uc eA citar ege 10 TEN IRS -F HEM Gets ae ol A ke 5 OteIMenree nilllkemeey ae Ge Rc cay ete cesses cldyos eal oti stere eR ees. nas eats 28 [DYED AHEM, wield Res, can SEAR eee EERE ee Ie eC ere noe ae eee 57 Four adult female mice were placed on this ration, with results as shown below. Before being placed on this ration, the mice were fed Ration 1 for 19 days. Days. No. 30 @. No. 30a 9. No. 35 9. No. 35a 9. gm. gm. gm. gm 1 24.2 25 .2 25.0 20.5 7 24.4 24.3 23.7 20.0 14 24.3 24.8 24.8 22.6 21 24.2 25.2 24.8 22.2 28 2212 24.2 24.0 21.0 35 24.7 295 25.3 22.4 42 23.5 25.9 25.6 PM Way it 49 28.0 28 .0 25.2 21.8 51 23.5 21. 0F 26h or 21.5 * Haten by a cat. ** Affected with lice; isolated and treated with carbolated vaseline. From these data it is obvious that Ration 3 satisfies the purpose for which it was compounded, at least for periods of 51 days. The casein of this ration was replaced by hydrolyzed casein in Experiment 6. Experiment 4.—In this investigation casein and its derivatives had to be subjected to heat; hence it was considered advisable to know the effect of heating on the nutritive value of these compounds. For this purpose purified casein was suspended in nitrogen-free water and boiled in a casserole on an open flame for 2 hours. Thereafter the water was evaporated on a water bath at a low temperature. Finally the 186 Nutritive Value of Diamino-Acids “boiled”? casein was dried in an electric oven at 40°C. Ration 4 was made up as follows: Ration 4. per cent oY Yow sto eee el:h=(2) 04 ee SAI ie th oe eerie s 2 10 Proteinchree mille: oe ee es cons oe se eee 28 Butter fates! «ec ee soe eee eee td Bice hice eee ec 5 Dexbriniiinc 43.5.5 6 oe Tar or ALi SIS cece ae 57 The ingredients were made up into a biscuit form, as described above. Four mice were fed this ration after having been on Ration 1 for 26 days. Days. No. 60 @. No. 60a 2. No. 53 9. No. 53a 9. gm. gm. gm. gm. 1 18.0 18.0 21.0 18.0 7 17.5 18.0 20.3 19.7 14 18.4 18.2 Ii et 20.0 21 22.0 20.6 19.0 21.3 28 21.9 20.3 19.9 21.3 35 20.5 19.0 21.4 19.2 42 23.2 20.7 20.0 20.0 49 23.2 21.8 21.8 22.2 These results indicate that casein suspended in water for 2 hours at 100°C. does not lose its nutritive value. Experiment 5.—This experiment was planned to determine the effect of heating moist casein in the autoclave for 1 hour at 15 pounds’ pressure. McCollum and Davis®’ recently reported that the nutritive value of casein is affected by heating it in this way. Their experimental animals were young growing rats. Ration 5 was made into a biscuit form out of the following ingredients: Ration 6. per cent Casein heated in an autoclave at 15 pounds’ pressure for 1hour.. 10 Protein=tree sri kst jn Wee heen tee BORG Fc 2acicc Me. 3 Rie Se 28 1 B01 Fok Ls 2 | BR UN oer? ie ee are Ake ah Aen RI oC 5 KB YES 7h ee ace eae geen a ee Nc rt ee A A EP MS Ss 57 The four mice used in Experiment 4, together with four additional ones, were placed on this diet with the following results: E. M. K. Geiling 187 2 No. 60 9.|No. 60a 9.)/No. 53 9. |No. 58a 9.| No. 81 9.|No. 8la 9.} No. 83 9.|No. 83a 9. Lol gm. gm. gm. gm. gm. gm. gm. gm. 1 23.0 21.8 21.8 222 25.0 Dona) 24.5 19.5 all 2125 21.8 20.7 Deo 23.5 2205 2AxS Wee 14 22.9 21.8 20.6 2206 24.2 23,0 PID) 16.5 21 23.5 22.4 Dark 28 22.2 21.0 20.9 21.1 PAU) 2135 20.5 19.3 35 PAV ete: PMs 20.7 21.8 17.0 19.3 PAY 19.4 42 PAVE) Pppat 18.5 Pale Ch ions 18.4* NORO*4 | Leone 49 PA Dae 19.0 PAS 14.6 19.2 22.0 16.3 56 19.5 PAS) 19.5 20.5 Died 19.3 BDAt 15.0 68 19.0 Des 19.2 21.0 52nd 19.0 Zo 15.0 70 19.6 ZOEZ 19.17 PAU ORE day. 18.0 22.0 13-0 77 19.7 20.6 PANES 22.8 17.6 20.0 15.0 84 16.8 19.4 PANS) 24.0 1D? 18.0 22 91 | Died 16.3 Dilern 24.2 Died 16.0 10.7 98 | 92nd Died 84th 18.0 10.1 day. 91st day. Died Died day 101st. 98th day. day * Added cystine and 10 per cent lard. ** Changed to unboiled casein and 10 per cent lard. + Butter fat containing casein added. t Weight on the 79th day. After Nos. 53 Q and 53a Q had been on the autoclaved casein ration for 70 days, they were fed a ration made of butter fat containing some untreated casein. The impurity of this butter fat was not discovered until the 21st day of feeding. The animals immediately gained in weight, thus showing the beneficial effect of the addition of only a small amount of unheated casein. With these subjects this ration was discontinued on the 91st day of observation. The remaining six animals gradually lost in weight and in each case ultimately died, thereby verifying the conclusion of McCollum and Davis that heating casein in an autoclave for 1 hour at 15 pounds’ pressure destroys its nutritive value. The decline in weight, however, was not as rapid as in the case of the - growing rats in their experiments. All the experimental mice were fed 2.5 gm. a day and took the food well, except in the declining stages, when the food was much scattered. The addition of cystine to the ration of Mouse 81 2 on the 188 Nutritive Value of Diamino-Acids 42nd day of observation exerted no beneficial effect, the mouse declining in weight steadily and dying on the 52nd day. A simultaneous addition of cystine to the ration of Mouse 8la 9 was accompanied by a slight increase in weight, and maintenance of weight for 3 weeks, although ultimately this animal died also. With Mouse 88 @ a change to a ration containing unheated casein on the 42nd day caused a rise in weight for 3 weeks, but for reasons unknown a subsequent decline and death on the 101st day ensued. A similar change in the ration of Mouse 88a 9 did not materially check the decline of this animal, which died on the 98th day of the test. The cause of the nutritive failure in these latter two cases may have been due to the unusually warm weather prevalent at that time. Experiment 6.—The mice from Experiment 3 were placed on a ration in which the purified casein was replaced by a hydrolyzed casein preparation digested for 2 months with pancreatin, and further treated as described above. The product contained 36.2 per cent peptide nitrogen, but did not give the biuret test, showing that the peptides were of the very simplest kind. The ration fed was made into a biscuit form and contained the following materials: Ration 6. per cent Hydrolyzed casein (2)mionths) 50.2... 2. oo. sey oe See eee 10 STC OSC iii eee ee nhl ie ae oS Su ee 5 Burtber fat seo ne IOS ce Se Oe De: Protein-=freewmilke 2 oc. ke sb aew oe na oes Aa eS ER ee 28 fl BY ¢ 1 21 0 ern Seo ny SE ee Pe ne a PON RN rtp mt 52 The sucrose was added partly to overcome the sharp taste of the hydro- lyzed casein. Later the sucrose was increased to 10 per cent. It is to be noted that the hydrolyzed casein in this experiment was not supplemented with tryptophane. The results obtained are given in the following table. A decline in weight will be noted in all cases, indicating that the hydrolyzed casein lacked or was deficient in some essential constituent. This substance was either destroyed or removed from the digested casein, either in the undigested portion or in the precipitate from the nine and a half volumes of alcohol. Another possibility is that it was not present in sufficient quantity relative to the actual food intake to satisfy the animals’ require- ments. This latter statement appears to be the most probable one, for some mice did maintain themselves for a while, notably K. M. K. Geiling 189 Days. | No. 30 9. |No. 71b 2. | No. 72 2. | No. 72b 9. | No. 7la 9.} No. 77 9. | No. 35a 9. gm. gm. gm. gm. gm. gm. gm. een 23.9 23.6 18.5 18.2 23.8 Pal abs GON 22220 23 .3 21.2 LES 17.0 21.4 18.5 14 || 25.7 23.5 22.0 LZ0s Iino 20 .8§ 18.4 21 | 24.5 22.0 20.5 16.0 15.0 20.0 18.0 28 | 24.0 2153 22.0 15.2§ 14.6§ 19.5777) 91822 35,|. 23.3 21.7 2ailie T4-8***) Tae Asse 20 0 Ves 42| 23.2 19.5* 20.0 20.1 18.6 49 | 22.5* 20.0 19.5 20.2 Died 56 | 21.5 1S. G2 | S71 5% 18.2 42nd 63 | 20.5 18.2 19.2 16.8 day (00 2028** "| 19.3 20.0 15.9 hi 242 20.2 2101 G20 84 | 24.6 ZO RT| A 2a ee 17.8 91 | 24.5 Wied 17.6 70 93). 24.4***| 20.3 Died 17.0 105°) 22.0 20.0 97th 112-|_ 23.0 19.3 day 119) |) 23-6 20.07 126} Died 19.5 133 | 119th 17.5 140 day. 16.0f * Increased sucrose to 10 per cent. ** Added cystine to the ration, 1 per cent. *** Changed to “hydrolyzed casein preparation’ made from casein digested for 34 months and containing 24.25 per cent peptide nitrogen, negative biuret test. + Changed to a ‘‘monoamino-acid mixture’’ plus cystine, in place of the hydrolyzed casein ration. t Changed to “‘monoamino-acid mixture’? plus the diamino-acids from gelatin and cystine. The animal refused food and was placed on the preliminary ration. § Added flowers of sulfur to the ration. Mice 30 @ and 72 9, and then declined slowly, possibly indieat- ing that the supply of this essential substance was slowly being depleted. The fact that casein is low in cystine led to the suspicion that the de- ficiency might be due to the lack of this amino-acid; consequently about 1 per cent cystine was added to the ration fed to Mice 30 9, 71b 9, and72 9. A marked change was noticed immediately; Mouse 30 2 increased more than 3 gm. in weight within 1 week after the addition of cystine and main- 190 Nutritive Value of Diamino-Acids tained this level for 21 days. The other two mice, 71b 2 and 72 9, in- creased from 18.6 and 18.7 to 21 and 21.7 gm. respectively, in the course of 28 days. A noticeable fact was that the mice were losing hair in the later stages on Ration 6, but this ceased as soon as cystine was added to the ration, and within a short time the mice had sleek coats again. These data indicate rather clearly that cystine was the missing constituent of the hydrolyzed casein, and lend support to the conclusion of previous investigators that cystine must be regarded as an essential amino-acid, both for maintenance and growth. Mice 30 9, 71b 9, and 72 2 were then placed on a hydrolyzed casein ration prepared by digestion with pancreatin for 3} months instead of 2 months. Mouse 72 9 died, after having been on the ration for 13 days. Mice 30 2 and 71b @ did not take to the food at first, but after a week they ate the ration fairly well. Mouse 30 9 maintained its weight for 3 weeks, when it died. Mouse 71b °, after maintaining its weight for 5 weeks, was changed to a ration made up as follows: Ration 7. per cent ‘‘Monoamino-acid mixture’’ prepared as described above...... 9 Gy BtIMes. S..0 eet Beetle: ee Rs are a RSA eS if Buttertat. sao. mca... Setees. 7 Rect Ponte alc Se ee 10 Protem-=tree milky 22.8, ce tanecet cis erst oe eee ee 28 Dext ian: 2 eee kc 2. ae oh a bya Reo ae aegis See 42 A steady decline set in, although the food consumption was good. The mouse dropped in weight from 20.0 to 16 gm. in 3 weeks, indicating that the material removed by precipitation with phosphotungstic acid in acid solution is either wholly or in part necessary for the maintenance of adult mice. Further experi- ments will be reported later confirming this point. The behavior of this mouse is interesting. It lived on hydro- lyzed casein, containing only amino-acids and the simplest pep- tides as the sole source of nitrogen, for 140 days. To ascertain whether the animal body is able to synthesize cystine when flowers of sulfur are added to a ration deficient in cystine, three mice, 7.e., 72b 2, 7la 2, and 77 9, after a rapid decline in weight on Ration 6, had flowers of sulfur added to their ration instead of cystine. The loss in weight continued, showing that the animal body is not able to utilize sulfur to synthesize cystine. These animals were then placed on a ration made from casein digested with pancreatin for 3} months instead of 2 months; Mice 72b 2 and 7la 2 refused food and were placed on the preliminary E. M. K. Geiling 191 ration. Mouse 77 2 maintained itself for 3 weeks and was then changed to the monoamino-acid ration (Ration 7). On this ration the mouse rapidly declined in weight for 4 weeks. The monoamino-acid mixture was then substituted by hydrolyzed casein digested for 3} months. Loss in weight ceased, and the animal maintained its weight for 3 weeks. The results secured with this animal also lend support to the conclusion regarding the indispensability of the diamino-acids for maintenance. Mouse 35a @, after maintaining its weight for 5 weeks on Ration 6, died suddenly. The greater nutritive efficiency of the casein hydrolyzed for 34 months as compared with that hydrolyzed for 2 months seems clear from the data of these experiments, though an explanation of this relation is difficult. The food consumption of the experimental animals was quite satisfactory, and declines in weight cannot be attributed to too low a food intake. Each animal was offered daily about 2 gm. of food which was usually consumed, except where noted. Several other mice were placed on hydrolyzed casein, but refused food and consequently lost weight rapidly. These results are not reported here. Experiment 7.—Another experiment was started in which the nitrogenous portion of the ration was made up of the monoamino- acid mixture from casein supplemented with the diamino-acids from gelatin plus cystine, histidine, and tryptophane. The food mixture was made into a biscuit form from the following materials: Ration 8. per cent Mionoannmo-acids from: CASEIN... 2.26... ee ee ese eee ese 8.0 Miamine-acids trom gelatin. .....0.......5......0..aa0n08-: 1.5 (CHARLTON « Sle ds OS COE ce ea rarer 0.25 VE GUSREGUIINES So Soe aca SE ee RU ae SS gE a 0.25 Mleracra Goren CMe ere eat t.padsvaisie) tocrare Ce oe sislvla"sve oe eu encio tala due ee @ 0.50 SUCRE ood Gere 0 Cae ORES ee eee 10.00 BU heTeia eet evel Pe eMter tly ki iirortheve wotsis-s cede dem oteetea sas ard 10.00 Protein-free milk............ Sore ee URE A A See 28.00 IDYesqtimai: Jo 3 Shc 5 35 OL YS gee See Se eee oe eee 42.00 192 Nutritive Value of Diamino-Acids Four mice received this ration; three refused it, but the fourth, 90 9°, ate it in sufficient quantities after 14 days, maintaining its weight from that time for 42qdlays. Below are given the changes in weight. Days. No. 90 2. gm. 1 . 27.5 i 24.5 14 22.0 21 20.2 28 21.0 35 20.7 42 20.8 49 20.8 56 23)2 This experiment, although carried out with only one mouse, tends to show that the ration fed is adequate, provided the animals take sufficient of it; furthermore, it helps to overthrow the ob- jection which might be raised that the monoamino-acid mixture lacked some essential monoamino-acids other than tryptophane and that the losses in weight obtained in Experiment 6 might also be due to their absence. Here maintenance was secured with a ration composed of the monoamino-acid mixture, supple- mented with the diamino-acids and tryptophane, showing that the loss in weight in the case of Mice 71b 2 and 77 @ on the mono- amino-acid mixture alone was due to the absence of the diamino- acids. Further proof of this will be given in the next experiments. Experiment 8.—To improve the taste of the rations and increase their calorific value, in this and subsequent experiments, more butter fat and lard were added. Furthermore, instead of mixing all the ingredients together and rubbing them into a paste, the following procedure was adopted. The lard and butter fat were melted on the water bath, and the amino-acid mixture was stirred into the melt. The starch, protein-free milk, sucrose, and lactose, previously well mixed, were then added, and the whole mass was well stirred. By adopting this method, apparently, the food mixtures were made more palatable, and the mice took more of the food although some difficulty was experienced in a few cases. The following ingredients constituted the ration used in this experiment: E. M. K. Geiling 193 Ration 9. per cent Monoamino-acid mixture from casein....../...........06-- 115 Cation ten saeco oe sg aloe oynsel os bv gees Seeale ober a 0.25 GEER OREL OY 8) GN OND RA es ae) ea hed ean 0.25 STICTOSE eT ee ree hee arck het hd seia aye aussi ws elo Se au eecree MSR ve oe 12.00 IE GY GLOSS BV 5 cet RUN HERES aro ee SE ae Ce CRS ree ie 00 PRO aNeS TemllliG, oo ak Raia Be Hen ob eee Eocene dud ocadanc 28 .00 SEDI Re ONS 6, Se dia PO TS i, bre a ea nC SORE Cot ae 14.00 TBYOTR RESP. TEN He t,o eat ieee aoc nel en ea ree ee a ge 18.00 Deal eee eee eS et yereeetc ce cpa seen ial Suc Aenean soneeattetscest 10.00 Before being placed on this mixture, the animals used in this and sub- sequent experiments were fed for about 2 weeks on a ration composed of 33 per cent skim milk powder, 18 per cent butter fat, 10 per cent lard, and 39 per cent starch. Five mice were placed on Ration 9 with the following results. Days. No. 208a 9. No. 203 9. No. 204a 2. No. 205a °. No. 205 9. gm. gm. gm. gm. gm. 1 24.0 23.2 31.0 22.0 22.0 7 20.0 18.6 22.0 18.0 18.0 14 18.6 1657 20 .0* We0F NO 21 ty 0* 16.7 18.07 1725 20.5 28 20.7 18.7 19.0 22.5 35 20.8 18.5 19.5 21.0 42 21.0 VAOT 18.57 19.57 * Changed to Ration 10: 9.25 per cent monoamino-acid mixture, 2 per cent diamino-acids from gelatin, 0.25 per cent cystine, and 0.5 per cent tryptophane. { Refused food and placed on preliminary ration. The animals were offered 2.5 gm. of food each day and took it satis- factorily, except where noted. It will be observed that there was a rather large decline in weight between the Ist and 7th days; thereafter the loss was gradual. Mice 208a @ and 205 9 made very pronounced gains when the diamino-acids were added. In the case of Mice 203 2 and 205a 2 the gain in weight was smaller and more gradual. The experiment was dis- continued after 42 days, because the animals were not taking the food well, and it was felt that the loss in weight of the experimental subjects was due to the fact that they were not consuming sufficient food to satisfy their energy requirements. ‘ The results of this experiment, together with those of Experi- ments 6 and 7, point rather clearly to the conclusion that all or 194 Nutritive Value of Diamino-Acids some of the constituents of hydrolyzed casein removed by the precipitation with phosphotungstic acid in acid solution, 7.e., the diamino-acids, are essential for the maintenance of adult mice. The fact that a similar ration with only 2 per cent of the mono- amino-acid mixture replaced by the diamino-acids from gelatin caused the mice to increase in weight and maintain themselves shows rather definitely that the loss in weight on the monoamino- acid food mixture was due to the absence of the diamino-acids, and not to the removal of some essential monoamino-acids which may have been in peptide combination with the diamino-acids. Experiment 9.—Having shown that either:all or some of the diamino-acids are necessary for the maintenance of adult mice, the next step was to ascertain which of the diamino-acids are esssential. In Experiment 6 it was shown that cystine was necessary; hence in the present trial it was added to all the rations employed. The rations of this experiment were similar to those used in Experiment 7, except that the nitrogenous portion was supplied by the monoamino-acid mixture plus cystine and tryptophane and one of the three diamino-acids, arginine, histidine, or lysine. Four mice were placed on a diet in which cystine and histidine were the only diamino-acids present. Two of the subjects refused the food and were taken off the experiment; the other two, 203a 2 and 203 @, did well on the ration and maintained their weight for 28 days. The food mixture was made up as follows. Ration 11. per cent Monoamino-acid mixture from casein.............-.2..000% 10.75 Cystine 1.5 Sere eee ee ee Ba ae ee 0.25 ‘Tryptophan oe os ee ee eee es eee ee 0.25 ERI SHIGITIO® 20. hn eae eer SNE Acie Lea rae Nee a Te 0.75 Sieroses sii o0. se Ue ee ee Petey ee ede eect, Se eens Sie eh eae 12.00 TSACE ORE Ake lack Cath ee Oe clk RRR ne PR LS 6.00 SEAT CIN Aah Nip: Poe eye i eee ET ee et top ae 14.00 Protein-tree mille: 245th ee ees Re bce ei eee 28 .00 EUG Seca Nate act te ae SE AY ee Sk tae 10.00 BUbtbertabes ] coher cote ee be eerie oleh cae eatin eee 18.00 After having maintained themselves on this ration for the period men- tioned above, the histidine was replaced by an equal amount of arginine, and two more mice, 9la 2? and 91 9, were placed on this second ration (Ration 12). The four animals maintained themselves satisfactorily E..M. K. Geiling 195 for 21 and 28 days, and thereafter the arginine was replaced by the same amount of lysine (Ration 13). Four other mice, 205a 2, 205 2, 207a °, and 207 2, making a total of eight, were fed this ration, and all declined in weight. Mouse 207 2 lost weight much more rapidly than its mate. After it was on the lysine ration for 28 days and had declined from 20.3 to 13.0 gm. in weight, the lysine was replaced by arginine, resulting in an increase in weight to 14.5 gm. Thereafter the arginine was again replaced by lysine, when decline set in once more. The eight mice were then changed to a diet in which the lysine was replaced by histidine (Ration 11), but unfortunately the animals did not take the food very well. However, they consumed sufficient food to maintain their weight for 14 days. The significant point about this part of the experiment is that as soon as histidine was added the mice ceased losing weight, and although they did not regain much weight, they nevertheless maintained themselves. The animals were then divided into two lots of four each; Lot 1, made up of Mice 205a 2, 205 2, 207a 9, 207 9, were fed the ‘‘monoamino-acid mixture’? prepared from the enzyme digest, supplemented with cystine, tryptophane, and histidine (Ration 11). These animals consumed suffi- cient of the food to maintain their weights fairly well for 14 days, but the food intake was not very satisfactory. The other lot of mice, 203a 2, 203 9, 9la 9, and 91 Q@ were offered a similar ration, except that the monoamino-acid mixture was prepared from an acid digest, and contained no peptides. This lot refused the food and, being in a low state of nutrition, two of the animals died (91a 2 and 91 9 ); the other two were placed on the preliminary ration. The changes in weight of the mice in this experiment were as follows. 196 Nutritive Value of Diamino-Acids Weight. Weight Weight. Weight. | | — on es ——————— es a - Food ot : Food OF : ; : Food ia |e |e 2 2 lee) @ [mee Sol egies Bh eakwlluphe a aunes i ces 80 wes Al 4 vl Z a A vA a Z | gm. gm. gm. qm gm qm. gm. qm. gm. qm. gm. | gm 1/24.0 422.08 22) 57120032 23 .27/20.8 27. b4\238752 7|22.5 117.0 24 22.5 |16.5 |20 19.5 18.2 |25.5 {21:0 |20.0 /18.0 14/22.3 |17.5 32 21.7 \15205)/18 19S OF L55. 121 5 \ | QOES Rie 21/221 15.0 3 20.5 {18.0 |17.5 |20.89)16.7 942.57:9|22 5 120.0 |22.5 28/21.7 |16.0 34.5 |21:0 113.0%22:5 |18.0 |15.5 117-5 — |20:07120 17/3520" 35122.5117.51 (32.5 |20.0 }138.7 |17.5 |17.55)14.7 419.0 18.5 {16.0 |23.0 42)/22.5 |17.5 25.5 119.0 |14.52/18.5 |17.5 115.0 [16:0 16.4415.5417.5 49/23 .5 |18.7 35.0) |18.2 11208 17.5) 16.7 fo oglivied ioe 56|23.7 |18.8 35.0 {18.0411 .7 418.5 1.588 63)23.5? 18.52 |35.0 {18.0 {11.7 |25.0 70}21.0 |16.0 35.0 {18.5 |11.0 |20.8 7718.54|15.24 {23.6 1.358 84 14.0% 6/23.5 91 14.5 2.188 | 1 Monoamino-acid mixture (enzyme digest) + cystine + tryptophane + arginine (Ration 12). 2 Monoamino-acid mixture (enzyme digest) + cystine + tryptophane + lysine (Ration 13). 3 Monoamino-acid mixture (enzyme digest) + cystine + iryptplae +. histidine (Ration 11). 4 Monoamino-acid mixture (acid digest) + cystine + try ptOSEEee = histidine (Ration 14). 5 Food intake per week for two mice. 6 No accurate account kept of food intake. 7 Scattered food badly. 8 Average daily food per mouse. ° 9 day period. From these data it is obvious that histidine and arginine are interchangeable and that either one of them in the presence of cystine is able to support maintenance for the periods given. Maintenance in the absence of both isnot possible, thus confirming the work of Ackroyd and Hopkins,® * except that their experi- ments were conducted with growing rats and the ones under discussion with adult mice. Whether lysine is dispensable cannot be so definitely concluded from these trials, although the indica- tions point strongly in that direction. Allowance must be made for E. M. K. Geiling 197 the possibility that there may be some lysine in the nitrogenous portion of the protein-free milk which was used in all the rations. The amount present at most must be very small, but even the presence of small quantities of an essential substance in a ration is sometimes quite significant. Should lysine not be necessary for maintenance, an excellent illustration would be on hand show- ing the qualitative difference between the requirements for growth and for maintenance; in the former case this amino-acid is essential. In this connection may be mentioned the work of Osborne and Mendel?’ with zein supplemented by tryptophane as the only source of nitrogen, except that contained in the protein-free milk. A rat maintained its weight on this ration for over 180 days. This finding is in harmony with the results reported here, as no in- vestigator has as yet reported the presence of lysine in zein. The behavior of Mice 203a 9 and 203 @ is particularly inter- esting. These animals were started on a diet with histidine and cystine as the only diamino-acids and maintained their weights for 28 days; arginine was substituted for histidine and mainte- nance continued for 28 days more, but when the arginine was removed and replaced by lysine loss in weight ensued immediately. SUMMARY. 1. Synthetic rations, adequate for the maintenance of adult mice, have been compounded with purified casein and with casein hydrolyzed with enzymes and containing only amino-acids and the simplest peptides, as indicated by a negative biuret test. Except for the presence of a small amount of nitrogen in the pro- tein-free milk used, this hydrolyzed product formed the sole source of nitrogen in the food mixture. 2. Casein suspended in boiling water for 2 hours does not lose its nutritive value. 3. Heating casein for 1 hour in an: autoclave at 15 pounds’ pressure impairs or possibly destroys its nutritive value. This is in agreement with the findings of McCollum and Davis. 4. If the diamino-acids are removed from hydrolyzed casein with phosphotungstic acid in acid solution, the residual amino- acids are inadequate for the maintenance of adult mice. This is not in agreement with the findings of Henriques and Hansen, but confirms those of Ackroyd and Hopkins. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. l 198 Nutritive Value of Diamino-Acids 5. Cystine appears to be necessary for the maintenance -of adult mice. 6. Arginine and histidine seem to be interchangeable in nutri- tion. Full grown mice are able to hold their weight, when either one of them, together with cystine, is present in the ration. In the absence of both, loss of weight results. This is in harmony with the work of Ackroyd and Hopkins in the case of growing rats. 7. Lysine does not appear to be necessary for the maintenance of adult mice. The author desires to thank Professor H. 8. Grindley for his interest in and general supervision over this:work. His special thanks are also due to Dr. H. H. Mitchell, at whose suggestion this work was undertaken, for his constant interest and helpful suggestions. BIBLIOGRAPHY. 1. Mitchell, H. H., J. Biol. Chem., 1916, xxvi, 231. 2. Henriques, V., and Hansen, C., Z. physiol. Chem., 1904-05, xliii, 417. 3. Osborne, T. B., and Mendel, L. B., Carnegie Institution of Washington, Publication 156, 1911, pt. 1, 12. 4. Abderhalden, E., Z. physiol. Chem., 1915, xevi, 1. 5. Hopkins, F..G., J. Chem. Soc., 1916, cix, 629. 6. Ackroyd, H., and Hopkins, F. G., Biochem. J., 1916, x, 551. 7. Abderhalden, E., and Einbeck, H., Z. physiol. Chem., 1909, 1xii, 322. 8. Abderhalden, E., Einbeck, H., and Schmid, J., Z. physiol. Chem., 1910, Ixvill, 395. 9. Kowalevsky, K., Biochem. Z., 1910, xxiii, 1. 10. Osborne and Mendel, J. Biol. Chem., 1914, xvii, 332. 11. Osborne and Mendel, J. Biol. Chem., 1916, xxv, 1. 12. Buckner, G. D., Nollau, E. H., and Kastle, J. H., Am. J. Physiol., 1915-16, xxxix, 162; Kentucky Agric. Exp. Station Bull. 197, 1916. 13. Osborne and Mendel, J. Biol. Chem., 1916, xxvi, 293. 14. McCollum, E. V., Simmonds, N., and Pitz, W., J. Biol. Chem., 1916-17, XXvill, 483. 15. Osborne and Mendel, J. Biol. Chem., 1914, xviii, 11. 16. Myers, V. C., and Fine, M. S., J. Biol. Chem., 1915, xxi, 389. More definite results in the same direction have been recently obtained | by W. H. Thompson (J. Physiol., 1917, li, p. 11) experimenting with rabbits. When arginine was injected intravenously, part of its guanidine residue, namely from 8 to 25 per cent (average 14.5 per cent), was utilized to form muscle creatine. Also, in all experi- E. M. K. Geiling 199 ments there was a marked increase in the creatinine of the urine during the hour immediately following the injection 17. Osborne and Mendel, J. Biol. Chem., 1915, xx, 351. 18. McCollum, E. V., and Davis, M., J. Biol: Chem., 1915, xx, 645. 19. Osborne and Mendel,’ pt. 11, 80. 20. Mitchell, H. H., and Nelson, R. A., J. Biol. Chem., 1915, xxiii, 459. 21. Van Slyke, D. D., J. Biol. Chem., 1911-12, x, 15. 22. Plimmer, R. H. A., The Chemical Constitution of the Proteins, London, 2nd edition, 1913, pt. 11, 52. _ 23. Denis, W., J. Biol. Chem., 1911, ix, 369. ° 24. Folin, O., J. Biol. Chem., 1910, viii, 9. 25. Wheeler, R., J. Exp. Zool., 1913, xv, 209. 26. McCollum and Davis, J. Biol. Chem., 1915, xxii, 247. . Osborne and Mendel, J. Biol. Chem., 1915, xx, 378. STUDIES ON ENZYME ACTION. XV. FACTORS INFLUENCING THE PROTEOLYTIC ACTIVITY OF PAPAIN. By EDWARD M. FRANKEL. (From the Harriman Research Laboratory, The Roosevelt Hospital, New York.) (Received for publication, May 25, 1917.) The plant protease, papain, has been the subject of numerous studies. The earlier literature on the subject has been reviewed in some detail by Mendel and Blood (1910), whose paper embodies the results of one of the few careful investigations of this ferment. The work of these authors is, however, almost entirely qualita- ‘tive in character. In extending the general plan of enzyme study that has been undertaken in this laboratory, it seemed advisable to reinvestigate some of the facts concerning this enzyme, em- ploying some of the more accurate methods that have been re-_ cently developed. This ferment lends itself very well to chemical study since it may be obtained in large amounts of fairly uniform activity. In this paper some observations concerning the purification of the active material and a consideration of the influence of acidity and the quantitative relationship between enzyme and substrate will be presented. The peculiar behavior of H&N in accelerating the action of papain has been studied further. Purification of Papain. In view of the method by which commercial papain is produced,! it seemed desirable that a more refined material be used in this work, especially since some of the contaminating material might have a deleterious action on the enzyme. It also might be 1 Pratt, D. S., Philippine J. Sc., 1915, x, 1. 201 202 Proteolytic Activity of Papain expected that a refined material would be more uniform in com- position and would therefore make the different experiments more comparable. The papain as obtained from Parke, Davis and Company was a light brown finely divided powder nearly all of which was soluble in water, giving a yellow to brown solution. The insoluble material settled rapidly and the solution could readily be re- moved from it by decantation. A few preliminary experiments were carried out with a view to determining to what extent jurification of the active material could be accomplished. 15 gm. of papain were ground up with 500 cc. of distilled water and allowed to stand over night. The next morning a portion of the suspen- sion was filtered through asbestos and treated as follows. A. 50 ec. of the filtered solution were treated with 150 cc. of acetone and the mixture was allowed to stand 2 hours. The precipitate that formed was centrifuged off and the mother liquor decanted. The precip- itate was then taken up in 100 cc. of water. This is referred to as Solution AS B. 50 ce. of the filtered solution were added to 100 cc. of acetone and then treated as in A, the water solution of the resulting precipitate being Solution B. C. 50 ce. of the filtered solution were added to 250 cc. of 95 per cent alcohol and then treated as in A, the water solution of the resulting pre- cipitate being Solution C. As controls, the original filtered and unfiltered solutions diluted with an equal volume of water were used. These solutions were D and E respectively. To test the activity of the solutions 5 cc. were allowed to act on 25 ce. portions of 1 per cent gelatin containing 0.2 per cent of tricresol as a pre- servative. The solutions were incubated at 37° for 17 hours and the increase in ‘‘formol’’ titration over the blanks was used as a measure of the activity. Formol titration. Solution. Substrate blank. | Ferment blank. erent: Action. A 2.65 0.35 4.00 1.00 B 2.65 0.30 4.00 1.05 C 2.65 0.45 4.90 1.80 D 2.65 1.00 4.60 0-95 E 21200 1.10 5.00 1.25 E. M. Frankel 203 From the above experiment it is apparent that the active material may be concentrated by the precipitation with two to three volumes of acetone or with five volumes of 95 per cent alcohol; the latter procedure in addition to effecting a concen- tration of the active material seems to remove some of the in- hibiting contaminants. It is also to be noted that the precipita- tion removes from the active material a relatively large amount of substance giving a formol titration. It is possible to fractionate commercial papain in this way so that about two-thirds of the ma- terial can be removed without appreciably impairing the activity of the remaining material. Using the information obtained in the above experiments, about 50 gm. of papain were purified for use in the experiments given below. 150 gm. of commercial papain were rubbed with 4,500 cc. of water and allowed to stand over night. The next morning 4,000 cc. of clear solution were siphoned off and without filtering were poured into 8 liters of acetone. The precipitate was allowed to settle. After standing 4 hours the clear supernatant fluid was decanted off and the precipitate filtered and washed with acetone. The precipitate was finally drained on a large Buchner funnel and then rubbed up with 800 cc. of warm water, and the turbid brown solution allowed to stand 36 hours in a tall cylinder, a layer of tolu- ene acting as a preservative. After standing, the clear supernatant liquid was siphoned off and poured into 4 liters of 95 per cent alcohol, and the precipitate filtered on a Buchner funnel. The filtration proceeded very slowly, taking 24 hours. The precipitate was rubbed with 95 per cent . alcohol and then with ether, and dried after filtration in a current of air. The drying was rather unsatisfactory, the material becoming light brown in color. The activity of the material was, however, very high, so that it is certain that the ferment is fairly stable in aqueous solution and pre- cipitated in the presence of acetone, alcohol, and ether. This observation is contradictory to some of the statements that appear in the literature regarding the deterioration of papain. Other experiments also confirmed the conclusion reached here. Papain allowed to stand over night at 37° seems to show little if any deterioration. Dialysis of the ferment in collo- dion bags results in a certain loss of activity, the bag contents becoming less active while the dialysate becomes slightly active, the sum of the two or the combined action of both being less than that of the untreated aque- ous solution that stood under the same conditions. The deterioration is accelerated by dialyzing at 37°. 204 Proteolytic Activity of Papain Optimal Hydrogen Ion Concentration for the Papain Action. Inasmuch as most ferments seem to have a definite range of acidity or alkalinity in which they exhibit their maximal activity, it seemed strange that papain should, as stated in the literature, act equally well in acid or alkaline solution. To throw more light on this point a series of experiments was undertaken to determine at which hydrogen ion concentrations papain was most active proteolytically. The data recorded below are typical of the results obtained in different experiments so the conclusion seems justified that papain, in common with other ferments, has an optimal hydrogen ion concentration, in this case approximately 10-> x. In all cases, the indicator method was used and the results are therefore not more accurate than a half a unit in the pH”. In the presence of proteins the indicator results are not entirely to be relied upon except for comparative purposes. The absolute hydrogen ion concentrations of the various solutions used cannot be given with certainty. A 2 per cent solution of gelatin. in water was treated with HCl and NaOH so that the solutions when tested with suitable indicators showed that they were of the hydrogen ion concentration desired. A 0.5 per cent solution of purified papain was divided into three parts and adjusted to 10%, 10°, and 10-°Nn. 25 ce. portions of the various gelatin solutions were measured out and treated with 5 ec. of the papain solution of the same range of acidity. Duplicate blanks were set up with the papain and the gelatin and triplicate mixtures were made containing the protein and ferment. Proteolytic Action of Papain at Various Hydrogen Ion Concentrations. pH. Action. Initial. 3 Final. 2 2 0.50 3 3.5 1.45 4 + 4.95 5 5 5.39 6 6 3.35 U 6 2.55, 8 6.5 1.50 9 a 1.25 2 The symbol pH is used interchangeably with the term hydrogen ion . concentration and denotes numerically the negative exponent of 10. E. M. Frankel 205 Of the latter, one was used to test the hydrogen ion concentration before and after the flasks were allowed to stand in the incubator. The period of incubation was 22 hours. The results given in the column under ac- tions are the formol titrations in cc. of 0.1 N alkali after correcting for all blanks. The above experiment shows fairly conclusively that papain exhibits its greatest activity at an acidity equal to the concen- tration of the hydrogen ion of 10~° nN; 7.e., slightly more acid than is necessary to cause methyl red to change from yellow to red. The method of experimentation used above gives results which are entirely in accord with those obtamed when the rate of cleavage of gelatin and egg white is followed at different hydro- gen ion concentrations. It is interesting to note the changes in hydrogen ion concentration that occur during the proteolysis. In those cases either side of the optimum acidity, the tendency is for the solution to become more acid or alkaline, apparently tending to bring the solution to the optimum acidity. This is rather peculiar in view of the fact that at all times the solutions contain a large quantity of material that might act as buffer. In fact as the digestion proceeds the buffer action should become more marked. since a greater number of amino and carboxyl ‘groups are present. The only explanation that is apparent at present must involve an assumption that postulates two different types of cleavage products, depending on the hydrogen ion con- centration. In one case we must assume the liberation of a pre- ponderance of basic amino-acids or peptides, in the other an excess of acid compounds. Having found that there was a definite hydrogen ion con- centration at which papain was most active proteolytically, the question to what extent the ferment was decomposed on standing with acid and alkali was raised. To throw light on this point, papain solutions were treated with different strengths of acid and alkali and then neutralized to methyl red (hydrogen ion conc centration 10-° N) and allowed to act on gelatin. The actions were compared with those of the untreated solution of papain at the same hydrogen ion concentrations. Suitable blanks for enzyme and substrate were run and the results corrected for them. Toluene was used as a preservative. 206 Proteolytic Activity of Papain Influence of Acids and Alkalies on the Proteolytic Activity of Papain. Concentration. ee Action. Remarks. 7 hrs. OVS NI SCIOS: fon ck ss co ee eee 4 0.10 | Incubation with Od ios ee eee 4 0.25 gelatin 41 hrs. Water. fhm. ss oe ee eee 4 4.95 0-1 Nialkalit 2. ee eee 4 0.20 02:56 2Shs Ac, 10 aoe Ee eee ee 4 0.05 O05 Neacids..ac. cache eee 1 2.30 | Incubation with OOD eee a ei ee ee eee ek 1 2.90 gelatin 18 hrs. Water:: coc tae ee ee 1 4.10 O202:nalkslite seen se cee ae 1 3.65 1 2.85 Oe ee pera teas sec eee The above experiment shows that the ferment is sensitive to both acid and alkah, the latter being less destructive. Quantitative Relationships between Papain and Its Substrate. In studying the changes that occur when an enzyme and its substrate react it is evident that while the enzyme is affecting the substrate, the latter is modifying the activity of the enzyme. It has been stated in the literature of the subject that the pro- teolytic activity of enzymes follows the simple mass action jaw, this conclusion being deduced from the study of the kinetics of hydrolysis. There are several reasons why it is perhaps a fruit- less task to try to formulate a statement of the kinetics of the reaction involving the enzymatic cleavage of a protein. First, the system involves two colloidal components and it is therefore unlikely that solution kinetics will apply, but instead adsorption phenomena may be the basic factors (Nelson and Vosburgh, 1917). Second, the cleavage of protein does not represent a reaction where one stage is completed before another begins but rather a complex of a number of simultaneous reactions. In order to determine what role the relative quantities of enzyme and substrate play in the action of papain on protein the follow- ing experiment was carried out. A series of solutions of gelatin of definite concentration were treated with acid until their hydrogen ion concentration as indicated colorimetric- | ) E. M. Frankel 207 ally was 10-5 n. Similarly a series of papain solutions were prepared. Mixtures of these as indicated in the table were incubated at 37° for 24 hours and the extent of cleavage was determined by the formol titration. Toluene was added as a preservative. All the data recorded are corrected for suitable blanks on enzyme and substrate. Proteolysis with Varying Concentrations of Enzyme and Substrate. Papain. ; Gelatin. mq. mg. mg. mg 125 250 500 750 Formol titration. mg 5 1.70 2.65 3.25 3.55 10 2.10 3.50 5.05 5.90 25 2.25 4.40 7.30 8.95 50 2.45 4.70 8.55 11.00 The results of the above experiment lend support to the view that in the cleavage of protein by papain there is a two stage reaction, the first involving a combination of enzyme and sub- strate, and the second the cleavage of this intermediate compound to give the enzyme and the split products of the protein. It will » be seen from the curves that when the amount of the substrate present is relatively small, the proteolysis is not proportional to the enzyme concentration but tends to a definite point, the _ addition of further enzyme producing little additional cleavage. In the case where the ferment is kept constant, the proteolysis depends on the ratio of substrate to ferment. If the ferment concentration is large the ‘“formol’’ titration after proteolysis is almost directly proportional to the quantity of substrate. With smaller substrate concentrations the action is dependent on the concentration up to a certain point, after which the addi- tion of more substrate causes little more action. These findings would indicate that a given quantity of enzyme can handle a given quantity of substrate after which the addition of either com- ponent leads to no further action. This brings the action of papain into the same class as urease, invertase, and lipase. In considering the relations of enzyme and substrate, it appeared of interest to determine to what extent the addition of more enzyme 208 Proteolytic Activity of Papain and substrate to a digestion mixture would affect the results. The results of the experiments are summarized in the following table. In the table indications are made of the amounts of fer- ment and substrate added on the Ist and 2nd day and the ‘“‘for- mol” titrations of the resulting digestion mixture at the end of the 2nd day recorded under Action, the figure given being corrected for all blanks. Influence of the Addition of Ferment and Substrate to a Digestion Mixture. 1 per cent 5 per cent papain solution.|gelatin solution. No. = 5 Action. Remarks. PAG ee ein lees . 3 < cs) iss a 7 _ = = A = a cc. cc. ce. cc. 1 5 0 5 0 | 2.40 2 5 5 5 0 | 2.55 | Slight increase due to additional quantity of ferment. 3 1) 410 0 5 0 | 2.70 | Twice the original quantity of fer- ment gives only slightly larger action than No. 2. 4; 5 0 5 5 | 3.50 | Addition of more substrate shows that active ferment remains but that action on 2nd day is smaller than on Ist. 5 5 0 10 0 | 5.30 | Twice the original quantity of sub- strate gives much larger actionthan No.4. Time factor may playa réle. 6 5 5 5 5 | 4.35 | Two individual additions of enzyme and substrate do not cause twice | the action in No. 1, due possibly | | toaretarding effect of the products. These experiments also support the view that there is a defi- nite quantity of ferment required for a given amount of sub- strate and that an excess causes little more action. The striking action of HCN on proteolytic activity of papain has been noted by Vines (1903) and Mendel and Blood (1910). The experiments of these authors were for the most part qualita- tive in character and served to show that HCN had a definite role as a specific activator of papain action, altering the char- acter of the reaction to such an extent that more extensive cleav- E. M. Frankel 209 Proteolytic Activity of Papain-HCN at Varying Hydrogen Ion Concentrations with Varying Concentrations of Papain. Experiment 1. | Experiment 2. Experiment 3. Experiment 4. Gelatin: ....... 500 mg. | Gelatin .. 500mg. | Gelatin.... 500 mg. | Gelatin..... 500 mg. ICING S 22 531 HCN SS BS HON s232- 67 fe ICN oe 5a (PAPAIN. a. 2201 PapainereeD) Papain.. 12.5 5 Papen’ i. sl pay Volume...... 35 ce. Volume.... 35 cc. Volume.... -35 ce. Volume.. .... 35 ce. pH. pH. pH. pH. ee Action: = Action.|— 2 | Action: | 2a | Action: Initial.} Final. Initial.| Final. Initia].} Final. Tnitial.| Final. | 2.) | 3.5 10.55 2 2 0.50 3 3.5 | 0.40 o Aron) |e 150 3 3.5 1.90 4 4.5 | 3.10 4 |4.5 | 6.00 4 4 9.30 4 4 10.45 5 5 3.90 Stale 6.00 5 5 8.70 5 5 9.70 6 5.5 |.4.85 6/2525" 6205 6 6 8.50 6 6 9.75 7 5.5 | 4.90 ekorOnle Onze te 6 9.40 8 WO | ek O @.0) | 6 8.95 8.5 | 6.5 8.80 PROTEOLYTIC ACTIVITY OF PAPAIN AT VARYING H ION CONCENTRATIONS IN THE PRESENCE AND ABSENCE OF HCN FORMOL TITRATION ron 3 4 5 6 7 8 9 pH CuHart 1. 210 Proteolytic Activity of Papain age of the protein resulted. In considering the nature of the active groupings in papain it seemed desirable to review some of the experiments of the earlier workers, using more modern methods and taking into consideration the factors of acidity and the quantitative relationships between the various components of the reacting system. The optimal hydrogen ion concentration for papain activity in the presence of HCN was determined in the same way as in the case where no HCN was used, the same ferment solution being taken, under much the same conditions. The data above are in marked contrast to those obtained where no HCN was present. Instead of finding a definite hydro- gen ion optimum for papain-HCN proteolysis we find that the ferment is equally active, or nearly so, over a wide range of acidity. The results obtained with lower concentrations of ferment leave the whole matter unsettled. Further work on this point involving more careful measurements of the hydrogen ion concentration by means of the gas chain method is planned. The role of the HCN is not at all clear. From experiments presented below, it would appear that the HCN combines with the papain and the substrate to form an intermediate compound which then undergoes cleavage much as the intermediary compound of papain and protein does. Under such circumstances it may be that the ternary HCN-papain-protein compound has different stability. ; To determine whether proteolysis in the system papain- protein-HCN follows the same general scheme as in the system papain-protein, the following set of experiments were carried out. Proteolysis with Varying Concentrations of Enzyme and Substrate. HCN Constant. Papain. Gelatin. mq. mg. mg. mg. 125 250 500 750 Formo] titration. mg. 5 2.55 4.50 7.50 9.60 | 10 2.75 4.95 8.80 12.10 25 3.10 5.75 10.45 14.45 50 3.35 6.25 11.45 16.25 E. M. Frankel 2AM PROTEOLYTIC ACTION WITH VARYING QUANTITIES OF ENZYME AND SUBSTRATE IN THE PRESENCE AND ABSENCE OF HCN FORMOL TITRATION Substrate Constent.Figures Enzyme Constant. Figures \in- indicate milligrams used in dicate milligrams used in different series. FORMOL TITRATION ioe) 208 30-40. 50 250 500 750 Milligrams of Papain Milligrams of Gelatin CHART 2. The general plan was the same as before. The solutions of gelatin and papain were adjusted to pH 5 and the requisite amounts added to each experiment. The HCN solution used was prepared by dissolving 13 gm. of Kahlbaum’s KCN in normal HCl, enough (200 cc.) of the latter being used to make the solution just neutral to methyl red. 2 cc. of this solution were used in each experiment. The volume of the digestion mixtures was 32 cc. Toluene was added as an additional preservative. The results given are corrected formol titrations. To determine the relation between the extent of proteolysis and the amount of HCN used, the following experiment was 212 Proteolytic Activity of Papain carried out. The same enzyme, substrate, and HCN solution were used as before, the quantities of the first two being kept constant while the third was varied. The results are given below. Proteolysis with Varying Quantities of HCN, Enzyme, and Substrate Constant. Papain 10 Mg., Gelatin 750 Mg., Volume 30 Ce. Formol] titrations. HCN. Titration. Action due to HCN. | Acticn per unit HCN. cc 0 5.30 1 10.75 5.45 5.45 2 12.10 6.80 3.40 5 12.80 7.50 1.50 10 13.90 8.60 0.86 The data presented in the two tables above indicate that pro- teolysis in the system papain-HCN-protein follows the same general laws as were noted above for the papain-protein system. Fixing two components we find that increase of the third tends to increase the total cleavage but not in proportion to the amount added. In fact there seems to be a tendency towards a definite maximum. The only explanation of this phenomenon that is apparent is one which assumes the existence of an intermediary ternary compound in which all three components are present in definite ratio. Any excess of enzyme or HCN over that neces- sary to give the proper combination seems to remaininthe system without taking part in the reaction. If an excess of substrate be present, it would seem as though some of the material were awaiting its turn to be used. In the experiments of Mendel and Blood various attempts were made to explain the action of HCN in papain proteolysis. They found that among other things methyl cyanide did not have the same effect as HCN, indicating that it was not the nitrile group that was involved. They noted that KCN had less action than HCN but this is undoubtedly due to the fact that the alkalinity of the KCN gave rise to an unfavorable hydrogen ion concentration. The only other substance that was found to be effect- ive in accelerating the action of papain on protein was hydrogen sulfide. This led them to suggest that possibly the reducing properties of the two_ substances were responsible for their action. E. M. Frankel aes If the reducing properties of HCN were responsible for its activity in papain proteolysis it might be expected that some of the HCN would be destroyed in the course of the digestion. The | following experiment was carried out to test this point. 5 gm. of gelatin were dissolved in 200 ce. of water containing 0.4 per cent of tricresol. The solution was divided into two equal parts and 10 cc. of 1 per cent HCN solution added to each. To one 50 mg. of papain were added, and the other was used as a control. Both flasks were incubated at 37° for 24 hours, and then acidified with 15 cc. of 15 per cent sulfuric acid and distilled in steam, the distillates being collected in alkaline solu- tion. The distillates were titrated with silver nitrate according to Lie- big’s method for cyanide determination with the following results: AgNOs: ce. ORCC PELCINESolutioneecatmccat cies © oc coer ree 19.35 Control gelatin: HON... 2. eect. c ccs routs: ies Ses ov eee 18.1 RAD AMP CINedigestioneey. sw eactee ay ee eae a aeaeeiaeriee aoe: 19.2 These results show clearly that HCN is not oxidized or con- verted into a compound that is not readily hydrolyzed by dilute acid. It is of course quite possible that the HCN enters into some combination that is not very stable in the presence of acid and can therefore be recovered completely with the method used. Further experiments on this point are in progress. It has been claimed that in the course of papain proteolysis free amino-acids are liberated. Mendel and Blood obtained evidence of the formation of tryptophane in papain-HCN di- gestion. Abderhalden and Teruuchi (1906) claimed that the ferment could effect the cleavage of glycyltyrosine. In our experiments we have not found it possible to hydrolyze glycyl- glycine, alanylglycine, glycylalanine, alanylalanine, or glycyl- tyrosine with papain either in the presence or absence of HCN. In the experiments on alanylglycine the hydrogen ion concen- tration was varied over a wide range with no change in the result. How far up in the scale it is necessary to go to effect cleav- age with papain is as yet unknown. Some experiments under- taken from a different point of view may be of interest in this connection. 200 ce. portions of 1 per cent gelatin and dried egg albumin solutions were adjusted to pH 5 and treated with 75 mg. of papain. The solutions were incubated at 37° and 25 cc. portions were withdrawn at the intervals TEE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. l 214 Proteolytic Activity of Papain noted and titrated by the formol method. The results given under Action are the formol titrations corrected for the titration of 25 ce. immediately on mixing the ferment with the substrate. When the solutions were nearly in equilibrium, several 25 ce. portions were withdrawn and treated with HCN. The formol titrations were then made after incubation for the stated period. ‘ Gelatin. Egg white. Time. i Action. Time. Action. az hrs. hrs 1a 0.35 3 0.35 3 0.65 5 0.35 43 0.80 22 0.75 23 2.00 46 eal 69 2.80 70 1.30 HCN added. a2 3.15 94 3.30 96 3.95 These experiments indicate that HCN is effective in renewing the proteolytic activity of papain even after equilibrium is apparently reached. Whether this effect is due to a cleavage of compounds of lower molecular weight or of some tinattacked protein is as yet unsettled. A priori, it would appear that the former view is correct. It is the plan of the author to investi- gate further this whole question of HCN action in certain fer- mentations since it offers a new point of attack in the study of the chemistry of these reactions. SUMMARY. A method of purification of crude papain is presented. The conditions of acidity for the optimum action of papain are found to be pH = 10~°. A consideration of the quantitative relations between papain and its substrate leads to the view that this ferment acts like urease, invertase, and lipase in forming an intermediary com- pound which is broken up into the cleavage products and liber=_ ates the enzyme. E. M. Frankel PANS: Investigation of the action of HCN in papain hydrolysis leaves this question still unsettled. There seems to be some difficulty in defining a hydrogen ion optimum for papain-HCN proteolysis. The quantitative relations of the enzyme, HCN, and protein lend support to the view that there is a ternary intermediary compound formed by the components which then breaks down. into cleavage products of the protein, enzyme, and HCN. It has been shown that HCN may be recovered almost quan- titatively from digestion mixtures, indicating that it is not util- ized in the reaction of fermentation. Papain, with or without HCN, seems to have no proteolytic effect on the dipeptides studied. HCN can renew proteolysis in papain digests that are almost in equilibrium. BIBLIOGRAPHY. Abderhalden, E., and Teruuchi, Y., Z. physiol. Chem., 1906, xlix, 21. Mendel, L. B., and Blood, A. F., J. Biol. Chem., 1910, viii, 177. Vines, S. H., Ann. Bot., 1903, xvii, 606. Nelson, J. M., and Vosburgh, W. C., J. Am. Chem. Soc., 1917, xxxix, 790. APPLICATIONS OF GAS ANALYSIS. I. THE DETERMINATION OF CO, IN ALVEOLAR AIR AND BLOOD, AND THE CO, COMBINING POWER OF PLASMA, AND OF WHOLE BLOOD. By YANDELL HENDERSON anp W. H. MORRISS. (From the Physiological Laboratory and the Department of Obstetrics and Gynecology of the Yale School of Medicine, Hew Haven.) (Received for publication, June 5, 1917.) In physics and in chemistry a multiplicity of methods of prep- aration terminate in a very few standard methods of measure- ment. Nothing in biochemistry corresponds to the place which the galvanometer and the chemical balance occupy in their sciences. On the contrary in biochemistry the methods of meas- urement are nearly as varied as those of preparation. As every method must be learned separately, the difficulties encountered by an investigator wishing to pass from one field of biochemistry to another are thus greatly increased. We believe that gas analysis should become one of the standard methods of measurement in biochemistry. At present most workers in this field make no use of gas analysis. Those who wish to determine the alveolar air use a Fridericia (1) apparatus or Marriott’s (2) color tubes; to determine the CO, combining power of the blood, the Van Slyke (8) mercury pump; for oxygen in blood the Barcroft (4) technique; for the study of the total respiratory exchange of man or animals, and for indirect calori- metry, the elaborate apparatus of Benedict. No two of these methods employ the same technique. The object of each is, however, as we believe, more easily and more accurately accomplished by means of methods terminating in gas analysis along standard and long tried chemical lines. One who learns to do gas analysis for any one purpose has the equiva- lent of all of these methods at his command. While the more refined forms of gas methods involve skill and experience, such refinements are needed only when extreme accuracy is desired 217 218 Gas Analysis. I and are for the most part quite unnecessary. A great deal can be done with sufficient accuracy with simple apparatus and technique. An apparatus with which one may determine the CO, content of the alveolar air and the CO: content and combining power of 22 a3 (eeTeee fen] ef) caffe] af fn nfinnfnny em Faen rota sipont tonto ay mpo Mey Oy \ Y. Henderson and W. H. Morriss 219 plasma or of whole blood is shown in Fig 1.! It involves practi- cally nothing new or which has not been used in equivalent form by a long line of investigators. It consists of a 25 cc. gas burette (A) with a bulb containing about 21 ec., and a tube below gradu- ated from 22 to 25 cc. in 0.02 cc., so that it can be read easily to 0.01 ec. Around the bulb of the burette is a jacket which is filled with water at room temperature, and a thermometer is placed in this fluid. At the top of the burette is a T-tube, one limb of which is connected by a rubber tube to a simple absorber (B) in which 10 per cent sodium hydroxide is placed. The com- pensating bottle (C) connected with the lower end of the burette by a rubber tube is filled with 1 per cent sulfuric or other acid. Spring clips or screw cocks (or hemostats) are placed at the points marked L, M, and N. CO» Content of Alveolar Air. The bladder of a football or other form of rubber bag is used as in the Higgins-Plesch (5) method of obtaining the alveolar air; the subject breathes into the bag four or five times and a spring clip is placed on its tube. To prepare the apparatus for an analysis the leveling bottle is lifted’ until the bulb of the burette is full. The nipple (D) is put on at L and closed. The bottle is lowered while the clip at M is opened until the NaOH solution in the absorber is drawn up to a mark on the capillary glass tube above it. The clip at M is then closed, and that at L is opened. ~ The bottle (C) is lifted until the burette is full, and a few drops of the fluid are run out (at L), the clip at N is closed, and the nipple, spring clip, and rubber tube, D-L, are taken off. The apparatus is now ready. The tube of the bag holding the air to be analyzed is connected (at L) and the clip on the tube of the bag and that at the lower end of the burette (N) are opened. As the fluid falls in the burette it draws from the bag a sample of air, the amount being determined by the height at which the leveling bottle is held. Between 24 and 25 cc. of air are taken; the clip on the tube of the bag (at L) is closed. 2 or 3 minutes are allowed for the sides of the burette to drain, then the bottle is held so that the surface of the fluid in it is at the same level as that in the tube of the burette and the volume of gas in the burette is read to 0.01 cc. The clip on the absorber (at M) is now opened. The bottle is raised and lowered four or five times 1 Tt can be obtained from the Emil Greiner Company, 55 Fulton Street, New York. 220 Gas Analysis. I so that the gas in the burette is driven over into the absorber and drawn back. The CO, in the sample of air is thus removed. The bottle is now lowered until the sodium hydroxide solution is drawn up into the capillary to the mark at which it stood at the beginning of the analysis, and the clip on the absorber (at M) is closed. After waiting the same length of time as previously for the sides of the burette to drain, the bottle is held so that the surface of the fluid in it is at the level of that in the burette and the latter is read off. The difference between the first and second readings divided by the first gives the percentage of CO.. To turn this percentage into the partial pressure of CO, in mm. of mercury it is aoe by a figure 40 mm. less than the prevailing barometric pressure; e.g., 5.5. per cent CO, or 0.055 X (760-40). This 40 mm. is the tension of water vapor in the lungs at body temperature. As regards the barometric pressure, it is sufficient for clinical purposes to use the mean presque of the locality, neglecting the daily variations. Duplicate analyses by this method should not differ by more than 0.02 ec. of COs, or 0.1 per cent, a sufficient degree of precision for all clinical and most scientific purposes. A Mod:fication of the Haldane-Priestley Technique. The foregoing account starts with the Higgins-Plesch technique for obtaining alveolar air. For clinical purposes this is the easiest and yet a sufficiently close method. It is generally recognized, however, that the figures so obtained tend to be too high, ap- proximating the gas tension of the venous rather than that of the arterial blood. The method originally employed by Haldane and Priestley—the single deep expiration through a wide tube and analysis of the last part of the expiration—although still the most accurate available, is difficult to apply on untrained or nervous subjects. In teaching elementary students one of us has observed that the chief trouble consists in the fact that the subjects are inclined to draw a deep inspiration before making the deep expiration through the tube. Of course this deep inspiration dilutes the pulmonary air with an excessive amount of fresh air and renders the CO: content too low. This difficulty may be avoided by means of the apparatus shown in Fig. 2. It consists of a vertical tube of a capacity of 50 to 100 ce. held by a cork (cut as shown in the figure) in a wide mouthed bottle containing acidulated water. The subject is instructed to suck the water quickly up the tube until a bubble of air passes in, and then to make the deepest possible expiration Y. Henderson and W. H. Morriss 221 through the apparatus, and to close the clip. The apparatus is then connected with the burette and a sample of air analyzed. When one is about to exert a suction he has no inclination to make a preliminary deep inspiration as he does when the first conscious effort'is to be expiratory. In our opinion methods of calculating the alveolar CO. by formulas (6) involving a figure for (or including) the dead space are invalid, for the reason that, as shown by Henderson, Chillingworth, and Whitney (7), confirmed by Haldane (8), and conceded now by Krogh and Lindhard (9), the dead space is a very variable and only roughly measurable quantity, which in reality can itself be estimated only by using some figure for the composition of the alveolar air. Fig. 2. 222 Gas Analysis. I Determination of COs in Blood. The blood is drawn directly into an all glass syringe containing a glass bead and a small but known volume of ammonium or sodium or potassium oxalate, or the blood is drawn first and the oxalate afterward. The blood and oxalate are mixed by inverting the syringe a few times. The blood may also be run directly from an artery into a test-tube under a layer of petroleum oil and then either defibrinated (by stirring under the oil) or treated with a few grains of powdered oxalate. Excess of oxalate is to be avoided, as it interferes with laking the blood later on. For the blood gas analysis a tube (which may be ealled the diffusion tube) such as is shown in Fig. 3 is used. When a rubber stopper is inserted in its large end it should have a capacity between 27.5 and 28.0 cc. A bit of rubber tubing 4 or 5 em. long is put on the capillary end and closed with a spring clip, or better with a bead valve consisting of a lead shot or piece of glass rod 5 to 8 mm. long inserted in the rubber tube. To be gas-tight the bead must be large enough to stretch the rubber considerably. Such a valve is opened by pinching the rubber tube over the bead with the fingers. A solution of 1 cc. of con- centrated ammonia in 500 cc. of distilled water is used, and another of 20 per cent tartaric acid. For an analysis 2 cc. of the dilute ammonia are put into the diffusion tube and 1 cc. of the blood is delivered with a pipette under the ammonia. The blood and ammonia are mixed and the blood laked by gently rotating the tube while held vertically. Then 0.5 cc. of tartaric acid solution is delivered with a long pipette below the laked blood. A 0.5 ce. pipette of sufficient accuracy is easily made of a.piece of common glass tubing by drawing this volume from a burette and marking with a file—half a drop more or less is unimportant. A rubber stopper is now inserted in the large end of the tube and the tube is laid horizontally and rolled rapidly for 3 minutes. The acidified blood spreads in a thin film which is continually renewed by the rotation and allows rapid diffusion of the liberated CO, into the airin the tube. For the rolling it is convenient to lay the tube in a rack such as shown in Fig. 4 and to connect it to a slow running motor by means of a belt. If the motor is not available the tube may be laid on a sheet of blotting paper or strip of cork and rolled back and forth by means of a strip of wood covered with blotting paper or cork (to avoid the heat of the hand). The essentials of this rolling are that rapid diffusion be obtained from the blood to the air in the tube without the production of any bubbles or foam. The analyzer is now arranged for a gas analysis exactly as above de- scribed for alveolar air; that is, the bulb is filled with fluid and pinch- cocks areclosed. Thesmall end of the blood gas tube is attached at L, while the lower end dips into a beaker of slightly acidulated water (see Fig. 5). The rubber stopper is now withdrawn under water, the pinch-cocks at L and N are opened, and by lowering the leveling bottle all the air is drawn from the blood gas tube into the burette. The pinch-cock or bead valve at L is closed and the burette is read. The air is passed into the absorber Y. Henderson and W. H. Morriss 22e Fig. 4. five times and the volume again read exactly as previously described. The decrease of the volume (not the percentage decrease) is the volume of CO, which has come from the blood. Every 0.01 cc. represents (with corrections) one volume per cent CO, in the blood. A correction must be made for the CO. which remains in solu- | tion. At temperatures of 14-18° the solubility is about 1.0, 18-22° about 0.9, and 22—26° about 0.8. To make this correction it is usually sufficient merely to add one-tenth to the CO, found in the analysis. This quantity is estimated by multiplying the CO, found in the analysis by a fraction of which the numerator is the volume of acidulated blood and the denominator the volume of air in the tube, and then multiplying again by the coefficient of solubility at the temperature of the analysis. An occasional analysis should be made without blood to de- termine the CO, which the ammonia solution may hold, and this correction is subtracted from each blood analysis. The ammonia should of course be kept tightly stoppered as it accumulates CO2 from atmospheric air if exposed. Duplicate analyses should agree to within 0.03 cc. of COr in 1 cc. of blood. This gives the CO, at the prevailing temperature and pressure. The correction to 0° and 760 mm. barometer may be obtained Pee) | ie 224 Gas Analysis. I from the table in a chemical calendar or calculated by multi- plying the volume of CO, at the prevailing temperature and pres- 273 X 760 sure by the fraction (temperature ©2783) X barometer. CO2 Combining Power of Plasma. Blood obtained as previously deseribed is centrifuged. The plasma is brought into gaseous equilibrium with alveolar air in the following way (virtually that of Van Slyke). The last half of a normal expiration is expelled through a wash bottle or large test-tube containing glass beads to absorb the moisture into a 250 cc. separatory funnel containing about 3 ec. of plasma. The funnel is immersed in a water bath at body tempera- ture and spun in such a way as to spread the plasma in a thin film over the glass surface. After spinning for about a minute another portion of alveo- lar air is introduced into the funnelfand the spinning repeated. It is well Hire. 5; Y. Henderson and W. H. Morriss 225 for beginners in this technique to make sure that the alveolar air used to obtain this equilibrium contains the correct percentage of CO, (about 5.5), by attaching the rubber bag to the opposite end of the funnel and analyzing a sample of the air which is blown through the funnel into the bag. An even better technique is to collect alveolar air in the rubber bag, to analyze it, and then to squeeze part of this air through the separatory funnel in which the plasma has been placed. The separatory funnel is placed in a water bath and spun and the process repeated as above described. After the plasma has been brought into equilibrium with the alveolar air 1 ce. is drawn into a pipette and placed in the blood gas diffusion tube under 2 cc. of dilute ammonia solution and mixed. Acid is added and the analysis for CO, made as previously described for blood. TABLE I. Comparative Results by Van Slyke Method and by Gas Analysis (Both Cal- culated by Means of the Van Slyke Table). CO:z in 1 ee. of Vol. CO2 in 100 ec. ae “used for plasma. pee a: paar A or, tion . Van Slyke Gas eae Van Slyke Ge method. | analysis. : method. | analysis. per cent CC; ce cc cc ce Incomplete abortion. | 5.84 0.67 0665/5 0201 56 55 sf 5.8 0.56 0.58 +0.02 45 47 Threatened abortion.| 5.9 0.69 0.68 | —0.01 57 56 After dilation and curettage. 5.4 0.59 0.64 | +0.05 47 52 Toxemia of preg- MGV eye cls gore «= 5.6 0.56 0.59 | +0.03 45 48 Before operation for prolapsus... ee toA0. 0.72 0.68 | —0.04 60 56 tee for protep- SUSs2 6.2 0.69 0.71 +0.02 58 58 ieeeinia. cor Lae TRENTOCN/S cs cE 5.4 0.52 0.54 | +0.02 40 42 ss Seth 0.56 0.53 | —0.03 45 42 Menorrhagia......... 6.2 0.66 0.66 0.00 54 54 Eclampsia. Maternal blood....| 6.2 0.32 0.34 | +0.02 22 23 Fetal ss 5.6 0.31 0.28 —0.03 21 18 Menorrhagia........:{ 6.7 0.58 0.60 | +0.02 47 48 Normal labor. Maternal... 52. 0.39 0.43 | +0.04 29 35 1 ET ie re 0.48 0.55 | +0.07 37 44 HPCIAINDSA 2... |: 0.37 0.36 | —0:01 27 26 226 Gas Analysis. I The degree of precision is about the same as with the Van Slyke apparatus. 246 “B1OY jetiajyoeq snorntul uw jo yueuIdojaAep oY} SIOAVJ YOY ‘qey pu utaqzoid ul You JoIp Surzedrysuod & OF WOTPOTI}Sot JO FNS ay} SB ‘umoas ay} JO Alperoedsa ‘yous} [BUTSOJUT OY} JO WOTFTPUodS dUsISAYUN UB OF BSBOSIP 9Y} OFNII}ZVW OF uolj;euvydxe gjqisneyd o1our &@ st4y “AAAnos Jo ABOLOTJ9 ay} ulejdxo 03 pavmios ynd savy Si0y}O pus YUN TOU uolyeurdxe ue ,,‘SouTul -BYIA,, UT JUSTOYap Jou SI JOIp STY} YY} UOTJUOJUOD INO soziseydure yor ‘syooM QZ Jo poled v Burmp FYs1oM Ss} peyqnop UvY} d10UI PUB YY[BoY poos pourequrvur oay Jo dnoss v ur sid BouINS 9UO JoIp YIU YSery puv yvo poT[or B UO MOY 97VIISNI|T 0} a10Y UMOYS av S9AIND oY, “ooeLdap 9]qBdT}OU B OF STBUTUB OY} pazoojyoud aAvy 0} pres oq youuvo yy “ATTEp plow oLI}1O JO UOTPAYSTUTUIpPE oy} Aq oUT}SOYZUT OY} UL SUISTUBSIOOAOTUT Jo yjMo1B oy} aesvanoosip 0} pozyduls}}T aM “fF LUVAD 0G BI g\ | él SJ 00¢ || Ng se NIV 247 E. V. McCollum and W. Pitz “pooj 94} Jo sorjsodoud pworsAyd oy eAOIdUIT 0} pus} JOU pIp ‘asunod jo ‘ayvoOzZUEq WINIPOg “pooy HUI pu’ yvoO 9Y} UL VOUBYSGnS oIYNqQIOOSTJUB UB JO Yor] UBY} JeYJO 9sSNBd 9UIOS WOIJ OS pIp AAANoS COTeAOp YOIYM oSOYy YBVY} JUOPTAY 4I SOYVUI SITY, “porlsed Suo] wv 19AO YYySIoM Apoq Ul posBvoiIOUl puv YA[VoOY JO 9IISOp UV} BV POUTeJUTVUT YOTYM S|[BNPTAIPUI o1v ‘F puB E SJABYH Jo osoy} ul sv ‘dnoas STy} UL *938OzZUEq WNIpPOS oY} YNOYIIM YoIp SITY} UO TeNSN ST UBY} JOSUO] YONUI POAT] S[TBULIUG VAY ot{} JO 99AY} OJ “QoaIyo 9A1}00}01d JYSIIS B Poeslotex9 sABY OF SIvodd¥ UOT}IPpe SIU, ‘e}vOzZUeq UINIPOS Jo yudo sad ZC YIIM JOY JOGO} YIU ysoay puv $}BO PoTjOI JO Jorp B SuIpesy Aq VIAdZOVq JO YIMOIS oY 0} 3du10}48 UB UIBSB SoyBIASN][T SIYT, “¢ LUVHO zeeatans: Eee eS 1seases lency D vnd Defici ” ny ¢ amine Wat 248 ‘AAINOS JO SUBIS JUSS poMoys puv syooa F 10978 porp “uaaos ayy JO 4SoT]BUs 94} “OMT, “yey puw Yoo]s AjjeuoTydaoxe aie AY} SyooM QT 10}j8 ynq [JoM Mois ATUO Jou Aoyy IOF JUSUT}BAI} STG} AG VoIdap a[qBIOpIsUoD B 0} peyyousq A;yUeIEdde a10M S[BUIIUB VSoyT, “UINdID oY} UT Sadej ulejer 03 AOUOpUd, OY} UOTZOV oTJIVYYvO Sy1 Aq yoRIOJUNOD 04 ule[eyyydjousyd jo sop paysn [pe A[quyms vB ALR] “NBII WY} SUIAIS Jo puv Y]IUI puv s}vo Jo yop vB ssid vous Sutped} jo syjnsor oyy SMOYS SIP], “9 LUVHD I 9 vl zl al :) 9 Mee ee ce BS . iy Bai 2 ca Zz 2 if 249 E. V. McCollum and W. Pitz ‘suvod Jud00d UIQITM peso due useq SsBy UNIO} FY} YOM UL OSES oY} UT ,,ASvosIp AoUoLloyap,, BST AAINOS yey} BVepl oy} ysuIeSe A[Su0I4s sonZiv puv ‘sou sqns orNqioostyuB uB A1Ie. ATqissod youuvo yorym “Wnyxporjyed pmb ‘oouvysqns v YIM yuoUTyReT] Aq ‘AAINOS YQIM YyBvop TBM STRUNTUL JO SOIND [BNJoV JUOSoIdoI Sosvo osoyT, “OSBESIP oY} WIOY} OAS YOTYA Y[MtU puB szvO Jo yorp oyy UO A]}01148 PONUIZUOS ITY “SoTqBJeSoA UaeId JO JoIp B UO JI Se YoaTs puv Auds sv ourvoaq puv Ajajo[du09 peroeaoooa AT[VNpvss pue ‘A[lep wunyepotjed pmb] wears aram Ay, *AAanos Jo OT4SIIOJOVIVYD SUNS oIdvyILOUeyY pue szUIOL UAT[OMS oY} PoMoYs pus Ap 4STE oY} UO YPM 04 o[qvuN o1OM F puBZ ‘SON “Yoo Y}p yy JO pus oY} JB S1OY}O OMY PUB SYaaM E JO pusd oY} 4B AAINDS JO polp 9UQ “WNIZGY) PY YITUA ojoYM Ysoaj puv 84vO poor poy or9.M Cnoad SITY} UT S[VUIIUB oY} [ POlleg UT "2 LUVHD moet || | aX — INA iency Diseases ” and Defie itamines y* ss 0 5 9 JO 19JOBIBYD 94} UO PUB Sadaj BY} JO JoJOvIvYO [voIsAYd BY} UO JOoTO SJ OF LOYYVA YN ‘aoUBYSqNS OTyNqGAOOSTZUB oyloeds B ZUTUIB{UOD S$} 04 BNP JOU ST OIYNGIOOSIZUB UB SB vdIN[ ddURIO Jo ADUBTIONJO oY} YVYy SoyBoTpul 19ded SIG} UI Sjrvyo 9Y} UT peyusseid 9oUSpIAS oY, “Sid BoUINS oY} UI WdaS S9INyVOj [BOTUIOZvUR IvTTNOOd 9y4 YIM qOB1} DATJSOSIP B IOJ AIOJOVISTZRS You a1¥ Satzaedord [eBorsAyd sqrt ynq ‘oyenbope AT]BoTUIEY) SI Yotp y[IUI pus ywvo ey, ‘“A[[eulIoU MOIS 0} B]qQv you ore Avy} AAINOS WOIJ oSvjUB.IEd GIy B S}99}0I1d sIy} YSnoyIpy ~“Ayrep vornt 9BUBIO YSIIJ JO “09 [ UIAIS PUG ITU PUB $}¥O PIT[OI pay ssid vouINZ MOj JO SoAAND OY} SMOYS SIG], "8 LUVHD 251 E. V. McCollum and W. Pitz ‘AAINOS Surdopaaosp yNoyyia [[oM Ayparey MOIS PUB SOOUBISGNS 9A1}90}01d Jo UOTPIPPE YNOYIIM YW pus s}vo oyey 04 9[Qv o1B STBUNTTUR Moy B puv ‘requinu e[quiepisuod B 4yoa}01d 0} dAIeS asTMOYT] (2 JUBYD) winyzepo.jed pnb] 40 (g yaeyD) uroyeyyydjousyd sv yons seourysqns jo sosop aiqeying “faanos ysuIese sTeUIUe 944 Syoo}0;d SIYL ‘oprxoupAy wNiIpos YIM pozipeaynou sv 49948] oY, ‘“Aprep ooml oFuvso JO owe ‘09 T QUAY Jolp YT puw yVo pa[[or B uo ssid vous AMO; JO SaAINd YZMOIZ oY} SMOYS SIT, 6 LUVHD ean u oa 8790 nA, Tou UO T ‘gouRysqns dIgNqsoostyuB [BoTjoyjodAY B BSuiystuuny AQ UBY OYZVA BAO [VUT}Se}UT 94} Burdytpour pues uUoljPeULUTI[a Surpre sy jo asneveq sjyoej01d aomnf oduv10 yey} WOISN[DUOD Vy} 0} Sn pa] SyudUTIedxe Ino Aq pop1Oye ByeVp-19y}0 oY, “Suryvoy ysnoiy} taaod aAtj00}01d at Sury}Aue ysoy] pvy sotnf osuvs0 oY} AaYJOYM Oploep 04 o[qISsoduat 41 SoyVUL JOIp Y[IUE pus yvO poy -IpOWUlUN UB UO S[BUITUB JO LOLAVYOd oY} UO UOTBIIBA OY], “NOY T AOJ ainssaid Spunod CT 7B 9AV[IOJNG UB UT Poyvay we0q psy yorya soml osuvs0 poztpVaynou JO °99 T YIM JoY}oSO} YOrp YIU puw 4vo UB PdAtoveI sdId vouINS sso, ‘OT LUVHD eo ra fo p es Wor aot paztyteiyznoe Hyon ie pe slaty es uo LOT ine: ee a ~ "a e-— EEE EE SSEEOSSS'SS:SS&C“C“C~ ~~ rr es 252 APSE oo eet St ee SLL remove Som SUIOT Stl “6 POMed JO suluUlseq oY) Iojje syooM 72 pelp 4] . “oymedds poos v py puB 41oyv ourBd0q 41 ynq ‘sBo] pury 94} UL SSOUJT}S OULOS POUTBUIL [[IYS a1oY], *Z% polled ur poaoadurt Ap} void WOTPIPMOS S}T JYSIOM UL UIRS YOU PIp WI YSNoyyy “T potuoeg JO pus OY} 4B UOT}IPUOD [BOTFTIO V UT OSTB SBA g BIg vouINy “VOIP 94} JO UOT}VoYTpoUT oY} Ioize SYOoM B UOTIPUOD YUdT[AOXE UT SBM 41 ynq °% ON jo 7BYy} UBY} IOMO+s SBA ATOAODAL S}T ‘qnoqe Woy} possvip pue sso] pury oy} Jo 9sn oY} 4SO[ pvy F SIg Bourne “UOTHTPUOS FUO[[IoXe WB UT SYooM YZ JO pus oy} 4B svM puv ATprdu) poAorduat [euTTUe oy} o1NyxTUT LeSNs pus ‘prow OTAPLO “F[BS oY} JO UOTPONpOAZUT 9yy AaVFY “AYMOUIP YILM poyyworq puv ‘CAanos ALM WOTpIpUoD puq AIOA B UL SBA Z BIg Bourn ‘got asuB1to jo ‘00% 0} JUOTVAINDS ainyxtuT ayy Jo JUNOUTY UB PpolI1Vd 8ZBO FO "WB OT 9,°aduUv10 OY} JO UOT}A0d STGIPS PYF UL oso} JO JUo}UOD oY} JO UOTZRIIUNT UT dn ope oiNn4xIM AvSNS puL ‘prov o1Ay1 ‘4[VS B YIM JOM Poxtu 910M $7BO pe[or oy} posed puooes oy} uy “AAanos pry IB ‘aLOJoroY} ‘Z pOMeg UL “Yao YIZT oy} [UN osvostp oyy dojaaosp jou pip Zz puB T “SON O[TYM “Yoo Y}p oy} JO puo oy} 4B AAINOS pvy (F PUB Eg “SON) Woy} JO OMT, ‘oABTDOING UT ul oinssoid Sspunod ¢T yB INOY T LOF popVoy oLoM 10}4V] OUT, “SJOLIVD pu’ $}¥O poT[OI poy a19M Anois sty} Jo speurrus oY} T poled Uy “TT LuvAD yl Zl 8 9 "| ie x u fQ Eiri + eyeryIo y D°HY (‘pAUB O TeTotstipy) TeS eel | et ea fs v a oa =soam N ot —— aS oa i We ai as PACTS > v > ° f) Lo] oO » a a Ks) Co) o Q Pwo » By) 2 wn THE INFLUENCE OF BILE ON PHENOL PRODUCTION, By HARRY DUBIN. (From the John Herr Musser Department of Research Medicine, University of Pennsylvania, Philadelphia.) (Received for publication, June 5, 1917.) We have found in our previous studies of phenol production that an increase in the formation of phenols, whether normal or pathological, resulted, with one exception, in an increased con- jugation.1 Here there was an increased phenol formation with an unchanged or even lessened conjugation, indicating that the bile plays some part in assisting the liver in its conjugating function. It was in the hope of throwing further light on this subject that the present investigation was undertaken. Our methods of procedure and analysis have already been described.!. After a period of normal observation, the dog was operated on? and a bile duct-ureter anastomosis was produced after the method of Pearce and Eisenbrey.’ Bile was continually present in the urine, no evidence of jaundice being present at any time. In one instance the bile was removed from the urine by ~ allowing it to stand with fullers’ earth for 2 hours. However, on analysis this clarified urine gave the same figures as were obtained with the original urine. At autopsy the anastomoses were found to be intact. In Dog 3, an external bile fistula was produced whereby we were able to collect the bile so that the urine was free from it. The external fistula was produced as follows. Small holes were cut for a distance of about 5 inches, and near one end, in the walls of a rubber tube about 15 inches long. Around these holes were tied a soft rubber bag, the rubber tube at this end being fitted with a small silver cannula. The silver cannula was inserted into the common bile duct and tied in place. 1 Dubin, H., J. Biol. Chem., 1916, xxvi, 69. 2 For rs pporative work, we are indebted to Dr. J. E. Sweet of the Department of Surgical Research. 3 Pearce, R. M., and Eisenbrey, A. B., Am. J. Physiol., 1913, xxxii, 417. 255 256 Bile in Phenol Production The other end of the tube was led out through a stab wound and clamped off, so that the bile collecting in the soft rubber bag, which remains inside the body, could be drained off. The animal was bandaged to keep the rubber tube in place. As early as 24 hours after the operation an increased phenol production with an unchanged conjugation was obtained. Table I presents a short résumé of work previously published, showing a comparison of results obtained under various patho- logical conditions, normal control periods being present in each case. The table shows the contrast in regard to the degree of conjugation in the bile dog as compared with the others. TABLE I. Phenol Elimination under Normal and Pathological Conditions. Phenols. Dog. = Free Total Free a mg. mg. per cent per cent ile INOrmale oe ce eee er Cee 167 217 76 24 Bilecductscuts cose en ee 263 329 80 20 2. INO TTA te onc aicters Macao ene ene tae 175 227 77 23 Intestinal obstruction............. 263 477 55 45 a}. INGE ae ae cic et eee 184 223 83 Ws Pancreatic ducts cut............... 203 293 69 31 4. Normal bate sijcrcie amyl > butyl > propyl > ethyl > methyl. Ethyl and methyl gave results differing little from that in pure water. This action of the higher alcohols can- not be explained on the basis of lowering of surface tension alone, since even when isocapillary solutions were used (Lillie, 1916) the order of effect was amyl > propyl > methyl. Further investiga- tion will be necessary: before many of the observed phenomena can be explained. 276 Light Production and Catalysis DISCUSSION. It has been observed by Harvey that the intensity of the light produced does not run parallel to the rate of evolution of oxygen. This has been confirmed by my experiments, since in some cases a violent bubbling off of oxygen occurred when little or no lumi- nescence was present. It is evident, too, that the effect of the colloidal metal was not due entirely to the large surface exposed and the consequent adsorption of oxygen, since widely different results were obtained with metal sols of approximately the same degrees of dispersion. The colloidal, metals which were most active in transferring oxygen from the hydrogen peroxide to the pyrogallol were those which are able to form unstable compounds with oxygen. Some interesting experiments along this line were described by Reed (1916, a), showing that when finely divided platinum (platinum black, incorrectly called ‘colloidal platinum’’) on an electrode is charged with oxygen by making it the anode, it can bring about oxidations directly. Platinum charged with hydrogen cannot oxidize directly, yet both are able to decompose hydrogen peroxide. The author (Reed, 1916, b) draws the conclusion that ‘‘although substances which act as oxidases or peroxidases usually decompose hydrogen peroxide, yet in the case of colloidal platinum they are quite independent:” It must be pointed out, however, that in the decomposition of hydrogen peroxide by colloidal platinum it is only the rate of a reaction already occurring which is changed, and that a sample of platinum black cannot be placed in contact with a solution of hy- drogen peroxide without immediately taking up oxygen. I have observed, for example, that if finely divided platinum which showed no direct oxidase action, 7.e., was unable to blue gum guaiac, was placed in hydrogen peroxide for a few minutes and then removed and washed with four changes of distilled water, it did show di- rect oxidizing action by blueing guaiac solution. This was not due to an adsorption of hydrogen peroxide as such, as has been suggested by some authors, since the same result was obtained with colloidal platinum sols by bubbling air through them or, more slowly, by letting them stand in contact with air. Plati- num, palladium, and silver sols, as prepared by the are dis- persion method in water, showed this presence of oxygen by blue- Bic. Goss pa ii Ff ing guaiac directly without the presence of hydrogen peroxide. If a stream of hydrogen was bubbled through the sol for a few hours, the colloidal metal lost the power to oxidize guaiac directly but produced oxidation by hydrogen peroxide, or, if let stand in air for only a few minutes, the colloid took up oxygen and regained its oxygenase activity. Bose (1900), after summarizing the work done previous to that time, concluded from the experimental evidence that the combi- nation of hydrogen and oxygen with platinum, palladium, and gold was not due to the formation of compounds, but to solid so- lution, together with a certain amount of adsorption. He noted that, while the metal had a greatly differing capacity for hydrogen and oxygen, the total of absorbed and adsorbed gas became greater with subdivision of the metal, and that the difference between the capacity for hydrogen and oxygen became less. Later investigations have supported this view regarding the con- dition of the hydrogen, but have indicated the formation of a series of oxides of platinum and palladium. Lorenz (1906) no- ticed that the E.m.F. of cells with platinum electrodes differed widely among themselves and suggested that the u.m.F. of the oxygen electrode was determined by the formation of an oxide. He found that the oxides of cadmium, copper, silver, and iron gave an E.M.F. which was identical with that of the metal when meas- ured under the same conditions. Wohler (1905) studied the oxides ‘of palladium and their behavior toward reducing agents, and found that palladium dioxide, although a strongly exothermal compound, was a better oxidizing agent, toward many substances, than molecular oxygen. He suggested that this behavior was due to the easy splitting off of atomic oxygen. Hydrogen perox- ide reduces palladium dioxide much more readily than platinum dioxide, but the reverse order is true of the monoxides, platinum monoxide being easily reduced by organic acids or by hydrogen peroxide. Palladium metal catalyzed the decomposition of hy- drogen peroxide more vigorously than palladium dioxide, which, in turn, decomposed the peroxide more rapidly than the monox- ide. The authors concluded that the decomposition of the hy- drogen peroxide was brought about, not by palladium monoxide or dioxide necessarily, but perhaps by an intermediate oxide. It appears from these experiments that, in the case of the col- 278 Light Production and Catalysis loidal metals, catalase activity and the ability to transfer oxygen, v.e., oxidase activity, are not at all independent but that both are dependent upon the ability of the metal to combine with oxy- gen to form an unstable complex which is a better oxidizing agent than either hydrogen peroxide or molecular oxygen. This combi- nation is probably due to adsorption of oxygen on the great sur- face exposed, together with the formation of oxides as the result of this adsorption, the oxide being soluble in the metal. The oxi- dation of the pyrogallol in this reaction mixture was brought about, therefore, by both the adsorbed and combined oxygen, the proportion of adsorbed oxygen becoming greater with decreased size of particles. The colloidal metal is continually supplied with oxygen by the hydrogen peroxide, whose normal decomposition according to the equation te = — H.0, @ H + HO. @ HO+0 is accelerated by the removal of oxygen atoms. The metal-oxy- gen complex furnishes ‘‘active’’ oxygen by dissociating in con- tact with the pyrogallol. SUMMARY. White light was obtained by the oxidation of pyrogallol with hydrogen peroxide in the presence of certain colloidal metal and metallic oxide sols, closely resembling that produced by lumi- nous organisms. The action of the colloidal metal was similar to that of the vege- table oxidase, except that in the former case the catalyzer was not destroyed. Visible light was produced by concentrations of colloidal plati- num as low as one part in 250,000. A platinum sol containing 0.0002 gm. per cc. produced a fair light at —5°C. and a bright light at 0°C. This catalysis was not due entirely to the high degree of disper- sion of the metal or oxide and the consequent large specific sur- face, but was also, in part, dependent upon the ability of the metal to form unstable compounds with oxygen. The production of light was inhibited, in general, by the pres- ence of protective colloids such as gelatin and egg albumin. B. 'C. Goss 279 Potassium stearate markedly increased the intensity of the light produced, the concentrations of the other substances remain- ing constant. A similar effect was observed in the presence of the higher alcohols such as capryl and amyl, suggesting the influ- ence of surface tension. The author wishes to express his appreciation of the assistance given by Drs. E. N. Harvey and Alan W. C. Menzies in this in- vestigation. A closely related paper by Harvey, who has noticed independently that oxidation with light production may be brought about by colloidal silver and platinum, appears on page 311. BIBLIOGRAPHY. Bach, A., Compt. rend. Acad., 1897, exxiv, 951. Bach, A., and Chodat, R., Ber. chem. Ges., 1903, xxxvi, 600. Bayliss, W. M., Principles of General Physiology, London, 1915, 583. Bose, E., Z. physik. Chem., 1900, xxxiv, 703. Dubois, R., Compt. rend. Acad., 1901, exxxii, 431. Dubois, Ann. Soc. Linn. Lyon, 1913, 11. Harvey, E. N., Am. J. Phystol., 1916, xli, 454. Kastle, J. H., and Loevenhart, A. 8., Am. Chem. J., 1901, xxvi, 539. Lorenz, R., and Hauser, H., Z. anorg. Chem., 1906, li, 81. Paal, C., and Amberger, C., Ber. chem. Ges., 1904, xxxvii, 124. Paal and Amberger, Ber. chem. Ges., 1907, xl, 2201. Paal, C., and Schwarz, A., J. Chem. Soc., 1916, a, cx, 307. _Paal, Ber. chem. Ges., 1916, b, xlix, 548. Radziszewski, B., Ber. chem. Ges., 1887, x, 70. Radziszewski, Ann. Chem., 1880, ecili, 305. Reed, G. B., Bot. Gaz., 1916, a, lx, 53. Reed, Bot. Gaz., 1916, b, xii, 409. Trautz, M., Z. physik. Chem., 1905, liu, 1. Ville, J., and Derrien, E., Compt. rend. Acad., 1913, elvi, 2021. Wohler, L., and Konig, J., Z. anorg. Chem., 1905, xlvi, 336. BLOOD FAT IN DOMESTIC FOWLS IN RELATION TO EGG PRODUCTION.* By D. E. WARNER anv H. D. EDMOND. (From the Poultry and Chemical Laboratories of the Storrs Agricultural Experiment Station, Connecticut.) (Received for publication, June 2, 1917.) The natural color of the body fat of most hens and the fat in egg yolk is yellow. This color is due to a pigment of the xan- thophyll group with a very little ‘‘carotin-like pigment,” as shown by Willstitter and Escher in their isolation of it in crystalline form from eggs,! and identical with plant xanthophyll as shown by Palmer.2 It was the disappearance of this color from the external parts of the hen’s body that led the writers to make the present investigation. There is a correlation between ege laying activity and yellow pigment in the domestic fowl. Blakeslee and Warner*® have shown that where the yellow in the ear lobes did not exceed 20 per cent in a given group of hens the per cent of birds laying at that time was high; but with birds having a higher amount of yellow there was a decline in per cent laying. The conclusion drawn by Blakeslee and Warner* was that ‘‘the laying removed the yellow pigment from the body for the production of eggs more rapidly than it could be replaced by the normal metabolism.’ With this fact in mind the writers believed that if the yellow pigment which was present in the hen’s body previous to its laying was transferred to the egg yolk, in like manner the body fat which contains the yellow pigment would be taken from the body by the blood to assist in the build- * By fat in this paper we mean the total ether extract composed chiefly of fats and cholesterol. 1 Willstaétter, R., and Escher, H. H., Z. physiol. Chem., 1911-12, Ixxvi, 214. ? Palmer, L. S., J. Biol. Chem., 1915, xxiii, 261. ’ Blakesiee, A. F., and Warner, D. E., Am. Naturalist, 1915, xlix, 360. * Blakeslee and Warner, Science, 1915, xli, 482. 281 282 Blood Fat in Fowls ing up of the egg yolk. If this is the case a hen which is laying heavily should have blood much richer in fat than a hen that is not laying. The object of this paper is to show the relationship of blood fat in fowls to (1) egg production, (2) presence of food in the ali- mentary tract, (8) color of legs, ete., and (4) sex. Warner® has shown that the average amount of fat found in the high producing hens and hens that were laying was 1.426 per cent; in the low producers it was 0.886 per cent. This, however, was preliminary work and involved only ten hens. These results were considered worthy of:a further study. In October, 1916, eighty-two hens and twelve cockerels belonging to the Storrs Agricultural Experiment Station were used for the following work. All of the hens used for this work were White Leghorns that had just completed their 1st or 3rd year of laying, and their trap nest records were at hand. Among this group of females were found birds having high, mediocre, and low fecundity. EXPERIMENTAL. In considering various procedures for the determination of fat in blood the methods of Kumagawa and Suto,® Rosenthal and Trowbridge,’ and Bloor’ seemed to require too much time and manipulation without compensating advantages. A modified Soxhlet method for fat, similar to the one used by Gettler and Baker,® was used for this work. About 5 gm. of blood were drawn from the basilar vein into a small test-tube. This blood was poured into a thin-walled lead dish of known weight, containing a mat of asbestos fiber, and the whole was quickly weighed. The dish and contents were then dried over phosphorus pentoxide 7m vacuo for 12 hours at 50— 60°C., and kept over phosphorus pentoxide until extraction was 5 Paper read before the Ohio meeting of the American Association of Instructors and Investigators in Poultry Husbandry, August, 1916; J. Am. Assn. Instructors and Investigators in Poultry Husbandry, 1916, iii, No, 1, p, 4. 6 Kumagawa, M., and Suto, K., Biochem. Z., 1908, viii, 212. 7 Rosenthal, H., and Trowbridge, P. F., J. Biol. Chem., 1915, xx, 711. § Bloor, W. R., J. Biol. Chem., 1914, xvii, 377. ® Gettler, A. O., and Baker, W., J. Biol. Chem., 1916, xxv, 218. D. E. Warner and H. D. Edmond 283 made. After an 18 hour extraction the ether was evaporated from the extracted fat and the extraction flask dried over phos- phorus pentoxide in vacuo to constant weight at 50-60°C. By the use of this process the extracted material was found to be completely free from blood pigments and only possessed the yellow color of fat when it was present in the larger amounts; otherwise the extract was colorless. The extraction as described above gave fat, cholesterol, and fatty acids, and probably also traces of other lipoids.? For the discussion in this paper the entire extracted material is called fat. The cholesterol was determined in the blood of a few laying and non-laying hens and also a few roosters, as shown in Table X. The average amount of cholesterol for these laying and non-laying hens was more uniform than the fat for these same groups, and no particular correlation could be traced. It is planned, however, to make a further study of the cholesterol and fat in the domestic fowl. The method used for the determination of cholesterol was that used by Gettler and Baker, as directed by Autenrieth and Funk?° and also by Bloor." As outlined by the above method, the fatty extract was taken up with 5 ce. of chloroform to which were added 2 ce. of acetic anhydride and 1 cc. of sulfuric acid, and made up to 10 ee. with chloroform. The solution was then kept in the dark for 15 minutes and compared ina Duboscq colorim- eter with a similarly treated standard solution of cholesterol in ‘chloroform. Table I shows that the total per cent of fat found in the blood of seventy hens was 28.467 per cent or a mean of 0.407. The per cent of fat found in the blood of hens is variable and the aver- age per cent of fat found in this group of Leghorn females is much lower than that reported by Ingles.!2 The per cent of fat ~ varies from 0.083 to 1.953, which is quite striking. The varia- tion is due largely to the fact that the birds that had a high per cent of fat in their blood were all laying at the time the samples of blood were taken, whereas most of the others were not laying. t¢ Autenrieth, W., and Funk, A., Miinch. med. Woch., 1913, 1x, 1248. 11 Bloor, J. Biol. Chem., 1916, xxiv, 227. 2 Ingles, H., Manual of Agricultural Chemistry, London, 1908, 259. The amount of non-protein reported by Ingles would, however, include fat, mineral matter, and possibly some organic substances. 284 Blood Fat in Fowls TABLE I. Fat in the Blood of Seventy Hens That Had About Completed Their 1st Year of Laying. No. Fat. No. Fat. No. Fat. per cent per cent per cent 15B- 9 0.083 15B-52 0.180 15B-32 0.258 15 0.085 46 0.181 5L-67 0.272 13 0.088 5L-71 0.183 5L-62 0.300 54 0.114 68 0.187 15B-97 0.307 . 16 0.123 15B-55 * 0.187 5L-60 0.312 | 5L-54 0.123 58 0.188 _ 15B-11 0.406 . 15B-56 0.124 20 0.189 5L-51 0.437 . 18 0.125 5L-64 0.191 15B—23 0.448 3 0.131 77 0.195 64 0.448 60 0.131 15B-80 0.198 5L-63 0.526 74 0.1338 31 0.202 15B-—24 0.541 49 0.1388 36 0.202 39 0.625 5L-53 0.144 37 0.204 5L-70 0.735 15B-51 0.154 66 0. 204 57 1.255 19 0.161 34 0.205 15B- 5 1.315 5L-66 0.166 25 0.223 89 1.319 15B- 1 0.167 62 0.228 51-65 1.388 83 0.167 98 0.228 15B-90 1.641 5L-52 0.167 104 0.230 5L-59 1.779 15B-96 0.169 82 0.231 15B-57 1.864 5L-55 0.175 77 0.241 43 1.915 15B-88 0.178 91 0.242 12 1.953 99 0.179 5L-69 0. 246 Total. 28 . 467 105 0.180 15B-28 0.253 Mean. 0.407 Table II shows the per cent of fat found in the blood of the 3 year old hens. These data show that the blood was very low in fat, giving an average of only 0.171 per cent. There is a marked difference between the amount of fat found in this group and that found in the blood of 1 year old hens. None of the 3 year old birds had laid, on the average, for 60 days at the time the blood samples were taken. This fact we believe warrants the conclu- sion that a low fat content signifies that the hen is not laying. Of the eighty-two hens there were, then, only seventeen that were laying when the blood samples were taken; the average number of eggs from these hens was 163.7, showing a much higher average egg production than for the birds that were not laying D. E. Warner and H. D. Edmond 285 TABLE II. Fat in the Blood of Twelve Hens That Had About Completed Their 3rd Year of Laying, and Days since Their Last Egg Was Laid. No. Fat. Days since last egg was laid. per cent 107 0.066 56 425 0.100 46 76 0.126 73 382 0.157 36 43 RR 0.161 92 401 0.166 51 430 0.168 63 386 0.176 35 102 0.194 61 404 0.208 66 116 0.228 79 110 0.297 62 Mota. se sos 2.047 720 Wiles a2 008 ae ete eee 0.171 60 TABLE III. Laying Hens. No. Yearly total. Fat. per cent 69 fez 0.246 15B-32 158 0.358 11 168 0.406 23 135 0.448 5L-63 147 0.526 15B-39 138 0.625 5L-70 176 0.735 57 144 1.2385 15B- 5 227 1.315 89 186 1.319 5L-65 142 1.388 15B-90 211 1.641 5L-59 107 1.779 15B-57 172 1.864 43 179 1.915 12 192 1.953 Motalboan Son ae. 2,615 W163) Meant: 56° 4e ees 162.8 1.109 286 Blood Fat in Fowls TABLE IV. Non-Laying Hens. > oo No. Yearly total. Fat. Days since last egg was laid. per cent 15B- 9 178 0.083 61 15 120 0.085 66 13 202 0.088 aS 54 113 0.114 110 16 139 0.123 66 51-54 148 0.123 36 15B-56 124 0.124 20 18 186 0.125 9 3 168 0.131 78 60 153 0.131 22 74 165 0.133 28 49 220 0.138 10 5L-53 144 0.144 65 15B-51 154 0.154 6 19 156 0.161 61 5L-66 134 0.166 55 15B- 1 166 0.167 60 83 102 0.167 58 5L-52 147 0.167 23 15B-96 146 0.169 17 5L-55 134 0.175 22 15B-38 156 0.178 36 99 189 0.179 6 105 101 0.180 35 52 10 0.180 212 46 110 0.181 83 5L-71 140 0.183 46 - 68 133 0.187 88 15B-55 150 0.187 22 58 198 0.188 13 20 141 0.189 26 5L-64 170 0.191 6 77 101 0.196 60 15B-80 149 0.198 8 31 113 0.202 51 36 150 0.202 39 37 167 0.204 33 66 72 0.204 82 34 171 0.205 24 25 134 0.223 38 D. E. Warner and H. D. Edmond 287 TABLE IV—Concluded. No. Yearly total. _ Fat. Daveaince last eee per cent 62 96 0.228 95 98 117 0.228 55 104 38 0.230 53 82 113 OR23i 54 Ud 166 0.241 10 91 171 0.242 22 28 132 0.253 51 5L-67 91 On272 55 5L-62 146 0.300 62 15B-97 152 0.3807 3 5L-60 149 0.312 BY 51 139 0.437 50 15B-64 0 0.448 365 24 147 0.541 32 Mean. ms... ..- 139.09 0.199 ole when the blood samples were taken. The average per cent of fat found in the blood was much higher for birds that were laying than for the group of birds that were not laying (Tables III and IV). This is very striking and quite agrees with the averages reported earlier.» One-half of the birds had over 1 per cent of fat in their blood, as shown in Table III. Table IV shows that the per cent of fat in the blood of non- producers was very much lower than the per cent of fat in the blood of laying birds. Every bird in this group had stopped lay- ing from 6 to 365 days previous to the time the blood samples were taken. The average egg production for this group of birds was 139.1, which is a much lower average than for the birds that were laying when the samples were taken. Not only this, but the average per cent of fat was much lower, while the average days since they had produced any eggs was 51.7. Only one individual in this group of birds had over 0.5 per cent of fat in the blood while others gave a reading as low as 0.083 per cent. Table V shows that if all the birds having over 1 per cent of fat in their blood are grouped together. the average egg produc- tion is found to be 173.4, or an average of over 55 eggs more than that of the birds that had yellow beaks, legs, and vents, THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 Blood Fat in Fowls bo CO (96) TABLE V. Data for Hens with Over 1 per Cent of Fat in the Blood, and with Pale Beaks, Legs, and Vents. No. Yearly total. Fat. Days eee per cent 5L-57 144 1.235 0 15B- 5 227 1.315 0 89 186 1.319 0 5L-65 142 1.388 0 15B-90 211 1.641 0 51-59 108 1.779 0 57 172 1.864 0 43 179 1.915 0 12 192 1.953 0 Totals sees: 1,561 14. 409 0 Means. oes cc... 173.4 1.601 0 and a low fat content. Further, this group of birds were all laying when the records were taken, and higher annual egg pro- duction would of course be expected from birds that had contin- ued.to lay until October, the time the blood samples were taken. As a rule, if a bird has been laying for some time the legs, beak, and vent undergo a physiological change, and where they continue to lay, these parts will become decidedly pale in color. Table VI shows that what is true of the external parts of the body is quite true of the blood. That is, the blood of birds hay- ing external parts that have bleached have on the average a higher per cent of fat in their blood. The fat content of the blood is correlated with egg laying activity, and there is also a slight cor- relation between the amount of fat in the blood and high yearly production. The amount of fat is also correlated with the fad- ing of the external parts. There is an average of 0.816 per cent of fat in the blood of hens having pale beaks, legs, and vents. When this is compared with the data for the group of hens hay- ing these parts yellow there is found to be only an average of 0.196 per cent of fat (Tables VI and VII). Table VII shows the reverse of Table VI in the average yearly total per hen, average per cent of fat, and average number of_ days since the last egg was laid. The average production was only 117.9 for the birds having yellow legs, beaks, and vents, D. E. Warner and H. D. Edmond 289 TABLE VI. Data for Hens with Pale Beaks, Legs, and Vents. No. Yearly total. Fat. Dayeisinte tastes per cent 15B-60 153 0.131 22 51 154 0.154 6 99 189 0.179 6 105 101 0.180 35 51-64 170 0.191 6 15B-77 166 0.241 10 32 158 0.258 0 if 168 0.406 0 23 135 0.448 0 51-63 147 0.526 0 7 176 0.735 0 15B-5 227 1.315 i) 89 186 1.319 0 5L-57 144 1.235 0 15B-90 211 1.641 0 57 172 1.864 0 43 179 1.915 0 12 192 1.953 0 MOtalen 20-45. 3,028 14.691 85 Means 25. 168.20 0.816 4.72 whereas the average production for the birds that were pale in ‘those parts was 168.2. The per cent of fat found in this group of birds was only 0.196, whereas in the group of birds having pale parts it was 0.816. The average number of days since the last eggs were laid for the group of birds having pale parts was 4.722, while for the group of birds in Table VII it was 75.21. This would indicate that if a bird had yellow legs, yellow beak, and a yellow vent, the per cent of fat in the blood would be low. It has been reported by Matthews that in man and animals the blood is much richer in fat after they have been eating than it is after fasting. This is not true with hens, as shown from Table VIII; the average per cent of fat for the birds that had been without food is a little higher than those that had not. The difference is not great, however, 0.009 per cent, but would indicate that fasting for 16 hours has no effect upon the amount of fat in the blood. | 290 Blood Fat in Fowls TABLE VII. Data for Hens with Yellow Beaks, Legs, and Vents. No. Yearly total. “Fat. , Days since last egg fr 3 was laid. hi per cent ; 107x 110 0.066 56 | 15B- 9 178 0.083 61 | 15 120 0.085 66 425x 120 0.100 46 15B-54 113 0.114 110 16 139 0.128 66 5L-54 148 0.123 36 76x 111 0.126 » 73 15B- 3 168 0.131 78 51-53 144 0.144 65 15B-19 156 0.161 61 401x 93 0.166 51 | 15B- 1 166 0.167 60 | 430x 107 0.168 63 15B-52 10 0.180 212 46 110 0.181 83 5L-71 140 0.183 46 15B-55 150 0.187 22 102x 113 0.194 61 5L-77 101 0.195 60 15B-31 113 0.202 51 404x 101 0.208 66 116x 83 0.228 79 15B-98 117 0.228 55 62 96 0.228 95 82 113 0.231 54 51-67 . 91 0.272 55 110x 128 0.297 62 5L-62 146 0.300 62 60 149 0.312 37 51 139 0.437 50 15B-64 0 0.448 365 Mest al hee ee ee EWEe 6.268 2,407 Meantinssc...: 117.906 0.196 (5221 Table IX shows the per cent of fat found in the blood of male birds. The figures for fat from these few birds indicate that the per cent of fat in the blood of male birds was more constant than in hens. This may be accounted for by the fact that the beaks, \ } 4 D. E. Warner and H. D. Edmond 291 TABLE VIII. Fat in the Blood of Hens Fasted for 16 Hours and Hens Not Fasted. Fasted. Not fasted. per cent per cent 0.133 0.164 0.154 0.187 0.178 0.188 0.179 0.198 0.202 0.228 0.223 0.231 0.228 0.241 0.230 0.242 0.253 0.258 0.541 0.307 0.625 0.448 1.915 1.864 4.861 4.753 Average...... 0.405 0.396 TABLE IX. Fat in the Blood of Twelve Male Birds. No. Fat. per cent 108 0.097 104 0.141 99 0.146 96 0.149 7 0.160 98 0.166 65 0.170 0 0.191 B-78 0.194 20 0.200 33 0.246 27 0.249 AROTEN S35 on oS ao oo eee ee 2.109 MIGBIING 54 Selo dads oO ee 0.176 legs, and ear lobes of male birds do not change in color and that there is no building of eggs within the body cavity. The reserve fat stored in the body is not used up in the same way that the female uses hers, and consequently there is no cause for the fad- 292 Blood Fat in Fowls ing in the parts mentioned above. The average per cent of fat in a male bird is 0.176, or about the same as the per cent found in the blood of 3 year old hens. The blood of Bird 108 was very low in fat. This is suggestive and leads to the investigation whether the male birds that are prepotent can be detected by the fat content of their blood. (By prepotent we mean a male bird which transmits to its female offspring high fecundity.) TABLE X. Fat and Cholesterol in the Blood of a Few Laying and a Few Non-Laying Hens, and Also a Few Roosters. No. Cholesterol. Fat. Laying hens. per cent per cent 26 0.019 1.915 9 0.047 0.526 32 0.076 0.258 28 0.082 1.864 61 0.110 1.388 76 0.114 0.246 16 0.121 1.953 64 0.149 1235 34 0.155 0.448 62 0.169 1.779 5 0.214 1.315 Meanteeear eter 0.114 aie Non-laying hens. 41 0.023 0.138 35 0.077 0.241 60 0.078 0.180 39 0.081 0.124 68 0.106 0.191 43 0.118 0.448 79 0.121 0.735 Meanie aces tae 0.086 0.294 Male birds. 95 0.069 0.097 96 0.075 0.166 94 0.089 0.170 93 0.110 0.146 Mean: .ccecerereee ee 0.086 0.145 D. E. Warner and H. D. Edmond 293 SUMMARY. The present paper is a contribution to the general problem of the relation between the amount of fat and fecundity. The data comprise the per cent fat valuations and the egg records for eighty-two hens. The egg records cover a period from November to October inclusive. Blood samples were taken from October 28 to Novem- ber 3, 1916. There is little or no correlation between the amount of fat in the hen’s blood and her yearly egg yield. On the other hand the blood of a hen laying at the time the sample is taken is much richer in fat than that of a hen that is not laying. The average per cent of fat found in the seventy hens that had just completed their Ist year of laying was 0.407; for the twelve old hens having just completed their 3rd year of laying the per cent was 0.171; for the twelve 14 year old male birds it was 0.176. From the data at hand it seems improbable that high producing hens can be selected merely by sampling their blood and analyz- ing for fat unless one is careful to take into account whether or not a hen is laying at the time of sampling, and also the season of the year. Depriving hens of food for 16 hours did not seem to lessen the ‘per cent of fat in their blood. There exists a close correlation between the color of the beak, legs, and vent and the per cent of fat found in the blood. The birds that had pale legs, pale beaks, and pale ani carried a very high per cent of fat in their blood and also had a high average egg production. The birds that had distinctly yellow legs, beaks, and ani gave a very low average egg production, and the per cent of fat carried in their blood was very low. This would show that: birds that were not laying were storing fat in the body cells, and consequently their legs, beaks, and ani would become yellow, the natural color for all American breeds and the Leghorns. The average per cent of fat found in the blood of the 3 year old hens was much lower than that found in the blood of the 1 year old birds. The per cent of fat found in the blood of male birds was more constant than that of 1 year old hens. There 294 Blood Fat in Fowls = ” did not appear to be very much difference between the per cent of fat found in the blood of male birds and that found in the blood of females not laying. It seems feasible to eae that the principal reason why the blood of laying hens is much richer in fat than that found in birds not laying is that the fat stored in the body tissues is taken up by the blood and carried to the egg yolk. TRITICO NUCLEIC ACID. By B. E. READ anp W. E. TOTTINGHAM. (From the Laboratory of Physiological Chemistry, Johns Hopkins University, Baltimore.) (Received for publication, June 7, 1917.) Recent investigations have definitely indicated that there are ‘two distinct types of nucleic.acid, one of which is found in animal tissues and the other in plants. While this distinction rests upon the examination of a large number of nucleic acids of animal origin, our knowledge of plant nucleic acid has been acquired from a study of but two substances; one of these is prepared from yeast (yeast nucleic acid) and the other from the wheat embryo (tritico nucleic acid). In 1902 Osborne and Harris! prepared tritico nucleic acid and showed that like all other nucleic acids it produces phosphoric acid, guanine, and adenine by hydrolysis with dilute mineral acids. But tritico nucleic acid was found to differ from all animal nu- cleic acids and to resemble yeast nucleic acid by its production of uracil instead of thymine, and pentose instead of levulinic acid. At the time Osborne and Harris’ paper appeared cytosine was not known, but Wheeler and Johnson? afterwards found cytosine among the hydrolytic products of tritico nucleic acid. So far as concerns their fundamental groups, therefore, the two plant nucleic acids differ sharply from all animal nucleic acids and are identical with one another. After Levene and Jacobs? had shown that by hydrolysis with ammonia at high temperatures yeast nucleic acid loses its phos- phoric acid and produces four nucleosides, it became of great in- terest to know whether tritico nucleic acid would behave in the 1 Osborne, T. B., and Harris, I. F., Z. physiol. Chem., 1902, xxxvi, 85. ? Wheeler, H. L., and Johnson, T. B., Am. Chem. J.; 1903, xxix, 505. ? Levene, P. A., and Jacobs, W. A., Ber. chem. Ges., 1910, xliii, 3150. 295 296 Tritico Nucleie Acid same manner. Levene and La Forge’ found this to be the case. They described a very efficient method of preparing tritico nucleic acid and found that the substance yields three of the four nucleo- sides of yeast nucleic acid (guanosine, adenosine, and cytidine). They made no attempt to find uridine. Hence, not only in their fundamental groups but in the forma- tion of nucleosides the two plant nucleic acids are seen to be iden- tical so far as experiments have been made, and we shall extend the coincidence to include uridine. It has very recently been shown in this laboratory® that when yeast nucleic acid is hydrolyzed with ammonia at low temperatures no phosphoric acid is removed and two nucleotides are formed; vig., guanine mononucleotide and adenine-uracil dinucleotide. In the following we show that these same nucleotides are produced by tritico nucleic acid. The guanine mononucleotide obtained from tritico nucleic acid does not differ either in its chemi- cal composition or in any of its chemical properties from the sub- stance formerly prepared from yeast nucleic acid. By mild acid hydrolysis it produces guanine but not adenine, and sets free its entire phosphoric acid. By treatment with brucine it forms a erystalline brucine salt which melts at 203°C. Adenine-uracil dinucleotide prepared from tritico nucleic acid is identical with that prepared from yeast nucleic acid. By mild acid hydrolysis it produces adenine but not guanine, and liberates only half of its phosphoric acid. By hydrolysis with ammonia at high temperatures it produces both adenosine and uridine. It forms with brucine a crystalline brucine salt which. has the compo- sition required for the formula Cy9He;5N 7O1sP2(C23He6N204) 4. 14H20, and melts at 175°C. The nitrogen percentage of the brucine salt shows that it con- tains a uracil group, not a cytosine group, and that uracil and uridine are not obtained from tritico nucleic acid as secondary products to cytosine and cytidine. The presence of four brucine radicles in the brucine salt leads to the disaccharide structure of the dinucleotide.® 4 Levene, P. A., and La Forge, F. B., Ber. chem. Ges., 1910, xliii, 3164. 5 Jones, W., and Read, B. E., J. Biol. Chem., 1917, xxix, 111. Read, B. E., ibid., 1917, xxxi, 47. 6 For argument see Jones and Read, J. Biol. Chem., 1917, xxix, 113. B. E. Read and W. E. Tottingham 297 O=P — O.C;H,02.C;H4Ns O HOW O=—E = O e C;H,O2 . C.,H3;N:202 HO Although it is not possible to prove rigidly the identity of two substances the exact coincidence in so many properties of tritico nucleic acid with yeast nucleic acid leaves little room to doubt that they are the same chemical compound. EXPERIMENTAL. 100 gm. of tritico nucleic acid prepared from the wheat embryo by the method of Levene and La Forge* were dissolved in 530 ec. of 2.5 per cent ammonia, and the solution was heated in an autoclave for 14 hours at 115°. The cooled product was treated with 530 ce. of absolute alcohol when the ammonium salt of guanine mononucleotide was precipitated, leaving the ammo- nium salt of adenine-uracil dinucleotide in solution. The two sub- stances were separated by filtration, purified, and the two nucleo- tides finally obtained as white powders by the methods already described in connection with yeast nucleic acid.® Guanine Mononucleotide—Commercial yeast nucleic acid evi- ‘dently contains various impurities which contaminate products that are prepared from it. This is not the case with tritico nu- cleic acid which produces guanine mononucleotide so free from foreign substances that its purification offers little difficulty. The substance which we obtained gave the following analytical data. I. 0.2634 gm. dried at 100° required 13.65 cc. H2SO, (1 ec. = 0.0037 gm. N). Menem OE a ag = 90087 fe), IIl..0.4566 “ “ “ “ gave 0.1415 gm. Mg2P20;. TV ete tee ce “SEE =p pOmlo Zn iF N P CHIC GGL... a 19 28 8.54 Found. MeaSeo56e4e sds any ee 19.17 iO ee 19.11 298 Tritico Nucleic Acid Hydrolysis of Guanine Mononucleotide with Dilute Mineral Acid. —Portions of the guanine mononucleotide were heated with twenty volumes of 5 per cent sulfuric acid and estimations of phosphoric acid and guanine were made with the product. I. 0.3487 gm.dried at 100° and hydrolyzed for 3 hrs. gave 0. Mie Mg:P:07. Ps 0. 3394 ac “ce cc 100° “ce a3 “ce SB ae “ 0. 1032 III. 0.4924 “ sa OOe 5 ie “lhr. “ 0.1806 “ guanine. IV. 0.7170 ce oe “ce 100° “ ce “ce 1 “ “ 0.2677 “ce “ Iz ‘ Guanine. Calculated. > o242.5: ©. ore 8.54 41 6 Found. DDS et see pete ee eo 8.57 fH Leer ara See ie eer ert oe eee 8.49 iH i] TS eo: oe EN ee 1s ett es SRA = | 36.68 OE ne he tos ee 37.30 These results show that the substance loses its entire phosphoric acid by mild acid hydrolysis and therefore contains no pyrimidine group. The final filtrates from magnesium ammonium phosphate gave no gelatinous precipitate of silver-adenine upon treatment with ammoniacal silver solution. The Brucine Salt of Guanine M. a LiL —A solution of 1 gm. of guanine mononucleotide in 5 ec. of hot water was treated with 2.5 gm. of brucine in 5 ee. of hot absolute alcohol. The erys- talline dibrucine salt was immediately deposited in characteristic rosettes, which after washing with hot absolute alcohol and re- crystallization from hot water melted at 203°C. Preparation of Guanosine from Guanine Mononucleotide—Guan- ine mononucleotide was heated with five parts of 2.5 per cent ammonia in an autoclave for 2 hours at 140°C. On cooling in the ice chest guanosine was deposited as a gelatinous mass. This was filtered on a Buchner funnel and recrystallized from hot water with the use of animal charcoal. Pure guanosine was ob-. ‘tained in long transparent needles which on analysis gave the following results. B. E. Read and W. E. Tottingham 2.99 Ii: 0.2382 gm. air-dried substance required 14.11 cc. H2SO; (lec. = 0.0037 gm. N). II. 0.8086 gm. heated at 120°C. lost 0.0905 gm. H2O N Calculate deetpactet cer to ee. 11.29 21.94 Found. Ih cee och See Rare ee 21.92 IM et ee ye fo uit a csie, volviye 11.19 Adenine-Uracil Dinucleotide—Specimens of adenine-uracil di- nucleotide prepared as described were submitted to hydrolysis with 5 per cent sulfuric acid. The product gave no precipitate of guanine on the addition of ammonia. The solution was treated directly with magnesia mixture and the free phosphoric acid determined as magnesium ammonium phosphate. The final fil- trate from magnesium ammonium phosphate gave the character- istic gelatinous silver-adenine precipitate upon treatment with am- moniacal silver solution; so that the particular structure of the compound as a dinucleotide composed of one pyrimidine group containing firmly bound phosphoric acid and one purine group con- taining a phosphoric acid radicle easily set free by hydrolysis with dilute mineral acid is thus fully established.7? Analyses for total and partial phosphorus gave results consistent with this structure. I. 0.6936 gm. dried in a desiccator gave 0.2442 gm. MgNH,PO,.6H20. ele 0.7803 ce “ce ce ee “ ce 0.2751 73 ce TU Ge S02 Fe “after complete destruction of organic matter 0.7762 gm. MgNH,PO,..6H20. We I: Il. Bulppuamee Used. 2. 2c ccc. be es 0.6936 0.7803 1.2024 dimevotibydrolysis.«......2.0.:.... 3 hrs. 3 hrs: Total. MgNH,PO,.6H.O obtained......... 0.2442 0.2751 0.7762 Amount calculated per gm.........| 0.3521 0.3525 0.6450 Pyrimidine correction.............. 0.0300 0.0300 _ From purine nucleotide............| 0.3221 0.3235 Half the total phosphorus......... 0.3225 7 For full discussion of this subject see Germann, H. C., J. Biol. Chem., 1916, xxv, 189. 300 Tritico Nucleic Acid The above results show that exactly half of the phosphoric acid of this dinucleotide is easily split and the other half is firmly bound. Preparation of Uridine from the Dinucleotide —25 gm. of di- nucleotide were dissolved in 125 ce. of 2.5 per cent ammonia and heated in an autoclave at 133°C. for 2 hours. The product was allowed to stand in ice water for several hours but no trace of guanosine was deposited. A portion of the autoclave product was warmed and treated with a hot aqueous solution of picrie acid until cold picric acid failed to give a precipitate with a drop of the fluid. Upon re- crystallization of the bulky gelatinous precipitate from hot water, adenosine picrate was obtained in characteristic transparent yel- low plates. The main portion of the autoclave product was treated for uri- dine by the method of Levene and La Forge.’ Snow-white crys- talline uridine was obtained, which melted at 158° (corrected), and gave the following analysis. 0.3746 gm. air-dried substance required 11.57 cc. H2SO, (1 ce. = 0.0037 gm. N). N @aleullatedesacece cece ee ere ne Oe es Ee Eee eee 11.48 1 to) hao Pernt Seen A eee en ne RSE Tae Ran RS Ree Pe Sa ee 11,438 Brucine Salt of the Dinucleotide—8 gm. of substance dissolved in 9 cc. of hot water were treated with a solution of 7.2 gm. of brucine in 15 ce. of hot absolute alcohol. After washing the pre- cipitated brucine salt thoroughly with hot alcohol, it was recrys- tallized from fifty parts of hot water. The characteristic crystal- line substance melted at 175°, and gave the following analytical data. II. 0.4300 “ gave 33.2 cc. N at 757 mm. and 28°C. III. 0.4458 “ CO SOSO er OSs peoee 21 Ges ; LY Saez “on complete oxidation of all organic matter 0.1098 gm. MgoP2O;. V. 1.3013 gm. gave on complete oxidation of all organic matter 0.1189 gm. Mge2P20;. VI. 0.3867 gm. gave 0.2750 gm. crystalline brucine, C23H2gN204.4H20 (no chloroform extraction was made of the mother liquor). B. E. Read and W. E. Tottingham Found. N P. 8.46 2.49 8.44 8.47 2.49 2.55 301 Brucine. CAA THE PROTEOCLASTIC TISSUE ENZYMES OF THE SPLEEN. By MAX MORSE. (From the Nelson Morris Memorial Institute for Medical Research of the Michael Reese Hospitul, Chicago.) (Received for publication, June 11, 1917.) Problem.—Two autolytic enzymes have been described in the spleen. The one is described (1-4) as being active in an alkaline medium, adsorbable by kieselguhr and extractible with 5 per cent NaCl solution. This is the a-protease of Hedin and his cowork- ers. The other enzyme which is capable of hydrolyzing proteins is the 6-protease of these investigators, having the characteris- tics of being salted out with ammonium sulfate and of being extractible with dilute acetic acid. It is not adsorbed by kiesel- guhr. The writer has attempted to determine the range of activity of autolyzing enzymes, and inasmuch as it appears that the spleen varies from other tissues in regard to the relation between en- zyme activity and hydrogen ion concentration (5), a study has been made concerning this relation in spleen enzymes. Methods. Spleens from the ox and the dog were used; the former, obtained from the hospital butcher, was about 48 hours old; the latter was prepared from a freshly killed dog. No characteristic differences in spleen enzymes were observed in the two cases. The digests were prepared in the same manner in which they have been prepared in the former studies on tissue enzyme action (6). The tunica was removed before the spleen pulp was ground. Suspensions were made in the various phosphate mixtures given below so that 40 gm. net weight of _ spleen tissue were made up to 100 gm. by means of the phosphate mixtures. These mixtures were prepared from Sérensen’s chart (7), the pH varying from 5.68 to 7.59 for beef spleen and to 7.98 for that of the dog. These mix- tures were then read on a Leeds and Northrup Type K potentiometer. A 0.1 -N calomel electrode, Leeds and Northrup standard cadmium element, and a Clark (8) oscillating hydrogen vessel with platinum electrode were 303 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 304 Spleen Enzymes used. The hydrogen was obtained electrolytically and purified by means of a palladium sponge filter of de Khotinsky.! a-Amino nitrogen determinations were made by means of the gasometrie method of Van Slyke. Aliquots were used from the filtrates of 2.5 per cent trichloroacetic acid precipitates for these determinations. Amino nitrogen per 40 gm. of fresh tissue. Reaction. Initial. 24 hrs. 6 days. pH gm. gm. gm. 5.68 OF 221+ 0.3438 6.55 0.221 0.322 6.58 0.221 0.290 6.80 0.221 0.268 6.94 0.221 0.238 7.08 0.221 0.238 7.59 0.221 ; 0.211 7.98 0.0053** 0.0053 0.0055 * Beef ** Dog : DISCUSSION. It will be observed that digestion occurs where pH <7.08. pH 7.07 is approximate neutrality at the temperatures at which these readings were made. A departure from neutrality towards alkalinity results in inhibition of hydrolysis as far as the proteins of the tissue are concerned. Apparently the enzyme or enzymes are active at an alkalinity of pH 7.5-7.9, as will be seen from the following experiment where 5 gm. of Witte’s peptone were intro- duced into a digest composed as the one given above for dog spleen in pH 7.98 phosphate mixture. Amino nitrogen per 40 gm. A of tissue. Reaction. Initial. 24 hrs. pH gm. gm. Controleiiecesc cap ls tat epee 7.98 0.054 0.054 ¥ + 5 gm. of peptone....... | 7.98 0.107 0.217 1 Throughthe generosity of Mr. Albert Kuppenheimer of Chicago the In- stitute has been able to procure physicochemical apparatus of various sorts. The present work was done by means of the apparatus for electromotive force, purchased from this fund. Max Morse 305 The peptone is digested to a marked extent, showing that a protein-hydroly zing enzyme (or enzymes) is active. It is difficult to explain, however, why, if the enzymes are active at neutrality and in an alkaline reaction, no, increase in a-amino nitrogen is obtained from the polypeptides which occur in partly autolyzed tissue, such as that obtained 48 hours after the animal’s death; for the writer has determined that peptones and higher amino- acid linkages occur to a considerable extent in liver from a meat market. On adding large amounts of peptone and albumose, as in these experiments, a correspondingly large increase in amino nitrogen is obtained over that which is present initially. The conclusion is unescapable that native peptones and other poly- peptides are not hydrolyzed by the tissue enzymes under the conditions of these experiments, although those introduced are digested. Inasmuch as these experiments show that as far as the native proteins of the spleen are concerned, no alkaline-acting enzyme is recognizable capable of hydrolyzing these proteins, the experi- ments are divergent from those of Hedin and others who have described an enzyme of this nature, the so called a-protease. The question arises as to whether the a-protease is a true auto- lytic enzyme, belonging to the spleen tissue, or whether it is not of the nature of Martin Jacoby’s heterolytic enzyme, and is de- rived from the leukocytes which abound in the spleen. For these ‘bodies, alkaline-acting enzymes have been described by Opie (9) and later by Jobling and Strouse (10). Leukoproteases have been described by Leber (11), Jochmann and Ziegler (12), and others, following the initial work of Miller (13), but the range of their activity in regard to reaction of medium is not known. CONCLUSIONS. An enzyme or enzymes exist in the spleen, capable of hydrolyz- ing peptone and also (as shown by Hedin) fibrin. The proteins of the spleen itself, however, autolyze only in neutral or acid, not in alkaline solution. The most rapid autolysis was observed in the most acid solution tested, pH = 5.68. From the standpoint of necrosis, the a-protease of Hedin can scarcely be operative since there is no evidence that it affects na- tive proteins. 306 Spleen Enzymes It is probable that the a-protease of Hedin is not an autolytic enzyme, but rather a heterolytic one, resident in the white blood cells, for such an enzyme has been described, definitely, for such cells. BIBLIOGRAPHY. 1. Hedin, 8. G., and Rowland, 8., Ueber ein proteolytisches Enzym in der Milz, Z. physiol. Chem., 1901, xxxii, 341. . Hedin, 8. G., Investigations on the proteolytic enzymes of the spleen of the ox, J. Physiol., 1903-04, xxx, 155. 3. Leathes, J. B., On the products of the proteolytic action of an enzyme contained in the cells of the spleen, J. Phystol., 1902, xxviii, 360. 4. Catheart, E. P., On the products of digestion of the proteolytic spleen enzyme acting in an alkaline medium, J. Physiol., 1905, xxxii, 299. . Morse, M., Enzyme and reaction of medium in autolysis, J. Biol. Chem., 1917, xxx, 197. 6. Bradley, H. C., and Morse, M., Studies of autolysis. I. The acceler- ating effect of manganous chloride on liver autolysis, J. Biol. Chem., 1915, xxi, 209. . Sdrensen, S. P. L., Enzymestudien. II. Uber die Messung und die Bedeutung der Wasserstoffionenkonzentration bei enzymatischen Prozessen, Biochem. Z., 1909, xxi, 131. Clark, W. M., A hydrogen electrode vessel, J. Biol. Chem., 1915, Xxill, 475. 9. Opie, E. L., The enzymes in phagocytic cells of inflammatory sere | J. Exp. Med: 1906, viii, 410. 10. Jobling, J. W., and Strouse, S., Studies in ferment action, II. The extent of leucocytic proteolysis, J. Exp. Med., 1912, xvi, 269. 11. Leber, T., Die Entstehung der Entziindung, und die Wirkung der ent- zindungerregenden Schidlichkeiten, Leipsic, 1891. 12. Jochmann, G., and Ziegler, C., Ueber das Leukozytenferment in Milz, Miinch. med. Woch., 1906, liii, 2093. 13. Miller, E., Uber das Verhalten des proteolytischen Leukocytenfer- mentes und seines ‘‘Antifermentes’”’ in den normalen und krankhaften Ausscheidungen des menschlichen Kérpers, Deutsch. Arch. klin. Med., 1907, xci, 291. bo Cr J (v/a) COMPARATIVE METABOLISM OF CERTAIN AROMATIC ACIDS. By CARL P. SHERWIN. (From the Laboratory of Fordham University Medical School, New York.) (Received for publication, June 11, 1917.) Fate of Phenylacetic Acid in the Organism of the Monkey. Many experiments have been reported showing a marked dif- ference between metabolism in the human body and in the lower animals. This alone has stimulated considerable interest on ac- count of the apparent lack of reason for such difference. In the last few years experiments have been carried out with animals more closely related to the human in order to compare their metabolism with that of both human beings and lower animals. Hunter and Givens (1) report the purine metabolism in the monkey as resembling that of the lower animals and not the human. In the urine of the monkey (Cercopithecus) allantoin accounted for 73 per cent of the nitrogen arising from the catabo- lism of the endogenous purines while the remainder appeared principally as purine bases and none as uric acid. Later work by the same authors (2) shows like results for another monkey (Cercopithecus callitrichus). Wiechowski (3) obtained results comparable to those of Hunter and Givens but was unable to find allantoin as the end-product of the purine metabolism of the chimpanzee. Baumann and Oviatt (4) found the urinary sulfur excretion in the monkey (Macacus) quite different from that of man. In the case of the monkey the ratio of inorganic to ethereal sulfates was approximately 34 to 1 while for man it bore the ratio 12, tomk. An interesting phase of this same question is the manner in which the different animal organisms detoxicate certain protein putrefactive products. The series of compounds resulting from the action of putrefactive bacteria on the aromatic amino-acids must be detoxicated and eliminated from the animal body as 307 308 Metabolism of Aromatic Acids soon as possible. This detoxication process seems often to differ for different species or at least results in the excretion of an en- tirely different end-product for the acid thus resorbed. Phenylacetic acid is one of those aromatic acids which yield different conjugation compounds in the urine after ingestion by men and animals respectively. Phenylacetic acid in combination with glycocoll is excreted normally by many of the herbivora as phenaceturic acid. Phenaceturic acid was discovered by Sal- kowski (5) in normal horse urine. Phenaceturic acid has been found in the urine of many different animals after feeding phenyl- acetic acid. Thus E. and H. Salkowski (6) ‘isolated it from the urine of dogs and rabbits, and Vasiliu (7) from the urine of sheep after phenylacetic acid feeding. E. Salkowski (8) believed he had found traces of the phenaceturic acid in normal human urine. Totani (9) isolated from the excreta of chickens phenylacetylor- nithuric acid after phenylacetic acid feeding. Thierfelder and Sherwin (10) found in human urine two compounds, phenylacetyl glutamine and phenylacetyl glutamine urea after various amounts of phenylacetic acid had been ingested but were unable to find even a trace of either phenaceturic acid or uncombined phenyl- acetic acid. Before beginning certain metabolism work on a monkey it was important to know whether the fate of the phenylacetic acid in the organism of the monkey would resemble that of the human organism or that of the lower animals. The combination with glycocoll resulted showing the metabolic process to resemble that of the lower animals, not only in the case of the phenylacetic but also for the p-hydroxyphenylacetic and p-hydroxy benzoic acid (11). EXPERIMENTAL. A 4.2 kg. female monkey (Macacus rhesus) was placed on a regular diet of milk, bread, bananas, and apples for several days in order to see if the urine contained traces of any phenylacetic acid compound, After no com- pound of this nature could be found in the urine, the monkey was fed 1 gm. of phenylacetic acid per day as the soluble sodium salt. The sodium salt dissolved in fresh milk at first seemed quite acceptable, but after the 3rd day it was refused. Any food or liquid possessing even the slightest odor of phenylacetic acid was entirely refused by the monkey. At this point forced feeding was resorted to and the water solution of the sodium salt introduced directly intothe stomach by means of astomach tube. The C. P. Sherwin 309 amount thus given varied betweenland2 gm. The physiological effect seemed to be much the same as that produced on human beings. Each dose was followed by a period of dullness and inactivity, while the ingestion of 2 gm. (0.47 gm. per kg. body weight) resulted in an entire loss of appetite for 24 hours. When the phenylacetic acid is ingested by men in quantities varying from 0.05 to 0.26 gm. per kg. body weight intoxication results much the same as that produced by alcohol. The monkey’s urine was collected for 36 hours after each dose of the acid, and the different portions were united and acidified with phosphoric acid until a distinct acid test with Congo red resulted. The concentrated urine was then extracted repeatedly in a liquid extracting apparatus with ethylacetate until no phenylacetic acid compound formed by evaporation of the last extraction. The ethylacetate extract thus prepared was evap- orated to one-half its original volume and placed on ice for 24 hours. As no phenylacetyl glutamine crystallized out at this concentration, the extract was again evaporated and the evaporation continued on subsequent days until the entire ethylacetate solution amounted to about 100 cc. only. At this concentration, there being no possibility of obtaining any phenylacetyl glutamine, the extract was evaporated carefully almost to dryness and then taken up with 50 cc. of water and boiled with charcoal to remove the pigment. After filtering this water solution from the charcoal, evaporat- ing, and allowing the concentrated solution to stand on ice, there appeared small rhombic crystals of phenaceturic acid. The mother liquor from these crystals was optically inactive, showing again the absence of phenylacetyl glutamine. After the feeding of 1 gm. of phenylacetic acid the amount of phenaceturic acid isolated from the urine amounted to only 0.82 gm. or 51 per cent of the expected yield. The phenaceturic acid was twice recrystallized from water, dried in vacuo, and then used for analyses. Melting point 142-143°. 0.1232 gm.-of the acid required 6.29 cc. of 0.1 N sodium hydroxide instead of 6.37 cc., the theoretical amount. 0.1331 gm. substance gave 8.50 cc. nitrogen at 21° and 748 mm. pressure. Calculated for CioHiu NOs: Found: 1M 9 oe, ong SO ERO IOI ea ee a 7.25 7.15 CONCLUSION. The metabolism of phenylacetic acid in the organism of the mon- key is the same as that found in the lower animals and entirely different from the metabolism of the same substance in man. Phenylacetic acid in the monkey is conjugated with glyco- coll and excreted as phenaceturic acid, while in man it is con- jugated with glutamine and excreted partly as phenylacetyl glut- amine and partly as phenylacetyl glutamine urea. 310 Metabolism of Aromatic Acids _ eS . Hunter, A., and Givens, M. H., J. Biol. Chem., 1912-13, xiii, 371. . Hunter and Givens, J. Biol. Cher., 1914, xvii, 55. . Wiechowski, W., Biochem. Z., 1909, xix, 368; Prag. med. Woch., 1912, . Baumann, L., and Oviatt, E., J. Biol. Chem., 1915, xxii, 43. . Salkowski, E., Z. physiol. Chem., 1885, ix, 229; Ber. chem. Ges., 1884, . Salkowski, E., and Salkowski, H., Ber. chem. Ges., 1879, xii, 653; Z. . Vasiliu, Mitt. landw. Inst. Univ. Breslau, 1909, iv, 703. . Salkowski, E., Z. physiol. Chem., 1885, ix, 229. . . Thierfelder, H., and Sherwin, C. P., Ber. chem. Ges., 1915, xlvii, 2630; . Results unpublished. BIBLIOGRAPHY. xxxvii, 275. xvii, 3010. physiol. Chem., 1882-83, vii, 161. Z. physiol. Chem., 1915, xciv, 1. Sherwin, Inaugural Dissertation, Tubingen, 1915. STUDIES ON BIOLUMINESCENCE. VUI. THE MECHANISM OF THE PRODUCTION OF LIGHT DURING THE OXIDATION OF PYROGALLOL. By E. NEWTON HARVEY. _ (From the Physiological Laboratory, Princeton University, New Jersey.) (Received for publication, June 8, 1917.) INTRODUCTION. It has been very definitely established from the observations of many investigators that the production of light by animals is a chemiluminescent phenomenon; 7.e., a luminescence accompany- ing a chemical reaction. As free oxygen is necessary for light production in all forms, the reaction may be classed among the oxidations. In several animals two substances (Harvey, 1) in addition to water and oxygen are necessary to produce light and at least one of these is oxidized. As a number of compounds of definite chemical composition are known which emit light dur- ing oxidation, it seemed desirable to study such reactions more in detail in order to determine under what conditions the light is emitted. Ina previous paper (Harvey, 1916) I showed that pyrogallolin weak solu- tions would produce light if oxidized by the plant peroxidases or by blood containing hemoglobin or by certain salts (KMnO:, K,Fe(CN)<;, and FeCl;) if H.O2 were present. These substances may be collectively spoken of as oxidizers. No light appeared in the absence of H.O, and no light appeared with pyrogallol and H,O, alone. All means of oxidizing pyrogallol with light production were also efficacious in blueing gum guaiac, but guaiac can be oxidized by many methods which fail to produce light with pyrogallol. The oxidation of gum guaiac was never observed to produce light nor was light ever observed during the oxidation by peroxidases of many other easily oxidizable hydroxy-phenyl and amino-phenyl compounds. No other per- oxide was found that could take the place of HO» in the luminous oxidation of pyrogallol by plant peroxidases. The peroxidase unquestionably acts by transferring oxygen from H.O» to the pyrogallol, and both H,O2 and peroxidase are used up in the process (2). It is not necessary that H,O2 : 311 312 Bioluminescence. VIII be decomposed with formation of bubbles of oxygen. KCN, acids, alkalies, and heat affect the light production in a characteristic way because they affect the peroxidase. It is consequently necessary, in order to gain a more accurate knowledge of the corditions under which the oxidation of pyrogallol leads to light production, to employ oxidizing substances of definite composition which are not affected by temperature, acid, ete. The present paper deals with the production of light by these inorganic substances. The first observation that pyrogallol would luminesce was made in con- nection with photography. Lenard and Wolf (6) noticed that a photo- graphic plate which was developed with a pyrogallol—Na»,SO;—K,CO; de- veloper glowed, if dipped in a saturated alum solution. They found also that the developer itself would light on mixing with alum solution, and thought that light was due to pyrogallol, which collected about the Al(OH); precipitated by K.CO; and was there oxidized. K»S,0; might replace the Na.SOs; of the developer (21). Trautz and Schorigin (8) also observed the luminescence of a photo- graphic plate as described by Lenard and Wolf and noted in addition that 30 per cent formaldehyde might take the place of the saturated alum solu- tion. In addition they noticed that a piece of raw leather, extracted with ligroin to remove fat, will luminesce if moistened with 30 per cent formal- dehyde and exposed to the air. Further work showed that both aldehydes and phenol derivatives luminesce when oxidized in many ways, as follows. Aldehydes. 1. With (a) neutral or (6) alkaline H2O.. (a) CHO (35 per cent) + HO: (3 to 30 per cent) = CO. + H:O + He. (b) 2CH.O + 2NaOH + H.O, = 2 HCOONa + 2 H.O + He. (Acetic, propionic, valeric, benzoic, and salicylic aldehydes and glucose give light under the same conditions.) 2. With Na.O» in ice water (formaldehyde, acetaldehyde, glucose, and vanillin). 3. With a warm saturated solution of NaClO; (formaldehyde). 4. With alkalies (simple aldehydes, sugars, alcohols, and many other substances). Phenols. 1. With warm neutral HO, or alkaline H.O2 at room temperatures (pyro- gallol, tannic, and gallic acids, benzkatechin, metol, eikonogen, and naph--. thol). 2. With alkalies (pyrogallol, tannic, and gallic acids and many others). According to Trautz and Schorigin, when 2 cc. of 10 per cent pyrogallol are mixed with 5 ce. of 10 per cent NazCOs, the solution glows at the surface: in contact with air; if some formaldehyde is added to this, the light lasts a long time, and if HO» is also added the light is much brighter. Trautz’ well known luminescent reaction with pyrogallol results when 35 ce. of 50 per cent K2COs;, 35 ce. of 10 per cent pyrogallol, 35 cc, of 35 per cent. ot) E. Newton Harvey 313 formaldehyde, and 50 cc. of 30 per cent H.O.2 are mixed. i -- rs see Pech Tier UCL eC PC) I aC wAhte Kee “(O(N O)a7*ND) AA woddor 3 sis tg : ae A190 ae = wately une e esas cue rrseersessssa(NTO)aTeuz) A OUulLy, a =e ait “ULES Sezsul e ehelle WO /enevarenere Wnaye Wane vralelielates (NO) O01 %20D) * SNOLLOT o) = = aa as SS ee ae eae ACE) at kof) +5 pvory its “MO]S e ; . eo AO A -+ “MOIS — patina SS Sena ey Shoe Sa ena c sh AS mrs CN) OPW) opruvAd0A1o} IOATIG 7, ee —_ : ‘sopRpId oa a +--+ — meee sss SURE SNOUTUUNT GoRIyxo (S2ppoorp1a DpoLvonN’T) AYOATST — - _ — Seah aA LS = (StUD YUU DINDYIO AJ) - uvozojoad op ++ meres Cuifopuabpry purpridhia) 55 poow.acyso " +--+ ++ — -- meres ss 38 (10QDYy DMDINULADY) JORLIXO pryynyeuued snourway 2 if OREO ON OER MONG! (-\/(0(6, 3 53 4+ - EE ee a ee TE Ce CE ELON On OC CUay HAS eh SOK KOI) ++ “TIBI Pan O80 0.01008 es BER eee Weve ee pane. Wiel ena *(DpDNIVYAL aworsa fy) Rs on = te ‘poory Peete eres pares tas sam hie SOUOdD last elaniD ‘A ay eae “Ue Serer e nts rrseere reams ety auererss (dg gpauell) JOeIjxe Pllouny ac Gg bE THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI. NO. 2 Vi Bioluminescence. 320 - We ay ig Ee eS a ty “quTegT Tou, iiss sec Abeer aan serene broke Jick ea mechs ‘yorlq unUye]g gee ws en = a —_ ose oo ONE Ciialessalie iene fa 3:20), @ialibi-env\reltoyalal an eirsits\rens Wala} sy'e,e) ss “(QIg) Y}1va [elosnyuy ze *MO|S ae AOA ae ens cae Mee a [lens elcaixe: Bie orale) ee eee space selene Vow (c/n) 6 ele) 856) Melis wee late) ©) \p) 6b, «a 6-4 0.560 *(9) yovyq ou0g Tg “MOJS A190 A =a as ‘mart By NR ROO (8 IO COCR OAON D5 Oa O08 ONCE YO 05920 att eCn *('O*qd) aprxo pee poy og 26 ~— 4 al Re Fs Wheyetetta ot 2)eicelcg in) ohn) ene cet srfettolierel anfertol «10. ee legecumte dopaue ay lailellejielnysNulvsie ielivie (Odd) aprxo ‘ 6Z +4 + - Lae A = Maiiieite ooh Shekel stp! sieenel evietesekomelieyey isin) ome io) 4/15) * hele .0uai s) Sma lokeDeaecereer *(®04d) aprxo1ed pay 8% ai *MOTS a 2 Be BUSHES crete) 2 fake collottej telco hie ne em= taal ta foMe.6) elieiisironel ene coyotes ‘(90°83 ) aprxo snoanoa0ur youl 1Z ns ae sae aes PE OG lsheixereceslce teneg en ctemepexeiaievinsiey'eiejse\le.e,'6 i= is) sce tsi eege (ouneenegs (03H) aprxo d1M9.19U poy 9% + + “qQUIe oT Bef) piled syscue) dcslies taeitel ms deiae lee elinteymilewelusive bie) aires Gua seinssi¥s(s 01> 91 ii 8] ie see smalen eae iene (O%8V) apixo I9ATIG GZ a8 tle es om | (SSIS ON ORO ORCS iC. CRON SC CE CHC CH OSC ORO NSC) OD DiCc (OND) “ "4 yori ¥S *MOTS ae fe Ad A ate Ot) NL Sasiten biel aurea caja, alisits jogo) co) pie (ace nites Suielareles oh stages op seh neens gs "(Q*ND) aprxo aioddoo pry 82 ae *MOTS ae : —" ileteeica tera) Ses ata cen ore) areisiclcolteis\seiislce) ate:eiege Tessbelis e\ene inns ‘@(HO) ND) aprxoipAy ioddoy GG me a = Ps a eis ae ce ee CD om tro tt (®Q%0,q) Oprxo olds poy 1Z Te =| = x. Steele taresann Soke: Mots » Ogi le seit ainsi (FQtaq) OpIxo oLUdF-snod19g7 Oz ate c= 1 is Pa eeliehisie cits age binin wieitx ie) selipiae: teria) sixaluaize) isiNigras site Mics: sah capris (®(HO) oq) aprxoipAy OLLI) 61 bE "MOIS = oo Se ei | louis Ge mul eke) sud lowte Twielioficlehevexelieehelisicslulnesisive 0) ¢e5 ee). sicuel anciab eis (QO?) aptxo Sno1loy yorvlg ST *MO|S eCIN = = — PN (Saran eee a a (aq) (uosoupAy Aq JudBBOI S yoOIOJY) [BJU UOT, LT =f + “qQUIB IT SET il feenssiatlexsyr sire) i8lros, stkasyeyieiteice)h.s) ols yal (oiie)) #¥ailatie Relies) o¥ ellie tte nsitslis\ionsas| Canale ieeed Caaua lent) csc sy dTTCIOIN QT a } "MOTS * = Peet iy lls teyisiig sii ie te jieice)seisol @Memelie) ace lain eli s¥ejelle is lelelohe neat (uy) (aopmod) [eqour osouvsuvyy CT te le 3 “quIe DF Wheira voit) a (sien bnparinte| ©, le sn aiiaijes(oi le (a)/alie oie) hobs !(| sCoMamotelia/ eXa\niin) 9010 (fQ*uyy) aprxo oTUvS UB, tI ar] ese Bea Sar Diguivea eee siolreaoxsel yey} J[By-ouo ore orNyXIUL [VU UT SUOTyeI}UBDUOD ‘ooUDY ‘[O[[esoIAd QT /W oar uoeleqryT | jo Sureny¢g go Scena YIM FYSUT ret tO#H Jud sad ¢g yavd [AO [oT[eBo1Ad YOT/W Javed T Jo oAN{xIUT OF popps oUINjOA [UNDG —— nt ‘papnqvUuog—l ATAVL E. Newton Harvey a2 1 solutions which blue guaiae readily (bromine water) give no light with pyrogallol + H2O2, and some substances which blue guaiac + HO: only slowly (potassium chrom alum and MnO.) give light with pyrogallol + H2O2. Generally the substances giving light with pyrogallol + HO: liberate visible bubbles of oxygen from H.0, but this is not necessary for light production, as indicated by the plant extracts or blood which give a good light after the. catalase has been destroyed. In my first paper I reported that blood containing hemocyanin gave no light but this is an. error due to the use of solutions too much diluted. Note also from Table I that extracts of luminous animals give no light with pyrogallol. Of special interest is the light produced by insoluble precipi- tates like MnO, and the colloidal metals Ag and Pt. Platinum black, with less surface, gives only a faint light. These exert a marked catalytic decomposition of H,Os and they are said to be wholly unchanged at the end of the reaction. Indeed light has been observed when HO: is dropped upon finely divided MnOz, Ag, or Pt (Dammer, 14) but with the concentrations of H,O. used in my experiments light only appears if pyrogalloi is also present. The Pt and Ag solutions were supplied by Dr. Goss of the Chem- istry Department, whose paper on light production by colloidal solutions appears on page 271. Potassium permanganate added to many organic substances is reduced to hydroxides, probably Mn(OH)3, and with certain organic substances, especially proteins, the Mn(OH); forms a colloidal solution. The colloidal Mn(OH); has a marked oxidizing action. This is unquestionably the explanation of Reed’s (16) supposed action of permanganate on peroxidases as indicated by the work of Bunzell and Hasselbring (15), a conclusion which I had already reached when their paper appeared. Reed supposed that the permanganate gave oxygen to the peroxidase converting it into a substance with greater oxidizing power, but the forma- tion of colloidal Mn(OH); explains his results satisfactorily. These colloidal Mn compounds will oxidize guaiac directly and give light with pyrogallol + HO, as will also a precipitate of Mn(OH)s. My results are recorded in Table I. Although gum guaiac is oxidized without H,O2, no light appears with pyrogallol unless H2O2 is present. An excess of HO. converts the brown colloidal Mn(OH); into colorless compounds which will then not blue guaiae. 322 Bioluminescence. VIII . . ° . { It is a curious fact that although the ferrocyanides give a good, light, the ferricyanides give none or at most a very faint light. Different samples of both ferrecyanide and ferricyanide solutions were found to vary in this regard and the variation proved to be: the result of exposure of the solutions to light and air. Ferro- eyanide particularly acquires the power to oxidize pyrogallol with bright light production after exposure to light and air. The phenomenon has nothing to do with phosphorescence (7.e., the giving out of light rays after exposure to light) but is no doubt the result of the formation of some substance in the ferrocyanide by photochemical action. It is presumed that this substance, whatever it is, has a marked power of transferring oxygen from HO, to the pyrogallol. Photochemical changes in ferrocyanides and ferricyanides are well known and very marked and, according to Haber (22), consist in the splitting of Fe(CN),.’’” to Fe’ -+ 6 CN’. Under the influence of light the iron separates and pre- cipitates as Fe(OH); or, if (NH4)25 be added, as black FeS. How- ever, other changes must also occur as neither iron salts nor Fe(OH) ; in colloidal or precipitate form can oxidize pyrogallol with anything like the light production obtained with K:Fe(CN)<« after exposure to sunlight and air. The following tabulation gives the results of my light experiments with K,Fe(CN), and K;Fe(CN)e. . Light production with m/100 Galersbarors pyrogallol + 3 per cent H2Os2. Solution, exposure to sunlight. Color Biter exposure to sunlight 1 week. : Soles cant Solution kept light. in the dark. M/20 K4Fe(CN), in | Light Deep yellow Bright. Negative air. yellow.| with brown- to very ish precipi- faint. tate. M/20 K4Fe(CN),. in 2 Light yellow. | Negative. sf hydrogen. M/20 K;Fe(CN), in | Yellow. | Dirty green Fair. We alr. with brown precipitate. M/20 K3Fe(CN), in Ze Dirty = green fs es hydrogen. with blue precipitate. E. Newton Harvey 320 2. Luminescence During Electrolysis ——All attempts to obtain light from the oxidation of pyrogallol by the nascent oxygen set free at the anode during electrolysis have failed despite the fact that various oxidizers were present. Bancroft and Weiser (17) have described light at the anode during electrolysis of certain salts. An especially bright orange light was observed during the electrolysis of NaBr with the mercury anode. The color is the same as that obtained when mercury vapor burns in an atmosphere of bromine to form HgBr and the light during electrolysis is no ‘doubt due to combination of the bromine set free by electrolysis with the mercury of the anode. A film of HgBr can be observed to form and under this film the mercury is supposed to be volatil- ized by the heating effect of the current. Light produc- Anode. Electrolyte. tion at anode or cathode. Platinized platinum. Half saturated K.SO, containing | Negative. M/20 pyrogallol. cs s Half saturated K,SO; containing s m/200 pyrogallol. Turnip juice + K.SO, containing M/200 pyrogallol. Pt with PbO: surface. Half saturated K,SO, containing vs m/200 pyrogallol. mae MnO; “ Half saturated K.SO, containing ae M/200 pyrogallol. Lead. Half saturated K.SO, containing M/200 pyrogallol. Platinum. MnSO, solution containing mM/200 ES pyrogallol. Reed (8) found that platinum black only blues guaiac if first exposed to nascent oxygen by making it the anode in some elec- trolyte. He used dilute HCl as the electrolyte, a fluid open to the objection that chlorine, whose powerful oxidizing action is so well known, is also liberated at the anode. This objection can be eliminated by using Na2SO, solution as electrolyte and I can confirm Reed’s result. A platinized surface made anode in a solution of NasSO, and then washed in water will blue gum guaiac directly but loses this power if made cathode in the same solution and then washed in water. Many other metallic sur- 324 Bioluminesecenee. VIII faces will blue guaiac if made anodes but none of these give light with pyrogallol if made anodes in electrolytes containing pyrogal- lol. The combinations in the table on page 323 were tried. The Pt cathode which becomes markedly alkaline was enclosed in a porous cup. The current varied up to a maximum of 2 amperes. In some experiments the Pt cathode instead of being enclosed in a porous cup was rotated rapidly but in no case did light appear. 3. Peroxides Giving Light with Pyrogallol and Oxidizers—The peroxidase of turnip Juice or KyFe(CN)¢ or any substance which gives light with pyrogallol and HO, may be.spoken of collectively as oxidizers. No substance of peroxide nature other than H2O2 will give light of any brilliancy or permanency with these oxi- dizers. Certain oxidizers will give a momentary light with such peroxides as NagQ2, BaOz, or perborates and persulfates but it is very faint and does not last as does the light with H,O.. The substances tried are listed in Table II. TABLE II. Peroxides Giving Light with Pyrogallol and Oxidizers. 3 |g b (sla oe B1¢ & eB, ie) x o-|o 3 BS 2 e5{3 2/81 3 | seiaea Oxidizer. (Equal parts added) < 2 5 a iB Eyal laren = > = 2 to a mixture of M/100 pyro- | 5 |Ha pe © S 2\5 a eS 2 |2 ¢ ] o |\cd = ac) aA = 3 [-) ~~ | =! gallol and the peroxide.) $0/5 Bs s} =| <4 3 so iS & lac ° z g |S £ Ke 6 a2)/4s|c a ° a a = ° ie >=} oRn sy & Noma ir) 2 a aq qd: = (sales; 6 = |Slal=s a | 8 |22 2 eS 5 a g g g es a m 33 mia lo Z Q = | A sa} Z leg Alaviserh oy MNCs socccad Go +) —| — = = =| = — = 1 per cent blood extract.| +| —| —| Faint = |= =) = a (acl ces flash om /20 K,Fe (CN). Oa =f | a a ail) el = a mil io m/100)KIMnO,.. 5... 2 =e =) = = = —| —| Faint! Fair | —| — flash.| flash. M/10 FeCl, he CEP OO CRP ONC + az = M/1OOK@r Os see: cere te Sie — = Na hypobromite........ +] —| —| Faint | Faint |\—| —| Fair | Fair | =| — : flash.| flash. flash.| flash. Ca hypochlorttes.<.222) +) =|, =|9e> — | <—| —| Paint S “Se MnO, vale veln, 6 pieie/ eka etna sts © + Mn(OH); sol in peptone} + _ = Colloidal Ag. 2242. - + } | E. Newton Harvey 325 4. Concentration of Pyrogallol and Oxidizer and Light Production. —Pyrogallol and H.O, and the oxidizer must naturally be present in a definite low concentration before light is visible. This con- centration is surprisingly low for pyrogallol—m/32,000 when potato juice is used as oxidizer (Harvey, 1) and m/512,000 with K,4Fe(CN).s. The latter dilution corresponds to 1 part of pyro- gallol in 5,000,000 parts of water. With increasing pyrogallol concentrations the intensity of the light increases up to a limit and then decreases. In the case of vegetable peroxidases this is no doubt due in part to precipitation of the peroxidase, as a visible coagulum is formed by strong pyrogallol in potato Juice. How- ° TABLE III. TABLE IV. Concentration of Pyrogallol and Concentration of K4sFe(CN). and Light Production with (after Light Production with (after Miz- Mixing) 0.75 per Cent H202 and ing) 0.75 per Cent H202 and m/400 M/40 K,Fe(CN )g. Pyrogallol. Character of light Toe ane ction Character of light from mixture. from mixture. Concentration of pyrogallol (after mixing). mixing). M/4 Negative. M/4* Negative to M/40 Faint. very faint. M/400 Good. M/8 Faint. m/4,000 Bright. M/16 Good. M/8,000 cs M/32 :: m/16,000 <3 M/64 Fair. m/32,000 Fair. M/128 < m/64,000 Faint. M/256 Faint. m/128,000 Ss M/512 Negative. m/256,000 Very faint. a m/512,000 Extremely faint. *If this solution is first ex- m/1,024,000 | Negative. posed to sunlight a good light —————————_—————— results. ever, this explanation does not hold for KyFe(CN).. In the ease of this oxidizer, we also find no light produced if too strong pyro- gallol is used. It is also necessary to employ an optimum con- centration of K,Fe(CN)., in order to produce light.’ Too great as well as too small a concentration of K,Fe(CN). fails to excite luminescence. H,O2 must be present in a certain low concentra- tion but there is apparently no upper limit in this case, at least to a concentration (after mixing) of 7.5 per cent. The actual concentrations necessary are given in Tables III and IV. An 326 Bioluminescence. VIII excess (at least ten times the amount necessary) of H2,O2. was present. What is the explanation of this absence of light with concen- trated pyrogallol or concentrated KyFe(CN).5 solutions? We know that oxidation of the pyrogallol goes on because the mix- ture turns dark. One immediately thinks of absorption of light by the concentrated solutions. It is obvious that no light will be visible unless the intensity of emission is greater than the ab- sorption. Both m pyrogallol and m/2 K,Fe(CN). are colored, the former a dirty light yellow, due to slight spontaneous oxida- tion, and the latter a clear light yellow. Neither of these solu- tions absorb light sufficiently to obscure the luminescence of mix- tures of M/200 pyrogallol and m/20 KyFe(CN).5 with H:Os, as we can very easily determine by viewing the luminescent mixture in a test-tube of 12 mm. diameter through at least a 40 mm. layer of m pyrogallol or mM/2 K,Fe(CN)>. Only when the lght of the luminescent mixture has markedly faded is the absorption noticeable. The oxidation products of pyrogallol, brown in color, do absorb light to a marked extent and the rapid fading of the luminescence of pyrogallol is no doubt largely due to the contin- ued formation of these products. Absorption by brown oxida- tion products, which are formed only slowly, could not explain the absence of immediate light production on mixing mM pyrogallol with HO, and m/20 KyFe(CN). or M/2 KyFe(CN).> with m/100 pyrogallol and H2Ov. The explanation must lie in some other direction. Perhaps the greatest amount of light is emitted with a definite optimum ve- locity of oxidation of pyrogallol. The velocity of oxidation of pyrogallol should increase with increasing concentration of pyro- gallol or of K,Fe(CN)s. We may suppose that with concentrated pyrogallol or concentrated KyFe(CN).> the velocity is above the optimum and although much heat is produced, none of the energy goes to light. If this supposition is correct we might expect to obtain light with concentrated pyrogallol or ferrocyanide by low- ering the temperature, since we might in this way lower the reac- tion velocity to the desired optimum. [Experiment has shown, however, that not lowering but on the contrary raising the temp- erature results in the greater production of light when either pyrogallol or ferrocyanide is present in high concentration. This EK. Newton Harvey BATT can be seen by a glance at Tables VI and VII. The phenomenon is akin to that observed during the oxidation and luminescence of phosphorus (23). White phosphorus will only luminesce in pres- ence of oxygen but if the oxygen pressure is too great, the luminescence ceases. At O°C. with water vapor present this “maximum luminescence pressure” of oxygen is 320 mm. Hg. and it increases 13.19 mm. Hg. for each degree rise in temperature. Thus there is a maximum and a minimum concentration of oxygen for the luminescence of phosphorus. In the case of pyro- gallol we have a maximum and a minimum concentration -of pyrogallol and also of the oxidizer for the luminescence of pyro- gallol. There appears to be no maximum for H.O,. The effect of temperature is to raise the maximum and is the same both as regards phosphorus and pyrogallol. The oxidation of phosphorus probably proceeds in several steps with the formation of intermediate oxides. If we assume that only one of these intermediate oxidations is connected with the production of light it is probable that a certain oxygen pres- sure and temperature will favor that reaction step at the expense of the others. This oxygen concentration and temperature will then correspond to the optimum for luminescence. OATIYVBON “JUTBT “QuIByy “Vale PG Fk ta por ke eae wine INM OL/W “qQUIBT “Vy sig “qyslig "poor PMH OS ea aed EGE TG ®OID OOT/W oO “IATLRBIN “BT “Quieyy “VR IT “IATL BBOIN €) sj 0/9 0) eh rere! oueMemanbiie re) 0 Bp) ths:.6 cd ele LOUD yy 0G/N™ = ‘ysey Jurey Fe i ‘, PNK Gi Reo CGC. YOUN OOL/ 2 “poor) "44 BlIg “4ysIIg ‘poor “OATPRION oWelo'.e: iesfo:se)ce1(5) 0 alley is)'elteltaiisy si etateraied (NO) 3M 0z/N if “IB 5 “IGT Pie nd (ek ee ae *qOV1}X9 pool Juso Jed T S “OATL VION “poor “poor ge a RM et CO. 2 1010 ‘gone drum, _ . z 0001-86 oGL 00S 00% 06-0 eS *IOZIPIXO AD *soin erode y, = q +) ‘ 4 tO*H Juay sad g + poypboshig OO, /W fo junowy yonby up yyun paxripy s] sazyprxg ay, “Uuoyonpodlg PYbVT pun ainjosad wa J, ‘A WIAVL fora) a N 329 E. Newton Harvey poor -yystaq ALO A Pp | “‘yysug “poor) “poor) “ATC “Jules AIDA “*)50Z 7B POPBINPES YPXIS-9UQ “YSBY JUTv AT “poor) “poor) “ATR “ATR AT “VUTB AT "OATYVION Safi a 2 “).0Z 7B PPPBANYCS J[VH 0001-86 ofh 00S o0f 00% oO oG-0 *(Surxrur 109jye) YYST] OF -agmnyerod ure], poesodxe 9(NO)9q'M Jo WOIyBIPUOIUOL) Qe quay lad g + 10))v60uig OOT/W yun paxryy st 9(NQ)2W7'M fo qunowy jonby uy ‘uoyonpotg 7YybvT pun ‘apunhoodiag fo W0VyDspUaIUO,) ‘gunqo.sad wa J, ‘ITA ATAVL “OATPVION “YSBp Arey “Ysey FySUIg “4yslig “yystag “Vystig “qystig -000‘'F/W "YSty WSL ” ” poe Wolers) eel “Fare OOF/N poe) “Wyse gles) Furey “que AUN SHt ” ‘OF/W *QUIBT “IB “QUIT “yurey Alo A “OATIBION “OAT}BION ‘OATYBBON © | “p/W 2001-86 | oS 208 208 202 201 oc-0 Sine JojjesorAd jo uol} *soinjesoduray, ~e1} 99000) Qe quay wad g + Jop)vboshg yyun paxipy ST 9(NO)29Ad'M 06/N fo yunowy ywonbg up ‘woyonpolg 1y5vT pun oyvbowkg fo WorDn.1jwaauoy) ‘gunjo.Lad wa J, ‘TA GTIAVL 330 Biolumineseenee. VIII intensity of the light will be increased and the duration of the light decreased with rise in temperature. Although no numerical data can be given, my experiments do indicate that with rise of temperature, for each oxidizer (Table V), and each concentration of oxidizer (Table VII), or each concentration of pyrogallol (Table VI) the duration of the light is shorter and its intensity is greater up to a certain definite temperature. Above this temperature the intensity as well as the duration of the light decreases. It is perfectly plain why potato juice gives no light at high tem- peratures, because its oxidizer is destroyed between 80° and 85°; but why should the light be fainter or absent at 98° in the case of KMnOx, or MnO, or any of the oxidizers of Table V which are not affected by a temperature of 98°C.? In harmony with the explanation already suggested we must suppose that it is a defi- nite (optimum) temperature and not the highest temperature which favors that particular step in the oxidation of pyrogallol involving the production of light. At the temperature indicated in the lower right hand square of Table VI, particularly, the oxida- tion proceeds in such a way that no light is produced. 6. Light Production in Absence of Peroxide-—When pyrogallol is oxidized by peroxidase and hydrogen peroxide, purpurogallin, a substance of doubtful composition, is said to be formed. Its formula is probably CywH sO; (4) or CuHs0; (5). This substance can be recognized qualitatively by its dark color. As purpuro- gallin is fairly insoluble in cold water, Bach and Chodat (2) deter- mined it by collection on a weighed filter, drying at 110°, and again weighing. Reed (3) made determinations by filtering, washing free of unoxidized pyrogallol, and titrating by 0.05M KMn0O,j. Purpurogallin or a similar substance is also formed when pyro- gallol is oxidized by many oxidizing agents; viz., by AgNOs, KMn0,, or NaNO, in acid solution, by quinone or platinum black or K;Fe(CN)., or by allowing a solution of pyrogallol and gum arabic to stand. Perkin (5) reports a very favorable yield by the electrolytic oxidation of pyrogallol in 15 per cent NasSO, with lead cathode and rotating platinum anode. I have observed no light during the action of these substances in moderate con- centrations (7.¢., in concentrations comparable to those previously described as giving light in the presence of H.O2) although the oo | E. Newton Harvey dol pyrogallol turned brown. The AgNO; was reduced to metallic Ag. The following combinations were tried. 1 cc. 2 per cent AgNO; + 1 cc. N/10 HNO; mixed with 2 ec. m/100 pyro- gallol. 1 ec. M/10 NaNO, + 1 ce. n/10 HNO; mixed with 2 ec. m/100 pyrogallol. Darcie Ore “mixed with 2 ec. m/10 pyrogallol. 1“ “ K;Fe(CN),+ 1 cc. n/10 HNO; mixed with 2 ec. m/100 pyro- gallol. Gum arabic solution and m/2 pyrogallol. The gum arabic and pyrogallol mixtures become brown slowly (during the course of 2 or 3 days) and needle crystals, sometimes eurved and forming rosettes, of a golden brown substance, sepa- rate. No light was observed to be given off during this process. The only substances investigated capable of giving light with weak pyrogallol alone are sodium hypochlorite and hypobromite (Table I), ozone, and acid permanganate, which gave light in the following combination. 1 ce. M/100 KMnO, + 1 ce. nN to N/10 HCI or HNO; mixed with 2 cc. m/100 pyrogallol. Ozonized oxygen (by silent electric discharge) bubbled through m/100 pyrogallol. The light with acid permanganate is at most a momentary flash. Neutral or alkaline permanganate gave no light with pyrogallol alone. Fahrig (19) had observed that ozone shaken with certain samples of water would luminesce but was unable to determine the cause of it. Otto (20) found that water carefully purified of or- ganic matter gave no light with ozone but that benzene (feeble light), alcohol, thiophene, milk, and urine did. I find that m/100 and m/10 pyrogallol, m/10 orcinol, m/10 resorcinol, m/100 escu- lin, fresh or boiled potato juice, and certain other fluids give light if ozone is bubbled through them. Each bubble, as it rises through the liquid is aglow. As already mentioned, if the pyro- gallol is too concentrated (m solution) no light appears. Pure oxygen gives no light if bubbled through the above solutions. Pyrogallol also turns brown if an alkaline solution is exposed to air. COs, acetic acid, and brown substances are said to be formed. Oxygen is absorbed, the amount depending on the amount of alkali present up to a certain concentration. Weyl 239 Biolumineseence. VIII and Goth (7) found that most oxygen was absorbed when 0.25 gm. of pyrogallol was present in 10 ec. of NaOH of specific grav- ity 1.03. Less pyrogallol or a greater or less amount of NaOH resulted in the absorption of less oxygen. Trautz and Schorigin (8) found that a mixture of 2 cc. of 10 per cent pyrogallol + 5 ce. of 10 per cent Na2CO3 luminesced at the surface but I have been unable to confirm this experiment or to observe light at room tem- perature during the darkening of pyrogallol + alkali under any conditions. Alkali of all concentrations from solid NaOH or NasCO; to N/6,250 NaOH has been added to various concentra- tions of pyrogallol but light has never been: observed. It is per- haps not surprising to observe no light in the more concentrated solutions, which turn black almost instantly, because of absorp- tion, but even in weak solutions (N/10 to n/6,250 NaOH + m/100 pyrogallol) no light has ever been observed even though H»O» also be added. Such solutions become only brown in color, the more deeply brown the greater the concentration of the alkali. The absence of light during the absorption of oxygen in alka- line pyrogallol is possibly due to the rapidity of the oxidation. The oxidation of the pyrogallol by K,Fe(CN). can be made to proceed so rapidly or in such a way that no light appears (Table | VI, the 98° column); on the other hand it may be that the oxida- tion product formed with alkali is different from that formed with H.O, and an oxidizer or that one step in the oxidation series is omitted. So little is known of the chemistry of purpurogallin that we cannot decide the question at the present time. It is apparent, however, that the conditions for light production are quite definite. The oxidation must proceed in a particular way. 7. Light Production in Non-Aqueous Solvents. Alcohol.—Light production occurs in fairly strong alcohol as indicated by the mixtures in Table VIII. The turnip juice plus alcohol will give light with pyrogallol so long as the peroxidase is not precipitated but if precipitation oc- curs, then only a very faint light is produced. Strong alcohol (95 per cent), however, does not prevent light production so long as an oxidizer not affected by the alcohol, such as MnOz, is em- ployed. Even hemoglobin powder (Eimer and Amend) gives a faint light with pyrogallol plus He.O. if suspended in absolute alcohol (No. 6 of Table VIII). K. Newton Harvey TABLE VIII. Light Production in Alcohol. Equal Parts of Solution A and Solution B Are Mixed. 339 Solution A. 1 ce. of 3 per cent H2Oz2 in 90 per Solution B. cent alcohol, plus: 4 No. Light production. hag In alcohol. Substance. In alcohol. per cent per cent 1 | m/100 | Absolute. | Precipitate formed by 90 Very faint.* adding 90 per cent alcohol to turnip juice. 2 | m/200 50 Same with 50 per cent 50 Faint. * alcohol. 3 | M/200 50 Filtrate from above 50 Negative. precipitation. 4| m/200 40 Precipitate formed by 40 Bright. adding 40 per cent alcohol to turnip juice. 5 | m/200 40 Filtrate from above 40 Sa precipitation. 6 | m/100 | Absolute. | Hemoglobin powder. | Absolute. | Faint. 7 | m/200 50 A se 50 Bright. 8 | m/100 | Absolute. | Powdered MnOsz. Absolute. | Fair. 9 | m/100 ee es H.Mn0Os. se Good. 10 | m/100 « M/10 FeCls. vg Negative. “11 | m/200 50 Mmy/Z0) 50 = * Light appears only some time after mixing. Acetone —Hemoglobin or MnO. powder both give light with 50 per cent acetone containing M/200 pyrogallol and 1.5 per cent H.02. The light from the hemoglobin is very bright. If benzoyl peroxide replaces the H2O2 no light appears. Ether, Chloroform, and Benzene.—Pyrogallol is slightly soluble in benzene, fairly soluble in chloroform, and easily soluble in ether. If pyrogallol (m/100 or less) in ether, chloroform, or ben- zene be shaken with an equal volume of 3 per cent H2O in water, the ether, chloroform, and benzene layer then removed, and added to hemoglobin or MnO, powder, light only appears with ether and hemoglobin, and it is very faint. Penzoyl peroxide in place of H2O2, although soluble in all three solvents, will give no light “Im any case. 334 Bioluminesecence. VIII 8. Effect of Acid and Alkali on Light Production.—The influence of acid and alkali on light production can best be studied by using as an oxidizer some inert substance such as MnO: not changed by acid or alkali in dilute concentration. Table IX gives the results. The MnO, powder was added to the HCl and NaOH in the con- centrations indicated in the table and mixed with an equal volume of 1 ee. of 3 per cent HO. + 1 ce. M/100 pyrogallol. The con- centration of acid and alkali in the final mixture is therefore one- half of that designated in the table. It will be observed that a small concentration of acid (v/160) inhibits the ight, whereas a definite concentration of alkali (w/40) increases the light to an optimum. TABLE IX. Effect df HCl and NaOH on Light Production by MnOx. ; Concen een of Light production. Concent ace of Light production. n/10 Negative. n/10 Fair. n/20 - n/20 Bright. n/40 fe n/40 Fair. n/SO Very faint. n/80 Faint. n/160 Faint. : n/160 ef Water. se Water. fe SUMMARY. 1. The literature on chemiluminescence of pyrogallol is reviewed and a classification of all known types of chemiluminescent reac- tions given. A light-producing system consisting of an oxidizer, a peroxide, and an oxidizable substance (pyrogallol) is considered in detail. 2. In addition to substances already recorded, chromates, bi- chromates, hypochlorites, hypobromites, hypoiodites, chromium and iron salts, colloidal silver, platinum, and oxides of manganese, bloods containing hemocyanin or chlorocruorin, and _precipi- tates of Fe.Fe(CN)<, Mn(OH)2 #5 Mn(OH)s, Mn0., Mn.Os, AgsO, and metallic silver and platinum black all give light with pyro- gallol and H.O.. Many other substances recorded gave no light. 3. Ferrocyanide solutions only give a bright light with pyrogal- lol + H.O: if first exposed to sunlight and air; ferricyanides give E. Newton Harvey 300 a faint light only, if exposed to sunlight in presence or absence of alr. 4. Perborates, persulfates, Na2O2, and BaOs, can take the place of H,O2 with some oxidizers but not with all. 5. No light has ever been observed during the liberation of nascent electrolytic oxygen at anodes of platinum black, manga- nese dioxide, or lead peroxide, in electrolytes containing pyro- gallol. 6. Pyrogallol gives light in absence of peroxide, only with sodium * hypochlorite and hypobromite, acid (but not neutral or alkaline) permanganate, and ozone. No light has been observed in pres- ence of alkali (with or without H,O.) although rapid oxidation takes place. 7. There is a maximum, minimum, and optimum concentra- tion of pyrogallol and also of potassium ferrocyanide for light pro- duction. Above the maximum and below the minimum no light appears. No maximum was found for H2Os, up to a concentra- tion of 7.5 per cent. 8. Increase of temperature increases the brightness of the light for each concentration of pyrogallol and of potassium ferrocya- nide up to an optimum; then a decrease occurs. The greater the concentration of pyrogallol or of ferrocyanide, the higher the tem- perature necessary to give light. Increase of temperature de- creases the duration of the light. '9. Pyrogallol can be oxidized with light production in fairly strong alcohol, acetone, and ether. 10. Acid prevents and alkali favors the luminescent oxidation of pyrogallol by MnO, + H2Osz. | BIBLIOGRAPHY. 1. Harvey, E. N., Am. J. Phystol., 1916-17, xlii, 318; 1916, xli, 454. 2. Bach, A., and Chodat, R., Ber. chem. Ges., 1904, xxxvii, 1342. 3. Reed, G. B., Bot. Gaz., 1916, xii, 53. 4. Thorpe, E., A Dictionary of Applied Chemistry, New York, 1913, iv, 421. 5. Perkin, A. G., and Perkin, F. M., J. Chem. Soc., 1904, Ixxxv, 243. 6. Lenard, P., and Wolf, M., Ann. Phys. u. Chem., 1888, xxxiv, 918. 7. Weyl, T., and Goth, A., Ber. chem. Ges., 1881, xiv, 2659. 8. Trautz, M., and Schorigin, P., Z. wiss. Photographie, Photophystk, u. Photochem., 1905, iii, 121. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 336 9. 10. Te 12. 13. 14. 15. 16. Lite 18. 19. 20. 21. 99 23. Bioluminesecence. VIII Trautz, M., Z. physik. Chem., 1905, liii, 1. Radziszewski, B., Ber. chem. Ges., 1877, x, 70, 321. suinchant, Compt. rend. Acad., 1905, exl, 1170. Dubois, R., Compt. rend. Acad., 1901, exxxii, 431. McDermott, F. A., J. Am. Chem. Soc., 1913, xxxv, 824. Dammer, O., Handbuch der anorganischen Chemie, Leipsic, 1906, i, 431. Bunzell, H. H., and Hasselbring, H., Bot. Gaz., 1917, |xiii, 225. Reed, Bot. Gaz., 1916, |xii, 233. Bancroft, W. D., and Weiser, H. B., J. Physic. Chem., 1914, xvii, 762. Weitlaner, F., Verhandl. zool.-bot. Ges., 1911, xli, 192. Fahrig, E., Chem. News, 1890, 1xii, 39. Otto, M., Compt. rend. Acad., 1896, exxiii, 1005. Lenard, P., and Wolf, M., Jahrb. Photographie, 1891, 289. Haber, F., Z. Electrochem., 1905, x1, 847. Abegg, R., and Auerbach, F., Handbuch der anorganischen Chemie, Leipsic, 1907, 11, pt. 3, 376. ; : ’ THE STRUCTURE OF THE PURINE MONONUCLEOTIDES. By WALTER JONES anp B. E. READ. (From the Laboratory of Physiological Chemistry, Johns Hopkins Medical School, Baltimore.) (Received for publication, June 12, 1917.) A careful examination of the literature on the structure of plant nucleic acid will show that excellent evidence has been fur- nished for the following conclusions. 1. Nucleic acid is a tetranucleotide; that is to say, it is com- posed of the groups of four mononucleotides. 2. The four mononucleotide groups of yeast nucleic acid are joined to one another through their carbohydrate groups, giving rise to a polysaccharide structure. The order of the mononu- cleotide groups in nucleic acid has been determined. These two deductions are based principally upon the prepara- tion and properties of various simpler nucleotides from yeast nu- -cleic acid,! and upon a comparative study of the rate at which phosphoric acid is liberated from them by mild acid hydrolysis.” 3. Each mononucleotide is composed of three groups, because by complete acid hydrolysis each produces three substances, ribose, phosphoric acid, and a nitrogenous compound. This has been found in the case of one of the four mononucleotides (guanine mononucleotide) by direct experiment with the substance itself but its extension to the other three mononucleotides, although indi- rect, is practically conclusive. 1 Guanine-cytosine dinucleotide, Jones, W., and Richards, A. E., J. Biol. Chem., 1915, xx, 25. Adenine-uracil dinucleotide, Jones, W., and Read, B. E., ibid., 1917, >o,0 bg MULTE Uracil-cytosine dinucleotide, Jones and Read, ibid., 1917, xxxi, 39. Guanine mononucleotide, Read, ibid., 1917, xxxi, 47. * Jones and Read, J. Biol. Chem., 1917, xxix, 123. 337 338 ; Purine Mononucleotides 4. The preparation of four nucleosides from plant nucleic acid proves that in each of the four mononucleotides the nitrogenous group is combined to a ribose group.® These statements would deal completely with the gross structure of nucleic acid as represented in the formula below, if the posi- tions of the phosphoric acid groups in the four mononucleotides were definitely known. ee HOW O=SP — 0.C;H,02.C;H.Ns HO O O=P = O . C;H,O 5 C.H3N.20>. ; HO” | O HO. | O—F =< O . C;H,O 5 C.H.sN;0 HO | ; 0 : HO | O=2 — O Gi C;H7O2 . C;H4N;O \. HO” Yeast nucleic acid. We propose to show the position of the phosphoric acid group in each of the purine mononucleotides. The three possible arrangements of the three groups in guan- ine mononucleotide are indicated in the following formulas. O its HO — P—O.C;H;0;.CsHiN;O on O II, C;H,0O,.O — P.C;H,N;O on 3 Guanosine, Levene, P. A., and Jacobs, W. A., Ber. chem. Ges., 1909, xlii, 2474; adenosine, 2703; cytidine, ibid., 1910, xliii, 3150. Uridine, Levene, P. A., and La Forge, F. B., Ber. chem. Ges., 1912, xlv, 608. Read, B. E., and Tottingham, W. E., J. Biol..Chem., 1917, xxxi, 295, W. Jones and B. E. Read 339 O | I. C;H,0,. C;H,N,O. P — OH | OH Guanine mononucleotide forms guanosine, so that its guanine and ribose groups must be directly combined. Moreover, the nucleotide is a dibasic acid that forms a crystalline dibrucine salt. Formula II is therefore excluded. A crucial experiment described below decides between Formu- las I and III. Guanine mononucleotide liberates guanine with much greater rapidity than it, liberates phosphoric acid. For- mula III is therefore excluded and Formula I alone remains. Similar reasoning would show that adenine mononucleotide has an analogous structure. But adenine mononucleotide has never been prepared. We have found, however, that adenine- uracil dinucleotide, which can be easily obtained, liberates its adenine far more rapidly than it liberates its phosphoric acid. As the two phosphoric acid groups of the dinucleotide are not di- rectly jomed to one another the results obtained may be safely applied to hypothetical adenine mononucleotide. Guanine Mononucleotide. Pure guanine mononucleotide was prepared from yeast nucleic ~ acid and allowed to dry in a desiccator with sulfuric acid. The product still contained about 7 per cent of moisture but its state of dryness and even its condition of purity are not essential to the experiments described, as we shall deal with ratios. Six portions from the same specimen of the nucleotide were weighed into small flasks provided with condensing tubes and each portion was treated with twenty parts of 5 per cent sulfuric acid. The flasks were then submerged in a boiling water bath and heated for various periods from 5 minutes to 3 hours. Each hydrolyzed product was made alkaline with ammonia while still hot, for the precipitation of free guanine, and after standing over night the guanine was filtered off, allowed to dry, and weighed. The filtrates from guanine were heated to the boiling point and treated with magnesia mixture for the precipitation of free phosphorie acid. After standing over night the crystalline 340 Purine Mononucleotides precipitates of magnesium ammonium phosphate were filtered off, allowed to dry, and weighed. In Table I the quantities of the constituents determined are all caleulated for 1 gm. of mononucleotide and the maximum amount obtained in any experiment is assumed to be 100 per cent. From this amount the corresponding percentages for the other experiments are reckoned. The difference in the rates at which guanine and phosphoric acid are liberated from the nucleotide is so great that it can be attributed neither to imperfections of method nor to analytical error. TABLE I. Magnesium ammonium G F phosphate. ame: Musee sae of used. 1ydrolysis. Berea P ce Obtained. nucleo Per cont | Obtained. nucleon eens 0.4234 5 min. | 0.0499 | 0.1178 18.8 | 0.0835 | 0.1972 50.0 0.5030 Seen 0.1348 | 0.2680 42.8 | 0.1747 | 0.3473 88.1 0.5000 SON ees 0.2004 | 0.4008 64.0 | 0.1901 | 0.3802 96.3 0.3487 ihr 0.1881 | 0.5395 86.2 | 0.1396 | 0.4002 101.5 0.3360 Die 0.2017 | 0.6000 95.8 | 0.1310 | 0.3899 98.9 0.3448 By 0.2158 | 0.6259 100.0 | 0.1360 | 0.3944 | 100.0 The results given in Table I are represented diagrammatically in Fig. 1. Its method of construction is so obvious as to require no explanation. lal GUANINE a | d | see Oe / ! | “Phosphorie ace: || | | | | : | I . / oe | | aoe | | i | | | | ' | | . | | | | | . | Time. of hydrolysis | a |i Cee ie i W. Jones and B. E. Read oat Adenine-Uracil Dinucleotide. Portions of the dinucleotide were heated with dilute sulfuric acid for various periods of time. As guanine was not to be taken into consideration the fluids were made alkaline with ammonia and the free phosphoric acid was determined by precipitating directly with magnesia mixture as described above. In a second series of experiments with the same specimen of dinucleotide, the hydrolyzed product was made alkaline with am- . monia and the adenine was precipitated with an ammoniacal solu- tion of silver nitrate. The gelatinous precipitates of silver-ade- nine were washed until the wash water contained no trace of either free or combined ammonia, when the filter papers with their precipitates were placed in Kjeldahl flasks and the nitrogen was determined. For each experiment 10 cc. of sulfuric acid, 5 gm. of potassium sulfate, and a few drops of aqueous copper sulfate were used, Adenine “ : . PhosP | | vA | | | / | | | | ha | | | | eee od Time of hyd rolysis| | | Hires 2: The results are given in Tables II and III and are represented diagrammatically in Fig. 2. The methods of tabulation and con- struction are the same as those used above for guanine mononu- cleotide with one exception. In the case of the dinucleotide a correction has been made for the small amount of phosphoric acid which is liberated from. the pyrimidine nucleotide group. Repeated experiment has shown that this corresponds to 10 mg. of magnesium ammonium phosphate per gm.-hour.! 4 Jones, W., J. Biol. Chem., 1916, xxiv, p. iii. Jones and Read, ibid., 1ON7,) xxixe e233 342 Purine Mononucleotides TABLE II. Magnesium ammonium phosphate. Nucleotide Time of used. hydrolysis. 4 Bee. - : _ ; Obtained 4 cricleotide,| earsotiod.|| muclootiden | cataha 0.7665 5 min 0.0726 0.095 0.001 0.094 25.0 0.9712 Is 0.1638 0.169 0.002 0.167 44.9 0.9754 30)“ 0.2330 0.239 0.005 0.234 62.9 0.8832 1 hr 0.2815 0.319 0.010 0.309 83.1 0.8258 2 hrs. 0.3192 0 387 0.020 0.367 98.7 0.9500 By & 0.3816 0.402 0.030 OnS72 100.0 0.9976 Aes 0.4067 0.408 | 0.040 | 0.368 98.9 TABLE III. Adenine nitrogen. Nucleotide Time of used. hydrolysis. \ Obtained. paces of total 0.5245 1 min. 0.0313 0.0597 69.1 0.5592 iy Ae 0.03938 0.0703 81.4 0.6016 S0b 0.05138 0.0853 98.8 0.5009 1 hr. 0.0436 0.0870 100.8 0.5756 Drees 0.0496 0.0861 99.7 0.5338 3 0.0461 0.0864 100.0 THE SIMILARITY OF THE ACTION OF SALTS UPON THE SWELLING OF ANIMAL MEMBRANES AND OF POWDERED COLLOIDS. By JACQUES LOEB. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, June 25, 1917.) L. 1. When dry animal membranes, like pig’s bladder freed from fat by prolonged treatment with hot ether (a week or 10 days in a Soxhlet apparatus), are put into water, they will swell; they will also swell and even slightly more when put into a salt solu- tion; but when the membrane is put first for about > hour into M/8 solution of NaCl and subsequently into distilled water, the swelling will be twice or three times as great as in either of the — other two cases. A further investigation of this problem sug- gested a possible connection with the structure of the membrane. An investigation of this phenomenon seemed desirable in view _ of the fact that Flusin' has tried to connect the phenomena of osmosis with the imbibition of the separating membrane. A series of pieces of dry pig’s bladder, weighing about 0.4 gm. each, were put into 50 ec. of M/8 NaCl, and left there for half an hour. One was put as a control into distilled water. After a half an hour each piece was freed from adhering solution by putting it between two sheets of filter paper and pressing gently; it was then weighed. The pieces which had been in M/8 NaCl were then put into NaCl solutions of different concentrations, namely, M/8, M/32, M/128, m/512, m/1024, m/4096, m/8192 NaCl, and into H.O, and the control in H,O was again put into H,O. After different intervals the membranes were taken out of the solutions, freed from liquid adhering to their surface, and weighed. Table I gives the results of one such experiment. 1 Flusin, Ann. Chim. et Phys., 1908, series 8, xiii, 480. 343 344 Salts and Animal Membranes TABLE I. Swelling of dried pig’s bladder in per cent of original weight of piece of bladder. After z In NaCl eae Control ; yr H20 M/8 M/8 m8 M/8 M/8 M/8 M/8 M/8 his. 1 187 186 190 191 184 185 173 183 149 Then transferred into NaCl M/8 M/32 M/128 | mM/512 | m/1024 | m/4096 | m/8192 | HsO H20 13 204 219 289 329 310 328 301 354 165 164 219 234 363 409 408 455 438 500 187 24 240 242 | 396 464 451 510 542 562 208 These figures show the following facts: A dried membrane kept permanently in H.O increases in weight 208 per cent in 24 hours, and when kept permanently in M/8 NaCl it increases in 24 hours slightly more, namely, 240 per cent. This was to be expected on the basis of the old experiments of Hofmeister. What the writer. did not expect is the fact that if a piece of membrane is put first for 30 minutes into M/8 NaCl and subsequently for 24 hours into a weaker NaCl solution or distilled water, it swells the more the lower the concentration. Thus a membrane kept for 30 minutes in M/8 NaCl increased its weight when put sub- sequently into H.O 562 per cent within 24 hours. The fact that a membrane kept permanently in M/8 NaCl swells so much less (240 per cent) must then be due to the fact that presence of the NaCl solution counteracts the swelling. We must, therefore, dis- criminate between two effects of the salt solution upon the swell- ing of a membrane, namely, first, a chemical reaction between the membrane and the salt, which would cause in itself a considerable swelling if it were not inhibited by the second effect of the solu- tion which counteracts this tendency to swell, and the more so the higher the concentration of the salt solution. We can elimi- nate this second factor by allowing the membrane to react with. the salt first and then transferring the membrane to H:O after having washed off the adherent salt solution. In a recent series of papers the writer? has pointed out the different behavior of 2 Loeb, J., J. Biol. Chem., 1916, xxvii, 339, 353, 363; 1916-17, xxviii, 175. Jacques Loeb 345 the membrane of the eggs of Fundulus when taken directly from a salt solution or when taken from distilled water or a solution of a non-electrolyte. -2. When an attempt was made to repeat the experiment just described on pig’s bladder with solid blocks of gelatine, either dry or containing varying quantities of water, they failed completely. Gelatine shows none of the peculiarities mentioned in Table I. A 40 per cent gelatine solution was prepared and after it had set, small pieces were cut out of the jelly and exposed to air for 21 hours during which time they lost about 50 per cent in weight. Then the same experiment as represented in Table I was repeated. Table II gives the result. TABLE II. Increase in weight of blocks of gelatine in per cent of original weight. After In NaCl ee Control H:O M/8 M/8 M/8 M/8 M/8 M/8 M/8 M/8 Ars. 4 23 24 ote 27 25 25 30 24 19.5 Then transferred into NaCl M/8 M/ 32 M/128 | m/512 | m/1024 | m/4096 | m/8192| H2O HO ; 13 51 52 56 54 51 52 58 50 44 19 198 193 196 190 176 176 158 170 | 162 24 246 236 239 231 214 216 208 214 | 188 The swelling in M/8 NaCl was slightly greater than in H.O, but the gelatine first treated with m/8 NaCl and then put into HO for 24 hours did not swell more than the piece kept perma- nently in M/8 NaCl. Thin sheets of gelatine behaved like blocks of the same material. It was thought that the salt had perhaps not entered sufficiently _ into the gelatine. To avoid this possibility the gelatine was dis- _ solved in M/8 NaCl instead of in H.O and after setting the pieces were cut out of the gelatine. When put into NaCl solutions varying from m/8 to M/8192 or into H.O the result was identical with the one expressed in Table II. Nor were the results differ- ent when the gelatine used was completely dried beforehand. 346 Salts and Animal Membranes 3. It was found that finely powdered gelatine behaved exactly like pig’s bladder and behaved differently from solid blocks of gelatine. This was true when the powdered gelatine used was the same as that used for making the solid blocks mentioned in Table II. Commercial Cooper’s powdered gelatine was put through a No. 60 sieve and again through a No. 80 sieve. The grains going through the former but not through the latter sieve served for the experiment. 2 gm. of such gelatine were put into a cylin- drical funnel, the bottom of which was covered with a round piece of filter paper. The upper surface of the powdered gelatine in the funnel was also covered with a round piece of filter paper in order to make it possible to pour water or salt solution on the gelatine without stirring up the particles too much. When 25 ce. of distilled water are poured on the gelatine, part of the water runs quickly through but part of the water is retained and the mass of gelatine swells. When the process is repeated, only a slight further swelling takes place, and after this no further swell- ing takes place no matter how much water filters through the gelatine. When instead of letting water run through the gelatine we let 25 ec. of M/8 NaCl run through, and repeat the process of filtermg m/8 NaCl through the gelatine, the latter swells also and even a trifle more than in the H,O experiment, but the maxi- mal swelling is also soon obtained. If, however, we allow first 25 ec. of M/8 NaCl to run through a mass of powdered gelatine and follow this with consecutive washings by distilled water, the mass will swell considerably more with each consecutive washing than in either of the other cases. The following experiment is the analogue of the one presented in Table I. Eight cylindrical funnels, each containing 2 gm. of powdered Cooper’s gelatine (size of grain between sieves No. 60 and No. 80) were prepared in the way described above, and 25 ce. of M/8 NaCl were percolated through each funnel. The mass of gelatine increased in each cylindrical funnel about 18 to 20 mm. in height. After this 100 cc. of a different solution were sent through each of the eight funnels and the additional swelling was ascertained in each mass at the end of the experiment. These solutions were: M/8, M/32, M/128, m/512, m/1024, m/4096, > m/8192 NaCl, and H.O. The results are. given in Table III. ee « F * Jaeques Loeb 347 TABLE III. Swelling of powdered gelatine in mm. height of a cylinder containing the powder. After In NaCl Contro}t He M/8 M/8 M/8 M/8 M/8 1/8 M/8 | M/8 = | a, (Cisse eee 19.5 20 20 19 17 18 18.5 20 | 15 Then 100 ec. of the following solutions were allowed to run through. NaCl H:O H20 M/8 M/32 M/128 | M/512 | m/1024 | m/4096 | M/8192 Additional swelling....| 8(?) 5 5 12 PA 31 40 64 14 The increase in swelling is expressed in terms of the height of the cylindrical mass of gelatine. It is obvious that when powdered gelatine is treated first with M/8 NaCl and subsequently with H.O it retains much more water than when it is perfused permanently with H,O or per- manently with M/8 NaCl. This can also be explained on the as- sumption that the phenomenon is due to two different effects, first the reaction between salt and gelatine which increases the sweil- ing, and second the inhibition of the swelling if the gelatine is washed in the salt solution. This inhibition reminds us of the inhibition of acid swelling of gelatine by the presence of salts, though the mechanism may be different in the two cases. From these experiments it ts obvious that pig’s bladder and pow- dered gelatine have a peculiarity in common which we were not able to discover in solid blocks of gelatine, namely, to swell considerably more when a short treatment with M/S NaCl is followed by a treat- ment with H.O, than when the mass is treated exclusively with H.O or exclusively with salt. The property which powdered gelatine and pig’s bladder have in common is that they consist of very small discrete particles, grains in the one and fibers in the other; while _ the gelatine in a block must be considered as a homogeneous mass or as one enormous particle. It should also be stated that this pe- culiar behavior of powdered gelatine is probably found in many powdered colloids; thus powdered ovomucoid kindly given to us by Dr. Lopez-Sudrez behaved exactly like powdered gelatine. - 348 Salts and Animal Membranes To make the demonstration complete we should add that the after effect of a previous salt treatment just described is found also if other concentrations of NaCl than m/8 are used, namely, m/4, M/2, M/1, ete., but that the after effect ceases when the NaCl concentration is too low, namely, below m/64 or m/128 NaCl. It should also be Said that if we leave powdered colloids or pig’s bladder permanently in a salt solution nothing comparable to this after effect is noticed even if the NaCl solution is very weak. In such cases the maximum swelling is soon reached. Table IV gives the results of such an experiment with powdered gelatine, which covers also the case for pig’s bladder. We have TABLE IV. Swelling of powdered gelatine in mm. height of a cylinder. NaCl | = = re H20 s|slelsfele/e/a/ S|) e/é Se SP Se ee ey | ES = = 3 s 3 After first 25 cc. of | BOLUGION 0 si01 es 23 \ 221221 DA22 120) 21) 2 20 )32205|e2il 19 Additional swell- ing after further 25 ec. of solution.| 11) 8] 6) 4) 4) 2) 2.5) 2 2 2, 2) 2 Additional swell- } ing after third 25 ec. of solution....| 0] 0} 0} O}| O| 0] 0 0 0 0 0 0 mentioned the fact that if only m/8 NaCl solutions are allowed to percolate through powdered gelatine the maximum swelling is soon reached and no further swelling takes place. This is true for all concentrations of NaCl tried. We only notice a slightly increased effect with the increase of the concentration. ihe The Effect of Different Ions. 1. The excessive swelling, observed in pig’s bladder, in pow-_ dered gelatine, or in ovomucoid when a short treatment with NaCl solution of not too low a concentration is followed by a treatment with distilled water, is not confined to NaCl, but is produced by = Jacques Loeb 349 many if not all neutral salts with a univalent cation; while the neutral salts with bivalent cations have no such effect. Pieces of pig’s bladder were put for 30 minutes into M/8 LiCl, NaCl, KCl, MgCh, CaCk, and SrCl.. Table V gives the result. TABLE V. Increase in weight of pig’s bladder in per cent of original weight in M/8 solutions of: After if ReD | NaCl | KCU | MeCh|, “Cake |Z sre |) Conte! hrs. 4 181 193 186 225 164 162 139 They were then transferred to distilled water. 5 306 335 325 169 138 149 145 4 592 547 454 189 138 161 151 225 662 598 434 195 134 165 145 Leaving aside minor differences, it is obvious that the previous treatment of the membrane with Li, Na, and K causes a consider- able increase in swelling, while this increase is lacking in the case of Mg, Ca, and Sr. The salts of the alkali earth simply prevent the subsequent increase in the swelling caused by NaCl or LiCl. Mg is less active than Ca or Sr, which was to be expected. Jt cannot be said that CaCl, or SrCly affect the swelling in the opposite way from that caused by NaCl, since the effect of CaCl, does not differ much from that of distilled water. All salts of Na cause the after effect though the quantity of swelling varies with different anions (Table Vi). TABLE VI. Swelling of pig’s bladder in per cent of original weight in solutions of different sodium salts. After = Te M/8 M/8 M/8 M/8 M/16 bly & ad m/l _| Control NaCl | NaNOs | NaCH:COO | NazSO« | NaeSO. | Naz tar) Na tar | “6 hrs. 5 188 209 157 195 222 L72 201 167 They were then transferred to distilled water. 3 325 312 290 345 464 336 418 189 5 638 464 505 644. 846 644 872 221 223 758 840 490 659 832 1,071 1,400 Das 350 Salts and Animal Membranes The experiment shows clearly that a short treatment of the membrane with any sodium salt causes a considerable further swell- ing in H,O after the free salt solution is leached out. One fact, however, stands out, namely, that the striking difference between the action of univalent and bivalent. cations is in no way repeated among the anions. This is in harmony with the writer’s first extensive experiments on antagonistic salt action on Fundulus in which he showed that the toxic action of high concentrations of salts with univalent cations could be inhibited by small quantities of salts with a bivalent cation, while no such valency effect could be found for the anions.* 2. The same difference in the effeet of the ante with univalent and bivalent cations upon the subsequent swelling of pig’s bladder in distilled water can be found in powdered gelatine. 2 gm. of powdered gelatine (of grain size 60-80) were put into each of a series of cylindrical funnels, and 25 ce. of M/8 LiCl, NaCl, KCl, MegCk, CaCh, SrCh, and BaCl were kept in contact with the gelatine for half an hour and then allowed to run off. Four times in succession 25 ec. of distilled water were then allowed to run through each cylinder. Table VII gives the swelling in mm. height of the cylinder. TABLE VII. Swelling in mm. height of cylinder of powdered gelatine under the influence of different chlorides. M/8 M/8 M/8 M/8 M/8 M/8 M/8_ | Control LiCl | NaCl KCI | MgCl | CaCk | .SrCh | BaClk | HO Swelling under in- fluence of 25 cc. salt solution...... 23 25 24 24 24 | 24.5) 24 | 24 Additional swelling under influence of 100 ce. HO (four washings).........| 15 33 25 4.5 4 3.5) | 2D eR The amount of swelling in the different salt solutions was too small to be discovered. In the after effect, however, the striking difference between the salts with univalent and bivalent cations 3 Loeb, Arch. ges. Physiol., 1901, \xxxviii, 68; Am. J. Physiol., 1901-02, vi, 411. Jacques Loeb Bi | (which was also noticeable in the case of the swelling of pig’s membrane) shows itself. There is practically no difference in the after effect of CaCl, upon the swelling of the membrane in distilled water and the swelling caused in distilled water without any previous salt treatment. 3. No difference of this kind in the after effect of a treatment with NaCl and CaCl could be discovered in solid blocks or sheets of gelatine. ie Antagonistic Salt Action. 1. In 1901% the writer showed that the injurious effects which a salt with univalent cation has upon the eggs of Fundulus, as soon as the concentration of the salt exceeds a certain concentration, can be inhibited by the addition of a very small quantity of a salt with a bivalent cation; while the addition of a salt with a bivalent anion had no such effect. The writer drew from these facts the conclusion that the antagonistic action of the bivalent cation in this case must be due to an action of the salts upon the state of colloids, but for a long time it was found difficult to imi- tate such an antagonistic salt action directly on colloids. Re- . cently analogies have been shown to exist by Schryver* on gels of cholate solutions, by Clowes for soaps,® by Lenk® for the swell- ing of gelatine, and by Fenn’ for the precipitation of dissolved gelatine by alcohol in the presence of different salts. A very striking antagonistic salt action can be demonstrated in the case of the after effect of NaCl upon the subsequent swell- ing of pig’s bladder and powdered colloids in distilled water. This swelling is inhibited if a comparatively small quantity of 4Schryver, S. B., Proc. Roy. Soc., Series B, 1914, lxxxvii, 366; 1916, Ixxxix, 176. Schryver, S. B., and Hewlett, N., ibid., 1916, lxxxix, 361. 5 Clowes, G. H. A., J. Phys. Chem., 1916, xx, 407; Science, 1916, xliii, 750. 6 Lenk, E., Biochem. Z., 1916, lxxiii, 15, 58. 7 Fenn, W. O., Proc. Nat. Acad. Sc., 1916, ii, 534, 539. In one of the ex- amples mentioned by Fenn, e.g., the precipitation of gelatine in mixtures of solutions of Na; citrate and CaCl, by alcohol, the precipitation depends to a large extent upon the formation of a supersaturated solution of cal- cium citrate which is precipitated by alcohol even if no gelatine is present. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 352 Salts and Animal Membranes CaCl. (or any other salt with bivalent cations) is added to the NaCl. | The experiment consisted in the following. Pieces of dry pig’s bladder were put for 30 minutes into the following solutions: 50 ec. M/8 NaCl 50:“ m/4 * 50) “aaa 1 cc. M/8 AG: + 49 ec. HO 2 “ m/8 + 48 <“ ns 50 “ M/4 “ 4 “ M/8 “c + 46 “ce “c 50 “ce M/4 “ 8 “ M/S “ce + 42 “ce “ 50 “ mM/4 “ 4+ 16 “ m/8 yy ate Ba 8 oe * 50 “ec M/4 “cc + 2D. “ M/8 “ce + 18 “ce “ As the reader will see, each of these solutions was M/8 in regard to NaCl but contained increasing concentrations of CaCl. After the membranes had been in the solutions for 830 minutes they were transferred to distilled water. It will be seen that the membrane previously treated with pure NaCl swelled considerably while the ones treated previously with NaCl + CaCl, swelled the less the more CaCl. was added. eee Vil. Antagonism between NaCl and CaCl: on the subsequent swelling of pig’s bladder when put into distilled water. Increase in weight in per cent of original weight. ; After Ce. M/8 CaCl in 100 ec. M/8 NaCl Coatral H20 0 1 2 4 8 16 32 hrs. 4 189 173 182 185 157 159 164 143 | The membranes were then all transferred to distilled water 3 336 305 339 280 155 155 160 21 345 345 (430?)} 269 137 154 167 145 It is obvious that the addition of 4 ee. of M/8 CaCl to 100 ce. of m/8 NaCl already inhibits markedly the subsequent swelling of the membrane in disiilled water and that the addition of 8 ec. of M/8 CaCl to 100 cc. of M/8 NaCl inhibits this swelling almost completely. If the reader will look over the first row of figures representing the swelling of the membranes while zn the salt solutions the an- _ Jacques Loeb By) tagonistic salt action will be seen to be extremely small if it exists at all. It should also be pointed out once more that the antagonistic action between NaCl and CaCl: is not the algebraic mean between a swelling effect of NaCl and a dehydrating effect of CaCh. Table V has already contradicted such an assumption. In membranes not treated previously with any salt the swelling in H2O amounted in Table VIII to 145 per cent, while in membranes which had been treated for 30 minutes with 100 cc. of M/8 NaCl + 32 ce. of m/8 CaCl, and were then transferred to H,O it amounted to 167 per cent. The addition of CaCl, served only to prevent the enormous increase in weight which a previous treatment with M/8 NaCl alone induces, namely, in this experiment 345 per cent, but it cannot be said that CaCl. and NaCl affect the membrane in an opposite sense. 2. The antagonism experiment just described is successful also with powdered gelatine. 2 gm. of powdered gelatine (size of grain 60-80) were put into a series of funnels (as described in the beginning of this paper) and first 25 ec. of m/8 NaCl with different quantities of CaClk (as described in the experiment with pig’s bladder) were allowed to run through the gelatine. Subsequently 75 ce. of distilled water were allowed to run through each cylinder. The result was striking, Inasmuch as in the salt solution the swelling was approximately the same in pure NaCl and in NaCl with CaCl; but in the subsequent treat- ment with H.O the swelling was enormous in the powdered gela- tine treated previously with pure m/8 NaCl, while this after ef- fect was prevented when 8 cc. or more of CaClh were added to 100 ce. M/8 NaCl (Table IX). F Mg and Sr act similarly to CaCle. The same antagonistic effect was observed with powdered ovomucoid. 3. The writer has not been able to find such an antagonistic action in the after effect of a previous treatment of solid blocks of gelatine with mixtures of NaCl and CaCh. The experiment in Table X was performed with solid pieces of dry gelatine. While in the salt solution the swelling was greater in both M/8 NaCl and m/8 CaCl, than in pure distilled water, in the subse- quent treatment with distilled water no after effect of the na- 354 Salts and Animal Membranes ture of that found in powdered gelatine or in pig’s bladder could be discovered. This difference in behavior between powdered gelatine and gelatine in a solid block cannot be ascribed to a difference in the chemical nature of the gelatine used in both cases since the results TABLE IX. Swelling of powdered gelatine in mm. height of a cylinder. * Ce. m/8 CaCl in 100 ec. m/8 NaCl Control ; H20 oe | eae ie Fe After 25 ec. salt solution had been BOC CUE eee ee omen tee 22 QO 235 22a 23a tei 19 Then 75 ce. distilled water were allowed to filter through. Additional swelling after the 75 ec. distilled water had perco- lated)cr ore Meer gree ce cep 30 | 37 | 29 | 20 | 3.5] 3.5) 5 3 TABLE X. Increase in weight of solid blocks of dry gelatine in per cent of the original weight of the solution. After Ce. m/8 CaCl in 100 ec. M/8 NaCl Controls. 100 ec. M/8 CaCls H20 1 101 |. 108 | . 99 | 1065], 113 |, <99-| “aia 96 The pieces of gelatine were then transferred to distilled water. i 3 246 265 243 264 | 265] 241 267 257 23 483 550 469 514 | 524 500 527 538 were the same when the blocks of gelatine were made by dissolving Cooper’s powdered gelatine in warm water and allowing the so- lution to set. The results were also the same when the solid blocks of gelatine were only half dry at the beginning of the ex- periment, or when they were exposed to the salt solution longer or less than 30 minutes. ‘ Jacques Loeb 350 LV. The Influence of the Size of the Particles of Powdered Gelatine wpon the Salt Action. The fact that the after effect of a treatment with neutral salts with monovalent cation is the same for pig’s bladder and for pow- dered gelatine, while it is different for gelatine in the form of one solid block, suggests that we are dealing with a surface effect of the salt. In order to test this suggestion powdered gelatine of four different sizes of particles was prepared, by sifting Cooper’s powdered gelatine through sieves with different openings, namely, with 50, 60, 80, 100, and 120 wires per inch. The first lot con- tained particles which went through 50, but not through 60 (des- ignated 50-60); the next those going through Sieve 60 but not through No. 80 (designated 60-80), and so on. 2 gm. of each of these particles were put into a funnel as described and first 25 cc. of M/8 NaCl were poured on each lot. In order to insure equal action of the salt the solution was kept in contact with the gelatine for } hour before it was allowed to filter. The swelling was measured and then 25 cc. of distilled water were-poured on the gelatine and the H.O was kept in contact with the gelatine for 20 minutes before it was allowed to filter through, and the swell- ing was measured again. This was followed by again pouring 25 cc. of distilled water into each funnel, keeping the H:O in con- tact with the gelatine for 45 minutes, then allowing it to run through, and then again measuring the swelling. Table XI gives the result. TABLE XI. Swelling of powdered gelatine of different grain size. Swelling measured in mm. height of the cylindrical mass of gela- tine. Size of particles. 50-60 | 60-80 | 80-100 | 100-120 Swelling after 25 cc. m/8 NaCl had been al- lowedtonpercolatenms 2... Sse cece eens 23 25 | 24 23 Additional swelling after first 25 ec. H,O had been allowed to percolate. . rete 3 6 8.5] 10 Additional swelling after the peoad 25 ¢ ce He Oo had been allowed to. percolate............... 19 2h | ead 38 356 Salts and Animal Membranes It is obvious that the smaller the particles the greater the re- tention of water and the greater the swelling. This would be expected if the after effect of the salt upon the swelling in dis- tilled water were a surface effect; for the total surface of a given mass of gelatine is of course the greater the smaller the size of the particles. It should be possible to calculate the increase in surface with the decrease of the size of the powdered granules if the phenom- enon were not complicated by a second variable which acts in the opposite sense as the size of the granules and which the writer believes to be the process of packing. The ‘‘packing”’ diminishes the free area of the particles. The excessive swelling of pig’s bladder under the influence of a previous treatment of NaCl is therefore a phenomenon which can be repeated in powdered gelatine but not in solid blocks of gela- tine, and hence must. be due to a difference in the structure of the two groups of systems, the gelatine block being homogeneous with only a comparatively small outer surface while the pig’s bladder and the mass of powdered gelatine or ovomucoid consist of small discrete elements with an enormous internal surface. The writer is inclined to believe that the salts combine with the gelatine and as a consequence modify the chemical affinity of the surface of the discrete particles for water. The result is a greater retention of water after the free salt solution has been replaced by distilled water. The gelatine or ovomucoid salts with univalent cations retain the water in such cases very power- fully, while the colloidal Ca salts do not possess this peculiarity. The mechanism of the swelling of powdered colloids or animal membranes like pig’s bladder in distilled water, after a previous treatment with a neutral salt with univalent cation, is different from the mechanism of swelling of a solid block of gelatine under the influence of acid or alkali. The latter case has been explained very elegantly by Procter.® 8 Procter, H. R., J. Chem. Soc., 1914, ev, 318. Procter, H. R., and Wil- son, J. A., ibid., 1916, cix, 307. ae ay Jacques Loeb aot V. The Antagonistic Salt Action on the Percolation of Water through Powdered Colloids. In 1905 the writer? suggested that the antagonism between salts with monovalent and bivalent cations was due to the fact that small quantities of bivalent cations prevented the diffusion of the salts with univalent cations through the animal membranes. This idea has since been generally accepted and has received sup- port by the work of many experimenters, especially the brilliant experiments of Osterhout!® on Laminaria. It seemed, therefore, of interest to see in which sense the salts with univalent and bi- valent cations influence the rate of diffusion or percolation of water through powdered colloids. We must keep in mind that the antagonistic salt effects de- seribed in this paper differ in an essential point from the antago- nistic salt effects on the living organism. The observations on the antagonistic action of salts on living organisms were all made while the living object was 7n the salt solution; and the only excep- tions from this rule are the observations on washed eggs.” The antagonistic effects described in this paper deal with the behavior of colloids after the salt solution has been replaced by distilled water and after all the free salt solution has been washed away. It is very easy to examine the influence of a salt treatment upon the rate of percolation of liquids through powdered colloids. 2 gm. of powdered gelatine or ovomucoid are put into a cylindrical funnel in the way described at the beginning and 25 ce. of a salt solution carefully poured on top of the mass. This solution runs through very rapidly as long as the particles are not too small. This is followed by pouring 25 cc. of distilled water upon the mass and this is then repeated. The rate of percolation becomes slower with each washing (due possibly to a denser packing of the particles) and after two or three washings with distilled water a definite effect of the previous salt treatment upon the rate of percolation can be discovered. This effect is exactly the reverse of the influence of the salts upon the subsequent swelling in ® Loeb, Arch. ges. Physiol., 1905, evii, 252. ‘© Osterhout, W. J. V., Plant World, 1913, xvi, 129; Proc. Am. Phil. Soc., 1916, lv, 533. DFQ 358 Salts and Animal Membranes distilled water. Thus a treatment of powdered gelatine with m/8 NaCl inereases the rate of swelling of the mass in distilled water (after the salt solution is washed off) but it diminishes the rate of percolation of distilled water through the mass. CaCl: neither favors swelling nor does it retard the rate of percolation; it may accelerate it slightly. It is thus easy to demonstrate an antagonistic salt action upon the rate of percolation of water through powdered gelatine. The same solutions as in Table VIII were used in the following experi- ment. 1 gm. of powdered gelatine was put into each funnel and at first the various mixtures of NaCl + CaCle solutions (25 ee. in each case) were allowed to run through. This was followed by repeated washings with 25 ce. of distilled water. Table XII gives the cc. of water which percolated from the funnels into a measur- ing cylinder after 25 ce. of H,O had been poured on the mass for the third time (third washing). TABLE XII. Rate of percolation of 25 cc. of distilled water through powdered gelatine after a previous treatment with the following solutions and two washings with 25 ce. H2O Ce.-of H2O0 which percolated in Ce. m/8 CaCle in 100 ce. M/8 NaCl H20 0 1 2 4 8 16 32 459 MING chicas hs 9 | 10.4 | 13.6 | 18-2 | 19.1 | 18.4 | 21.4) 22.4 UZ rss seen Sots Silo a 2 26* | 24 30* | 26.5 |:25.5)) 276 * Some of the water retained in previous washings had filtered through. The rate of percolation is slowest in the gelatine previously treated with m/8 NaCl; and here the swelling is greatest. This is natural since the swelling as well as the lowering of the rate of percolation have the same cause, namely, the retention of water by the powdered gelatine. The same oe can be made with powdered ovomucoid. The writer has not yet tried any ee on the influence of a previous salt treatment on the diffusion of water through a membrane of pig’s bladder, though he intends to do so. The experiments mentioned here bear a certain resemblance to the observations of soil chemists on the percolation of water through soil previously treated with salts. The writer’s attention Jacques Loeb 309 was called to this work by Professor Lipman in Berkeley, in whose laboratory the subject has recently been investigated by Mr. Sharp. It seems that many years ago A. Meyer first ob- served the fact that if soil had been soaked with certain salts it became impermeable for water, after the salt had been leached out. Schlésing and Van Bemmelen showed that the phenomenon was connected with a greater degree of suspensibility of the soil after such a treatment.’ Soil treated with m/8 NaCl becomes almost impermeable for water after the salt solution is washed out. No measurable swelling of the soil follows when the soil is first treated with M/8 NaCl, then with distilled water until all the salt solu- tion is driven out, and the soil becomes highly impermeable. The impermeability in this case is much greater than in the case of powdered gelatine or ovomucoid and the suspicion is justified that the impermeability of the soil after a treatment with NaCl is at least partly due to a denser packing of the particles. This variable may also be at least partly responsible for the retarda- tion of percolation of water through powdered gelatine after a previous treatment with NaCl. CaCl. does not retard the subsequent percolation of water through soil and it is easy to demonstrate the antagonism be- tween NaCl and CaCl: upon the subsequent rate of percolation of water through soil after the salt is leached out. 10 gm. of finely powdered garden soil were put into each of a ‘series of cylindrical funnels. Then 25 ce. of the antagonistic salt mixtures were poured on the soil and the time measured until 20 cc. of the solution had diffused into a measuring cylinder put un- der the funnel. Then 25 cc. of H,O were poured into each funnel and again the time for 20 ¢c. of liquid to run through the soil was measured, and this was repeated three times. Table XIII gives the results. It may be possible to make practical use of this action of Ca (which seems to be the same for all bivalent cations) for rendering impermeable soil permeable for water. The writer does not wish to enter into the cause of this behavior of soil beyond mentioning that if all organic matter of the soil is 1 Sharp, L. T., Proc. Nat. Acad. Sc., 1915, i, 563; Univ. Cal. Publ. in Agricult.. Se., 1916, 1,291. 2 Van Bemmelen, J. M., J. prakt. Chem., 1881, xxiii, 388 360 Salts and Animal Membranes TABLE XIII. Time in minutes for 20 ce. of liquid to run through 10 gm. of soil in a cylindrical funnel. Ce. m/8 CaCl in 100 ce. M/8 NaCl. Con- trol 0 1 2 4 s | 16 | 32 | Hs 25 ec. salt 34 37.5 35 DORON ese | oa 34 43 2a suds Oma (ee 66 50 46.5} 40 | 34 33.5) 44 25es So tah ets She ae S5, 1,185 | 1,185 66.5} 48 68.5 * T.e., over night. destroyed by ignition a treatment of such soil with NaCl will no longer call forth the striking inhibition of the percolation of HO after the salt is leached out; but that an addition of some finely powdered organic colloid (powdered dry oak leaves, gum tra- gacanth, powdered gelatine, or ovomucoid) can restore to some extent this effect of a previous washing with NaCl. A mixture of finely powdered marble and powdered colloids acts like a mix- ture of ignited soil and organic colloids. SUMMARY OF RESULTS. 1. Dried pig’s bladder, freed from fat, when treated for a short time with a solution of a salt with univalent cation swells consid- erably more when subsequently put into distilled water, than it does if it remains permanently in the same salt solution or when it remains permanently in distilled water without a previous salt treatment. 2. It is assumed that this increased swelling of the membrane in distilled water after a previous treatment with one of the salts with univalent cation is due to an interaction between the salt and a constituent (probably protein) of the membrane; when the bladder remains permanently in the salt solution the latter pre- vents the swelling which takes place as soon as the salt solution is replaced by H.O or a very weak salt solution. 3. A treatment of the membrane with salts with a bivalent cation (Mg, Ca, Sr, and Ba) does not induce the excessive swell- ing when the membrane is subsequently exposed to distilled water. Neither does such a treatment induce a dehydration of si Ce tes aie. TO in Jacques Loeb 361 the membrane. Membranes previously treated with salts with a bivalent cation swell when afterwards put into distilled water approximately to the same extent as membranes that have not been treated with any salt. 4. The addition of about 8 ce. of M/8 CaCl. to 100 ce. of m/8 NaC prevents the after effect which a treatment with a pure M/8 NaCl solution produces. It should be noticed that CaCls does not influence swelling in the opposite sense from that of NaCl, but that it renders the after effect of the treatment with NaCl impossible in some other way. 5. It is impossible to repeat these effects of a previous salt treatment upon the subsequent swelling in distilled water with solid blocks of gelatine, or with sheets of gelatine. 6. It is, however, possible to repeat them with powdered gela- tine or with powdered water-insoluble ovomucoid (and probably a large number of other powdered colloids). 7. The fact that pig’s bladder behaves in regard to these phe- nomena like powdered colloids but not like solid blocks or sheets of gelatine suggests that the salt effects described in this paper are due to an action upon the surface of colloidal particles (fibers in the case of pig’s bladder). 8. This suggestion is supported by the fact that the effect of a previous treatment with m/8 NaCl upon the subsequent swelling _ of a given mass of powdered gelatine in distilled water is greater when the size of the particles is smaller and hence the total internal surface greater. 9. It follows from all this that the mechanism of the swelling described in this paper is of a different nature from that observed in solid masses of gelatine under the influence of acid or alkal. 10. Observations upon the rate of percolation of water show that the effect of salt upon the subsequent rate of percolation of distilled water through the powdered gelatine varies inversely with the rate of swelling. A previous treatment with M/S NaCl solution retards the percolation of water through the powdered gelatine, while a previous treatment of the mass with M/8 CaCl, has no such effect. The addition of a small quantity of CaCh to NaCl prevents the subsequent retardation of the rate of per- colation of water as it prevents the swelling. 11. It has been known that a treatment of soil with NaCl 362 Salts and Animal Membranes renders the soil almost impermeable to water after the salt is leached out. In this case, however, no swelling of the soil seems — to take place and the writer is not certain whether the influence of a salt treatment upon the percolation of water through pow- dered gelatine and ovomucoid is identical with or only analogous to that upon the percolation of water through soil. THE METABOLISM OF SULFUR. II. THE INFLUENCE OF SMALL AMOUNTS OF CYSTINE ON THE BALANCE OF NITROGEN IN DOGS MAINTAINED ON A LOW PROTEIN DIET.* By HOWARD B. LEWIS. (From the Laboratories of Physiological Chemistry of the University o Pennsylvania, Philadelphia, and of the University of Illinois, Urbana.) (Received for publication, June 13, 1917.) Recent studies of the metabolism of protein with the aid of the more accurate methods of analysis, developed largely by Folin and by Van Slyke and their coworkers, have shown that amino- acids are absorbed as such from the alimentary tract, and circu- late in the blood or are stored in the tissues. This has led to a realization of the fact that the réle of the proteins in nutrition ts a ‘function of their component amino-acids, and that the adequacy or inadequacy of any individual protein is dependent upon the quantitative relationships of the amino-acids of its molecule. Attention accordingly has been centered on the problem of the ability of the organism to synthesize or dispense with the various individual amino-acids. Certain proteins, such as gelatin, have long been known to be lacking in specific amino-acids. Other proteins not completely lacking in any essential amino-acid may yet contain some of these amino-acids in the molecule in such small amounts as to render a high protein intake necessary to supply a sufficient quantity of the requisite amino-acids. This point of view has been clearly defined by Osborne and Mendel (1) in their studies on the growth requirements of the white rat. These observers have shown that growth on diets contain- ing different proteins in approximately the same percentage relation may vary considerably, provided the percentage of protein in the diet is near the minimum requirement for these animals. Ona diet containing a higher * I wish to express my indebtedness to Professor A. E. Taylor of the University of Pennsylvania for courtesies which have facilitated the con- tinuance of this research at the University of Illinois. 363 364 Metabolism of Sulfur. II percentage of protein, these variations in nutritive efficiency are less marked or disappear altogether. They have also for certain proteins dem- onstrated the specific amino-acids the quantity of which determines the minimum for that protein. Thus rats failed to grow at a normal rate on diets containing less than 15 per cent of casein, while the addition of cystine to food containing 9 per cent of casein without further change at once ren- dered the diet decidedly more adequate for growth. Similar results were obtained with edestin to which lysine was added, and later (2) with lysine and gliadin. Recently by similar methods of study Hogan (3) has shown that for the proteins of the corn kernel, tryptophane is the first limiting factor and lysine the second. Of the amino-acids, tyrosine (or phenylalanine (4) ), tryptophane, ly- sine, and cystine are generally recognized as essential amino-acids which must be present preformed in adequate amounts in the diet. Ackroyd and Hopkins (5) have recently reported experiments which indicate that either histidine or arginine but not both must be present in the diet for normal nutrition. The claim of the indispensability of cystine preformed in the diet is based largely on indirect evidence or analogy, rather than on direct experimental evidence. With the exception of the work of Osborne and Mendel, in which the addition of cystine to casein, a protein notably low in cystine, was demonstrated to lower the percentage of casein required for normal growth, little experimental work on this point is available. The high cystine content of the proteins of the epithelial tissue, the constant loss of this sulfur-rich protein through the hair, skin, ete., especially in - the lower mammals, and the inability to use inorganic sulfates for protein synthesis and growth have led to a belief in the indispensability of pre- formed cystine in the diet. Practical feeding experiments with swine (6) have also furnished evidence of a possible réle of cystine. The offspring of sows fed during pregnancy with blood albumin (a protein rich in sulfur and in which a large part of the sulfur is presumably present as cystine) were larger, with heavier and darker coats of hair than the controls. Me- Collum and Davis (7) have suggested that the loss of sulfur from casein when heated may be associated with the decreased efficiency of such casein as a foodstuff. Holt in a recent address (8) hag made the suggesticn that the success attained in infant nutrition with whole milk from cows may be due not to the higher percentage of protein as such in whole milk, but to the fact that by thus increasing the percentage of protein in the diet, the infant’s actual amino-acid needs for growth, especially in lysine and cys- tine, may be more nearly satisfied. Abderhalden (9) attempted without success to remove the cystine from completely hydrolyzed protein by pre- cipitation with glacial acetic acid in order to determine the‘‘biologische Werltigkeit’”’ of cystine by feeding the resultant product to animals. In acceptance of this belief in the indipensability of cystine, many workers have added cystine to the products of hydrolysis of proteins before feeding, to replace the cystine destroyed in the course of the hydrolysis (Abderhal- den (4) and Totani (4), Geiling (5) ). igi. Hens. 365 The experiments reported in the present communication rep- resent an attempt to approach the problem of the cystine (and sulfur) requirement of the organism of the dog from the standpoint of the protein minimum, making use of variations in the nitrogen balances as indications of any change in the adequacy of the diet. The general plan of the experiments has been as follows. Dogs were maintained on standard diets of low protein content but of ample calorific value. After control periods on these diets, small amounts of cystine (0.5 to 1.0 gm. of cystine daily) were added to the standard diets, and the influ- ence of the cystine was noted as evidenced by changes in the nitrogen bal- ances. In certain experiments as additional controls, nitrogen in the form of glycocoll, a dispensable amino-acid, and of tyrosine, an essential amino- acid, was added to the standard diet in amount equivalent to that added in the form of cystine. The animals used were females, and were accus- tomed to laboratory conditions, one of the animals (Dog A) having been in the writer’s possession for over 3 years, the subject of frequent nutri- tion experiments. The animals were kept in the usual metabolism cages and the urine was separated into 24 hour periods by daily catheterization. There was no evidence of cystitis at any time. Separation of the feces into periods was accomplished by the use of carmine. In order to give consist- ency and bulk to the feces, neutral calcium phosphate (Merck’s ‘Blue - Label’) was added to the diet, the pure calcium phosphate being chosen rather than the usual bone ash in order to avoid as far as possible the pres- ence of sulfur in the diet other than the sulfur of the protein. The source of the protein of the diet was beef heart, trimmed from the adjacent fat, finely ground, and carefully mixed to ensure uniformity. In one experiment (Dog G), it was necessary to substitute for the beef heart, finely chopped steak from which connective tissue and fat had been removed as far as pos- sible before grinding. The cane sugar, lard, and starch were pure com- mercial preparations. The purity of the cystine, which was prepared from wool by acid hydrolysis according to Folin, was established by analyses for nitrogen and sulfur. Nitrogen was determined by the Kjeldahl-Gunning method. For the determination of sulfur in the food a weighed sample was evaporated to dryness in a porcelain evaporating dish with concentrated nitric acid on the water bath, the dried residue treated with Benedict’s copper nitrate- potassium chlorate oxidation mixture, evaporated, and treated as in the method for urine. The standard daily diets are shown in Table I. In the experiments with Dog A, Series I, and Dogs C and G, the cystine was added without altera- tion of the diet. In the experiments with Dog A, Series II, and Dog B, in order to keep the nitrogen intake constant, when the amino-acids were fed, an equivalent amount of nitrogen in the form of beef heart was removed from the diet. The starch was first made into a smooth paste with water, the other ingredients were added, and the whole was thoroughly mixed. 366 Metabolism of Sulfur. I] TABLBD I. Standard Diets. Dog A. Dog B.| Dog €.| Dog G. Series | Series Suchosejagm: facie Sahai eee eee 100 50 90 90 70 Starchwugnite: Sis cnesse eee eo eee 40 30 20 40 40 Tair Sg niee:astes So qsfsertaee 8 ae ear See 50 40) 50 50 50 Calcium phosphate, gi... 2.6.3 ete 5 5 5 5 10 Wiatersce kg. gm. gm. gm. gm. gm. 1 13.54) 1.27 | 0.211) 1.62]) 0.28 |— 0.638 ; 2, 13.47) 1.27 | 0.21N 1.57) 0.28 |— 0.57) Standard diet, 40 I 3 13.47) 1.27 | 0.211) 1.82) 0.28 |— 0.83 gm. beef heart 4 13.39] 1.27.) 0.211) 1.62) 0.28 |— 0.63 and 0.5 gm. cys- 5 | 13.34] 1.27 | 0.211] 1.59] 0.28 |— 0.60] tine. | 6 13-37) 12.27 (O22), 2407) 102282 |—. £08 AO Uae eae 7.62 | 1.266} 10.29) 1.68 |— 4.34 IAVOTEBE 5... .az0re acer 1.27 | 0.211) 1.72) 0.28 |— 0.72 |< “7. | 13,228 OT 150,077). 23.18|ou8s. = Ban | 8 3.01) 1.08.| 0.077} 2.52] 0.38 |— 1.82 Ul ed 13.02) 1.08 | 0.077) 2.64] 0.38 |— 1.94} Standard diet and 10 3.06] 1.08 | 0.077| 2.71) 0.388 |— 2.01} 40 gm. beef heart. 11 13.04] 1.08 | 0.077} 2.39) 0.38 |— 1.69 12 3.01) 1.08 | 0.077| 2.68) 0.38 |— 1.98 Tota) eee 6.61 | 0.462} 6.12) 2.28.|—11.79 Average........... | 1.10 | 0.077) 2.69) 0.38 |— 1.97 a ae | 13 | 13.04 1.14 | 0.211) 1.69 14 | 12207 eta WO .211)), 1.50 As Period I. Animal refused to eat; experiment discontinued. H. B. Lewis TABLE VI. Dog A. Weight 15 Kg. 373 Period. | Day. 1 2 3 : 4 5 6 MOt Ales... 8 0. Average ..... 7 8 y il 10 11 12 Totaliene se sh Average ..... 13 14 15 Ill 16 ites 18 Mowealins x...,.°2 Average ..... 19 20 21 IV 99 23 24 Potalnso oe: Average ..... i | Fecal | Nit oe: Fae eae pea bale Diet. tro- 5 : jon Sulfur.| gen pene | oa gm. gm. gm. gm. gm. 1.14 |:0.082} 2.21] 0.16-|—1.23 1.14 | 0.082} 2.06} 0.16 |—1.08 1.14 | 0.082) 2.20} 0.16 |—1.22| Standard diet and 40 gm. 1.14 | 0.082} 2.40) 0.16 |—1.42| beef heart. 1h 14110. 082|" 2.99] 0.16 :|—1.2 1.14 | 0.082} 2.33) 0.16 |—1.35 .84 | 0.492) 13.42) 0.96 |—7.54 1.14 | 0.082} 2.24) 0.16 |—1.26 1.14 | 0.276) 1.92) 0.22 ;—1.00 1.14 | 0.276) 1.46) 0.22 |—0.54) Standard diet, 37 gm. beef 1.14 | 0.276] 1.41] 0.22 |—0.49| heart, and 0.75 gm. cys- 1.14 | 0.276} 1.33) 0.22 |—0.41 tine. 1.14 | 0.276} 1.21) 0.22 |—0.29 TT4* | 02276) Leal70% 227)|—0225 6.84 | 1.656} 8.50) 1.32 |—2.98 1.14 | 0.276} 1.42) 0.22 |—0.50 1.14 } 0.082} 1.38} 0.24 |—0.48 1.14 | 0.082] 1.85] 0.24 |—0.95} As Period I. 1.14 | 0.082} 1.94] 0.24 |—1.04 1.14 | 0.082} 1.94) 0.24 |—1.04 1.14 | 0.082, 1.88] 0.24 |—0.98 1.14 | 0.082} 1.90) 0.24 ;—1.00 6.84 | 0.492} 10.89] 1.44 |—5.49 1.14 | 0.082) 1.82) 0.24 |—-0.92 1.14 | 0.076} 1.84] 0.22 |—0.92} Standard diet, 37-gm. beef 1.14 | 0.076). 1.85) 0.22 |—0.93| heart, and 0.47. gm. 1.14 | 0.076) 1.87| 0.22 |—0.95| — glycocoll. 1.14 | 0.076) 4.87} 0.22 |—0.95 1.14 | 0.076) 1.86) 0.22 |—0.94 1.14 | 0.076) 1.87) 0.22 |—0.95 6.84 | 0.456} 11.16) 1.382 |—5.64 1.14 | 0.076) 1.86) 0.22 |—0.94 374 Metabolism of Sulfur. II TABLE VI—Concluded. Intake. Uri- Je ; 7 Period. ||| (DB¥-4|\eaeseean ea tee aes Ee eed Diet. pees Sulfur.| gen gen. ance. gm gm. gm gm gm 25 1.14 | 0.076} 1.94).0.21 |—1.01 26 | 1.14 | 0.076) 1.94) 0.21 |—1.01| Standard diet, 37 gm. beef Vv 27 | 1.14 | 0.076} 1.91) 0.21 |—0.98| heart, and 1.12 gm. ty- 28 | 1.14 | 0.076) 1.86) 0 —0.93) rosine. 29 1.14 | 0.076; 1.69} 0.21 |—0.76 30 1.14 | 0.076] 1.68} 0.21 |—0.75 Total ........| 6.84 | 0.456] 11.02) 1.26 |—5.44 Average .....| 1.14 | 0.076) 1.84) 0.21 |—0.91 31 1.14 | 0.276} 1.03) 0.22 |—0.11 32 1.14 | 0.276) 0.81) 0.22 |+0.11| As Period If. VI 33 1.14 | 0.276) 0.70) 0.22 |+0.22 34 1.14 | 0.276) 0.75) 0.22 |+0.17 35 1.14 | 0.276) 0.83} 0.22 |+0,09 36 1.14 | 0.276} 0.95) 0.22 |—0.03 Total ........| 6.84 | 1.656] 5.07) 1.32 |+0.45 Average .....| 1.14 | 0.276; 0.85) 0.22 |+0.07 37 1.14 | 0.082} 0.96) 0.25 |—0.07| As Period I. 38 1.14 | 0.082} 1.27) 0.25 |—0.38 VII 39 1.14 | 0.082} 1.55) 0.25 |—0.66 40 1.14 | 0.082} 1.59) 0.25 |—0.70 41 1.14 | 0.082} 1.66, 0.25 |—0.77 42 1.14 | 0.082 ue 0.25 |—0.70 Total ........| 6.84 | 0.492). 8.62! 1.50 |—3.28 Average .....| 1.14 | 0.082 144 0.25 |—0.55 H. B. Lewis 37D TABLE VII. Dog. B. Short Haired White Female Bull Dog. Nitro- Uri- Fecal | Nitro- Period.| Day. |Weight.| gen rr nitro- |gen bal- Diet. intake. | M!tro gen. ance. gen. kg. gm. gm gm. gm. 1 | 12.36) 1.22 | 1.99) 0.24 |—1.01] 43 gm. beef heart and 2 | 12.29} 1.22 | 1.91] 0.24 |—0.93} standard diet. I 3) 12-238] 122") 2211) 0.24 |= 3 4 | 12.20} 1.22 | 1.95) 0.24 |—0.97 5 | 12.21) 1.22 | 1.83) 0.24 |—0.85 6 | 12.10} 1.22 | 2.05) 0.24 |—1.07 Mo Galleries eee pes ass 7.32 | 11.84] 1.44 |—5.96 AV GT Apes 52 yo 1.22 | 1.97) 0.24 |—0.99 7 | 12.00} 1.22} 1.82} 0.21 |—0.81| 40 gm. beef heart plus 0.9 8 | 12.04) 1.22] 1.39} 0.21 |—0.38| gm. phenylalanine and 9 | 12.00) 1.22.| 1.87) 0.21 lo 86 standard diet. II 10 | 11.89} 1.22 | 1.69) 0.21 |—0.69} 1.0 gm. tyrosine substi- 11 | 11.91) 1.22 | 1.68} 0.21 |—0.68| tuted for phenylalanine. 12 | 11.98] 1.22 | 1.80} 0.21 |—0.80 “LG ee 7.32 | 10.25) 1.26 |—4.22 INVIETAQC. £0. ees: 1.22 | 1.71) 0.21 |—0.70 13 | 11.90) 1.22 | 1.55) 0.24 |—0.57| 40 gm. beef heart, 0.7 gm. 14 11.88) 1.22 1.30] 0.24 |—0.32 cystine, and standard Il 15 ~| 11.82) 1.22.) 1.32) 0.24 |—0.34| diet. TGS L588) 1222 || 1527) 0.24./—0.29 17) 14-93) 1.22 | 1.10) 0.24 |—0.12 VS OSie1.22 |° 1.14).0.24-|—0.16 Patalteen <5 hese 7.32 | 7.68) 1.44 |—1.80 AV OTH Ota) Aco trcres 1.22 | 1.28) 0.24 |—0.30 376 Metabolism of Sulfur. IT TABLE VIII. Dog G. Long Haired Collie. Weight 15 Kg. F Uri- Nitro- Fecal : Ponodal aac nary itro- | Nitrogen Diet. esod.| Day. | gen | mie | THI [balance i gm. gm. gm. gm. 1 | 1.16 | 3.27] 0.37 | —2.48] Standard diet plus beef heart 2 1.16 3:04| 0.37 | —2.25 40 gm. 3 1.16 2289|On372)) —2el0 I AW A.16)) 2583)70.37 |) —2708 5 | 1.45 | 2.81] 0.37 | —1.73| Hamburg steak 43 gm. Go| Wea 27 9GOrs7 | e—1 7A Total .......| 7.54 | 17.68] 2.22 |—12.36 Average .....| 1.26 .95| 0.37 | —2.06 7 | 1.54] 2.44) 0.3 —1.22) Standard diet, Hamburg steak S| Mi basl i 7410-32") 0252 43 gm., plus cystine 0.75 gm. * 9 | 1.54] 1.67] 0.382 | —0.45 TD) S10" 1 0.545)° E267)" 02324) =0°45 11 1.54 1.86} 0.32 | —0.64 12 1.54 1.94) 0.32 | —0.72 Motalies.ees- 9.24 |} 11.32] 1.92 | —4.00 Average ... 1.54] 1.89) 0.32} —0.67 13 | 1.54 | 1.97] 0.26 | —0.69} Standard diet, Hamburg steak 14 | 1.54] 2.67| 0.26 | —1.39| 43 gm., plus glycocoli 0.45 gm. Ill 15 1.54 3.11) 0.26 | —1.83 16 .| 1.54 | 2.92| 0.26 |.—1.64| 17 1.54 2.96|-0.26 | —1.68 otaleocss <2 2| 772 40 |elonoo| Meo ei e2e Average .....| 1.54 2.73) 0.26 | —1.45 more H. B. Lewis BF BIBLIOGRAPHY. . Osborne, T. B., and Mendel, L. B., 1915, J. Biol. Chem., xx, 351. . Osborne and Mendel, 1916, J. Biol. Chem., xxv, 1; xxvi, 1. . Hogan, A. G., 1917, J. Biol. Chem., xxix, 485. . Abderhalden, E., 1915, Z. physiol. Chem., xevi, 1. Totani, G., 1916, Biochem. J., x, 382. . Ackroyd, H., and Hopkins, F. G., 1916, Biochem. J., x, 551. See also Geiling, E. M. K., 1917, J. Biol. Chem., xxxi, 173. . Evvard, J. M., Dox, A. W., and Guernsey, S. C., 1914, Am. J. Physiol., XERIV Oll 2. . McCollum, E. V., and Davis, M., 1915, J. Biol. Chem., xxiii, 247. . Holt, L. E., 1916, Arch. Pediat., xxxiii, 18. . Abderhalden (4), fp. 22. . Lee, F. S., Scott, E. L., and Colvin, W. P., 1916, Am. J. Physiol., xl, 474. . Janney, N. W., 1916, J. Biol. Chem., xxv, 177. NUTRITION INVESTIGATIONS UPON COTTONSEED MEAL. III. COTTONSEED FLOUR. THE NATURE OF ITS GROWTH- PROMOTING SUBSTANCES, AND A STUDY IN PROTEIN MINIMUM. By ANNA E. RICHARDSON anp HELEN S. GREEN. (From the Nutrition Research Laboratory, Department of Home Economics, The University of Texas, Austin.) (Received for publication, June 21, 1917.) In the first of a series of studies upon the nutritive value of cottonseed flour,! we reported: “Our results indicate that cottonseed meal does not contain sufficient mineral for growth, is not actively toxic, contains efficient protein, and per- haps fat-soluble growth-promoting substances, similar to those of butter fat, but in less adequate quantities.”’ A second paper? further demonstrates the efficiency of the pro- tein of cottonseed flour for the normal growth, development, and reproduction of the albino rat. It is the purpose of this paper to consider the content in cottonseed flour of growth-essential factors other than protein and mineral, and to report the results of studies of the protein minimum of cottonseed flour. A preceding report showed that when rats received 50 per cent cottonseed flour in a diet containing no additional growth-pro- moting substances other than those furnished by cottonseed flour, considerable growth was experienced for 135 to 205 days and that when the amount of cottonseed flour was increased to 70 per cent with no additional growth-promoting substances there has been normal growth for 90 days and continuous growth for 165 to 205 days. These results clearly indicate that 50 per cent cottonseed flour furnishes a considerable amount of growth-pro- 1 Richardson, A. E., and Green, H. S., J. Biol. Chem., 1916, xxv, 307. * Richardson and Green, J. Biol. Chem., 1917, xxx, 243. 379 380 Nutritive Value of Cottonseed Meal. III moting substances although probably not enough for normal growth, whereas 70 per cent of the flour furnishes an amount suf- ficient at least for normal growth during 90 days in spite of the fact that this diet is decidedly deficient in mineral content.’ However, from these experiments we could draw no conclusions as to whether this growth was due to the fat-soluble or water- soluble growth factor, or to both, or deduce anything as to the relative amounts of these two substances. We reported also that 50 per cent fat-free, ether-extracted flour in a diet made up with lard and starch failed to induce as pro- _ nounced growth or maintain animals without complete failure for as long a period as did the unextracted flour, due probably to the absence of the ether-soluble, growth-essential substance. How- ever, animals grew for 145 to 190 days and then maintained weights on the ether-extracted flour for 240 days, which seems to indicate in the light of the results reported in the present paper that there must be present in the ether-extracted flour a consider- able amount of the water-soluble growth-promoting factor. A series of carefully controlled experimental diets designed to throw light on the relative amounts of fat-soluble and water-solu- ble growth factors in cottonseed flour has been fed for several weeks. Composition of Diets. per cent per cent A. Ether extract cottonseed flour........ 4.35 12 Casein <1), Cet tem” Lesage 2a eee 18. 18 Tare 8, £52 2c aot) ete eee 17.65 10 Minera lsMuxture, DL. s. .c222 +0 ssc eee tO) 5 Starehe ate eee cee eee 30 30 eaCtOse fence cas ne ene FM, Deere 25 25 3 Ash analysis of cottonseed flour (Golaz). per cen SiO eee ane oa ae ie ee SOE eee s cena 0.14 COTA lees oe ae tas ns OEE None SO eres eres Rete ee ters ef UNE MER et meg ee eae, oth Scene ees 0.06 [OG ie. )e eee eRe AS Be Beh ice 10. tk ae mere > fs Qt sO cee ee ed ere eeh agterd classe ine Thies Sind! oem aoe 2.01 a, Ce a ee Rae aint bis Brak le cain Seam ae 0.26 MgO) coi iiaepee ee epithet 2 a aeibh ede ws. Ae ee 0525 A. E. Richardson and H. S. Green 381 B. Ether extract cottonseed flour........ 4.35 12 Ibtiwol: Sc eles BES OO See 17.65 10 Chai, 2 aes ea ee 18 18 Mineral Mixture III.. es op 5 Starch and water extr Gt ‘of Cotvene BECCBHOUT Roe ccltnd oc se tuleekas 20 20 STALCHEEE IRE so cine atin Bs booting gee 35 35 (CES Ese se on 18 IBUbGeIa Ge eet et es eases le ee 12 TUNG bint AS oe aes Se pe ae ee MOR Pe Foe 12 MineralyMixture* DDE... ...... 02. sce 5 Starch and water extract of cotton- See dutOur sete See ee eee 20 Starch oe eee eee ti eee 33 Dyes Gaseie nes oon, eee Ee: 18 t 6/0200 [ene en ge eR Cd Bil Us Re 24 Minerale Mixture: UG eee ao Starches Mn: Sitar |G Orne eee oe £8 53 The casein used in these diets is purified according to the method described by McCollum and Davis.’ The mineral employed in the rations is supplied by Mineral Mixture III], described by the same authors.° Mineral Mixture IIT. per cent ISIE OMS. 2a ie ane eae ne an ale ae ar Sein Ae oe 3 12-3 K,HPO;: Dodd cto Sei CION an diated. 4. oUt, Ges clomleto Sis. cmudplob waas 28 08 CaH4(PO.4)2.H.O BOSE) CEOS LUNG EXO CONOO DC Des AE DOs 3,5 Bits fob kere 0.74 AR OUR ATYOTOUS):<.)2%:45°%./)03'2 wolgin os Sd o's Fades Prete aE 1.56 Wale OUTER SAE gt sei tesa nel ieee Pa eS ste or 545 IN ie (AIM GT OUS) sf a. eect toe ee Sages sts Weta ae Sle (GIS), JIE YO1 BEN ECE} sr alts ete ee res ce ce MPa A 46.80 Fe CE A lg Peed A el rea 1.64 In Diet A the water-soluble growth factor is supplied by lac- tose of the purity of ordinary reagents.° Diets A and B receive the fat-soluble growth factor contained in the ether extract of cottonseed flour. The extract is made by percolating the flour until the ether comes through colorless. The ether is then driven from the extract by evaporation over a hot 4 McCollum, E. V., and Davis, M., J. Biol. Chem., 1915, xxiii, 233. § McCollum and Davis, J. Biol. Chem., 1915, xxi, 625. § McCollum and Davis, J. Biol. Chem., 1915, xxiii, 181. 382 Nutritive Value of Cottonseed Meal. III water bath at 60°C. These two diets were first made up with 4.35 per cent of the ether extract, an amount equivalent to 50 per cent flour in the diet. After 11 days the amount was increased to, 12 per cent to enable a comparison of the relative amounts of fat-soluble growth-promoting substance contained in butter fat and in cottonseed flour fat. Diets B and C receive the water-soluble growth factor con- tained in the water extract of cottonseed flour dried on starch. The cottonseed flour is extracted for 1 hour with frequent stirring with ten times its weight of distilled water. The liquid is filtered free from the flour, acidified with dilute acetic acid, and heated to boiling to coagulate any possible soluble protein. It is then filtered, evaporated down at 60°C., and dried on starch over a water bath at 60°C. with the aid of electric fans. The water ex- tract of 800 gm. of cottonseed flour dried on 200 gm. of starch weighs 120 gm., the weight of both the extract and starch being 320 gm. Therefore 1 gm. of the starch plus extract is equivalent to 2.5 gm. of cottonseed flour, and 20 per cent of this substance in the diet is equivalent to 50 per cent cottonseed flour, which is the amount of flour generally employed in our experimental ra- tions previously discussed. Diet C receives the fat-soluble growth factor from butter fat. From Chart 1 it may be seen that stock rats growing normally at the age of 36 days when placed upon Diet A supplying the water-soluble growth-promoting substance from lactose, and con- taining only 4.35 per cent ether extract of cottonseed flour, equivalent to 50 per cent flour in the diet, do not receive sufficient fat-soluble, growth-promoting substance from this amount of ether extract to continue normal growth. When the amount of ether extract is increased to 12 per cent, comparable to the con- ~ tent of butter fat which induces normal growth and well-being of the animal, all other nutritive factors being favorable, these same animals although retarded in growth during 11 days on the diet containing 4.35 per cent ether extract now resume a normal rate of growth. When animals receive in their diet the fat-soluble accessory of butter fat and the water-soluble accessory of cottonseed flour equivalent to 50 per cent flour in the diet, as in Diet C, it will be seen from Chart 1 that rats grow perfectly normally. A. E. Richardson and H. 8. Green 383 When both the water-soluble and fat-soluble substances are supplied from 20 per cent water extract and 12 per cent ether ex- tract respectively, of cottonseed flour, it will be seen from Chart 1 that rats grow normally although their growth was retarded dur- ing 11 days by the insufficient amount of fat-soluble growth fac- tor supplied by only 4.35 per cent ether extract. Chart I indicates the behavior of animals on a diet deficient in both the water-soluble and fat-soluble growth-promoting sub- stances. When placed on Diet D the growth of these animals is immediately retarded and after 4 weeks, they begin to lose in weight. These results indicate that 50 per cent cottonseed flour in a diet furnishes sufficient water-soluble accessory for normal growth, but does not furnish enough fat-soluble accessory. As compared with 12 per cent butter fat, 12 per cent ether extract of cotton- seed flour equivalent to 138 per cent of flour is apparently as efficient in supplying the fat-soluble food accessory. Earlier studies of cottonseed flour!’ reported the efficiency, both for growth and normal reproduction of the albino rat, of diets in which cottonseed flour furnished the only source of pro- tein. These diets furnished an abundance of protein, 25 per cent, as well as the other essentials to normal growth of the rat. To test still further the efficiency of cottonseed proteins as compared with those proteins of known physiological value, a series of ex- periments was made to determine the protein minimum of cotton- seed flour. The experimental diets contain all the essentials to growth with varying amounts of protein, 4 to 18 per cent as follows. per cent: DIME OOtGONSCEOHOUL:. - os oog0e it eatsal ess cess sivas nemeesne 36 iRrotem-—tree milk. ..3........: sa Sr SN AE ee ee ee See 10 SUE DEUBUA Gp eae oes tie cre. Sicccio Wa sco Lciotsd olaraheusyer cia Popaiat nce 12 ILATGL. . o satsois orchard cho pees tae ene ee eee gee 16 SHOIRGID, oe acalS ie 5 PE ae ee ee 26 PRMOGIOMCECE, HOUE: 2. 5 ok. co cients coc dacs aabdossesccesesten 24 Protem-tree milk. ..2....5...:.... deat pee ee © Ds. Sefton 10 JURE EOD. © oni UR rr 12 JORIS, cystic fed Bese eal ene leg 16 SUA NPE oak As ec ooeidlp ieee se eee wena 38 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 384 Nutritive Value of Cottonseed Meal. III 14. Cottonseed flourt..2 ke eee eae eee eee 8 Butier:fate). Sot 2.2 eee eee Me ee? ection Pe ee hs 12 Thar ..9:<2ss.c arte de A oe ER os cao ie oa eee 15 Proteim=free amulllk.., .. cece he ae cs ee 22 Starch ethic. ce: os ee ee en ns, ae ae 43 45. Cottonseed hour) s/n ee eee ee hens dee oo Ee 18 Butter dat verre ates a eee ee ea nae 2 12 Protemn=freewmilk 5-0. a Ee ieee nes eee 22, iain dase cicee orien acas. rsh. 6s RTE ROE RE tee has ey, 2x ee 15 Stare octet ets sic Fine ee eee tae 33 46>*Cottonseed our...) 5 a. ee ee hn ee eae 12 BuUttieraaiGe wk ven Bees cise ete. CASI era a 12 BT eee. psihe ttc oon tone TNO. oo oo ee eee 15 ‘Proteim=free; mills 5 oeee. ch sesoe iets .c soe Oe oe eee 22 Starch oe ee cis cis tes sacs on eens Pare Se ee ee 37 On a diet in which 18 per cent protein is furnished by 36 per cent cottonseed flour as in Diet 23, rats have grown practically normally and are still alive at the age of 410 days. As may be seen from Chart II, the females have grown larger than the aver- age female, whereas the males have been very slightly under size. Three of the four females on this diet have reproduced, Rat 1027 giving birth to three litters, and Rat 1029 to two litters. Of the six families produced on this diet only six young, Nos. 2060-5, from two families, have survived. At the age of 148 days the one male rat is a little above four-fifths of the average weight at that age, while at the age of 159 days the females are slightly under ayerage weight. When 12 per cent protein is furnished by 24 per cent cottonseed flour as in Diet 24 the majority of the animals have not grown normally (Chart III) although Female 1028 attained normal full growth quicker than the average individual. Only one of three females ever reproduced, No. 262, giving birth to three females, Nos. 2029, 2030, 2031, all of which are alive at the age of 269 days, fine appearing animals but below average size. On Diet 45 furnishing only 9 per cent protein from cottonseed flour, all the animals are decidedly under weight at the age of 165 days. But Female 310 at the age of 155 days gave birth to eight fine looking young, all of which she devoured (Chart IV). These results do not altogether agree with the work reported by aa A. E. Richardson and H. S. Green 385 Osborne and Mendel.’ They obtained normal growth with 9 per cent cottonseed protein although they report no reproduction. Similarly, as seen by Chart IV, on Diet 46 which furnishes only 6 per cent cottonseed protein our rats have been able to show very little growth during 108 days, the animals averaging only 8 to 9 gm. increase in weight, whereas Osborne and Mendel have ob- tained considerable growth on 6 per cent cottonseed protein. With Diet 44 containing only 4 per cent protein, an amount smaller than any considered by Osborne and Mendel, rats have behaved as indicated by Chart IV. When first placed on this diet there is a decided loss in weight for several days, after which there is an almost successful attempt at maintenance for about 50 days. These animals are still continued on this diet and though stunted are in good condition, extremely active, and have fine coats of fur. SUMMARY. 1. 20 per cent of the water extract of cottonseed flour dried on starch, equivalent per gm. to 2.5 gm. of cottonseed flour, 7.e., 50 per cent cottonseed flour in the diet, contains sufficient water- soluble food accessory for normal growth. 2. 4.35 per cent of the ether extract of cottonseed flour equiva- lent per gm. to 11.5 gm. of cottonseed flour, z.e., 50 per cent cot- tonseed flour in the diet, does not contain sufficient fat-soluble food accessory for normal growth, but 12 per cent of the ether extract appears quite as efficient in supplying enough of the fat- soluble accessory for normal growth as does an equivalent amount of butter fat. 3. 18 per cent cottonseed protein when supplied with ade- quate amounts of all other necessary nutritive factors induces . practically normal growth of the male rat, and better than aver- age growth in the female, and fairly normal reproduction, with high mortality among the second generation. At the age of 148 days the male of the second generation is about four-fifths aver- age size and the female slightly under size. 4. 12 per cent cottonseed protein does not induce perfectly normal growth. On this diet one female has borne three young, 7 Osborne, T. B., and Mendel, L. B., J. Biol. Chem., 1917, xxix, 289. 386 Nutritive Value of Cottonseed Meal. III all of which are alive at the age of 269 days, although below aver- age size. 5. Normal growth has not been obtained on 9 per cent cotton- seed protein but at the age of 155 days one animal, No. 310, has borne a fine looking litter of eight young, all of which she devoured. 6. Very little growth has been obtained with 6 per cent pro- | tein, the average gain in weight during 108 days being 8 to 9 gm. 7. With only 4 per cent cottonseed protein rats have fallen off in weight when first placed upon this diet but have almost suc- cessfully maintained their weight for 50 days thereafter. { 110) saree ye ee. aaa dina (in is CuartlI. X indicates the point at which the rats were given the experi- mental diets. With animals on diets A and B, XZ indicates the period during which the diet contained 4.35 per cent ether extract of cottonseed flour. Z indicates the point at which the amount of ether extract was in- creased to 12 per cent. Rats receiving 20 per cent water extract of cotton- seed flour with all other nutritive factors favorable in Diet C grow nor- mally. Rats receiving 12 per cent ether extract of cottonseed flour with all the nutritive factors favorable in Diet A, period YZ, grow practically normally. Ratsreceiving both water and ether extracts of cottonseed flour » in Diet B, period Z, grow normally. Rats on Diet D lacking in both water- soluble and fat-soluble food accessories do not. continue normal growth. ‘ 30 DAYS ce CuartII. Indicates the behavior of animals on Diet 23, furnishing 18 per cent cottonseed protein. Rats 2060, 2061, 2064, and 2065 are of the second generation on this diet raised since weaning on Diet 23. pie eee -}-—— Pea or eae é: Zan oO DAYS AGE 20 DAYS Cuart III. Indicates the behavior of animals on Diet 24, furnishing 12 per cent cottonseed protein. Rats 2029, 2030, and 2031 are of the second generation on this diet raised since weaning on Diet 24. 387 Ps 388 Nutritive Value of Cottonseed Meal. III pater mrad YOR Tree . Cuart IV. The curves of Rats 283 and 292 indicate the behavior of animals receiving 9 per cent cottonseed protein in Diet 45. Rats 318 and 285 have received 6 per cent cottonseed protein in Diet 46. Rats 335, 342, and 351 received only 4 per cent cottonseed protein in Diet 44. THE AVAILABILITY OF THE ENERGY OF FOOD FOR GROWTH.* By C. R. MOULTON. (From the Department of Agricultural Chemistry of the University of Missouri, Columbia.) (Received for publication, June 4, 1917.) Food which enters the animal body has a certain amount of total energy, called the heat of combustion. This energy is different,for different feeds. Of this total energy the animal loses part by way of the feces in the undi- gested food residues, part by way of the urine in incompletely oxidized bodies, and part by way of combustible gases voided. The amount over and above these losses is called the metabolizable energy. Not all of this metabolizable energy is available for the uses of the animal body in either maintenance or growth. There is a loss due to the work of digestion, mastication, and movement of the food through the digestive tract. There is also a further loss due to a stimulated metabolism upon the absorption of digestible substances from the alimentary tract. There may be a slightly greater muscular activity due to the increased food consumed. All this energy is converted into heat and lost from the body. What is left of the metabolizable energy after these second losses are accounted for is called the net, or available, energy. This may be used for production of work, or may be stored in the animal body in the form of protein, fat, or other body substances. In connection with the general ‘‘ Use of Food Experiment”’ con- ducted at the University of Missouri Agricultural Experiment Station since 1907, some data have been obtained upon the rela- tive amount of the energy of the food which may be recovered in flesh gained. The animals used were mature beef steers, 2 or 3. years old, of the Shorthorn breed. They were as nearly alike in body weight, previous method of treatment, and type as it is possible to select beef steers. One animal was somewhat heavier than the other two. The ration used in the work consisted of five parts of mixed grain to two parts of alfalfa hay. The mixed grain * Read before the Division of Biological Chemistry of the American Chemica! Society, April 12, 1917. 389 390 Energy of Food for Growth was eight parts maize meal (corn chop) and one part old process linseed meal. The digestibility of the ration was determined by digestion trials. The cost of maintaining the animals at constant weight was determined by extended maintenance trials. Greater de- tails will be found in previously published work.! After the maintenance trial one animal, Steer 18, was slaughtered for analy- sis. Two other animals were fattened, one to full prime condi- tion, Steer 48, and the other to a condition 40 to 50 days under prime, Steer 121, when they were slaughtered and analyzed (Table I). TABLE I. Composition of Animals. Weight. Steer 18. | Steer 121. Steer 48. Warm empty weight, gm........... 302,183 508,513 744,708 Perscention watetec-s on: ce. ose aio 50.02 41.73 Weight ofmyater.gim.. 2. fo 3 82025: 173,259 254,339 310,750 Pericentrotelate meer cece ore 18.03 29.72 41.25 Weieht- onset, Gio recn. noi en 54,479 151,131 307,164 4 Per*cent, of nitrogente = s2\s2chsas « 2.96 P| 2.07 v4 Weight of nitrogen, gm............ 8,955 12,776 15,391 ‘ Weirht of protein, on.1% 0... 55,968 79,847 96,194 Percent GUase esc se ook et boes 5.70 4.14 3.45 | Weight: of ashrigmis.-. 0eeercis ace 17,211 21,064 25,697 Per cent of phosphorus............ 1.07 0.74 0.62 Weight of phosphorus, gm.......... 3,230 3,769 4,648 Since Steer 121 during maintenance, or at the beginning of the full fed period, weighed less than the check animal and Steer 48 weighed more it is necessary to calculate their composition as- suming the same percentage composition as the check animal had. Table II shows the results together with the calculations for the composition of the gain. In calculating the thermal equivalent of the fat and protein gained it was necessary to use the data of other investigators. For protein the value of 5.6776 calories per gm. was used. This 1 Trowbridge, P. F., Moulton, C. R., and Haigh, L. D., Missouri Agric. yi if Exp. Station Research Bull. 18, 1915. \ Cakt roulton 391 is the value found by Kohler? for the lean muscular tissue of beef cattle from which the fat had been removed by ether and a cor- rection made for the fat in the residue as determined by the Dormeyer*® method. For fat the value of 9.4889 calories per gm. was used. This is the average of four results for beef fat quoted by Fries,* namely, those of Stohman and Langbein, Stohman and associates, Gibson, and Danilewsky. Steer 121 stored up 926,359 calories in the fat gained and 141,252 calories in the protein gained. This is a total of 1,067,611 cal- ories. Steer 48 stored up 2,355,601 calories in fat gained and TABLE II. Composition of Gain. Energy Stored in Flesh Gained. Steer 121. Steer 48. : Composition : Composition Flesh gained . Flesh gained are (estimated). Sa (estimated). (enna ; gm, per cent gm, per cent Warm empty weight........... PAU TPA | 5 po ac AlE900)|F ee sec \NVGHIGT Rs ae ee ere een 84,174 39.76 123,372 29.52 Be mene aria es als a ie eocane 97,626 46.11 248,246 59.40 INGCTOSCME A: ee oe esc: ee 3,981 1.88 5,707 WES PROGINS is). fe ih's lee ticch cn. 24,879 i765) 35,667 8.54 LAGI 6 Jos eae a 4,161 1.97 7,084 1.70 BhoOsphOnuss mae see aoe 597 0.28 l5D 0.28 Energy in fat, calories.......| 926,359 2,355,601 fs sprotems. .‘ 141,252 202,502 Total energy stored, “ ....... 1,067,611 2,558,103 202,502 calories in protein gained. This is,a total of 2,558,103 calories. In order to make these gains and store this energy these animals consumed a large amount of feed. Steer 121 consumed over 2,000 pounds of digestible organic nutrients and Steer 48 con- sumed nearly 8,000 pounds. The equivalent metabolizable energy was found by the method of Armsby.® For this ration it is 3,803 ? Kohler, A., Z. physiol. Chem., 1900-01, xxxi, 479. 3 Dormeyer, C., Arch. ges. Physiol., 1896-97, lxv, 102. ‘ Fries, J. A., U. S. Dept. Agric., Bureau Animal Industry, Bull. 94, 1907, 13. ) 5 Armsby, H. P., and Fries, J. A., J. Agric. Research, 1914-15, iii, 451. 392 Energy of Food for Growth calories per kg. of digestible organic matter, or 1.72 therms per pound. eves more cere 3 FOG OBE ck cae tae ae ke eee O05), PDCeMEXULACO NT .- -erere eee 2-3 Cooked meat, beef, ham, am. : fish: %, 332425: Ree 180-200 Oleomargarine (in casoid bis- 5 hettucela-dack eer eee 30 CUD) Seis eR eg a rts 18 Celery (thrice boiled)...... 15-30 Lard (in casoid biscuit)........ 12 Cauliflower (thrice boiled)... 150-180 Buttericecss toe eee 75-90 Nelson and Burns found the normal urinary calcium oxide excretion in man to range between 0.18 and 0.62 gm. per day. In our diabetic the alkali seems not to have influenced the urinary calcium output. 5 This case is included in the paper of Underhill, F. P., J. Am. Med. Assn., 1917, lxviii, 497. _-°~ M. H. Givens and L. B. Mendel 433 That the balance of calcium in the diabetic ordinarily is nega- tive seems to be established from the recent work of Kahn and Kahn who made careful analyses of the intake and output of frve diabetics. The literature of the subject is given in their paper. SUMMARY. Administration of base or acid produced no significant effect upon the balance of nitrogen, calcium, magnesium, and phos: phorus in the dog. Administration of hydrochloric acid increased the urinary ex- cretion of calcium and thereby altered the relation of calcium to magnesium in the urine. The calcium contained in milk was more effective than soluble calcium lactate in producing calcium retention. Administration of large doses of alkali bicarbonate to a human diabetic did not decrease the urinary output of calcium. BIBLIOGRAPHY. Albu, A., and Neuberg, C., Physiologie und Pathologie des Mineralstoff- wechsels, Berlin, 1906. Atwater, W. O., and Bryant, A. P., U. S. Dept. Agric. Bull. 28 (rev.), 1906. Bertram, J., Z. Biol., 1878, xiv, 354. Caspari, Berl. klin. Woch., 1897, xxxiv, 126. ‘Dubois, M., and Stolte, K., Jahrb. Kinderheilk., 1913, Ixxvii, 21. Gaehtgens, C., Z. physiol. Chem., 1880, iv, 36. Gerhardt, D., and Schlesinger, W., Arch. exp. Path. u. Pharm., 1899, xli, 83. Granstrém, E., Z. physiol. Chem., 1908-09, Iviii, 195. Hart, E. B., and Steenbock H., J. Biol. Chem., 1913, xiv, 75. Kahn and Kahn, Arch. Int. Med., 1916, xviii, 212. Magnus-Levy, A., Arch. exp. Path. wu. Pharm., 1899, xlii, 163. Malcolm, J., J. Physiol., 1904-05, xxxii, 183. Mendel, L. B., and Benedict, S. R., Am. J. Physiol., 1909-10, xxv, 23. McCrudden, F. H., J. Biol. Chem., 1911-12, x, 187. Nelson, C. F., and Burns, W. E., J. Biol. Chem., 1916-17, xxviii, 237. Ruedel, G., Arch. exp. Path. u. Pharm., 1894, xxxiii, 79. Secchi, R., Biochem. Z., 1914, lxvii, 143. Sherman, H. C., and Gettler, A. O., J. Biol. Chem., 1912, xi, 323. Tereg and Arnold, Arch. ges. Physiol., 1883, xxxii, 122. Le iy STUDIES IN CALCIUM AND MAGNESIUM METABOLISM. II. THE EFFECT OF DIETS POOR IN CALCIUM. By MAURICE H. GIVENS. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven.) (Received for publication, June 20, 1917.) Albu and Neuberg have stated that the kind and intensity of calcium exchange is extraordinarily dependent upon the quality of the nourishment. Forster fed a dog washed meat and showed by analysis of blood and tis- sues that the dog had lost calcium. The loss from the tissues did not equal the excess of output over intake, therefore the bones must have yielded some calcium. Perl found a normal urinary calcium excretion in the dog on acalcium- poor diet of bread and a little condensed milk. The output was increased on a heavy meat-lard diet. Goitein found that rabbits on a calcium-rich diet of oats and bone meal put on calcium and magnesium; with a diet less rich in calcium, as oats alone, the animals were in calcium and magnesium balance; and with a “ ealeium-poor diet, as maize alone, the rabbits might be in nitrogen and magnesium balance but negative calcium balance. Boekelman and Staal studied the effect of a calcium-poor diet and a calcium-rich diet on calcium excretion in man. The calcium-poor diet was an ordinary mixed ration while the comparison diet had its calcium con- tent a little more than trebled by the addition of a liter of milk. In their four subjects the urinary calcium increased on the calcium-poor diet and decreased on the calcium-rich diet. Patterson fed rabbits on oat-meal and maize, a diet which, as he says, “admittedly leads to calcium starvation.’’ He found the amount of cal- cium in the blood in a normal proportion to the rest of the ash; in the bones, however, the relation of the calcium to the total minerals was re- duced. He believes: ‘‘The bones can, without doubt, act as reservoirs of calcium (and probably of magnesium).”’ The most intensive study made of the effect of diet on the calcium and magnesium excretion of dogs is that of Kochmann and his pupil Petzsch. From a study of the effects of protein, fat, and carbohydrate on the alkali- earth metabolism of dogs they concluded that the calcium balance was 435 436 Studies in Metabolism. II apparently influenced by the amount of these organic nutrients, in that an addition of any one of the three substances to the basal ration would draw out lime from the organism in a noticeable degree. They thought the amount of calcium necessary for a maintenance of calcium balance de- pended upon the character of the food intake and had to be determined for each diet. Unlike the case of calcium metabolism they found magnesium exchange unaffected by protein, fat, and carbohydrate. The phosphorus metabolism likewise is believed to be influenced by the protein intake. These workers did not use a mixed diet. Protein, fat, and carbohydrate respectively were superimposed upon a diet of protein alone. In our ex- periments we have endeavored to provide a constant caloric intake on a mixed diet. TABLE V.* Urinary Alkali-Earth Excretion, Mg. per Kg. Wea oU| eGa0 MgO Diet. Author. kg. 11 0.7 0.8 No food. Mendel and Benedict. 11 Les. 3.0 *s Givens. 14 0.5 2 Ca-poor. Mendel and_ Benedict. 20 1.0 2:5 rs Gaehtgens. 14 1.0 4.0 oh Givens. 20 1.5 1S = Gaehtgens. 15-5 1.8 ef Tereg and Arnold. 16.4 2.0 4.0 s Kochmann. 10.3 Da? 8.8 fc Gs 13.5 225 & Ruedel. 24 216 5.0 sé Secchi. 13 3.0 See) sf Givens. 14 3.0 U0 ff ef 9.3 3.0 7.6 sf Kochmann. 4 3.0 10 “ Heiss. 14.5 See Bow ss Givens. 12 3.5 4.1 £6 ee DRS 5 10 - Kochmann. 33 183 Ca-rich. Tereg and Arnold. 22 ie sf Perl. if 3.4 8 es Maleolm. 9.6 4 5 on * 6.5 4.6 12 i Kochmann. 14 6 a ce Secchi. 7 16 13 i Kochmann. * The tables are numbered consecutively through the three papers. M. H. Givens 437 The results of Kochmann are included in the following résumé (Table V) of all of the available literature on the urinary alkali- earth excretion of adult dogs. The statistics fail to show any definite relation between the diets and urinary calcium and mag- nesium. The tendency is for the excretion of these elements to increase as the intake of them increases. From the literature on the subject we are led to conclude that a diet poor in calcium is not conducive to a storage of either cal- cium or magnesium despite an abundance of nitrogenous food. This is further established by our own investigations. The fact that the body may show a satisfactory nutritive bal- ance with respect to one essential element while being depleted of another deserves more emphasis than it has received in the past.!. There are available for comparison in our series of ex- periments, the results obtained on five animals; viz., Dogs B,?F,? J,4 FL, and M: (Table VI). With Dogs B, F, and M, there re- TABLE YI. Nitrogen, Calcium, and Magnesium Metabolism, Average Daily Analyses. Dog Ms. Urine. Feces. Balance. Period. | Diet. ash a as a ee ee ee ts a Weight- N Ca | Mg N Ca | Mg N Ca Mg gm. mg. | mg. | gm. mg. | mg. gm. mg. mg. kg. 1 Az | 5.98 | 34 | 49 | 0.60 | 270) 30 | +0.91 | —159} —17 | 13.8 2 a 5.41 | 23 | 45 | 0.72 | 220) 26 | +1.30 | —99} —11 | 13.8 3 - 5.52 14.0 4 By 5.46 | 36 | 55 14.2 5 S 5.07 | 83 | 47 | 14.3 sulted from the calcium-poor diets a negative balance of calcium and magnesium. The relation of calcium to magnesium in the urine was not altered appreciably in any of the dogs. Whether this would have been the case if the experiments had been con- 1 Tilustration of the significance of this will be found in the compilation of Forbes and others, Ohio Agric. Exp. Station Bull. 295, 1916, and Bull. 308, 1917. -? Table IT, p. 425. 3 Table III, p. 426. 4 Table VII, p. 442. 5 Table VIII, p. 442. 438 Studies in Metabolism. II tinued over a sufficiently long period cannot be foretold. The excretion of lime in the feces was reduced owing to the decreased intake of the element. No apparent relation between nitrogen and calcium output is evident in any of the dogs. In almost every instance the nitro- gen balance has been positive while the calectum balance has been negative. This might be expected from the nature of the diet, ; which, though rich in nitrogen was poor in calcium. The relation between the outputs of calcium and magnesium on the diets here under discussion is by no means constant. Thus with Dog B the urmary magnesium excretion was greater than the calcium. This decided difference may be due to the fact that this old laboratory dog had probably been on a calcium-poor diet for years, receiving no more lime than was actually needed to maintain her health. The relation of calcium to magnesium in the urine was almost always about 1:1 in all of the other animals. Calcium always exceeded magnesium in the feces as 2:1 or more in both Dogs B and F. These two animals excreted the most of their phos- phorus through the kidneys and its variations did not seem to influence the calcium output. From the data submitted it is evident that diets poor in calevcum are not conducive to positive calcium balance even when an abun- dance of nitrogenous food is available. - My thanks are due Professor Lafayette B. Mendel for his advice and criticism. BIBLIOGRAPHY. Albu, A., and Neuberg, C., Physiologie und Pathologie des Mineralstoff- wechsels, Berlin, 1906. Boekelman, W. A., and Staal, J. P., Arch. exp. Path. u. Pharm., 1907, lvi, 260; = Forster, J., Z: Biol., 1876, xii, 464. Gachtgens, C., Z. physiol. Chem., 1880, iv, 36. Goitein, S., Arch. ges. Physiol., 1906, exv, 118. Heiss, E., Z. Biol., 1876, xii, 151. Kochmann, M., Biochem. Z., 1911, xxxi, 361; 1912, xxxix, 81. Kochmann, M., and Petzsch, E., Biochem. Z., 1911} xxxii, 10, 27. Maleolm, J., J. Physiol., 1904-05, xxxii, 183. M. H. Givens 439 Mendel, L. B., and Benedict, S..R., Am. J. Physiol., 1909-10, xxv, 23. Patterson, 8S. W., Biochem. J., 1908, ii, 39. Perl, L., Arch. path. Anat., 1878, Ixxiv, 54. Ruedel, G., Arch. exp. Path. u. Pharm., 1894, xxxiii, 79. Secchi, R., Biochem. Z., 1914, Ixvii, 143. Tereg and Arnold, Arch. ges. Physiol., 1883, xxxii, 122. t a : fi ass - 7 é & : 26 Rey oe a , ~ P i ee - ’ ; v 7 . W at &, . ‘ . . f é Ne: " . * : ' ‘ ‘ . \ . ” a F ? \. on ° ' ¥ 4 * i) % . \ , . . i mel a ery STUDIES IN CALCIUM AND MAGNESIUM METABOLISM. III. THE EFFECT OF FAT AND FATTY ACID DERIVATIVES. By MAURICE H. GIVENS. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven.) (Received for publication, June 20, 1917.) The influence of fat on the mineral metabolism is of prime importance in the nutrition of growth. Steinitz has shown that in infants a high fat diet may lead to a negative balance of the alkalis and a reduction of the alkali earths. On a diet rich in cream the absorption of calcium dropped from 76 to 34 per cent. Cronheim and Mueller have shown that in children on an ordinary milk diet the fecal calcium is not solely combined as soaps. Some of it is ex- creted in this form but there is a great excess in other compounds. In other words, on a moderately rich fat diet only a small amount of the — calcium is removed by the fatty acid. Bahrdt has confirmed the statement of Cronheim and Mueller. Rothberg found that a nourishment rich in fat gave a negative calcium balance. In the same children Birk found that the magnesium balance was also negative. Meyer says that the results of Steinitz and Rothberg and Birk are not contradictory, but that the condition of the child greatly influences the results; as, for instance, if there were a diarrhea there be would less chance for absorption in the intestines. In experiments referred to in Paper II of this series Kochmann found that addition of lard or dextrose to a diet of protein alone increased the urinary calcium excretion. From his study of the influence of protein, fat, and carbohydrate he concluded that calcium balance was dependent upon the kind and amount of nourishment. Experiments which were being conducted in the laboratory by Dr. J. F. Lyman on the utilization of certain fats, fatty acids, and their derivatives offered a further opportunity to study me calcium excretion under these conditions. Two male dogs were used, so that all experiments were made in duplicate. In the control periods the dogs received a stand- ard diet of 250 gm. of clean beef, 50 gm. of cracker meal, 40 gm. 441 442 Studies in Metabolism. III TABLE VII.* Nitrogen and Calcium Metabolism, Averages per Day.—Dog J. ~< Urine. Feces. | Balance. — ao 2 Substituted for lard. Be 2 Nellvca | on 4 -caleee gm mg gm. mg ea 1 | — 8.83) 19 | 18 | 0.60) 257 |+1.00|—171| 95 2 | Ethyl palmitate. 8.68) 20} 18 | 1.33) 393 |+0.41/—314| 57 3 9.61) 32] 24] 0.69) 329 |+0.07|—257| 97 4 | Glyceryl palmitate. | 7.99} 23] 25 | 0.51) 364 |+1.93/—214) 93 5 9.97} 34] 30 | 0.65):329 |—0.19)—264| - 93 6 | Palmitie acid. 9.15) 27 0.76) 472 |+0.52/—400) 81 7 8.89) 31 0.66} 72 |+0.88|—485| 97 8 | Glyceryl palmitate + Ca lactate. 11.74| 27 0.52) 122 |—1.82/+107| 92 9 9.08} 21 0.62| 307 |+0.73|—228] 98 10 | Starch and sucrose. |10.23/) 16 0.61) 221 |—0.44;—168 * The tables are numbered consecutively through the three papers. ** My thanks are due Dr. J. F. Lyman for the use of these unpublished figures. TABLE VIII. Nitrogen and Calcium Metabolism, Averages per Day.—Dog FL. os Urine. | Feces. Balance. = 2 2 Substituted for lard. ES fj N Ca Mg N Ca N Gal S gm mg. | mg. | gm. | mg gm mg oak 1 LOZ) al 20 | 0.70} 300|—0.40)/—214| 94 2 | Ethyl palmitate. 9.08] 17 22 | 0.91} 285|+0.44/—207| 48 3 9.47) 29 25 | 0.84] 336/+0.15|—264| 96 4 | Glyceryl palmitate. | 8.68] 29] 35 | 0.62) 350|+1.13)/—278) 93 5 9.14) 33 33 | 0.34] 378/+0.64/—314| 93 6 | Palmitic acid. 9.25) 29 0.67} 492/+0.50|—428| 79 7 9.84 37 0.76] 705|+-0.17|—700| 95 o) Glyceryl palmitate + Ca lactate. 9.11} 26 0.64|1,380|+0.68| —50| 88 9 | Cridiarot 0.53) 307|+0.79|—236| 98 ‘10 | Starch and sucrose. | 8.52) 26 0.77| 336|)+1.13|—257; — * My thanks are due Dr. J. F. Lyman for the use of these unpublished figures. M. H. Givens 443 of lard, 10 gm. of agar, and 400 ce. of tap water. In the other periods an equal quantity of ethyl palmitate, glyceryl palmitate, palmitic acid, and glyceryl palmitate plus calcium lactate (8 gm. daily for 2 days, then 10 gm. daily for 2 days) respectively were substituted for the lard. In the last period the calorie equiva- lent of 40 gm. of fat was replaced by a mixture of cooked starch and sucrose. The results are summarized in Tables VII and VIII. The present conception of the digestion and utilization of fats and other comparable esters of fatty acid would lead one to ex- pect that if they are hydrolyzed in the normally functioning ali- mentary tract the resulting fatty acid will either be absorbed promptly or excreted as insoluble soap with the feces. The ex- tent to which absorption occurs may therefore depend not only upon the digestion of the esters but also upon the degree to which alkali earths are simultaneously present in the intestine to ren- der the fatty acids insoluble and unutilizable. Conversely the loss of alkali earths through the bowel may likewise be promoted by the presence of large quantities of fatty acids. The extent of digestion and utilization of palmitic acid and its derivatives in Dr. Lyman’s experiments will be recorded elsewhere. A study of the data here presented shows, with respect to the deportment of the calcium, that when the utilization is poor the loss of calcium is proportionately larger. This is exemplified as a rule in the following data selected from the tables. TABLE IX. Relation of Calcium Excretion to Fat Utilization. | Fati n food. Fat utilization. | “Daily Ca output per cent mg. Dog J. Lard. 95 257 Ethyl! palmitate. 57 393 Lard. 93 329 Palmitic acid. 81 472 Dog F. Lard. 94 300 Ethyl palmitate. 48 285 Lard. 93 378 Palmitic acid. 79 492 444 Studies in Metabolism. III Similarly the negative caletum balance was smaller in those cases where the utilization of the fat was more satisfactory. Although the quantity of calcium fed as calcium lactate in one of the periods was undoubtedly sufficient to induce a storage of lime on the basal diet (Table VII, Period 8) this could not be accomplished when the fat utilization was poor (Table VIII, Period 8). It is evident from the data presented that poor utilization of fats or fatty acids may increase the excretion of lime in the feces and prevent the storage of calcium even when the caleiwm intake is com- paratively abundant. My thanks are due Professor Lafayette B. Mendel for his ad- vice and criticism. BIBLIOGRAPHY. Bahrdt, H., Jahrb. Kinderheilk., 1910, Ixxi, 249. Birk, W., Jahrb. Kinderheilk., 1907, lxvi, 300. Cronheim, W., and Mueller, E., Biochem. Z., 1908, ix, 76. Meyer, L. F., Jahrb. Kinderheilk., 1910, 1xxi, 1. Rothberg, O., Jahrb. Kinderheilk., 1907, Ixvi, 69. Steinitz, F., Jahrb. Kinderheilk., 1903, lvii, 689. —— THE RELATION OF THE QUALITY OF PROTEINS TO MILK PRODUCTION. Iil.* By HE. B. HART ann -G: C. HUMPHREY. WITH THE COOPERATION OF BARNETT SURE. (From the Departments of Agricultural Chemistry and Animal Husbandry of the University of Wisconsin, Madison.) (Received for publication, July 2, 1917.) In 1915! we presented data showing marked inequality in the efficiency of the protein mixture of rations for milk production. This difference in efficiency we attached to the character and quantitative proportion of the proteins constituting the mixture. We emphasized the fact that the mammary gland, in its construc- tive capacity for milk proteins, was not independent of the quality of the proteins in the ration and further that the “nu- — tritive ratio” or plane of protein intake may be varied with varia- tion in the source of the proteins. With milk proteins consti- tuting about 70 per cent of the proteins of the ration for a cow— the remaining nitrogen coming from the roughage, corn stover— “it was possible to maintain a positive nitrogen balance and the . production of 35 pounds of milk per day with a nutritive ratio of 1:8. - This is an exceptionally ‘‘wide” ratio. Where the pro- teins were drawn from the corn or wheat kernels and constituted the same proportion of the corn stover roughage ration as indi- cated above, no such positive nitrogen balances could be main- tained. We pointed out that during the negative nitrogen balance increased tissue autolysis resulted, and for a brief time at least there was no decrease in the milk proteins or milk solids elaborated. In 1916? we showed that with corn stover and corn meal as the basal ration there were appreciable differences in the efficiency * Published with the permission of the Director of the Wisconsin Ag- ricultural Experiment Station. 1 Hart, E. B., and Humphrey, G. C., J. Biol. Chem., 1915, xxi, 239. ? Hart and Humphrey, J. Biol. Chem., 1916, xxvi, 457. 445 446 Quality of Proteins for Milk Production. III with which certain protein concentrates could supplement such a ration for milk production. With the concentrates furnishing ap- proximately 50 per cent of the total digestible proteins of the ration and the nutritive ratio fixed at approximately 1:8 (the total proteins constituted about 10 per cent of the dry matter of the ration) the gluten feed was measurably inferior to oil meal, distillers’ grains, casein, or skim milk powder as a protein supplement in the particular mixture used. In the proportion used and with corn stover as the roughage none of these concen- trates was able to maintain the animal in nitrogen equilibrium. ~ A possible exception was milk powder. However, the distinctive point we wish to emphasize is that the negative nitrogen balance was much greater in the case of gluten feed than with the other materials used. The efficiency of the proteins of distillers’ grains was attributed to the presence of the embryo of the seed. We pointed out that the comparison would probably hold only for the mixture studied and that a different behavior might be expected should the basal ration be varied. Our expectations were entirely confirmed by the data to be presented. When the roughage of the ration was changed from corn stover to clover hay (medium red, Trifoliwm pratense) but the corn kernel maintained as the basal grain, the inefficiency of the gluten feed as a supple- ment disappeared, and among the four protein concentrates studied, namely, gluten feed, oil meal, distillers’ grains (Ajax), and cottonseed meal there was little, if any, difference in efficiency. - Furthermore, positive nitrogen balances were maintained during most of the periods of observation (16 weeks) on the nutritive ratio of 1: 8:5. EXPERIMENTAL. The plan was to use a basal ration of clover hay, corn silage, corn meal, and starch, to which would be added the protein con- centrate. The basal ration was maintained constant in relation to its source.and proportion of nutrients for any individual in the different periods, the only variable in succeeding periods being the concentrate and the amount of starch. These were supplied in such quantities as to make the plane of protein intake and net available energy uniform in the several periods, After 16 weeks of observation with the various concentrates fed at a nutritive E. B. Hart and G. C. Humphrey 447 plane of 1: 8.5 the nutritive ratio was increased to approximately 1:5 for a period of 3 weeks by the addition of casein. This was done for the purpose of noting the effect of a high protein intake on both the quantity of milk secreted and its composition. This casein addition raised the total protein intake from approxi- mately 12 per cent of the dry matter of the ration to 16 per cent. Table I illustrates the proportion of the various feeds in the ration when an animal was receiving daily approximately 50 pounds of material. TABLE I. Source and Proportion of Nutrients Used. TOG Re cdiss = crocedn oct arias 4 Bees Cues eee Oil meal. | Cottented lbs. : lbs. | lbs. lbs. Clowertaiy®,.am Nees as73, daa eee 8 8 8 8 Cornnmetagest: artistes ocean 28 28 28 28 Gornsme Whe ese ke ne 6 6 6 6 (ONCEMUTALE Ss seq ie aoe eae 4 B.0t 3 2.88 POMC UER OE 8 Nod) oer. canes Seal 4 4.63 5 ail Three Jersey cows of good milking capacity were used. Two were pure bred and one a grade. They were not with calf. The animals were milked twice daily and exercised two or three times a week. Their weight was taken weekly. The plan was to place ~ each animal on any one of the rations for a period of 4 weeks with an immediate change to one of the other rations, thus in- volving each animal in 16 to 20 weeks of observation. A feeding period of 1 week preceded the quantitative collection of urine and feces. - The urine and feces were analyzed daily for nitrogen, while a weekly analysis was made of a 7 day composite sample of milk. Our earlier observations had shown that when a nutritive ratio of 1: 8 was used (equivalent to a plane of digestible protein intake of approximately 7 per cent, or 9 to 10 per cent of total protein) and the nutrients were drawn from corn stover, corn meal, and certain protein concentrates, a positive nitrogen balance could not be maintained. The daily production of milk in these earlier records was 35 to 40 pounds. Because of these facts it was planned to use rations with a nutritive ratio of 1:8 as it was essential that THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 2 La TABLE IL. Composition of Rations. Nitro- | Total gen. Produc- c Digestible protein; tion. Weight. nutritive ratio. Gluten feed ration. lbs. | percent) gm. therms Glover Lay: soccer cco 8 1.90) 69.00} 2.75 Cornisilagesaie- = Sse ene 28 0.38) 48.00) 4.64 ; Gorn ‘meal tectNe ae... cee 6 1.65} 44.90} 5.38] 2.2 lbs. diges- Gluten tieed rg: reat: 2 cee 4 3.98} 70.20} 3.17| tible protein. Saree tt otis. ook Set ee ee 0.06) 1.00) 4.00 Totals eee eee Aer Oe 233.10} 19.94 1=8.9 Distillers’ grains ( Ajax) ration. _ Clover ihayeeee\ co. ai ore 8 1.90} 69.00} 2.75 Wornysilageer sna cesmine otis 28 0.38) 48.00] 4.64 Cornvmesl ca .2 a cei ei. ne 6 1.65) 44.90) 5.38 Distillers’ grains...............| 3.37} 4.65) 71.40] 2.66 DEATCHox Rp choe McLean and Van Slyke, J. Am. Chem. Soc., 1915, xxxvii, 1128. 483 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XX XI, NO. 3 484 Chlorides in Blood out the mixture is complete in 10 minutes if the flask is shaken occasionally. Our metaphosphorie acid, a J. T. Baker Chemical Company preparation, is free from chlorides. The method, with the suggested modification, is described in full for the sake of completeness, though the operations after re- moving the blood proteins are simply those of McLean and Van Slyke, as also are the solutions used in the titration of the pro- tein-free filtrate. We proceed as follows. Take 2 ec. of plasma (or whole blood) in a 25 ec. volumetric flask. Add about 20 cc. of water and then slowly and with stir- ring add 1 ce. of a freshly prepared 25 per cent solution of meta- phosphoric acid. Fill to the mark with water, shake well, and let stand for 10 minutes with oceasional shaking. Filter and take 10 cc. of filtrate in a 25 ce. volumetric flask, add 5 ce. of the m/29.25 silver nitrate solution, 5 ce. of 10 per cent magnesium sulfate solution (to facilitate the flocking out of the silver chloride), fill to the mark with water, shake, and let stand for 5 minutes. Filter or centrifuge the liquid and take 20 ce. of the clear solution in a small Erlenmeyer flask, add 4 cc. of the nitrite, citrate, starch solution, and titrate with the m/58.5 potassium iodide solution. The calculation is as follows: 1.25 (8 — cc. KI used) X } X 100, or, simplified, 156 (8 — ec. KI used) = mg. NaCl per 100 cc. of blood or plasma. This modification makes the admirable micro titration method of McLean and Van Slyke still available, and the procedure for removing the blood proteins is, we think, somewhat simplified. Another point in its favor is that, if chloride determinations are to be done on whole blood it may be unnecessary in some cases to use a separate portion of blood, but the titration may be done on the filtrates obtained in the determination of non-protein ni- trogen according to Folin and Denis,’ or on the filtrates obtained in blood urea determinations according to a method soon to be published from this laboratory. In this case, since 5 cc. of blood are diluted to 50 cc. the above mentioned formula for the caleu- lation does not hold true. If 10 ce. of a non-protein nitrogen or urea nitrogen filtrate were used, the formula would be: 1.25 (8 — ec. KI used) X 100 mg. NaCl per 100 cc, of blood. 2 Folin, O., and Denis, W., J. Biol. Chem.,, 1916, xxvi, 491. 4 G. L. Foster 485 Table I shows how perfectly the results by the modified pro- cedure agree with those of the McLean-Van Slyke method. Old method. New method. KI KI Sample. Normal human plasma I........... os a ES | ee ee ee ve = Wy Epa ene Pe $8 “ SS | Rei mean ee ects DlASIMa LUD... soi asm once eres: 4: FO A NOTA es tit el 8 ec ha a a eee? OG Wi Deere Si ea ere ees ° pia) i il Ieee so eae ure ease ater as ald () RRR ha ee ee Beef whole blood IV................ ? “ “ “ IV “ “ cc IV “ “ “ IV. “ce “ “e V “ec “ “ We “ “ “ce V * Non-protein nitrogen filtrates. ** Blood urea filtrates. bo bh Or ow vw Go co Seer =) i=) or (=) 24. bo ee ~I oe) NaC] per 100 ce. mg. 620 623 601 598 594 587 590 595 592 503 500 502* 506* 472 474 abt A75** I am indebted to Professor Folin for his advice and encourage- ment in this work. Fs THE FUNCTION OF MUSCULAR TISSUE IN UREA FORMATION. By RALPH HOAGLAND anp C. M. MANSFIELD. (From the Biochemic Division, Bureau of Animal Industry, United States Department of Agriculture, Washington.) (Received for publication, July 9, 1917.) INTRODUCTION. The formation of urea in the animal body is a question of fundamental importance in protein metabolism. Certain fea- tures of the problem have been worked out with some degree of satisfaction, but there is still uncertainty both as to where and how urea is formed in the body and as to the nature of the proc- esses Involved. The present status of our knowledge on the subject may be briefly summarized as follows. The end-products of protein digestion—chiefly amino-acids— are absorbed directly into the circulatory system from the walls - of the small intestines, and are then transported to the various body tissues. The body cells select such amino-acids as are needed for their repair or upbuilding, and quantities in excess of these needs are deaminized with the subsequent production of ammonium carbamate and ammonium carbonate, which are transformed into urea, and removed from the blood by the kidneys. The deaminized residue of the amino-acid molecule may then be utilized for the production of energy or it may be transformed into reserve material—glycogen or fat. In case of the break down of the protein constituents of the cell, it is prob- able that the amino-acids formed undergo changes similar to those which take place in the amino-acids absorbed from the digestive tract. The important question, so far as the purposes of this paper are concerned, is: Where and by what means is urea formed in the body? | 487 488 Muscular Tissue and Urea Formation There is abundant proof that the liver performs an important function in urea formation. Satisfactory evidence that it is the sole source, or even the most important source, of urea pro- duction in the body is lacking. On the other hand, the prevail- ing theory is that urea formation is in all probability a function of body cells in general rather than of the liver cells in particular. It seems worth while to quote briefly the views of a few leading physiological chemists on this subject. Otto von Firth! states: ‘‘Im Ganzen neigt man gegenwirtig mehr und mehr der Meinung zu, dass die Faihigkeit, Harnstoff zu bilden, nicht ein Vorrecht der Leber, vielmehr, ebenso wie die Fahigkeit, Eiweiss zu ver- brennen, eine der allgemeinen Eigenschaften lebender Zellen ist.’’ Abderhalden?'says: ‘‘Aus dem vorliegenden Beobachtungen ergibt sich der Schluss, dass die Leber Harnstoff bildet. Der Versuch am itiberleben- den Organ hat das eindeutig bewiesen. Wahrscheinlich kann er auch von allen anderen K6rperzellen erzeugt werden. Allerdings fielen bis jetzt die Versuche an anderen iiberlebenden Organen als der Leber negativ aus. Die Beobachtungen an Individuen mit pathologisch veriinderter Leber und vor allem die Tierversuche, bei denen die Leber kiinstlich schwer geschi- digt wurde, zwingen uns jedoch zu der Annahme, dass die verschiedensten Gewebe Harnstoff bilden kénnen.”’ Folin® states: ‘‘The hypothesis that the urea forming process is thus probably largely a matter of muscle metabolism predicates of course noth- ing as to the nature of the process. Inno way does it invalidate or weaken the prevailing view that deaminization rather than oxidation represents the first step in the formation of urea from amino-acid nitrogen. Ascribing to the muscles the greatest share in the urea formation represents, therefore, in no essential point a return to the earlier teachings of Pfliiger or of Voit.”’ Bayliss? has this to say concerning the formation of urea in the body: “In the second place, certain experiments tend to show that the liver has not much more power of deamination than the other cells of the organism. The experiments of Lang (1904) and of Miss Bostock (1911) have shown that tissues in vitro are capable of deaminating amino-acids to some extent, but that the process is not quite the same as that in the living organism, since 1 Von Firth, O., Probleme der physiologischen und pathologischen Chemie, Leipsic, 1913, u, 101. 2 Abderhalden, E., Lehrbuch der physiologischen Chemie, Berlin, 3rd edition, 1914, i, 586. 3 Folin, O., and Denis, W., Protein metabolism from the standpoint of blood and tissue analysis. Third paper. Further absorption experiments with especial reference to the behavior of creatine and creatinine and to the formation of urea, J. Biol. Chem., 1912, xii, 161. * Bayliss, W. M., Principles of General Physiology, London, 1915, 265. R. Hoagland and C. M. Mansfield 489 amides are more readily acted on in vitro than are amino-acids, while the contrary is the case in the organism. ‘‘We may conclude that amino-acids are supplied to the tissues, and, with the exception of that small part used for repair or growth, are deami- nated there. “The great activity of the liver in the conversion of ammonia to urea makes it probable that the main part of the ammonia from the tissues is converted into urea in that organ.”’ The basis for the theory that urea formation is a function of body cells in general rather than of liver cells in particular is, of course, founded upon the idea that the metabolism of the individ- ual cell is complete in itself. Thus, m case of muscle cells, it is well known that they contain several proteolytic enzymes which actively break down the cell proteins with the production of amino-acids and ammonia, if not urea. Hoagland, McBryde, and Powick® found that when muscular tissue of an ox was in- cubated for 100 days under aseptic conditions at 37°C. the amino-acid content of the tissue increased 741 per cent and the ammonia content 519 per cent. While the two constituents in- creased at somewhat nearly the same rate, yet at the end of the period the amino-acid nitrogen constituted 18.9 per cent of the total nitrogen of the tissue, while the ammonia nitrogen constituted only 1.6 per cent. Unfortunately no study was made of the urea content of the muscular tissue in those ex- periments. The results of the experiments indicate clearly that muscle cells have to a high degree the ability to break down their protein constituents into amino-acids, and to a lesser degree, into ammonia. The lower production of ammonia was probably due to certain limiting factors which are discussed by the authors of the above paper. These changes are in harmony with the theory that ammonia is an intermediate product in the transfor- mation of amino-acids into urea. Now since muscle cells break down their protein constituents into amino-acids and ammonia, it is reasonable to expect that those cells might transform the ammonia into urea. At any rate, that is the idea which led the writers to undertake the investigations which are to be reported in this paper. Again, it may be that the change is carried no further than ammonia by the muscle cells, and perhaps by other ® Hoagland, R., McBryde, C. N., and Powick, W. C., Changes in fresh beef during cold storage above freezing, U. S. Dept. Agric., Bull. 433, 1917. 490 Muscular Tissue and Urea Formation body cells as well, and that the ammonia is carried to the liver where it is transformed into urea. The well known marked ability of the liver to change ammonia into urea harmonizes with this view. Direct experimental evidence that urea is formed elsewhere in the body than in the liver is comparatively meager. The problem is a difficult one to solve. It necessitates either the complete removal of the liver from the circulation of an animal and a study of urea production in the body under such conditions, the difficulties of such a procedure being obvious; or it requires the identification of urea-forming enzymes in'body tissues other than the liver. There is considerable indirect evidence that urea may be formed elsewhere in the body than in the liver, but the value of such evidence is uncertain, and it will not be presented. The following direct experimental evidence supports the view that urea formation is a function of body cells in general. Matthews and Nelson® conducted experiments in which solutions of amino-acids were injected, in one case into the muscular tissue of dogs with Eck fistulas, and in the second case into dogs from which all the abdominal viscera except the kidneys and a small portion of the liver had been re- moved. In the first instance there was a slight increase in the ammonia and urea content of the urine after 2 to 24} hours. In the second instance there was always an increase in the ammonia content of the urine at the end of the Ist hour. The increase in the urea, which generally appeared from 1 to 2 hours later, was not so constant, only being manifest in about 90 per cent of the experiments. In some instances the only result obtained was a marked increase in ammonia accompanied by a decrease in urea. Fiske and Sumner’ injected solutions of amino-acids into animals in which the liver and kidneys had been removed from the circulation, and as checks, ran controls in which only the kidneys were removed from the cir- culation. The experimental period amounted to 3 hours or less. On the whole, it was found that there was an appreciable accumulation of urea both in the blood and the muscular tissue after the injection of amino-acids, and that the accumulation was practically as great when the liver was excluded from the circulation as when it was in its normal relations to the other organs. Asa result of their experiments the authors conclude that the liver is not the chief site of urea formation from amino-acids. With the exception of arginase, urea-forming enzymes have not been identified in body tissues. * Matthews, S. A., and Nelson, C. F., Metabolic changes in muscular tissue. I. The fate of amino-acid mixtures, J. Biol. Chem., 1914, xix, 229. 7 Fiske, C. H., and Sumner, J. B., The importance of the liver in urea formation, J. Biol. Chem., 1914, xviii, 285. oe R. Hoagland and C. M. Mansfield 491 EXPERIMEN TAL. The purpose of the experiments which are to be reported in this paper was to determine whether or not urea-forming enzymes are present in muscular tissue. The general plan of the experiments was to obtain sterile samples of muscular tissue by aseptic methods and incubate the tissue in sterile containers for various periods of time, urea being determined in the fresh tissue and in the mcubated samples. The general method of procedure was as follows. Methods. Fat steers were slaughtered at a local abattoir by the customary methods under the direction of the authors. The operation was carried on with as great dispatch and under as clean conditions as possible. Prior to slaughter the killing floor was washed with hot water and then with a solu- tion of mercuric chloride (1:1,000). The animal was stunned, hoisted from the floor, and the large blood vessels of the throat were severed. When bleeding was completed the carcass was dropped on to the clean killing floor and wet down with hot water. The proposed lines of incision for the removal of the hide were first scrubbed with hot water and soap and then ~ with a hot solution of mercuric chloride (1: 1,000). The knives and saws used during slaughter were immersed in hot water and then placed in a hot solution of mercuric chloride from time to time duringslaughter. One hind quarter from each animal slaughtered was selected for experimental “purposes. The hide was removed from this quarter with as great care as possible to avoid contamination of the freshly exposed tissues. During the removal of the hide, cheese-cloth that had previously been soaked in mercuric chloride solution was wrapped around the surface of the hind quarter as rapidly as the hide was removed; and finally when it had been separated from the carcass, the quarter of beef was wrapped with dry cheese- cloth and paper and immediately and rapidly transported to the laboratory. At the laboratory the outer coverings were removed from the quarter of beef, and it was placed in a special room where samples of the muscular tissue were taken for incubation and analysis. Method of Taking Samples. The room in which the samples of muscular tissue were taken was about 10 ft. square and provided with a false cheese-cloth ceiling at a height of 10 ft. A window which opened into the room was fitted with a cheese-cloth screen. Prior to taking samples the room was thoroughly washed and then sprayed with a 3 per cent solution of liquor cresolis compositus. 492 Muscular Tissue and Urea Formation The operators wore sterile gowns, head cloths, and rubber gloves while taking the samples. The bichloride cheese-cloth was first removed from the quarter of beef and the exposed surface was then wet down with a solution of bichloride of mereury and wiped dry with a sterile cloth. The proposed lines of in- cision were sterilized by means of a large steel spatula previously heated to a bright red color. Sterile knives and forceps were used in taking all sam- ples. The seared lines were first cut through to a depth of about 0.5 cm. and a second knife was used to trim back sufficient surface tissue to expose the area of muscular tissue necessary in order to get a sample of the de- sired size. A third knife was used to cut out a block of muscular tissue, which was immediately taken up by means of a pair of forceps and placed in a sterile glass covered dish; additional samples were taken in a similar manner. Every possible precaution was taken to prevent contamination of the samples taken for incubation. The dishes in which the samples were incubated are 10 cm. high and 10 em. in diameter and are provided with loosely fitting glass covers. They are of the type known as ‘‘dressing jars.’”’ As soon as the desired number gf samples had been obtained the dishes were sealed by means of strips of adhesive tape, which were then painted with melted paraffin so as to make the seal practically air-tight. The dishes were weighed empty, after the introduction of the samples of tissue, and at the end of the incubation periods. The dishes were placed in an incubator and held there at 37°C. for various periods of time. The samples of muscular tissue taken for study were secured from the following muscles: biceps femoris, semimembranosus, semitendinosus, and vastus externus. Bacteriological Examination of Samples. The incubated samples were examined from time to time and those which showed apparent evidences of bacterial contamination were discarded. Samples which appeared to be sterile were selected at intervals and sub- jected to a careful bacteriological examination to determine whether bac- teria were present or not. Chemical results are reported only on samples proven to be free from bacteria. The following procedure was employed in the bacteriological examination of the samples. The tape was first removed from the jar and the edges of the cover were flamed. Since considerable quantities of juice had exuded from the sam- ples of muscular tissue during incubation, the juice as well as the tissue was examined for the presence of bacteria. 0.5, 0.25, and 0.1 ce. portions of the juice were plated with standard agar. Portions of tissue were first taken from the surface of the sample, it was then cut in two, and similar cultures were taken from the interior, sterile knives and forceps being used. Three portions were taken from the surface and a like number from the interior. One portion from each surface was placed in peptonized beef R. Hoagland and C. M. Mansfield 493 broth for the development of aerobic bacteria and two portions were intro- duced into tubes of glucose agar for the development of anaerobic organ- isms. The tubes of agar had first been boiled to drive off traces of air and then allowed to cool to approximately 40°C. The portion of tissue was introduced into the tube and as soon as the material had settled to the bot- tom the tube was placed in ice water for the agar to harden so as to exclude air and encourage the growth of anaerobic organisms. Smears were also made on cover glasses and examined for bacteria. All cultures were in- cubated for at least 6 days, and in the absence of growth the samples of muscular tissue were accepted as being sterile. Methods of Analysis. The muscular tissue was freed as far as practicable from visible fat and connective tissue, and finely ground in a meat grinder. In case of the incubated samples, from which considerable juice had exuded, the tissue was first ground and then thoroughly mixed with the juice. Analytical work was always started im- mediately after the sample had been prepared for analysis. Urea was determined by the well known urease method. Du- plicate determinations were made on all samples, and the average result is reported. The procedure employed was as follows. 25 gm. of the ground tissue were introduced into a 200 ce. Erlenmeyer flask, and 100 cc. of absolute alcohol added. The flask was stoppered and shaken to break up lumps of tissue and the shaking was repeated at inter- vals during an extraction period of about 20 hours. The alcoholic extract was then decanted into a beaker and the tissue residue was transferred to a mortar and ground in the presence of sand. The ground tissue and alco- holic extract were returned to the flask and extraction was continued for 2 to 3 hours or longer with occasional shakings. The contents of the flask were then filtered on an asbestos filter with the aid of suction, and the flask and filter were washed with hot 95 per cent alcohol. The filtrate was trans- ferred to a beaker and evaporated to about 10 cc. on a steam bath, but never to dryness. The contents of the beaker were transferred to a 25 ec. volumetric flask and made to volume. A 10 per cent solution of a dry commercial preparation of urease con- taining the proper quantity of mixed phosphates to insure the maximum activity of the enzyme was made up just before each set of determinations was to be made. The apparatus used for the determination of urea and ammonia in the extracts of the tissue is of the same general character as that commonly used for the determination of ammonia by the Folin aeration method, and it need not be described in detail. Test-tubes 6} inches long and 1} inches 494 Muscular Tissue and Urea Formation in diameter and having a capacity of about 100 cc. were used to hold the tissue extract. 25 ec. of N/50 sulfuric acid, three drops of octyl alcohol, 7 dréps of a 0.05 per cent solution of methyl red, and 50 ec. of water were introduced into each absorption cylinder. 10 cc. of the concentrated tissue extract were introduced into each of two 100 ec. test-tubes, one portion for the determi- nation of ammonia, the other for urea and ammonia combined. To the tube containing the first portion were added 15 ce. of water and 7 drops of octyl alcohol, and to the second tube were added 2 cc. of a 10 per cent solution of urease, 7 drops of octyl alcohol, and 13 ce. of water. The tubes were connected with the apparatus and allowed to stand 45 minutes for the urease to convert the urea into ammonia, when air was drawn through the apparatus at a moderate rate for 3 to 5 minutes. Suction was then turned off, the test-tubes were disconnected, and 12 gm. of potassium carbo- nate were added to each tube containing tissue extract. The tubes were immediately connected with the apparatus and air was drawn through it, slowly at first, then as rapidly as practicable, for a total period of at least an hour. The excess of acid in the absorption cylinders was titrated against N/50 NaOH. In making calculations correction was made for the ammonia liberated from the aliquot portion of the tissue extract which had not been treated with urease. The ammonia liberated from urea by the action of urease was calculated in terms of percent- ages of urea in the original tissue. In cases where the samples of muscular tissue had lost weight during incubation, the data have been corrected for that loss. Following the procedure which has been described it was found that when known quan- tities of urea were added to samples of muscular tissue the amount present could be determined with a high degree of accuracy. Experiment 1.—A fat steer was slaughtered at a local abattoir and one hind quarter was transported to the laboratory, the procedure requiring 1 hour and 25 minutes. Samples were taken for incubation from the fol- lowing muscles: biceps femoris 11, semimembranosus 9, vastus externus 7, and semitendinosus 3, total 30. The weights of the samples ranged from 90 to 409 gm., the average weight being 226 gm. Such large sam- ples were taken because a study was being made not only of possible urea-forming enzymes but of other muscle enzymes as well. Samples were also taken from each muscle for the determination of urea in the fresh tissue. The samples taken for incubation were placed in an incu- bator at 37°C. 8 hours and 20 minutes after the death of the animal. The analyses of the fresh tissue were started 7; hours after the animal was killed. The incubated samples were examined macroscopically at the end of 24, 48, and 72 hours, and subsequently at less frequent inter- R. Hoagland and C. M. Mansfield 495 vals. Samples showing positive evidences of contamination were dis- carded. Usually, though not always, samples which showed no macro- scopic evidence of bacterial contamination after incubation for 7 days would prove to be sterile upon careful bacteriological examination. ae opi) 73) Oo (S Oo 1S) pit psa per cent|per cent 50| Biceps femoris muscle. Lhr. 10min. |0.734)0.170\0.904 52 i 8 as 3 hrs: 0.564/0.17110.735 |—18.7 53 ss cs es Are 30 “ 10.664/0.208/0.872 |— 3.5 oA 4 ne ; S GS 30 “ |0.430/0.134/0.564 |—37.6 *56 a a Ss 20 “ |0.29810.362/0.660 |—27.0 57 # se oe 1K) ae 30 “ |0.148/0.48710.585 |—35.3 58 ss a = 1 15 “ |0.172/0.312/0.484 |—46.4 61 fs ad SS 2days Ilhr. |0.097/0.464/0.561 |—37.9 64 s “ be Te 5 hrs. 0.379/0.363 |—59.8 51] Semimembranosus “ eh 10 min. |0.978)9.153}1.131 55 os oe 6hrs. 30 “ {|0.257/0.386)0.643 |—43.1 * Cultures show slow-growing anaerobic organism after 6 days. DISCUSSION OF RESULTS. The experiments will first be discussed separately, and then as a whole, in order to make clear the significance of the results obtained. Experiment 1.—In this experiment the importance of determin- ing the carbohydrate content of the muscles immediately after the death of the animal was not fully appreciated, so the first analysis was not started until 7} hours after slaughter. It does not seem worth while to pay particular attention to the changes in the glycogen or the dextrose content of the muscles, but rather to the changes in the sum of the two constituents expressed in terms 508 Glycolysis of Muscular Tissue of dextrose. These changes are shown most clearly in the column headed ‘‘change in total carbohydrates,” where the changes are expressed in percentage of the amount of carbohydrates present in the first sample of each muscle analyzed. The sample of biceps femoris muscle incubated 2 days and 3 hours shows an apparent increase in total carbohydrates amount- ing to 5.3 per cent, which is practically within the limit of experi- mental error, while the samples incubated for periods ranging from 3 days and 4 hours to 15 days and 2 hours show,decreases in total carbohydrates amounting to approximately 20 per cent. The sample of this muscle incubated 22 days and 3 hours shows an increase In carbohydrates amounting to 116 per cent. The vas- tus externus muscle shows an increase of 48.6 per cent in total carbohydrates after mcubation for 29 days and 3 hours, which confirms the increase in the carbohydrate content of the biceps femoris muscle. The sample incubated 36 days and 6 hours con- tains 11.3 per cent less carbohydrates than the first sample of this muscle analyzed. Taken as a whole, the results of this experiment show (1) samples of tissue from both the muscles examined had appreciable glycolytic properties, the decreases in total carbohydrates varying from 11.3 to 21.1 per cent; (2) samples from both muscles appar- ently had the ability to synthesize carbohydrates. Experiment 2.—The results obtained in the first experiment sug- gested the idea that in all probability the glycolytic properties of muscular tissue are most marked for a short time after death, and that these activities are brought to a close by the development of certain conditions associated with rigor mortis. For this reason samples of muscular tissue were analyzed as soon as possible after the death of the animal, in this case 70 minutes, and later at the end of 53 and 103 hours. In this experiment the glycogen was very rapidly transformed into dextrose, the change being prac- tically complete at the end of 55 hours. The changes in the total carbohydrate content of the samples of muscle during incubation are of great interest. All of the samples, which were incubated for periods ranging from 5} hours to 34 days and 3 hours, show large relative decreases in total carbo- hydrates, varying from 55.4 per cent in the sample incubated 2 days and 1 hour to 91.5 per cent in the sample incubated 27 days 1 g < R. Hoagland and C. M. Mansfield 509 and 1 hour. The sample incubated only 53 hours shows a de- crease of 81.1 per cent. By referrmg to Table II the column headed “glycogen and dextrose calculated as dextrose,” it will be noted that there is first a large decrease in the total carbohydrate content of the sam- ple incubated 5} hours; while in the sample incubated 105 hours there is a considerable increase as compared with the 53 hour sample, but still a marked decrease as compared with the first sample analyzed. This increase in the carbohydrate content of the sample incubated 10} hours, as compared with the sample incubated 53 hours, confirms a similar increase observed in Ex- periment 1. Finally the samples of biceps femoris muscle incu- bated for 27 days and 1 hour, and 34 days and 3 hours contain less carbohydrates than the sample incubated 53 hours. On the whole, the results of this experiment show that the samples of muscular tissue examined had very marked glycolytic properties. It is of particular significance that glycolysis took place so rapidly immediately after the death of the animal. The fact that there was a decrease of 81.1 per cent in the carbohydrate content of the sample of muscular tissue analyzed 53 hours after- the animal was killed, as compared with the amount present in the muscle 1 hour and 10 minutes after slaughter, mdicates the rapidity with which the change took place. Experiment 3.—The importance of analyzing the samples of tissue at frequent intervals following the death of the animal is at once apparent from an examination of the data presented in Table III. The changes in the glycogen content of the samples of the biceps femoris muscle during incubation are of particular interest. The sample analyzed 1 hour after the death of the animal contains 0.672 per cent glycogen, while the one analyzed 1 hour and 45 minutes later, or 2 hours and 45 minutes after the animal was killed, contains 0.930 per cent, or a relative increase of 38.4 per cent over the amount present in the first sample. The samples incubated for 4 hours and 15 minutes, and 6 hours and 20 minutes show gradually decreasing percentages of glycogen as compared with the sample incubated 2 hours and 45 minutes, but, on the other hand, larger percentages of glycogen than the sam- ple analyzed 1 hour after slaughter. The samples incubated 10 hours and 30 minutes, and 27 hours contain less glycogen than = 510 Glycolysis of Muscular Tissue any of the others. These facts are of much importance in that they indicate that muscular tissue has the ability to synthesize glycogen. Since there was no decrease in the dextrose content. of the muscle during the increase in glycogen, and because ,proof is lacking that glycogen can be formed from fat, it appears that the glycogen must have been formed from proteins. The total carbohydrate content of the biceps femoris muscle shows, first, a relative increase of 33 per cent in the sample incu- bated 2 hours and 45 minutes as compared with the sample in- cubated 1 hour,’and then increases of 21.4 and 12.0 per cent in the samples incubated 4 hours and 15 minutes and 6 hours and 20 minutes, respectively. The samples incubated for periods ranging from 8 hours and 30 minutes to 27 hours show decreases varying from 11.1 to 35 per cent. If the percentage changes in the total carbohydrate content of the samples of muscle be re- ferred to the amount present in the sample incubated 2 hours and 45 minutes as a basis, it will be found that the samples incubated for periods ranging from 4 hours and 15 minutes to 27 hours show the following decreases in carbohydrates: 8.7, 15.7, 33.5, 51.0, and 33 per cent. Unfortunately samples of the semimembranosus muscle were not analyzed at as frequent intervals as were the samples of the biceps femoris muscle. The samples of the semimembranosus muscle incubated for periods of 20 days and 4 hours and 28 days and 3 hours show decreases in total carbohydrates amounting to practically 60 per cent of the amount present in the sample ana- lyzed 1 hour after slaughter. Judging by the results obtained in the other experiments, it is probable that the decreases in the carbohydrate content of the muscle just noted took place within a few hours after the death of the animal. Taken as a whole, the results of this experiment indicate (1) samples of the biceps femoris muscle apparently had the ability to synthesize glycogen and at the same time to increase their content of total carbohydrates; (2) samples of the biceps femoris muscle showed marked glycolytic properties, over 50 per cent of the carbohydrates disappearing within a period of 8 hours; (8) samples of the semimembranosus muscle also had marked glyco- lytic properties. Experiment 4.—Samples from both the biceps femoris and R. Hoagland and C. M. Mansfield Silt semimembranosus muscles showed appreciable glycolytic proper- ties. The decreases in the total carbohydrate content of the muscles did not proceed at a constant rate, the changes taking place most rapidly during the first 12 hours after the death of the animal. The total carbohydrate content of the biceps femoris muscle had decreased 18.7 per cent 3 hours after the death of the animal, 46.4 per cent after 12 hours and 15 minutes, and 59.8 per cent after 7 days and 5 hours. The semimembranosus muscle decreased 43.1 per cent in total carbohydrates in a period of 6 hours and 380 minutes following slaughter. Taken as a whole, the results of this experiment show that the samples of muscular tissue examined had appreciable glycolytic properties, the disappearance of sugar taking place most rapidly within 12 hours after the death of the animal. These results con- firm those obtained in the previous three experiments. SUMMARY OF RESULTS. The results of the experiments which have been reported show that muscular tissue from a normal ox has appreciable glycolytic properties, and that these properties are most active for a short. time following the death of the animal. The factors which limit the glycolytic changes which take place in dead muscular tissue under such conditions as prevailed in these experiments have not been determined as yet, but in all probability the changes are limited by conditions associated with the development of rigor mortis. Muscular tissue also appeared to have the ability to synthesize carbohydrates, both glycogen and dextrose, presumably from proteins. These findings show that appreciable glycolysis may take place in muscular tissue without the aid of any extract from pancreatic tissue. These results do not, on the other hand, indicate that the pancreas does not play an important part in the glycolysis that takes place in living muscular tissue. In the experiments which have been reported the animals were bled as completely as pos- sible, but the tissues were not washed free from blood, since this procedure did not seem practicable under the conditions of the experiments. ‘It is barely possible, but highly improbable, that the small amount of blood which remained in the muscular tissue 512 Glycolysis of Muscular Tissue may have supplied an activator or coenzyme previously secreted by the pancreas, which stimulated glycolysis. Rather, the au- thors are of the opinion that the pancreas functions somewhat differently in sugar metabolism. They are of the opinion that in all probability the pancreas plays an important part in the actual formation of the glycolytic enzymes in living tissues, but that active glycolysis may take place in dead tissues in the ab- sence of any extract of the pancreas. The possibility that the decrease in the copper-reducing action of the clarified extracts of the incubated samples of muscular tis- sue might be due to the formation of disaccharides, as suggested by Levene and Meyer (29), has been considered. In a considerable number of cases the copper-reducing action of clarified muscle extracts was determined both before and after inversion, but in no case was there any evidence of the formation of disaccharides. Attention has been called to the somewhat irregular character of the changes in the carbohydrate content of the muscular tissue during autolysis. In a small degree such irregularities may be due to unavoidable errors, but for the most part they must be due to other causes. Considering the complex nature of the metabolism of sugar in the living organism, it is reasonable to expect that the changes which take place in the carbohydrates of dead muscle may be equally or even more complex. In the nor- mal living organism the various forces work in harmony so as to dispose of the absorbed sugars according to the needs of the body; in the dead tissues the equilibrium of forces has been broken up and other factors enter into play, many of which are not clearly understood, so that the changes which take place under such conditions cannot compare either in orderliness or in extent with those that take place in the living organism. Glycolysis has been measured in the experiments which have been reported by the disappearance of carbohydrates, The ques- tions naturally arise: What has become of the carbohydrates? Have they been oxidized completely to carbon dioxide and water, as in the living organism, or has the oxidation been incomplete? The work of Stoklasa suggests that the sugar present in the muscu- lar tissue may have been oxidized only to carbon dioxide and alco- hol. In order to secure information on this subject the following experiments were conducted. R. Hoagland and C. M. Mansfield 513 I]. The purpose of the experiments which are to be reported was to determine whether the glycolysis that takes place in muscular tissue results in the production of carbon dioxide and alcohol. The general plan of these experiments was to secure sterile samples of muscular tissue by aseptic methods as soon as possible after the death of the animal, and to incubate the tissue in sterile media in suitable fermentation tubes so as to collect any gas that might be formed. Special large size fermentation tubes, of the so called ‘fish hook” type, having a capacity of 360 cc., were used. Samples of muscular tissue were taken in the special room previ- ously described. Every precaution was taken to avoid contami- nation of the samples of tissue. HN .CH,.CH,.CH2.CHNH;. COOH. NH; NH; C— NH ——_—_> © — Nid HN .CH,.CH,.CH,.COOH. HN .CH,.COOH Therefore it is important to determine whether the animal organism possesses the power to convert glycocyamine into crea- tine. Considerable work has been done on this problem but the question has not been conclusively decided. Czernicki,! using the zine chloride precipitation method, obtained an increase of creatinine in the urine of rabbits after feeding glycocyamine but not when glycocyamidine was fed. This author does not regard his work as decisive. Jaffé? also determined the sum of creatine and creatinine in the urine and muscle of rabbits after feeding and injecting glycocyamine. He also employed the Neubauer zinc chloride method. This author found that glycocyamine, when * The experimental data are taken from a dissertation submitted by Harry M. Hines as a partial requirement for the degree of Master of Science, State University of Iowa, Iowa City. 1 Czernicki, W., Z. physiol. Chem., 1905, xliv, 294. | 2 Jaffé, M., Z. physiol. Chem., 1906, xviii, 430. 549 550 Origin of Creatine. II injected or fed, led to an increase in the creatine and creatinine fraction both in muscle and urine. Dorner,’ a pupil of Jaffé, repeated the latter’s experiments but employed the more accurate colorimetric method. He found an increased excretion of crea- tine after the administration of glyecocyamine, but also found that muscle zn vitro possessed the power to methylate glycocyamine. Mellanby! criticizes Jaffé’s technique and Dorner’s experiments. He finds no consistent increase in the creatine content of chicken muscle after the administration. of glycocyamine, but regards his work as inconclusive. More recently Palladin and Wallenburger® have apparently obtained relatively large increases in creatine when rabbit’s muscle was permitted to act on glycocyamine at incubator temperature, and also when rabbits were injected with glycocyamine. Our work on thissproblem may be divided into three parts. At first it was necessary to see whether creatine could be accurately determined in the presence of glycocyamine according to the conventional methods. We found that an acid solution suffi- ciently strong to convert creatine into creatinine would usually effect a partial conversion of glycocyamine into glycocyamidine, and as the latter gives the picric acid reaction, the results for creatine were too high. Both Jaffé and Dorner separated creatine and creatinine from glycocyamine by exhaustive extraction with aleohol. We find that 95 per cent alcohol will extract sufficient quantities of glycocyamine to interfere with the creatine deter- minations. The solubility of glycocyamine in 95 per cent aicohol is 0.02 per cent, that is 20 mg. dissolve in 100 ce. of boiling 95 per cent alcohol. The second part of our work was concerned with the action of freshly hashed muscle on glycocyamine in vitro. Under these conditions, an increase of creatine was never encountered. Finally we resorted to injection, perfusion, and feeding experi- ments on rabbits, dogs, and men. While the results of these experiments were not uniform, we obtained, in some cases, an increased elimination of creatine after the injection of glycocy- amine. 3 Dorner, G., Z. physiol. Chem., 1907, li, 225. 4Mellanby, E., J. Physiol., 1908, xxxvi, 447. 5 Palladin, A., and Wallenburger, L., Compt. rend. Soc. biol., 1915, Ixxviii, 111. This reference was only obtainable in abstract form. L. Baumann and H. M. Hines 551 EXPERIMENTAL. The glycocyamine was obtained through the interaction of guanidine and chloroacetic acid according to the procedure of Ramsay.’ The yield was 65 per cent, based on the amount of chloroacetic acid used. The guanidine required for this reaction was synthesized from calcium cyanamide, according to the excellent method of Levene and Senior.7 Determination of Creatine in the Presence of Glycocyamine. Palladin and Wallenburger determined the creatine in muscle by heating the extract on a water bath for 3 hours in an approxi- mately 0.6 N hydrochloric solution as suggested by Riesser.’ Experiment A.—100 ce. of a 1 per cent solution of glycocyamine in ap- proximately 0.6 N hydrochloric acid were heated on a boiling water bath for 3 hours, then diluted to 200 ec. To one of two flasks, each containing 5 mg. of creatinine, were added 5 ce. of the glycocyamine solution (25 mg. of glycocyamine), and the intensity of the color was compared, after the addition of picric acid and alkali. The color of the glycocyamine flask was deeper than the control. This difference was equivalent to 0.8 mg. ~ of creatinine for 25 mg. of glycocyamine or 3.2 for 100 mg. Experiment B.—To two of four flasks, each containing a known quantity of creatine dissolved in 275 cc. of approximately 0.6 N sulfuric acid were added 100 mg. of glycocyamine. After heating on the boiling steam bath for 3 hours, the intensity of the Jaffé reaction was determined by means of the colorimeter and expressed as creatinine. With glycocyamine 38.97 and 39.36 mg., mean 39.16 Without ok SOLS) Pe aDEN Pe aiagets) Difference, for 100 mg. of glycocyamine, 3.31 mg. Experiment C.—A similar experiment in which 0.66 N sulfuric acid was used, showed an increase corresponding to 3.9 mg. of creatinine per 100 mg. of glycocyamine. Experiment D.—A fourth experiment in which diluted muscle extract and 0.66 N sulfuric acid was used showed that 100 mg. of glycocyamine increased the color proportionately to 4 mg. of creatinine. Solubility of Glycocyamine in Boiling 95 per Cent Alcohol.— An excess of analytically pure glycocyamine was boiled with §5 per cent alcohol under reflux for 1 hour, then filtered through a hot steam jacketed funnel into a tared flask of known volume. ° Ramsay, H., Ber. chem. Ges., 1908, xli, 4385. 7 Levene, P. A., and Senior, J. K., J. Biol. Chem., 1916, xxv, 628. 8 Riesser, O., Z. physiol. Chem., 1913, Ixxxvi, 415. 52 Origin of Creatine. II After evaporation of the alcohol on the steam bath, the flask was dried to constant weight and weighed. It was found that 57 ce. of boiling 95 per cent of alcohol dissolved 0.0115 gm. of glycoey- amine or 100 ee. 0.020 gm. Weyl’s Reaction—Both Jaffé and Dorner placed considerable importance on Weyl’s test as a means of determining glycocyami- dine in the presence of creatinine. We were unable to obtain sharp reactions in the concentrations with which we were dealing, so we preferred to control our experiments as described below. Experiments in Vitro.—F¥reshly hashed rabbit muscle or dog liver was weighed into a flask contaming glycocyamine dissolved te a Glycocy- ; ‘ Glycocy- Weight of | Test | amine | Creatine. | Waahict | geet | amine, | Creatine. gm. | hrs. mg. per cent gm. hrs. mg. per cent Experiment E. Control. E | 7.293 24 40 . 0.577 7.529 24 40 0.546 7.724 24 40 0.561 8.672 24 40 0.564 7.996 48 33 0.596 en! 48 40 0.596 6.426 48 40 0.595 7.061 48 40 0.600 (BBY 48 40 0.563 Averages. co...) ..eeee eee. WnorS AVCTABO =. he. pt sea 0.576 Experiment F. : Control. 7.146 24 50 0.461 7.081 24 50 0.470 6.784 46 50 0.470 6.093 46 50 0.471 Average: >. )2Jnaeeee aire 0.465 Average. ot 51 24.528 hee 0.470 Experiment G. ‘Control. Weight of | Weight of liver. liver. 6.370 | 36 | 50 0.0567 | 6.129 36 50 0.0587 9.039 “| 42 50 0.0444 | 8.384 42 | 50 0.0520 7.147 | 63 | 50 0.058 8.105 106 50 0.056 1 AVOCTSEC. occ oe cot ote 02053 AVeLAGE:....... /p...+ ae nUSe * Glycocyamine added at the time of the determination, ta «s. — . L. Baumann and H. M. Hines 553 in isotonic Henderson’s phosphate solution. After the addition of toluene, the flask was placed in the incubator for from 24 to 96 hours and the creatine content determined according to the method of Janney and Blatherwick.® The absence of bacterial growth was confirmed by culture. The controls which had re- mained in the incubator were analyzed after the addition of an equivalent amount of glycocyamine. In this way the color due to glycocyamine was accounted for. As is evident from the tables, our results do not indicate methylation of glycocyamine in vitro by muscle or liver. Injection and Feeding Experiments. Rabbits were placed in a metabolism cage on a diet of carrots. The urine was gathered in 24 hour periods and diluted to a defi- Experiment H.—A gray rabbit weighing 2,540 gm. was used for this experiment. At the conclusion of the experiment, the muscle contained 0.452 per cent of creatine, 72.55 per cent of water, and 3.71 per cent of nitrogen. Experiment H. ; Glycocy- Glycoey- Day. |Creatinine.|Creatine.*| | amine Day. | Creatinine./Creatine.*| amine injected. injected. mg. mg. mg. mg mg mg. 2 84 2 0 12 97 0 120 3 120 6 0 13 100 11 400 4 93 10 0 14 99 39 350 5 77 2 0 15 102 44 350 6 110 2 0 16 108 54 350, 7 62 22 0 17 93 62 350 8 83 9 0 18 102 61 100 9 86 1 0 19 99 54 200 10 112 1 0 20 100 51 0 11 91 0 > 0 21 112 40 0 22 106 15 0 Average... 91.8 5.5 Average.101.6 39.2 23 107 2 0 24 106 4 0 * Creatine is expressed as creatinine. 9 Janney, N. W., and Blatherwick, N. R., J. Biol. Chem., 1915, xxi, 567. 554 Origin of Creatine. I] nite volume. During the second period, they were injected with known amounts of glycocyamine subcutaneously. Creatine was determined by heating the urine on the water bath with 0.66 N sulfuric acid for 3 hours. A quantity of glycocyamine greater than that which could be present in the urine of the injected animals was added to nor- mal rabbit urine and its effect on the creatine estimation de- termined. It was found that this quantity of glycocyamine (400 mg.) increased the daily creatine from 10 to 12 mg. If we allow 12 mg. for the effect of dissolved glycocyamine in the urine and 6 mg. for the average daily excretion during the preliminary period, we still have an excess of more than 21 mg. to be ascribed to the influence of glycocyamine. Experiment I.—A female pregnant rabbit!® was placed in a cage on a diet of carrots. At the beginning of the experiment it weighed 2,435, at the end 2,564 gm. The creatine content of its muscle at the conclusion of the experiment was 0.448, the water content 74.22, and the nitrogen 3.58 per cent. Experiment I. Glycocy- Glycoey- Day. |Creatinine./Creatine.*} amine Day. | Creatinine.jCreatine.*} amine injected. injected. mg. mg. mg. mg. mg. mg. ] 60 6 0 10 112 4 300 Dee 106 5 0 11 92 5 150 “3 88 11 0 12 76 37 500 4 84 3 0 13 116 39 300 5 $4 7 0 14 71 30 300 6 124 12 0 15 111 30 300 (4 7 1 0 16 74 17 0 8 11 2 0 9 12 0 0 Average. ..89 5 Average... 93 24 17 103 0 20 * Creatine expressed as creatinine. 10 The pregnant condition of this animal was determined when the experiment was in progress. L. Baumann and H. M. Hines 555 If we allow 12 mg. for the glycocyamine effect and 5 mg. for the daily average excretion during the preliminary period, 7 mg. remain which we are inclined to ascribe to the effect of the glycocyamine. We are unable to account for the low concentration of creatine in the muscle tissue. Experiment J.—The subject of this experiment was a male rabbit weighing 2,174 gm. at the beginning and 2,155 gm. at the end of the experi- ment. The diet consisted of carrots. Experiment J. Glycocy- Glycocy- Day. Creatinine.|Creatine.*| amine Day. |Creatinine./Creatine.*} amine injected. injected. mg. mg. mg. mg. mq. mg. 1 74 f 0 9 105 10 150 2 78 0 0 10 68 vi 300 3 oT 4 0 li 96 4 150 4 50 6 0 12 85 25 300 5) 59 1 0 15 Lost. 300 6 100 8 0 14 71 bl 300 2 63 3 0 15 87 16 300 8 116 1 0 16 105 19 0 17 | 3) 25 0 Average... 74 3.4 Average. . 83 if 18 90 0 0 * Creatine is expressed as creatinine. After making the necessary deductions it is found that the excretion of creatine is increased during the second period by 1.6 mg. per day. Experiment K.—The subject of this experiment was a bitch weighing 12 kilos. She was. placed in a metabolism cage and fed on a mixture of skimmed milk powder, bread, lard, and agar. The experiment was started after the dog had been on this diet for 4 months. The urine was collected in 24 hour periods and made to a definite volume with the wash water from the cage. At the close of the experiment the dog weighed 15 kilos. 17 days later a small piece of the vastus externus muscle was removed from a right hind leg under anesthesia. It contained 0.35 per cent creatine, 556 Origin of Creatine. II] 73.71 per cent water, and 3.68 per cent nitrogen. 2 months later the cor- responding muscle from the opposite side contained 0.318 per cent of crea- 29 tine, 73.39 per cent of water, and 3.52 per cent of nitrogen.!! Experiment K. Glyeocy- Glycocy- Day. | Creatinine.|Creatine.*| amine in- Day. | Creatinine.|Creatine.*) amine in- jected. jected. mg. mg. mg. mg. mg. mg. 1 386 1 0 pel) 362 14 400 2 370 15 0 11 345 64 600 3 382 21 0 12 396 61 830 4 380 23 0 13 398 60 800 5 378 12 0 14 380 90 800 6 345 3 0 15 365 122 800 7 393 29 0 5 364 14 0 9 364 14 Average 373 14.5 Average| 374 68.5 * Creatine is expressed as creatinine. In previous experiments we found that 100 mg. of glycocyamine when heated with 0.66 N acid for 3 hours reduces alkaline picric acid proportionately to 3 mg. of creatinine. Assuming that all of the glycocyamine which was injected appeared in the urime we would deduct 24 mg. for the glycocyamine effect and 14.5 mg. for the daily average creatine elimination during the preliminary period. This leaves a daily average elimination of 30 mg. to be ascribed to the effect of glycocyamine injection. Experiment L.—A man weighing 62 kilos, aged 23 years, and in perfect health, was placed on a practically creatine-free diet consisting of bread, milk, cereals, vegetables, and fruit. The urine was carefully collected in 24 hour periods. Creatine was determined by heating the urine on the water bath for 3 hours with n sulfuric acid. Glycocyamine was adminis- tered per os. 41 In previous unpublished experiments we have found that the creatine * content of dog muscle is markedly diminished after the animals have been kept in cages on the milk and bread diet for several months. L. Baumann and H. M. Hines 5bG Experiment L. Day. Creatinine.|Creatine.* Rees Day. Creatinine.|Creatine.* Evecey, mg. mg. mg. mg. mg. mg. 1 1,397 26 None. 10 1,358 81 1,000 2 ar6 Alo; ne 11 1,362 24. 2,000 3 1,378 23 Hs 12 1,342 33 2,000 4 1,398 27 os 13 1,282 83 2,000 5 1,320 15 MG 14 1,342 53 3,000 6 1,214 0 i a 1,386 13 te 8 1,367 ay ce 9 1,256 0 “f Average| 1,348 19.5 Average 1,337 55 * Creatine is expressed as creatinine. As considerable glycocyamine was undoubtedly present in the urine, the increased excretion of creatine during the second period is more apparent than real and is probably due to the presence of the guanidine acetic acid. Experiment M.—A brown bull bitch weighing 12.3 kilos was fed on a diet consisting of skimmed milk powder, dry bread, lard, and agar, for 4 months prior to the operation. On March 7 a piece of muscle was removed from the right hind leg under aseptic precautions, and ether anesthesia. On March 14 the animal was given 200 mg. of glycocyamine twice a day and this was repeated for the next 8 days. On March 28 a corresponding piece of muscle was removed from the opposite side. The specimens of muscle were ex- amined for creatine, water, and nitrogen. . Experiment M. Creatine . Creatine. | per gm. of Water. Nitrogen. dry muscle per cent mg. per cent per cent Before glycocyamine............... 0.30 9.2 67.31 3.72 After SF Dot! sO E 5 eet 0.31 EO 71.84 3.44 We are unable to account for the difference in water content in the two samples of muscle, as the weight of the dog remained almost constant throughout the experiment. The increase of creatine is not convincing as we have no means of saying how much glycocyamine has been stored in the muscle tissue. 558 Origin of Creatine. II Experiment N.—A small dog, 9 months of age, weighing 4.5 kilos, was decerebrated under anesthesia and a piece of the vastus.externus muscle was removed from one hind leg. Approximately 2 gm. of glycoeyamine dissolved in isotonic acid phosphate solution were injected into the jugular vein. 53 hours later a similar piece of muscle was removed from the oppo- site side. The creatine and water content of the muscle before the in- jection were 0.305 and 72.46 per cent, respectively, and after the injection, 0.316 and 71.47 per cent, respectively. In terms of dry substance, the creatine content of both samples is identical. Experiment O.—The right hind leg of a dog weighing 15 kilos was per- fused for 1 hour with 1.7 gm. of glycocyamine dissolved in 270 cc. of per- fusion fluid. This consisted of 1.7 gm. of glycocyamine dissolved in 100 ce. of water, containing 3.28 ec.-of 5 N hydrochloric acid. This was rapidly neutralized with an equivalent amount of sodium hydroxide solution and added to 150 ce. of the dog’s own defibrinated blood. The pressure of the perfusate during the experiment fluctuated between 80 and 100 mm. of mercury. Only 110 ec. of fluid were recovered. The hamstring muscles were removed from both hind legs at the close of the experiment, and creatine and water determined. Creatine was determined according to the method of Janney and Blatherwick.? The perfusion technique has al- ready been described.!? Owing to its slight solubility in water, glycocy- amine is difficult to perfuse in concentrations greater than that employed in this experiment. Experiment N. Creatine per Creatine. em. of dry Water. muscle. per ceat mg. per cent i] Perfusedmiuscles esse eos ee eee 0.382 14.8 14.2 UWnperfused> “Sau 2e seo Ses cet ee ec 0.374 13.4 72.12 The muscle of the perfused limb weighed exactly 1 kilo. If we assume that all of the glycocyamine contained in the perfusate was retained by the muscle tissue then 25.8 gm. of dry muscle tissue should contain 170 mg. of glycocyamine in addition to the creatine. If we deduct the chromogenic equivalent of this amount of glycocyamine,!* which is 11 mg., from the apparent 2 Baumann, L., and Marker, J., J. Biol. Chem., 1915, xxii, 49. ? Janney and Blatherwick use 1 N acid for the conversion of creatine into creatinine. When 100 mg. of glycocyamine are heated with this concentration of acid for 3 hours, and alkaline picrie acid is added to the neutralized solution a color develops which is equivalent to 5.5 mg. of creatinine. mt e. L. Baumann and H. M. Hines 559 creatine value for the perfused muscle, we find that 25.8 gm. of dry muscle contain 371 mg. of creatine or 14.4 mg. per gm. This increase is worthy of note. We hope to have opportunity to repeat this experiment. CONCLUSIONS. 1. Our experiments do not offer any evidence for the methyla- tion of glycocyamine by muscle or liver tissue in vitro. 2. The injection of glycocyamine into rabbits and dogs may be followed by an increased excretion of creatine. a THE PRODUCTION OF CREATINURIA IN NORMAL ADULTS. By W. DENIS anp A. S. MINOT. (From the Chemical Laboratory of the Massachusetts General Hospital, Boston.) (Received for publication, July 28, 1917.) As a result of numerous investigations made in the thirteen years which have elapsed since the publication of Folin’s method for the determination of creatine and creatinine, it is now a gen- erally accepted fact that creatine does not occur in the urine of normal men living on a creatine-free diet. Until a few years ago it was believed that this statement applied also to normal non-pregnant women. In 1911 it was however pointed out by Krause! that the urime of normal women frequently contains small amounts of creatine. This creatinuria is, according to this investigator, associated with the sexual cycle, is always present after menstruation, and while in some individuals it may disappear during the intermenstrual period in others it persists. In work recently published from this laboratory,’ it has been shown that in children it is possible to increase or decrease the amount of creatine excreted in the urine by increasing or decreasing the quantity of protein (creatine-free) in the food. As a result of these experiments the suggestion was made that creatinuria in normal children is probably due to the high level of their protein intake. If this theory be true, it would seem possible that if the protein intake could be increased to a sufficiently high level, creatinuria in normal adults might also be obtained. In order to test out this hypothesis we have carried our experi- ments on four normal adults, two men and two women, along the same general lines and by the same methods used in investigating this subject in connection with children. 1 Krause, R. A., Quart. J. Exp. Physiol., 1911, iv, 293. ? Denis, W., and Kramer, J. G., J. Biol. Chem., 1917, xxx, 189. 561 562 Creatinuria Our first experiments were carried out with two normal women who had previously acted on several occasions as subjects for metabolism studies. Just previous to becoming subjects for this investigation these women had, in connection with another prob- lem, been living for about 8 days on a creatine-free diet of low protein content. The urimes passed on the last day of this period were examined for creatine but with negative results. Both sub- jects were then placed on a creatine-free diet containing the largest amount of protein they could take. The diets used, which were the same for both subjects, were as follows: High Protein Diets. 1 2 3 OSGeo sae A Es 12 6 12 Gelatin, gine as sc S55 ae ee 50 50 Cheese (neufchatel), gm. ....... 200 50 200 IME SCO iter et cate CO eae 500 500 200 Breads igi: ss2 feek.c<0 cee ae se 300 400 SUE LEe uae a Sea een 50 50 DUttCIn tae. eta eee ce 50 50 hemonste IFS ek eee ae 2 Oranve: 6s eiec tee cee 1 Tomato and lettuce salad....... One portion (not weighed). Low Protein Diet. Bread, Gm. 324 ase 50 Baked apple, gm...200 Cream, 40 per cent, Corn meal, gm..... 75 Apple and celery CC: 1...) tee 200 Potato, gis ae 00 salad. Qitjss. ee 200 1 orange. Lactose, gm........100 Butter, gm......... 50 1 banana. Bacon fat, gm.:.... 100. ~=—-1: lemon. The low protein diet was eaten by Subject I in the form given above; in the case of Subject II the corn meal was replaced by 200 gm. of potato. From the results presented in Table I it will be seen that it was possible in the case of these two women to induce creatinuria by high protein feeding. After 5 days on the high protein Diet 1 Subject I was allowed to return for 3 days to the food which she ordinarily consumes, which consists largely of bread, vegetables, — i >| ; W. Denis and A. 8. Minot 563 and fruit, with a small amount of meat (approximately 50 gm.) once a day. Even this diet was apparently sufficiently low in protein to cause an immediate disappearance of all creatine TABLE I. I. Normal female, 38 years old, II. Normal female, 22 years old, weight 91.0 kg. weight 53.6 ke. eo : ES a Diet. is ie ES A Diet. ag | 2° 3 25 | ee 3 © bo 2.5 z © oo 2s 2 a AY oe) HH Ay ©) May gm gm gm gm gm. gm 14-15 5.02) 1.12 | 0 15-16 etoe LEN (0) 8.46] 1.12 | 0 1 16-17 12.6 = = 1 16.87} 1.13 | 0 1 17-18 LOE 76 lal | ORT2 1 24°58) 1.25 | 0.13 1 18-19 16.27} 1.20 | 0.09 1 20.06) 1.22 | 0.37 1 19-20 20.35) 1.33 | 0.16 1 24.43) 1.21 | 0.35 1 20-21 20.00) 1.46 | 0.21 il 19.60) 1.19 | 0.22 1 21-22 12.28! 1.21:| 0 ] 12.66] 1.28 | 0.24 | Mixed.* 22-23 7.98) 1.18 | 0 Mixed.* 18.45) 1.12 | 0.18 1 23-24 9.80) 1.13 | 0 J 21.88) 1.16-| 0.138 1 24—25 TO e210 2 20-59) 1185-0. 15 1 25-26 15.41] 1.23 | 0 2 Ale 23) U2 Oss 1 26-27 1225) ES! 0210 2 22160) Te tS) || Ooh? 1 27-28 11.94) 1.10 | 0.12 2 = = = if 28-29 14.97] 1.19 | 0.13 2 17.22) 1.05 | 0.14 Low. 29-30 8.25] 1.14 | 0.10 Low. 7399) W510) | (0-10 ad 30-31 6.66) 1.21 | 0 oe 6.58} 1.10 | 0.10 ‘ 31—June 1 6225) L17_|| 0 e 7.75| 1.08 | 0.09 ve June 1-2 5.48} 1.11 | 0 oe 4.65} 1.00 | 0 * 2-3 5.90) 1.11 | 0 ce o. COlel O70 * 3-4 OOS elle 0 “ 5.88} 1.16 | 0 Ss 4-5 14.61) 1.38 | 0.08 3 12.95) 1.19 | 0.09 3 5-6 18.69) 1.42 | 0.15 3 TF .26|5 W138") 20228 3 6-7 21.08} 1.26 | 0.20 3 23 .85| 1.09 | 0.3¢ 3 7-8 20.50} 1.28 | 0.22 3) 5 22:22) 1.33 | 0.20 3 * Ordinary mixed diet with meat, once a day. from the urine. An attempt was then made by feeding the high protein Diet 2 to obtain an idea as to the general level at which creatinuria could be induced but even on this moderate protein intake creatine again appeared. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL, XXXI, NO. 3 564 Creatinuria A low protein diet was then consumed for 5 days with the resultant disappearance of all creatine after the 2nd day, and this was in turn followed by a 5 day period on a diet containing approximately the same amount of protein as that contained in the high protein Diet 1, but differing from Diet 1 in that it con- tained only an extremely small amount (probably not more than 2 or 3 gm.) of carbohydrate. By this experiment it was believed that it might be possible to obtain evidence as to the effect, if any, produced by the ingestion of carbohydrate on creatine excretion. As will be seen creatine appeared on the 2nd day as was the case in the period in which the high protein Diet 1 (which contained 300 gm. of bread) was taken and remained at about the same level as that reached during the first period.® An attempt was made to follow this period of high protem and low carbohydrate feeding by one containing the same amount of protein but with a much larger amount of carbohydrate than was given in the first period. This, however, proved to be impossible, as the large amount of carbohydrate consumed (300 gm. of lac- tose-and 300 gm. of bread) produced in both subjects such marked digestive disturbances that they were compelled to abandon the experiment. The experimental periods on Subject II were essentially similar to those just described for Subject I except that the high protein Diet 1 was continued for a much longer period than in the case of Subject I, and that no period of moderate protein feeding was introduced. Having been successful in the production of creatinuria in women by forced protein feeding we next attempted to duplicate our results in men. Our third subject was a man 25 years old, weighing 62.6 kg. For a period of 14 days he consumed the following ration daily. Whole ‘egggh. een eo eee 6 Bread, GIN: vio. NEN: ae 200 Hee whites i. = es tee 4 Macaroni,“ 2.....0).2. ssa 50 Gelatin-gi7. sche ee 50 ~=Potato, Seba core ab pet a Cheese (American), gm......... 100 Sugar, (Oe sie « eee 40 Mille: cé: % lea oo eee 100 Butter, i eae } 0 -, 40 3 During this period the urine was examined daily for 2cetone bodies by means of the Scott-Wilson reagent. In the case of Subject I a minute W. Denis and A. 8. Minot 565 During this period the urine was frequently examined for crea- tine but invariably with negative results. The mixed urines for the period were found to contain 251.1 gm. of nitrogen, or an average excretion of 18.0 gm. per day. An attempt to increase the protein intake was unsuccessful. For our next experiment we used Subject IV, a man 20 years old, weighing 57.7 kg., who seemed possessed of a more capacious appetite than Subject III. For 12 days this man was kept on the same rations as had been used for Subject III. The mixed urines for the period were found to contain 227.9 gm. of nitrogen, or an average daily excretion of 19.1 gm. Frequent attempts to find creatine in these urines were, however, invariably rewarded with negative results. It was therefore decided to increase the protein intake; and for 5 days the ration consisted of: LO LSPIS | Shs Berra seer racic aan oe a0!” «Chocolates gn...2..5...52-- mote. Ol) NMI EGOta ys an ee ain eal 00> Bread ged Ai fro 150 Gelatin Gite tices ease eK 50 Sugar aren RL 5 areata 50 The determination of the urmary nitrogen gave the following - results for these 5 days: 22.9, 19.4, 32.7, 34.5, and 33.5 gm. In spite of the high level of nitrogen consumption no creatine could be detected in the urine at any time during the 5 days during which the experiment was continued. In the present state of our knowledge of creatine metabolism the above results are difficult of interpretation, The hypothesis that the ease with which creatinuria can be induced in normal women may be due to their relative lack of muscular develop- ment (as compared with that of men) immediately presents itself, but the results recorded above cannot be interpreted as furnishing evidence in favor of such a theory. SUMMARY. In two normal women it was found possible to cause creatine excretion by feeding a high protein (creatine-free) diet, and to trace of diacetic acid was found on the 3rd day and a larger amount (about 10 mg. calculated as acetone) on the 4th day. In the case of Subject II traces of acetone bodies appeared only on the 4th day. 566 Creatinuria cause the creatinuria so produced to disappear by the consumption of a low protein diet. In two men to whom a similar experimental procedure was applied we were unable to produce creatine excre- tion even when a sufficient amount of protein was consumed to cause the urinary nitrogen to rise to 34.5 gm. per day. EXPERIMENTAL STUDIES ON GROWTH. IX. THE INFLUENCE OF TETHELIN UPON THE EARLY GROWTH OF THE WHITE MOUSE. By T. BRAILSFORD ROBERTSON anp M. DELPRAT. (From the Department of Biochemistry and Pharmacology, Rudolph Spreckels Physiological Laboratory, University of California, Berkeley.) (Received for publication, July 19, 1917.) Object of the Experiments. Previous investigations have shown! that the administration of tethelin, a lipoid extracted from the anterior lobe of the pitui- tary body,” leads to marked retardation of the growth of the white mouse during the first 10 weeks of the third or ‘“‘adolescent” growth cycle, a retardation which is succeeded by pronounced acceleration. The effect of these phenomena is to completely distort the form of the curve of growth.in the third growth cycle, giving the appearance of great prolongation and enlargement of the second growth cycle (which normally terminates at 5 weeks,’) and an acceleration and curtailment of the third growth cycle. In the experiments referred to, the administration of tethelin was initiated when the animals were 5 weeks old and the second growth cycle had already begun to merge into the early stages of the third. The observed effects are therefore to be interpreted as effects of tethelin upon the third or final cycle of growth during which maturity of the sexual organs is attained, together with possible effects upon the residual growth still attainable by a continuation of the second growth cycle. It appeared of con- siderable importance to supplement these results by ascertaining the effects of tethelin upon the first (infantile) and second growth 1 Robertson, R. B., J. Biol. Chem., 1916, xxiv, 397. Robertson, T. B., and Burnett, T..C., J. Exp. Med., 1915, xxi, 280. 2 Robertson, J. Biol. Chem., 1916, xxiv, 409; Endocrinology, 1917, i, 24. 3 Robertson, J. Biol. Chem., 1916, xxiv, 363. 567 568 Studies on Growth. IX cycles of the white mouse. The following experiments were accordingly undertaken. Methods. The method employed was similar to that utilized by Robert- son and Cutler‘ in determining the effects of lecithin and choles- terol upon the growth of suckling mice. Since, during the first 3 weeks of extra-uterine development, mice are dependent upon the mother for their nutrition, administration of substances directly to the young by. mouth is attended with difficulties, while the manipulations incident to hypodermic or intraperitoneal admin- istration might very conceivably exert an effect upon the welfare and thus indirectly upon the growth of the animals which might mask or distort the effects which are the object of inquiry. We have therefore sought to influence the growth of the young dur- ing the period of lactation by administration of the tethelin by mouth to the mother, fully recognizing, however, that the inter- pretation of the results obtained during this period is compli- cated by the possibility that the quantity and quality of the mother’s milk may be somewhat affected by the administration and also by the possibility that the tethelin may be utilized or destroyed by the tissues or in the mammary glands of the mother so that it may fail altogether to reach the young. The decided effects upon the growth of suckling mice obtained by Robertson and Cutler as a result of administering cholesterol tothe mother, however, encouraged us to adopt this method in the present investigation. During the period of growth subsequent to weaning (3rd to 5th weeks) and comprising the greater part of the second growth cycle the mother was removed from the cage containing the young which thereafter received the tethelin directly with their food. The administration of tethelin was discontinued at the end of the 5th week (z.e., on the 35th day after birth) but the animals, there- after fed upon a normal diet, were weighed on the 42nd and 49th days after birth in order to determine the residual or continued effects of the tethelin previously administered. Sixty litters of mice were divided into three groups, A, B, and ©. The division was made at the birth of the litters, the litters ‘ Robertson, T. B., and Cutler, E., J. Biol. Chem., 1916, xxv, 663. T. B. Robertson and M. Delprat 569 being alternated so that the first litter came in group A, the second in group B, the third in Group C, and so on. In this way the groups obtained consisted of nearly equal numbers of initially similar animals. The litters were kept in separate cages during the period of the experiment and the mother was supplied with an abundance of rolled barley and water and occasionally with fresh lettuce leaves and dried bread. In addition to this the control animals (Group A) each received daily 1 ec. of mixed yolk and white of egg, another group of animals (B) received the same amount of egg mixture to which, however, were added 10 mg. of tethelin dissolved in 0.2 ce. of distilled water, while the third group (C) received the same amount of egg mixture daily to which, after 14 days, 7.e., after the termination of the first or in- fantile growth cycle, were added 10 mg. of tethelin dissolved in 0.2 cc. of distilled water. The animals were weighed daily (with occasional omissions of 1 day) to the nearest cg., each mouse in all the litters being weighed separately. The litters were all kept in the same room and under identical conditions. The dosage of tethelin was not modified with the growth of the animals. The dosage per gm. of body weight therefore diminished very rapidly with age. Excluding the uncertain proportion, very possibly, as we shall see, amounting to the whole of the substance administered, which may have been destroyed or appropriated by ‘the tissues of the mother during the 21 days of lactation, the average dosages per gm. of the young amounted approximately to the following. Age. Approximate dosage per gm. of young. days mg. 0 leds 7 0.67 14 ORD 21 0.40 28 0.31 35 0.26 This may be compared with the dosage, varying between 0.30 mg. at 5 weeks and 0.15 mg. at 1 year, administered to the mice employed in the experiments previously reported. 70 Studies on Growth. IX or DISCUSSION. The results obtained are shown in Table I and depicted graph- ically in Fig. 1, in which the continuous curve denotes the growth curve of the normal animals, the broken curve that of the ani- mals which received tethelin from birth to 35 days, and the dotted curve that of the animals which received tethelin from 14 to 35 NORMAL 17] GRAMS ie TETHELIN FROM BIRTH . TETHELIN FROM 14 Days 1 DAYS 10 20 30 40 50 Fic. 1. The influence of tethelin upon the early growth of white mice. days of age. It will be seen that prior to 14 days of age the ad- ministration of tethelin to the mother was absolutely devoid of effect upon the growth of the young, the two curves of growth being so nearly identical as to be indistinguishable from one an- Age. yy i~) ONOnmhwnrHoaoe Fs co T. B. Robertson and M. Delprat 571 TABLE I. A. iB: C. Normal. Tethelin. Tethelin after 14th day. No. weighed. ree No. weighed. ae No. weighed. ates qm. gm. gm. 118 1.47 88 1.61 126 1.47 iil L783} 102 1.82 107 1.67 116 1.96 103 2.01 106 1.85 100 2.28 81 Qo 99 228 84 2.53 90 2.51 108 2.47 78 2.76 81 DESO 89 2.76 79 3.06 60 3.14 60 3.02 91 3.35 82 53005) 100 BA! 80 3.66 83 3.59 85 3.54 91 3.86 83 SH10 85 BT 82 4.11 64 4.04 79 4.09 13 4.36 75 4.15 82 4.29 67 4.25 74 4.44 bls 4.39 65 4.58 55 4.72 49 4.75 82 4.44 78 4.64 7 4.74 75 4.74 65 4.79 76 4.93 87 4.81 69 4.85 77 4.94 76 5.12 52 5.39 76 9.33 64 5.16 63 5.44 74 Sedo 60 5.28 63 5.83 ed, 5.65 65 5.99 54 6.26 55 6.09 74 5.89 70 6.64 83 6.51 67 Gad. 65 6.92 75 7.05 79 7.03 63 7.19 79 7.38 66 Waoe 51 8.00 76 7.86 63 ach 66 8.06 71 8.36 63 8.00 61 8.67 73 8.23 62 8.65 48 9.02 49 8.84 74 8.55 63 8.88 78 9.30 65 9.49 61 9.27 63 10.02 73 9.62 66 9.60 60 10.20 67 10.04 46 10.36 66 10.41 Yh 10.52 50 10.91 70 10.98 54 10.67 56 11.02 64 11.10 52 11.14 43 laleal 43 ileal 65 11.08 57 11.61 73 IDL See 140 14.85 52 13.91 69 13.83 80 17.43 42 15.90 55 15.59 572 ‘Studies on Growth. IX other in the diagram. After 14 days a noticeable acceleration of growth occurs, and that this acceleration is a genuine effect of the administration of tethelin is indicated by the fact that the growth curves of the two tethelin-fed groups overlie one another so as to be almost indistinguishable from one another in the dia- gram, while the normal (continuous) curve diverges from them as indicated in the figure. This acceleration persists until the cul- mination of the second and initiation of the third cycle after which a decisive retardation, as observed in the previously re- ported experiments, occurs. This retardation is exhibited not- withstanding the cessation of the administration of tethelin at 5 weeks, corresponding with the conclusion of the second cycle. TABLE II. Variability. Kee A. B. OH so ae Normal. Tethelin. Tethelin after 14th day. days per cent per cent per cent 0 18.4 le 212 7 20.8 725 21.8 14 20.3 20.6 222 21 2oah 20.2 20.1 28 26.2 26). 1 25.0 35 26.2 23.2 22h, 42 20.5 21.6 20.1 49 16.0 18.5 16.8 The effects of the administration therefore persist for a consid- erable period after its discontinuance. The variabilities of the weights of the three groups of animals at 7 day intervals computed in the manner described in previous communications,° are enumerated in Table II. No decisive effect of the administration upon the variability of the animals is ob- servable, although the animals receiving tethelin display a slight tendency to diminished variability, the mean variability of the normal group during the period from and including 21 to 49 days being 23.5 per cent while the mean variabilities of the tethelin- fed groups during the same period were 21.9 and 22.3 per cent. 5 Robertson, J. Biol. Chem., 1916, xxiv, 363, 385, 397; Am. J. Physiol., 1916, xli, 535. T. B. Robertson and M. Delprat Sid No effect was observed of the administration of tethelin upon the date at which the eyes of the young opened, which is a very sharply defined criterion of development. Since the animals of Group C did not receive tethelin until the 14th day, the animals of both Groups A and C serve during the first 14 days of develop- ment as controls with which the development of Group B may be compared. It will be observed (Table III) that the average date at which the eyes opened.in the young of group B is inter- mediate between the dates at which the eyes opened in the young of the two control groups. In view of the absence of effect upon growth in weight during the first 14 days, the lack of effect of administration of tethelin to the mother upon structural develop- ment as evidenced by the opening of the eyes was to be anticipated. TABLE III. Opening of Eyes. Class. qilien Gen Opodo wien eye operat days per cent MONON TI boos iets cen eRe wes ek Saeed 14.4 7 3) (CLG HOYT) te oe ea 14.0 6 © (Tethelin after 14th day). ...-......... 13.6 9 BL BeOS hi el ae il 13.8 8 The comparative invariability of the period at which the open- ing of the eyes occurs has been commented upon elsewhere.* § As will be observed on comparing the figures enumerated in Tables _ IL.and III the variability of the period at which the eyes open is correlated with the initial variability in weight of the young at birth, the least variable group in weight at birth displaying the least variable period at which the eyes open, while the most variable group in weight at birth (C) displays the most variable period at which the eyes open. Hence the attainment of this stage of development, although relatively so invariable a phe- nomenon, is nevertheless in some measure affected by the much more variable phenomenon of growth in weight. § Daniel, J. F., Am. Naturalist, 1912, xlvi, 591. 574 Studies on Growth. IX SUMMARY. From the total lack of effect of the administration of tethelin to the mother upon the growth of suckling young we may infer either that tethelin exerts no effect upon growth during the first (infantile) growth cycle or else, which is more probable, that the tethelin administered to the mother is not secreted as such to any appreciable extent by the mammary glands of the mother. The administration of tethelin to the young subsequently to the 14th day, when their eyes are open and they have access to food other than that supplied to the mother, results in a noticeable acceleration of growth during the second growth cycle (2nd to 5th weeks), followed, upon initiation of the third cycle, by a marked retardation which evidences itself despite the fact that the administration of tethelin is discontinued at the end of the 5th week. Variability of the period at which the eyes open is correlated with the initial variability in weight of the young at birth. 34 ieee THE BLOOD LIPOIDS IN NEPHRITIS. By W. R. BLOOR. (From the Laboratories of Biological Chemistry of the Harvard Medical School, Boston.) (Received for publication, July 28, 1917.) At the time when bleeding was a common therapeutic practice nephritis was one of the conditions in which milkiness of the plasma was occasionally observed (1), and blood examinations in recent years give support to these earlier reports—indicating that there may be a disturbance of fat metabolism in this disease. Thus Watjoff (2) found, in a case of nephritis, microscopically visible fat which stained with osmic acid. Bonniger (3) reported blood fat (total lipoids) high. Erben’s (4) analyses showed increased values for fat and lecithin in a subchronic case. Greenwald (5) found high lipoid phos- phorus in some of his series of nephritics. Chauffard, La Roche, and ~ Grigaut (6) reported hypercholesterolemia in chronic nephritis with li- pemia (milky plasma) in a case of uremia, and Widal, Weill, and Laudat (7) found lipemia frequently in nephritis. Henes (8) observed that the blood cholesterol was increased and that the increase was greatest in the most severe cases. He called attention, however, to the fact that in one fatal case with uremia the cholesterol value sank shortly before death. J. Miiller (9) gave the following high values for the blood lipoids in a case of nephritic lipemia (blood taken post mortem): Total ether extract 3.6 per cent; neutral fat 2.15; cholesterol 0.84; lecithin 0.69. Schmidt (15) ‘found in most patients with hypertension, in whom kidney function was not far from normal, that the cholesterol values were high, while if marked functional deficiency existed the values were normal or below. Epstein and Rothschild (10) found in chronic parenchymatous nephritis, particu- larly in the edematous stage, that the blood lipoids were very high— cholesterol up to 1.23 gm. per 100 ce. of blood, while in uremic cases, especially those with high nitrogen retention, the lipoids were much diminished—cholesterol as low as 0.08 gm. per 100 cc. The origin of the high lipoid they believed to be ingested or mobilized fat representing a condition of non-utilization, as evidenced by the fact that on a low fat diet the lipemia disappeared. Denis (11) found a notable increase in blood cholesterol in nephritis in only one case out of about fifty of various types examined. She suggests that the lack of high values in her series may have been due to a diet which contained little cholesterol. 875 eo a6 Blood Ligeids im Nepivitis Wie De gubiisitted rests ane thus quite canticiime amd while ums Gaesers such as he mature ef the Get amd the stage of the tsese im atiitem w simestaay pen. witeh eivtonsiy ins not lesen em noe acceunt im meh of ie work reported) agpesc we Reve a fee am the pend wales, there i: ene Sieve ap Slow Gist aimee: im ike metabolism may be & amire af negieics ami Gist 2 Gorter stimiy of the blend pois im Gs eaten = demiie A sve o@ Heed sang: fom ers Gress. SD as to sete eimennsry lipem aad] Ge ele ie nests commie wih Ges af mor] inivadeads aeemdiy Busse} 1ouet oid “dmit Went Geiaei Wl Leuicisiter ai te Respite as SHOE 25 ehamet ap git ciaawes preiiired hy sii, . _ Waits. Cam Se mam whe metizedis fier ims af te Died God: Smee Ge fies pobiiestiom 2 wie are Invest ae as. Priam. @ tie Senaie jor Anes Shew wees ae a he made of Gon wii Gieed a pipes shot 1 iw 92 ee. at Ginet? ae rege I & dew: item tie vem mip 2 Se HIS Gwe Gees of sae see Ce, Pe On Bei a snail Task and wel Gekem One Gop @ eGirsie for exch ce, fas Geen Sod op peeves dime = ite bined = welll deel, HUE gis Gp avon bemeisss whieh may wie pee Sl meme eireie Sassi. Osaiene bes heen jou ce prodiee hemolysis more ire ULSI tiem airs a) Giteesfiore ae be (ese desreitle for pee Siest Feo dite well med seme op tite fask. 3 oe. ae mee wei? af wit ¢ pee aod rm Sowly wah Sime te 7 i Mec a dehelete @ 2 1) ce eee fis. = = mater af tie Heed is ioesiered to 2 sredmied o pS ie sie evuinacion of thee aes is ofsem poorly dome amd they 2? wir Wome | The evel of te wii God ed ot i corpnssie Inver is them rend off and the percentage of corpus saictiace? EF arenmsteness permit. the bleed may be deo = = W. R. Bloor BW directly into a pipette by means of a needle and short piece of rubber tubing, and run from the pipette into the centrifuge tube, or if the lipoid values of the whole blood only are desired the blood may be run directly from the 3 ce. pipette into the alcohol-ether. ' It is desirable to get the blood into the alcohol-ether as soon as possible after drawing so as to avoid possible changes in the lipoids by standing. After it is once precipitated it may be allowed to stand in the stoppered flasks for a week or more until a number of samples have collected or until it is convenient to proceed to the next step. Values do not appear to be affected in any way by this step. The filtered extracts when kept in tightly stoppered bottles in a cool place in the dark have been found to keep unchanged for at least 6 months so that the procedure offers a convenient way of collecting samples and storing them until it is convenient to make the analyses. Total Fat—In the determination of total fat the following changes have been made. To the biood extract after saponifica- tion are added 5 ce. of alcohol-ether, and the whole is raised to boiling. The liquid is then removed from the heat and after’ active bubbling has ceased the ether vapor is blown off, the beaker meanwhile being gently shaken. The 5 cc. of standard solution are heated in the same way (a few grains of coarse sand are added to promote even boiling). This treatment removes most of the ether, which if allowed to remain tends to produce bubbles in the water solution and also to cause differences in color (brownish tints) in one or other solution, due to slight . differences in the aggregation of the precipitated material. Both standard and test solutions are thus treated throughout in as nearly the same way as possible so that the changes in volume produced by the loss of the ether are the same in each. To each of standard and test solutions are now added 50 ce. of distilled water, the solutions well stirred to ensure complete solution of the soaps, and the readings made according to the original directions. If. the drying of the saponification mixture has been carried too far or if the temperature at the end of the drying has been too high, low values will be obtained, due probably either to baking of the mixture on the bottom-of the beaker or to partial destruc- tion of the cholesterol by the hot concentrated alkali. For this “eruournoud ‘stj1aydou = aynoy “*BIULOL() “‘stqiyqdou oynoy ‘woudsAqy ‘noudsiq ‘stqiaqdou Iv[NIVULO[S aInoy *SHAIVUIIYY NAN = ial rm on o.oo CG St st oO N mMonotne Neatn ee ate anton a N 28 ~~ a “soposndiop sPOOTSTOUM Guna 8S) «Oe esl in pec | baton O20.) Pesta) eT BON Maee icc 290 NC Oe | HA Celt Boal ul EO) Ge} 31 2 ee eT 6.0.) 6: || "91 OsOulranenl |. asl 9°1 6 oa Oee 4) ner 90°T | €8°T | 08 T 64'°0 | 06°T | Sa°T alee are! 4 bd o | S610 o'o Sele S.919'S So Gerd S S “OVGT O}TS ‘ONST- ‘O)9T ‘O1GS" “OFT O/SG eso oS “0)99" “OLY “OS” ‘0}09° ‘O/SE" 0/98" ‘OFS: “O\SF- “O\SP- O& 09 gg GS S SO OS SOLOS GIG SIZ LIG G0G £06 GGG 10G SIG LEG 006 VGG LOZ S 90 €.O 026 S I ‘OSVIDAV [VULION “BUISB[ “sepasndiog | “soposndio09 | “RI “sojosndio,) “poolg ef0qM “poorq efOUM cused > 2 *seposndio+y a poo[q efoGM *saposndi0¢ » Woouny Aoupryy [O1e}se[OyL) “sprodTy [e}0} vurselg , Ues0I}1U Ulo}01d-uON uly4109'T splow A44By 1eIOT, ‘JOIOJSOOYO AYR] 1890, ‘syilyda vy UL sprodvy poolg 578 *jeqidsoyy [Bsloudr) SPJosNYoLssvyy OY} JO SpLOOOI OY} WOIJ POUTB}GO 919M SoN[vA 9ST], y ‘o'0 IBsNs 5 poolg ‘“posvaso S p -ul sojyvydsoyd ips ‘yyvop ‘eIuIeIQ |Z8T | 0 |S8'0i8'% [OT | €'T] 60} GE} GG [ET OF OSS OST O/6T O\0L OST O)SS O}E9 O/E9O/E9°O) ET V0G O 8 b “yyeop : pues “eu0d ‘CQz é aInsseid poorg izez | O |69'O/S'S 12 T | 6 T] OT] OT] 9T [6L O/8% OAT OST O/T 0/09 O\SG O|0E O|T9 O\ZF O|6r 0) FT 616 W.O GG : *19}B] YOOM T ? we : qyVop pus BVUIOD |+68 z6'0\8'S |OT | ST] FT] L224] OG [9% OILE O}LT O}ES' O}TS O/6E O/ES 0/6G O}FS 0/29 0j/6S 0) 6E 8co H 1G Q “yyeop “eu0D oT LAG 1Z'0 96 °0 09°0 D 02 : “Yyeep é ‘sIsoplo’ poyxivyy 169 | € [G8 00'S |6'0 | ST] 2:0] OF | 8'T | 0'O)SF O/0E O/ZT OjZZ '0j09 O/9T OlZE O}SF O}F9 O)ZE 0} 9E G06 IW 61 r ‘YyRoep “BUIOH coole'e ITT | 2:1] 2:0 | 8%] ST | O'Ol@h Ol1Z Ol1Z O|1Z O}FL O|FZ 0|9E" O/0S" 0/29 O|E9" 0} GGG >», LT ; ‘oroudsAp AI9A {TIT |G [F8'Oi0'S JOT |S] Tt] O@) OT i ee a CSS cele o& IGG IW 91 c “BUOpa : ‘syyrzydou oynoy |ep | F Day rT 610 lee"o L¥°0 €26 HST ‘ ie | tt, 580 Blood Lipoids in Nephritis reason the drying should be done on a water bath rather than on an electric stove. The use of 50 cc. of water instead of the original 100 ce. gives a solution which is more easily read and when most of the ether is removed as above there is no trouble with bubbles. Cholesterol—Very few changes have been made in the method as described. A gentle air blast is used to hasten evaporation of the blood extract and to prevent over-heating, which produces a brownish tint and destroys part of the cholesterol. The direc- tions for the production of the color have been slightly modified. Because of the high price of satisfactory acetic anhydride the amount used has been reduced from 2 ee. to 1 ee. for each de- termination. The smaller amount has been found entirely ade- quate for the amounts of cholesterol determined. The acetic anhydride obtainable at the present time is frequently of poor quality, either colored or developing a color in the determination. Distillation has been found to improve it. The temperature at which the color is produced is within a degree or so of 22°C., rather higher than lower. This temperature is the ordinary room temperature in most laboratories. Lecithin.—The strychnine molybdate precipitation is now used altogether instead of the less convenient and less sensitive silver precipitation. The directions for the use of this reagent in the determination of lecithin have been given previously (13). A dilute sulfuric acid (one part of concentrated acid with three parts of water) is used for neutralization instead of the concen- trated acid which was originally recommended because of the necessity with the silver precipitation of keeping the volume of solution small. A number of determinations indicate that, with the molybdate precipitation, the use of cane sugar solution and second heating recommended in the earlier directions is probably not necessary and it has been omitted in the later determinations. The results of the determinations of the blood lipoids are given in Table I, the values being expressed in gm. per 100 ec. of blood. The cases are arranged in the table in approximately the order of their severity and, for completeness, the values for non-protein nitrogen (in mg. per 100 ce. of blood) and kidney function (as determined by per cent excretion of phenolsulfophthalein), ob- tained from the records of the Massachusetts General Hospital, are included. a W. R. Bloor 581 RESULTS AND DISCUSSION. Total Fatty Acids —High in both plasma and corpuscles. Lecithin——In the plasma lecithin was generally normal or below while in the corpuscles it was frequently high and especially so in Cases 222, 205, 204, and 219, where the blood samples were taken shortly before death. Cholesterol—Practically normal throughout. Fat.—In the plasma this value was, with two exceptions, much above the normal. In the corpuscles it was frequently abnor- mally high and the high values were most marked in those cases with the most severe symptoms. Total Fatty Acids: Lecithin.—In the plasma this ratio was gen- erally much above normal (due to excess of fat). In the cor- puscles it was occasionally high. Lecithin: Cholesterol—This ratio was normal in most cases in both plasma and corpuscles but high in those cases with more severe symptoms. Total Lipoids —Generally above normal. The percentage of corpuscles was generally below normal and in some of the severe cases very much below, as in the last two of the series. The plasma in this series of cases was free from visible fat although occasionally muddy with material in coarse suspension. The most marked abnormalities observed in the blood lipoids in this series were then as follows: 1. High values for total fatty acids in both plasma and corpuscles. 2. High fat in the plasma (with occasional high values in the corpuscles) which in general was most marked in those cases with more severe symptoms. 3. Frequent high values for lecithin in the corpuscles, which were very marked in fatal cases where the samples were taken shortly before death. 4. High values for total lipoids in the plasma. The high total fatty acids in the plasma and corpuscles and the high lecithin in the corpuscles with normal cholesterol are the conditions observed in the blood in alimentary lipemia (13) and 582 Blood Lipoids in Nephritis therefore strongly suggest that the abnormalities observed in the lipoids in nephritis are due to a retarded fat assimilation in the blood. ‘The extent of the abnormality would in that case depend on at least three factors—the extent of the retardation, the amount of fat present in the blood when the process began, and the amount of fat entering the blood from the alimentary canal. The great variations in the blood lipoids in nephritis reported in the literature are explainable as the result of differences in these factors. On the basis of work already done on the blood lipoids it seems possible to distinguish between “acute” and ‘‘chronic’”’ disturb- ances in the blood lipoids—acute disturbances such as occur in alimentary lipemia, which are characterized by increased fat and . lecithin, and chronic disturbances characterized also by increased cholesterol, of which the best example is diabetes. The excess of lipoids in both these examples frequently results in milkiness of the plasma. The abnormalities in nephritis observed in this work would, on this basis, be classified as acute disturbances, but there is some evidence in the literature to show that the condi- tion may become chronic with high cholesterol values. The best example of this chronic condition reported is that of Miiller already noted (9). In this case the distribution of the blood lipoids was similar to that found in diabetes—high fat (glycerides) with cholesterol increased almost parallel with the fat, while lecithin, although much above the normal value, had still not increased to anything like the extent of the other two. The plasma in this case was milky. Other cases reported in the literature had high cholesterol and lipemia (6, 7). No milkiness of the plasma was observed in any of the series reported in this paper although the plasma was occasionally muddy from matter in coarse sus- pension, which was probably not fat. As to the cause which produces these disturbances in the lipoids in nephritis, the most frequent other condition in which abnormalities of the blood lipoids are common is diabetes and a prominent symptom of severe diabetes is ‘‘acidosis’””—a decreased “alkali reserve” in the blood. Recent investigations (14) have established the fact that acidosis is frequently a feature of severe nephritis, and in most of the cases of this series acidosis was prob- ably present, as evidenced by the dyspnea, low carbon dioxide _- W. R. Bloor 583 tension in the alveolar air, and coma. Since an adequate alkalinity of the blood and tissues is necessary for their normal functioning it seems very probable that the retardation of fat assimilation found in nephritis is one manifestation of a gen- eral phenomenon brought about by a decreased blood and tissue alkalinity. SUMMARY. The abnormalities in the blood lipoids in severe nephritis were found to be high fat in plasma and corpuscles and high lecithin in the corpuscles. The cholesterol values were practically nor- mal. These abnormalities are the same as are found in alimen- tary lipemia and for this reason are regarded as the result of a retarded assimilation of fat in the blood, which in turn is thought to be one manifestation of a general metabolic disturbance brought about by a lowered ‘‘alkali reserve’ of the blood and tissues. BIBLIOGRAPHY. . Fischer, B., Arch. path. Anat. u. Physiol., 1903, clxxii, 30. . Watjoff, S., Deutsch. med. Woch., 1897, xxiii, 559. . Bonniger, M., Z. klin. Med., 1901, xlii, 65. . Erben, F., Z. klin. Med., 1903, 1, 441. . Greenwald, I., J. Biol. Chem., 1915, xxi, 29. . Chaufiard, A., La Roche, G., and Grigaut, A., Compt. rend. Soc. biol., 1911, lxx, 108. — SB Ot Hw CO OD 7. Widal, F., Weill, A., and Laudat, M., Semaine méd., 1912, xxxii, 529. 8. Henes, E., Deutsch. Arch. klin. Med., 1913, exi, 122. '9, Miiller, J., Z. physiol. Chem., 1913, lxxxvi, 469. 10. Epstein, A. A., and Rothschild, M. A., J. Biol. Chem., 1917, xxix, p. iv. 11. Denis, W.,.J. Biol. Chem., 1917, xxix, 93. 12. Bloor, W. R., J. Biol. Chem., 1914, xvii, 377; 1915, xxii, 133; 1916, xxiv, 227, 447; 1917, xxix, 437. 13. Bloor, J. Biol. Chem., 1916, xxiv, 447. 14. Peabody, F. W., Arch. Int. Med., 1915, xvi, 955. 15. Schmidt, H. B., Arch. Int. Med., 1914, xiii, 121. THE DYNAMICS OF THE PROCESS OF DEATH. By W. J. V. OSTERHOUT. (From the Laboratory of Plant Physiology, Harvard University, Cambridge.) (Received for publication, July 25, 1917.) The writer has found that by measuring the electrical conduc- tivity of tissues placed in toxic solutions the process of death ean be followed in the same manner as the progress of a reaction in vitro. Studies on a considerable variety of toxic solutions have shown that in them death proceeds as a monomolecular reaction. In the case of Laminaria in NaCl, it was observed! that the reaction behaves as if it were “‘inhibited’’ at the start, as shown by the fact that the velocity constant was fairly regular except at the start, where it was clearly below the average value. At that time - the writer was unable to suggest a satisfactory explanation of this “inhibition.”’ Later studies? have afforded a clue to the expla- nation. In these studies it was found necessary to assume that - death should be regarded as a series of reactions of the type O-—-A—M—B. It is assumed that M is a substance which determines the normal permeability and electrical resistance (and perhaps other normal properties) of the protoplasm. As long as the tissue re- mains in its normal environment M is formed as rapidly as it is decomposed but when the tissue is placed in NaCl (which is toxic for Laminaria) the reaction O— A ceases, while the reactions A — M —B proceed at an increased rate. As a result the quan- tity of M decreases to zero, at which point the tissue is regarded as dead. This point can be determined by measuring the elec- trical conductivity of the tissue as can any intermediate point (e. g., half dead, etc.). 1 The methods of measurement and of calculation are explained in Science, 1914, xxxix, 544. 2 Osterhout, W. J..V., Proc. Am. Phil. Soc., 1916, lv, 533. 585 586 Dynamics of Process of Death We may assume for convenience that in sea water (the normal environment of Laminaria) the concentrations’ of A and M are constant at 8.853 and 0.2951 respectively, and that on trans- ferring to NaCl the velocity of the reactions A — M, and M — B, increases to 0.018 and 0.540 respectively. We assume that these reactions are monomolecular and irreversible (or practically so). We can then calculate the amount of M (this amount will for con- venience be called y) at any time, 7’, after the tissue is transferred to the solution of NaCl. As explained in a previous paper,? we may employ for this purpose the formula: = 0.2051 (o— Ke Ka —Kil _ 5—-K: y = 0.2951 (e~ K27) + 8.853 =) (e — e— E:T) in which y is the amount of M, 7 is the time of exposure to the solution of NaCl, e is the basis of natural logarithms, and Ky; and Ky are the velocity constants of A — M and M — B respectively. 100%' 50 10¢ 200 MINUTES Breast Electrical resistance of Laminaria in a solution of NaCl. Ex- perimental values (0, 0) and calculated curve. The results of a series of such calculations are given in Table I, together with the values obtained in a recent series of experi- ments. The calculated and observed values are plotted in Fig. 1. It will be seen that the agreement is very satisfactory. 3 These values were used because they had been employed in previous calculations. The values of y are multiplied by 305 and 10 is added. W. J. V. Osterhout 587 TABLE I.* Net Electrical Resistance of Laminaria in NaCl 52 mM. The Resistance in Sea Water (the Normal Environment) Is Taken as 100 per Cent. Resistance. Ks Time premeep ae [i ae Observed. Caloulater a0 jeteor onect ver. | ron caleulated min. per cent per cent . 10 87.50 87.76 0.0065 0.0064 20 73.01 74.96 0.0077 0.0071 30 62.51 64.26 0.0078 0.0073 40 55.30 55.32 0.0075 0.0075 50 48.81 47.86 0.0073 0.0075 60 40.21 41.62 0.0079 0.0076 70 36.79 36.41 0.0075 0.0076 80 32.41 32.06 0.0076 0.0076 90 27.52 28 .43 0.0079 0.0077 100 24.69 25.39 0.0079 0.0077 110 23.00 22.86 0.0076 0.0077 120 22.82 20.74 0.0071 0.0077 150 16.51 16.26 0.0076 0.0077 180 14.54 13.65 0.0072 0.0077 PAVED T OUP eee iota cia giGle cits ahacieisis Mesh ss wae file 0.0075 0.0075 * All readings were made at 15°C., or corrected to this temperature. If we were unaware that this curve represented two consecu- tive reactions, and supposed it to represent a simple monomolecu- lar reaction (M — B), we should calculate its velocity constant by the usual formula:4 1 a k= po (eS) 7 p 8 a-—2 If we make this calculation, employing for this purpose the calculated values given in the third column of Table I, we obtain the values of the velocity constant K;3 given in the fifth column of Table I. It is evident from an inspection of these values that the ve- locity constant K; falls below the average value at the start. The amount by which it falls below the average value will depend on the relation K, + Ke. When K, and Ke are nearly 4 Common logarithms are used for convenience.! We put a = 100 — 10 and a—2z=y-— 10. 588 Dynamics of Process of Death equal, the velocity constant falls a good deal below the average value at the start, but as the difference between them is increased the velocity constant Ks; will be found to fall less and less below the average value at the start.2 This is easily shown by assuming various values® of K, and Ko. From this it follows that we can tell something about K, + Ke from the experimental values of Ks. It is evident that in the present case the experimental vakies of K point to the relation K, + Ky, = 30 (or K,; + Ky = 30). This relation was actually assumed by the writer in a previous paper in order to fit, not the NaCl curve, but antagonism curves? in various mixtures of NaCl + CaCl. It is therefore a striking confirmation of the general correctness of the underlying assumption that we are also able by this assumption to fit the NaCl curve so closely. In general, where a chemical reaction is slower at the start than is expected, we may suspect that we have to do, not with a simple reaction, but with consecutive reactions of the kind here deseribed.§ This explanation also applies to a considerable number of other cases of toxie action. It is of interest that in all these cases death behaves as a re- action which is continually going on but at a very slow rate until accelerated by the toxic agent. We have assumed this accelera- tion to consist partly in the increase of the velocity constant and partly in the stopping of the reaction O — A, causing a decrease in the substance (V/) to which normal permeability (and perhaps other normal properties) are due. It may prove to be generally true that death behaves as a monomolecular reaction, more or less inhibited (or accelerated) ° It should be noted that we get the same result (as regards K; falling below the average at the start) when A, + K. = 30 as when K2 + K, = 30. With certain relations of K, + Kz the constant K3; may be above the average value at the start. ®* When the values of K,; and Ky» are changed, the concentrations of A and M must also be changed in such a way that Cone. A + Conc. M = K, + K2 if we wish the concentrations of A and M to remain constant in the normal environment. 7 The manner in which the relation K,; + Kz influences the forms of these curves cannot be discussed in this paper. ® Mellor, J. W., Chemical Statics and Dynamics, London, 1909, chapt. vi. — W. J. V. Osterhout 589 at the start. The assumption of consecutive reactions affords an explanation not only of the inhibition (or acceleration) at the start but also of the fact that up to a certain point the reac- tion appears to be reversible. The latter fact will be fully dis- cussed in a subsequent paper. SUMMARY. In the case here described, death proceeds as a monomolecu- lar reaction which is somewhat “inhibited” at the start. This is easily explained if consecutive reactions are involved in the process of death. This explanation also applies to many other cases of toxic action. In all these cases death behaves as a reaction which is con- tinually going on and which is accelerated by the toxic agent. ad, Oc 17 ¥ iii ite . i \ | he THE STRUCTURE OF YEAST NUCLEIC ACID. By P. A. LEVENE. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, July 26, 1917.) Levene and Jacobs have formulated the structure of yeast nucleic acid as a tetranucleotide. The facts that led up to their formulation were: 1. The formation of four nucleosides on neutral or ammonia hydrolysis. 2. The formation of simpler nucleotides on hydrolysis with dilute mineral acids. 3. The presence of a phosphorus to nitrogen ratio which agreed quite well for the tetranucleotide theory. 4. The ratio of amino to total nitrogen in nucleic acid was in. harmony with the ratio required by the four bases, guanine, ade- nine, uracil, and cytosine. The mode of linkage between the individual nucleotides was at that time not determined, and in the graphic formula represent- ing the nucleic acid the linkage of the nucleotides was of a pro- visional character and arbitrary in nature. Following that, the studies of Levene and Jacobs on thymus nucleic acid led to methods which permitted the separation and the study of individual mono- and dinucleotides composing the complex nucleic acids. Levene and Jacobs then returned to the study of yeast nucleic acid, applying the experience gained on the thymus nucleic acids. They then directed their attention to the nucleotides obtained by acid hydrolysis of the yeast nucleic acid. A large quantity of the material was prepared and trans- formed into the brucine salt. Other work, however, made a demand on their energies, and the work on the nucleotides was somewhat neglected. However, in course of the present aca- demic year the work has been resumed. There were on hand 125.0 gm. of the brucine salts when the work was begun. 591 592 Yeast Nucleic Acid Meanwhile Dr. Walter Jones and his coworkers published several important publications on the structure of yeast nucleic acid. The basis of their work is the conception of the nucleic acid molecule as expressed by Levene and Jacobs. The work of Jones deals specifically with the mode of linkage between individual nucleotides; Jones accepts a tetraribose of the structure ( (C;Hi00;)4—3H2O) as the nucleus of the molecule. In this nucleus all the carbonyl groups of course must be free, since this is demanded by the existence of nucleosides. Dr Jones bases his conclusion on two arguments: one is a proof by analogy; namely, the wide distribution of polysaccharides in nature; the other is the discovery by him of dinucleotides in which four of the hydroxyls of the phosphoric acid are free. The first argument does not seem valid, since according to present knowledge all the polysaccharides in nature have a glucosidic structure, and the one assumed by Jones to be present in the molecule of nucleic acid can be constructed only through ether linkage. Just because a proof by analogy is lacking, all the greater rigor is required from the experimental evidence. There have appeared three papers by Jones and his coworkers dealing with the subject. In one, Jones and Richards claim to have cleaved the molecule of yeast nucleic acid into two large fractions, guanine-cytosine dinucleotide, and adenine-uracil di- nucleotide. However, the authors admit that the dinucleotides were not isolated in pure form. In a second paper Jones and Germann claim to have accom- plished by ammonia hydrolysis the same cleavage as Jones and Richards had accomplished by enzymes. However, from the fraction named guanosine-cytosine only guanylic acid was obtained. The adenine-cytosine fraction analyzed for the di- nucleotide and a brucine salt obtained from it analyzed satis- factorily for the same substance. The uracil nucleotide was not traced. In a third paper Jones and Read, following closely the directions of Levene and Jacobs, prepared the pyrimidine nucleotide fraction previously obtained by these authors. Jones and Read converted it into the brucine salt and found the analytical data of this substance to agree with that of cytosine-uracil dinucleotide. P. A. Levene 593 Thus the principal argument of Jones and his coworkers in favor of the dinucleotide structure of their substances is based on the analysis of the brucine salts. The alkaloid brucine was in- troduced into the study of nucleotides by Levene and Jacobs. The advantage of this reagent, as seen by them, consisted in the fact that it permitted fractionation of the brucine salts of nucleo- tides not only out of water, but also out of ethyl or methyl aleohol. The separation of the cytosine and thymine nucleo- tides of the thymus nucleic acid was based on this property of the brucine salts. Not much importance was attributed to the analyt- ical data of the brucine salts. Indeed, Jones did not exaggerate the value of the analytical data of the brucine salts when he criti- cized the work of Tannhauser. Employing the same ammonia hydrolysis Tannhauser thought he had isolated a trinucleotide. Jones took exception to the conclusion on the basis of the fact that the percentage composition of the brucine salt was not much different from that of the brucine salt of guanylic acid. Levene and Jacobs, in their previous work on the brucine salts of nucleotides, formulated the following requirement as a test of the individuality of a nucleotide. First, a constant composition of - the brucine salt on fractional crystallization, and second, a con- version of the brucine salt into a barium salt, which furnished analytical data agreeing with the theory for the assumed sub- stance. This seemed the minimum requirement. But even admitting that all the dinucleotides of Jones and coworkers actually exist, does this fact necessarily force the con- clusion of a tetraribose nucleus? There are theoretically pos- sible not only two but six ways of linkage between two nucleotides. They are: OS OH Oe — C;H,0, base (— H) O= P— C;H,O3 base (— H) va O eee an 0 vf 7 | O=P— C;Hs0,4 base (— H) O=P- C;Hs0,4 base (— H) oH” OH” Tu U 594 Yeast Nucleic Acid ce OH. O = P — C;H,0; base (— 2H) O = P — C;H;0; base (— H) ri | of | OH O = P—C,;H,0, base (— H) we 6 OH” OH AN O — C;H;0; base (= H) on” III IV OH OH O =P — O,H,0, base (— 2H) O = P— CsH,0, base (— H,O) OH” Yo on” ay OH ' OH cane < O= P—C,;H,0O; base (— H) O aoe ——- CsHs0, base (— 2H) oH” OH” V VI Forms III and V are possible of existence only when at least one of the two bases has two NH groups in the molecule and form VI when one base has two NH groups and the other, one NH and one OH groups. In the first three forms of dinucleotides there are less than four readily ionizable hydrogen atoms. They should form di- or tribasie salts. But we have seen in guanylic acid the readiness with which a nucleotide forms a basic salt. Furtbermore, dinucleotides of type II were shown to be present in thymus nucleic acid. The remaining three may all function as tetrabasic acids. Forms IV and V may be regarded as the two most probable structures for dinucleotides forming tetrabrucine salts. A de- cision between these two alternatives must be based on either experimental data or on valid theoretical reasoning. These possibilities must be particularly borne in mind, be- cause of the two following facts: first, in purine nucleotides the phosphoric acid is less firmly linked to the nucleoside than in pyrimidine nucleotides; and second, pyrimidine nucleotides form a dinucleotide more resistant toward hydrolytic action of acids than the purine nucleotides. The latter fact was first shown by P. A. Levene 595 Levene and Jacobs for thymus nucleic acid, and is now shown to hold also for yeast nucleic acid. These differences in behavior are surely due to differences in structure. If a diribose is proven to exist in all dinucleotides, then the difference will have to be explained by the differences in the position of the hydroxyl groups which form the oxidic linking of two molecules of ribose. The final decision on the details of the structure of dinucleotides has to be reserved. All these considerations make it a difficult task to express the structure of yeast nucleic acid in a graphic formula without any arbitrary elements. If such a one is desired, it can only be expressed for the present in the following manner. OH. O=P—C,H;0, C;H,N;0 OH” OH O=P—C,H,0, CiENZO 5 OH” —H,O | — 2H20 OBL O=P—C,H30,4 C,H;N.0, ; OH” OH O=P—C,H30, C;H,N; OH” In the present communication it is desired to report on the cytosine-uracil dinucleotide prepared some time ago. The crude material was prepared nearly 5 years ago and the dinucleotide was isolated about 4 months ago. The result was not com- municated because more light on the mode of linkage was de- sired. However, because of the recent publication of Jones and Read, our results are now presented. By means of fractional ex- traction with methyl alcohol the crude brucine salt was sepa- rated in several fractions, each different in composition. The fractions that came closest in their composition to that of the brucine salt of the cytosine-uracil dinucleotide were converted into a crystalline barium salt. The analytical data of the sub- stance agreed with the theory for the assumed dinucleotide. The ratio of the amino to total nitrogen of the salt agreed with the theory for the barium salt of the dinucleotide. The melting THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 3 596 Yeast Nucleic Acid point of our brucine salt was 200°C. (corrected). The physical constants of our substance are slightly different from those of Jones and Read. This is not surprising since Levene and Jacobs have already shown that the crude barium salt is altered in com- position by reprecipitation, and it is now also shown that various fractions of the crude brucine salt vary in composition. Further- more, we are not aware of a single instance in which a partial hydrolysis leads to products of only one phase of hydrolysis. As a rule, under such conditions, substances of different degrees of cleavage are formed. Indeed, in the partial hydrolysis of thymus nucleic acid Levene and Jacobs have shown the presence simultaneously of thymine-cytosine dinucleotides and of the mononucleotides of the same bases. Whether the substance of Jones and Read or ours is nearest to the pure dinucleotide remains to be established. As regards the mode of linkage of this dinucleotide, structure IV may perhaps be accepted, however, not with absolute cer- tainty, as structure V cannot be excluded without further proof. However, if structure V were to be accepted it would necessitate the acceptance of a union between a hydroxyl group of the sugar and one of the nitrogen atoms of the base. Taking into considera- tion the structure of the two bases uracil and cytosine, (1) HN-—CO (6) (1). N—CNH, (6) (2) OC CH (6) and (2) 0C oer (5) (3) HN—CH (4) (3) HN=—CH (4) and further, taking into consideration the fact that in cytosine the NH, group is unsubstituted and that therefore cytosine cannot serve as a connecting link between two nucleotides, then uracil would have to be accepted as the bridge between two riboses of the dinucleotide. Future work will have to decide between the two structures. This applies also to the pyrimidine dinucleo- tide from thymus nucleic acid. EXPERIMENTAL. The bulk of the material was the brucine salts prepared about 5 years ago from the barium salts described by Levene and Jacobs. Pet. P. A. Levene 597 50.0 gm. of the brucine salt were extracted with 1,500 cc. of boiling methyl alcohol. The insoluble part was successively ex- tracted until the insoluble residue did not perceptibly lose in weight on two successive extractions. The alcoholic extracts on standing at 25°C. formed a crystalline deposit. This was fil- tered, and the filtrate concentrated to dryness. The residue was recrystallized out of water. Out of 50.0 gm. there remained 14 gm. as the difficultly soluble part. This fractionation was carried out on two lots of 50.0 gm. each of the old material, and on two smaller samples prepared in course of this year. In the following table is given a summary of the analytical data of the various samples. Sample No. ¢ |u| N |e M. P. [a], Insoluble. Gontracted TRON = 9 ih eae 56.56|5.89/6.55)2.85)) 178°C. | 0.0210-0 _ _ 4 gp Decomposed 1X0.05 25-C: 7) 3 leo Re ca te ae 56. 20/6 .09)6 .62)2.64 3) a ee eee 56 .72/6 .18/6.11/3.17 Soluble. (Contracted de 58. 18|6.37|8.05]2. 7011 18°°C- gine Decomposed 1X0.05 200°C. Contracted 5 58.9916.68/7.89|2.79]1_ 185°C. | 0.18X5-0 _ 13 ge ’ Decomposed 10.05 200°C. Oars ected sa 3: 58 .47/6 .93)/9 .05|2.71 . Theory for dinu- cleotide. ...... ./59.36/6.73/8.19|2.80 Samples 5 and 6 were dissolved in warm water with the aid of a slight excess of ammonia and shaken in a separatory funnel with chloroform to remove all brucine. To the brucine-free solution barium hydroxide was added in slight excess over the quantity theoretically required to neutralize the acidity of the nucleotide, and the solution was repeatedly evaporated to dryness under diminished pressure until all ammonia was removed. The residue 598 Yeast Nucleic Acid was then dissolved in water and neutralized to litmus with sul- furie acid, filtered from barium sulfate, and concentrated under diminished pressure to a very small volume until an insoluble white precipitate began to form. On cooling, the precipitate increased. It was filtered, dissolved to a clear solution, and again concentrated to a small volume. This time a granular precipitate settled out which under the microscope consisted of striated plates and needles. 0.0955 gm. substance required for neutralization 4.93 cc. 0.1 N acid. 0.1909 <“ cS gave 0.0426 gm. MgoP.O;.. 0.0955 “ - “0.0464 “ BaSOx,. 0.0966 “ i “0.0798 gm. CO, and 0.276 gm. HO. 0.0020 “ My “in Van Slyke micro-apparatus 0.59 ec. N at 14°C. and 748 mm. Goa One Found: Om Ce UME Bt Ae Orie. Vie RT 22.69 22.53 13 ieee on Ate eh eed ya ted Ae Ae 2.62 3.19 IN ei Se ies cots Rao ena nek 7.63 7.23 12S BI Fare ee eo ey BEE ae aR el 6.76 6.24 1 Be BS kee) er ehh Cae 2 tke Da ian fos ena io 28.72 28 .59 NETS NG: Sete eee ee ee eee 1.53 1.70 The optical rotation of the substance in 2.5 per cent HCI was: *« +0.50° X 2.5 a SS Ip 5000 = + 12.5° for the barium salt = + 18.52° for the free dinucleotide. BIBLIOGRAPHY. Jones, W., and Richards, A. E., J. Biol. Chem., 1914, xvii, 71. Jones, W., and Germann, H. C., J. Biol. Chem., 1916, xxv, 93. Jones, W., and Read, B. E., J. Biol. Chem., 1917, xxix, 123; 1917, xxxi, 39. Levene, P. A., and Jacobs, W. A., Ber. chem. Ges., 1911, xliv, 1027; J. Biol, Chem., 1912, xii, 411. THE REMOVAL OF NITRIC ACID FROM SOLUTIONS OF ORGANIC COMPOUNDS. By P. A, LEVENE anp G. M. MEYER. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, July 26, 1917.) The lack of a convenient method for removing nitric acid has restricted to a minimum its use in organic and biological chem- istry. In this laboratory we have felt the disturbing presence of nitric acid particularly in the preparation of the salts of anhy- drotalonic and anhydrogalactonic acids, and of similar substances. In inorganic analysis several methods for removing nitric acid are used. The majority of them are based on the process of re- duction. Unfortunately the reduction is frequently carried out under conditions that would act destructively on the organic material, if any were present. The problem as it presented itself to us consisted in the selection of a method of reduction which could be carried out in a solution with a reaction in the neigh- -borhood of neutrality and at ordinary temperature. Besides it was desired to select such reagents as were readily removable from the solution. Of the reducing agents generally recommended for this pur- pose the following were employed: zine dust, zinec-copper couple,! iron and sulfuric acid,? and aluminum amalgam.? By means of zine dust or zinc-copper couple in solutions approaching neu- trality the reduction proceeded very slowly and never reached completion. On the other hand, by means of aluminum amalgam the reduction is completed in 6 hours. When the process of re- duction of nitric acid by means of aluminum amalgam was rec- ommended as a method for the quantitative determination of nitric acid, it was assumed that the acid was reduced completely ‘ Williams, M. W., J. Chem. Soc., 1881, xxxix, 100, 144. * Alberti and Hempel, Z. angew. Chem., 1892, 101. ’ Ormandy, R., and Cohen, J. B., J. Chem. Soc., 1890, lvii, 811. 599 600 Nitrie Acid to ammonia. Under the conditions here described only about 40 per cent of the nitric acid was reduced to ammonia; the remain- ing 60 per cent escaped the solution undoubtedly in the form of lower oxides of nitrogen. In this connection it was interesting to note that the rate of disappearance of nitric acid was much higher than that of ammonia formation, thus showing that in the course of the reduction the proportion of the lower oxides com- pared with that of NO; was continually increasing. Because of this it was deemed important to show experimentally that when the solution became free from nitric acid it was also free from Ne2Qs. Theoretically one could not expect to find an absence of nitric acid where nitrous acid was present, since, in aqueous solution, nitrites always assume the following equilibrium: 3MNO, + H20 = MNO; + 2NO + 2MOH. However, to remove all possible doubt, the reduced solution was analyzed for both nitric and nitrous acids. Nitric acid was estimated by means of “nitron.’’* At the end of the experiment a test for nitric acid was also made by means of brucine. The test was always negative. The absence of nitrous acid was shown by means.of a solution of potassium permanganate. After standing with the reduced solution for 1 hour no reduction of the permanganate solution could be detected. Hence it was proven that aluminum amal- gam could be employed conveniently for removing nitric and nitrous acids. 4 Diphenyl-endanilo-dehydrotriazol, with which it forms an insoluble precipitate. C.H.N NG.Hs fos HC < > Co) Bee NG.Hs 3usch, M., Ber. chem. Ges., 1905, xxxviiil, 681. Gutbier, A., Z. angew. Chem., 1905, xvili, 494. AP P. A. Levene and G. M. Meyer | 601 When the process is applied to the removal of nitric acid from organic mixtures it becomes necessary also to remove the intro- duced reagents or their transformation product. The following procedure was adopted. 1. The acidity of the solution is determined by titration of a small sample and the solution is then neutralized by means of barium hydroxide. 2. About 2 gm. of freshly te: aluminum amalgam are added for each gm. of nitric acid. The reduction is allowed to proceed for 8 hours or over night. The solution is aerated during the entire time of reduction. 3. The mixture is filtered from the mercury and aluminum. A slight excess of barium hydroxide is then added to the filtrate, and the mixture is concentrated under diminished pressure, to remove ammonia. Generally the process is completed after the evaporation has been repeated twice. 4. The barium is removed quantitatively and the filtrate is ready for further operations. EXPERIMENTAL. Determination of Nitric Acid with Nitron.—The solution of barium nitrate containing the equivalent of about 100 mg. of nitric acid was acidified with sulfuric acid and filtered from ' barium sulfate. To the filtrate 12 to 15 cc. of 10 per cent nitron solution in 5 per cent acetic acid were added and then the mix- ture was cooled in an ice bath. The precipitate was transferred to a weighed alundum crucible, washed with ice cold water, dried at 105-110°, and weighed. The weight of nitron nitrate < 0.168 = nitric acid. Determination of Ammonia.—Ammonia was determined by dis- tilling aliquot portions of the solution, made alkaline, into 0.1 N acid. In those experiments in which the nitrate solutions were aerated during the reduction, the diluted acid solution into which the air was passed was added to the sample taken from the reaction flask, so that the determination in all experiments rep- resents the total quantity of ammonia which was formed during the period indicated. Aluminum Amalgam.—Sheets of aluminum foil about 4 by 6 inches are passed.through a flame to remove the grease and im- 602 Nitric Acid mersed in a shallow bath of about 3 per cent solution of mercuric chloride. In a few minutes the surface of the foil is covered with mercury. The foil is immediately washed in running water and is at once transferred to the nitrate solution. Reduction of Nitrate Solutions—Dilute solutions of barium nitrate were placed in flasks with aluminum amalgam. These flasks were fitted with rubber stopper and tubing so that the ammonia which might be generated would be collected in wash bottles containing dilute acid. In some experiments the ammonia was drawn into the dilute acid by a slow current of air. Experiment 1.—4 gm. of barium nitrate were dissolved in 200 ce. of water and 2 gm. of aluminum amalgam were added and allowed to remain for about 2 hours without aeration. After 12, 36, and 60 hours, samples were withdrawn for analysis. The results are shown in Table I. TABLE I, Time. ee en ahe NHs found. Theory. hrs. qm. gm. per cent 0 0.520 12 0.0774 14.9 36 0.1449 27.9 60 0.0483 37.6 } Experiment 2.—8 gm. of barium nitrate were dissolved in 400 cc. of water and 4 gm. of aluminum amalgam were added. This solution was aerated and the ammonia was determined after 12, 36, and 60 hours. The results are given in Table II. TABLE II. Time. ate eee NH: found. Theory. hrs. gm. : gm. per cent 0 1.0400 12 0.2734 27.0 36 0.1049 31.63 60 0.0219 38.00 Experiment 3.—2 gm. of barium nitrate were dissolved in 100 ce. of water, 1 gm. of aluminum amalgam was added, and aerated ay, +4’. P. A. Levene and G. M. Meyer 603 for 14 hours, after which time the nitric acid test with nitron was negative as well as with brucine. Experiment 4.—8 gm. of, barium nitrate were dissolved in 400 ce. of water and 6 gm.of aluminum amalgam were added. After 4 hours the nitric acid test with the nitron and brucine was negative. Experiment §.—50 ce. of 2.3 per cent nitric acid and 1 gm. of aluminum amalgam were allowed to stand for 16 hours. After 4 hours, 15.9 per cent acid was not reduced and after 16 hours 4.6 per cent nitric acid remained. Experiment 6—10 gm. of barium nitrate were dissolved in 500 ec. of water and 7 gm. of aluminum amalgam were added. Then additional amalgam was added every 2 hours as indicated in the table, and at each 2 hour interval a sample of the liquid was withdrawn for the determination of nitric acid and ammonia. As Table III shows, no nitric acid remained after 6 hours although only 40 per cent of the nitric acid was obtained in the form of ammonia. TABLE III. Nitric acid determined with nitron. Ammonia (NHs). S | ne g | A BS = Z ° ‘38 28 Present. Reduced. Present. Psa, £e 33 — s we S oe Onis BZ o's = S20] 8s #|8& SS) || pers $ 3 el mi S-s ae a 8 ae eae | is) [os aie Hil< 6) Z i Z Z hrs.| gm. am. gm. per cent |per cent| gm. gm. gm. oe ae 0 | 7.0) 5.660 0.000 100.0 |000.0 |1.527 Pe AAD) DhaasSY7 4.268 24.0 | 76.0 0.1205)0.1205) 7.90) 7.60 3 | 2.0} 0.558 5.102 9.8 | 90.2 0.1475|0.2680} 9.68/17 .60 4 | 0.5} 0.0045 5.655 0.79} 99.21 0. 1665/0 .4345|10.90)28.50 5 | 0.5} Trace. | Nearly all. | Trace.| 99.99 0.1904/0.6249/12.48/40.96 6 | 0.5} None. 5.660 None .|100.00 Experiment 7.—10 gm. of barium nitrate were dissolved in 500 ce. of water and 7 gm. of aluminum amalgam were added. The solution was aerated and allowed to stand for 14 hours. The mixture was filtered and the filtrate tested for nitric and nitrous acids. Nitric acid was found absent both with brucine and nitron. 20 ce. of 0.1 N potassium permanganate acidified with An . | 604 Nitric Acid | sulfuric acid were added to 20 ce. of the filtrate and the unused permanganate was titrated, according to the method of Volhard, with potassium iodide. No nitrous poly was found. THE PREPARATION OF LYXOSE. By E. P. CLARK. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, July 26, 1917.) Of the methods for preparation of lyxose the one introduced by Ruff and Ollendorff! is the most convenient. This consists in the oxidation of calcium d-galactonate by means of hydrogen peroxide, using ferric acetate as a catalyst. The details of the directions as given by Ruff and Ollendorff are not sufficient for the preparation of this pentose on a large scale. In the course of the past year over 3,000 gm. of the sugar were prepared in this laboratory, and the process has been gradually improved so that finally a yield of 195 gm. of pure crystalline lyxose was obtained from 1,500 gm. of calcium galactonate. The conditions as finally adopted are reported here in the hope that they may prove useful to other workers. EXPERIMENTAL. 500 gm. of calcium galactonate were dissolved in 2 liters of boilmg water, and 3 liters of 3 per cent hydrogen peroxide added. The solution was cooled to about 35° and 75 ce. of ferric acetate solution? were added, which soon caused a vigorous reaction. After the reaction was completed, which was indicated by the solution acquiring a deep purple color, it was allowed to cool. The solution was filtered and evaporated in vacuum to about 1,200 ee. To the concentrated solution 4 liters of 95 per cent alcohol were added with constant stirrmg. This precipitates a gummy mass which is hard to handle but if a current of air is blown through the suspension for a short time, all the gummy particles settle out, leaving a clear solution. This solution was 1 Ruff, O., and Ollendorff, G., Ber. chem. Ges., 1900, xxxili, 1798. 2 National Formulary, 3rd edition, Baltimore, 1906, p. 219. 605 606 Lyxose filtered with suction. The gum remaining in the jar, which was drained as dry as possible, and the precipitate on the filter, were then dissolved in about 900 cc. of hot water. This may be readily accomplished in 4 or 5 minutes by heating the mixture, with constant stirrimg, to about 60-65° by means of live steam. After the gum had dissolved, the hquid was cooled to room temperature and precipitated with 4 liters of 95 per cent alcohol, as above. Three lots of 500 gm. each were treated in this way and the combined residues from them were reoxidized in the following manner: They were dissolved in several liters of hot water, allowed to cool, and filtered. The filtrate was evaporated to dryness in vacuo to remove all the alcohol. The residue was then dissolved in about 2 liters of hot water by means of live steam and 5 liters of hydrogen peroxide were added. The solu- tion was cooled to 35° and 80 ce. of ferric acetate solution were added. After the reaction was complete the solution was fil- tered and concentrated, then precipitated with 95 per cent alco- hol as previously described. The residues were dissolved in 1 liter of water and again precipitated with 4 liters of 95 per cent alcohol. The combined alcoholic extracts resulting from the above pro- cedure were then evaporated zn vacuo to 1 liter. 95 per cent alcohol was added with constant stirring until a permanent pre- cipitate was formed; about 1.5 liters of alcohol were required. This solution was poured into 9 liters of absolute alcohol with constant agitation. The precipitate formed was filtered off, drained as dry as possible, and the filtrate evaporated 7m vacuo to a thick syrup (about 700 cc.). This syrup was taken up in 8 liters of absolute alcohol and 3.5 liters of dry ether slowly added with constant stirring, which precipitates a further quantity of calcium salts and other reaction products. The filtered solution was evaporated in vacuo to 500 cc., seeded with a few crystals of lyxose, and allowed to crystallize in a desiccator. Often the syrup can be made to crystallize spontaneously without seeding by seratching the inside of the beaker. The crystals were filtered with suction and washed first with absolute alcohol and then with dry ether. The yield was gener-- ally 150 to 165 gm. of pure dry sugar. lLyxose may be readily Wits EH: BP. Clark 607 recrystallized with little loss from four to five parts of boiling absolute alcohol. No attempt was made to work over the mother liquors from the crystallization of lyxose, as it was found that they could be used directly for the preparation of lyxosimine. The syrup gave a yield of lyxosimine corresponding to about 55 or 60 gm. of lyxose. Yields of pure crystalline lyxose from lots of 1,500 gm. of calcium galactonate have been obtained as high as 195 gm., but in these cases a corresponding diminution of lyxosimine has been obtained from the mother liquors, so that almost invariably the total yield corresponded to about 210 gm. of lyxose from 1,500 gm. of calcium galactonate. RN CHONDROSAMINE AND ITS SYNTHESIS. By P. A. LEVENE. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, July 26, 1917.) In previous publications! the conclusion was reached that chon- drosamine had the structure of.one of the two lyxohexosamines: OH OH | | sete a) 5 CH | iH oy CNH, NH: CH fe) | re) | OH CH or WnOE: Gr | | | CH CH | | COH H COH | | CH,OH CH,0H The relationship of chondrosamineto galactose was originally based on the following facts: Chondrosamine formed with phenylhydrazine an osazone in- distinguishable from galactosazone. Chondrosamine, on oxida- tion with bromine or with mercuric oxide, formed chondrosaminic acid.’ This acid on oxidation with nitric acid (subsequent to deamination) yielded anhydromuciec acid. On the other hand direct oxidation of chondrosamine (subsequent to deamina- tion) led to an optically active dicarboxylic acid which was as- sumed to be anhydrotalomucic acid. Furthermore, chondrosamine on oxidation with bromine or with mercuric oxide (subsequent to deamination) gave rise to anhydrotalonic acid. Chondrosaminic acid under the same treatment formed the isomeric anhydro- 1 Levene, P. A., and La Forge, F. B., J. Biol. Chem., 1913, xv, 159; 1914, xviil, 127, 240; 1915, xx, 434. : 609 610 Chondrosamine galactonic acid. In a measure the proof seemed sufficient to establish the configuration of the new aminohexose. Yet a scrutiny of the evidence makes it clear that only one point of evidence was absolutely beyond dispute; that is, the identity of the chondrosamine osazone with galactosazone. True, the two anhydrotetrahydroxyadipie acids, one obtained from chondrosam- ine and the other from lyxohexosaminie acids, seemed identical. However, this conclusion was based on their melting points being identical and on the absence of optical activity in one and the other. These points may be considered ample proof, but addi- tional evidence was desirable. Again the anhydrotalonic and anhydrogalactonie acids had been prepared only in form of their brucine salts, and although the differences in their optical rota- tion were in full agreement with the assumed structure of the two salts, yet more direct evidence in support of the assumption was much wanted. The most convincing proof possible would be of course the synthesis of all the substances derived from chondrosamine, using lyxose as the starting material. The synthesis of the sugar has now been accomplished, and with it for every known derivative of the natural sugar a corresponding derivative of the synthetic substance has been obtained. Thus the problem of the configuration of chondrosamine is definitely solved. Chondrosamine Hydrochloride——This was prepared last year by the reduction of lyxohexosaminic acid. Its optical rotation was then found to be [a]; = + 62.69 to + 91.10°. The rotation of chondrosamine hydrochloride was determined by Levene and La Forge, [a],;= + 129.50 to 93.82°. This difference in rota- tion and the fact that chondrosaminic acid differed in its rotation from the synthetic lyxohexosaminiec acid led originally to the conclusion that the natural and the synthetic sugars were epimers. However, the fact that the rotation of the two sugars reached the same value at the state of equilibrium suggested the pos- sibility that they were a and 6 forms of the same sugar. Hence the rotation of the recently prepared chondrosamine hydro- chloride was redetermined, and was found identical with that of the synthetic product in the proximity of [@];= + 57 to 98°. Many samples crystallized under different conditions were tested, always with the same result. This was a surprising and puzzling P. A. Levene 611 finding; and although there was no doubt in our mind as to the correctness of the early measurement, since all readings in this laboratory are taken by two observers, yet to eliminate all pos- sible doubt the rotation of the original material was again meas- ured and was found as originally recorded [e]}= + 129.0 to 95.0°. It is interesting to note that in the early work of Levene and La Forge the first form occurred on three occasions whereas in subsequent work it could not be obtained again. It is also worthy of note that the value of the molecular rotation of the end carbon atom of the two amino sugars calculated according to Hudson’s formula is in good agreement with the value calcu- lated by Hudson for the end carbon atom of other hexoses. Chondrosaminic Acid—The observations on chondrosaminic acid and on the acid obtained synthetically originally seemed puzzling and confusing. By the action of hydrocyanic acid on arabinosimine, Fischer and Leuchs? obtained pure glucosaminic acid. Because of this it was thought that lyxohexosaminic acid obtained by the action of hydrocyanic acid on lyxosimine was also a uniform substance. The synthetic acid originally obtained in this manner had [a],;= — 3.58 to — 20.7°. which differed from chondrosaminic acid with [a]}= — 16.15 to — 29.2°. On the other hand, the reduction of the synthetic acid led to the same chondrosamine, which on oxidation gave rise to chondrosaminic ‘acid. The confusion, however, was cleared up when it was found that the synthetic sugar on oxidation formed the same amino acid as the natural sugar. Parallel measurements of the optical rotation were made on samples of the acids from the natural and synthetic sugars and were found identical, [e];= — 17.94 to — 31.89°. In the light of this, lyxohexosaminic acid has to be regarded as a mix- ture of the epimeric acids. Indeed, the material prepared this year had [e]}= — 1.85 to — 8.78°., which differed from the [a], of the original synthetic acid. Obviously under varying condi- tions of experiment different proportions of the epimeric acids are formed. a, a-Anhydromucic Acid, and a, a,-Anhydrotalomucic Acids.— Inactive anhydrotetrahydroxyadipic acids were obtained from chondrosamine and from the synthetic lyxohexosaminic acids. 2 Fischer, E., and Leuchs, H., Ber. chem. Ges., 1903, xxxvi, 24. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 3 612 Chondrosamine So long as the latter was considered an individual substance dif- ferent from chondrosamini¢ acid, the observation needed special interpretation. Since, however, it was demonstrated that the synthetic acid was a mixture of two epimers, one being chondros- aminic acid, the puzzling element of the observation was elimi- nated, the identity of the two anhydrotetrahydroxyadipic acids became evident, and their structure could be no other than that of anhydromucic acid. The structure of the optically active tetrahydroxyadipic acids is manifested from the fact that the natural sugar gives rise to the acid with exactly the same properties as that derived from the synthetic sugar. Since the structure of. the latter is obvious from its origin the structure of the former as anhydrotalomucic becomes certain. a,a;-Anhydrotalonic and a,a-Anhydrogalactonic Acids.—Since the structure of the a, a-anhydrotalomucie and a, a;-anhydro- mucie acid has been established it became an easy task to prove the structure of the two monocarboxylic acids. The anhydro- talonic should be convertible into anhydrotalomucic acid, and an- hydrogalactonic into anhydromucie acid. The brucine salts of two acids were previously obtained, one with a melting point of 218° and [e]>= — 12.4, and the other with a melting point of 244° and [a];= — 9.40. It was then shown that the former on oxidation gives anhydrotalomucic while the latter forms anhy- dromucic acid. The former was assumed to possess the structure of anhydrotalonic, the latter of anhydrogalactonic acid. The assumption was fully justified and stands correct in the light of the new evidence. It may be mentioned in this place that formerly the formation of anhydrotalonic acid from the synthetic lyxohexosaminic had seemed puzzling. In view of the fact that lyxohexosaminie acid is recognized as a mixture of two epimers, the source of the anhy- drotalonic is to be looked for in one of the two epimers, namely, in epichondrosaminie acid. Pentacetyl Derivatives. As a final point proving the identity of the two sugars may be mentioned the fact that they yield the identical pentacetates. 30th the a and the # forms obtained by Hudson and Dale from the ee P. A. Levene 613 natural product were then obtained from the synthetic sugar. It may be noted in this place that Hudson and Dale designated the more soluble pentacetate of chondrosamine with [a]} = 101.3° as the 6 form and the more insoluble with [e]} = 11.0° as the a form. This nomenclature was perfectly justified when it was assumed that chondrosamine was /-ribohexosamine. Inasmuch as chondros- amine is now recognized as d-lyxohexosamine, the form with the higher rotation has to be named the a form, [e]>} = 101.3°, and the one with [a] = + 11.0, the 8 form. The mutual relationship of the derivatives of chondrosamine on one hand and of the derivatives and of the parent substances of the synthetic sugar on the other is represented in the following diagram. Anhydrotalomucic acid. Anhydrotaloniec acid. 7 iN x aN Chondrosamine. ' Lyxohexosamine (synthetic Ye 7 chondrosamine). a Chondrosaminic acid. Lyxohexosaminic acid (synthetic). a NS N\ Anhydrogalactonic acid. Anhydromucie acid. ~<-————-_ ¢ EXPERIMENTAL. Preparation of Synthetic Chondrosamine Hydrochloride.—The process was essentially the same as previously described? A slight improvement was introduced in that all solutions were con- centrated at a temperature of the water bath not exceeding 40°C. It was found convenient to recrystallize the sugar by dissolving it in a minimum amount of water and adding ethy! alcoho! sat- urated with hydrochloric acid. About 35.0 gm. of the synthetic ? Levene, P. A., J. Biol. Chem., 1916, xxvi, 143, 155. 614 Chondrosamine sugars were prepared. The specific rotation of the substance was the following. Initial. Equilibrium. 0 + 1.50 X 2.0392 25 + 2.50 2.0392 a], = —————— = +59.30° lal, = = + 98.80° le); LO. OnIGT TT ee la], ix0.0516 a and 8 Forms of Chondrosamine Hydrochloride.—A sample of chondrosamine recrystallized out of water and ethyl alcohol saturated with hydrochloric acid. Initial. : Equilibrium. e + 1.54 & 2.0385 2 + 2.28 X 2.0392 (o's |S —————— = + 53.14 a| = ——__——_—__—- = 90.42 | }p 1 X 0.0516 a | Ip 1 X 0.0514 a A sample was dissolved in a minimum amount of water, glacial acetic acid was added until the substance began to crystallize, the mixture brought to a boil, and filtered. A solution was made of exactly 2.5 ec. volume. Initial. Equilibrium. 0 + 3.04 X 2.5 5 4.5 X 2.5 Qs = eeomeeael = ee = ———_ —"= + 90:0° le], 1 <:0.125 zs le]; 1X 0.125 af A sample crystallized out of a minimum amount of aqueous hydrochlorie acid. Initial. Equilibrium. o + 2.60 X 2.5 3» + 4.80 X 2.5 = = °. =S C—O 2 o ie iseodora ye la], TsCONo2u= ee No attempt was made to explain the slight discrepancies in the solutions since the principal object was to determine condi- tions controlling the formation of either one of the forms. As” all attempts to obtain a sample with the original rotation failed, the rotation of the original material was redetermined. Dr. J. Lépez-Sudrez and Dr. G. M. Meyer controlled the reading. Initial. Equilibrium. 0 +250 25 : 1.80 205 la], = ix 0.0500 -' 720 la], = 1 X 0.500 = + 96.0° Calculating on the basis of Hudson’s formula, the molecular ee 215.5 = 8,400. The value found by Hudson for hexoses was in the neighborhood rotation of the end carbon atom = ee eS PO ee cA: Levene 615 of 8,000. The original form is to be regarded as the a and the new as the 6 form. Chondrosaminic Acid from Natural and Synthetic Sugars.—This acid was obtained from both the natural and synthetic sugar by oxidation with mercuric oxide. The conditions given by Pringsheim and Ruschmann‘ for oxidation of glucosamine had to be modified. 4.0 gm. of the synthetic sugar were dissolved in 62.0 cc. of water, to the solution 20.0 gm. of mercuric oxide were added, and the mixture was warmed on the water bath for 6 minutes. The re- action product was filtered immediately, the filtrate was freed from mercury by means of hydrogen sulfide, the filtrate from the sulfide concentrated to a small volume, under diminished pres- sure, when the acid crystallized in the distilling flask. The sub- stance was recrystallized once. It did not melt, but turned light brown at 190°, the same as the acid obtained from the natural sugar. A parallel measurement of the rotation of each product gave the following results. Natural product. Initial in 2.5 per cent HCl. Equilibrium. — 0.90 X 2.5 — 1:60.X< 2:5 [a]; = Ha aieiad = — 17.94° [o]; = ——_——_— = — 31.89° 1 X 0.1254 1 X 0.1254 Synthetic. Initial in 2.5 per cent HCl. Equilibrium. 0 — 0.94 X 2.5 25 — 1:60 X 2:5 Q| = ————_—_- = — 17.94° Qa) = ——___—__ = —_ 31 .87° | Ip 1 X 0.1256 I> 1 X 0.1256 . Lyxohexosaminie acid. 0 + 0.37 X 15.0 25 leis 15-0 Q) = —_§{_—__ = + 1.85° Qe = = — 8.75° Es 50 “Eur ir & a zZ 2.0 * 1.4946 The analysis of the synthetic acid gave the following results. 0.1050 gm. of substance gave 0.1426 gm. CO: and 0.0618 gm. H.O. Calculated for CeHizNOs: Found: rar ee. Sh ng does See bes 36.92 37.03 Leese on RSS 6.66 6.58 Thus the identity of the two chondrosaminic acids is estab- lished. * Pringsheim, H., and Ruschmann, G., Ber. chem. Ges., 1915, xlviii, 680. 616 Chondrosamine Anhydrotalonic Acid.—The acid was previously obtained in small quantities as a brucine salt. Tbe brucine salt was prepared in larger quantities from the natural sugar and from the sugar obtained from lyxohexosaminic acid. Lots of 30.0 gm. of chon- drosamine hydrochloride were dissolved in 150.0 cc. of water. To the solution 30.0 gm. of silver nitrite and a few drops of hy- drochlorie acid were added. The mixture was allowed to react for 6 hours. The silver chloride was then removed by filtration, the filtrate was placed on a boiling water bath for 5 minutes, then treated with a shght excess of hydrochloric acid, and again fil- tered. To the filtrate 65.0 gm. of bromine were added and the mixture was allowed to stand at room temperature for 3 days. The remaining traces of bromine were removed by shaking the solution with mercury, and the hydrobromic acid by means of lead carbonate and silver carbonate. Finally the nitrie acid still present in the solution was removed by the aluminum amal- gam method. The resulting solution was transformed into the brucine salt in the usual manner. The solution of the brucine alt was concentrated to a small volume when, on cooling, the brucine salt was crystallized out. When first obtained the sub- stance was readily soluble in methyl! aleohol. However, on re- peated recrystallization its solubility diminished, so that towards the end it dissolved only in a large volume of boiling methy] alcohol. The salt then crystallized in large polygonal prisms. The melting point was 218°C. and the optical rotation was the following. [a]? - =, 0.63° x 10.0 ey = 12-4° > 2X 2.522 The analysis of the substance gave the following results. 0.1063 gm. of substance on drying under diminished pressure at the temperature of xylene vapors lost 0.003 gm. of water. Calculated for C29H33s N2010+ H20: Found: |: 0 ae ee i rr pers. RM Mee oc occa Sell 3.58 0.1025 gm. of the dry substance gave 0.2284 gm. CO: and 0.0569 gm. H.0. Calculated for C29H3N2O010: Found: Ce tee ee et ee ee 60.80 60.78° — BL... occv ache dice Sac CR ee a oe estes 6.10 6.15 P. A. Levene 617 From the synthetic lyxohexosaminic acid the substance was prepared in the following manner. 30 gm. of the acid weré dis- solved in a solution of 200.00 cc. of water and 40.0 ce. of 10 per cent hydrochloric acid. 40.0 gm. of silver nitrite were added. The following morning 10.0 gm. of silver nitrite and 10.0 cc. of 10 per cent hydrochloric acid were added. After 30 hours from the beginning of the experiment, the silver chloride was removed by filtration and other remaining silver by hydrogen sulfide. From the filtrate the nitric and nitrous acids were removed by the aluminum amalgam method. It had the following composition. 0.1012 gm. of substance on drying in a xylene bath under diminished pressure lost 0.0038 gm. of H,0. 0.0974 gm. of substance, dried, on combustion gave 0.2158 gm. CO» and 0.0562 gm. HO. Calculated for C29Hs6 N2010H20: Found: ET @ ey VE Sas ee ee re eine cid Mice Oe SA d aves onal Sno * Calculated for C2o9HasN2O10: Found: China 5 S68 Scr SLE cas EIS aie Go ne eae Se 60.80 60.42 TEL Sis ade’ cio ee tT Ie BG 1 1S EER one ean Dae Pee 6.10 6.45 The substance had a melting point of 218°C. and the following rotation. Gk ee 0.655° X 10 Sac 2X 2.66 Thus the identity of the acid obtained from natural chondro- samine and of the one obtained synthetically was established. In order to prove its identity with a, 6-anhydrotalonic acid it was oxidized by means of nitric acid. 21.0 gm. of the brucine salt were freed from brucine by means of barium hydroxide and chloroform. After the removal of barium, the solution was con- centrated to 40.0 cc., an equal volume of concentrated nitric acid was added, and the solution was boiled over a free flame until a lively evolution of red fumes set in. The solution was then transferred to a large clock glass and evaporated to dry- ness. The residue was dissolved in water and again evaporated to dryness. This operation was repeated. The reaction prod- uct was converted into the calcium salt. The salt was recrystal- 618 Chondrosamine lized twice. Each time the calcium salt was decomposed by a little less than the calculated amount of oxalic acid, and then reconverted into the calcium salt. 0.1008 gm. of the air-dry substance on drying in a xylene bath under diminished pressure lost 0.0130 gm. of H:0. Calculated for CeHsOsCcs+3H20: Found: HoQ) a0 Reaches Sao as Bere he ee ee 12.68 12.90 0.0878 gm. of dry substance gave on combustion 0.0936 gm. CO, and 0.0274 gm. H.O. 0.0896 gm. of the substance gave 0.0206 gm. CaO. Calculated for CceHs07;Ca+H:20; Found; Cio ek eet Ee gee a ee 29.03 29.07 ST iceipdt Date e deh ie else Ria ee 3,22 3.49 Cain Bei ssid cote te ow eee enoomne 22.99 The optical rotation of the substance in a 10 per cent solution of HCI was the following. ’ [a] ists2 0302) <2250 = = —7.5° > * 1X 0.100 ? Thus the anhydrohexonic acid obtained from lyxohexosaminic acid yields on further oxidation an optically active anhydro- tetrahydroxyadipic acid; hence it possesses the structure of anhy- drotalonie acid. Anhydrogalactonic Acid.—10 gm. lots of chondrosaminic acid were treated in the same manner as lyxohexosaminic acid in the above experiment. The brucine salt obtained in this manner melted at 244°, and had the following optical rotation. — Se ar Gk — 0.47° X 10.0 ss 2 X 2.508 The composition of the substance was the following. 0.0988 gm. of the substance, dried in a xylene bath, gave 0.2196 gm. CO, and 0.0576 gm. EO} Calculated for C29H2sN2010: Found: Oe os sie Biavens, 6 bo 0 wid eee eas Gitte Se etek ee eo 60.80 60.73 | re er re Been, Meee ee oh On clare 6.10 6.53 10.0 gm. of the brucine salt were freed from brucine as in the above experiment. The brucine-free solution was evaporated P. A. Levene 619 , to 25 ee. and then diluted with an equal volume of concentrated nitric acid and boiled over free flame until the volume was re- duced to about 20.0 cc., then 10.0 cc. of nitric acid were again added, and the solution was again boiled, over free flame. When the solution was concentrated to 20.0 cc. it was transferred to a clock glass and concentrated on a boiling water bath to dryness. The substance immediately crystallized. It was redissolved in water and again evaporated. The operation was repeated once more. The final crystalline residue was dissolved in hot ace- tone and very little ether was added, when a small amorphous precipitate formed. This was removed by filtration, and the fil- trate was allowed to crystallize. The crystals were filtered, re- dissolved in hot acetone, and allowed to crystallize. The final product melted at 205°C. 0.1600 gm. of the substance dissolved in 2.5 cc. showed no optical activity. On the basis of this the identity with anhydrogalactonic acid of the monohydroxy acid obtained from chondrosaminic acid is established. a,a,-Anhydrotalomucie Acid from the Synthetic Sugar.—Q gm. of the synthetic lyxohexosamine hydrochloride were dissolved in _ 80.0 ec. of water containing 1 cc. of hydrochloric acid. 8.0 gm. of silver nitrate were added and the mixture was allowed to stand for 6 hours. It was then filtered. The silver was removed from the filtrate by means of hydrogen sulfide. The clear filtrate from the silver sulfide was concentrated to 20.0 cc., cooled to 0°C., and diluted with 20.0 cc. of nitric acid also cooled at 0°C. The solution was allowed to stand over night. It was then boiled over a free flame until a lively evolution of yellow fumes set in. The solution was then transferred to a clock glass, evaporated on a water bath, and the calcium salt was then prepared in the usual way. The salt was reprecipitated twice and had the following composition. 0.1054 gm. of substance on drying in a xylene bath lost 0.0130 gm. H.0. Calculated for CsHsOsCa+2H20: Found: Hols ote ee 12.68 12.33 620 Chondrosamine 0.0924 gm. of the dry substance gave on combustion 0.0982 gm. CO, and 0.0298 gm. H.O and 0.0924 gm. CaO. “ Calculated for CeoHsO7Ca+H20: Found: OTR Pa rer Hay chs Ac ons Satan che oa as 29.03 28 :98 |S SRR tre hin 5 mets coo dont 3.22 3.60 OGL. Binsers 5 Whe Metiets cFooe oe IT ce eee SOR eee 22.58 23.05 The optical rotation of the substance is the following. Initial. Equilibrium. 20 ar 0.322 >< 25 20 ame 0.32° x< 2.5 a) = ————_. = — 8.0° Ql). = ———— = — §.0° [ | [ ik 1 x 0:100 : 1 x 0.100 in 10 per cent HCl. Hence the substance is a,ay-anhydrotalomucie acid. It is identical with the substance previously obtained from natural chondrosamine. Pentacetyl Derivative of the Synthetic Chondrosamine.—The substance was prepared from the natural chondrosamine by Hudson and Dale.’ Practically the same conditions were followed in this work. In 30.0 ee. of acetic anhydride 5.0 gm. of zine chloride were dissolved. To this solution 5.0 gm. of the hydrochloride were added and the mixture was warmed gently until a lively reaction developed. The reaction was kept up for 2 minutes, then the reaction product was poured into 100 ec. of water cooled to 0°C. The mixture was neutralized with potassium bicarbonate, trans- ferred to a separatory funnel, and extracted with chloroform. The chloroform extract was washed with water. Over the chloroform a layer of crystals appeared which were insoluble in water. The crystals consisted of the more insoluble fraction of the pentacetates. The chloroform extract was evaporated to dryness in vacuo, the residue was recrystallized out of alcohol, and from the mother liquor a second crop of crystals was obtained. The top fraction consisted of the pure a form ‘while the most soluble form was practically the pure 8 form. The a form turned slightly brown at 232° and melted with decomposition at 237°C. (corrected). Its optical rotation in chloroform solution was the following. 5 Hudson, C. S., and Dale, J. K., J. Am. Chem. Soc., 1916, xxxviii, 1431. 2 P. A. Levene 621 2» +0.07 x 20.0 te les “Sxo.082 ~ + *75 | A similar fraction from natural chondrosamine had a melting point of 235°C. and the following optical rotation. 2 0.09 X 20.0 (4 SIO | Ip 2 X 0.075 9 Hudson and Dale found for their 8 form a melting point of 235° (with decomposition) [e], = + 11.00°. The 8 form was apparently not quite pure, but taking into con- sideration the small quantity of starting material it is rather surprising that each form could be separated with so little difficulty. The melting point of the 8 form was very sharp at 197°C. and the optical rotation in chloroform solution was 2 +0.90 X 20.0 Qa ee lo], = SCSI Ti ee oa Hudson and Dale found for the @ form the melting point. 182-183°C. and [a], 101.3°. The composition of the pentacetyl derivative was the following. 0.1014 gm. of the substance gave 0.1834 gm. CO, and 0.0560 gm. H,0O. Calculated for CeHsNOs(CHsCO)s: Found: eee eet 2) 5 Rega Meee 52 sos) Se 49 .49 49 .32 THE RELATION BETWEEN THE CONFIGURATION AND ROTATION OF EPIMERIC MONOCARBOXYLIC SUGAR ACIDS. Ill. THE PHENYLHYDRAZIDES. By P. A. LEVENE anp G. M. MEYER. (From the Laboratories of The Rockefeller Institute for Medical Research.) , (Received for publication, July 26, 1917.) In previous publications! it was demonstrated that the direction of the optical rotation of the a carbon atom of a pair of epimeric monocarboxylic sugar acids is determined by its configuration. When the hydroxy! of the a carbon atom is in the same position as in d-gluconic acid (on the right) the direction of the rotation is right, while when it is in the same position as in d-mannonic acid the direction of the rotation is left. The rule was found valid for phenylhydrazides, brucine, strychnine, and sodium and calcium salts. In the early phase of the work it seemed that in salts of different acids with the same base and possibly in all derivatives of the sugar acids, the value of the molecular rotation of the a carbon atom was constant. If that were true the magnitude of the rotation of the a@ carbon atom of a given salt would serve as an index of its purity. Following our observation, Hudson’ succeeded in demonstrating that in phenylhydrazides the mag- nitude of the rotation of the 6, y, and 6 carbon atoms of several monocarboxylic acids were insignificant as compared with that of the a carbon atom, so that the direction of the rotation of the phenylhydrazide could be determined by that of the a carbon atom. Because of this Hudson suggested that, vice versa, the configuration of the a carbon atom in a given sugar acid may be 1 Levene, P. A., J. Biol. Chem., 1915, xxiii, 145. Levene, P. A., and Meyer, G. M., zbid., 1916, xxvi, 355. 2 Hudson, C.S., J. Am. Chem. Soc., 1917, xxxix, 462. . 623 624 Sugar Acids. III determined by the direction of the rotation of its phenylhydrazide. Wherever the phenylhydrazide is obtainable this undoubtedly is correct. In our first note on this subject only one pair of epimeric phenylhydrazides was discussed, and further work was in prog- ress when Hudson’s article appeared. We did not discontinue our work for the reason that it has furnished some interesting results regarding the numerical value of the a carbon atom in different derivatives of the same acid. As already mentioned, it was expected that the molecular ro- tation of the a carbon atom would be constant in all compounds of such acids, the purity of which could not be questioned. of — 2.98°. «0 5.101 X — 0.10° leo = “Sau x05 7 7 88° This indicates little or no purification. No better results were obtained with the acetyl derivative. The material before acetyl- ation had a value of [a], of —2.30°, and the product after saponi- — fication of the acetyl derivative showed [a], = —2.34°. Hydrolysis of Cerasin Obtained by Fractionation. The fraction most soluble in methyl alcohol was further frac- tionated out of pyridine and chloroform, which gave a product with a rotation of —2.24°. P. A. Levene and C. J. West 641 ee oacolex — 017° pe ee tt aoe 0.400 X 1 30 gm. of this material were hydrolyzed with 450 cc. of 10 per cent hydrochloric acid and 75 ce. of 95 per cent alcohol. The acids obtained from this hydrolysis, worked up in the usual way, had the following composition: 0.1167 gm. substance gave 0.3293 gm. CO» and 0.1324 gm. H,0O. Calculated for Found: CoaH2O2: CoH 00s: Ch so eas Baty eae cree eel ® a pare 78.20 75.33 76.95 Bl pe lear | Slee In Ha dale RS iar a ray A 13.16 12.50 12.70 The fraction I a, (page 639) was extracted with dilute acetone (one part water to ten parts acetone). The insoluble part had a rotation of —3.45°. » 5.2834 X — 0.28° [a]; = woes oo 20 gm. of this material were hydrolyzed by boiling for 6 hours with 150 cc. of 98 per cent alcohol containing 7.5 cc. of sulfuric acid. On standing over night at room temperature, scales of ethyl lignocerate separated out. - 0.1129 gm. substance gave 0.3240 gm. CO, and 0.1318 gm. H,0. Calculated for Found: Co6H202: KOMP Ses shy. ee na Oia 78.8 78 .26 1B. Soee oceeenee PAS SSB Cos ee on ove eae 13.10 13.16 15 gm. of the same sample were then hydrolyzed with 150 cc. of 10 per cent hydrochloric acid and 15 ce. of 95 per cent alcohol. The fatty acids were obtained through the barium salts. The acids were then transferred into the ethyl ester by boiling in a solution of 100 ce. of 98 per cent alcohol containing 5 gm. sulfurie acid for 6 hours. On standing, scales separated out, with the following composition. 0.1150 gm. substance gave 0.3264 gm. CO» and 0.1306 gm. H;O. Calculated for Found: 26 5202: Oy sue us 1a Sa a ea ee 78.80 77.40 642 Cerebrosides. IV Hydrolysis of Cerasin Obtained by the Benzoylation Process. 3 gm. of the material with [a]; = —3.25° were heated in a sealed tube with 50 ce. of 3 per cent sulfuric acid at 100—105° for 18 hours. The acids were isolated through the barium salts and had the following composition. 0.1044 gm. substance gave 0.2940 gm. CO, and 0.1152 gm. H,0. Calculated for . Found: CosHasO2: C25H5003: OF ee STs a eet Pee sat Meare 78.20 75 38 76.80 TS srg Pe re eae nn 13.16 12.50 12.385 Acetyl Phrenosin. Acetyl phrenosin was first prepared by Thierfelder.6 20 gm. of phrenosin, 20 gm. of fused sodium acetate, and 200 cc. of acetic anhydride were heated to gentle boiling for } hour. The excess of acetic anhydride was removed by distillation in vacuum and the semisolid residue dissolved in ether and water. The ether solution of the acetyl derivative was washed with water, with dilute’ alkali, and again with water, dried over sodium sulfate, and the ether concentrated on the water bath. The residue was taken up in hot dry methyl alcohol, in which it is very soluble, and from which it separates in a granular condition upon cooling. After the second crystallization from methyl alcohol, the yield was 14 gm. of a product with [@],=—10.4°. The third crystal- lization gave a product with [e],=—11.07°, which was not changed on further purification. Acetyl phrenosin melts somewhat unsharply at 41-48°. Thier- felder gives the melting point as 39-41° and the rotation as —3°. Our product, dissolved in a mixture of equal parts of chloroform and methyl alcohol (by volume) showed the following rotation: Ge « 7.0708 X — 0.39° 1; P 0.4980 x 0.5 2 6.5070 X — 0.43° _ fe ee iT .08° Lalo = 9 5085 x 0.5 a 110% a ~ P. A. Levene and C. J. West 643 0.1052 gm. substance gave 0.2568 gm. CO, and 0.0892 gm. H.0. 0.5000 gm. substance neutralized 4.35 cc. 0.1 N HCl. 0.500 gm. substance, used for an acetyl determination, required 28 cc. 0.1 n HCl, and in a second experiment, 28.2 cc. 0.673 gm. substance, in 22.56 gm. chloroform produced a rise in boiling point of 0.108°. 0.943 gm. substance, in 22.56 gm. chloroform, produced a rise in boiling point of 0.158°. Calculated for hexacetylphrenosin Found: R CeoHios NOt: (OEP Ae 5s Sips aati DO te RRC LN SO 66.68 66.57 1S HS | tn Gloom co aa oo ete 9.80 9.49 IN epee estoy a3 oe Reka eho SMa eects (ERE SR 1.29 1.22 DOE ARG cimiadiy niin Hee AIRR SN GRIER ADS ea ee 23.8 24.08 24.24 BY Co) (Eta fa exh gts econ NP ee A ae ea 1079 1003 960 One experiment was made to determine whether the same derivative could be prepared by the use of acetyl chloride. 20 gm. of phrenosin were dissolved in 150°ce. of pyridine, the solu- tion was cooled to 0°C., 16 ec. of acetyl chloride were added in portions, and the mixture was allowed to stand 2 days at 0°C. The pyridine hydrochloride was filtered off, the solution washed with water, dilute hydrochloric acid, and then with water, dried with sodium sulfate, and concentrated. The residue was crystal- lized repeatedly from methyl alcohol. After the fourth crystal- lization, it had a rotation of —8.41°, and melted at about 4°. 6.121 — 0.33° [a ee 6.1218 X_— 0.33° 8 Ae Dd” 0.4794 X 0.5 The analysis indicates that it is probably a mixture of acetyl derivatives. It was not purified further. 0.1008 gm. substance gave 0.2484 gm. CO, and 0.0896 gm. H,0. Calculated for Hexacetyl- Triacetyl- phrenosin phreno-in Found: CeoH20sNOis: CssHo9NOn: *! So BES iio oe 66.68 67.94 67 .22 STREP ee ae ee my hs 9.80 10.46 9.96 Acetyl Cerasin. 20 gm. of a cerasin with ee —2.8°, 20 gm. of sodium acetate, and 200 ce. of acetic anhydride were boiled under a reflux for 3 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 3 644 Cerebrosides. IV hour. The reaction product was worked up as described above, and the product then twice crystallized from methyl alcohol. This material did not show a change in rotation upon further purification from methyl alcohol. It is slightly more insoluble in methyl alcohol than is acetyl phrenosin, as stated by Thier- felder.° Acetyl cerasin melts at 54—56°, and in a mixture of equal parts of chloroform and methyl alcohol gave the rotation: = ee leo [ay 5.9310 X — 0.60° ss 0.4324 X 0.5 0.1035 gm. substance gave 0.2576 gm. CO» and 0.0932 gm. H.O. 0.500 gm. substance neutralized 4.4 ec. 0.1 N HCl. Calculated for pentacetyleerasin Found: Cs7Hio NOt: Cr See ORT ARE Soe ole Oe Se eee a 67.87 67.75 1G pe cee ASD Seta irs aS Pet Roe RM i I tp 10.10 10.00 INES gs ee errr Cee CY ce Me i UR ee oa A 1.40 1.23 As stated on page 639, the acetyl derivatives are not a suitable means of separating phrenosin and cerasin. Benzoylphrenosin. 10 gm. of practically pure phrenosin were dissolved in 100 ce. of pyridine, cooled to room temperature, and treated with 12 ce. of benzoyl! chloride. The reaction mixture was cooled under the water tap and then allowed to stand at 0°C. over night. The next day the pyridine hydrochloride was filtered off, the solution concen- trated in vacuum at 50°C., and the resulting oily liquid poured, with stirring, into a large volume of about 2 per cent sodium hydroxide solution. The benzoylphrenosin separated as an oil on the sides and bottom of the dish. After washing with water and hydrochloric acid, the material was taken up in ether and dried with sodium sulfate. An alternative procedure is to dis- solve the pyridine containing oily material in ether, then shake with dilute acid and alkali, then with water, and dry this solu- tion. After removal of the ether the product is taken up in hot methyl alcohol. From a concentrated solution, the benzoyl de- rivative separates as an oil, which solidifies on cooling. From a sufficiently dilute solution it separates as colorless nearly crystal- P. A. Levene and C. J. West 645 line material, which is easily filtered. It may also be obtained in a granular form from ethyl alcohol. It is easily soluble in acetone, benzene, petroleum ether, chloroform, acetic acid, and pyridine. In an experiment in which 80 gm. of phrenosin were used, the benzoyl derivative was purified by crystallization from methyl alcohol, then from acetone (in which about 50 per cent of the material was lost), and then from methyl alcohol until the rota- tion was constant. The melting point as observed on this material was fairly sharp at 65-66°. In methyl alcohol and chloroform the rotation is: _ 7.1484 X + 0.75° alee = Seon ne | Ip 0.5062 * 0.5 ae 6.1518 X + 0.92° Te lee a ee 281-979 le], 0.5320 X 0.5 qe » _ 6.1060 X + 0.91° : Mit. [eo], = osnox0s 1 2 Sample I was obtained from material which contained at least 75 per cent cerasin. Samples II and III were obtained from material considered nearly pure phrenosin. The analysis and molecular weight indicate that this material of constant rotation was probably a tribenzoylphrenosin. 0.1004 gm. substance gave 0.2704 gm. CO» and 0.0792 gm. H.O. 0.1028 <“ PPO i 84 a eae err) 00866: — 6 “8 0.1000 “ - omnivore merce crn OORTG.. oS 0.5000 “ neutralized 4.40 cc. 0.1 N HCl. 0.5000 “ oe ae AO SasOn Ll. — 0.500 “ “ used for a benzoyl determination required 11.6 ec. 0.1 N NaOH, and in a second experiment EOC Ee: 0.255... “ es in 23.21 gm. chloroform raised the boiling point 0.026°. 0682.“ es as above, raised the boiling point 0.108°. 008° “ P CPP eas a FS § “f ce ONGOe Calculated f abeuacvipkaonosin Found: CesHiosNOv: CD be \ OI ina ae A 72.65 73.45 73.89 73.81 PIE e087. les ea 9.28 8.83 9.43 8.69 Nis. oe Se rr ae 1.23 1230) MS? OPIETAC 0) a) Sareea y 1598 24.36 24.99 VEGI ae Pet 1,139 1,037 1,066 1,092 Average....... L380, O86 1d bce EE hy 1,065 hog Ai 646 Cerebrosides. IV An attempt was then made to prepare a hexabenzoyl deriva- tive. 40 gm. of the above mentioned material were dissolved in 300 ce. of pyridine and treated with 400 ec. of benzoyl chloride. After working up as usual, the product was taken up in methyl alcohol. An oil formed which would not solidify even after standing several weeks. Since the material with constant rota- tion served our need the experiment was discarded. Phrenosin from Benzoylphrenosin. About 10 gm. of sodium were dissolved in 300 cc. of methyl alcohol and 20 gm. of benzoylphrenosin added with shaking (the benzoylphrenosin may also be added in acetone solution). A white precipitate gradually forms upon boiling (immediately, if an acetone solution is used). After boiling for 2 hours, the re- action product is cooled in the ice box, the precipitate filtered off, recrystallized from methyl alcohol, then from glacial acetic acid, and finally from a mixture of equal parts of methyl and ethyl ~ alcohols. The phrenosin crystallizes practically quantitatively for upon concentration of the mother liquor only sodium benzoate was obtained. The rotation of this material in pyridine corresponds well with that found for the phrenosin obtained by fractionation of the top fraction of the cerebroside mixture. Rosenheim gives +3.78° and +3.70°. In pyridine the rotation was: » _ +8.0000 x 0.12° Po gaa aL a lo], 0.5932xX05 °° Cinnamoylphrenosin. 10 gm. of phrenosin were dissolved in 100 cc. of pyridine and 15 gm. of cinnamoyl chloride added. After standing over night at 0°, the filtered solution was concentrated and worked up as described above. The resulting product was purified by crystal- lization from methyl or ethyl alcohol, or from acetone. Twice recrystallized from methyl alcohol, and once from acetone, the substance analyzed as a tricinnamoylphrenosin. Tricinnamoyl- phrenosin is slightly less soluble in organic solvents than the benzoyl derivative. It melts at 69-70°. P. A. Levene and C. J. West 647 2 _ 7.2060 X 0.75° CY = ae = oe Le Ip 0.4976 X 85 2 0.1028 gm. substance gave 0.2718 gm. CO» and 0.0868 gm. H.O. 0.1016 “ ss ES 0:27 32ers SOLOS 70) ie 0.500 <“ cote hoe aan 1.70 1.68 The optical rotation of the substance, in a mixture of equal parts of chloroform and methyl alcohol (by volume) was: [ar = 9.5312 X 0.26° = = + 14.00° . 0.3550 X 0.5 Hydrolysis of the Mixed Cerebrosides. 1.5 gm. of the mixed cerebrosides were heated with 75 cc. of 3 per cent sulfuric acid for 24 hours in a sealed tube at 105°. The base and acids were separated and prepared for analysis as described in a previous article. The acids had the following composition: 0.1012 gm. substance gave 0.2978 gm. CO, and 0.1098 gm. HO. Calculated for Found: CosH 500s: CosHagQOo: Ore ee nls Fg ene THB) 78.20 75.40 1B cS a ae Cr Ee Tee eo 12.50 13.20 12.18 Thus the acid was apparently nearly pure cerebronic acid, C2;H50s. The base was transformed into the sulfate and gave the following fig- ures on analysis. 0.0990 gin. of substance gave 0.2264 gm. CO, and 0.0910 gm. H:0. Calculated for Found: (CizH3sNOv2)2H2S0a: RO ro eS ee ics ater Lode 61.08 62.36 le 5 ce AS RETOR SG 5 SiG a ee ee 10.78 10.29 Crude sphingosine sulfate, previous to crystallization, very frequently gives analytical data as in the present experiment. There is little doubt that the base of the kidney cerebrosides is sphingosine. 652 Cerebrosides. V Liver Cerebrosides. Desiceated and pulverized liver tissue was allowed to stand over night with 95 per cent alcohol and then filtered. The residue was repeatedly extracted with boiling alcohol, each ex- traction lasting about $ hour. The combined extracts, on stand- ing in the refrigerator, at 0°, gave a dark, nearly black deposit, which corresponds to the “‘white matter” of the brain extracts. This deposit was extracted in the cold progressively with ace- tone, alcohol, and ether. The still dark but quite dry mass was fractionated into two parts by dissolving it in hot pyridine and allowing it to cool to room temperature. The mother liquor containing the cerebrosides was concentrated and poured into acetone. The precipitate thus obtained was still very dark. For further purification it was boiled with hot alcohol; a small part remained insoluble. The solution was decanted and a concen- trated solution of barium hydroxide was added as long as a pre- cipitate formed. ‘This mixture was allowed to stand in the ice box, and the precipitate which formed was repeatedly extracted with boiling alcohol. The extracts, upon cooling to 0°, gave a precipitate which had the appearance and properties of the brain cerebrosides. This product was then repeatedly extracted with ether, when analysis showed that it was still contaminated with large amounts of neutral fat. The purification was then con- tinued by. extraction with ether and by crystallization from acetic acid. The product, however, persisted in containing neutral fat. Finally, the product was dissolved in hot methyl ethyl ketone, from which it settled out on cooling. This was repeated three times, the product then having the composition of cerebrin. Because of the great losses connected with the purification it was not possible to obtain sufficient material for hydrolysis. The test for galactose was positive and there is little doubt that we are dealing with a characteristic cerebroside mixture. 0.1006 gm. substance gave 0.2522 gm. CO: and 0.1010 gm. H,0. Caleulated for phrenosin: Found: ee ee ema s sigala § Gok do Soleo &o8' Or 69.65 68.36 |: ee ern men einen Ge he ie od Sake, ec cemeyee 11.24 - iieZs P. A. Levene and C. J. West 653 Egg Cerebrosides. Egg yolk (dried commercial egg yolk was used in all the work) was thoroughly extracted with acetone at room temperature, to remove egg: oil. The material was then extracted with boiling alcohol, repeatedly, as in the former preparations. The com- bined alcoholic extracts were concentrated to a small volume, and repeatedly treated with acetone, to complete the removal of the egg oil. The acetone-insoluble fraction (lecithin, cephalin, cerebrosides, and saturated phosphatides) was extracted with ether. A small part did not go into solution. The ether sus- pension was centrifuged, the insoluble material suspended in acetone, filtered off, and again extracted with ether. The in- soluble material then corresponded to “white matter” previously mentioned, and is the material analyzed by Stern and Thier- felder® and considered impure diaminomonophosphatide. It was later given the name alben by Bing and Ellermann.? This was fractionated out of pyridine as described above. The cerebroside fraction was crystallized out of glacial acetic acid and the neutral fat removed by repeatedly extracting with acetone at 50°, and finally by crystallization from methyl ethyl ketone. This material possessed all the physical properties of the mixed: cerebrosides, and gave the galactose test with orcin. 0.1020 gm. substance gave 0.2476 gm. COz and 0.0974 gm. H,O. 0.5000 ‘ “s neutralized 6.75 ec. 0.1 N HCl. Calculated for 5 phrenosin: Found: Cee ey are ee maT Ys ae UG Li 69.65 69.32 DE incr Sey etd Eset BOs et EE Ol a ce 11.24 11.19 IN| geass AU ea sos Re ee Gi ep ee ee 1.70 1.89 The optical activity of the material in a mixture of methyl! alcohol and chloroform was: 12.4370 X 0.11° Ee 0.3062 X 0.5 ape 8 Stern, M,, and Thierfelder, H., Z. physiol. Chem., 1997, liii, 370. ° Bing, H. J., and Ellermann, V., Biochem. Z., 1912, xlii, 289. 654 Cerebrosides. V Hydrolysis of Egg Cerebrosides. 1.5 gm. of the substance were heated in a sealed tube with 75 ce. of 3 per cent sulfuric acid for 24 hours at 105°. The acids and bases were prepared in the manner described above. 0.1000 gm. substance gave 0.2708 gm. CO: and 0.1070 gm. H,0. Calculated for Found: CosH5003: Gosia. coat Sa Ra cee aes 75.33 74.52 rere es ee ee ere es ce See 12.50 12.09 The base was analyzed as the sulfate. 0.0905 gm. substance gave 0.2024 gm. CO, and 0.0810 gm. H,0. Calculated for Found: (C17Has NOz)2H2SOa: ( Opes 2 is ae Nee Gey AEE as ek ered 13 61.08 60.99 | Rees ee ee Teer ee re Ae Sree Bsc | Vers: 10.01 INDEX. CETO-ACETIC and 6-hydroxy- butyric acids, intravenous in- jections of, 59 Acid and base, effect on calcium and magnesium metabolism, 421 —., citric, fermentation of Asper- gillus niger, 15 ——, fatty, derivatives, effect on calcium and magnesium metab- olism, 441 -——, guanylic, and its preparation from yeast nucleic acid, 47 ——, hydrochloric, effect on mineral excretion of dogs, 461 ——, nitric, removal from solutions of organic compounds, 599 ——., nitrous, effect of temperature on reaction of lysine with, 527 ——, tritico nucleic, 295 ——, uric, content of the blood of new-borns, 261 ——, , In urine and blood, modi- fications in colorimetric deter- mination of, 165 ——, yeast nucleic, preparation of guanine mononucleotide from, 47 ; , structure, 591 Acidosis, tables for finding alkaline reserve of blood serum in health and acidosis from total COs, or alveolar CO, or pH at known CO, tension, 519 Acids, aromatic, comparative metab- olism of, 307 ——,, diamino-, in proteins, nutri- tive value for maintenance of adult mice, 173 ——,, B-hydroxybutyrice and aceto- acetic, intravenous injections of, 59 Acids, monocarboxylic sugar, rela- tion between configuration and rotation of, 623 Adrenalin hyperglycemia, relation to decreased alkali reserve of the blood, 471 Alkali reserve of the blood, de- creased, relation to adrenalin hyperglycemia, 471 Alkaline reserve of blood serum in health and acidosis, tables for finding, from total COs or al- veolar CO: or pH at known CO, tension, 519 Alveolar air and blood, CO: in, and CO, combining power of plasma and whole blood, determination of, 217 Ammonia formation in soil, effect of salts, 411 Anemia, blood lipoids in, 79 Aromatic acids, comparative metab- olism of, 307 Aspergillus niger, citric acid fer- mentation of, 15 BASE and acid, effect on calcium and magnesium metabolism, 421° Baumann, L., and Hines, H. M. The origin of creatine. II, 549 Bile and phenol production, 255 Bioluminescence, 311 Blood and alveolar air, CO2 in, and CO, combining power of plasma and whole blood, de- termination of, 217 — fat in domestic fowls in rela- tion to egg production, 281 lipoids in anemia, 79 nephritis, 575 655 656 Blood, modification of MeLean-Van Slyke method for the determina- tion of chlorides in, 483 —— of new-borns, uric acid con- tent, 261 serum, tables for finding alka- line reserve of, in health and acidosis, from total COs, or alveolar CO». or pH at known COs, tension, 519 ——sugar, determination of, in reference to its condition in the blood, 533 ——and urine, modifications of colorimetric determination of uric acid in, 165 Bioor, W. R. The blood lipoidsin nephritis, 575 —— and MacPuHerson, D. J. The blood lipoids in anemia, 79 Boaert, L. J. A note on modifica- tions of the colorimetric de- termination of uric acid in urine and blood, 165 (CALCIUM, effect of diets poor in, on calcium and magnesium metabolism, 435 ——and magnesium 421, 435, 441 Carbon dioxide in alveolar air and blood and CO, combining power of plasma and whole blood, 217 ——-——, total or alveolar, and pH at known CO; tension, tables for finding alkaline reserve of blood serum in health or acido- sis, from, 519 Catalysis, light production at low temperatures by, with metal and metallic oxide hydrosols, 201 Cerasin, 635 Cerebrosides, 627, 635, 649 ——, conditions for hydrolysis, 627 ——of the kidney, liver, and egg yolk, 649 metabolism, “ Thee” Chlorides in blood, modification of MecLean-Van Slyke method for determination of, 483 —— in body fluids, determination, 55 Chondrosamine and its synthesis, 609 Citric acid fermentation of Asper- gillus niger, 15 Criark, E. P. The preparation of lyxose, 605 Corpus callosum and intradural nerve roots (man and dog), relative amount of sheathing substance in, 395 Cottonseed flour, nature of growth- promoting substances, 379 — meal, nutrition investigations on, 379 Creatine, origin of, 549 Creatinuria in normal adults, 561 Curriz, J. N. The citric acid fermentation of Aspergillus ni- ger, 15 Cystine, influence of small amounts on the balance of nitrogen in dogs maintained on a low pro- tein diet, 363 DEATH, dynamics of process of, 585 Deficiency diseases and the vita- mine hypothesis, 229 Devprat, M. See Ropertson and DELPRAT, 567 Denis, W., and Minot, A. S. The production of creatinuria in normal adults, 561 Determination of blood sugar in reference to its condition in the blood, 533 chlorides in blood, modi- fication of McLean-Van Slyke method for, 483 — body fluids, 55 —— — CO, in alveolar air and blood, and the CO, combining power of plasma and whole blood, 217 ‘Index Determination, colorimetric, of uric acid in urine and blood, modi- fications of, 165 ——, quantitative, of dextrose in muscular tissue, 67 Dextrose in muscular tissue, quan- titative estimation of, 67 Diamino-acids in proteins, nutri- tive value for maintenance of adult mice, 173 Diet low in protein, influence of small amounts of cystine on nitrogen balance of dogs on, 363 poor in calcium, effect on cal- cium and magnesium metab- olism, 435 che ——, vitamines in, 149 Dinucleotide, uracil-cytosine, 39 Dusin, H. The influence of bile on phenol production, 255 FPMOND, H. D. See WARNER and Epmonp, 281 Egg production, relation to blood fat in domestic fowls, 281 —— yolk, cerebrosides of, 649 Enzyme action, 97, 201 Enzymes, proteoclastic tissue, of the spleen, 303 FALg, K. G. Studies on enzyme action. XIV. Further experi- ments on lipolytic actions, 97 Fat of blood in domestic fowls in relation to egg production, 281 and fatty acid derivatives, effect on calcium and magne- sium metabolism, 441 Flour, cottonseed, nature of growth- promoting substances, 379 Food, availability of energy of, for growth, 389 Foster, G. L. A modification of the McLean-Van Slyke method for the determination of chlo- rides in blood, 483 657 FRANKEL. E. M. Studies on enzyme action. XV. Factors influenc- ing the proteolytie activity of papain, 201 (As analysis, applications of, 217 Gertine, E. M. K. The nutritive value of the diamino-acids occurring in proteins for the maintenance of adult mice, 173 Gryetin, H. R. See Perers and GEYELIN, 471 Givens, M. H. Studies in calcium and magnesium metabolism. Il. The effect of diets poor in calcium, 435. III. The effect of fat and fatty acid derivatives, 441 ——and MeEnpe.t, L. B. Studies in calcium and magnesium metabolism. I. The effects of base and acid, 421 Glyeolytie properties of muscular tissue, 501 Goss, B. C. Light production at low temperatures by catalysis with metal and metallic oxide hydrosols, 271 Green, H. S. See RicHARDSON and GREEN, 379 GREENWALD, I., and Wetss, M. L. The fate of inosite administered to dogs, 1 Growth, availability of energy of food for, 389 ——, experimental studies, 567 Growth-promoting substances of cottonseed flour, 379 Guanine mononucleotide preparation from yeast nucleic acid, 47 Guanylic acid and its preparation from yeast nucleic acid, 47 and its HA4LPIN, J.G. See Hart, Hat- PIN, and STEENBOCK, 415 658 Harpine, V. J., and Mason, E. H. The estimation of chlorides in body fluids, 55 Hart, E. B., Haurin, J. G., and Sremnspock, H. The behavior of chickens restricted to the wheat or maize kernel, 415 ——and Humpnrey, G. C. The relation of the quality of pro- teins to#milk production. III, 445 See Sure and Harr, 527 Harvey, E.N. Studies on biolum- inescence. VIII. The mech- anism of the production of light during the oxidation of pyrogallol, 311 Henperson, Y., and Morriss, W. H. Applications of gas analy- sis. I. The determination of CO: in alveolar air and blood, and the COz combining power of plasma and of whole blood, 217 Hines, H. M. HIngEs, 549 Hoacianp, R. The quantitative estimation of dextrose in mus- cular tissue, 67 —and Mansrietp, C. M. The function of muscular tissue in urea formation, 487 —— and——. Glycolytic proper- ties of muscular tissue, 501 Humpurey, G. C. See Hart and HumPusrey, 445. Hydrochloric acid, effect on mineral excretion of dogs, 461 Hydrogen ion concentration of the ileum content, 269 —— —— —— at known CO, ten- sion, tables for finding alkaline reserve of blood serum in health and acidosis, from, 519 Hydrosols, metallic oxide, light production at low temperatures by catalysis with, 271 See BAUMANN and Index 8-Hydroxybutyrie acid and aceto- acetic acids, intravenous in- jections of, 59 Hyperglycemia, adrenalin, rela- tion to decreased alkali reserve of the blood, 471 [LEUM content, hydrogen ion concentration, 269 Inosite administered to dogs, 1 JONES, W., and Re awreab. Structure of the purine mon- onucleotides, 337 —and—. _Uracil-cytosine di- nucleotide. 59 KIDNEY, cerebrosides of, 649 Kinassury, F. B., and Sepewicx, J. P. The urie acid content of the blood of new-borns, 261 Kocu, G. P. The effect of different salts on ammonia formation in soil, 411 Kocu, M. L. See Kocn and Kocu, 395 Kocu, W., and Kocu, M. L. Con- tributions to the chemical dif- ferentiation of the central nervous system. IV. The rel- ative amount of sheathing substance in the corpus eallo- sum and intradural nerve roots (man and dog), 395 Kuriyama, S., and Mrennet, L. B. The physiological behavior of raffinose, 125 ] EVENE; PA Chondrosamine and its synthesis, 609 —. The structure of yeast nucleic acid, 591 ——and Meyer, G. M. Cerebro- sides. III. Conditions for hy- drolysis of cerebrosides, 627 lnidex Levene, P. A., and Meyer, G. M. The relation between the con- figuration and rotation of epimeriec monocarboxylic sugar acids. III. The phenylhy- drazides, 623 —and—. The removal of nitric acid from solutions of organic compounds, 599 —and West, C. J. Cerebro- sides. IV. Cerasin, 635. Y. Cerebrosides of the kidney, liver, and egg yolk, 649 Lewis, H. B. The metabolism of sulfur. II. The influence of small amounts of cystine on the balance of nitrogen in dogs maintained on a low protein diet, 363 Light production at low tempera- tures by catalysis with metal and metallic oxide hydrosols, 271 — — during the oxidation of pyrogallol, mechanism of, 311 Lipoids of blood in anemia, 79 — nephritis, 575 Lipolytie action, experiments, 97 Liver, cerebrosides of, 649 Lors, J. The similarity of the action of salts upon the swell- ing of animal membranes and of powdered colloids, 343 - Lysine reaction with nitrous acid, effect of temperature on, 527 Lyxose, preparation of, 605 MACPHERSON, D.J. See Boor and MacPHerson, 79 Magnesium and ealcium metab- olism, 421, 435, 441 Maize or wheat kernel, behavior of chickens restricted to, 415 MANsFigLp, C. M. See Hoaguanp and MansFIExp, 487, 501 Mason, E. H. See Harpinc and Mason, 55 659 McCuenpon, J. P., SHEDLOvV, A., and Tuomson, W. The hydro- gen ion concentration of the ileum content, 269 ——, ——, and Tables for finding the alkaline reserve of blood serum in health and acidosis from the total CO, or the alveolar CO: or the pH at known COs, tension, 519 McCoiuum, E. V., and Pirz, W. The vitamine hypothesis and deficiency diseases. A study of experimental scurvy, 229 McGuiean, H., and Ross, E. L. Methods for the determination of blood sugar in reference to its condition in the blood, 533 Meal, cottonseed, nutrition in- vestigations on, 379 MENDEL, L. B. See Givens and MENDEL, 421 See Kurtyama and MENDEL, 125 See OsporNE and MENDEL, 149 Metabolism, calcium and nesium, 421, 435, 441 —, comparative, of certain aro- matie acids, 307 — of sulfur, 363 Metallic oxide hydrosols, light production at low temperatures by catalysis with, 271 Meyer, G. M. See LEveNE and Meyer, 599, 623, 627 Milk production in relation to quality of proteins, 445 Mineral excretion of dogs, effect of hydrochloric acid, 461 Minot, A.S. See Denis and Minor, 561 Monocarboxylic sugar acids, re- lation between configuration and rotation of, 623 mag- THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXI, NO. 3 660 Mononucleotide, guanine (guan- ylie acid), and its preparation from yeast nucleic acid, 47 Mononucleotides, purine, structure, 337 Morriss, W. H. See Henperson and Morriss, 217 Morsr, M. The proteoclastic tis- sue enzymes of the spleen, 303 Moutrton, C. R. The availability of the energy of food for growth, 389 Muscular tissue, function in urea formation, 487 — —, glycolytic properties of, 501 , quantitative estimation of dextrose in, 67 NEPHRITIS, blood lipoids in, 575 Nerve roots, intradural, and corpus callosum (man and dog), rela- tive amount of sheathing sub- stance in, 395 Nervous system, chemical differ- entiation of, 395 Nitrie acid, removal from solutions of organic compounds, 599 Nitrogen balance in dogs on low protein diet, influence of small amounts of cystine, 363 Nitrous acid, effect of temperature on reaction of lysine with, 527 Nucleic acid, tritico, 295 —— ——,, yeast, preparation of “guanine mononucleotide from, 47 —— —— ——,, structure, 591 Nutrition investigations upon cot- tonseed meal, 379 ()SBORNE, T. B., and MEenpDEL, L. B. The réle of vitamines in the diet, 149 OsterHouT, W.J.V. The dynamics of the process of death, 585 Index PAPAIN, proteolytic activity of, 201 Peters, J. P., Jr., and Grye.in, H. R. The relation of adren- alin hyperglycemia to de- creased alkali reserve of the blood, 471 Phenol production, bile, 255 Phenylhydrazides, 623 Pirz, W. See McCetivum and Pirz, 229 Plasma and whole blood, determi- nation of CO. combining power, 217 Protein, influence of small amounts of cystine on nitrogen balance of dogs on diet low in, 363 minimum, 379 Proteins, nutritive value of di- amino-acids in, for mainte- nance of adult mice, 173 ——, relation of quality of, to milk production, 445 Protevclastic tissue enzymes of the — spleen, 303 Proteolytic activity of papain, 201 Purine mononucleotides, structure, Sau Pyrogallol, light production during oxidation of, 311 influence of AFFINOSE, physiological be- havior of, 125 Reap, B. E. Guanine mononucleo- tide (guanylic acid) and _ its preparation from yeast nucleic acid, 47 —— and TorrinGHaM, W. E. tico nucleic acid, 295 See Jonrs and Reap, 39, 337 Ricuarpson, A. E., and GREEN, H. S. Nutrition investiga- tions upon cottonseed meal. III. Cottonseed flour. The nature .of its growth-promot- ing substances, and a study in protein minimum, 379 Tri- Index Rosertrson, T. B., and DELPRAT, M. Experimental studies on growth. IX. The influence of tethelin upon the early growth of the white mouse, 567 Ross, E. L. See McGuican and Ross, 533 ALTS, effect on ammonia forma- tion in soil, 411 ——, similarity of action on swell- ing of animal membranes and of powdered colloids, 343 ’ Seurvy, experimental, 229 Sepewick, J. P. See KinGsBuURY and SzepewIick, 261 Serum of blood, tables for finding alkaline reserve of, in health and acidosis, from total CO: or alveolar CO, or pH at known CO, tension, 519 SuepLtov, A. See McCuienpon, SHEDLOV, and THOMSON, 269,519 Suerwin, C. P. Comparative me- tabolism of certain aromatic acids, 307 Soil, effect of salts on ammonia for- mation in, 411 Spleen, proteoclastic tissue enzymes of, 303 STEENBocK, H. See Hart, Hat- PIN, and STEENBOCK, 415 STEHLE, R. L. A study of the ef- fect of hydrochloric acid on the mineral excretion of dogs, 461 Sugar acids, epimeric monocarbox- ylic, relation between configu- ration and rotation of, 623 —— of blood, determination of, in reference to its condition in the blood, 533 Sulfur metabolism, 363 Sure, B., and Hart, E.B. The ef- fect of temperature on the re- action of lysine with nitrous acid, 527 ‘TETHELIN, influence on early growth of white mouse, 567 661 Tuomson, W. See McCuienpon, SHEDLOV, and THomsoNn, 269, 519 Tissue enzymes of the spleen, pro- teoclastic, 303 ——, muscular, function in urea formation, 487 ——, , glycolytic properties of, 501 . , ——, quantitative estimation of dextrose in, 67 TottTineHam, W. E. TOTTINGHAM, 295 Tritico nucleic acid, 295 See ReaD and URACIL-cytosine dinucleotide, 39 Urea formation, function of mus- cular tissue in, 487 Urie acid content of the blood of new-borns, 261 in urine and blood, modi- fications in colorimetric deter- mination of, 165 AN SLYKE-McLEAN method for determination of chlorides in blood, modification of, 483 Vitamine hypothesis and deficiency diseases, 229 Vitamines in the diet, 149 WARNER, D. E., and Epmonp, H. D. Blood fat in domestic fowls in relation to egg produc- tion, 281 Weiss, M.L. See GREENWALD and WEIss, l West, C. J. See LEvVENE West, 635, 649 Wheat or maize kernel, behavior of chickens restricted to, 415 Wiper, R.M. Intravenous injec- tions of B-hydroxybutyrice and aceto-acetic acids, 59 and YEAST nucleic acid, preparation of guanine mononucleotide from, 47 , structure, 591 7 nN QP The Journal of biological 501 chemistry J77 vol.31 COPer Biological & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY ap ed} ew ot sy », 2, ee Sige (cei aig 4 rH Hh - '