? pear e te $S485 pos ast! an a . ay eee ca s- Saat pled sees fcotine: Par tas tr Liste IE eee i Caan et Oe ah bp oar aeh Serer: rs F : an roe axe bis Fe ts) aereriy. peee 7 oy bes eric) heath erank Shee ks sere itate oe = ~a e Re hn : Pt rereeh: Gleneneenyem: wie pester coe eee : een br at as : ery Re ve Satrechgac hes eS enh na oh os, Shideeppaisned che a rte weed vet sats PAN o te ihe te Gs Eee te stay Soe aeteee eter ra quiets segs tae Fowl nge tea ge Fe, teaeh Reecadensa gts 5 Fae eter bow ee ree zi eat She sfc tC 1 Boe hat Match etdes een -sokahare by barat: ttre acest 2 gets vane we yaa Seat Es Pe steten ese ease « a gee cave. Scala dani aeg eeett sees co Beets raheshio be Srereeds aus eerrner ls} Sueiaeeeeces y abe b- #9 = sheet t Art| peeves a Oe) ome: 4 . =e reyees repre ‘ ’ tapierecbiate ¥ Shon srses F183 a35% mY - i ma ah die MAT ey 7) 2 nd : mb teat - Sietramery seeks 4 he dede » ti iets ree? OF y Aes + = he pins \.G Oren Theo ui ~ny eerie, “ . +s ihetoy mp ptf ht rt Wea oe teterts |, pbberer ri) Bitar 7 Maen Gna ere Ke fhe , BS MU Viranrvens bya at . epeepeetrry pt z * De¥ v Be cades caeanee Messees BISA! SeulFadeen'd cdews y ta eke a baal Vaete Si aR ew As? oe tt Bae che ee welt Tory % ; ane Maen a Maen Mee! reat) ss Wet-b-atir tahoe, ade shay iecae prep) Sra 062 2" oe eae ts 4 Srhkideteracpr ase? « ehiVab bats he-te Som Eotyake: rositink-ath Ce histeoer ware Alga Ades y= veer wee. & Spas rir 1 a hyveitett? fete iAdy on Lperdededeer peste ee SMebstsdrgntes 0.914 =A Poh: tine: re oe : Eek, sane Ve bese Sinha ne ligi tea ry Rite: a Ps. “3 tn 24> pase oem ay ets, gonath “pis diese pyc rs ¢ es ore we Sehads Area F bast h, Mane Rois Pyetocsarrerst ite lacs reat Usiegviw 7 ot an Dea tede vi, audi Usisadeie ‘ ay Om gis Sang va i meray ee: Pestanet wag tine ry Eee Tet paittestonts % SATS ee Etats Ad Crreses AG pas 8 Ge AS > peeve perresen) Pe “> Such aens aise teletoaean tars ~t -% 145 2 Ee kote ie Le nee arte Be pana ee ~~ Paton athe Pathe Pact deiqe A pe Ss = oe ate ake aPricache S porte ea Setedjee +h Bigeye Picea Renate Chee <> preterit. . oy bes oe Ee Ge ~ orc, Vraretereren, eats Ly feta « “ Pan ae yee ADAP NTA 3 eee er ea ies 4 a) a4 a4] te Ta , —— THE JOURNAL OF BIOLOGICAL CHEMISTRY FOUNDED BY CHRISTIAN A. HERTER AND BUSTAINED IN PART BY THE CHRISTIAN A, HERTER MEMORIAL FUND CFFICIAL ORGAN OF THE AMERICAN SOCIETY OF BIOLOGICAL CHEMISTS EDITED BY STANLEY R. BENEDICT, New York,N. Y. LAFAYETTEB. MENDEL, New Haven, Conn. HENRY D. DAKIN, Scarborough, N. Y. DONALD D. VAN SLYKE, New York, N. Y. WITH THE COOPERATION OF THE EDITORIAL COMMITTEE OTTO FOLIN, Boston, Mass. WALTER JONES, New York, N. Y. L. J. HENDERSON, Cambridge, Mass. GRAHAM LUSK, New York, N.Y. ANDREW HUNTER, Toronto, Canada THOMAS B. OSBORNE, New Haven, Conn. WALTER W. PALMER, New York, N. Y. A. N. RICHARDS, Philadelphia, Pa. L. L. VAN SLYKE, Geneva, N.Y. VOLUME XLIX BALTIMORE 1921 447) Copyricur 1921 BY THE JOURNAL OF BIOLOGICAL CHEMISTRY oe —— ote - 4 "S$ -peelee { ata PUB LIBHED BY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH FOR THE JOURNAL OF BIOLOGICAL CHEMISTRY, INC. WAVERLY PRESS Tax Witvtams & Wiixins Company Ba vtimorg, U.S. A. ' 3 é ; CONTENTS OF VOLUME XLIX. No.1. November, 1921. Van SLYKE, Donatp D., and Stapie, Witt1amC. The determination penn Carte DIOOd . ...; . . sc... s,s eae nana dees ee ee 1 Srapie, Witt1am C. A mechanical shaker and other devices for use with the Van Slyke blood gas apparatus... /..............-.¢.0: 43 Netson, Erwin E., and Greenr, CuHartes W. The chemical com- position of the ovaries of fresh water gar, Lepidosteus............ 47 GREENE, CHARLES W., and Netson, Erwin E. The chemical com- position of the skeletal muscle of the fresh water gar, Lepidosteus. 57 OsBoRNE, THomMAS B., WAKEMAN, ALFRED J., and LEAVENWORTH, Cuartes S. The proteins of the alfalfa plant................... 63 Howe, PautE. The use of sodium sulfate as the globulin precipitant in the determination of proteins in blood......................-- 93 Howe, Paut E. The determination of proteins in blood—A micro ane HRNRICH ne PEE... SE es ues ews ¢ see oS : ee 109 Howe, Patt E. An effect of the ingestion of colostrum upon the com- position of the blood of new-born calves....................05- 115 Fiemine, Witt1AM D. Vitamine content of rice by the yeast method. Organic nitrogen as a possible factor in the stimulation of yeast... 119 WitzeMANN, Ep@ar J. The catalytic effect of ammonia on the oxi- dation of butyric acid with hydrogen peroxide................... 123 Suarrer, Poitier A. Antiketogenesis. III. Calculation of the keto- genic balance from the respiratory quotient.................... 143 Fiske, Crrus H. Observations on the ‘“‘alkaline tide’’ after meals. I. 163 Fiske, Cyrus H. Inorganic phosphate and acid excretion in the post- 0 en Re aa 9 cS cS 171 MclIivarye, T. C. A buffer solution for colorimetric comparison.... 183 Jones, Mantua R. The calcium content of blood plasma and cor- go os: LIV DEES 2 ne ogee A 187 Buatuerwick, N. R. Observations on blood fat in diabetes........ 193 See WW ormnaspieraies A... ee Se cee 201 Devprat, G. D., and Wuiprie, G.H. Studies of liver function. Ben- zoate administration and hippuric acid synthesis................ 229 Buunt, Karuarine, Netson, Atta, and Oteson, Harriet Curry. The basal metabolism of underweight children................... 247 “No.2. December, 1921. Noreaarp, A., and Gram, H. C. Relation between the chloride con- tent of the blood and its volume per cent of cells................. 263 ili 1V Contents Gram, H.C. A new method for the determination of the fibrin per- centage in plood andeplasman..e ees oes os. hee 279 Couture, J. B. A further study of the respiratory processes in Mya arenaria-and, other marine mollupea.;\.). aes ees 8s vee se eee 297 Das. VW. oultates!vn« bloodigtry...csrieetcs cinerea aie vs selon eacra ee 311 Gipson, R. B., and Martin, Frances T. Some observations on crea- tine formation in a case of progressive pseudohypertrophie mus- CUllan Ty sbEO PMY 2,..o:<.ge sc eosin SOS ae Een ine raee 319 Dunn, Max S., and Lewis, Howarp B. The action of nitrous acid on COISCLIN MS Abs 25. lative wig «0.4/6.8 cee eae S25 thy of ROE RRR TTT Scene enn eee BPA Dunn, Max S., and Lewis, Howarp B. A comparative study of the hydrolysis of casein and deaminized casein by proteolytic enzymes. 343 HuspsBarp, Rocrer 8. Note on the determination of 6-hydroxybutyric Husparpd, Roger 8. Determination of the acetone bodies in urine... 357 Hupsarp, Roger 8. Determination of the acetone bodies in blood.. 375 Hvssparp, Roger 8., and Wrieut, Froyp R. Blood acetone bodies after the injection of small amounts of adrenalin chloride......... 385 Casori, F. A. Some nutritive properties of nuts. Il. The pecan nut hi as a source of adequate protein. ..44..25 J2s2a56 oe ee eee 389 Saiptey, P. G., McCotium, E. V., and Srmmonps, Nina. Studies on experimental rickets. IX. Lesions in the bones of rats suffering from uncomplicated beri-beri. Plates 1 to 4................... 399 Keeton, Rogert W. Ammonia excretion following experimental administration of acids via the stomach and peripheral vein..... 411 Wana, Cut Cor. The composition of Chinese edible birds’ nests and the nature of their proteins: «.......0ists.c oaks hee ie 3 ee 429 Wane, Curt Cur. The isolation and the nature of the amino sugar of Chinese edible. binds) nests... . |. aoe e en ee eee 441 Lusk, Grawam. Animal calorimetry. Eighteenth paper. The behavior of various intermediary metabolites upon the heat PEOGUGHLON.. ..... Aeadcline in ee Volume of sbeaieea cette OQ: seaboreD after Not NasOOs chamber | untrapped |} from 2nd left in SOLA CUS station over Hg solution | extraction | untrapped padi zed when COz2 left in ° solution was re- chamber. | untrapped | after Ist J eed placed by solution. /extraction.*| | ¢—d 100 = cc ce. cc cc. ce. ce. ag fell 1 0.672 50 0.17 0.002 0.002 0.000 0). 0 2 0.672 50 0.25 0.004 0.003 0.001 0.1 3 0.672 25 0.29 0.004 0.004 0.000 0.0 4 0.672 5 0.20 0.005 0.003 0.002 0.3 5 0.672 by 2.50 0.033 0.030 0.008 0.4 6 0.672 2 0.20 0.007 0.002 0.005 Oar ai 0.672 1 0.16 0.012 0.002 0.010 1.5 8 0.672 1 0.10 0.012 0.CcOL 0.011 1.6 9 0.672 0.78 0.18 0.014 0.003 0.011 1.6 10 0.672 0.78 0.20 0.014 0.003 0.011 6 11 0.672 0.74 0.18 0.013 0.003 0.010 1.5 ips 0.336 0.40 0.10 0.007 0.001 0.006 1.8 13 0.336 0.40 0.2 0.009 0.0016 | 0.0084 2.2 14 0.336 0.40 2.50 0.024 0.018 0.006 1.8 15 0.336 0.40 2.50 0.023 0.018 0.005 ; 1.5 / om coz * Calculated as a ¢__ A + (a CO == 1) S the micro-apparatus (Van Slyke, 1917) in the original or an en- larged form. In such an apparatus the water solution may be entirely separated from the gas phase before the vacuum is re- leased, and reabsorption of CO, therefore made impossible. We have found, however, that with care to prevent oscillation of the mercury-water layer after atmospheric pressure is reached, such constant results are obtained with the simpler, usual apparatus and technique (see Table I), that we prefer to continue its use, merely introducing the factor 1.017 to correct for the reabsorp-_ D. D. Van Slyke and W. C. Stadie 29 tion. This correction involves a change of about 1 volume per cent in the results of an ordinary blood or plasma CO, determination. 7. Calculation of CO, Results—In deriving the formula by which results were calculated in the original paper (Van Slyke, ¢ S 1917, p. 358) the factor an aco, Was used to calculate the frac- 5 tion of the gas remaining dissolved in the water when equilibrium was reached in extraction. (S = cc. of water solution, 50 = ce. volume of extraction chamber, a’co, = ec. of COs, measured at the prevailing temperature, dissolved by 1 ec. of water in equilib- rium with CO, gas at 760 mm. tension.) This factor repre- sents an approximation that is entirely exact only when a’co, = 1, which occurs at 18°. The deviation between the above approximate solubility correction factor and the exact factor increases, (1) as a’co, becomes greater or less than 1, and (2) ME ee nas ‘ as the ratio z increases. For the conditions under which the 5 approximate factor was used, however, (temperature range = S < ’ 15\= 30°55 = 0.05) the error introduced does not exceed 1 part per 1,000. Consequently the numerical factors arrived at by the original approximate equation do not require correction for the sake of accuracy. It seems, however, desirable to present the exact general equation which expresses the relationships between gas and liquid phases under conditions such as those prevailing in the apparatus. V, = volume of CO, obtained by one extraction and measured at atmospheric conditions of ¢° temperature and B mm. barometric pressure. Vor, 70 = total volume of CO:, reduced to 0°, 760 mm., in the solution analyzed. S = volume of water solution in apparatus. A = volume of chamber occupied by gas and solution during ex- traction (50 ec. in our apparatus). A — § = volume of gas phase during extraction. = absolute temperature, = ¢ + 273. aco, = solubility coefficient of CO, in water, the cc. of CO: measured at 0°, 760 mm., dissolved by 1 cc. of water in equilibrium with CO, under 760 mm. tension. 30 Blood Gases fi ae 2 a'co, = aco. X 5737 distribution coefficient of CO. between gas and water = cc. of CO: measured at t°, B mm. dissolved by 1 cc. of water in equilibrium with pure CO, at #®, Bmm. (a’co, was referred to as aco, in the original paper.) w = vapor tension of water. p = partial pressure of CO, in apparatus when equilibrium is reached in the extraction. x = volume of CO, gas, measured at 0°, 760 mm., held in solu- tion when equilibrium is reached. The total CO, content of the solution analyzed is obtained by reducing the volume of CO; extracted to standard conditions by (B — w) 273 760 T the volume thus corrected the volume zx of CO. remaining in solu- tion. Thus: multiplication with the usual factor , and adding to Baw. 973 7600 * T (1) Vo; 70 = Vt x Since the volume of gas dissolved is proportional to its partial pressure, solubility, and the volume of the solvent, we have aD (2) & = 76g Sco: Since pressure varies inversely as gas volume (8) Pa ve Bw Aas oe 7 (4) p= (Bw) fea Substituting this value for p in (2), and the value thereby found for x in (1), we have B-w 273 ya —w Saco: ©) Vor, veo = Vi reg oY “zag, ee . ~B-—w/273 cen.) 6) Vor, 160 = VieZ69 = A-S, D. D. Van Slyke and W. C. Stadie 31 ; 273 : Since %o, = co, X TP? equation (6) may be expressed as _ B-—w. 2% S (7) Vo, 760 — eee 760 x ar € i reece or B-w S a (8) 0°, 700 =" 769 (1 + 0,00367 H\1 TA —S 2 C0: FO oO a Factor correcting for Factor correcting atmospheric pressure for unextracted and temperature. CO;. In the approximate equation used in the original paper the ; 1 factor correcting for unextracted CO, was ee When 1 — 5 %co: S @'co, = 1 (at 18°) this becomes identical with 1 is 3 a’ cos A both factors then reducing to 4 aie .The values of the combined factor eae : 760 (1 + 0.00367 #) ( + Tats aco.) multiplied by 1.017 to correct for reabsorption of CO, during release of the vacuum, are given in Table XIII at the end of this paper for the calculation of results. Expressing B-—w 760 (1 + 0.00367 ¢) to be multiplied by as f, we have given its values for B = 760, s for other barometric pressures. Be- 760 cause the vapor tension, 1, is not quite the same fraction of B — w when B is other than 760, this usage is not absolutely exact. At a barometric pressure B = 740 it introduces an error of plus 0.1 per cent, at 700 mm., an error of plus 0.4 per cent. For barometric pressures outside the range 740 to 780 mm., therefore, one must use the customary tables for values of B-—w 760 (1 + 0.00367 ¢) per 1,000. For work at ordinary altitudes, however, the factors as given in Table XIII are sufficiently exact. in order to avoid errors exceeding 1 part ao Blood Gases Determination of Carbon Monoxide. In the method of Van Slyke and Salvesen (1919) it appears slightly more accurate to use 1.36 volumes’per cent as the correc- ~ tion for nitrogen gas rather than the 1.2 volumes per cent correc- tion found by the above authors. Still more accurate results may be obtainable by absorption of the CO with ammoniacal, cuprous chloride solution as recently described by O’Brien and Parker (1921). : In the determination of carbon monoxide it is advisable to use the finer bore apparatus and to magnify the gas volumes by re- ducing the pressure as described at the beginning of this paper, since the amounts of gas measured are likely to be much smaller than those measured when carbon dioxide or oxygen is determined. The blood should be trapped in the lower bulb of the apparatus before releasing the vacuum, as it 1s undesirable to mix blood with pyrogallo! and cuprous chloride. Determination of Methemoglobin. In the determination of methemoglobin (Stadie, 1920) the total blood pigment is determined colorimetrically, and the methemo- globin is estimated by subtracting from the total pigment the oxyhemoglobin. The latter is determined by the oxygen capac- ity method. The oxygen capacity portion of the methemoglobin determination therefore requires revision in its method of calcu- lation, as outlined above. Instead of utilizing Table I of Stadie’s paper the oxygen capacity of the blood is determined and eal- culated as described above, to make complete allowance for the nitrogen gas content of the blood. The volume per cent of oxygen is then multiplied by 0.746, the number of gm. of hemo- globin that combines with 1 ec. of oxygen. The result is the number of gm. of hemoglobin per 100. ec. of blood. The change in the mode of calculation affords sufficient increase in accuracy to be justified, but the error involved in the former method is too slight to invalidate any results that have been obtained by it. The absolute error is 0.5 volume per cent in oxygen capacity (equivalent to 0.37 gm. of hemoglobin per 100 ec. of blood) or about 2.5 per cent of the amount normally pres- ent. In the methemoglobin calculation this error is partly neutralized by the fact that the same percentage error is intro- D. D. Van Slyke and W. C. Stadie 33 duced into the colorimetric estimation of total pigment by mak- ing the standard methemoglobin solution from blood in which the hemoglobin had been estimated by the same oxygen capacity method. For example, in Table IX the model calculation given on page 240 of Stadie’s paper (1920) is reproduced together with the calculation as corrected in this paper. It is seen that the change in methemoglobin calculated is but 0.1 gm. per 100 ce. of blood. TABLE IX. Methemoglobin Calculation. |Hemoglobin per 100 ce. of blood. | Calculated as in original paper (Stadie, 1920). Calculated as described in this paper. Hemoglobin strength of standard blood ealcu- lated from oxygen capacity................... 15.0 14.6 Total blood pigment colorimetrically deter- paren — 4+ Of stamdard.............:....ce. 5 Hemoglobin determined by oxygen capacity.... 10.0 Methemoglobin by difference................... 2.5 Determination of All the Gases in One Blood Sample. Both carbon dioxide and oxygen, as well as the residual nitrogen, may be determined in one sample of blood with but little more expenditure of time and effort than is required to determine any one of the gases alone. For liberation of the gases both acid and ferricyanide are added to the blood in the apparatus, and all the gases are extracted and measured together. The COs, is then absorbed by dilute alkali, leaving O. and No. The O. may then be determined by absorption with pyrogallol, or estimated by subtracting the average N» content of blood from the sum of O. + Na». The essential point in determining the gases together was found to lie in the use of proper amounts of the reagents employed to set free the carbon dioxide and oxygen, particularly in the use of & minimum amount of acid. If potassium ferricyanide and a large excess of acid are added to blood, both oxygen and carbon dioxide are quantitatively freed, but the ferricyanide forms such a heavy precipitate with the blood proteins that it is inconven- THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 1 34. Blood Gases ient, although possible, to handle the resulting suspension in the apparatus. The precipitation of the proteins by ferricyanide occurs, however, only when the reaction is strongly acid. If but the minimum amount of lactic acid necessary to free the CO, with certainty is added, the ferricyanide-protein precipitate formed is so small in amount and so finely divided that it does not interfere with the determination. Furthermore, the acid can be trapped in the lower bulb of the apparatus before the blood is admitted, and can then be mixed with the blood after the appa- ratus has been evacuated. The blood is then in the lower part of the evacuation chamber, and such precipitate as does form does not touch the measuring tube at the top, which remains clean and clear. The precipitate, unlike the black deposit formed by action of ferricyanide on the mercury in the ammoniacal solution TABLE X. Amounts of Acid Required to Free COz in Human Whole Blood. Lactic acid added to 1 ce. of blood. CO: found in blood. millimols vol. per cent 0.05 41.5 0.10 52.8 0.10 53.3 0.20 52.8 used in the original oxygen determination, is instantly soluble in 0.1 N alkali. Consequently any particles that adhere to the walls of the apparatus are removed as a rule by a single washing with dilute alkali solution. The minimum amount of acid that will entirely free the CO. of whole blood for extraction under the conditions of our analyses was determined by preliminary experiments, some of which are recorded in Table X. The amount needed was found to be 1.0 ec. of 0.1 N lactic acid for 1 ec. of whole blood. Half as much fails to set the CO, completely free. The minimum amount of ferricyanide required in the pres- ence of acid is the same as in ammoniacal solution; vz., 5 mg. of potassium ferricyanide per 1 ce. of normal blood (Table II). In avoiding protein precipitation, however, the amount of ferri- cyanide present appears to be of much less importance than the D.. D. Van Slyke and W. C. Stadie 35 acidity of the mixture; if the acidity is too high, the minimum necessary amount of ferricyanide causes an inconvenient precipi- tate, while if the acidity is not too high an excess of ferricyanide may be used without such inconvenience. We consequently employ 10 mg. of ferricyanide per cc. of blood, twice the neces- sary amount. The analysis is performed as follows: Apparatus Used.—In order to obtain with 1 ce. of blood results for carbon dioxide and oxygen accurate to within 1 per cent of the amounts measured, it is desirable to use the fine bore type of apparatus described at the beginning of this paper, although, with care, results sufficiently accurate for many purposes may be obtained with the ordinary apparatus. Details of Determination.—First 2 cc. of 0.05 n lactie acid are admitted into the chamber of the apparatus. The acid is freed of air by evacuation and shaking in the usual manner. The air- free acid is then trapped in the bulb below the lower cock of the apparatus, and 1.9 cc. of water, with a drop of octyl alcohol, are similarly introduced and freed of air. The blood is now stirred and a sample drawn into an Ostwald pipette calibrated between two marks to deliver 1 cc. About 1.5 cc. of the air-free water in the gas analysis apparatus are run into the cup at the top, and the blood sample is at once run beneath it. We usually slightly open the cock below the cup while the pipette is draining, so that most of the blood flows directly into the chamber of the apparatus, the layer in the cup at no time being more than 2 to 3 mm. deep. The layer of water, even though it be somewhat acidified, prevents, because of the relatively slow rate of diffusion through it, the loss of CO, from the blood. All the blood, fol- lowed by the water layer above it, is now admitted into the cham- ber of the apparatus. When about half the water layer has been run in, the half remaining in the cup is stirred with a rod in order to get into suspension a few corpuscles that have lodged on the bottom of the cup. 0.05 cc. of solution containing 20 gm. of potassium ferricyanide per 100 cc. is added to this last portion of water, which is then admitted into the chamber. The 0.05 ce. of ferricyanide solution may be conveniently measured as 1 drop from a pipette, which has been found by trial to deliver thus 0.05 to 0.06 ce. of the solution. 26 Blood Gases. The chamber is evacuated until the mercury has fallen to the 50 cc. mark. The 2 ce. of 0.05 N lactic acid trapped in the lower bulb of the apparatus are now admitted and mixed with the rest of the solution. At this moment a small amount of brown precipitate forms, but not enough to interfere with the subse- quent manipulations. The oxygen and carbon dioxide (and car- bon monoxide if present) are extracted by whirling the solution about the wall of the evacuated chamber until, when the vacuum TABLE XI. Comparison of Carbon Dioxide and Oxygen Contents of Normal Venous Blood Determined Separately by Former Methods, and Together by the Present Combined Method, Respectively. Blood kept under paraffin oil. All determinations made within 3 hours after blood was drawn. - Volume of Potassium Re Method. blood for CO2 O2+ Ne. O2 ferricya- Hoe analysis. nide used. vol. vol. vol. per cent | per cent | per cent* O: alone. | Van Slyke, 1918. 2 211 19.6 40 On >“ cy ab etnias U2 [sis 2 20.9 19.4 40 CO, “ Wi co OLE 1 65.1 CO, “ i ae, LOL Ge 1 65.5 CO, “ e Sol Oz 2 65.0 O2 + COs. | Present. 2 64.4 20.5 19.0 10 Oz + GOs. a 2 65.0 20.9 19.4 10 O2 + COz e 2 63.5 20.9 19.4 15 O2 + CO, oe 2 64.4 20.9 19.4 30 O2 + CO, ss 2 64.2 20.9 19.4 70 * The Ne was determined and found to be 1.5 volumes per cent. It is deducted and subtracted from the O» + Nz to estimate the Os. is released and the gas measured, no increase in volume is observed. Complete extraction is usually attained in 1 minute when the apparatus is shaken by hand. When the mechanical shaker devised by Stadie (1921) has been used, 2 minutes shaking after the first evacuation, followed by 30 seconds shaking for the check evacuations, has been our usual routine. As a rule no increase occurs after the first evacuation. Increase after the second indicates a leak in the apparatus, due usually to improper lubrication of the upper cock. —_---S -——— D. D. Van Slyke and W. C. Stadie 37 When extraction of the gases is complete the water solution is trapped below the lower cock, as in the original CO, determina- tion, and the gases are measured at atmospheric pressure. The gases measured are COs, Oz, Neo, and sometimes CO. The CO, is absorbed by running 0.5 nN NaOH from the cup down the side of the measuring tube until no further shrinkage of the gas volume occurs; about 0.5 cc. of the hydroxide solution suffices. TABLE XII. Comparison of Carbon Dioxide and Oxygen Contents Determined Separately by Former Methods, and Together by the Present Combined Method. All determinations made on 1 cc. samples in a fine bore apparatus Oxygen. CO, | Method. Former. | Present. | Former. | Present. vol. vol. vol. vol. per cent | per cent | per cent | per cent Blood aerated with alveolar air........ 18.1 17.9 58.7 58.2 “Sedmuted with saline.............. 9.4 8.8 38.7 39.5 Venous blood about 60 per cent satu- oTdita) yale, Oa eee 13.6 13.9 35.8. 36.4 0.02625 m Na.CO; (calculated 58.8 vol- HIMESH CCHEOOs). foi 52-508 eee eee | 58.7 59.1 | As the volume, S, of solution in the apparatus is twice as great as in the method as originally described, the factors used for calculation of carbon dioxide are different. For the present conditions, with 5 cc. of total water solution extracted in the apparatus, the factors are given in Table XIII. If no CO is present, the residual mixture left after absorption of CO, may be measured as O. + No, and from the O2 + N: con- tent reduced to 0°, 760 mm. (by Table XIII), the average N» content of blood, viz. 1.36 volumes per cent, may be subtracted in order to estimate the oxygen; or the oxygen may be measured - directly by absorption, as described by Van Slyke and Salvesen. In the latter case about 0.5 cc. of pyrogallol solution is run in from the cup, and is followed by a little water to clean the tube and give a clear meniscus for reading. 28 Blood Gases TABLE XIII. Factors for Calculation. S = 1.017 f (i + 50— 8 " coy)s Bw Air, { measured at} factor by which the volume of CO:z 760 (1 + 0.00367 t) room temperature obtained after 1 extraction is Tem- | factor by which and pressure, multiplied in order to obtain the pera- gas measured aco dissolved by volume of COz, reduced to 0°, ture. moist at ¢°, : 760 mm., contained in the B mm. is reduced solution analyzed. to 0°, 760 mm.* He a S = 2.5 ce. S = 5.0 ce. °C cc. cc 0.982 X 7 1.002 x — | 1.061 x — 15 * 760 | 1.075 | 0.052 | 0.105 760 25 | 0.888 “| 0.828 | 0.043 | 0.086 | 0.942 “| 0.986 + 26 | 0.883 “| 0.808 | 0.042 | 0.084 | 0.936 «(O28 BS 27 =| 0.878 “| 0.789 | 0.041 | 0.083 | 0.931 “*. | 0298 ti 28 | 0.873 “| 0.772 | 0.040 | 0.081 | 0.924 “| 0.964 es 29 | 0.868 “| 0.755 | 0.040 | 0.080 | 0.918 | 0.957 “ 30 | 0.863 “| 0.738 | 0.039 | 0.078 | 0.912 “| 0.950 * * To calculate O2 or hemoglobin when O, + Ne volume is measured, multiply gas volume by f, to reduce to 0°, 760 mm., and by such factor as is necessary (100 when 1 cc. of blood is used, 50 when 2 ce. are used) to bring results to volume per cent basis. Then for a. O2 content, subtract..... 1.36 vol. per cent Ne b. O2 bound by hemoglobin in venous. blood, subtractes.s soo 15 “ “« “ N; + dissolvedis e. O2 bound by hemoglobin in arterial blood, subtract.:peaseeeeee ib i nd 7 d. Oz: bound by hemoglobin in blood saturated : with air at 20°, subtract kth, ee oe “ “ “cc “co “ “ D. D. Van Slyke and W. C. Stadie 39 TABLE XIlI—Concluded. Per cent of normal hemoglobin (Haldane scale) = —s = 5.41 d. Grams of hemoglobin per 100 cc. of blood = 0.746 d. : ; 100 b 100 ¢ Per cent of total hemoglobin saturated with O2 = en or a Volumes per cent O2 unsaturation = d —c ord — b. b and c may be determined with slightly greater accuracy with the aid of Table IV. The values for f given in the second column are for baro- metric readings corrected for temperature (see remarks on p. 31). The values for a’... are obtained by multiplying by 1 + 0.00367 t the values for a, given by Bohr and Bock (1891). { The dissolved air is given as measured at room temperature. It is subtracted from the air + CO: volume, measured after one extraction of plasma or aqueous carbonate solution, in order to obtain the CO, which ; pe bee S : : is then multiplied by 1.017 ( 1+ 50_S Be ) in order to obtain the total volumes per cent of CO, in the solution analyzed. When whole blood is analyzed, the air correction cannot be used, because of the O2 present, and the CO: must be determined by absorption with NaOH solution. The volume of gas absorbed is then multiplied by the above factor. The factor 1.017, being empirical (see p. 27), may vary slightly for different apparatus. If carbon monoxide is present, the oxygen must, of course, be absorbed by pyrogallol. The residual gas is the CO + 1.36 volumes per cent of Ns. The CO is calculated by reducing the mixture of CO and N; to 0°, 760 mm. (Table XIID), and sub- tracting 1.36 from the result in volumes per cent, or the CO may be absorbed by cuprous chloride (O’Brien and Parker, 1921). The results of some determinations are given in Table XI and XII. Examples of Calculations. 1. Total Oxygen Content of Venous or Arterial Blood. a. From Oz and Nz Measured Together. Blood sample.......... 2:00 Gee: O. + N2 measured..... 0.405 “ at 25°, 750 mm. 0.405 X 0.888* X = Ae ee eer 0.3534“ “ 0°, 760 “ 0.3534 X 50 = O.+N:2 per 100 cc. blood. 322s <= ance rea | Oe PEPER ENERO 000, aid o%in'e was 5s 2 os Poo ee aa Dee ET OTOR TPR... cw ceca ts see SGses Se AD ce * Factor from second column of Table XIII. 40 Blood Gases b. From Oz Determined by Absorption with Pyrogallol. Blood sample......... 2.00 cc. Oz + Ne measured.... 0.405 “ at 25°, 750 mm. N. after absorption of Oo... 200 eee OL0sL<* : * 25°. 7o0s a Oz by difference...... O:3T4- | 25° STO 0.374 X 50 = QO: per 100 ce. blood. 18:70 ‘“ “ 25°, 750 “ 18.70 X 0.888* X (20 = Ox 10055 wie "OF 760n aes 760 2. Oxygen Combined with Hemoglobin in Venous or Arterial Blood as Drawn. a. Venous Blood. Total O2, calculated as above.............. 16.37 cc. at 0°, 760 mm. O; invphysical’solution. |. cesses. eee Ono; “ “(OF Fern O2 combined with hemoglobin............. 1627“ 8 0? 7Gn ee b. Arterial Blood. Total O2, calculated as above.............. 16.37 cc. at 0°, 760 mm. ©, in physical solutions «,. ....<./\-nsaeue eee 0.3275 “O°, 7o0mee O. combined with hemoglobin............. 16:05..“* ,* OF; 760m 8. Oxygen Combined with Hemoglobin in Blood Saturated with Air at Room Temperature (Oxygen Capacity). a. From Oz and Nz Measured Together. Blood sample......... 2.00 cc. O: + Nz measured.... 0.495 “ at 20°, 767 mm. 767 0.495 X 0.910* X aA So) Ne a ae 0.455 “ “ 0°, 760 7 50 X 0.455 = O2.-+ Ne per 100 ce. blood... a eee 22.75. “ 0? S760 Physically dissolved O2 + N2 per 100 ce. blood She Gaiieees 2.10%» $9 00% OQ Combined O: per 100 cc. blood.. 20.65 “ “ 0°, 760 “ b. From O2 Determined by Absorption with Pyrogallol. ° Blood sample......... 2.00 cc. O. + Ne measured.:.. 0.495 “ at 20°, 767 mm. N». after absorption Of Osseree Pa ee ee 0.033 “ “ 20° RG ee O2 by difference...... 0.462 “ 20° 7a *Factor from second column of Table XIII. { These figures are approximate and are the average for arterial and venous bloods. For the accurate values of oxygen in physical solution see Table II]. —— D. D. Van Slyke and W. C. Stadie 41 767 0.462 X 0.910* X 760 xX 50 = Total Q2 per 100 ec. blocd Aes tes 21.22 cc. at 0°, 760 mm. Physically dissolved Oz per 100 cc. blood. 0.58 “ “ 0°, 760 “ Combined O2 per 100 ce. nlood sk oe ae 20: 64s) eS OF. F605 (4 4. Calculation of Hemoglobin Content from Oxygen Capacity. Combined O: per 100 cc. blood, calculated as above, 20.64 ec. at 0°, 760 mm. 20.64 X 0.746 = hemoglobin per 100 cc. LOO « oo. 8 eto Se = 15.40 gm. hemoglobin, 20.64 185 X 100 = per cent of Haldane’s aver- age normal hemoglobin....... = 112 per cent. §. Calculation of Oxygen Unsaturation of Blood. Oxygen combined with hemoglobin per 100 ce. of blood after saturation of blood with air at room temperature (oxygen capacity) 2 eS) ree 20.64 ce. at 0°, 760 mm. Oxygen combined with hemoglobin of blood as PlerytiE (ICWATIONEZ»(G)).,. a: ers oe ie BS D ° | z 5 oR a Ga 10 < < < 6) Az) A C27 | Lepidosteus platystomus. | 7.75 |13.52) 2.59/0.87| 0.43/0.094| 0.26) 75.3 C31 | Lepidosteus platystomus. | 2.23 |14.34 2.81/0.89} 0.40/0.077| 0.30) 79.7) 79.5 C37 | Lepidosteus platystomus. | 4.40 |14.02, 2.76\0.43) 0.30\0.064) 0.27) 78.3) 77.4 C35 | Lepidosteus | ~J] ~I cs osseus. 6.397|15.17| 2.66/0.85) 0.34/0.071| 0.28) 74.9) 71.7 C41 | Lepidosteus osseus. 7.75 |14.64| 3.17/0.92| 0.51/0.091| 0.26] 73.5] 71.7 42b| Lepidosteus osseus. 13.19 |14.45) 4.61 1.33) 0.31 0.066) 0.21) 66.4) cf * The calculations are in terms of parts per 100 gm. of moist sample. t Part lost. t Not determined. DISCUSSION. Lipovds. The total lipoids (ether-soluble fats) of gar muscle vary from 2.2 to 13.19 per cent, Sample 42b being much the richest in lip- oids. This sample came from a large specimen of Lepidosteus osseus from the Mississippi River at New Boston, Illinois, taken in November. All other samples are July and August fish which show relatively lower lipoid content. Clark and Almy (1918) have pointed out that the fat content of a number of the food fishes is highest in the autumn. 60 Skeletal Muscle of the Gar, Lepidosteus Protein. The protein content of the muscles of the gar is uniformly about 14 per cent, the average of the determinations being 14.35 per cent. This figure is very much lower than that for the ovary of this species, the average protein content of which is 25 per cent. When one recalls the activity of the muscle, and the relatively passive condition of the ovary, this is a somewhat surprising observation. The figure is quite comparable to that for the muscles of the spawning salmon, 14 per cent, but is considerably less than that for the muscles of the normal salmon entering fresh water, 20 per cent figured on a fat-free basis. If a protein storage occurs in the gar, it is evidently on a much lower level than in the salmon. Organic Extractives. The organic extractives are a measure of tissue metabolism. This is indicated by the werk of Hatai (1917), showing that the tissues of the growing white rat contain a greater proportion of extractives than those of the adult. Greene (1918) showed that during the period of migration of the king salmon, during which there is vigorous muscular activity, the organic extractives remain comparatively constant or actually increase, notwithstanding the decrease in the mass of muscle or in its diminishing percentage of protein. The water-soluble solids and the organic extractives, Table I, are quite uniform if Sample 42b, which is again an extreme, be left out of account. The average for the total solids, 3.10 per cent, is quite comparable to that given by Greene (1919) for salmon muscle (average of fifteen samples 4.05 per cent). The total non-protein nitrogen of the extractives varies from 0.30 to 0.51 per cent. Suzuki and Joshimura (1909) found for the tunny 1.09 per cent, a much larger value, and Buglia and Costantino (1912) report values of 0.47 and 0.57 per cent for certain teleosts of the Mediterranean. The amino nitrogen is fairly constant, varying from 0.164 to 0.194 per cent. C. H. Greene (1919) has shown for salmon muscle that the fluids are an important factor in determining the amounts of amino nitrogen held in the tissues. For the salmon, C. W. Greene and E. E. Nelson 61 when the fluids present have a content of 100 mg. of amino nitro- gen per 100 gm. of water, they act as though saturated, since the amino nitrogen shows no further change during the migration. When the figures in the present series are calculated per 100 gm. of water the following results are obtained: Sample. Amino N mg. Se eee 25 sss a SNe ae Vines 0 oe en uicls Goer Oe cite Meet 111 SO TET cree dsc ood a ae olde Dee oars De aeeeee 82 (Cede e sso 20 tad oe ee pa ees ra nena 98 ONT sn 565 oo SO te tee oe eee a Vo oem ere Fo 126 DEL pcoot aod 1) 0 SO ee eee oO eee or Aron ore eee oe 96 WG, oc spas cue. Oe eee ee waa Oke 100 ATER, 300.35 Se eer ae i219 ING SERRE es fr 102 The average figure of 102 mg. is in interesting agreement with Greene’s figures, though the variation is considerable. Creatine varies from 0.21 to 0.30 per cent, or again leaving out of account Sample 42b, from 0.26 to 0.30 per cent. Myers and Fine (1913) have pointed out that the value for any species is constant and specific, as far as the adults are concerned. They observed that with kittens the value increased with age. If 42b was a sample from an older fish than the rest, as has been assumed through the paper, the determination of creatine should show a higher value than the rest, not a lower one. No explana- tion is offered for this discrepancy. Suzuki and Joshimura (1909) found the following values for creatine: Fish. Creatine. per cent FERIA EMT Te So. 6a. edie aie tus 3 ete sie eros el es tueie eatie/ere sieges 0.100 DIRE onc oweo BAA n Dern earn ana aac 0.300 So. cca dbac sete nees sted sea 0.320 Okuda (1912) reports values from 0.42 to 0.75 per cent in various teleosts. C. H. Greene (1919) considers Okuda’s values too high. In the original series of analyses creatinine was determined, but in view of the fact that Grindley and Woods (1906) deny the presence of creatinine in perfectly fresh muscle (confirmed by Mellanby, 1908), and show that merely evaporation of the watery extract of muscle on the water bath is sufficient to convert a large amount of the creatine into creatinine, the determinations have not been included in the tables. Suzuki and Joshimura 62 Skeletal Muscle of the Gar, Lepidosteus (1909) in experiments on the isolation of various substances from the extractives of fish muscle, did not isolate creatinine. Kruken- berg (1881) who reports creatinine in fish muscle, was probably in error on this point. Food Value of Gar Muscle. "The preceding analyses show that the gar flesh compares favor- ably with that of those species which contain the moderate amounts of stored lipoids. The gar flesh contains from 2.23 to 13.19 per cent of total lipoids, an average of about 7 per cent. The caloric value of the total fats averages a slightly higher value than that of the total proteins. The proteins average 14.37 per cent for the six samples and there is but little individual variation in the series. The gar flesh is on the whole a very palatable food of good caloric value. BIBLIOGRAPHY. Atwater, W. O., Rep. U. S. Com. Fish and Fisheries, 1888. Balland, A., Compt. rend. Acad., 1898, exxvi, 1728. Buglia, G., and Costantino, A., Z. physiol. Chem., 1912, 1xxxtii, 437. Clark, E. D., and Almy, L. H., J. Biol. Chem., 1918, xxxiii, 483. Greene, C. H., J. Biol. Chem., 1918, xxxiii, p. xii. Greene, C. H., J. Biol. Chem., 1919, xxxix, 457. Greene, C. W., Bull. U. 8S. Bureau Fisheries, 1913, xxxiii, 69. Greene, C. W., Tr. Am. Fisheries Soc., 1915, xlv, 5. Greene, C. W., Am. J. Physiol., 1916-17, xlii, 609. Greene, C. W., J. Biol. Chem., 1919, xxxix, 435. Grindley, H. 8., and Woods, H.8., J. Biol. Chem., 1906, ii, 309. Hatai, 8., Am. J. Anat., 1917, xxi, 23. Hollande, A. C., Bull. sc. pharmacol., 1918, xx, 405. 1909, xviii, 497. Krukenberg, C. F. W., Jahrb. Tierchem., 1881, xi, 340. Lehmann, F., Allg. Fischerei-Ztg., 1900, xxv, 91; abstracted in Z. Un- tersuch. Narhungs- u. Genussmittel, 1900, 111, 475. Lichtenfelt, H., Arch. ges. Physiol., 1904, ciii, 353. Mellanby, E., J. Physiol., 1908, xxxvi, 447. Myers, V. C., and Fine, M. §., J. Biol. Chem., 1913, xiv, 9. Okuda, Y., 8th Internat. Cong. Applied Chem., Original Communications, 1912, xvii, 145. Paton, C. N., 18th Ann. Rep. Fishery Bd. Scotland, 1898, iv, 143. Polimanti, O., Biochem. Z., 1915, Ixix, 145. Suzuki, U., and Joshimura, K., Z. physiol. Chem., 1909, lxii, 1. Williams, K. I., J. Chem. Soc., 1897, Ixxi, 649. THE PROTEINS OF THE ALFALFA PLANT.* By THOMAS B. OSBORNE, ALFRED J. WAKEMAN, AND CHARLES 8. LEAVENWORTH. (From the Laboratory of the Connecticut Agricultural Experiment Station, New Haven.) (Received for publication, September 13, 1921.) From a purely economic standpoint the green forage plant is the raw material from which all production is primarily derived. From a scientific standpoint it is of the greatest interest because in its leaves both carbohydrates and proteins are synthesized. Thus, the chemistry of the living plant presents problems of both practical and scientific importance. It is true that we know a multitude of products derived from plants and we know much of the chemistry of these, but this knowledge consists mostly of isolated facts which contribute comparatively little to a knowledge of the chemical make-up of the plant as a whole. Less is known of the chemistry of the proteins of the living plant than of any of its groups of constituents. Rouelle! in 1773 announced that the glutinous matter, 7.e. protein, which up to that time was known to exist only in the seeds of wheat, was present also in other parts of various plants. Later, he separated this glutinous substance from the juice of hemlock by heating to a moderate temperature and filtering out the coagulum, the protein nature of which was proved by the products of destructive distillation. In 1789 Fourcroy? gave an extensive account of the occurrence of coagulable protein in the juices of various parts of many plants, and described the method by which he obtained prep- *The expenses of this investigation were shared by the Connecticut Agricultural Experiment Station and the Carnegie Institution of Wash- ington, Washington, D. C. 1Rouelle, J. méd. chir., pharm., 1773, xxxix, 250. 2Fourcroy, A. F., Ann. chim., 1789, iii, 252. 63 64. Proteins of Alfalfa Plant arations of what he supposed to be pure plant albumin. This observation of Fourcroy was the first to demonstrate the pres- ence of two kinds of protein in plants. A number of different investigators between 1799 and 1805 found albumin in the juices of many plants and in the sap of trees, but they did not add much to the information obtained by Fourcroy. After 1805 no more attention was paid to the proteins of green plants presumably because, after Einhof* undertook his investi- gations of the seeds of cereals, the relative ease with which pro- teins could be obtained from seeds emphasized the difficulties of their extraction from green plants. It is a very remarkable fact, and not at all to the credit of science during all these years, that, apart from microchemical observations no attempt had been made to study the protein constituents of living plants until we undertook our investigation of the spinach plant.‘ Various green forage plants contribute large amounts of pro- tein to the ration of farm animals but practically nothing is known of the chemistry of these, even the proportion of protein in these plants not being yet established. It is true that the agricultural chemist states the percentages of protein in his analyses of green fodders, but these are made by indirect methods founded on assumptions unsupported by satisfactory evidence. A serious gap therefore exists in our current knowledge of the chemistry of nutrition which makes it impossible to apply to the practical problems of feeding on the farm what has been learned of the nutritive value of the proteins of the cereals as well as of the protein concentrates. Special conditions are presented in dealing with the proteins of a living plant which deserve consideration. First: the microscope shows that most of the protein is con- tained in cells, whose walls must be ruptured before the indiffusible proteins can be extracted. Second: the living plant contains enzymes which rapidly convert one constituent into another and might be expected to profoundly alter the proteins before these could be extracted from the tissues and separated from solution. 3Hinhof, H., Neues allgem. J. Chem., 1805, v, 131; 1806, vi, 62. Osborne, T. B., and Wakeman, A. J., J. Biol. Chem., 1920, xlii, 1. Osborne, Wakeman, and Leavenworth , 65 Third: the living plant contains a large number of constitu- ents of many different types of known and unknown nature and these must be separated from the protein in order to obtain preparations of even roughly approximate purity. The investigation reported in this paper embodies attempts to apply the methods developed in studying the proteins of the spinach leaf to alfalfa. Further experience with these methods has led to improvements, whereby it now appears to be possible to learn not only much that is new respecting the proteins of alfalfa but also respecting other constituents about which little is known. By using suitable mills and presses relatively large quantities of the clear undiluted juice of the alfalfa plant were obtained, and also the water-soluble constituents were almost completely extracted within so short a time that autolytic changes were re- duced to a minimum. Chlorophyll, fats, phosphatides, etc., were removed by ex- tracting with cold alcohol and ether, and so obtained uncon- taminated with water-soluble substances which also might be soluble in these solvents. The residue was then extracted at room temperature with dilute aqueous NaOH solution, but it was found that, although this solvent readily dissolves nearly all known types of protein, only a small part of the nitrogenous substances was thereby removed. Nearly all of the residual nitrogen was, however, extracted by boiling for a few minutes with 60 per cent alcohol containing 0.3 per cent NaOH. Preliminary experience in developing these methods was gained during the course of the summer, so that it was late in the autumn before we were in a position to apply them satis- factorily on a relatively large scale. In order to continue this work during the winter alfalfa plants, cut about 3 months after sowing, were loosely packed in bags and placed in a cold room at —18°C. within 2 hours after cutting. The plants, thus quickly frozen, were kept at this low temperature until immediately before use. When removed from the cold storage room they appeared as fresh as when first cut. As their behavior during extraction was like that of freshly gathered plants, we believe that they had suffered no change sufficiently great to materially THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 1 1d) o hee al Proteins of Alfalfa Plant affect the results of the experiments described in this paper. Since the results of the experiments with these frozen plants were more complete than those obtained with freshly gathered plants we describe them first in order that the reader may more easily understand our work. The still frozen plants were put six times through the meat chopper and then the pasty mass was passed several times through a ‘“Nixtamal’’ mill. The pulp, which weighed 22 kilos, was - pressed in the hydraulic press, the press-cakes reground with 1,500 cc. of water and again pressed. This process was re- peated. The washed cakes were then ground and kept under alcohol until the next day. When the juice, obtained by directly pressing the ground plants, first ran out it was somewhat turbid and green from suspended chlorophyll grains. Later it ran free from chlorophyll and other suspended particles and was ‘collected separately. This latter portion, which weighed 5,415 gm., represented the un- diluted fluids of the plant. To this was added 18.7 per cent, by weight, of alcohol which produced a voluminous precipi- tate that was at once collected on large folded filters and allowed to drain out over night. The process thus far described was brought to this point within 6 hours. The part of the press-juice which was green and turbid weighed 7,144 gm. This was filtered during the night through felts of paper pulp, whereby it was freed from chlorophyll and other sus- pended matters. The next morning there was a slight flocculent precipitate in the previously clear brown filtrate; this was re- moved by centrifuging. The treatment to which these two parts of the alfalfa juice were subsequently subjected will be described later. The washed press-cakes, which had been brought into strong alcohol within 6 hours after the plants were first ground, were next extracted with about 9 liters of 93 per cent alcohol at room temperature three successive times and pressed out each time with the hydraulic press. After practically all of the alcohol-soluble constituents were thus removed, the press-cakes were ground in the ““Nixtamal’”’ mill with a liberal quantity of 0.3 per cent NaOH solution and pressed in the hydraulic press. After thrice treating in the same way, Osborne, Wakeman, and Leavenworth 67 the press-cakes were boiled for about 5 minutes with 60 per cent alcohol, containing 0.3 per cent NaOH, drained on filters, and the residue pressed in the hydraulic press. After repeating this treatment the press-cakes were finally extracted with 93 per cent alcohol and dried in a current of warm air. The green alfalfa plants were thus successively extracted with water, alcohol, dilute aqueous alkali, and with dilute alkaline aleohol. During these processes samples of extracts and residues were taken for analysis in order to determine the amount of solids, nitrogen, and inorganic matter removed at each stage. By allowing for the portions thus taken out we obtained the following data, which must be regarded as representing only the approximate proportions soluble in the different solvents. Ash-free solids. Nitrogen. Ash (CO>-free). gm. |percent| gm. |percent| gm. | percent ingihe alialfaetaken...2.......... 4 449 211.6 506 .4 PeMmmeTE WEED. 2... ke 1 898} 42.7 | 92.6 | 43.8 | 359.0) 70.9 So per cent aleohol.:............. 283) 6.4) 4.3] 2.0 9.3) 1.8 0.3 per cent aqueous NaOH...... 230) 5.2 | 14.8] 7.0 * 0.3 per cent alcoholic NaOH..... 761) 17.8 | 88.2 | 39.3 Wixtracted residue................ 1,292) 29.1 | 11.1 | 5.3 4 464/101 .2 206.0 | 97.4 * Satisfactory determinations of ash could not be made in the alkaline extracts, nor in the residue after these solvents were used. We thus find that over 60 per cent of the ash-free solids and 83 per cent of the total nitrogen were extracted by water and hot alkaline alcohol. The small proportion of nitrogen soluble in aqueous alkali is striking, and shows the presence of relatively little nitrogenous substance insoluble in water and having the solubility in alkalies characteristic of most native proteins. Although the hot alkaline alcohol dissolved only 17 per cent of the solids of the alfalfa plant, it extracted almost 40 per cent of the nitrogen, most, if not all, of which belongs to protein. The small proportion of nitrogen left in the residue shows that the cells were ruptured so completely that nearly all of the nitrog- enous substances were extracted by the various solvents. 68 Proteins of Alfalfa Plant The Water-Soluble Constituents of the Green Alfalfa Plant. The brown press-juice, obtained as described on p. 66, may be considered to be free from suspended solids because the chloro- phyll grains are so small and so difficult to separate by filtration that, if these are removed, as evidenced by the absence of any green deposit on centrifuging at high speed, appreciable quantities of other suspended matters cannot be present. This brown press-juice contains a considerable amount of colloids which renders the solution opaque in thick layers. By transmitted light the juice is translucent and has a deep yellow color; by reflected light it is dark brown. The fresh juice is strongly acid to litmus, 19.0 ec. of 0.1 N NaOH solution being required to make 100 cc. approximately neutral to litmus and 13 ce. more to give a perceptible blue tint. At this reaction a considerable precipitate separates (Ca3P,03?). The amount of solids contained in the juice undoubtedly varies with the state of development of the plant. One sample from plants gathered early in November, which had been sown August 19th, contained 9.26 gm. of solids per 100 cc. The juice of plants cut from the same field a few days later, and kept frozen for about 4 months, contained 10.52 gm. of solids per 100 ce. This somewhat higher content in solids is probably due, in part at least, to loss of water by evaporation during storage in the frozen state. After standing for a few hours the clear brown juice slowly deposits a precipitate, which ultimately becomes considerable. A relatively large precipitate separates at once on slightly acidify- ing with acetic acid. The ‘‘Colloid Precipitate.”’ As already stated, the colloids are precipitated by adding about 18 per cent by weight of aleohol. As the precipitate thus pro- duced contains several substances of pronounced colloidal prop- erties it will be hereafter referred to as the ‘‘colloid precipitate. ”’ Whether these colloidal constituents are chemically combined with one another, or are simply mixed together in the precipitate, has not been learned. The chief constituent is protein, but together with this is a considerable amount of caleitum phosphate Osborne, Wakeman, and Leavenworth 69 and of organic calcium salts, including one, or more, coloring matters which have not yet been identified. The colloid precipitate from about 12 liters of the clear brown juice of plants cut immediately before pressing was washed by centrifuging with water, 50 per cent alcohol, and finally with 74 per cent alcohol. Thus obtained in a highly hydrated state it was a compact gray-colored mass. In order to remove calcium, phosphoric acid, and other sub- stances soluble in acid alcohol, it was suspended in 5 liters of 74 per cent alcohol containing 0.1 per cent hydrochloric acid, centri- fuged, and treated four times more, each time with 3 liters of the acid alcohol. The united extracts, when made slightly alkaline to litmus with 1 per cent sodium hydroxide solution, yielded a large gelati- nous precipitate resembling calcium phosphate, which will be referred to later as the ‘‘neutralization precipitate.’’ This could not be removed by filtering, or by centrifuging, until heated to boiling. The clear, slightly alkaline filtrate and washings were then concentrated to 750 cc. in vacuo below 50° of the bath. Analysis of aliquots of this deep yellow-brown solution showed that the alkaline filtrate contained 20.66 gm. of solids, of which 16.3 gm. were NaCl, leaving only 4.36 gm. of other substances. It also contained 0.453 gm. of nitrogen. When this alkaline filtrate was made neutral to litmus with 0.1 n HCl, a turbidity formed which became a small flocculent precipitate on doubling the quantity of added acid. A part of the original solution when freed from alcohol and saturated with (NH,4).SO,, gave a slight precipitate which, however, gave no biuret reaction. In view of this fact, and also of the small amount of nitrogen in this solution, it is evident that little, if any, protein had been removed from the “colloid precipitate’ by the acid alcohol. The results of this extraction in terms of dry solids were as follows: gm. per cent Insoluble in 74 per cent alcohol containing 0.1 per cent ner ST eee ht te es ee. os 5 Shige te 72.0 75.9 eMeniralization precipitate”............-.......25.-. 18.4 19.4 Filtrate from ‘‘neutralization precipitate”’............ 4.5 4.7 TD MEDIA co's 5c RAO REISS cat tien RMP ea crs ote 94.9 100.0 70 Proteins of Alfalfa Plant It thus appears that about one-fourth of the “colloid precipi- tate” was soluble in the dilute acid alcohol of which four-fifths was precipitated at a slightly alkaline reaction. The nitrogen was distributed as follows: gm. per cent Insoluble in 74 per cent alcohol containing 0.1 per cent FICH. ee a eee 9.75 91.7 ‘Neutralization precipitate’... 77.50.05 aoe ola et 0.43 4.1 Filtrate from ‘‘neutralization precipitate’”’............ 0.45 4.2 Total cco silences «0 nd SR ee 10.63 100.0 According to these figures the dry matter of the original ‘‘col- loid precipitate’’ contained 11.2 per cent of nitrogen. The ash was distributed as follows: gm, per cent Insoluble in 74 per cent alcohol containing 0.1 per cent TACT. go nS ee oe eae 1.10 9.8 “Neutralization, pretipitate’”’ >. .7s. Smeets = a. toe 9.82 | 87.5 Filtrate (estimated by difference)....................- 0.30 pdf Dotals. cn c2 54:4 bc dae Oils ns ittne ee Ee Coe 22, 100.0 The original ‘‘colloid precipitate” accordingly contained 88.2 per cent of organic and 11.8 per cent of inorganic matter. A portion of that part of the ‘‘colloid precipitate’ which re- mained after extracting with the acid alcohol was washed with dilute and strong alcohol, dried at 107°, and analyzed with the following results: Preparation 1. per cent ASH: 28 yoo ake eee WA aids SANs ROAR esse 39 Ash-free is, Nitrogen... ......c scone pis eI: one toie.s pic. or 14.77 Phosphorus. . . .)'s'. yeas Cena he oi. <0 se 2 0.18 Pentosans™. . .. tess ce peed d 2s ols 1.84 *Estimated from phloroglucide after distillation with HCl. This preparation gave all of the characteristic color reactions of the proteins and, from the experiments later described, there can be very little doubt that nearly, if not all, of its nitrogen belonged to protein. Osborne, Wakeman, and Leavenworth 71 Another portion of the moist substance, equal to 6.15 gm. dried at 107°, was suspended in water and twice as much 0.1 N NaOH as was needed to impart an alkaline reaction was added to the suspension. This produced a thick viscid fluid which when centrifuged for some time at high speed, separated into a clear yellow layer at the top, gradually changing to a thick trans- parent jelly at the bottom, there being no line of separation at any point. The whole was poured into several volumes of ab- solute alcohol and 5 cc. of 20 per cent NaCl solution were added. The substance flocked out at once and then could be separated by centrifuging. After suspending the sediment twice more in absolute alcohol it was converted into a granular condition. Washed with ether and dried at 107° it weighed 5.77 gm., equal to 94 per cent of the substance taken. This preparation, No. 2, contained 8.18 per cent of ash and 15.85 per cent of nitrogen, cal- culated ash-free. This treatment with alkali and alcohol had, therefore, raised the nitrogen content by more than 1 per cent. Another portion of this substance, equal to 16 gm. dried at 107°, was extracted five successive times by boiling each time for 1 hour with absolute alcohol, the insoluble part being separated by centrifuging. The first and second alcohol extracts contained only 0.0366 gm. of solids. After thus extracting the substance very thoroughly with boiling absolute alcohol, it was twice ex- tracted by boiling with ether for 1 hour each time. This re- moved only 0.1868 gm. which, on drying in a beaker exposed to the air, formed a varnish-like film which did not redissolve in ether or in alcohol. The amount removed by ether was equal to only 1.1 per cent of the dry substance. After thus freeing the ‘‘colloid precipitate” from everything soluble in water, acid alcohol, boiling absolute alcohol, -and boiling ether, a portion was dried at 107°. This preparation, No. 3, contained 1.23 per cent of ash and, ash-free, 15.04 per cent of nitrogen, only a little more than before extracting with hot alcohol and ether. In order to further purify this substance by reprecipitation, 10 gm. were suspended in water and 0.1 Nn NaOH was added until the reaction was neutral to litmus, which required 100 cc. With 25 cc. more it became alkaline. After further adding 75 cc. of 0.1 nN NaOH the mixture was warmed on the steam bath for a few minutes. On centrifuging a large part was deposited as a jelly, leaving the solution somewhat turbid. ie Proteins of Alfalfa Plant The latter was decanted and the jelly mixed with 50 ee. of 0.1 n NaOH and heated for 1 hour on the steam bath. Under this treatment the jelly slowly dissolved. The clear solution was united with that obtained by first centrifuging, and then very slightly acidified by adding 0.1 n HCl until the precipitate sepa- rated sharply. In order to remove alcohol-soluble substances which might have been liberated by heating with alkali, this hydrated precipitate was suspended in alcohol, 35 cc. of 0.1 NaOH were added, and the solution was reprecipitated by 28 ce. of 0.1 Nn HCl. The precipitate was first washed with 50 per cent, and then with strong alcohol, digested with absolute alcohol and finally with ether, and dried at 107°. This preparation, No. 4, contained 0.98 per cent of ash and, ash-free, 15.91 per cent of nitrogen. This figure is nearly 1 per cent higher than that found in the substance which had not been dissolved in hot alkali. The filtrate and washings from this precipitate, both aqueous and alcoholic, were united, concentrated zn vacuo to small volume, and subjected to a fractionation, the purpose being to learn the nature of the non-protein substances which it contained. Owing to the difficulties encountered in dealing with small quantities of substances of unknown nature nothing definite was learned beyond the fact that, apart from a relatively small proportion of protein and PQ,, the substances contained in these filtrates were chiefly coloring matters soluble in strong alcohol. The outcome of this preliminary examination of the ‘‘ colloid precipitate”? shows that this is a mixture of substances of pro- nounced colloidal properties. In alcohol containing a little hydrochloric acid about 25 per cent of its dry solids are soluble, none of which is protein. Four-fifths of this part are precipitated by making the solution alkaline to litmus. This precipitate looks and behaves like calcium phosphate, which doubtless is its chief component, but together with this is a considerable quantity of one or more calcium salts of organic substances, as the following data show: gm per cent Organie matter...: 2... See ee 8.62 46.8 inorganic matter.) /4...:/)). Sees eee 9.82 53.2 BOWEL 56 s:0\0') os Ware's bongs tee ee: ee 18.44 100.0 Osborne, Wakeman, and Leavenworth 73 . The ash contained 3.15 gm. of Ca and 2.56 gm. of POs, together equal to 5.71 gm., or to 58.1 per cent of the ash. The PO, is equal to 4.18 gm. of CasP2O0s containing 1.62 gm. of Ca, leaving 1.53 gm. of Ca equal to 3.82 gm. of CaCO 3. The sum of the calcium phosphate and carbonate is 8.00 gm., or 1.83 gm. less than the total ash. As the balance of this precipitate was used for other experiments these figures could not be con- firmed, nor a direct determination of CO. be made; hence the nature of the unaccounted for balance of 1.83 gm. was not dis- covered. From these data it appears that acid alcohol removes chiefly calcium salts of phosphoric acid and coloring substances which, in the free state, are very soluble in alcohol. When concen- trated alcoholic solutions of the latter are poured into water colloidal solutions are formed, from which pigment is pre- cipitated by adding a little NaCl. The sodium salt of the colored substance is very soluble in water with a deep orange-yellow color. Its tinctorial power is high. On adding acid, these yellow solutions become colorless at about the neutral point to litmus, and remain clear unless the solution contains relatively much of this substance. What relation this coloring matter has to apparently similar substances heretofore obtained from vegetable sources we hope to ' determine in the future, but the properties above noted indicate flavone or flavone-like pigments. Having thus found that the ‘“‘colloid precipitate” contains a considerable amount of calcium salts of organic substances which in the free state are much more soluble in strong alcohol than in water, a relatively large quantity of another preparation of the ‘‘colloid precipitate’? which had been washed thoroughly with 50 per cent and finally with 75 per cent alcohol was sus- pended in 2,000 cc. of water and centrifuged, in order to remove adhering alcohol. It was next suspended in 2,000 cc. of water containing 5 gm. of HCl and again centrifuged. The clear solu- tion thus obtained gave no biuret reaction, nor any precipitate when a sample was saturated with (NH,).SO., which shows that no protein had been dissolved by the acid. When again extracted with 0.25 per cent aqueous HCl and centrifuged an opaque colloidal suspension resulted which could 74 Proteins of Alfalfa Plant * not be decanted from the sediment. After adding about 60 gm. of NaCl and 5 gm. of HCl to the 4,000 ee. of solution and heating to 50°, a good separation was secured on centrifuging. The clear acid extracts were united, made faintly alkaline to litmus, heated on the steam bath for about half an hour, and then centrifuged. When dried at 107° this precipitate contained: per cent Loss“onignition . ...,.)..)c. 3: eee See ee ee. eee 29 .45 Ne een So PIAS o ec © 70.00 Nitrogen: ...o...64.64ss os Ge Roe ee EE eee eee 0.55 The nitrogen was equivalent to only 1.84 per cent of the ash- free substance, indicating a relatively small proportion of nitrog- enous substances in this ‘‘neutralization precipitate.” The ash contained: per cent POR See » SIGS. oe Bite SE eee Se eee ener 18.94 Pb clsjeswes eae hd Be side ae ds © SEE a eee oe 36 . 04 1 Cee SE Ro 8 no cao CAE wes oc hoe 0.78 These figures correspond to per cent Ca3P.08 AO Oro Ea a OG 2S tors Aoi eo Sebo a4 Bones 30.90 CaCO awcicoiuisqse +> hbed ae see ee eee 60.15 MeO... otic ccs ais as tle 2 2 hc 1.30 Totaloc ie 6.6 6000s a ee ie 92.35 This ‘‘neutralization precipitate’? consisted essentially of calcium salts of phosphoric acid and organic substances, the pro- portion of the latter being larger than was obtained from the extract made with acid alcohol. The residue of the ‘‘colloidal precipitate,’ after thus extracting with dilute aqueous HCl, was suspended in water and shaken out with ether three successive times. The ether extracts were washed with water and evaporated in vacuo. The residue was taken up in absolute alcohol and likewise concentrated in vacuo. This process was repeated until water was removed. The residue was then extracted with absolute ether in which 0.5474 gm. was soluble. This is an almost insignificant proportion of the rela- tively large quantity of the ‘‘colloid precipitate” from which it a Osborne, Wakeman, and Leavenworth 25 was derived. Unfortunately the dry weight of the portion taken was not determined, but was certainly more than 40 gm. The residue of the ‘‘colloid precipitate’? was next extracted three times with absolute alcohol. Owing to the colloidal condi- tion of the solids it was difficult to obtain a good separation by centrifuging until a little NaCl was added. The 10,000 cc. of deep brown extract were concentrated in vacuo to 400 ce. and by centrifuging freed from a small amount of solids which separated during concentration. The clear solution was evaporated in vacuo and the residue extracted with absolute alcohol, in which most dissolved. Thus freed from sodium chloride, the clear alcoholic solution, when evaporated, left a very dark red-brown, amorphous residue weighing 3.32 gm. which was readily dissolved by 50 ce. of absolute alcohol. When this solution was poured into 150 ec. of distilled water a colloidal suspension resulted which did not separate on standing over night. The addition of a little HCl did not cause any separation, but the further addition of a little NaCl produced a large flocculent precipitate, leaving a clear, but deeply colored, solution. Examination of this pig- ment has, as yet, revealed no difference between it and the similar substances extracted by acid alcohol from the ‘colloid precipi- tate.” It seems probable that this method of isolating the pig- ments may facilitate their further study. After extracting the ‘“‘colloid precipitate’ with water, aqueous HCl, ether, and alcohol a portion was dried at 107° and analyzed with the following results. Preparation 5. per cent PE MES alee dish came diate seine 0.62 Ash-free substance. (oi) 2G a oe a re Ae, 14.60 EYED? CRM 8 2b cs ss cya: te cis arlene: nisiolo's ALI a ee 0.97 Lo MET. DES ee rll te 0.83 FS CELTIC el 8 nee Ree PR Aaa | 0.95 Another portion, equal to 13.5 gm. of the dry solids, was sus- pended in 950 cc. of water and 131.1 ec. of 0.1 N NaOH were added until the suspended solids separated sharply at the iso- electric point. The clear, almost colorless solution, when evapo- rated to dryness, left a residue weighing 0.7623 gm., consisting of NaCl. The NaOH added was equivalent to 0.7669 gm. of NaCl, 76 Proteins of Alfalfa Plant or to 0.4785 gm. of HCl, or to about 3.5 per cent of the dry solids of the preparation taken for this experiment. It thus appears that during extraction with HCl a notable amount of this acid had combined with the protein and that no water-soluble organic substance had been liberated at the isoelectric point. In order to determine whether, or not, substances soluble in alcohol, but insoluble in water, had been liberated by thus re- moving the combined acid the solids were next suspended in absolute alcohol and centrifuged. The resulting clear solution, however, contained only 0.0620 gm. of solids, thus showing the absence of a notable quantity of substance soluble in alcohol, but insoluble in water, liberated at the isoelectric point. To learn whether, or not, anything could be extracted by hot water the residue was next digested with water for an hour on the steam bath. After removing the suspended protein by centri- fuging, the somewhat colloidal solution was treated with 8.5 ce. of 0.1 n NaOH and about 0.8 ec. of 20 per cent NaCl solution which caused the colloidally suspended matter (protein) to separate as a flocculent precipitate, leaving the solution clear when centrifuged. This precipitate, A, was treated as de- scribed below. The clear solution when evaporated left a residue, chiefly NaCl, weighing 0.2830 gm. After thus extracting with hot water the residue was washed once by centrifuging with water. The washings contained only 0.1056 gm. The 0.3886 gm. thus extracted contained 0.1525 gm. of ash (NaCl) and 0.2361 gm. of organic matter. It is clear that only an insignificant amount of organic substance was thus ex- tracted by hot water. The residue was next boiled with absolute alcohol and centrifuged. The alcoholic washings, on evaporation, left only 0.0276 gm. of solids. . The residue of Preparation 5, which had thus been freed from combined HCl and everything else soluble in hot or cold water as well as in alcohol, was dried at 107° and analyzed with the following results: Preparation 6. per cent ABD, s.0 560/00 0:5 'e 4.06 s(t eee tT eae ee win Scie 0.74 Ash-free substance. Total nitrogen‘: 2 ee celeste ss eked eee 15.45 Amide 4!» +2 SeRPRS aerate ae alee ff Phosphorus. ..../: 3 teae cee cess 0 ale ee 0.081 Suir... 0. icc wierd eee ne ee eee 1.02 Osborne, Wakeman, and Leavenworth ae The nitrogen content of this preparation was practically the same as that of Preparation 7, similarly obtained, which, ash-free, contained 15.34 per cent of nitrogen. The small precipitate, A, p. 76, which separated from the hot water extracts after adding a little more alkali and some NaCl, as above noted, was dissolved in 20 cc. of 0.1 n NaOH, 100 ce. of water were added, and the solution was centrifuged. A part separated as a jelly from which the clear solution was decanted. To this solution 18.6 cc. of 0.1 N HCl were added, and the precipi- tate produced was washed with water and alcohol and dried at 107°. This preparation, No. 8, weighed 0.64 gm. and contained 0.63 per cent of ash. The ash-free substance contained 15.59 per cent of nitrogen; 7.e., only 0.24 per cent more than did Prepara- tion 6 which had not been dissolved in alkali. These data show that the residue of the ‘‘colloid precipitate, ”’ which remains after extracting with water, aqueous hydrochloric acid, ether, and alcohol, consists chiefly of a hydrochloride of protein which is entirely insoluble in water. Such unusual prop- erties suggest combination with some non-protein group. When this protein hydrochloride is suspended in water at the room temperature none is dissolved, as shown by the absence of a biuret reaction in the filtered solution. At temperatures approaching 100°, however, it is converted into a clear yellow jelly which does not pass into a true solution, even after long heating. After cooling, the addition of an excess of hydrochloric acid, or of sodium chloride, converts this jelly into a coarse floccu- lent precipitate resembling an ordinary protein coagulum, such ‘as is commonly produced by heating. This settles rapidly, leaving the solution clear. A little coloring matter is thus re- moved, as shown by the strong yellow color of the filtered solution when made alkaline with NaOH. When suspended in water and treated with NaOH to the iso- electric point the protein separates completely, leaving the solu- tion water-clear. On evaporation a residue is left which is practically all sodium chloride. The combined hydrochloric acid is likewise converted into sodium chloride by sodium acetate. A quantity of the moist protein hydrochloride, equal to 9 gm. dried at 107°, was suspended in water and freed from chloride by bringing to the isoelectric point with NaOH. The undissolved 78 Proteins of Alfalfa Plant protein was collected by centrifuging, suspended in 300 ce. of water saturated with toluene, and used for the following ex- periments. Experiment I. 10 ce., containing 0.8 gm. of protein + 5 cc. of water. Experiment II. 10cc. +3 cc. of water + 2 cc. of 0.1 n HCl. Experiment III. 10 ce. + 5 ec. of 0.1 n HCl. These mixtures, which contained 0.3 gm. of the protein, were heated simultaneously in a steam bath. Suspension III gelati- nized first, Suspension II very soon after, and Suspension I appeared to be unchanged. After cooling Nos. II and III be- came firm jellies, which did not flow on inverting the test-tube, and on standing gradually contracted, leaving a little clear serum above. After 3 days there was much less serum in No. II than in No. III, suggesting that the acid was slowly hydrolyzing the protein, or possibly a combination of the protein with some non- protein complex. Experiment IV. 10 cc. + 2 cc. of 10 per cent HCl + 3 cc. of water. When heated on the steam bath the protein was precipitated like a heat coagulum, leaving the solution water-clear. This solution gave a strong biuret reaction showing the protein to be slightly soluble, perhaps owing to hydrolysis, when heated with acid of this strength. After washing the precipitate with water it became slimy when the excess of acid was removed. When this was returned to the tube and heated as before, it formed a clear transparent jelly. On continued heating the jelly was slowly converted into a solution, which finally became so clear that it caused no Tyndall effect. This solution gave a flocculent precipitate with only 2 ee. of 0.1 Nn NaOH, which was completely redissolved by 1.2 ce. more. The ready solubility in dilute alkali, when contrasted with the behavior of this protein when treated with alkali before heating with acid, suggests that it exists as a conjugated protein which is hydrolyzed by the hot acid. See Experiment VI. Experiment V. 10cc. +3 ce. of 20 per cent NaCl + 2 ce. of 0.1 n HCl. Heated on the steam bath the effect was like that of No. IV, showing that the ions from NaCl prevent the formation of the Osborne, Wakeman, and Leavenworth 79 jelly as do those from HCl. After washing out the excess of salt and acid the protein was readily converted into a jelly when 5 ec. of 0.1 N HCl were added, and the mixture was heated on the steam bath. On further heating it behaved just like Suspen- sion IV. Experiment VI. 10 cc. + 15 cc. of water + 10 cc. of 0.1N NaOH at room temperature. This produced a viscid solution which was diluted to 100 ce. and centrifuged at high speed for sometime, whereby a small amount of gelatinous substance was deposited. Although the solution appeared to be nearly clear it filtered exceedingly slowly through paper and was quite opaque in a beam of sunlight. After heating in the steam bath for about 4 hours it could be readily filtered through paper. A larger quantity of Preparation 5 was freed from combined HCl by adding NaOH to the isoelectric point and centrifuging. The deposit was centrifuged once with water and then suspended in 200 ce. of water and 25 cc. of 0.1 Nn NaOH were added. This produced a thick, opaque ‘‘solution,”? which was diluted to 400 ec. and centrifuged at high speed for some time. The nearly clear solution was decanted from a voluminous deposit of jelly and again centrifuged and decanted from a much smaller deposit. By suspending the jelly deposits in water and centrifuging re- peatedly, as before, two parts were obtained, one representing the gelatinous fraction, the other the more soluble part. The solu- tion of the latter, which could be filtered through paper fairly easily, was then precipitated by 17 cc. of 0.1 Nn HCl. This prep- aration, No. 9, was washed with water and with alcohol and dried at 107°. It contained 0.76 per cent of ash and, ash-free, 15.25 per cent of nitrogen. The jelly fraction was suspended in water and a little NaCl added. On centrifuging it separated as a coherent deposit which was washed with water and alcohol. Dried at 107° this preparation, No. 10, contained 2.05 per cent of ash, and, ash-free, 15.40 per cent of nitrogen. Although the solubilities of these two parts were so markedly different as to indicate the presence of two different proteins their nitrogen content was substantially the same, and also like that of the product from which they originated. Other data must be 80) Proteins of Alfalfa Plant obtained before the difference in solubility here noted can be explained. Another sample of ‘‘colloid precipitate” was extracted four successive times with 93 per cent alcohol containing 0.1 per cent HCl and then treated with dilute sodium hydroxide, which pro- duced a very turbid gelatinous ‘‘solution.”’ This was centrifuged at high speed and the solution decanted from a jelly-like deposit. The latter was again suspended in dilute alkali and centrifuged, this process being repeated several times. The decanted solu- tions were united, filtered clear through paper pulp, and made neutral to litmus with dilute HCl. The resulting precipitate was washed with water, dilute and strong alcohol, and then with ether. Dried at 107° this preparation, No. 11, which weighed only 1.5 gm. and represented only a small fraction of the original sub- stance, contained 1.4 per cent of ash and, ash-free, 15.33 per cent of nitrogen. Having found the protein similarly obtained from spinach leaves to be readily soluble in hot dilute alkaline alcohol this solvent was applied to the alfalfa protein. The leaves of freshly cut alfalfa plants were removed by hand and kept over night in the ice chest. The next morning 3,600 em. of these were ground with 10 liters of water and then pressed in the hydraulic press. The green juice was filtered through paper pulp, which removed the chlorophyll, and the clear brown juice treated with about 20 per cent by weight of aleohol. The voluminous precipitate was washed, first with quite dilute and then with stronger alcohol, until adhering mother liquor was removed. A part of this highly hydrated precipitate, equal to 19.7 gm. dried at 107°, was suspended in 1,000 ce. of 0.2 per cent NaOH solution, an equal volume of alcohol added and heated to 80° for 2 to 3 minutes. By centrifuging at high speed a yellow, gelati- nous residue was removed which, when washed with alcohol and ether and dried at 107°, weighed 1.96 gm., and contained: per cent SDE ha poeta Nitrogen...... 2.71 (= 5.69 per cent of the ash-free substance.) The clear, alkaline filtrate when neutralized with HCl gave a flocculent precipitate, which was washed first with 50 per cent EEE Se EE a Osborne, Wakeman, and Leavenworth 81 and then with stronger and finally with absolute alcohol and dried at 107°. This preparation, No. 12, weighed 11.91 gm. and contained 0.62 per cent of ash. The ash-free substance contained: Preparation 12. Protein. | Nitrogen. per cent per cent SMTA TED TUTORS rnin ye ee MIE 0.96 5.86 Humin EE orci aces ae ec ees oe Te 0.60 3.67 Basic peers). Ss). UL a ee 3.76 22.98 RE Pe te gS sag. She nis CUE I ee 11.04 67.49 SRO MI NU CAOONE FOE Pie... s,s 5. ss dn'p.e x anton eae 16.36 | 100.00 The percentage of nitrogen in Preparation 12 is about 1 per cent higher than that found in Preparations 8, 9, 10, and 11 which had been dissolved in aqueous alkali at room temperature. This fact indicates, that by the treatment with hot alkaline alcohol, some non-protein substance had been removed, possibly in consequence of hydrolytic action of the alkali. The following table gives the nitrogen found in the several fractions produced by the above described procedure. Grams. N recovered: Mean tesiduc, insoluble in alkali....:...0.....05. 2.52. 0.0600 ye | W mm alcoholic washings of above..................... 0.0541 2.4 Hem protem precipitated by.HCl..................... 1.8920 84.6 Peemeeiinen te TOM ADOVE. ..............0000cecesnecene 0.2300 10.3 Ee ee ee 2.2361 100.0 These figures show that most of the nitrogen belongs to pro- tein soluble in the hot alkaline alcohol and precipitable by acid. Another preparation of the ‘‘colloid precipitate” was made in a similar way from plants cut a few days before blossoming. A portion of the moist preparation, containing 31.7 gm. of dry solids, 3.655 gm. of N = 11.54 per cent, and 3.552 gm. of ash = 11.12 per cent, was heated in the steam bath for a few minutes with 60 per cent alcohol containing 0.2 per cent of NaOH and 82 Proteins of Alfalfa Plant then centrifuged for some time at high speed. The gelatinous deposit was treated three successive times with hot dilute alkaline aleoho' as before, and then washed with dilute and stronger a!co- hol. Dried at 107° this residue weighed 4.32 gm., equal to 14.3 per cent of the ‘‘colloid precipitate” from which it was derived. It contained 52.7 per cent of ash, equal to 68.1 per cent of the inorganic matter of the ‘‘colloid precipitate,”’ and 2.51 per cent of nitrogen, equal to 3.1 per cent of the nitrogen in the ‘‘colloid precipitate.” This nitrogen was equal to only 5.3 per cent of the organic matter in this preparation, indicating the presence of a large proportion of non-protein matter in this residue. Cf. similar product on p. 80. The united alkaline solutions decanted from the foregoing residue were neutralized with HCl and the precipitate produced was washed with water, alcohol, and ether, and dried at 107°. ~ This preparation weighed 17.72 gm., equal to 56 per cent of the ‘colloid precipitate,’’ a considerably smaller proportion than was recovered in the precipitate in the corresponding experiment last described. The nitrogen content of this preparation, No. 13, was, however, practically the same as that of Preparation 12, being 16.21 per cent of the ash-free substance, as against 16.36 per cent. The rest of the ‘‘colloid precipitate,’’ from a part of which Preparation 13 was obtained, was worked up in the same way and yielded 233 gm. of Preparation 14 which contained 0.72 per cent of ash and 16.27 per cent of nitrogen, calculated to the ash- free substance. 50 em. of the air-dried preparation, No. 14, equal to 44.68 gm., ash- and moisture-free, were boiled for 24 hours with 25 per cent of sulfuric acid and tyrosine, histidine, arginine, and lysine determined according to the slightly modified method of Kossel.® Preparation 14. Protein. per cent per cent AV NOSIME: js: s15) 602.4 6 ce aL Histidine.,...........-4.8c eee eee 2.56 (containing N = 0.69) ATeinime: of: {Svea eee ee AL Hf = Re) Lysine... oc... eee 3.34 ( e £6" = 0764) Total. N occcies Setonde ote Oe Le eo eee 3.62 'Cf. Osborne, T. B., Leavenworth, C. 8., and Brautlecht, C. A., Am. J. Physiol., 1908, xxiii, 180. Osborne, Wakeman, and Leavenworth 83 The nitrogen contained in the basic amino-acids thus found agrees closely with the basic nitrogen found in Preparation 12, namely 3.76 per cent. The Filtrate from the ‘‘Colloid Precipitate.”’ The filtrate from the precipitate produced by adding 18 per cent of alcohol to the alfalfa press-juice yields another volumi- nous precipitate when its alcohol content is raised to about 40 per cent by weight. This precipitate is a mixture of protein and inorganic salts, but as this treatment with alcohol renders nearly all of the latter insoluble in water or dilute salt solutions, it has been impossible, as yet, to learn anything of its charac- teristics. We have, however, obtained evidence of a small proportion of heat coagulable protein, as well as of proteose, in the filtrate from the ‘“‘colloid precipitate.’”’ Thus, 46 kilos of fresh alfalfa plant, equal to about 10 kilos dried, were thoroughly extracted with water and the clear extract was freed from the ‘‘ colloid precip- itate” by adding about 23 per cent by weight of alcohol. The filtered solution was then concentrated, heated to boiling, and acidified with acetic acid. The coagulum which separated was thoroughly washed with boiling water and then with alcohol and ether. Dried at 107°, this weighed 37.5 gm., and contained 43.0 per cent of ash. The ash-free substance, however, con- tained only 1.9 gm. of nitrogen, equal to 12.5 gm. of protein (N X 6.25), or to only 0.12 per cent of the dry alfalfa solids, and this on the assumption that all of this nitrogen belongs to protein. In another case 23 per cent of alcohol was added to the press- juice from alfalfa leaves containing 929 gm. of dry solids. The filtrate from the ‘‘colloid precipitate’? thus produced was con- centrated till the aleohol was removed, heated to boiling, and slightly acidified with acetic acid. The coagulum weighed 4.0 gm. when dried at 107°, and contained 0.4803 gm. of N and 0.5427 gm.*of ash. The nitrogen in this coagulum was equal to 2.7 gm. of protein (N X 6.25), or to 0.29 per cent of the dry solids of the leaves. A part of the filtrate from this coagulum was saturated with (NH,4)2S04 which produced an oily precipitate having the physical properties characteristic of proteoses. As this precipitate could 84 Proteins of Alfalfa Plant not be removed by an ordinary filter it was necessary to use a felt of paper pulp, which yielded a clear filtrate. The precipi- tate was redissolved by extracting the filter thoroughly with water, the filtrate again saturated with (NH,).SO., and the oily precipitate treated as before. After twice more reprecipitating in this way the clear solution was made up to a definite volume, and total, as well as ammonia nitrogen determined in aliquots. The difference, 0.225 gm., was assumed to belong to proteoses equal to 1.406 gm. (N X 6.25). Since this was derived from 38 per cent of the total filtrate from the coagulum, this quantity is equivalent to 0.40 per cent of proteose possibly present in the solids of the dried leaves. These figures, which must be accepted with great reserve, indicate that the alfalfa solids may possibly contain about 0.3 per cent of coagulable protein and 0.40 per cent. of proteose, surprisingly small proportions to be found in such physiologically active tissues as are those of a green plant. By far the greater part of the protein of the juice of the alfalfa plant is, therefore, represented by the protein in the ‘‘colloid precipitate.” The Alcohol-Soluble Constituents of the Green Alfalfa Plant. After extracting the 22 kilos of green alfalfa (see p..66) with water the final press-cakes were ground in the Nixtamal mill with 8 liters of 93 per cent alcohol, pressed in the hydraulic press, and the cakes twice more treated in the same way with alcohol. These extracts were concentrated to relatively small volumes and solids and nitrogen were determined in aliquot parts of each. Solids. Nitrogen. gm. gm. WIrstiextraeh.. sic. Se.6:.6 ee eee aa 99.6 2.25 Second “ec .cc ne ache ee eee 131.7 1253 hind; ‘ey ON Se er re er 51 9 0.49 Motel oS oi. secs os cts co kee er eee racic 283).2 4.27 Only 6.4 per cent of the solids of the dry alfalfa and 2.0 per cent of its nitrogen were thus extracted by alcohol. Osborne, Wakeman, and Leavenworth 85 The Alkali-Soluble Constituents of the Green Alfalfa Plant. A part of the moist residue of the alfalfa plant, which had been extracted with water and then with alcohol, contained 482 gm. of dry solids, 20.58 gm. of nitrogen, and 22.55 gm. of ash (CO,- free). This was ground with 5 liters of 0.8 per cent NaOH solu- tion and pressed in the hydraulic press. The press-cake was again ground up with about 4 liters of 0.3 per cent NaOH and pressed as before. This process was repeated twice. The fourth extract contained scarcely anything precipitable by HCl. The 18 liters of united extracts contained 2.70 gm. of nitrogen, 45.5 gm. of solids, and 3.5 gm. of ash (the two latter corrected for the added NaOH). These quantities are equivalent to 230 gm. of ash-free solids, 14.80 gm. of nitrogen, and 19.2 gm. of ash, calculated on the basis of the amount of alfalfa originally ex- tracted. The nitrogen in this solution, therefore, was equal to only 7.0 per cent of the total alfalfa nitrogen, and the ash-free solids to 5.2 per cent of its organic matter. When acidified with dilute HCl this alkaline extract yielded a voluminous precipitate which was washed by centrifuging the dilute and stronger alcohol, and then extracted with absolute alcohol and ether. This preparation, dried at 107°, weighed 13.5 gm., equal to 74 gm. from the entire lot of alfalfa plants originally extracted, or to 1.5 per cent of the dry matter thereof. It contained 0.55 gm. of ash and 1.305 gm. of nitrogen, equal to only 59 per cent of protein (NX6.25) in the precipitate. The ash-free sub- stance, therefore, contained 9.37 per cent of nitrogen, equivalent to 7.15 gm., calculated back to the alfalfa taken, or to 3.4 per cent of the alfalfa nitrogen. When distilled with HCl this prepara- tion yielded phloroglucide equivalent to 7.54 per cent of pentosan in the ash-free substance. Since this precipitate also gave strong xanthoprotein and biuret reactions there is little doubt that it consisted chiefly of a mixture of protein and carbohydrate. Analysis of aliquot parts of the filtrate from the precipitate produced by HCl showed that it contained 32.5 gm. of solids, other than NaCl, 1.24 gm. of nitrogen, and 0.0953 gm. of ammonia, equivalent, respectively, to 178, 6.8, and 0.52 gm. calculated back to the amount of alfalfa originally taken. The ash-free solids were equal to 3.6 per cent of the dry alfalfa and the nitrogen 86 Proteins of Alfalfa Plant to 3.2 per cent of its total nitrogen. The ammonia in the filtrate from the acid precipitate was equal to 7.7 per cent of the total nitrogen extracted. Whether this ammonia resulted from hy- drolysis of amide nitrogen of the protein, or from some non- protein nitrogenous substance, was not determined. The remainder of the filtrate was saturated with (NH4).SOu,, the precipitate produced filtered out and pressed on filter paper, redissolved, and again precipitated and pressed. When dissolved in a liter of water total nitrogen and ammonia nitrogen were determinéd in aliquots. The non-ammonia nitrogen thus found by difference, was equal to 0.4278 gm., or to 2.7 gm. of protein (N X 6.25), in the entire solution. This quantity is equivalent to 14.7 gm. protein (N X 6.25), or to 0.29 per cent in the alfalfa originally taken for extraction. Extraction with Alkaline Alcohol. After thus extracting with aqueous alkali, which might be expected to remove most of the residual protein, the press-cakes were ground up with 10 liters of 60 per cent alcohol containing 0.3 per cent NaOH, heated to boiling on the steam bath for a few minutes, and filtered hot. The residue was pressed in the hydraulic press and again treated in the same way. The first extract was much more deeply colored than the second, and likewise yielded a much larger precipitate on the careful addition of dilute HCl. The two precipitates were washed with 50 per cent and stronger alcohol, as well as with absolute aleohol and ether. That from the first extract weighed 54.2 gm., dried at 107°, and contained 1.37 gm. of ash and 7.63 gm. of N, equal to 14.42 per cent of the ash-free substance. The precipitate from the second extract weighed 12.0 gm., dried at 107°, and contained 1.15 gm. of ash and 1.17 gm. of nitrogen. The ash-free substance, therefore, contained only 10.74 per cent of nitrogen. These two precipitates, together, contained nitrogen equivalent to 23 per cent of the alfalfa nitrogen, or to 6.09 per cent of protein (N X 6.25), in the alfalfa plant. The united filtrates from these precipitates contained 80.5 gm. of ash-free solids, equal to 8.9 per cent of the alfalfa; also 6.38 gm. Osborne, Wakeman, and Leavenworth 87 of nitrogen equal to 10.0 per cent of the ash-free solids, or to 16.5 per cent of the total nitrogen in the alfalfa. Since heating with the alkaline alcohol might be expected to convert some of the amide nitrogen of the protein into ammonia, a part of the filtrate from the precipitate produced by adding HCl to the second alkaline alcohol extract was distilled with magnesia, and free ammonia equal to 2.2 per cent of the total nitrogen in this alkaline alcohol extract was found. This pro- portion of ammonia nitrogen was even less than that similarly found in the cold aqueous alkaline extract. We thus find that alkaline alcohol extracted solids, equivalent to 16.0 per cent of the alfalfa, and nitrogen equal to 39.3 per cent of its total nitrogen. The precipitate produced by neutralizing the first alkaline alcohol extract probably consisted chiefly of protein, as indicated by the 14.42 per cent of nitrogen in the ash-free substance, and by strong biuret and tryptophane reactions. It is almost certain that this crude precipitate contains some non-protein substance, probably mostly carbohydrate, but we have not, as yet, had an opportunity to determine this. We expect to subject this prod- uct to a critical examination and hope soon to be able to learn something of its relations to the protein of the “colloid precipi- tate”? which it resembles in many ways. The corresponding precipitate from the second alkaline alcohol extract contained, ash-free, only 10.7 per cent of nitrogen, and probably represents a still more impure preparation of the same protein. The filtrate from the precipitate produced by HCl in the first alkaline alcohol extract was deeply colored and contained 52.8 gm. of ash-free solids, equal to 5.8 per cent of the entire lot of alfalfa extracted. This relatively large amount of the alfalfa solids had escaped extraction by water, strong alcohol, and aqueous alkali at room temperature. That it became soluble in water after heating with alkaline alcohol suggests that it was pro- duced by hydrolysis during heating with the alkaline alcohol. Addition of neutral lead acetate to this filtrate gave a large precipitate containing much PbCl, together with a considerable quantity of organic lead salts. This precipitate was removed by centrifuging and decomposed with an excess of HCl. The addi- 88 Proteins of Alfalfa Plant tion of several volumes of absolute alcohol gave a deep yellow- brown solution which, after removing the PbCh, contained 23.3 gm. of solids (ash-free), largely soluble in absolute alcohol, and completely soluble in strong alcohol. This quantity is equal to 127.7 gm. from the total alfalfa originally extracted, or to 2.6 per cent. This represents only a part of the substances of this type, because the corresponding filtrate from the second extrac- tion and the alcoholic washings of the precipitates, which were not treated with lead acetate, doubtless contained a further considerable quantity of the pigment. Since substances having similar properties, and resembling the pigments derived from flavone, were also obtained from the ‘‘colloid precipitate” it appears that the alfalfa plant may contain a very considerable proportion of this group of substances. The fact that after the alfalfa had been previously extracted very thoroughly with water, alcohol, and then with aqueous alkali, heating with alkaline alcohol rendered so large an amount of the pigments soluble in water, as well as in alcohol, is strong evidence that both the protein and the pigment were derived from some combination, or combinations, hydrolyzed by the hot alkali. That simultaneously a correspondingly large proportion of pro- tein was also rendered soluble in dilute aqueous alkali suggests that, in the plant, the protein and pigments may be combined. Such pigments may occur in green plants in much larger proportion than has heretofore been supposed, because apparently it has been overlooked that a large proportion of them may be combined with protein, as a complex insoluble in the solvents usually employed for extracting them. The methods employed in this investiga- tion seem to be well suited for obtaining these pigments in a condition fit for further chemical study. We expect to continue our investigations along this line, and later we may be able to present results of interest to the chemist, as well as to the plant physiologist. The Extracted Residue. The residue from the extractions just described was nearly colorless and consisted chiefly of the fibrous structure of the plant. Dried at 107°, it contained 29.1 per cent of the dry, ash- free, solids of the alfalfa, but only 5.3 per cent of its nitrogen. In Osborne, Wakeman, and Leavenworth 89 view of the difficulties attendant on grinding large quantities of whole plants fine enough to rupttre all of the cells so small an amount of undissolved nitrogen is surprising. SUMMARY. It is possible to grind the fresh green alfalfa plant so thoroughly that practically all of the contents of its cells subsequently can be extracted by water, alcohol, dilute aqueous alkali, and hot alkaline alcohol, applied in the order named. Water extracts over 45 per cent of the dry matter of the plant, nearly 43 per cent of the ash-free solids, nearly 44 per cent of its nitrogen, and almost 71 per cent of its inorganic constituents. By subjecting the ground plants to high pressure relatively large quantities of the undiluted juice of the plant can be ob- tained as an almost clear dark brown liquid, free from chloro- phyll, or other suspended particles. This juice contains about 10 per cent of solids, a part of which is in colloidal solution. The addition of about 20 per cent of alcohol causes the latter to separate as a flocculent precipitate, which can be readily filtered out. The filtered solution contains much nitrogen, but very little protein, probably less than 1 per cent. Some of this protein can be coagulated by heating the acidified solution, but more of it has properties characteristic of proteoses. Most of the protein in the aqueous extract is in the precipitate produced by alcohol, which contains the substances previously in colloidal solution. In addition to protein, which forms up- wards of 70 per cent of this precipitate, there are also present calcium phosphate and calcium salts of organic substances which can be extracted from the protein by alcohol containing HCl in which the protein is insoluble. The organic substances appear to be largely pigments which resemble the flavone derivatives already known to occur in many species of plants. The protein combines with HCl without passing into solution at room temperature. When suspended in water this hydro- chloride of the protein is converted into a jelly on heating, but does not dissolve, until heated for a long time with an excess of the acid. It behaves in a similar way toward dilute alkalies, but in this case passes into the gelatinous state at room temperatures. 90 Proteins of Alfalfa Plant Appreciable amounts do not dissolve, however, until the solution is heated. The precipitate of the colloids contains about 11.5 per cent of nitrogen. After extraction with 75 per cent alcohol containing a little hydrochloric acid the nitrogen content is raised to nearly 15 per cent. By suspending the protein hydro- chloride, which results from the treatment with HCl, in water and adding dilute NaOH solution the protein separates at its isoelectric point in a flocculent condition. Practically nothing is thus removed except combined HCl. The nitrogen content of the protein is thus raised to nearly 15.5 per cent. When next treated with an excess of alkali at room temperature and precipitated by HCl the nitrogen content remains unchanged. No appreciable amount of substances soluble in alcohol, or in ether, are thus liberated by either the acid or alkali. When dissolved by heating with an excess of dilute alkali and precipitated with HCl preparations containing about 16.3 per cent of nitrogen were obtained. The nature of the non-protein complex thus removed has not yet been determined, but the deep yellow-brown color of the alkaline solutions, and the yellow color developed by adding alkali to the filtrate from the precipi- tate produced by acid, indicate the presence of the same pigment found in the acid alcohol extracts of the mixed colloids. Since this pigment is not removed by treatment with alkali at room temperatures, but is removed by heating, it seems probable that the protein is combined with this pigment, from which it can be separated only by hydrolyzing the complex. After removing the water-soluble constituents, 6.4 per cent of the alfalfa solids and 2.0 per cent of the nitrogen were ob- tained by extracting several times with alcohol. This extract contained nearly all of the chlorophyll, together with other sub- stances, the nature of which has not yet been ascertained. Dilute sodium hydroxide solution extracted only 6.9 per cent of the alfalfa nitrogen, a part of which was protein precipitable by slightly acidifying the extract. The precipitate also contained some pentosans. : After thus thoroughly extracting with water, alcohol, and dilute aqueous sodium hydroxide solution, the residue was boiled for a few minutes with 60 per cent alcohol containing 0.3 per cent NaOH. This treatment extracted 17.8 per cent of Osborne, Wakeman, and Leavenworth. 91 the ash-free solids and 39.3 per cent of the nitrogen of the alfalfa. About 60 per cent of this nitrogen belongs to protein, which can be precipitated by the cautious addition of acid. Some, or all, of the remaining 40 per cent may have been derived from the same protein in consequence of changes caused by the hot alkali. The protein in the residue of the alfalfa, after extraction with water and alcohol, behaves much like that found in the ‘‘colloid precipitate” produced by adding a little alcohol to the aqueous extract. That this, likewise, may be a complex containing protein and a flavone-like pigment is indicated by the liberation of a relatively large proportion of the latter simultaneously with the protein, when the alfalfa residue is heated with the alkaline alcohol. After the successive extractions here described the residue was . equal to 32 per cent of the solids of the plant, but contained only 5.6 per cent of its nitrogen. Consequently we can conclude that by the methods we have employed nearly all of the cells of the plant were ruptured so that their contents could be extracted by suitable solvents. The procedure here outlined seems especially suited for separat- ing the various groups of substances soluble in the several solvents under conditions particularly favorable for further investigation. We hope in the near future to be able to supplement this pre- liminary investigation with more detailed studies of these differ- ent groups. a Hy) hy (inhib 4 aa. We at i he ‘ pale rh Wai: ‘ ee ‘ ‘TR WG (05% 7 THE USE OF SODIUM SULFATE AS THE GLOBULIN PRECIPITANT IN THE DETERMINATION OF PROTEINS IN BLOOD. By PAUL E. HOWE. (From the Department of Animal Pathology of The Rockefeller Institute for Medical Research, Princeton, N. J.) (Received for publication, September 1, 1921.) In connection with a study of blood under normal and patho- logical conditions we proposed to use the method of Cullen and Van Slyke (1), for the determination of fibrin, globulin, and albumin nitrogen of blood plasma. A serious objection to the method is the use of ammonium sulfate as the globulin precipi- tant. This is so for two reasons, (a) because of the use of an ammonium salt which must be removed before determining the globulin nitrogen and (b) because of the physical difficulties involved in the removal of this nitrogen with magnesium oxide; particularly with reference to the ‘‘bumping”’ of the mixture. Both of these factors were realized by Cullen and Van Slyke, who, however, showed their method to be accurate, and believed that they had found a way of preventing the extreme bumping. We have corroborated their method as far as duplication of results is concerned but found that Merck’s highest purity magnesium oxide is not always suitable for these determinations. Satis- factory results were obtained only by constant shaking of the Kjeldahl digestion flask during the distillation of the ammonia and the early part of the digestion. The use of a non-nitrogen-containing precipitant in place of the ammonium sulfate, it seemed, would remedy both of the defects indicated above. The salt most commonly replacing ammonium sulfate for the precipitation of the total globulins is magnesium sulfate. Sodium sulfate has also been shown to be a satisfactory precipitant for globulin. Preliminary trials with these two salts indicated that, with regard to clean pre- 93 94 Sodium Sulfate as Globulin Precipitant cipitation and rapid filtration, sodium sulfate was far superior to magnesium sulfate. When magnesium sulfate is used to pre- cipitate proteins, the precipitates are gelatinous in character and the solutions filter slowly. The chief objection to the use of sodium sulfate is the necessity of working at temperatures above 34°C. for precipitation at the highest concentrations of the salt. The solubility of NasSO, + 10 H.O increases gradually up to approximately 10°C. and then rapidly to 34°C.; above 34°C. the anhydrous salt is in equilibrium with water and the solubility of the salt decreases gradually. Working at incubator temperatures conditions with regard to solubility are those which obtain in general with magnesium sulfate and ammonium sulfate; 7.e., a gradual change in solubility with each increment of temperature. The use of sodium sulfate for the precipitation of proteins is not a new procedure, nor is the use of it in the determination of the blood proteins a new process. Pinkus (2) realized the advan- tages of sodium sulfate in the study of proteins. He outlined the possibilities for its use and showed that (1) sodium sulfate pos- sesses at 30°C. the same protein-precipitating power as ammonium sulfate. (2) When the anhydrous salt is used instead of the hydrated salt, at the temperature indicated, it precipitated glob- ulins at the point of half saturation (about 25 per cent) and albumin at full saturation (about 50 per cent), it also allows of a fractionation of the proteins of Witte’s peptone. (38) The use of sodium sulfate presents the following advantages: (a) the color reactions are hardly at all interfered with, (b) the nitrogen of the precipitate may be estimated directly according to method of Kjeldahl, (c) it is easy to obtain solutions containing little salt (5 per cent) by cooling, and (d) the salt itself is practically non- toxic. (4) By adding to protein solutions enough anhydrous sodium sulfate to absorb all of the water, a product is obtained which can be kept without change in the protein and is easily workable. Porges and Spiro (3) confirmed the observation of Pinkus with regard to the precipitation limits of sodium sulfate: that the limits in terms of percentage of a saturated solution are approximately the same as those pertaining to ammonium sulfate. These authors used sodium sulfate in the quantitative separation of the serum proteins. They present evidence pointing Es Paul E. Howe 95 toward the presence of three globulins, a euglobulin and two pseudoglobulins, with precipitatiom limits at 28 to 36 per cent, 33 to 42 per cent, and 40 to 46 per cent of saturated sodium sul- fate estimated from their published chart. Haslam (4) made use of sodium sulfate in the separation of albumoses. Homer (5) has recently used sodium sulfate in the concentra- tion of antitoxin and finds that antitoxin may be concentrated with sodium sulfate without denaturing by heat fully as well as with ammonium sulfate with heat. The removal of the salt is, moreover, simplified when sodium sulfate is used. Miss Homer constructed curves showing the percentage of protein precipi- tated by increasing amounts of ammonium sulfate and of sodium sulfate under various conditions of acidity and heat treat- ment and found them to be similar. She did not find critical points in the curves for the precipitation of eu- or pseudoglobulin or of serum albumin from undiluted serum. Mellanby (6) had previously failed to find critical zones with ammonium sulfate or magnesium sulfate. The following limits for the precipita- tion of the serum protein at 35-40°C. were found: pseudoglobulin precipitation complete at 11.5 to 18.5 per cent of anhydrous sodium sulfate, albumin completely precipitated at 32 per cent of anhydrous sodium sulfate. The precipitation of the individual proteins began at, and was complete at, lower concentrations of ammonium sulfate or of sodium sulfate than was required for their precipitation from their respective separate solutions in saline solution. There was also a greater overlapping of the precipitation limits for the individual proteins than was found for their separate solutions. The limits given hold for concen- tration of protein between 6.5 and 10 per cent. When the concentration of protein was reduced to 2.5 per cent, 34 gm. of sodium sulfate in 100 ce. were required for precipitation. Further dilutions required an increase in the concentration of the pre- cipitating salt. The concentration of the electrolyte required for precipitation was affected by the reaction of the plasma, but a variation in reaction between pH 5.3 and 9.3 has only a slight effect on the result. Spiro (7) has studied the effect of the addition of neutral salts upon the hydrogen ion concentration of protein solutions. The addition of either sodium sulfate or magnesium sulfate to a pro- 96 Sodium Sulfate as Globulin Precipitant tein solution between the proportions of 1 part of salt to 9 parts of protein solution and 9 parts of salt and 1 of protein solution caused a change in the hydrogen ion concentration of the mixture from pH 7.73 to 7.38 for sodium sulfate and pH 7.73 to 7.27 for magnesium sulfate. On the other hand, the addition of ammo- nium sulfate causes a change of hydrogen ion concentration of from pH 7.73 to 5.91. In the light of the results obtained by Miss Homer and also Kauder (8) and Hofmeister (9) with regard to the variation in the quantity of salt required for the precipitation of proteins it became necessary to determine the optimum concentration of sodium sulfate for the precipitation of the total globulins. Dilution of serum has experimental evidence in its favor; Porges and Spiro and particularly Wiener (10) have diluted the serum before precipitation. Wiener, using ammonium sulfate, came to the conclusion that the accuracy of his results was enhanced by dilution; he obtained a smaller quantity of globulin under such conditions. Contrary to the findings of previous workers we found a critical zone and indications of more than one criti- cal zone by using concentrations of sodium sulfate which dif- fered from the preceding member of the series by 1 per cent of the anhydrous salt. Determination of the Critical Zone for the Precipitation of Total Globulins.—A series of 50 ec. portions of sodium sulfate at 37°C. containing increasing quantities of sodium sulfate was measured into or prepared in wide mouth glass-stoppered bottles. 5 cc. of blood plasma were added to each bottle and the bottles placed in the incubator at 37°C. After 3 to 12 hours in the incubator the solutions were filtered and an aliquot, 25 cc., of the filtrate was taken for analysis. Nitrogen was determined by the method of Kjeldahl.t. The results for the most complete experiments are given in Chart 1 together with other data concerning the range of constant nitrogen values. Later, in our desire to make use of sodium sulfate as the pre- cipitant of protein in the determination of globulin when using small quantities of plasma or serum, it seemed desirable to work at higher dilutions. In this work two procedures were adopted: 1 The addition of potassium sulfate is not necessary since sodium sulfate can take the place of potassium sulfate in the digestion. j i ) f j « 241 ; ai a ( aoe Ye ome il) Deeks stint, = , oe iA a aa ‘ i ho 7 , ae rey ‘ nf vs day ; + ke TAL = os ~ OS oe Ratist 4 Thats Ph | (mig “ai : i ir ts ' Paul E. Howe 97 (a) The plasma was diluted 1:10 with 0.8 per cent sodium chloride solution and 5 cc. of the diluted plasma, equivalent to 0.5 ce. of plasma, were added to 10 cc. of the required concentration of sodium sulfate; this gave a dilution of 1:33. (b) To 15 cc. of a given solution of sodium sulfate 0.5 ec. of plasma was added which gave a dilution of 1:32. The results obtained with these two procedures indicated that there was not an essential increase in accuracy by diluting the plasma before adding it to the salt solu- tion. In later work, therefore, the salt solution was added directly to the serum. Evidence showing the effect of dilution upon the precipitation of blood proteins was obtained from three sources: (a) by coagulat- ing the proteins of different samples of plasma at increasing concentrations of sodium sulfate (Chart 1), these data served to establish the probable critical zone of precipitation; (6) by precipitating the same sample of plasma over a short series of increasing concentrations of sodium sulfate which covered the probable critical zone and in increasing dilution (see Table I); and (c) by using a short series of concentrations of sodium sulfate with a number of different plasmas. Data from the last set of experiments are not included in this paper but appear in part in the succeeding one since they are to be considered only as confirma- tory results and are of particular value in connection with the relative accuracy of the macro method and the micro method described there. Data are presented in Chart 1 which indicate that at 37°C. a zone exists between 20 and 22 per cent? of anhydrous sodium sulfate in which an increase of 1 per cent of sodium sulfate does not cause an increase in the quantity of protein precipitated which is greater than the error of the method (a maximum analytical error of approximately 1.5 per cent). On either side of this zone a change of 1 per cent in the concentration of the salt gives a result which is distinctly greater or less than the results obtained at the critical zone. The critical zone between 20 and 22 per cent of anhydrous sodium sulfate was indicated by the results obtained in our first work 2 By 22 per cent of sodium sulfate we mean 22 gm. of sodium sulfate contained in 100 cc. of solution. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL, XLIX, NO. 1 98 Sodium Sulfate as Globulin Precipitant Par AE | 20 25 30 Na, SOx. s.525 iO 15 20 2s Cuart 1. Curves showing the quantity of nitrogen, as grams of nitrogen per 100 cc. of serum or plasma, remaining in the filtrate after precipitation with increasing concentrations of sodium sulfate at 37°C. Curves I and II are for plasma and Curve III for serum at a dilution of 1:9.1. Curve IV is for plasma and Curves V and VI are for serum at a dilution of 1:32.3. The abscissz are in terms of concentrations of anhydrous sodium sulfate expressed as per cent and the ordinates are expressed as grams of nitrogen in 100 ce. of blood. Paul E. Howe 99 at a dilution of 1:10. These values do not hold absolutely for all degrees of dilution nor for all samples of blood as can be seen from Table I and Chart 1. With increasing dilution of the plasma there is a tendency for the critical zone to shift in the direction of a higher concentration of salt, 21 to 23 per cent of sodium sulfate. Furthermore, at the higher dilutions the amount of protein precipitated is less than at the lower dilutions. These facts have been noted previously for various salts by other investi- gators. The range of salt concentrations over which there is not a marked increase in precipitation for a small increment in salt may extend only over 2 per cent of sodium sulfate. On the TABLE I. Data indicating the effect of the dilution of blood serum or plasma upon the precipitation of total globulins by sodium sulfate at 37°C. Results are expressed as gramis of nitrogen per 100 cc. of plasma and represent the quantity of nitrogen remaining in the filtrate after precipitation. Dilutions. Na2SO4 1:1 1:10 1:20 1:30 1:40 per cent 19 0.499 0.511 0.516 0.536 0.553 20 0.486 0.487 0.492 0.518 0.516 21 0.459 0.428 0.411 0.443 0.437 25: 0.437 0.416 0.393 0.430 0.4389 23 0.380 0.383 0.387 0.412 0.393 24 0.357 0.354 0.336 0.399 0.369 25 0.352 0.315 0.303 0.341 0.343 other hand, a sharp break in the curve of precipitation with in- creasing salt concentration has always been observed at approxi- mately the range of concentrations designated as the critical zone. The variability of the critical zone with dilution and to a less extent with the sample of blood raises a question as to the value of the results and as to the concentration of sodium sulfate which will probably be in the critical zone if only one precipitation is adopted. Our observations have been in most cases on blood which is freshly drawn. In some cases determinations have 3 As a convenience in measuring, 5 cc. of plasma were added to 50 ce. of salt solution which gave an actual dilution of 1:94. I 100 Sodium Sulfate as Globulin Precipitant been repeated on old plasma and have shown that in many cases 4 to 6 days later results obtained upon precipitation of the globulin are identical with those obtained with fresh blood. The fact is stated in support of the critical zone but not in support of the practice of analyzing old plasma. From the data presented and from that obtained in the analyses of blood in which two or more concentrations of sodium sulfate have been used, the precipita- tion of blood plasma or serum in a solution which contains 21.5 gm. of anhydrous sodium sulfate per 100 cc. of solution, or 21.5 per cent, is most likely to correspond with the critical zone. At dilutions of plasma of 1:10 or higher the concentration given falls in the critical zone. The critical zone may be on either side of 21.5 per cent of the salt, but this concentration has been one of at least two concentrations in which the quantity of globulin precipitated is essentially equal. Ordinarily results obtained at 21 and 22 per cent of sodium sulfate fall in the critical zone. Where it is desired to be assured of the presence of the critical zone determinations have been made at both 21 and 22 per cent of sodium sulfate. If both results at these concentrations agree it is assumed that the critical zone has been reached. If there is a marked difference in the results a precipitation is made at 23 per cent. In no case have we failed to obtain results in which one of the two pairs of concentrations showed an essentially con- stant degree of precipitation. Ocular evidence has often been obtained of a change in the quantity of protein thrown down when the precipitation took place in test-tubes under which conditions the volume of precipi- tated protein shows a marked increase at the critical zone. Comparisons of the protein precipitated by sodium sulfate at 21 to 22 per cent of sodium sulfate with that obtained with one- half saturated ammonium sulfate and magnesium sulfate at dilutions of 1:10 indicate that slightly less protein separates out when ammonium sulfate is used and more protein when magnesium sulfate is added. The results with magnesium sulfate have been repeatedly confirmed in other experiments. The data are con- tained in Table II. Pinkus (2) gives 18.8 gm. of anhydrous sodium sulfate per 100 cc. as the point of complete precipitation of the globulin of ox serum } dilute, horse serum ,!, dilute, serum globulin Paul E. Howe 101 7 per cent; and 20 gm. per 100 ce. solution for horse serum qv dilute and blood (ox) + dilute. The results given above for serum diluted approximately 1:10 agree with that of Pinkus. On the basis of our results we feel justified in accepting the results obtained at 21.5 per cent of sodium sulfate as representing the pre- cipitation of all globulins of blood serum or plasma. The results of Porges and Spiro pointed to the possibility of other critical zones at approximately 12 to 14 per cent and at 18 per cent of sodium sulfate. Working with increasing concen- trations of sodium sulfate we found critical zones at approximately the above concentrations (see Chart 1). These zones correspond roughly with the point of beginning precipitation in the chart TABLE II. Data relating to the quantity of nitrogen remaining in solution after precipitation with concentrations of sodium sulfate at the critical zone for total globulin and with one-half saturated ammonium sulfate and saturated magnesium sulfate at room temperature. Results are expressed as grams of nitrogen per 100 cc. of blood and represent the amount of nitrogen remaining in the filtrate after precipitation. peas Pig. Goat. Horse. per cent ‘ Sodium sulfate at 37°C............ 21.3 0.393 0.399 0.443 Aramonium sulfate. ....00.:....... 0.526 0.443 0.477 Mirresramine Sok. cxcee sss... | 0.323 0.385 of Porges and Spiro. They are not, however, always as definite as the critical zone at 21.5 per cent of sodium sulfate. The zone at 12 to 14 per cent of sodium sulfate would correspond to euglob- ulin which is usually precipitated by dilution and acidification with carbon dioxide or acetic acid and sometimes with saturated sodium chloride. The zone at 17.4 per cent of sodium sulfate would correspond to the second globulin of Porges and Spiro which is sometimes designated pseudoglobulin I. Relation between the Quantity of Protein Precipitated at 13.4 Per Cent of Sodium Sulfate and with Sodium Chloride or Acidi- fication with Carbon Dioxide.—It was assumed that the protein precipitated at the critical zone at about 13.4 per cent of sodium 102 Sodium Sulfate as Globulin Precipitant sulfate was probably euglobulin. If this was true the quantities of protein precipitated at this concentration and with other methods of precipitation should agree over a wide range of con- centration of protein. Comparisons were, therefore, made of the quantity of protein precipitated by 13.5 per cent of sodium sulfate at 37°C., and by saturation of the diluted serum or plasma, with carbon dioxide, and with sodium chloride at room termpera- ture. For the carbon dioxide precipitations 5 or 0.5 cc. of serum or plasma were pipetted into a cylinder and 50 or 15 cc. respec- tively of distilled water added after which carbon dioxide was passed through the solution slowly for from } hour to 2 hours TABLE III. Results obtained by precipitation of different sera with (a) 13.5 per cent sodium sulfate, (b) saturated sodium chloride, and (c) saturation with carbon dioxide. Results are expressed as grams of nitrogen per 100 cc. of serum and represent the amount of nitrogen remaining in solution after precipitation. : Total nitrogen. Sodium sulfate. Sodium chloride. Carbon dioxide. 0.658 0.595 0.621 0.573 0.665 0.665 0.661 0.578 0.693 0.600 0.569 0.621 0.695 0.613 0.615 0.652 0.696 0.630 0.615 0.652 0.731 0.595 0.661 0.600 0.863 0.718 0.792 0.735 0.901 0.665 0.718 0.569 1.046 0.915 0.929 0.915 255 1.000 0.997 0.957 according to the procedure of Robertson (11). The solution was then filtered and aliquot portions of the filtrate were taken for analysis. For saturated sodium chloride similar quantities of serum or plasma were taken and 50 or 15 cc. of saturated sodium chloride added after which solid sodium chloride was added in excess. These solutions were permitted to stand for 12 hours with occasional shaking. Table III contains comparative results with the three methods. The data cover a considerable range of protein concentrations and include some sera which contain practically no euglobulin. The average difference between the results obtained with sodium Paul E. Howe 103 sulfate and with saturated sodium chloride is a precipitation of 0.027 gm. less of protein nitrogen per 100 ce. of blood with sodium chloride than with sodium sulfate. With relation to carbon diox- ide, there was a precipitation of 0.021 gm. more protein nitrogen per 100 ce. of blood by carbon dioxide than with sodium sulfate. The average difference between results is not much beyond the experimental error but the general trend of the results is charac- teristic of the procedure. Similar results have been obtained in a larger number of cases with sodium chloride and sodium sulfate which give an average difference, 0.018 gm. of protein nitrogen, which is approximately the same as the one given above. Com- parisons of the quantity of protein precipitated at 12.5, 13.5, and 14.5 per cent of sodium sulfate have shown repeatedly that the amount of protein precipitated at 13.5 and 14.5 per cent of sodium sulfate is essentially the same and is always more than that precipitated by 12.5 per cent of sodium sulfate. The selec- tion of 13.5 per cent of sodium sulfate at 37°C., we believe, as nearly represents the euglobulin fraction as can be determined by such quantitative procedures. There is an advantage in using sodium sulfate instead of car- bon dioxide or sodium chloride in that very constant conditions can be maintained. With carbon dioxide losses by evaporation and frothing are to be contended with, while with sodium chloride there is always the possibility that saturation is not complete. The Presence of Pseudoglobulin, Precipitated at 17.4 Per Cent of Sodium Sulfate, in Blood.—The presence of two pseudoglobulins in blood has often been discussed. Porges and Spiro have pre- sented evidence in favor of the occurrence of two pseudoglobulins as the result of their precipitations from dilute serum with mag- nesium sulfate, sodium sulfate, ammonium sulfate, and other salts. Haslam (12) has definitely concluded that there are not more than two serum globulins in blood serum, euglobulin and pseudoglobulin. His statement is based upon results obtained after repeated precipitation of globulins with ammonium sulfate. The problem was, therefore, in such a state that it did not seem that our somewhat indefinite, but still suggestive, critical zone at about 16.4 to 17.4 per cent of sodium sulfate was significant. We are not now assured of the significance of the precipitation at 17.4 per cent of sodium sulfate. Additional evidence has been 104 Sodium Sulfate as Globulin Precipitant obtained, however, which points very strongly toward the pres- ence of a protein or protein complex whose precipitation is com- plete at approximately 17.4 per cent of sodium sulfate. The evidence rests upon two facts, in addition to that already in the literature, (a) the indication of a critical zone at 16.4 to 17.4 per cent of sodium sulfate, already referred to, and (b) the absence of protein which is precipitated at 16.4 to 17.4 per cent of sodium sulfate in certain bloods. A precipitate first occurs at 18.4 per cent of sodium sulfate. Under suitable conditions the blood of the same animals will contain, within a few hours, large quantities of protein precipitable at 13.5 per cent of sodium sulfate, or by carbon dioxide, or saturated sodium chloride, and at 17.4 per cent of sodium sulfate. Other work associated with this and confirming this observation will be presented in due time. The presence of a protein in this case rests upon the acceptance of the definition of a euglobulin as one which is precipitated by acidification with carbon dioxide in dilute solution. That the amount of protein precipitated by carbon dioxide from blood is practically the same as that obtained under two other conditions has just been discussed and points to the separation of at least a mixture .of fairly definite composition. This being so the failure to obtain a precipitate until at least 18.4 per cent of sodium sulfate has been added to blood serum under some conditions and the presence of a precipitate under others is an indication of the existence of a protein or protein complex precipi- table between the limits of 13.5 and 17.4 per cent of sodium sulfate. The quantity of protein precipitated between 13.5 and 17.4 per cent of sodium sulfate in one series of experiments is roughly equal to that precipitated at 13.5 per cent. There are bloods in which the quantity of protein precipitated between these limits is greater than at 13.5 per cent of sodium sulfate. Judging from the volume of precipitate the greatest proportion of the precipi- tation may occur between 16.4 and 17.4 per cent of sodium sul- fate. This fact made us hesitate between 16.4 and 17.4 per cent of sodium sulfate as the percentage which would represent the approximate completion of the precipitation of pseudoglob- ulin I. On the other hand, there are cases in which no precip- itate occurs at 17.4 per cent of sodium sulfate. When the crit- Paul E. Howe 105 ical zone is most marked at the range of concentrations under consideration, 17.4 per cent of sodium sulfate is always one of the two concentrations involved and we have, therefore, adopted this concentration as the point which we consider as best repre- senting the quantity of euglobulin plus pseudoglobulin I present in blood. In case plasma is being studied the value at 17.4 per cent will represent fibrinogen, euglobulin, and pseudoglobulin I. DISCUSSION. Data have been presented which indicate that as increasing quantities of sodium sulfate are added to diluted serum or plasma or serum at 37°C. the amount of protein thrown out of solution increases. There are at least three points in such a series at which an increase of 1 per cent of sodium sulfate does not produce as large an increase in precipitation as will be caused by the concen- tration of salt preceding or succeeding these concentrations. These points, or critical zones, are at 13.5 to 14.5, 16.4 to 17.4, and 21 to 22 per cent of sodium sulfate. This is particularly true of the concentrations 13.5 to 14.5 and 21 to 22 per cent of sodium sulfate. At 16.4 to 17.4 per cent of sodium sulfate it is not al- ways possible to demonstrate a critical zone. Basing our conclusions on the quantity of protein precipitated as determined by analysis of the filtrate from such precipita- tions* two of these critical zones agree very closely with other methods for determining proteins. Results at the zone at 13.5 to 14.5 per cent of sodium sulfate agree closely with ‘those ob- tained with saturated sodium chloride and carbon dioxide. At 21 to 22 per cent of sodium sulfate the results are approximately those obtained with magnesium sulfate and ammonium sulfate both of which have long been accepted as a means of completely * In using the procedures described it is assumed that the aliquot taken from the filtrate after precipitation contains a true proportionate amount of the unprecipitated protein and that there has not been any adsorption of the unprecipitated protein by the precipitated protein nor by the filter paper. The results of Spiro (13) indicate that precipitation is probably not complete and that there is a small amount of protein remaining unpre- cipitated. We have found that when diluted serum is filtered that there is a small loss of nitrogen in the process which may be slightly greater than the experimental error. } 106 Sodium Sulfate as Globulin Precipitaat precipitating globulins. For the acceptance of the zone at 17.4 per cent of sodium sulfate the evidence rests in part upon the ab- sence of protein precipitable up to that concentration in certain bloods. Whether or not results obtained by precipitation of proteins from a mixture of proteins with salts represent separations of pure proteins is an open question. The considerable mass of literature on this subject is in favor of the opinion that the protein thrown down is a mixture of proteins; (a) present as compounds, (6) due to the adsorption of other proteins by the precipitated protein, or (c) because the precipitation limits overlap. One fact stands out in our experiments, however, which is applicable to any concen- tration of salt; under constant conditions of temperature and con- centration of salt a constant amount of protein is precipitated. Robertson (11) in developing his refractometric procedure for the determination of globulins substantiated his use of ammonium sulfate for the precipitation of total globulins upon the constancy of the results obtained with definite concentrations of salt. From our subsequent work with a method developed from our findings presented here we agree with Robertson in referring to precipi- tation with ammonium sulfate, that ‘“‘if the proportion of this substance is different in the serum of different individuals or species, we may be fairly confident, therefore, that the quanti- tative relations of the globulin and albumin,”’ or intermediate globulin, ‘‘groups are different in these animals.” Our work has covered a number of species of mammals having blood of varying protein content and the limits given have held in every case. On the other hand, in carrying out a problem involving the determination of serum or plasma proteins we would strongly advise the use of two succeeding concentrations of salt at the critical zone until it had been determined that the critical zone was present at the concentrations of salt used in that particular case. SUMMARY. 1. Sodium sulfate solutions at 37°C. may be used to precipitate the proteins of blood into fractions corresponding to those usually separated by carbon dioxide or saturated sodium chloride, eu- globulin, and magnesium sulfate or ammonium sulfate, total globulin. Paul E. Howe 107 2. Critical zones in the curve representing the precipitation of protein with increasing salt coneentration have been located at 13.5 to 14.5, 17.4, and 21 to 22 per cent of anhydrous sodium sulfate at 37°C. For the purpose of estimating the quantity of protein present at these zones, 13.5, 17.4, and 21.5 per cent of sodium sulfate is recommended. 3. Evidence has been presented which tends to substantiate the presence of two globulins in blood serum in addition to euglob- ulin; pseudoglobulin I and pseudoglobulin II whose precipita- tions are complete at approximately 17.4 and 21 to 22 per cent of sodium sulfate respectively. BIBLIOGRAPHY. . Cullen, G. E., and Van Slyke, D.D., J. Biol. Chem., 1920, xli, 587. . Pinkus, S. N., J. Physiol., 1901, xxvii, 57. . Porges, O., and Spiro, K., Beitr. chem. Physiol. u. Path., 1903, iii, 277. Haslam, H. C., J. Physiol., 1905, xxxii, 267. Homer, A., Biochem. J., 1919, xiii; 278. . Mellanby, J., J. Physiol., 1909, xxxviii, 288. . Spiro, K., Biochem. Z., 1913, lvi, 11. . Kauder, G., Arch. exp. Path. u. Pharmacol., 1866, xx, 411. . Hofmeister, F., Arch. exp. Path. u. Pharmacol., 1888, xxiv, 247. . Wiener, H., Z. physiol. Chem., 1911, Ixxiv, 29. . Robertson, T. B., J. Biol. Chem., 1912, xi, 197. . Haslam, H. C., Biochem. J., 1913, vii, 492. . Spiro, K., Beitr. chers. Physiol. u. Path., 1904, iv, 300. Se ee fieene * “~~ 7 { iI Pane se. ; Bie 4 Ae ‘ie ett. Pe A i ry fet meet 2 7! d = ogfemayteay | ae eh , et 57 a = Lays Pe any ‘ ° weil fi y Ee | yt Ui ' - ‘ ’ —_ Sy 7 4 er P : Le \fu * ‘ i \ y hes “4 L . ot 4 . ‘ih M a a = | j A? ¥ . ‘ 4 i a ’ THE DETERMINATION OF PROTEINS IN BLOOD—A MICRO METHOD. By PAUL E. HOWE. (From the Department of Animal Pathology of The Rockefeller Institute for Medical Research, Princeton, N. J.) (Received for publication, September 1, 1921.) The use of sodium sulfate instead of ammonium sulfate in the precipitation of globulins makes it possible to determine the proteins of plasma or serum in small quantities of blood. The basis for the selection of sodium sulfate has been discussed in the preceding paper (1). Robertson (2) described a proce- dure for the determination of blood proteins in small quantities of blood with the aid of the refractometer. Cullen and Slyke (3) have proposed a procedure which gives consistent re- sults and which does not require any special apparatus beyond that to be found in any laboratory. Their method, however, required 5 cc. of plasma for each constituent of the blood deter- mined. The determinations described below can be performed with the usual laboratory apparatus and require but 0.5 cc. of plasma or serum for each determination. The procedures involve the precipitation of fibrinogen with calcium chloride, the globulins with definite concentrations of sodium sulfate at 37°C., and non-protein nitrogen with tri- chloroacetic acid. In the case of fibrinogen and non-protein ni- trogen the technique of Cullen and Van Slyke is followed. The globulins are precipitated by adding a concentration of sodium sulfate which is greater than the required percentage by the amount of sodium sulfate necessary to produce the desired percentage when added to the blood sample. The solutions are prepared by dissolving the required quantity of sodium sulfate in a little less than the final volume, which requires heat for the higher percentages, and then diluting to volume at 37°C. All precipi- _ tations and filtrations with sodium sulfate are carried out in the 109 J 110 Determination of Proteins in Blood incubator or hot room. The following concentrations of sodium sulfate are needed: 14, 18, and 22.2 per cent. When 15 ce. portions of these solutions are added to 0.5 ce. of blood the final concentrations are approximately 13.5, 17.4, and 21.5 per cent of sodium sulfate respectively. At 13.5 per cent of sodium sulfate euglobulin is precipitated, at 17.4 per cent euglobulin and pseudo- globulin I are precipitated, and at 21.5 per cent all globulins are thrown out of solution. In case blood plasma is used fibrinogen is present in each case and the nitrogen representing this protein must be deducted. Suitable blanks must be made for each deter- mination. The volume of solution used, 15.5 ec. (15 cc. of salt solution ‘plus 0.5 cc. of blood), permits duplicate determinations to be made on each precipitation. This does not insure against errors in precipitation, but it has been our experience that simultaneous duplicate precipitations almost invariably agree. When it is desired to make duplicate precipitations it is advised that 13.5 and 14.5 per cent, 16.4 and 17.4 per cent, and 21 and 22 per cent of sodium sulfate, final concentrations, be used. With these concentrations the values obtained with each pair should agree within experimental error, except perhaps in the case of 16.4 and 17.4 per cent of sodium sulfate. Precipitations are made in test-tubes or 50 cc. centrifuge tubes and then closed with a rubber stopper. ‘The filtrations are con- ducted in the incubator using a dry 9 em. filter paper. It is desirable to wet the filter paper with a small amount of the solu- tion to be filtered before pouring on the bulk of the solution con- taining the precipitate. The funnels are covered with watch- glasses. 1 inch test-tubes held in test-tube racks are convenient for filtration. With these tubes and the cover-glasses a reason- ably tight filtration system is obtained. For measuring, the accurately calibrated Ostwald pipettes and the 15 ec. graduated pipettes introduced by Folin are used. The nitrogen determinations are conducted in large Pyrex test-tubes in general according to the original micro procedure of Folin and Farmer (4), and the distillations are carried out, according to the procedure of Folin and Wu (5) in their system of blood analysis, without cooling the distillate. In distilling, a Pyrex connecting tube is used which carries a distilling head and Paul E. Howe jaa has an enlargement on the tube which dips into the acid to guard against mechanical transfer of alkali and back suction. For titrations we use standard hydrochloric acid and sodium hydroxide which are approximately 0.05 and 0.025 n respectively. Our burettes deliver 25 cc. and are graduated to 0.05 ce. Methyl red is used as an indicator. The determinations are as follows: Plasma is collected so that it contains 0.5 per cent of potassium oxalate. Both plasma and serum are centrifuged until clear. Total Nitrogen.—0.5 cc. of plasma or serum is placed in a large Pyrex test-tube and the 2 cc. of concentrated sulfuric acid, 1 drop of 5 per cent copper sulfate, and a quartz pebble are added; the solution is digested over a free flame until clear, and then 7 to 10 minutes longer. Cool 3 to 5 minutes, add 25 to 30 ec. of ammonia-free distilled water, a small amount of taleum powder or powdered pumice stone, and concentrated sodium hydroxide solution sufficient to neutralize the acid, and distill into standard acid. In place of using two 0.5 cc. portions of blood 15 ec. of 0.8 per cent sodium chloride solution may be added to one portion of 0.5 ce. and two 5 ce. portions of the diluted plasma taken for analysis. Fibrinogen.—0.5 ec. of plasma is measured into a tube, 14 ce. of 0.8 per cent sodium chloride solution at room temperature are added, then 1 cc. of 2.5 per cent calcium chloride, a small crystal of thymol, and the tube is stoppered. The tube and contents are allowed to stand until the fibrin is formed and then filtered on a dry filter. Two 5 ce. portions of the filtrate are taken for analysis. Euglobulin.—0.5 ce. of plasma or serum is measured into a tube, 15 cc. of 14 per cent anhydrous sodium sulfate at 37°C. and a little thymol are added, and the tube is stoppered, shaken, and allowed to stand for at least 3 hours, or until the precipitate has settled. The solution is then filtered through a dry filter and two 5 ec. portions are taken for analysis. The results represent euglob- ulin in the case of serum and fibrinogen plus euglobulin in the case of plasma. Euglobulin Plus Pseudoglobulin I—The procedure is the same as for euglobulin except that 18 per cent sodium sulfate is used. 112 Determination of Proteins in Blood Total Globulins.—The procedure is the same as in euglobulin except that 22.2 per cent of sodium sulfate is used. Non-Protein Nitrogen.—0.5 cc. of plasma or serum is measured into a tube and 15 ce. of 5 per cent trichloroacetic acid at room temperature are added. The remainder of the procedure is the same as in euglobulin. The calculations of nitrogen are those ordinarily associated with Kjeldahl determinations. The volume of solution from which the aliquot portions for analysis are taken is 15.5 ec. We have expressed our results in terms of grams of nitrogen in 100 ec. of blood. As the result of the analytical procedures the following results can be calculated for serum: Total nitrogen. Euglobulin nitrogen = Total nitrogen — nitrogen in filtrate from 13.5 per cent sodium sulfate precipitation. Pseudoglobulin I nitrogen = Nitrogen in filtrate from 13.5 per cent sodium sulfate precipitation — nitrogen from 17.4 per cent sodium sulfate precipitation. Pseudoglobulin II nitrogen = Nitrogen in filtrate from 17.4 per cent sodium sulfate precipitation — nitrogen in filtrate from 21.5 per cent sodium sulfate precipitation. Total globulin nitrogen = Total nitrogen — nitrogen in filtrate from 21.5 per cent sodium sulfate precipitation. Albumin nitrogen = Nitrogen in filtrate from 21.5 per cent precipitation — non-protein nitrogen. Non-protein nitrogen = Nitrogen in filtrate from trichloro- acetic acid precipitation. For plasma the euglobulin is estimated by subtracting the filtrate nitrogen from. the nitrogen in the filtrate after the pre- cipitation of fibrinogen. Table I contains some comparative data on the Cullen-Van Slyke procedure and the micro method. Determinations using sodium sulfate and magnesium sulfate in the Cullen-Van Slyke method are included. Determinations for pseudoglobulin I are not included in the table since they were not made at the time the plasma was analyzed. The results are expressed in terms of the nitrogen remaining in the filtrate after precipitation with- out calculating the various fractions. Paul E. Howe TABLE I. 113 Comparative results obtained with the method of Cullen and Van Slyke “macro’’ and the micro method. Results are expressed as grams of nitro- gen per 100 cc. of blood remaining in the filtrate after precipitation. Sodium sulfate. iS-o per cent... 20.9 ml * 22.0 “ 23.2 “ “ Trichloroacetic acid.| 0.06 Ammonium sulfate.. Magnesium sulfate.. ore CO Ne Pie. | bene . Howe, P. E., J. Biol. Chem., 1921, xlix, 93. . Robertson, T. B., J. Biol. Chem., 1918, xxii, 2338. . Cullen, G. E., and Van Slyke, D. D., J. Biol. Chem., 1920, xli, 587. . Folin, O., and Farmer, C. J., J. Biol. Chem., 1912, xi, 493. . Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. Horse. Cow. Macro. | Micro. | Macro. | Micro. | Macro. | Micro. | Macro. | Micro. L647, (27 205) 05 1.10 1.31 | 1.39 } 0.91 | 0.92 | 0.98 | 0.97 | 1.08 1.01 | 1.00 | 0.80 | 0.81 | 0.84 | 0.83 | 0.90 1.01 0.74 0.84 Ook 0.41 0.39 0.43 0.29 0.39 0.40 0.27 0.40 0.38 0.27 0.31 0.37 0.07 | 0.03 | 0.02 | 0.05 | 0.07 | 0.04 0.53 0.44 0.48 0.32 0.39 BIBLIOGRAPHY, oe Wii Raearee el oe al ea TN tsi © 5 TR ORSEMED fa Vo" i "ATS ; , : ¢ 5 a oe ~ $y ee iia) : aah > =F Be tied Pr Roa - 7 a fi “ 4 4 ' ye P + } ier, 24 r ty salt A. ae ! i 3 7 IN 7 : - [MliAeh toe na ¢ P ay hes is was ‘ hal Ais a os y , y a ry .) eth 0 i> OL i: left. found. ranted used. used. per cent per cent (er, cc. cc. cc. per cent 32 22.0 45 .2 0.0 10.0 0.0 35 0.340 33 24.8 48.4 0.0 5.0 5.0 35 0.192 34 23.0 60.0 5.0 5.0 0.0 35 0.170 3 16.2 58.0 5.0 5.0 5.0 30 0.153 36 17.9 59.2. | 1070 5.0 0.0 30 0.102 37 15.0 61.6 0.0 5.0 10.0 30 0.229 Comparison of the results in (32), (83), and (34) shows that the sodium hydroxide is somewhat less active than the potas- sium hydroxide when used in conjunction with ammonium hy- droxide. This is confirmed by the results in (36) and (37). Other unpublished experiments on the influence of these two bases on the spontaneous decomposition of peroxide show about the same degree of difference in their action, when used in the absence of ammonia. The results in (34) in which one equivalent of all three bases was used stand between (36) and (37), which is to be expected if the small difference in their behavior is correct. Biologically these results are of interest because they show that if under any conditions fatty acid oxidation in the organism resembles these oxidations either the potassium or sodium salt of mixtures would be attacked with about the same ease in the presence of ammonia. The results in this section are of especial interest in another way because they show that the catalytic influence of ammonium, sodium, and potassium as the bases is just the reverse of their influence as the phosphates, upon the peroxide oxidation of a number of simple organic compounds. Further experiments upon mixtures will be done in developing these unpublished results on phosphates. Partial Interpretation—The experiments in the preceding section bring out a number of facts concerning the interpreta- tion. (a) The ammonium and potassium butyrates here under consideration are the salts of a weak and a strong base, respec- tively, with a weak acid. At these dgutions there would be con- siderably more free butyric acid produced by hydrolysis when —— —_—— ee x? E. J. Witzemann iar ammonia only is used. That this free acid is not the only sig- nificant factor in the catalysis isshown by the fact that the simul- taneous presence of both bases promotes the greatest oxidation. (b) That the mere presence of many butyrate ions does not de- termine the oxidation, is shown by the fact that in the presence of ammonium hydroxide alone, in which, of all the experiments, there is the smallest concentration of butyrate ions, the oxida- tion proceeds much faster than when potassium hydroxide only is used where the butyrate ion concentration is greatest. (ce) That the butyrate molecules do not determine the oxidation is shown by the fact that oxidation is slowest where the concentra- tion of these is greatest; namely, in the solutions containing potas- sium hydroxide alone. (d) That the capacity to decompose peroxide spontaneously may be a factor is suggested by the fact that potassium decomposes peroxide much more slowly than ammonium hydroxide. That this is not the sole factor is shown by the experiments with sodium bicarbonate which decomposes peroxide well but does not facilitate these oxidations. (e) The concentration of the hydroxyl ion has no definite relation to the velocity of the oxidation as observed with these three bases. The known facts about these oxidations are not yet sufficient for a satisfactory interpretation. The statements given in the next two paragraphs suggest a partial interpretation. The oxidation of butyric acid in the presence of ammonia de- scribed above may belong to the so called “coupled” or induced oxidations. The formation of a peroxide of ammonia may represent one phase of this coupling. That such a peroxide of ammonia may be obtained under suitable conditions was shown by Melikoff and Pissarjewski.’ If this peroxide is the intermediate concerned it is much more unstable than the peroxide of potas- sium since when equal concentrations of the two bases are placedin solutions of hydrogen peroxide the ammonium hydroxide decom- poses the peroxide much more rapidly than the potassium hy- droxide does. The conceptions of an ideal catalyst and the theory of dislocation of Béeseken!® offer a concrete partial interpreta- tion of this part of the effect of ammonia. According to this ® Melikoff, P., and Pissarjewski, L., Ber. chem. Ges., 1897, xxx, 3144. 10 B6eseken, J., Kon. Akad. Wetensch. Amst. Versl., 1914-5, xxiii, 291; Rec. trav. chim. Pays-Bas, 1920, xxxix, 622. 138 Oxidation of Butyric Acid view of catalysis that catalyst will be best which has the greatest tendency to form the least stable complex with the substrate. If it is granted that the data justify this much of an interpre- tation for the interaction of ammonia and peroxide they fail to indicate how such a complex activates the oxidation of butyric acid. Sweeping hypotheses on this subject would be undesirable, especially in view of the fact that equally anomalous data are already in hand showing that when these bases are used as phos- phates the effects on the oxidation of butyric acid are quite different. Possible Biological Significance of the Ammonia Effect. — The experiments showed that butyric acid, whether free or combined - with sodium or potassium, is peculiarly susceptible to oxidation by peroxide when ammonia is also present, and that under these conditions relatively little carbon dioxide but much acetone is obtained. ‘This high degree of acetone formation depends upon the simultaneous presence and the interaction of ammonia, per- oxide, and butyric acid. Without entering into details it would seem that these three components play an active role in metabo- lism, particularly in the hepatic metabolism, and that acetone as acetoacetic acid is also formed in this organ. It will be recalled, for instance, that Embden and Kalberlah" found that the liver is the only organ that, on perfusion in the surviving state with normal blood, gives acetoacetic acid. Moreover there is in the liver an active traffic in ammonia equivalents involved in its important deaminizing and urea-forming functions. That it can actually supply this component of this reaction system better than other organs has recently been brought out by experi- ments of Keeton” of this laboratory, who found, in harmony with old observations, that when an inorganic acid is administered by stomach to dogs it is partially excreted as the ammonium salt, and causes an increase in the urinary ammonia nitrogen, both relative and absolute. When, however, the acid is injected into a peripheral vein, although it may cause some absolute increase of the urinary ammonia, it causes little or no increase in the quan- 1 Embden, G., and Kalberlah, F., Beitr. chem. Physiol. and Path., 1906, viii, 121. Embden, G., and Engel, H., Beitr. chem. Physiol. u. Path., 1908, x1, 323. 12Keeton, R. W., J. Biol. Chem., in press. E. J. Witzemann 139 tity of ammonia nitrogen relative to the total nitrogen. This difference in the behavior of acid administered by the portal and the peripheral routes, respectively, is interpreted as due to the fact that ammonia is an available base in the liver and enters into the neutralization of acid when the acid is neutralized in the liver, but plays little part in the neutralization of acids else- where in the body. It may not seem amiss to suggest, since the same type of oxi- dation as that observed in the above described test-tube experi- ments is known to occur in the liver in a high degree, that this oxidation in the liver may possibly be favored as in the test-tube by ammonia. The question has long been discussed as to whether the forma- tion of acetoacetic and $-hydroxybutyric acids in the body rep- resents a purely abnormal type of fatty acid oxidation, that only occurs under certain conditions (as in diabetic acidosis, etc.) or whether it is a step in the normal oxidative breakdown of the fatty acids, having an even number of carbon atoms, which is obscured in health because of the rapid occurrence of subsequent events. The first view would make acetoacetic acid an abnormal product, the latter would make it a normal product, which in ac- idosis of the diabetic type fails to break down further. Neu- bauer" in his well known study of the breakdown of the fatty and amino-acids favored the view that the formation of acetoacetic acid is a normal step. The above experiments and discussion would tend to favor the view that the formation of acetoacetic acid may be considered as a normal step in the oxidation of fatty acids, in harmony with Neubauer’s conception, and particularly in media in which am- monia is available. Moreover the special occurrence of this type of oxidation in the liver might possibly be brought into relationship with the special availability of ammonia in this organ, it being less characteristic of the oxidation of fatty acids elsewhere in the body, where relatively less ammonia is available. According to this line of thought a certain special physiological function of the body, namely its power to form acetoacetic acid, would be associated with a special characteristic of ammonia 13 Neubauer, O., Deutsch. Arch. klin. Med., 1908-09, xcv, 211. 140 Oxidation of Butyric Acid or its equivalent and the suggestion would follow that where acetoacetic acid is formed in the body ammonia may be present and acting in this characteristic way. The problem of why in health the evidence of acetoacetic acid formation is suppressed, 7.e. the problem of ‘‘antiketogenesis”’ (or “ketolysis’’) is clearly another matter. SUMMARY. Experiments were performed with the object of determining the reason for the difference in the results obtained in the oxida- tion of butyric acid with hydrogen peroxide in the presence of different alkaline substances. In the presence of potassium hydroxide in amounts varying from 0.20 to 4.0 equivalents no appreciable oxidation of butyric -acid took place, as was proved by the almost quantitative re- covery of the unchanged acid. In the presence of ammonium hydroxide in amounts varying from 0.20 to 10.0 equivalents much oxidation took place. The amount of oxidation in the presence of ammonia was found to increase with increase in the ammonium hydroxide, other things being equal, until more than 4.0 equivalents of ammonium hydroxide were present, after which it decreased somewhat. This decrease with large excess of ammonium hydroxide was due to the spontaneous liberation of oxygen by the action of am- monium hydroxide on the hydrogen peroxide, before it could be utilized in oxidation. If one equivalent of both ammonium hydroxide and potassium hydroxide is used more oxidation takes place than if two equiv- alents of either of these bases are added. The type of oxidation observed in these experiments was mainly of the 6 type or the conversion of butyric acid into acetone and 1 molecule of carbon dioxide. Consequently the statement in the above paragraph is equivalent to saying that the presence of ammonia in a solution containing potassium butyrate and hydrogen peroxide catalyzes the oxidation of butyric acid to acetone and carbon dioxide. The additive effect of ammonium and potassium here described in favoring oxidation constitutes a chemical analogy to the bio- logical effects of certain mixed salts. E. J. Witzemann 141 The experiments on the additive effects of sodium and ammon- ium show that this pair acts nearly the same as potassium and ammonium. There is just enough difference in the behavior of potassium and sodium to constitute a confirmation of unpublished observations of specific effects of these bases as phosphates. A few suggestions for a partial interpretation were made. It is clear, however, in this case, as in the phosphate effect on glu- cose previously reported, that the rédle of alkalinity and alkali is not primary in this oxidation. A satisfactory interpretation is particularly desirable in view of the possible biological signifi- cance of the results. Finally in discussing the possible biological significance of these results it is suggested that the ammonia effect here de- scribed may be the agency by which the normal oxidation of fatty acids is brought about in the liver. This suggestion rests upon the fact that in the liver the substances required for this effect are all available and that this organ also normally shows the greatest tendency to form acetoacetic acid. ANTIKETOGENESIS. III. CALCULATION OF THE KETOGENIC BALANCE FROM THE RESPIRATORY QUOTIENTS. By PHILIP A. SHAFFER. (From the Laboratory of Biological Chemistry, Washington University School of Medicine, St. Louis.) (Received for publication, August 29, 1921.) In the first paper of this series (1) it was shown that the oxida- tion of glucose by hydrogen peroxide in alkaline solution brings about the rapid disappearance of acetoacetic acid if the latter be present; and this ‘‘ketolytic”’ reaction was described as an in vitro analogy to the well known ‘‘antiketogenic”’ effect of food carbohydrate in preventing or abolishing the appearance of the “acetone bodies” as metabolic end-products in man. In asecond paper (2) the attempt was made to calculate the metabolic mixture of different subjects, in terms of the relative molecular amounts of all substances which are convertible in the body into aceto- acetic acid and its related acetone and hydroxybutyric acid (ketogenic substances), and of all substances which are conver- tible into glucose and have an opposing, antiketogenic action. The calculation was based upon certain assumptions, some of which have experimental justification while others are of the nature of first trial guesses. The main assumptions are the following: 1. Each molecule of fat is convertible into (a) 3 molecules of acetoacetic acid and (6b) 0.5 molecule of glucose, or its equivalent of antiketogenic derivative. 2. Protein is convertible, (a) into antiketogenic glucose or its equivalent to.the extent of 3.6 gm. for each gm. of urine nitrogen and (b) into acetoacetic acid for each molecule of leucine, phenyl- alanine, and tyrosine, it being calculated that each gm. of urine nitrogen corresponds to approximately 10 millimols of ketogenic substance. (c) The amino-acids, valine, lysine, histidine, and tryptophane are neutral as to ketogenesis. 143 144 Antiketogenesis. III 3. Carbohydrate exerts its antiketogenic action in the form of glucose or other hexose, 1 gm. of which is therefore equivalent to (1,000+180=) 5.56 millimols of antiketogenic substance. In the paper referred to, the calculation of the metabolic mix- ture was made on subjects under such conditions as to permit the additional assumptions, that the amount of carbohydrate burned was the amount fed, and that the fat catabolized was represented by the difference between the estimated total calories and the sum of the calories from carbohydrate and protein, the protein being indicated by the nitrogen excretion. Such a calculation of the total ketogenic and antiketogenic substance in the metabolic mixture being oxidized by a number of different subjects appeared to show that definitely abnormal amounts of acetone bodies first appeared when the molecular ratio of ketogenic to antiketogenic substance exceeded 1:1. This fact is interpreted as indicating that the avoidance of the appearance of the acetone bodies is due to the ‘‘ketolytic’’ decomposition of acetoacetic acid as fast as it is formed, by its reaction with a product of glucose oxidation (or related substance from certain amino-acids and from glycerol), there being for this purpose nearly always in normal subjects on ordinary diets an abundance of glucose and other antiketogenic derivatives undergoing catabo- lism. Whenever the rate of production of ketolytic material falls below the rate of the catabolism of ketogenic substances as happens when the normal subject greatly reduces the carbo- hydrate intake, and in the diabetic when his power of metaboli- zing carbohydrate is sufficiently low, there is a deficit of ketolytic substance and in proportion to this deficit, acetoacetic acid accumulates, is In part converted into acetone and hydroxy- butyric acid, and in the three forms is excreted as abnormal end- products. According to this conception the starvation acidosis of any subject and the often more severe acidosis of diabetes are alike the result of, and in proportion to the unusual ratio between the rates of the catabolism of ketogenic substance on the one hand, and of the formation of the necessary ketolytic substances on the other. It is the purpose of the present paper to describe another method of making a similar calculation of the metabolic mixture EE P. A. Shafte. 147 and of the ketogenic-antiketogenic balance, from the respw, e/o'x lo a seie woe -| 5,310 | 10,200 600 300 600 300 5,610 | 10,800 900 rgoriminal fatecs.tes sls. sisa seeks coe 5,610 | 10,400 600 ATG Ta eo oe rrr 400 300 PMOL EGO ee ons dance ees 400 200 I6s BOSTON OES Ae 100 The reaction may be written as follows: (1) 100 mols fat + 300 mols H.O + 50 mols O2— 50 mols glucose (100 X 874 gm.) (50 X 180 gm.) + 300 mols fatty acid + 100 mols H.O (300 X 278 gm.) If the fatty acid be then oxidized, (2) 300 mols fatty acid + 7,560 mols O02 5,310 mols CO, + 5,100 mols H.0 The respiratory quotient for the complete reaction, oxidizing the fatty acids and converting the glycerol into glucose, would be 5,310 CO: 7,560 + 50 O2 In the above reaction the heat liberated is 874 gm. fat X 9.461 calories (3) = 8,269 calories Deduct 90 gm. glucose 338 .6 7,930.4 The caloric value of a liter of oxygen, when the fatty acid of fat is oxidized and the glycerol converted to glucose (at R.Q. 0.698) is therefore 7,930 (75.6 + 0.5) X 32 gm. X 0.6998 It may be further calculated that in undergoing this reaction 1 gm. of fat (containing 278 X 3 + 874 = 0.9542 gm. of fatty acid) liberates 7,930 + 874 = 9.074 calories and accordingly 1 gm. of fatty acid furnishes 9.409 calories.* = 0.698 = 4.653 4 This figure includes also the heat involved in the hydrolysis of the glye- eride and in the conversion of glycerol to glucose. 150 Antiketogenesis. III TABLE II, Respiratory Quotients for Total Glucose and Fatty Acid. From the total CO, and O, subtract amounts corresponding to the metab- olism of the non-carbohydrate quota of protein during the respiration period. CO2 Oz Calories. gm. liters gm, liters Non-CH quota of pro- tein forl gm. of urine N. 4.067) 2307, 4.648 3.252 12.97 logs |= .60927) .31597 | .66727 | .51215 .11294 1 gm.=0.5089 liter |1 gm.=0.6998 liter log = .70662 log = .84497 The remainders represent the metabolism of fatty acid and glucose and their ratio (in liters) is the ‘‘fatty acid: glucose respiratory quotient.’’ Caloric value of 1 liter Molecular ratio of Fatty acid: spose Per cent of calories from the mixture glucose respiratory burned. quotient. ee ee eee Fatty acid, Calories. | Logarithm.| Glucose. Fatty acid. Glucose (1) (2) (3) (4) (5) (6) 0.698 4.653 66777 0 100.0 0.70 4.656 66801 a7 99.3 36.8 0.71 4.669 66922 4.0 96.0 6.2 OL 72 4.682 67043 7.3 92.7 3.2 0.73 4.695 67164 10.6 89.4 2ANT 0.74 4.707 67274 13.9 86.1 1.6 0.75 4.720 67394 17.2 82.8 1.24 0.76 4.733 67514 20.5 79.5 1.01 0.77 4.746 67633 23.8 76.2 .83 0.78 4.758 67742 21 72.9 0.70 0.79 4.771 67861 30.4 69.6 0.60 0 80 4.784 67979 33.6 66.4 0.52 0.81 4.797 68097 37.0 63.0 0.44 0.82 4.810 68215 40.4 59.6 0.39 0.83 4.823 68332 43.7 56.3 0.34 0.84 4.835 68440 47.0 53.0 0.28 0.85 4.848 68556 50.3 49.7 0.25 0.86 4.861 68673 53.6 46.4 0.22 0.87 4.874 68789 56.9 43.1 0.19 0.88 4.887 68904 60.2 39.8 0.17 0.89 4.899 69011 63.5 36.5 0.15 0.90 4.912 69126 66.8 33.2 0.13 0.91 4.925 69241 70.2 29.8 0.11 P. A. Shaffer 151 TABLE II—Concluded. Caloric value of 1 liter Molecular ratio of Fatty acid: ars: Per cent of calories from the mixture glucose respiratory burned. quotient. a fs Fatty acid Calories. | Logarithm. Glucose. | Fatty acid. Glucose () (2) (3) (4) (5) (6) 0.92 4.938 69355 73.5 26.5 0.093 0.93 4.950 69461 76.8 23.2 0.078 0.94 4.963 69574 80.1 19.9 0.065 0.95 4.976 69688 83.4 16.6 0.052 0.96 4.989 69801 86.7 13.3 0.039 0.97 5.001 69906 90.0 10.0 0.029 0.98 5.014 70018 93.3 6.7 0.019 0.99 5.027 70131 96.7 3.3 0.009 0.0 1.00 5.040 70245 100.0 The data above calculated for the theoretical respiratory quotients of fatty acid and of glucose and the corresponding ca- loric values of oxygen are given in Table II, together with inter- polated values between the extremes. Corresponding to each quotient, in Columns 4 and 5 are given the percentage of the total fat-glucose calories derived from glucose and from fatty acid. From the latter figures one obtains the relative molecular amounts of each, that is their molecular ratio, by multiplying each per- centage, by the corresponding fraction of a molecule of fatty acid and of glucose which is equivalent to 1 calorie, and dividing the product for fatty acid by the product for total glucose. In accordance with the assumption that each molecule of fatty acid gives rise to 1 molecule of acetoacetic acid, and that each mole- cule of glucose is equivalent to 1 of antiketogenic substance, the ratio as above obtained is the ketogenic-antiketogenic ratio (Column 6) except that it does not include the ketogenic value of protein. The latter, however, may be included very simply, since the ketogenic value corresponding to 1 calorie from protein proves to be almost exactly the same as the ketogenic value corresponding to 1 calorie from fatty acid; and the percentage of calories from (total) protein may therefore be added and calcu- lated with the fatty acid calories. The values in question are stated below. ( 52 Antiketogenesis. III we : Calories. | Ketozenic| miltimols glucose). Protein, 1 gm. urine nitrogen =.............. 26.5 10 20 Hach calorie from: proteim)—- 4.5 4se heer 0.377 | 0.755 Batty acid; 1 gm. =... ... feik cee 9.509 3.6 0) Hach calorie trom) tatty acid’ — ee eeeeeeas 0.378 | O Glocases, Lomi.j=) oe. eee eee Bis SRLS 3.76 0 5.56 Hach calorie tromeclucosei—s sane eee 0 1.478 Per cent of calories from Ketogenic value of 1 fatty acid + percent of ] X F. A. (or protein) calories from protein calorie = (0.377)/ u Ketogenic of 1 glucose calorie = (1.478) = ae PMG aT eRe TiS o > o ee =~ —_ Tro ee : Antiketogeniec value Antiketogenic Per cent of calories from x 3 total glucose It is convenient to treat the second factors in numerator and denomi- 7 0. nator separately as a single fraction, ——— 1478 0.255. As examples of the complete calculation of data of individual experiments the following may be given. (1) Urinary nitrogen excretion per minute = 12.7 mg. CO2z Oz RQ: ce. cc. ‘otal per minute.) ....::...3..cc bh eee eee 232 312 0.745 127 mig IN Xe O7cc COs a. noes. 26.3 IST iaers INP BAY GOs OF SR CRSA Petit. o antares Aes From fat and C H, including glucose from pro- CEI) 02. ALOR COR Ss ee RE), 205.7 | 270.7: Os760 0.271 liters O2 X 4.733 calories (at R.Q.0.76)..= 1.284 calories per minute from fat and C H, including glucose from protein. Non-C H quota of Bree 0.0127 gm. N per MANY te XK 1297 oan na se eee = 0.165 calories Total. 7.02100 Gias pee eeepc eee ne eee 1.449 calories per minute 0.0127 gm. X 26.5 = 0.336 calories from total protein. P.:A. Shaffer 153 33 i : v6 x 100 = 23.2 per cent of total calories from protein. 0.760 “F.A—G. R. Q.” corresponds to 79.5 per cent of F.A.-G. calories from fatty acid, and 20.5 per cent of F.A.-G. calories from total glucose. 79.5 + 23.2 20.5 (2) Urinary total nitrogen per minute = 14.2 mg. X 0.255 = 1.28 ketogenic ratio. CO2 Oz R.Q. OT SESE ys 246 324 0.765 Lo a Se aed yet. ol C0 re 29.4 O07 ail Re A717 Bi crc | @ sR 46 .2 From fat and C H, including glucose from pro- ee ee ia ede se Seley sss 216.6 | 277.8 | 0.780 278 liters O. X 4.758 calories (at R.Q. 0.78) = 1.323 0.0142 gm. N X 12.97 calories = 0.184 Total calories per minute = 13507 0.0142 X 26.5 = 0.376 calories from total protein. 0.376 == 1 Sei in. 1.507 x 100 9 per cent from total protein 0.780 F.A.-G.R.Q. indicates 72.9 per cent fatty acid calories and 27.1 per cent total glucose calories. 72.9 + 24.9 71 X 0.255 = 0.92 ketogenic ratio. The above calculation, which is rather tedious when applied to many data, is fortunately not necessary to obtain the information which it yields. It so happens that the two corrections involved; (a) the increase in the R. Q. from the glucose quota of protein and the consequent effect in lowering the ratio, and (6) the effect of the ketogenic influence of protein in raising the ratio, although not equal, are in opposite directions, and one may get almost the same result by ignoring the protein and calculating the total R. Q. as though only fatty acid and glucose were being burned. In this way almost all calculation is avoided. In Table III the ratios so obtained from the total R. Q. (in Column 2) are compared with the ratios corrected as above illustrated for both effects of protein, when the protein metabolism amounts to 10, 15, 20, and 25 per cent of the total energy exchange. From this comparison it is evident that, considering the technical difficulty 154 Antiketogenesis. III in exact and reliable determinations of respiratory quotients, the ketogenic ratios obtained directly from the total respiratory quotients are probably as nearly correct as can be expected from data of this character. We shall, therefore, in the following analy- sis of experimental data use only the total R. Q.; although it may be stated that we have also made the fully corrected calcula- tions on nearly all of the data presented. Perhaps it may be also explained that while the foregoing somewhat lengthy method of calculation of the ketogenic ratio from respiratory data is being abandoned as unnecessary, the short cut is permissible only because of the longer analysis of the underlying considerations. TABLE III. Ketogenic Ratios and Respiratory Quotients at Different Levels of Protein Metabolism. 10 per cent 15 per cent 20 per cent 25 per cent protein calories.|protein calories.|protein calories.|protein calories. Ketopene a ee Total R.Q. D = . : “Table. [p.a.-c.| ted tea g,] Batio pa c.| Ratio beac) Basie R.Q. K R.Q. oi R.Q. a R.Q rr A (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 0.72 3.2 0.726) 2.7 | 0.728) 2.65 | 0.731) 2.6 | 0.734] 2.4 0.73 Pye NF 0.736] 1.96 | 0.739} 1.9 | 0.742) 1.85 | 0.746] 1.78 0.74 1.6 0.746) 1.5 0.750) 1.44 | 0.753} 1.43 | 0.756} 1.42 0.75 1.24 0.757] 1.18 | 0.761) 1.18 | 0.765) 1.12 | 0.768) 1.14 0.76 1.01 0.767} 0.98 | 0.772) 0.94 | 0.777) 0.93 | 0.780) 0.92 0.77 0.83 0.778] 0.82 | 0.782} 0.80 | 0.788] 0.78 | 0.791) 0.80 0.78 0.70 0.788} 0.70 | 0.792) 0.70 | 0.799) 0.68 | 0.802} 0.69 0.79 0.60 0.80 | 0.59 | 0.804! 0.59 | 0.810) 0.57 | 0.813} 0.60 0.80 0.52 0.811} 0.5 0.816} 0.50 | 0.821) 0.5 0.827} 0.50 0.81 0.44 0.821} 0.43 | 0.826} 0.44 | 0.882) 0.43 | 0.8388] 0.44 According to the data in Table III a total respiratory quotient of 0.76 indicates the oxidation of a metabolic mixture made up of approximately equimolecular amounts of ketogenic substances (fatty acids or ketogenic amino-acids), and of antiketogenic derivatives of amino-acids, glycerol, or carbohydrate, expressed in terms of glucose. Expressed in the same terms, a respiratory quotient of 0.73 indicates approximately 2 molecules of ketogenic to 1 P. A. Shaffer 155 of antiketogenic substance in the mixture; while a quotient of 0.80 indicates only 0.5 molecule of ketogenic to 1 of antiketogenic glucose equivalent. With these values in mind we may inspect any respiratory data and provided we assume that the latter are truly representative of metabolic reactions, the ketogenic ratio of the subject may be determined. The question we have attempted to. answer is: What is the ratio at the time when definitely abnormal but small amounts of acetone bodies first appear in blood, urine, or breath? After an examination of many experiments by others as well as by ourselves there appears to be no doubt that the mixture in all subjects, at the threshold of ketonuria is, according to res- piratory data, that which corresponds to equimolecular mixtures of ketogenic and antiketogenic substances, the latter being calculated in terms of glucose equivalents. This is the same conclusion as that reached by a different method of analysis described in the earlier paper. In the application of the foregoing calculation to experimental results it is essential that the respiratory data be accepted as reliable and as indicating the actual oxidation in the tissues; and our experience in this laboratory has fully impressed us with the dangers in assuming that this is so. A large number of ob- servations with both Benedict unit spirometer apparatus and with large spirometers and gas analysis have been made on normal subjects by the writer and associates, and on hospital patients by my colleague, Dr. Olmsted, and while many of our results are consistent and probably correct, the R. Q. often varies in consecu- tive periods and for that reason is perhaps questionable; and it seems preferable to choose for this first analysis results from other laboratories, the accuracy and reliability of which are least open to question. Our own experiments will be reported in later papers. Fasting Normal Subject “L” of F. G. Benedict (9). The observations on this subject from the Carnegie Nutrition Laboratory are well known and require no comment. In Tabie IV are given the various total respiratory quotients of the sub- ject, before, during, and immediately after the long fast. On the 2nd fast day the subject excreted 0.5 gm. of hydroxybutyric acid, and on the following days the amount varied from 1.4 to 156 Antiketogenesis. IIT 7.0 gm. On the Ist and 2nd days after the resumption of food 0.8 and 0.5 gm. were excreted.’ The Ist and 2nd days of fast and the days after taking food may therefore be taken as the border-line of ketosis. It is evident from the table that acetone bodies first appeared when the R. Q. reached 0.75 or 0.76, and were disappearing when, after food, it reached 0.79 or 0.80. A re- spiratory quotient of 0.76 corresponds to a ketogenic ratio of about TABLE IV. Respiratory Quotients and Ketogenic Ratios of Fasting Subject ‘‘L.”’ F. G. Benedict (9). Total R. Q. Table 50, p. 345. Total Hydroxy- se akt:,2;,, |Ketogenic Bue Date. Day. Bed Ree ao for age? termined. calorim- appara- whole - eter, tas day, Table 38, night. morning. p. 396. p. 283. Apr. 10-11 | Food. 0.8L |. /0:8£ ):0.81 0.4 11-12 BS 0.88 0.89 | 0.885 0.16 << 12-13 43 0.86 0.89 | 0.875 0.18 “ 13-14 “ 0.81 0.82 | 0.815 0.4 25 Tals; Fast 1 OF 7Se) VOnon 0 S78 0.765 | 0.9 noe wal G ete) 0.75 (B76) |) MAD STACE 0.755 1.4 0.5 armel ene 0.738 | -0:75 | -0.74 02748) ik 725) 2S aS Te bat 0.74 0.75 0.745 | 0.73 22 3.5 e 19 STO 0.75 Deve Ih. Osis} Oni 3.0 2.1 eA) nae 0 0.68 0.74 | 0.71 Ong 3.0 3.5 ea ail Sah 7s 0.71 0:75 0.73 0.73 2.2 2.8 sie DD, Peas 0.73 0.74 | 0.7385 | 0.734 1.9 1.6 ee Do eran) 0.75 0.75 0.75 O57 250 | ean 3.5 St) DA lo} 0.72 0.76 0.74 OF726) |) a7 3.5 oe) 25 SS sO iz 0.75 0.735 0.733 1.8 1.4 May 18-14 se 30 0.72 OETeEy |) OR 3.0 5.4 “14-15 i io! 0.72 0:72.) 0:42 3.0 4.5 Food. “16-17 ik 0.80} 0.78} 0.79 0.6 0.8 “17-18 2 0.97 | 0.94] 0.96 0.03 0.5 1:1. As one would expect, the quotients fluctuate somewhat during each day, and assuming that they are “metabolic quo- tients’? the ketogenic ratio of the metabolic mixture doubtless also fluctuates; but it is clear that acetone bodies first appeared when the quotient first dropped to 0.76 or below. x P. A. Shaffer iGyé Normal Subjects on Non-Carbohydrate Diets (Higgins, Peabody, and Fite, 10). During 4 days each subject ate large amounts of protein and fat but no carbohydrate. On the Ist diet day small amounts of acetone bodies appeared in the urine. Respiratory metabolism was determined twice daily by spirometer and gas analysis. Table V contains a comparison of the total respiratory quotients, the corresponding ketogenic ratios, and the total acetone body ex- eretion expressed as hydroxybutyric acid per day, for two of the subjects. In each case the afternoon R. Q. of the Ist diet days (0.745 and 0.715) correspond to ketogenic ratios greater than 1, after which time the quotients indicate much higher ratios, in general agreement with the large excretion of hydroxybutyric acid. The conclusion appears justified that both subjects first excreted acetone bodies when the ketogenic ratio exceeded 1. Subject “Mrs. MCK” of Means (11), and Folin and Denis (12). A very obese woman was observed during three fasts, the res- piratory data being reported by Means and the urine analyses by Folin and Denis. The data essential to our discussion are brought together in Table VI. During the diet period preceding the first fast and during that fast the respiratory quotients are low (0.74 and from 0.71 to 0.68), indicating ketogenic ratios higher then 1 even before the fast, when presumably there was no ketonuria. Also during the next diet period the R. Q. remained low in spite of much food carbo- hydrate (148 to 239 gm. per day), being 0.74 to 0.76. The same, or lower, quotients were found in each of the next two diet periods and are difficult to explain. However, the quotients in both the second and third fast periods are wholly in accord with results from the preceding subjects. Acetone bodies first appeared when the quotient dropped below 0.76 indicating a ketogenic balance greater than 1. Diabetic Case No. 740 of Joslin (18). This case of severe diabetes with marked acidosis has been re- ported as an example showing “that if a diabetic is fasted, aci- dosis disappears and this is in marked contrast to the behavior TABLE V. Respiratory Quotients and Ketogenic Ratios. Normal subjects on high protein, high fat, non-carbohydrate diet (Higgins, Peabody, and Fitz, 10). Respiratory Date. period. Subject. H.L.H. |June 1/| 8 a.m. 4 p.m 2a. ans 4 p.m 8 San) 4 p.m SO 4 | Seams 4p.m Gay ey ciel 4 p.m SS Gol Seams 4 p.m Be Wirbe AST lle rer Basan 4 p.m Ce Sal Oy ar 4p.m “« 9} 8 a.m 4 p.m 10) || 8 a.am 4 p.m “AL | 8 "anm 4 p.m. 6 12\\-Stanuie 4 p.m Ketogenic ratio. Ss eee Total 7a || AEROS RQ/g [gs] Bane oe £5 24 hours. gC | BE om |a8 ie 1S) gm. 0.82 | 0.4) 04 0.86 | 0.2) 0.2 0.82 | 0.4] 0.3 0.745| 1.4) 1.3] 2.22 0.73 | 2.1] 1.4 ON eG les ont: 0.725] 2.6] 1.8 0.685) © 18.3 0 705)15.0 0 700)/40.0 25.6 0.680) © 3) be 0.765) 0.9 1.85f 0.82 | 0.4 0.785) 0.7 0.835} 0.3 0.715} 4.0 0.93 0.730} 2.1 0.700/40.0 2.6 0.740} 1.6 0.675] © Hos OE725/F2.0 0.670) © 1525 0.680) © ileal 0.690) © 1.0§ Diet. Mixed. Non-C H. << “ce + 100 ce. whiskey. Mixed, 468 gm. C H. “ “ce “cc “c + 180 ce. whiskey. Mixed, 386 gm. C H. * for 2.5 hrs. PLOT 21.5. brs: t for 2 hrs. § for 22 hrs. 158 P. A. Shaffer 159 of normal men, for they present acidosis upon fasting with no increase in the respiratory quotient as the fasting proceeds.”’ TABLE VI. Respiratory Quotients and Ketogenic Ratios. Subject, Mrs. MCK. from Means (11), and Folin and Denis (12). Date. ee naa pone: ideo eet (Means). A in Piper gm. Feb. 23 0.74 1.6 ee! 0.74 1.6 ti “« 95 0.71 6.0 0.34 “ Bee “o.26 ae eae “ 98 0.68 24.63 Mar. 1 0.74 m6 “« 9 0.76 1.0 “« 3 0.76 1.0 iA: 0.74 eG Food. eo 5 0.75 12 “« 6 0.75 1-2 cy 7 0.76 1.0 « 9 0.75 2 “« 10 0.76 1.0 0 Orit 0.75 12 0.03 | AST 0.75 11222 2.68 Second fast. « 13 0.72 3.0 8.60 “ 44 0.71 6.0 17.34 “ 15 0.73 a “« 16 0.73 met “ 47 0.73 2.1 ae « 19 0.78 0.7 « 20 0.79 0.6 0 “cc 1a ae a Bi he fast. “< 93 0.72 3.0 20.09 “94 0.71 6.0 See 25 ONS 2 Food. « 96 0.79 0.6 Since it is the writer’s contention that as regards the amount of acetone bodies formed the diabetic behaves essentially the same as any other subject who metabolizes the same mixture, it is 160 Antiketogenesis. III desirable that we attempt an analysis of this case. Table VII contains the total respiratory quotients, the corresponding keto- genic ratios, and the amount of total acetone bodies, expressed as hydroxybutyric acid, (including acetone and acetoacetic acid but not including breath acetone) which was actually ex- creted. Although as shown by a detailed calculation and as noted by Joslin the R. Q. is usually somewhat higher than is to be explained by the materials known to be available to the sub- TABLE VII. Severe Diabetic with Decreasing Acidosis on Fasting. Joslin, Case 740 (13). Tota’ Date. Total R. Q. Ketogenic ratio. |hydroxybutyric acid excrete gm. Apr. 15-16 0.72 3.2 24.9 Gai 0.73 2 18.9 comemaledesles 0.72 3.2 11.8 2 ely) 0.735 1.9 14.0 =) 19-20 0.755 1.0 7.9 “ 21-22 0.75 1.2 6.5 6 24-25 0.736 179 5.4 21-28 0.745 1.4 4.7 May 1 0.7 1.0 4.1 ject, if one assumes that the data represent metabolic quotients, it is evident that the acidosis decreased (but did not disappear) as the ketogenic ratio approached 1. According to the results from other subjects the continued excretion of about 5 gm. of hydroxybutyric acid -per day is hardly to be expected with a respiratory quotient of 0.76 and suggests that the quotients are actually slightly higher than the true metabolic quotients, or that the quotients during the respiration periods were higher than at other periods of the day. With an existing acidosis, respiratory quotients are apt to fluctuate and their interpretation is difficult and uncertain. Whatever the explanation of this discrepancy, it is evident from an analysis of Joslin’s data (not included in our table) that the gradual rise of the respiratory quotient, the lowering of the ketogenic ratio which that rise indicates, and the decline of the acidosis, were not caused by P. A. Shaffer 161 any increase in sugar burning power but were caused by the decrease in total metabolism resulting from the fast, and in the amount of ketogenic material (fat and protein) in the meta- bolic mixture. The metabolism of ketogenic substances was merely slowed down by undernutrition to the point where it no longer markedly overbalanced the already low rate at which the body was able to provide antiketogenic substance for “neu- tralization.’”? This appears to the writer to be the probable explanation of the beneficial effect of fasting and undernutrition in causing a decrease of diabetic acidosis. The point of view will be developed in a later paper. There are many other results from diabetics in the literature which are more or less suitable for similar examination and we have studied a number of them, especially in the publications of Benedict and Joslin from the Nutrition Laboratory and of Du Bois and associates from the Sage Laboratory at Bellevue Hospital. It is evident from these and earlier results as has been repeatedly noted, that the respiratory quotients are lower the more severe the grade of diabetes. The quotients determined by even the most expert investigators, however, often vary considerably from hour to hour and their interpretation is diffi- cult. Moreover, most of the cases on which sufficient data are given had more or less severe acidosis and under such circum- stances, factors resulting from acid production may so affect the respiratory quotient as to make it an unfavorable basis for calculating the ketogenic balance. A detailed consideration of other individual cases will accordingly not be attempted at present. Attention may be called, however, to the summary of Benedict and Joslin (14) which shows (p. 113) that the average quotients of all of their observations on severe diabetics, with marked acidosis, was 0.73 which corresponds to a ketogenic ratio of about 2, while the average quotient of the moderately severe and mild diabetics with little or no acidosis was from 0.73 to 0.77, the latter value corresponding to a ketogenic ratio of about 0.8. This general finding is, therefore, quite in harmony with the conclusion that a R. Q. of 0.76, corresponding to a ketogenic ratio of 1 represents the border-line or threshold of ketosis. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. I 162 Antiketogenesis. III SUMMARY AND CONCLUSIONS. A method is described by which the ratio of ketogenic to anti- ketogenic molecules in the metabolic mixture of a subject may be calculated from the respiratory quotient. Evidence is presented which appears to point to the conclusion that a molecular ratio of 1:1, which corresponds (according to the method of calculation) to a respiratory quotient of 0.76, is the limit for the avoidance of the excretion of acetone bodies. With quotients higher than 0.76 the catabolism of the antike- togenic glucose (or its equivalent from protein and glycerol) is great enough to remove acetoacetic acid as fast as it is formed, presumably by a ‘‘ketolytic” reaction analogous to the in vitro reaction previously described. BIBLIOGRAPHY. . Shaffer, P. A., J. Biol. Chem., 1921, xlvii, 433. . Shaffer, P. A., J. Biol. Chem., 1921, xlvii, 449. . Loewy, A., in Oppenheimer, C., Handbuch der Biochemie des Menschen und der Tiere, 1911, iv, i, p. 278. 4, Magnus-Levy, Arch. Physiol., 1904, 379; Z. klin. Med., 1905, lvi, 83. 5. Lusk, G., Arch. Int. Med., 1915, xv, 939. 6 i wd . Lusk, G., Science of nutrition, Philadelphia, 3rd edition, 1917, 61. . Bornstein, R., and Meyerhoffer, W., Landolt-Bérnstein physikalisch- chemische Tabellen, Berlin, 3rd edition, 1905, 419. 8. Williams, A. B., Riche, J. A., and Lusk, G., J. Biol, Chem., 1912, xii, 349. 9. Benedict, F. G., A study of prolonged fasting, Carnegie Inst. Washing- ton, Pub. 203, 1915. 10. Higgins, H. L., Peabody, F. W., Fitz, R., J. Med. Research, 1916, XXxIv, 263. 11. Means, J. H., J. Med. Research, 1915, xxxii, 121. 12. Folin, O., and Denis, W., J. Biol. Chem., 1915, xxi, 183. 13. Joslin, E. P., Am. J. Med. Sc., 1915, cl, 490. 14. Benedict, F. G., and Joslin, E. P., Metabolism in severe diabetes, Carnegie Inst. Washington, Pub. 176, 1912. OBSERVATIONS ON THE “ALKALINE TIDE” AFTER MEALS. I. By CYRUS H. FISKE. (From the Biochemical Laboratory, Harvard Medical School, Boston.) (Received for publication, September 2, 1921.) Since the original observations by Bence-Jones'? (confirmed some years later by Roberts*) showing a temporary decrease in the acidity of urine after meals, some investigators have re- corded their inability to find such variations consistently. Al- though most of the earlier reports on this phenomenon were con- cerned with acidity as determined by titration, the first! of Bence- Jones’ communications on the subject was confined to tests made with litmus, and the recent tendency*®* to associate the term “alkaline tide,” directly or by implication, with variations in the Cy of urine is accordingly a legitimate one, with the his- torical facts in its favor. To these two distinct uses of the term ‘‘alkaline tide’’ might be added a third, viz. the decrease in “‘acidity”’ or increase in ‘“‘alkalinity’’ shown by the double titration method recently used by Leathes’ and others, were it not for the fact that this method does not fulfill the purpose for which it is intended. The method is a determination of the ratio between the amount of alkali required to titrate a given sample of urine to the turning point of phenolphthalein and the amount required to reach the turning point of methyl orange.® The results obtained by means of it are supposed to rep- 1 Bence-Jones, H., Phil. Tr. Roy. Soc. London, 1845, exxxv, 335. 2 Bence-Jones, H., Phil. Tr. Roy. Soc. London, 1849, cxxxix, 235. 3 Roberts, W., Edinburgh Med. J., 1859-60, v, 817, 906. 4See Hopkins, F. G., and Hope, W. B., J. Physiol., 1898-99, xxiii, 271. 5 Hasselbalch, K. A., Biochem. Z., 1912, xlvi, 403; 1916, lxxiv, 18. 6 Denis, W., and Minot, A. S., J. Biol. Chem., 1918, xxxiv, 569. 7 Leathes, J. B., Brit. Med. J., 1919, ii, 165. 8 A similar procedure had been used earlier by others, but with quite a different object; viz., to determine what proportion of the total amount of weak acids in urine is neutralized (Auerbach, A., and Friedenthal, H., Arch. Physiol., 1903, 397; Henderson, L. J., and Palmer, W. W., J. Biol. Chem., 1913, xiv, 81). 163 164 Alkaline Tide resent the ratio of primary to secondary phosphate, and if this were true they could evidently be translated directly into terms of Cy. This claim, however, rests clearly upon the assumption that phosphate is the only weak acid present in urine, and that is well known to be erroneous; nor is the error introduced by this assumption a constant one, and of such a nature as not to invalidate comparative results, for the phosphate content of urine may undergo large and rapid variations without any significant change in the amount of other weak acids.1° From this it is clear that ratios thus obtained by the double titration method are not a reliable substitute for Cg determinations, and it is there- fore difficult to see any real significance in such ratios, in connection with the problems to which the method has been applied, beyond the informa- tion that could be more easily and much more accurately acquired by meas- uring the Cy directly. It is interesting to observe the extreme error into which others have fallen in consequence of the idea that urine behaves, in such titrations, as if it were a pure phosphate solution. Collip and Backus" have gone a step farther and, assuming that the sum of the two titrations gives ‘“‘apparently accurate’’ figures for the phosphate content of urine, have been led to con- clude that forced respiration causes a pronounced increase in the excretion of phosphate by the kidneys, although under these conditions the urine is often alkaline and may contain considerable bicarbonate. The figures published by Collip and Backus represent, not phosphate alone, but (ap- proximately) the sum of the phosphate and all other weak acids, including carbonic. Leathes!? himself (with Broadhurst) has also investi- gated the same question, and found no rise in phosphate excretion, presum- ably because he actually determined the phosphate. As far as I am aware, no one has denied that the Cy of the urine ordinarily diminishes for a time after meals, but the litera- ture on the subject indicates that an unequivocal drop in the titratable acidity under such circumstances is not so general an occurrence. Whether or not this apparent inconsistency is due simply to the comparative paucity of observations on the Cu of urine after meals, it is obvious that titratable acidity and Cy may vary independently, and even in different directions, when any considerable change occurs in the quantity of buffer substances present. Consequently, after a meal rich in phos- ® Michaelis has made the same assumption (Michaelis, L., Die Wasser- stoffiionenkonzentration, Berlin, 1914, 109). 10 Fiske, C. H., J. Biol. Chem., 1921, xlix, 171. 1 Collip, J. B., and Backus, P. L., Am. J. Physiol., 1920, li, 568. 12 Broadhurst, H. C., and Leathes, J. B., J. Physiol., 1920-21, liv, p. XXVIii. C. H. Fiske 165 phate, the Cx may conceivably fall without any parallel change in titratable acidity, provided always that the urine does not become actually alkaline to the indicator used in the titration. Although acidity titrations have been made in only a few of my experiments under circumstances where this state of affairs is likely to arise, something closely approaching it did occur in one instance (Table I). Here the titratable acidity at 9-10 p.m. (3 hours after the meal) is only a little less than in the period preceding the meal, but the Cy has fallen to about one-tenth of its previous value. : TABLE I. Experiment 20. Meal at 7.15 p.m. (for further details see Table II). Urine per hour. Time. Volume. pH Acidity. cc. cc.0.1N 5- 6 p.m. 60 5.25 11.0 Gare 34 5.15 10 sa (a one 21 5.10 12es 8-9 *“ 31 ‘ai0) | 1B} HW < 37 6.20 — 9.0 10-11 “ 32 Ues ey The observations reported in this paper (Table II) are taken from experiments (all on the same subject) made primarily for other purposes. They are of interest, in the first place, because most of them show the nature of the variations that may occur in the Cy of urine from hour to hour under the influence of food, whereas longer periods have ordinarily been used by others. Since the changes are often very sudden and transient, they may sometimes completely escape observation when the periods are several hours in duration. (The sharp fall in Cy in the 38rd hour after food in Experiment 37, for example, would have been entirely missed in 2 hour periods.) Some of these experiments are included to show that with this subject the urine is likely to become alkaline or nearly so at some time after a full meal consisting of meat, with or without other 166 Alkaline Tide food, but their main purpose is to point out certain difficulties in deciding whether a similar thing happens after a meal of small proportions. Concerning the experimental details, beyond what is stated under analytical methods and in the table, nothing need be said except that anything approaching vigorous exercise was always carefully avoided. Methods. The hydrogen ion concentration was determined colorimetrically by a modification of the dilution method somewhat different from that ordinarily used. In each instance, 1 per cent of al hour sample of urine (or its equivalent) was diluted to 10 ec. and compared with a fresh standard (made by adding carbonate-free standard alkali to acetic acid, monopotassium phosphate, or borate-KCl mixture,!4 and diluting to a concentration of 0.01 mw). The accuracy of the standards so prepared was frequently controlled by the electrometric method. The indicators used were methyl red, brom-cresol purple, phenol red, and cresol red. This form of the colorimetric method measures the Cu of the urine, not as secreted, but after dilution to a uniform basis. While this may in some ways be a disadvantage, the results are more useful for certain calculations that have been necessary in investigations to be reported later (e.g., the determination of the ratio of primary to secondary phosphate, and of the total amount of base combined with phosphoric acid), since the high dilu- tion of the solutions diminishes the possibility of complications due to incomplete ionization and the presence of electrolytes. The acidity was determined by titrating with standard alkali from a micro-burette to match a standard 0.01 m phosphate mixture (pH 7.4), prepared as described above. The indicator was phenol red, and no oxalate was used. The possibility of error from the effect of calcium was, however, eliminated in another way; v7z., by repeating each titration in the presence of twice as much water. As it happened, no difference between the two titrations could be detected in the case of any of the urines with which this paper is concerned, but sometimes such differ- 13 Henderson, L. J., and Palmer, W. W., J. Biol. Chem., 1912-13, xiii, 393. 44 Clark, W. M., and Lubs, H. A., J. Biol. Chem., 1916, xxv, 479. 16 Clark, W. M., The determination of hydrogen ions, Baltimore, 1920, 63. C. H. Fiske 167 ences are encountered, and all that is necessary then is to con- tinue the dilution until it has no further effect. This is satis- factory even with phosphate solutions containing more’ than enough calcium to combine with all the phosphate. Cy of Urine after Meals. The principal facts brought out by the results in Table II are these. The urine usually becomes quite suddenly less acid (and sometimes alkaline) in the 2nd or 3rd hour after a meal. When the meal is a full one (Experiments 14, 15, 20, 36, 38, 44, and 45) the Cy at that time is ordinarily much lower than it is likely to be otherwise. But when the meal is small (Ex- periments 1, 30, 32, and 37) it is usually impossible to decide whether the “tide” is due directly tc the meal under considera- tion or to a delayed effect of a previous meal, or whether it is in fact anything more than an apparent alkaline tide, representing in reality the recovery from a temporary increase in acidity (observed also by Hasselbalch®) immediately following the meal. The reason for this uncertainty may be seen in part from the records for the control experiments (Nos. 20 and 22), which were made following the same breakfast eaten (at about 8 a.m.) in all the others (puffed rice, milk, and coffee), but without further food before 7 p.m. Under these circumstances the pH under- goes minor variations for several hours and does not begin to show any inclination to approach 5.0 progressively until 4 or 5 p.m. The cause of these irregularities doubtless lies in the combined influences of the fixed alkali and potential acid in the food eaten for breakfast, and of the variations in Cy that occur in the morning (and later) when no food is eaten. 71%? It may be fairly concluded that the interpretation of similar observa- tions previously published by others is likewise uncertain for the same reasons, and that one cannot be sure of the meaning of any but large variations in the Cy (or titratable acidity) of urine after 16 Data on the urinary volume are included also (Table III) because of the relation that appears sometimes to exist between volume and acidity. It will be seen, however, that a decrease in Cx in these experiments isac- companied by a drop in volume quite as often as by diuresis. Alkaline Tide 168 “uooU 47% (daZjoo pus ‘1044nq ‘pReiq) younT , “ule % ~ZI ‘urd ZI-O1 ‘urd QI-8 ‘urd g ~9 ‘urd 9 +> “OUILT, ISIY OY} IOJ soinsy oyy, eg | 09°¢ | 099 G7 9 Pp a Ciiy it Ge | O€'S | 00'4 || 99°9 | aa°4 » I1-01 94°9 | 02°9 » O1-6 Gp'¢ | ¢o°S | 089 || aT a | Oss » 6-8 009 | oT g 3 Bk GT'9 | 069 GIG | STS | g9'¢ | 08'9 ee es) Gg} 009 94 | GFE 06% » 9-G g8'4 | 96°9 GG 0F 9 999 | 939 | 089 | Ge9 029 » &-> GG 0S'9 969 | G2°¢ | 039 | G69 G89 | OT 9, 1 ah ore Gh'G | OF 9 | 99°9 | 96°F | 089 | GG | OT 4 | Ob4 » £-B G&'¢ | ST'9 | 93°9 | g0°9 | O9'G | 989 | GEL | O44 Te ae | OF | 089 | 08°9 00°9 | Gh'¢ | $89 | a8'a wd T -ZI i 08'S | $h°9 | SF'9 G9 GL'¢ | OFS "Wl ZI-11 0¢'9 | G0"9 og’ | STS » IL-OL og'¢ | og'¢ GZ g "W'S OT-6 | | | “OUTIL, OOT 0g OOr 00T OOT 0g | OOT ouoN 00g 002 O0T O0T [egeoreeer: yo “ay 10d poysazur 10984 é G SPL 0€ LZ ST I Of CL | ST Cr I O& GE | 0€ GI Sai ad tg 1 iettaiea sy pe eee urd ‘[va yy oF bP «98 488 02 | rd Lg 3 08 I ST UR pape a "ON quourodxay ‘y[lu puw ‘99 fJoo “00 90% pus | ‘ard ‘1a3jnq pue pwaiq ‘so[qByjosdoA ‘yeayy | QUON ‘19}jNq “WIS OZ ‘pBaIq WS OZT ‘qBoul “UIs YCZ |°*"**** SBJYBIIG 0} UOTZIPps Ul [vay ‘ed A} poovj-pjoq Ul alv Avp oY} UI 109¥] S[VAU Joyyv SINOY 9 ‘our fo yd Tl ATAVL ‘(90Y09 “99 0OZ pu ‘YIU “dd OZ ‘e011 poynd ‘ws 07) s}uewTsedxe [[v Ul ‘We g yNOGe 7v YSBJYVOIg C. H. Fiske 169 meals except under circumstances that definitely exclude the influence of food taken earlier.!7 Whether or not it may be correct to say that the alkaline tide is due to the secretion of hydrochloric acid by the stomach,!$ it is certain that the acidity of urine after meals is influenced by TABLE III. Volume of Urine (Cc. per Hour). Experiment No........ 14] 15] 1 | 30 | 32 | 37 | 22 | 20 | 38 || 36 | 44 | 45 Le eee Time. Time. 9-10 a.m 87|207 1-11 =“ 75| 87 11-12 m 225)155 570 55| 54) 83 12-1 p.m 44) 47) 78/383 61| 49] 72 1=2 “ 71| 69) 66/325) 35/131) 79) 72 2-3 “ 66| 62/197/496; 31/159) 71/111 a 4° 56) 80)249/275| 30) 51) 60) 22 Ae 167|246|427| 30/209/116) 25 84) 74, 4-6 p.m. 5G" 58 35/186) 34| 60 G7 84/103) 34) 65 68/116} 6-8 p.m. aS “ 21| 55 S79 31| 37|| 40/114) 85) 8-10 p.m. 9-10 “ 37| 37 10-11 “ 32) 38!| 36) 22) 36) 10-12 p.m. W-i2 “ 34 73/102) 49} 12— 2 a.m. 17 The possibility of confusion from such “‘overlapping”’ effects was recognized many years ago by Roberts.? He also called attention to the necessity of examining the urine at very frequent intervals. 18 Dodds has recently found that the alveolar CO, tension, which nor- mally rises after meals, did not do so in a man whose stomach had been re- moved, and this is an important point in favor of the gastric secretion theory (Dodds, E. C., J. Physiol., 1920-21, liv, 342; see also Bennett, T. I., and Dodds, E. C., Brit. J. Exp. Patit., 1921, ii, 58). The contradictory evi- dence put forward by Hasselbalch is not at all convincing, being based upon the absence ef the alkaline tide after meals in subjects on a carbohydrate- freediet. Asa matter of fact, the tide did not disappear in his experiments until the 4th day on this diet, and its absence then can easily be accounted for by the acidosis. It is worth noting that Roberts* as long ago as 1860 undoubtedly produced acidosis in his subject by means of a purely ‘‘animal”’ diet, since he also observed that the alkaline tide diminished progressively in intensity in the course of 3 days on such food. 170 Alkaline Tide various factors operating at the same time. The maximum alkalinity after a protein meal is often reached at a time when the excretion of sulfate and phosphate (and therefore the rate of production of sulfuric acid, and perhaps of phosphoric acid as well) has reached or is approaching a maximum, and any de- crease in acidity occurring then must be in the face of this addi- tional acid production. It is safe enough to suppose that the intensity of the tide (or its appearance at all) will depend not only upon the amount of hydrochloric acid secreted by the stomach (if that is one of the factors involved), but also in an independent way upon the composition of the food. Consequently, even if such a tide is not always evident after meals, that fact alone does not detract from the importance of the phenomenon when it does occur. INORGANIC PHOSPHATE AND ACID EXCRETION IN THE POSTABSORPTIVE PERIOD. By CYRUS H. FISKE. (From the Biochemical Laboratory, Harvard Medical School, Boston.) (Received for publication, September 2, 1921.) The introduction, in the middle of the 19th century, of Liebig’s titration method for the determination of phosphate in urine stimulated the first investigations on the subject of diurnal varia- tions in phosphate excretion. As part of a comprehensive program instituted in connection with the Verein fiir gemeinschaftliche Arbeiten zur Férderung der wissenschaftlichen Heilkunde, two students (Winter! and Mosler?), working under the direc- tion of Vogel, observed that the rate of phosphate excretion in men on ordi- nary diets reaches its lowest point in the morning. Their experimental days were divided into 3 or 4 periods of varying length, and the result of this was that the time of occurrence of the minimum was not well defined. Shortly afterwards these observations were repeated by Beneke* with uniform 3 hour periods (except at night) and he found the lowest phosphate output (averaging 24 mg. of phosphorus per hour) to occur between 7 and 10a.m., with 39, 54, 52, and 45 mg. per hour in the succeeding 3 hour periods, and 39 mg. per hour aaa the night. Some years later, von Haxthausen® arrived at a similar conclusion, the uranium titration method having meanwhile been devised, and many others since have called attention to the low rate of phosphate excretion in the morning after the first meal of the day (Zuelzer,? Edlefsen,*° Speck,}° 1 Winter, A., Dissertation, Giessen; cited by Beneke, F. W., Arch. wis- sensch. Heilkunde, 1854, i, 667. 2 Mosler, F., Dissertation, Giessen; cited by Beneke, F. W., Arch. wis- sensch. Heilkunde, 1854, i, 670. 3 Beneke, F. W., Arch. wissensch. Heilkunde, 1854, i, 391, 571. 4de Jager,® in reporting Beneke’s figures, has retained an error in the - original in the average for the second period. 5 de Jager, L., Z. physiol. Chem., 1898, xxiv, 303. § yon Ficathageeae Boks ean dissertation, Halle, 1860; cited: by Edlefsen.® 7 Zuelzer, W., Virchows Arch. path. Anat., 1876, lxvi, 223, 282. 8 Edlefsen, G., Cent. med. Wissensch., 1878, xvi, 513. 9 Edlefsen, G., Deutsch. Arch. klin. Med., 1881, xxix, 409. 10 Speck, Arch. exp. Path. u. Pharmakol., 1881-82, xv, 81. 171 172 Phosphate and Acid Excretion Ott,1! Roeske,” de Jager,®> Sherman and Hawk,!3 Hawk,'4 Hawk and Cham- berlain,'® Cathcart, Kennaway, and Leathes,'* and some others with less definite results). The observation has been extended also to dogs receiving but one meal in the course of 24 hours (Feder,17 Vogt,!8 Loeb,!9 Wolf and Osterberg?°). The earliest investigators of this subject were engaged in breaking ground in a field that had not previously been touched. They could hardly, at first, have been in a position to plan their experiments in such a way as to avoid misinterpretations due to the influence of food. Their observa- tions on the curve of phosphate excretion accordingly occasioned no sur- prise, for their results with sulfate were essentially the same, and that was just what they expected on the basis of the erroneous conception then prevailing concerning the distribution of phosphorus in proteins.2! They were naturally, therefore, inclined to ascribe their findings chiefly to the ‘gradual fall in phosphate excretion during the night (inasmuch as that was the longest period without food), and slow absorption was to them a suffi- cient explanation for the failure of the phosphate curve to rise immediately after the first meal. No one since, in fact, has seriously questioned this way of accounting for the facts, with the exception of Loeb!®, and his specu- lations on the subject were not supported by experimental evidence of any sort. My attention was directed some time ago to the matter of variations in inorganic phosphate excretion during the Ist day 1 Ott, A., Z. physiol. Chem., 1886, x, 1. 12 Roeske, G., Dissertation, Greifswald, 1897; cited by Sherman and Hawk.}3 13 Sherman, H. C., and Hawk, P. B., Am. J. Physiol., 1900-01, iv, 25. 14 Hawk, P. B., Am. J. Physiol., 1903-04, x, 115. 4 Hawk, P. B., and Chamberlain, J.S8., Am. J. Physiol., 1903-04, x, 269. 16 Cathcart, E. P., Kennaway, E. L., and Leathes, J. B., Quart. J. Med., 1907-08, i, 416. 17 Feder, L., Z. Biol., 1881, xvii, 531. 18 Vogt, H., Beitr. chem. Physiol. u. Path., 1906, viii, 409. 19 Loeb, A., Z. Biol., 1910-11, lv, 167. 20 Wolf, C. G. L., and Osterberg, E., Biochem. Z., 1912, xli, 111. 21 This, together with a general lack of knowledge about the nature of the phosphorus compounds in foods (particularly meat), undoubtedly in- fluenced the interpretation of many later investigations in the field of phos- phate metabolism, including the researches on the nitrogen-phosphorus ratio initiated by Bischoff (Bischoff, E., Z. Biol., 1867, iii, 309). The term ‘‘phosphoprotein”’ (or its equivalent, ‘‘nucleoalbumin’’) is even now too often loosely used, although Plimmer and Scott have failed to find evi- dence for the existence of any such substance in animal tissues, with the possible exception of the pancreas (Plimmer, R.H.A., and Scott, F.H., J.Chem. Soc., 1908, xciii, 1699). ah C. H. Fiske 173 of fasting by observations on 1 hour urines collected during the morning. These revealed a surprising and apparently unaccount- TABLE I. Fasting since 7 p.m. of Preceding Day. Experiment 57. 100 cc. of water per hour. Urine (per hour). Time. Volume. Inorganic P. ce. mg. 7-8 a.m. 36 24.7 8-9 “ 25 22.0 9-10 <“ 22 ARG} 1O=10, *¢ 75 12.6 11-12 m. 136 eee 12—- 1 p.m. 85 15.4 jie os 40 18.4 DAES 95 25.0 Sa Awe ist 63 25.8 4-5 “ 178 26.6 aay ae 114 24.7 Gs 7“ 38 Zak ema se - 34 220 au 26 22.3 9-10 “ 31 (22.5 Urine (per hour). pane ; 3 2% | : - Ss : = Saas ee Se eo eo ae aes ce mg. | mo. | mo. | ox] ots | ots | om] ons 8-10 a.m PANE OV .3| 320 16.4) SUG 122A TOE? | ee la2oRs 10-12 m. 106 | 6.00) 13.1} 31.3 14.4) 42 \ 22:3) 950 ON e262 12-2 p.m 63 | 5.80) 16.9) 27.2] 11.9] 5.4 | 19.4) 7.4 | 7.0 | 26.4 2= 4. <8 79 | -5.75| 25.4) 26:2 12.0) 82S TFS PStOoneae6 4-6 “ 146 | 5.75| 25.7) 22.4) 14.7] 8.3 | 16-0) 9.2 110.0 | 26.0 G15! 36 | 5.50 23.1) 2199) 13.3) 7-4: 15.6) 8.3 | 9.9 | 25.5 Sands Fs 2 5.55! QA 2020) 1378) (22 | 14.3] 8.6 | 9.4 | 23.7 able increase in the phosphate output, reaching its maximum at about the middle of the afternoon (see Experiment 57, Table I). Further inquiry immediately showed that this rise begins 174 Phosphate and Acid Excretion soon after the drop long known to occur during the night has reached its lowest point, which may be as low as 6.3 mg. of phos- phorus per hour under these conditions.” A curve of this same general form has now been found, without exception, in about 40 such experiments on four subjects, and some of the observa- tions on three of these will be published later in another connec- tion. ‘The present paper will be confined to experiments selected from some 25 performed on the fourth subject, involving deter- minations, not only of phosphate, but also of several other factors concerned in the question of acid excretion (which likewise is subject to variations of considerable magnitude during the post- absorptive period), the object being to attempt to learn the reasons for these changes as a whole. No vigorous exercise was taken during the course of any of these experiments. It will readily be seen that the situation just described,” occurring when the rising part of the phosphate curve cannot be accounted for by food recently eaten, bears no necessary relation to the similar course of events that many have observed (at various times in the last 70 years) following breakfast, inasmuch as one could not have been predicted from the other. Neverthe- less, it is not at all impossible that the explanation of both, inso- far as they are qualitatively alike, will ultimately prove to be the same. One other matter respecting the phosphate curves themselves 22 A preliminary report on this point, made at the meeting of the Amer- ican Society of Biological Chemists in December, 1919, was published in March, 1920 (Fiske, C. H., J. Biol. Chem., 1920, xli, p. ix). Similar obser- vations have since been recorded by Broadhurst and Leathes in a paper presented before the Physiological Society in the following July, and pub- lished in December (Broadhurst, H. C., and Leathes, J. B., J. Physiol., 1920-21, liv, p. xxviii). 23 Feder!’ published two fasting experiments (on dogs) intended toserve as controls for his meat feeding experiments. Provided that certain assump- tions are made about the time of day represented by his periods, the results of these experiments give a curve not unlike those reported in this paper. Feder, however, dismissed them as accidental variations. Edlefsen® like- wise found in one experiment a rise in phosphate excretion between 12 m. and 6 p.m. of the 2nd day of fasting in man, with lower figures for the 6 hour periods preceding and following. This is undoubtedly the same phe- nomenon that I have described, but the 1st fast day of Edlefsen’s experi- ment did not show it, nor would mine if 6 hour periods had been used. } C. H. Fiske ote requires to be mentioned. In all those experiments in which the level of inorganic phosphorus excretion, in its rise during the late morning and early afternoon, eventually exceeded 25 mg. per hour, there followed (whenever the experiment was con- tinued for a few hours) a further slight fall to between 20 and 25 mg. In the single experiment that was continued later than 6 p.m., the curve finally became horizontal for some hours (Table I), and essentially the same thing happened in an experiment on another subject (not now included), which lasted until midnight. These oscillations in the curve of phosphate excretion (which an inspection of the tables will show to be entirely unconnected with the volume of urine), naturally suggested some relation to the alkaline tide observed by Hasselbalch™* during the morning in subjects who had been without food since the previous noon, and an investigation was accordingly begun, involving simul- taneous observations on both these things. The subject of these experiments ate, throughout, an ordinary mixed diet di- vided into the customary three meals a day, with no attempt to maintain constancy of composition, since this was readily shown to be unnecessary for the purpose. It was shortly found that, in this subject under such circumstances, no very marked alkaline tide occurred during the morning, whether the last meal had been taken on the evening before or at about the middle of the preceding day (Tables I and VI). In order to intensify this tide, with the hope of helping to unravel its relation tothe phos- phate curve, the device was adopted of drinking varying quanti- ties of milk at about midnight of the day before. The temporary drop in the Cy of the urine during the morning was then ordi- narily much more pronounced, and the more so the larger the amount of milk taken (Tables II and III). The effect was similar when the food eaten at that time included other things besides milk (Table IV). The variations observed in the excretion of inorganic sulfate can furnish no basis for explaining the alkaline tide by alterations in the rate of sulfuric acid production, since the tendency through- out is for the inorganic sulfate content of the urine to decrease gradually to an approximately constant level, with no significant rise in the latter part of the day, and since even the decrease is 24 Hasselbalch, K. A., Biochem. Z., 1912, xlvi, 403; 1916, Ixxiv, 18. 176 . Phosphate and Acid Excretion TABLE I. 200 Cc. of Milk at Midnight. Urine (per hour). Time. Volume. | pH Inorganic P. Experiment 31. Fasting. 200 cc. of water every 30 minutes. cc. mg. 8- 9 a.m. 40 5.00 278) 9-10 “ 215 5.85 12:2 OSL 474 6.80 6.3 fi—t2"m: 397 6.60 7.6 ,12= 1pm. 421 6.40 12.5 (OME 358 6.05 1h nee 358 6.05 1sat | Saas 330 5.85 25.7 2 300 5.50 27.8 : Experiment 35. Fasting. 100 cc. of water per hour. 6- 7 a.m. ae 5.05 oe (etch ate 29 5.10 24.4 8-9 “ 95 15 As) 18.9 9-10 “ 67 6.05 13.6 10-11 “ 130 7 AW) 14.6 11-12 m. 166 WAS 16.7 Experiment 25. Fasting. 100 cc. of water per hour. Urine (per hour). ° Time. ‘ 2 Inor- Phos- = ve, | pt | apie fwrte | JRE, | morc] AND) tan ce. mg. mg. mg. cc.0.1 M|cc.0.1 N|cc.0.1N 8- 9 a.m. 34 15.20 | 24.7 | 29.2 | 24.4 || 8.0 | 20.8] 152 9-10. * 360) |sormlo: falda ee:. 7 |) 21-9 4.3 | 1768131396 10-11 “ OY AlrossO niles P2374) 2156 3.6 | 16.9 | 13:5 IAPs scat, 65 Hat) |p Mesaey |) PALE ieikeye 4.0 | 15.4 | 11.3 IS I jo) AG Pe 20 be?) 2420) | 1625 4.9 |.17.2 | 10.3 eee Gy 3% |) SelOn) WO 1 |, 2258.) 16.9 6.2 | 16.3°| 10.6 Zen 52a ORL OM eeosoal zane le lOne 7.7 | 18.4 | 10.2 3-4 “ 57 SMOMEZS ou 2ecotiedy so 92+) TS PLOss 4-55 yes 44 5p OO Ne28e razon helG29 9.2 | 14.8 | 10.6 5-6 “ 25 4.90 | 28.5 | 19.7 | 16.1 9.2 | V4 ed Oe: GC. i. Fiske 177 TABLE II—Concluded. Urine (per hour). A, cat eae : Time. 2 22 |e £ o 2 5 o | 2 Nee a 2 > | ae g a i Se ad = ais! BA = 5 Rf eA eee w(t = meses ese df iG ee g a a | = Ue > 2, = Z 4 ey < nD < < Experiment 50. Fasting. 200 cc. of water per hour. eH PED NTO Wa Mee | O1m O1N 0.1N Bday Stay 7- 8 a.m. 37 | 5.40) 27.8) 32.1) 28.1)-8.9 | 22.9) 17.6)11.8 | 34.7 B= B“ 35 | 5.45) 20.3) 29.5) 26.1) 6.5 | 21.1] 16.3) 9.1 | 30.2 a-ip “ 75 | 5.45) 15.8} 29.3) 23.4! 5.1 | 20.9) 14.6] 8.7 | 29.6 10-11...“ 204 | 5.75] 12.5} 28.0) 21.3)) 4.0 | 20.0] 13.3) 7.7 | 27.7 11-12 m. 115 | 5.75} 14.2) 20.0) 19.5], 4.6 | 14.3) 12.2) 7.1 | 21.4 12— 1 p.m. 282 | 5.95} 16.5} 19.6} 18.8) 5.3 | 14.0) 11.7] 7.5 | 21.5 i -taa 193 | 5.85) 17.2) 22.3) 18.1] 5.5 | 15.9] 11.3] 7.9 | 23.8 Fi ila 152 | 5.65) 23.0) 23.1) 17.6) 7.4 | 16.5) 11.0] 9.9 | 26.4 oer 153 | 5.55) 28.2] 24.6) 19.2) 9.1 | 17.6) 12.0]11.8 | 29.4 accompanied by a parallel change at least as great in the am- monia excretion. A certain correspondence is evident be- tween the phosphate and the Cy, although the latter sometimes continues to fall after the phosphate has begun to rise again. The parallelism between phosphate and titratable acidity is much closer and fails at only one point, vz. the drop in acidity at the trough of the wave is likely to be more pronounced than the associated change in phosphate, and this corresponds with the occurrence of a fall in Cg, which is most marked at about that point. To judge from present tendencies in discussions of neutrality regulation in the animal body, many would be inclined to believe 25 Organic acid determinations by the method of Van Slyke and Palmer (Van Slyke, D.D., and Palmer, W.W., J. Biol. Chem., 1920, xli, 567) in several of these experiments yielded results for the various urines of any one experiment that were hardly distinguishable from one another. They were of the same order of magnitude, when reduced to the same basis, as those reported by the originators of the method for 24 hour urines, and since they cannot be regarded as anything but approximations the figures are omitted. They have served, however, to eliminate organic acids as a factor in the present problem. 178 Phosphate and Acid Excretion that the phosphate output in these experiments falls because there is less acid requiring to be neutralized, and rises again later because of increased acid production; in other words, that the amount of phosphate excreted is determined by the amount of other acids calling for removal from the body. For such an TABLE III. 500 Cc. of Milk at Midnight. Experiment 21. Fasting. 50 cc. of water per hour. Urine (per hour). Time. : Volume. pH Inorganic P. ce. mg. 6— 7 a.m. 24 5.50 35.9 [8S 38 5.35 35.0 8-9 “ 55 5.40 29.9 9-10 “ 45 5.35 24.6 10-11 “ ov 6.50 17.2 11-12 m: 170 6.85 19.4 12-1 p.m. 73 7.05 32-7 Experiment 65. Fasting. 200 cc. of water every 2 hours. Urine (per hour). s. Time. E 3 ae z aye 2 ne S| Bce eee | 2? ae ee ec. mg. | mo. | mo. || ote |o1x|o1n|o1n| 01% 6- 8 a.m. 65 | 5.50] 40.0] 23.5] 23.1|| 12.9] 16.8] 14.4] 11.8] 28.6 8-10 “ 26 | 5.55| 27.6| 20.0| 20.2|| 8.9] 14.3] 12.6] 10.0] 24.3 10-12 m. 55 | 7.25| 18.3| 8.7| 18.4|| 5.9| 6.2| 11.5| 1.4] 7.6 12-2p.m. | 141 | 6.95] 21.0] 6.6] 15.2|| 6.8| 4.7| 9.5] 2.8] 7.5 Pail a 78 | 6.50) 33.1] 8.4] 14.9]/ 10.7] 6.0} 9.3) 6.6) 12.6 interpretation there is, as far as I know, no evidence whatever, and it is easily shown (Table V) that the general form of the curve of phosphate excretion in the postabsorptive period is not at all altered by the administration of sodium bicarbonate, whether this is given some hours before the beginning of the experiment (Experiment 40) or at a time when the curve has begun to rise (Experiment 39). C. H. Fiske 179 TABLE IV. Meal at Midnight. Experiment 9. Fasting. At12m. (preceding), 1 egg, bread and butter, apple pie, and 500 cc. of milk. 100 cc. of water per hour. Urine (per hour). Time. 3 g2 |e | Z s : 2 |% | gs ll¢e| 8 |ie| 2 | 22 5 = 2 ad 2s 8 a3 3 a} S se 9 cs Sailgsa/ 8 |sea] 3s = - ry Re a a ee 7) < cc CL ML i lk nig of N O1 N O1N Of N 7-8 a.m. 36 | 5.30) 47.6) 35.9) 35.7)| 15.3] 25.6] 22.3] 16.0) 41.6 8-9 “* SORA tok. o) oh.5| 22 Olle eee ale l ies place ae 9-10 “ 390852201037 8) 31.5] 276i) 1222 2225l 17-3) 12 2he4 eo 10-11 “ 74°) 6220) 21.2) 17 .2| 26-7)) 6.8) 12:3) 1627) 6.0) 418-3 11-12 m. 257 | 6.60} 20.6} 16.0] 23.0]| 6.7) 11.4) 14.4) 4.8] 16.2 12-1 p.m. 186, 6.75] 19-9) 11.1) 19.21); 6:4) 7291220) S24 ties aaa < LOO kG 475|02378)) 11.3] 19.7) Toa) Salis Sse eG J= 3,“ 239 | 6.60) 29.8) 10.5) 15-8}|} 9.6) 7.5) 9.9) 5.4! 12.9 3-4 “ E70 WGeSso} ol 4) 10.1) 14.41 1021) 7-2) 9.0)5 6-0) 2 Au 130 | 6.10} 30.3) 14.4] 16.6]} 9.8) 10.3] 10.4) 8.1] 18.4 5G 52 | 5.30) 25.0} 20.2) 14.7]| 8.1) 14.4) 9.2) 9.0) 23.4 TABLE V. Effect of NaHCO; on Phosphate Excretion. Urine (per hour). Time. Volume. | pH |Inorganic ize Experiment 39. Fasting. 200 ce. of milk at midnight. 100 cc. of water per hour. cc. mg 11-12 m. 49 7.25 7.2 12-1p.m.| 117 6.95 14.8 1—- 2 p.m. 70 7.75 18.7 | 1p.m. 10 gm. of NaHCO; per os. Zoe 46 8.35 25.3 Experiment 40. Fasting. 200 cc. of milk at midnight. 5 gm. of NaHCO; in 200 ce. of water at4.a.m. 100 cc. of water per hour. 9-10 a.m. 72 7.65 15.2 1p-11, * 59 7.65 aD 7 fam. | 94 7.60 13.9 12—- 1 p.m. 87 7.50 21.9 180 Phosphate and Acid Excretion The observations as a whole, insofar as they can be accounted for at all by those factors that have been determined in these experiments, can be interpreted only upon the basis of a decrease (followed by a rise) in the rate of production of phosphoric acid during the morning, or, what is perhaps more probable, although its effect on the composition of the urine might well be the same, by an active retention of phosphate (phosphoric acid or primary phosphate), which later in the day is “released.’’ TABLE VI. Phosphate Excretion between Midnight and Noon. Experiment 44. Fasting since 2 p.m. 100 cc. of water every 2 hours. Urine (per hour). Time. we pai ee SE es et ee Volume. pH Inorganic P. ce. mg. 12— 2 a.m. 102 5.60 33.0 -4 123 5.65 31.0 4-6 “ 33 5.45 28.4 6-8 “ 32 5.60 18.7 8-10 “ 71 5.80 1 fan 10-12 m. 52 5.50 15.6 One experiment (Table VI) must be referred to again because of its bearing on the possible validity of this interpretation. In this experiment, which continued from midnight until the next noon, no food having been taken since 2 p.m., if will be seen that the phosphate output underwent only a gradual drop between midnight and 6 a.m., but then fell suddenly in the next 2 hours. If the falling curve as a whole were to be accounted for solely on the basis of the gradual excretion of phosphate derived from the last meal, it would be difficult to explain this marked irregularity. The explanation (phosphate retention) that I have offered as the one most probable from the facts available is not estab- lished by the experiments now reported, nor do I believe that it will by ztself completely account for the situation. The problem is now being investigated in several directions with the hope of throwing more light upon it. C: H. Fiske 181 EXPERIMENTAL. Most of the figures given for inorganic phosphate were ob- tained with the colorimetric method of Bell and Doisy,?* which has proved to be quite accurate enough for the purpose at hand.?? The analyses in Experiment 25 and those preceding it in the _series were made by the titration method recently described.”’ and this has been used also in many isolated instances as a check on the colorimetric method. The results have served to show?’ that the latter is entirely satisfactory, when an accuracy of about 2 per cent is sufficient, under all the conditions met in these experiments, whether the phosphate excretion is high or low. Inorganic sulfate was determined by precipitation with ben- zidine after removing the phosphate;?2 ammonia by aeration (into 5 ce. of 0.02 n HCl diluted with water) followed by titration (methyl red); Cy and acidity in the manner described in the pre- ceding paper.?® The urines, immediately after collecting, were diluted to some convenient volume (most often 100 cc.) and the analyses begun immediately, except with samples obtained late in the day, which were preserved over night with chloroform in the cold room. The sulfate determinations were sometimes all post- poned until the following day, but the phosphate was then re- moved (with magnesium carbonate, etc.) while the urines were fresh; the alkaline filtrates resulting keep for some time. 26 Bell, R. D., and Doisy, E. A., J. Biol. Chem., 1920, xliv, 55. 27 Fiske, C. H., J. Biol. Chem., 1921, xlvi, 285. 28 Fiske, C.H., J. Biol. Chem., 1921, xlvii, 59. 29 Fiske, C. H., J. Biol. Chem., 1921, xlix, 163 artl > es) at Sn a ae i 3 see to mee % vie? 7 aie Oo a Aint ae = a: S a beet ie Rea jase tyke Res HF ‘ t ay ek =e Pome sha) sf awe ’ es ny Seer his ‘ P aiheis Ne aa. inh Tp Ry Y hie 4 ‘ J ; Cie ihe A BUFFER SOLUTION FOR COLORIMETRIC COMPARISON. By T. C. McILVAINE. (From the Department of Soils, West Virginia University, Morgantown.) (Received for publication, September 23, 1921.) In the use of standard buffer solutions for colorimetric compari- son where more than a restricted range of reaction is required, it has been necessary in the past to make use of several solutions. The author has developed a system requiring but two stock solu- tions and covering a range of from pH 2.2 to pH 8.0 which ap- proximately includes the limits of reaction for arable soils and physiological media. “The materials used are as follows: 0.2 Mm disodium phosphate! and 0.1 m citric acid, combined in such volumes as to make 20 ce. of the mixture. The disodium phosphate employed was recrystallized three times. A 0.2m solution was prepared by titration against HCl, using methyl orange as indicator. Although the titration end- point was not distinct, the error in the resulting mixtures was usually not appreciable, seldom exceeding 0.01 pH. On account of the variable water of crystallization content of the phosphate salt? the stock solution was standardized by titration in order to have a reproducible method. Upon exposure to the air for a period of 2 weeks the water of crystallization of disodium phos- phate is reduced to 2 molecules. If, after exposure, a quantity of the salt is kept on hand in a closed container and the correct weight of the salt required to produce the proper concentration of stock solution determined by titration, it is possible at any time to make up the stock solution by simply weighing out the salt. The citric acid was recrystallized at least twice before using. 1 Molds may develop in phosphate solutions under suitable conditions, but this source of trouble has been obviated by Martin* by shaking the solutions with a little calomel for a few minutes and then filtering. 2 Martin, C. J., Biochem. J., 1920, xiv, 98. 183 184 A Buffer Solution The 0.1 m stock solution was standardized by titration against NaOH solution which had been prepared with boiled water and protected from CO:. Barium hydroxide was also used as recom- TABLE I. pH required. 0.2m NasHPOs. 0.1 m citric acid. ce. cc. 222 0.40 19.60 2.4 1.24 ; 18.76 2.6 2.18 17.82 2.8 3.17 16.83 3.0 4.11 15.89 See 4,94 15.06 3.4 5.70 14.30 3.6 6.44 13.56 3.8 Cel) 12.90 4.0 Tech 12.29 4.2 8.28 11.72 4.4 8.82 11.18 4.6 9.35 10.65 4.8 9.86 10.14 5.0 10.30 9.70 5.2 10.72 9.28 5.4 11.15 8.85 5.6 11.60 8.40 5.8 12.09 Ceti 6.0 12.63 eon 6.2 13.22 6.78 6.4 13.85 6.15 6.6 14.55 5.45 6.8 15.45 4.55 UD) 16.47 3.53 Uo2 17.39° 2.61 7.4 18.17 1.838 7.6 18.73 1.27 7.8 19.15 0.85 8.0 19.45 0.55 mended by Sorensen and given by Clark, but sodium hydroxide was found to be as accurate and more convenient. The correct weight of citric acid required to make the stock solution can also be determined by titration. 3 Clark, W. M., The determination of hydrogen ions, Baltimore, 1920. "TE pay Suajqewique) 2qND +—"0dH*enN W20 ploy 24319 WIO —~> € 2 1 fe) 02 61 8) 4 gl GI vl €l 2\ W Ol ezeydsoyud sussuasu0G ~~~" ~~~ = 8 HOPN-2}e4U}ID SUasUdIOG JDH-2}euz!D SUssSUua0S apRyeoR Sajodje/A HORN-27eydsoyd sqnq pur yur> HOCN -ate)euZYd Sqn q pue yuelD IOH -ayereyzud sqny pue yued BAUND PIDe D14zI9-ayeyUdSoUd 6 + il 9) Wd ur uoiDeay 185 186 A Buffer Solution The pH values of the mixtures were determined electrometri- cally by use of the chain; Hg | HegCl | NKCL | saturated solution of KCl| H,|Pt. No allowance was made for liquid potential. No attempt was made to maintain a constant temperature. How- ever, the temperature of both calomel electrode and buffer solu- tion were taken into account in computing the pH values. Clark’s® extension of Sorensen’s values for the normal calomel electrode was used with the necessary interpolations. Three extra calomel electrodes were employed for checking the accuracy of the one in general use. A Leeds and Northrup type K potentiometer and type R sensitive galvanometer were used for making the elec- trometric measurements. The electrode was of the platinum wire variety. The hydrogen was generated electrolytically and passed first through an acid permanganate solution, next through a hot tube, and finally through‘a wash bottle containing distilled water. A graph was constructed in which the pH values (determined electrometrically) of various mixtures of phosphate and citric acid solution, so arranged that the total volume was in all cases 20 cc., were plotted against the volumes of the two solutions. By interpolation, using the curve so obtained, it was possible to arrive at the proper volumes of the two solutions which when mixed would give 20 ce. of a solution having any desired reaction. The values given in Table I were obtained in this way and checked by actually preparing the solutions and measuring the pH values by the electrometric method. In all cases the variation of the observed from the calculated pH was 0.01 or less. Fig. 1 gives the titration curve for the foregoing system of buffers. For comparison, there is also given the titration curves of well known standard buffer solutions.*® The shape of the phosphate-citric acid curve of Fig. 1 indi- cates that the mixtures are well suited for colorimetric determina- tions of pH. THE CALCIUM CONTENT OF BLOOD PLASMA AND CORPUSCLES IN THE NEW-BORN.* By MARTHA R. JONES. (From the Department of Pediatrics, University of California Medical School, San Francisco.) (Received for publication, August 28, 1921.) In a recent communication by Jones and Nye (1) the distribu- tion of calcium and phosphoric acid in the blood of normal children was reported. As found in that series of observations the average calcium content of the blood of normal children ranging in age from 4 weeks to 14 years was as follows: whole blood, 9.4 mg. per 100 cc.; corpuscles, 8.7 mg.; and plasma, 10.0 mg. So far as we are aware, no data have been published on the calcium content of plasma and corpuscles in the blood of the new-born. The obser- vations herein reported on the calcium content of whole blood, corpuscles, and plasma in infants ranging in age from 4 hours to 12 days are a continuation of the calcium-phosphorus studies previously reported from this department. Technique. Lyman’s (2) nephelometric method was used, the technique described by Jones and Nye being employed except in a few minor details. In previous work in this laboratory the plan of checking our technique by making determinations on whole blood, cor- puscles, and plasma and comparing the actual whole blood value with that found by calculation from its component parts was adopted and followed in this series of observations. The blood was taken by means of syringe from the superior longitudinal sinus about 4 hours after feeding. Approximately 15 cc. were collected, 1 drop of a saturated solution of sodium citrate to 5 ce. of blood being used to prevent coagulation. As a precaution against * Part of the expense of this investigation was borne by a grant from the William H. Crocker fund for research in pediatrics. 187 188 Calcium in the Blood of the New-Born an exchange of ions between the corpuscles and plasma, the portion of blood to be used for these determinations (10 ec.) was introduced quickly through a short piece of glass tubing into a graduated tube containing 2 drops of the citrate solution and 1 ce. of paraffin oil was stirred gently, and centrifugated immediately. When more TABLE I, Calcium Content of the Blood of Normal Infants from Birth to 12 Days of Age. oe of Average calcium values etermi- | Average per 100 ce. Sex. Age. lanalasnilieci| Oe |S aIReSe| days per cent mg. mg mg. Boys. 0-2 5 50. 1 8.7 5.8 PA 2-4 6 50.5 8.9 5.4 12.3 4-6 6 49.1 8.8 5.4 12.3 6-8 6 45.7 9.0 Roe 12.3 8-10 6 47.2 8.9 Noll Pas 10-12 6 40.8 9.4 5.0 Pas Averagetan:. Sos cck 0-12 35 48.6 9.0 5.3 12.3 Girls. 0-2 6 54.8 8.2 4.9 12.2 2-4 49.2 8.5 4.9 12.0 4-6 5 49.0 8.6 4.7 12.2 6-8 5 45.7 8.8 4.7 12.3 8-10 6 43.9 8.9 4.5 12.1 10-12 5 42.8 9.1 4.7 1pAee AVETALES. shoe aie 0-12 33 47.4 8.7 4.7 12.2 Boys and girls. 0-2 11 55.0 8.4 5.4 12.2 2-4 12 49.9 8.7 5.2 12.2 46 il 49.1 8.7 Sel 12.3 6-8 11 45.7 8.9 5.0 12.3 8-10 12 45.5 8.9 4.8 1233 10-12 11 41.9 9.3 4.9 12.4 AVeYAGE acs cca 0-12 68 48.0 8.8 5.0 12.3 than one determination was made on an individual, the same centrifuge tube was used each time and the hematocrit reading carefully taken, due allowance being made for the amount of citrate solution which was added. Each sample of blood was analyzed immediately after collection. —— kid Martha R. Jones 189 After a few analyses according to the procedure described by Jones and Nye the calcium content of the corpuscles was found to be so low that it was necessary to change the dilution in order Per cent of red cells Mg. of calcium per 100cc. 0-2 a4 4-6 6-8 8-10 10-12 Days after birth Fic. 1. Graphic representation of the calcium content of the blood of normal infants from birth to 12 days of age. to obtain satisfactory nephelometric readings with the standard used. 3 cc. of corpuscles were laked with an equal volume of distilled water instead of 9 cc. and 5 cc. of the 1:1 dilution were added to 20 ce. of the trichloroacetic acid solution. Determina- 190 Calcium in the Blood of the New-Born tions were made on washed and unwashed corpuscles, and since the differences between the values obtained were insignificant, washing with 0.9 per cent NaCl was discontinued. 68 determina- tions were made on 22 infants, 12 boys and 10 girls. In 12 cases the first sample of blood was taken as soon after birth as possible, within 12 hours, and at intervals of 3 or 4 days thereafter until four analyses had been made. In 7 cases, the first sample was taken during the 2 to 4 day period, and in the remainder (3 cases) on the 9th day. TABLE II. Averages and Variations in the Calcium Content of the Blood of Normal Infants Ranging in Age from 4 Hours to 12 Days, and of Norma! Children from 4 Weeks to 14 Years. Calcium per 100 ce. Red plese Whole blood.| Corpuscles. Plasma. — mg. mg. mg. per cent ee Infants. i. -).i 8.1 4,2 11.4 37.1 Children......: (e I aol 30.1 fers at infantsa: eae 8.8 5.0 128 48.0 pee ee eChildren | eee 9.4 8.7 10.0 38.2 Hich GMB Bac aot 10.3 6.9 13.2 64.5 6 Children....... 12.4 12.0 12.4 44. The results of the analyses are given in Table I and Fig. 1. For convenience in comparison, averages and variations in the calcium content of the blood of infants and older children are given in Table II. DISCUSSION, An examination of Tables I and II shows that the average calcium content of blood plasma is higher in the new-born than in older children. This is in agreement with the findings of Meigs, Blatherwick, and Cary (3) who showed that. in heifers the calcium content of plasma is highest at birth and tends to become lower with advancing age up to 6 months. The value for whole blood (8.8 mg. per 100 cc.) is slightly less than that (9.5 mg.) reported by Brown, MacLachlan, and Simpson (4) in normal Martha R. Jones 191 infants under 1 year of age, while the average content of corpuscles (5.0 mg.) is markedly less than»that (8.7 mg.) found by Jones and Nye in older children. It is interesting to note (Table I) the constancy of the plasma values throughout the series, the average for each period varying less than 0.4 mg. per 100 ce. from the general average. Apparently in man the drop in the plasma content does not occur during the first 12 days of life. In general, the corpuscle values tend to decrease’ slightly during the first few days, the difference, however, in the majority of cases was well within the limits of experimental error of the method and much emphasis should not be laid uponit. On the other hand, there is a tendency for the whole blood values to increase. This is what one would expect with relatively constant values for plasma and corpuscles and a marked decrease in the percentage of cells. As Jones and Nye found in older children, the calcium values in the new-born tend to run higher in boys than in girls. The significance of this difference is not understood. Probably the most interesting feature of the hematocrit curves is their step-like character. This is shown most strikingly in the general average. The sudden rise in the boys’ curve in the fifth period can be accounted for by the very high percentage of cells in one infant, from whom the first sample of blood was taken on the 9th day. A few preliminary observations on the fate of the red blood cells during the first few days of life indicate that the decrease in the percentage of corpuscles is due to a relative increase in plasma volume rather than to a destruction of the cells themselves. Work on this subject is now in progress. SUMMARY. A series of observations on the calcium content of the blood of normal infants ranging in age from 4 hours to 12 days is reported. 68 determinations on 22 infants, 12 boys and 10 girls, were made, the average values being as follows: whole blood, 8.8 mg. per 100 cc.; corpuscles, 5.0 mg.; and plasma, 12.3mg. The average for plasma is higher than that reported in older children while corpuscle and whole blood values are less. The 12 days included in the series of observations were divided into six periods of 2 days each, and the results of the analyses made during each period averaged and plotted. The plasma values 192 Calcium in the Blood of the New-Born remained constant throughout, while there is a tendency for the corpuscle averages to decrease and those of whole blood to increase. The average percentage of red blood cells dropped from 55 to 41.9 during the 12 day period. To Dr. Bradford F. Dearing of the hospital staff I wish to express my sincere thanks for his interest and cooperation through- out this investigation. BIBLIOGRAPHY. 1. Jones, M. R., and Nye, L. L., J. Biol. Chem., 1921, xlvii, 321. 2. Lyman, H., J. Biol. Chem., 1917, xxix, 169. 3. Meigs, E. B., Blatherwick, N. R., and Cary, C. A., J. Biol. Chem., 1919, OS Qi, Ie 4. Brown, A., MacLachlan, I. F., and Simpson, R., Am. J. Dis. Child., 1920, xix, 413. ae OBSERVATIONS ON BLOOD FAT IN DIABETES. By N. R. BLATHERWICK. (From the Chemical Laboratory of the Potter Metaboiic Clinic, Cottage Hospital, Santa Barbara.) (Received for publication, August 30, 1921.) The former conception that diabetes is the result of a deranged carbohydrate metabolism is now known to be only a partial truth. It is today recognized that the metabolism of carbohydrate is faulty; also the utilization of fat and protein. The metabolism of fat in this condition is particularly interest- ing and any study of the factors entering into the incomplete combustion of fats leading to ketonuria is desirable. As a result of the investigations of Bloor (1916), and of Joslin, Bloor, and Gray (1917), the extent of the increase of blood lipoids in diabetes has been demonstrated. Their data revealed that the greatest increase was in total fatty acids, secondly in cholesterol, and lastly in the lecithin fraction. This pioneer work with the satis- factory method of Bloor is of fundamental importance in arriving at a better understanding of the behavior of fats itt diabetes. Newburgh and Marsh (1920) have recently suggested a new method of treatment of diabetes, designed to overcome the undernutrition incident to the Allen and other similar diets. Accordingly, a low protein, low carbohydrate, high fat diet is advocated by these authors who reported seventy-three cases to have been successfully treated by this method. Their criteria of success were (1) a sugar-free urine; (2) no acidosis; (3) nitro- gen balance maintained; and (4) capability of resuming the ordinary activities of life. Their diet is so constructed that it contains about 0.66 gm. of protein per kilo of body weight. This system although revolutionary in character, offered obvious desirable features if workable, and has been given a thorough trial in this Clinic. If the high fat diet does in fact allow the patient to be free from the acetone bodies, this should also be reflected in a constancy of values for blood fat. 193 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 1 194 Blood Fat in Diabetes Methods. Total blood fat (fatty acids plus cholesterol) was determined by the method of Bloor (1917), the methods of Folin and Wu Patient B Response to incomplete Per Cent Fasting 2.80 Blood Fat 2.50 2.20 1.90 1.60 1.30 1,00 0.70 0.36 0.30 0.25 0.20 0.16 0.10 Dist. 4n Crans\|-a-- a ee Carbohydrate “ : Bs Prote Epes Rate Fat 20 ¢ ° Smal n 10 3 E 5 =z 9/8 9/10 9/13 CHartT 1. Male, 22 years of age. Sugar discovered in urine July, 1920. Best weight 145 pounds: now (Sept. 8, 1920) 79 pounds. Admit- ted on verge of coma. Sept. 8, 1920, he excreted 111 gm. of sugar. The urine of this date showed a + + + acetone and + diacetic reaction. The urine of Sept. 12 contained 18.6 gm. of sugar and was negative for acetone bodies. Lipemia was present in the first blood drawn. (1920), and later of Shaffer and Hartmann (1920-21) were used for blood sugar determinations. All blood samples were taken in the postabsorptive condition before breakfast. 195 N. R. Blatherwick *SoIpOd 9U0490% JO UOT}O1OXO 10U BIMMOIATS JOYPIOU SI O14} JOIP SITY UG “AlVEp Fey Jo "ud JCT pus ‘urezo1d jo "wi gg ‘oyvapAYOqAv Jo “Ud gt BuIyvo (ci ‘Bny) Mou st puv ‘OT oUNL OUTS doIJ-1BSNS UdIG SLY OF] ‘ozlavBnsop 03 pavy AIOA SVM JuOTyed oY, “[wUILOT dAOge AT}BOIZ OSLO OU UT SUM 4B POOTY “Fuasoid 19M 90J99B JO SodvI} porsod oy} Jo ysou Buring “ws 79g poyoroxe oy ‘TT “Ady ug “TZ “Avy wo ‘m3 PT OF BZ Goq UO oovIy v WIOIJ YUNOW UT SurAIvA ‘A[SNONUT}UOD IedNs poyoroxo quoryed siyy, “(WS pp) esoon[s Yue. sod Z'9 pourvyuod ours jo uowroeds oy ZI ¥ ‘OL “Gat poyrupe uoyM ‘Spunod OTT OF 8OZ Wor] YSIOM UL SsO'TT “UOT}BINp sivoXk g JO soyoqviqT “plo Sivek FQ ‘OVI, *Z LUVHD Tt/t +/¥ 8¢/¢ 12/e *1/e 8/2 92/2 12/2 91/2 YH in 4 es v9 06 484 . O42 EJ ? : UTte4Odd ‘ =e oo: th peeeerr Se = a tr pane a 0470 4¥q pootg nus aed W yueT4Va 196 Blood Fat in Diabetes Patient W Non-effect of food fat upon blood fat Per Cent 0.75 Blood Fat 0.70 0.€5 ae 0.60 0.55 0.30 2 cl 0.25 q Blood Sugar ; 0.20 > 0.15 : 0.10 Beau Carbohydr “TB0 Ba . Protein 150 iV 120 ‘ Sh .Z 2/14 2/21 3/8 3/14 3/21 3/28 4/14 ¥ Cuart 3. Male, age 49. Diabetes complicated with chronic pulmonary tuberculosis, chronic arthritis, and chronic nephritis. Diabetes of 4 years duration, with resultant loss in weight from 194 to 120 pounds. When ad- mitted Feb. 10, a 12 hour specimen of urine contained 7.1 per cent glucose (119 gm.). The sugar excretion of this patient was 6.1 gm. on Feb. 14. He was sugar-free thereafter. A-+ + + acetone reaction was obtained on the above date and none or mere traces afterwards. Blood fat values in this case are also not much above normal. *potsod oy} JO JopulvUlel OY} SULINP pops0d901 SVM OUON "PT “qo [UN 9U0Q00% JO S00¥1} O1NMI pu PUNO] svM IvBns OU ‘OT ‘ULL I9JJY ~O9U0J90B Jo 90vI} B PUG IvdNS YUoD Jod GG poUTeyUOD g ‘ULE JO OUTIN o"uL, “2 ‘Uee poyIuIpy “4arvyo Aq poloAOD [VAIOJUI BULInp suolyvsodoO OMY UI POAOWOA sJOVILYV OTYOqQVIp [v10ye][IG ‘porrod sy} Sutinp spunod QZ 0} GET WOIJ FYSIOM Ul SSOT ‘UOT}PVINp SIvos ZY Jo SoyoquIqT ‘sivok [Pp ose ‘osInuU ‘o[VUIOT "Pp LUVHO ‘oe Li/v &/¥ 82/e 12/e bT/e 8/¢ bi/e Te/1 72/7 A/T O/T B/T 5 wn Val ie a8 On ee = 4 | Za : od eee 7 ie gg A ae a @ : je Ze By. Fy ‘Sb ¢ th, , ws “a 9 = 06 ors - a | O2T o Bl ~ a OST _ a A sed _ ie , “OST Uys Aug ; Zi aqerpAyoquey , eed Ul 4eTd y n 10} ie) ame i ae . —$ $ —_$ $$$ = ———— ase 02°0 eee: 1g pee a es ersae) Avang poorT| in 09°0 at} pe a a atic err 7 080 484 poolg ee 198 Blood Fat in Diabetes Explanation of Charts. In the charts are given percentage values for blood fat and blood sugar, and the diet in grams of carbohydrate, protein, and fat for the day preceding the blood sample. The diet was often constant over a long period, as will be seen by reference to the charts. Any supplementary data will be found in the legend of each chart. RESULTS. Chart 1 shows the response of a severe diabetic to partial fasting, such a restriction in diet as is generally prescribed when a patient exhibits marked ketonuria and lowered alkali reserve. In the course of 5 days, the values for blood fat and blood sugar are seen to have nearly reached the normal levels. The CO, capacity on admission was 26.5 ec., which became 74.8 cc. per 100 cc. of plasma on the last day charted. The drop in blood fat here found is not always to be expected. Bloor (1914) showed that the blood fat of fasting dogs may decrease or increase, depending upon the original nutritive condition of the animal. Probably the result here shown is the ordinary response of a hyperlipemic individual to such a restriction in diet. Charts 2, 3, and 4 show the behavior of blood fat and blood sugar of patients living on diets similar to those recommended by Newburgh and Marsh. The outstanding feature is the con- staney of values for blood fat with increasing amounts of fat in the diet. This indicates that these patients were utilizing the large amounts of fat in a satisfactory manner. SUMMARY. A study of blood fat in relation to the fat in the diet shows that cases of mild and moderate diabetes are apparently able to utilize satisfactorily large amounts of fat, as indicated by con- stancy of the blood fat level and by the absence of acetone bodies in the urine. It remains to be learned whether or not such high fat diets continued for a considerable period will prove to be an overstrain on the fat burning mechanism. N. R. Blatherwick 199 BIBLIOGRAPHY. Bloor, W.R., J. Biol. Chem., 1914, xix, 1 Bloor, W. R., J. Biol. Chem., 1916, xxvi, 417. Bloor, W. R., J. Biol. Chem., 1917, xxxi, 575. Folin, O., and Wu, H., J. Biol. Chem., 1920, xli, 367. Joslin, E. P., Bloor, W. R., and Gray, H., J. Am. Med, Assn.,1917, lxix, 375. Newburgh, L. H., and Marsh, P. L., Arch. Int. Med., 1920, xxvi, 647. Shaffer, P. A., and Hartmann, A. F., J. Biol. Chem., 1920-21, xlv, 365. -“ LIPEMIA., By W. R. BLOOR. (From the Department of Biochemistry and Pharmacology, University of California, Berkeley.) (Received for publication, September 28, 1921.) The blood plasma in the postabsorptive period is normally clear although containing appreciable amounts of material of a fatty nature which is insoluble in water—cholesterol, choles- terol esters, lecithin, and probably also some fat. The introduc- tion of more fat into the blood, whether by feeding or by mobili- zation from the fat stores, generally produces a milkiness nor- mally lasting only a few hours, but which in certain conditions, mainly pathological, persists for considerably longer periods. The milkiness of the plasma is termed lipemia. In human beings the plasma contains normally 0.5 to 0.8 gm. of lipoid per 100 ce. and lipemia ordinarily appears when values above these figures, but generally below 1 gm. per 100 cc., are reached. In diabetics, values above 1 per cent with clear plasma are often found (1) and occasionally remarkably high values are reached, as in Case H in Table I, where a value of 4.35 gm. of total lipoid per 100 ce. with clear plasma was found. Allen (2) has recently reported a case of hepatic cirrhosis with ‘‘total fat’’ values of 3.63 per cent with clear plasma, and similar cases of ‘‘masked”’ lipemia have been frequently noted in the literature (3). The masking may be an unstable condition, since on standing for a time (24 to 48 hours) milkiness may develop in a plasma which was clear when drawn.! This latter phase of the subject, although interesting, has been very little investigated and will not be dealt with in the present discussion, which will be confined to lipemia as de- fined above—a more or less prolonged milkiness of the plasma. Alimentary Lipemia. _ During the absorption of fat from the intestine the fat of the blood of most animals increases and eventually a more or less 1 Bloor (1), p. 429. 201 202 Lipemia pronounced lipemia is produced, the extent of which depends on ’ several factors, of which the most important are probably the rate of absorption from the intestines and the rate of disappear- ance from the blood. The rabbit is unique among the ani- mals investigated in that lipemia apparently cannot be pro- duced in it by fat feeding (4). There can be no question about the absorption of fat in these animals and the explanation of the non-appearance of lipemia must lie in the fact that the rate of disappearance from the blood is equal to or greater than the rate of absorption. In most animals if the increase of fat in the blood is considerable, it is followed by increases in the other blood lipoids. There is an increase of lecithin (5 to 8) and some- times also of cholesterol (6), although tae latter is not always found (7), which may be explained by the observation that the increase in cholesterol often does not appear till late in the period of absorption (9). There appears to be a definite se- quence in the changes in the three lipoids in the blood during fat absorption—the presence of fat if sufficiently large in amount or sufficiently persistent causing an increase of lecithin, and this in turn is followed by an increase in cholesterol. However, as was pointed out by Bang (10), animals vary a great deal both in their reaction to ingested fat (production of lipemia) and in the effect which fat in the blood has on the other blood lipoids, and it is not always possible to demonstrate these changes. Hueck and Wacker (11) have found that prolonged feeding of choles- terol may also produce a lipemia (rabbits) with secondary in- creases of lecithin and fat. In the study of the blood in alimentary lipemia, examination in most cases has been confined to the plasma but in the few instances where the distribution of,the lipoids between plasma and corpuscles has been studied (8) increases of lecithin and fat have been found in the corpuscles as well, in fact the main increase of lecithin in the blood has been found there. This participation of the corpuscles in the utilization of fat appears to be only tempor- ary since in the persistent lipemia noted below the corpuscles appear to have little or no part. Persistent Lipemia. The lipemia produced by a single fat feeding normally dis- appears within 24 hours and persistence after this time is prob- eee en W. R. Bloor 203 ably to be regarded as pathological. (This statement cannot apply, of course, to instances where the ingestion of fat is more or less continuous—as in suckling animals or in animals sub- jected to forced feeding for fattening or other purposes, in both of which the lipemia may be continuous.) By far the greatest number of cases of persistent lipemia have been found in diabetes, although nephritis and chronic alcoholism furnish some examples. In this condition when examinations of lipoids other than fat have been made it was found that both lecithin and cholesterol were increased along with the fat—that there was a “‘lipoidemia’”’ along with the lipemia (12). Since the discovery by Boggs and Morris (13) that a high grade lipemia can be produced in rabbits by bleeding, considerable study has been devoted to this type of lipemia and the data obtained are of especial value since the lipemia has been followed throughout its whole course (14, 15). In hemorrhagic lipemia also, all the blood lipoids are found to be increased. The purpose of the present paper is to present additional data ‘on the blood lipoids in persistent lipemia especially that of diabetes and of hemorrhage, and to discuss this and earlier data with the idea of reaching some conclusions as to the mechanism and probable cause of: persistent lipemia, and its relation to normal fat metabolism. The Lipemia of Diabetes——In diabetes, under the older forms of treatment, persistent lipemia was a common phenomenon and it was early recognized that the milky appearance of the plasma was due to ‘‘fat.”” As to the nature of this fat very little was known until the beginning of the present century when Fischer (3) found that it contained abnormally large amounts of choles- terol, a result which Klemperer and Umber (16) confirmed, and noted in addition that the content of lecithin was also high. Investigators since that time have in general confirmed these findings although not always (17). Examination of the blood of diabetics without visible lipemia (blood plasma clear) has shown that all the blood lipoids are generally abnormally high es- pecially in the severer cases (1), so that the diabetic may be said to be predisposed to that disturbance of fat metabolism. of which the visible sign is lipemia. Up to the present time the data on blood lipoids in diabetes have been obtained by the 204 Lipemia analysis of single samples and no attempt has been made to follow the course of the lipemia by examination of repeated samples taken during the course of the lipemia. It was felt that such a study would be of value not only from its relation to fat metabolism but also as a basis for the rational treatment of such abnormal conditions. Satisfactory material for examina- tion is not readily obtained and the author is greatly indebted to Drs. E. P. Joslin and A. A. Hormor of Boston for supplying not only blood samples taken under necessarily exacting condi- tions but also for furnishing other data of interest. The methods used were the published ones for total fat, cholesterol, and leci- thin, and need not be given here. The results of the analyses are given in Table I. For purposes of comparison the normal average values for blood lipoids in man are given at the top of the table also, and in the case of Subject Cl, the analysis of a blood sample taken some time previously when his blood plasma was clear. Discussion of Table I.—The first case presented is the most inter- esting, both because the lipemia is the most pronounced and because it has been more extensively studied than the others. The patient was a severe diabetic, who had been under observation and careful dieting for some time. On June 10th he broke diet, eating among other things large amounts of milk and cream, On June 12th the first sample of blood was taken, the patient meantime having taken no food since the time of breaking diet. In this first sample is to be noted the very high values for all the lipoids, the total lipoid being approximately eight times his own normal value and fourteen times the average for normal men, while lecithin is approximately four and cholesterol eight times the values found in normal men. The greatest increase is thus in the fat, as has always been found to be the case in lipemia. Other notable points are the low corpuscle percentage (diabetic anemia) and the low values for the ratio a The high cholesterol urinary ammonia and the low carbon dioxide tension in the alveolar air are also significant. (The data in the last column were supplied by Dr. Joslin.) The next two samples taken June 14th and 15th show higher values for the lipoid constituents, the total lipoid value on the latter date reaching its maximum of W. R. Bloor 205 about 13 per cent. Whether these higher figures represent an actual increase in the fatty matter of the blood (which could come only from the fat stores, since no fat and very little food of any kind was taken), or whether it is due to a concentration of the plasma is open to question, but might well be the latter, since the blood corpuscle percentage in the last sample was 39, an increase of about 33 per cent over that of the first sample taken, (29 per cent). The lecithin in the last (third) sample is five times, cholesterol nine times, and total lipoid twenty times the values for normal men. In succeeding samples the total lipoid (plasma) diminishes rapidly, the lecithin less rapidly, and the cholesterol relatively slowly, showing that the fat is the first to decrease as it was first to increase, the lecithin next, and the cholesterol last. In the last sample examined, taken about a month after the beginning of the lipemia, the lecithin value was still about 30 per cent, the cholesterol about 60 per cent above, and the total fatty acid over 100 per cent above the normal values for thisman. Although the plasma was now clear, calculated values for fat (fatty acid not combined as lecithin or in cholesterol esters according to the usual assumptions) in the plasma of the -last sample gives a value of over 0.5 per cent which, since that amount of free fat would result in milkiness, indicates that the fat or fatty acid exists in the blood in some form of combination Lecithin Cholesterol in plasma is also much above the normal due to the slow rate of disappearance of the cholesterol. The addition of fat to the diet toward the end of the lipemia and increasing its caloric value did not cause any increase in the lipemia (with the possible exception of cholesterol) which continued to diminish about as before. In the second person studied (Subject Sc) much the same phe- nomena are observable. The beginning high values for total fatty acids, lecithin, and cholesterol diminish slowly during the 3 weeks of observation; the values for the ratio Boies remain cholesterol (except in Sample 566) much below normal, and the calculated value for fat in the clear plasma of the final sample (about 0.5 per cent) indicates as before the presence of unknown compounds other than those ordinarily believed to be present. ipemia L 206 “FYSTIS “Oro -BIp {70'S “HN ‘Aurvoso vuseyd |g¢"0 0L°0 OOT‘ET| 06Z |006‘T/OOE‘TOOS jOTT‘T| 006 OFE‘Z\008‘OT/00G‘2) 6& | ST, LoL ‘g0R1} OIZOOVIP = —* ‘g¢ “OO qepoosTe ‘T'¢ “HN +60 ‘ze8ns poojq ‘fAuvelo vusetg |¢p'0 |0L°0 |0&Z‘6 | OFZ [OOS TONS‘ LOST‘ TOTS | 026 000‘F|00Z‘2Z 00Z‘9) Z | FL», | Gat ‘WU FZ OOD IBpOaATe { 1Z "0 ‘IBSNs poolq {6°¢ *HN ‘!+-+ ‘oreoerp ‘Auer BUISV[q ‘“pooj 1a4ye ABp puodvag |eGg'O |€2°0 |OG8‘8 | OZF |009‘T/09Z‘ TOO‘ TOFS | 026 000‘ F\000'L |OOT‘9| 6% | SI PUNE | FT ‘OL oun JoIp VyxoOlg ‘UBUL SIY} JOJ Son[VA [VULION |08"O |ST'T |OOL‘T | 02% |O8F jOZE (OSS [OSE | OFF [OIF |09F OFF | GP | F “URL | SF TO yoolqng “UO IOF OSVIOAG [VUION |96'0 |PF'T |029 o6t loz jotz ‘oor lozz | oog O9E |O8E j09E ‘Bu | ‘Ow-| ‘Ow | Ow | Ow | “Ow | ‘Om | “Ou | -Bm | Om | 107? ‘QSVIOAG [VUIIO PP ise Ree | bol Pee lp loPolee| e | | § ae | ge | ee | ‘syIvULEy a Re a sy ‘a}8q ‘ON = Jor10qs0,04, Dae “29 0OT ine | ‘09 QOT sod uIqy1007] ptodry Jad [o10}s9[04O 09 0OT ted urqqtoaT prow A}48} [8307 [810 “pymadyT 92j9qQDLq] TI TTaVL 207 Se ag ee ey “TROO VUE |19°0 060°% 08 oss 080‘T Gh | cl » | Pg ‘OT'O ‘1e3ns poorTq ” ” ” F9'0 |76'0 |029°S | 022 jOS8 jog9 jo99 Jore | 06S losr |oo9'T lozt‘T| se | 8 » | 12g ‘Apnoyo A]PYSI]s8 vusey_ |9°0 |98'0 \o6r‘z | ze loz8 loro lose loge OSS [00F jO&P'T 000'T| Ib | F UNL | ggg "S10 ‘183ns poorq {skup z 10} 4R] jo "wid Og ‘AX[UA ATPYBT]S VuUseT |06°0 |12'0 loez‘z | OFZ OL looz OVE [099 | OFS 00S [O€Z'T 096 | 48 | Te ,, | 99¢ "e1'0 ‘1e3ns poorq f4orp 9o1J-78J WO SXvp Z ‘AYU vUIsETY |Gh'0 |8G°0 |Ozz‘e | OSF l0zS‘T|006 OOF |069 | 02S 0S8 jO8c'T 046 | a | Ge ,, | gO ‘ZO ‘av3ns es pooTq *Fuoonsuvry ‘Axium vouseyT |6G°0 |89°0 logs‘z | ore loog‘tloes lors looz | oF9 ze‘ tlooe‘T Oze‘T] 26 | 22 ABI | TOS | ‘oa qool aa g yoolqng 3 “8T'0 “ABSns pooyq ‘1vopo vuseyg |29'0 |08'0 |oz0‘z | O1Z lo9z lozs jo9z lots | ooF lote loot‘t LO) A al ay Ha 0 a fam "yuo 10d . 9g ‘JOA94s0TOYO pourquiod fet "¢ Ee ‘rv3ns poorq ‘yusoseredo __,, 99°0 |04°0 |091°% | O8€ \008 jos9 jose joes | ost gz looz‘T logo‘tl 9¢ | 9 4, | OFT ‘0 ‘orgaourp s Axion A[pUTey BUST |2/°0 082'% 092 06 008 ‘T @ Arne | ser *(7BUL -10U) ued Jed g¢ tales pourqaios {Tg “Oo «RKpooATe ‘0 “oreourp Aprox ‘Urq, vurserg?lo9'0 |o2°0 lo6e'z | ore lore lose loge loze | oge OZI TOS T \0G2'Tl Ze | 8S 4, | OST Ze FOO ajo “OAe <0 ‘ojooulp ‘Ayu vUIseTA |8h°O |¢9°0 |09z'F | Oze jose‘ tloge logs lor9 | 009 ogo‘ tlooz‘z logo‘z 6e | zz» | ost "Te FOO avjooaye | os *0 ‘orjoourp :Axprur “Joryy vumseTd |eh"0 |19°0 joge‘9 | 00e |o9r‘ logo‘ 199 lozo | oF9 lozz‘tlooz't loose! ge | ozoung | ger Rs etme SPN en Tape 2. | S59 SA ae cae |B La ht IK¢ ‘IVOTD VUUSLTT | 0'9 | O'S |FFS | OOT | 2 096 | CIT |\OF9 OGL | GST |Z OE} ZT s,s, LI “FT pue “eT ‘21490 Uaye} 919M pooyq sofdures “09 CT __,, ” SF | 2s |Si% | 002 | SZ |096 | 08 |OL9 jO9T | ST |S°9G| OT ,, cI “IB9]O oy, PF € (09 GL |0&6 096 |OF9 OFT See Olea ae ess IT "AMIIUI ,, SiG) Gees SSS OST | 8 |OSZ‘T| 008 jOFZ joGE GT 10 047) OL =; OT ” ” ” ¥911§ IP 00Z | OOT |O8%‘T} OTE |00S |O0F9 cI. |G 61 | 6 ” 6 ” ” ” 0'2] 2% |066 OLT | OLL |00Z‘T} OFZ |OE9 [098 CT sola es ” 8 ‘ ‘AypIuT ATOA 4 IOS else OLL | OS |OSZ‘T| OFZ |009 {029 OSm GSO ” L ‘AYU, 9912'F [06 | OLT | OG [OZI'T] O12 009 j0E8 | OF |6°9T | 9 ” 9 ” ” ” LP \S1I8 CT OZ OOL og jO' St] G - ¢ “AY[IUL YVYMOUIOS —,, FAIA 061 ore | oe le oli Fr =» |F a ” ” O'F ESE €G 06 O00€ 0€ |G 13) € ” € = ” ” 9 F119 /19% | OFT | SL |OF9 | OS jOSh (06% | O€ |L'8Z| @ ” G 5. “IBID BUISB] 8°¢ |8S¢ ST GL 0ce | OF (10°98) LT “POT = ‘Bu | ‘Bu | ‘Bw | ‘bu bu | ‘Bw Bu “99 as "SOTT] GLE FUSION “VW FIQqVy i 5 Be |ooe B |e 5 = 5 a | 2 9, 3 2 be & H 8 ® g = 5 S 4 = = — — oy = oO D ® ® . D Z 2 a V3) 2 2 2 *SYIVUIOY, | a ee ee aqeq aes 0193890 ol a kote d ; d 99 QT 10d aes SF jousysqoug | wryso0"] Poo ad 28. oe N me ‘prumadry oboy.iowapyy] ur sprodvy poo.g Tl GIaAVL 213 W. R. Bloor LOTION eorpmr 2 UIY}109T Ts ca Sa a ae sgt BT eee st SS oe, gt ee ee oe ‘Ivao ,, GG] 1's |OOT | StI] St [008 | 9¢ |osz {99 SI |¢'te| ¢ ‘ady | IT “AY[IUI ATJUIVT, 29 | 9. |\86E | 091 | 0G 1000 T) 08 (O22 Ison 0c i2e8e) Jeo spe Ot “AXIOM: 5, Io | 8'l |a9 | Gat | 88 [008 | OO lore logs 108 l9°6L| # , 16 “AIM YVYMOUIOS —,, 98°T |929 OOT 9ST Ogee” leap I9°OT | eo ns 1 is 5 LP | S'S ere -| SFL | FF |009 | ZIT joze joes | 08 IT'St| cz =» IZ “AypIUr ATJUTVy ,, 0°% |8ze | ' oP 68 Geom (22 TSS Te an tO “Appnur A1,qSI][s vurse[g | FS | O'S [6st | OST | 2% jos | z8 loze |FoL | Gz IF'st| oz » |g "ST Av] Uae} pOoofq "09 GZ yeruediy ‘Apnoyo _,, $°S |FLT v6 08 CST heerrlOree Give ase riee 9 tf S61 Se |Spre | Cat | See l0SPs | SSe OSs O0T Ieee Ie) Zt 5 Ne ‘Appnu vuse[_ | 9°F | 9°% |6FT | OIL | €@ [FOS | 09 [gee |90r | Of |¢'se)| or ‘ae | T ‘SOIT FS WUSIOM “OD 91949" ; 5 9°9|/¢°¢|t6 | 06 | 9L [009 | 9¢ lore ‘Ivaloy, L°6| L2°& |Z91 | OFT | ZT |09g‘T! #9 00g Ayjrut ApyuIey,, 0°6 | O'S JOFE | OFT | OG [008‘T| OST [009 Fy “0 0°S | 0'€ JOST‘T| OFT | S8 j00z‘T| 092 008 » 5 02 O9T |29¢ |OZE‘T| OST |or9 ” ” €°9 | 8°& [SST‘l] O6T | 06 |00Z‘T| OFF [029 ” ny 6°9 | 6°S |SZI‘T} O6T | 06 [008‘T| 092 [008 i 5 0'S | O'S |OTZ‘T| OFT | OOT JO9T‘T| 00S [099 “AYU, L'9 | O'F |060‘T] SFE | 09 |026 | OFZ |099 9 iS en 9° | ZF \669 | OFT | SF |O80‘T! O6T jog9 “Ayr AToyw1opour __,, T'S | 9°S [O29 | OFT | FF [0%Z | O9T l00¢ ‘AYIA ATUIvy ,, 1% |2og | 08 | a joz8 | 88 jose » 9 Z€1 SL |98t | 002 | a l0F9 | #9 loz + UIA F148 09 MOT S . ff GE | 6'L 9ST | 002 | 8 [0F9 | F9 Osh jOOT | OF Ig'Ez| ce » ‘IvoTD BUISvT | 1° Og1 | 99 [029 009 [OZT | 08 |F'tE] 12 “0 ‘SOIT 6S FUFIOM “a JqqVy lipemia 214 “Ay[Iod AT}UIey ,, 0°€ | 2°% |00L | 002 | FOL 1009 | 88z 006 j0o0s | 08 |F'0Z SI » | OL *“AYTUL AppUTVy ATOA ,, 1%|8°% FOL | 022 | HOT [009 | 962 |G66 j00OS | Gz jO9T 8 » |6 “Ivapo,, ¥%| S'S |F6r -| 022 | 2 |SzG | 91z |00O'TlOgs | GZ 16°21 | Fr » |8 “ANTM ATurey AOA, Zo | 9°S |99¢ | 002 | 26 \OMP | 8FZ Oz [068 | GZ |I'eL| T “PO | Z ” i ¢°21¢°% |8h | OSL | 08 |96F | 002 [002 |00E | Oz |S'ST| 6 » | 9 » » 6°21 2° OSS | ZhL 1 02 |zer | Fez IFGO |F8E | OOS |2°SL| 82 » |S » »” G'€ | 0°S [61S | SZI | 99 [OF | 002 jOzB |S8E | Sze [6°22 | LO» | F ” 5 ce | 0's |6s¢ | 0&1 | 99 j9cmr | 002 jo8z |scr | Oze |g°0e| GZ » | > 33 8'% | 6'S |08¢ | ZFI | 99 |9IP | EL \09Z lose | 0% |T'1E| H » 1% ‘IBop VUISB[d | S'S | ZS |6IS | FST | 9L jO9E | SAT JOS8 [Sse | OOS |e'sh | ee “4dog | T "SOTTT ZT VGSIOA, “30 ‘Appnur _,, 99°F | 2% 1602 | SZI | OF (008 | 68 Ore |OFT | FL [8'0E! 26 » |8 “AMIGO ATPUIVZ ,, 9°S | 60 |2Z9F | OLT |2G9T |096 | ZT jO0e josz | 0% |g'0c| 6. » |2L “AY[IUE ‘UTYZ ,, rT |908 C6 eI YOU |5@t. eee | Sl 5, 1-9 “AY[IU ,, G°G'| O'S 00h | OST | GB |Zes- | Zz lose 199s |0z |€°12| Zi -» |S “AMM “UIY} ,, €°S |ber OL 091 O08 Oc 1S2ce OT Ge ” » ” » - CPP Ole. Serr i6s.. 809" lero. Ge. Ics) cst |S fa) Sr a 1s » » » ” : 9° |T6T 9g SZ 15 a OF a cal al st oer ean 3 ‘uorsuodsns osivo0o yyIM Appnur vuIse[g | ¢°s OST 00S | FIT |O1Z SI |o'9¢| eI ‘Ady | T “Bu Bus "Bu Bu "Bu Bus “Bu "20 ee | | “SOTE] SS FGSIOM “A H1qGqRY Q as] ae] Q ac) Q ae) Q ac) w Q Pp Ba RC oaea Recs diese (ee ag B Bele ce. \re: | See Pig Seetaage (ae Tone tile = ch, 8 is 3 fh . & : 09 2 & a a ob ; & 3 *SYIBUIOYY 278d 8 99 QOT red 0.19}89]0 geo] - . ; “Sa Se eae ooo pis 8) ‘papnjyaUuog—ii ATAV.L W. R. Bloor 215 beginning in three out of four experiments. In Rabbit B the values at the height of the lipemia are much higher than at the beginning, but in this case the values for the ratio were very low (for rabbits) at the beginning. In general much more marked increases in the lipoid values of the corpuscles are to be noted in these experiments than in those of Horiuchi. These are most marked in the lecithin, the cholesterol often showing relatively little change and the total fatty acid still less. In the dog, even with a reduction of red blood corpuscles from 45 to 13 per cent of the blood, no definite lipemia could be pro- duced and the only marked change in the blood lipoids to be noted was in the lecithin which was considerably increased in both plasma and corpuscles. Values for cholesterol and total fatty acids remained quite constant. It is significant that the : ithi : ; ratio sie Ga is consistently above normal throughout the cholesterol bleeding in the dog, while in the animals in which lipemia is pro- duced by bleeding it is almost always below normal which relates the low values of the ratio to the lipemia and not to the bleeding. Another feature of importance is that lecithin is more markedly increased than fat or cholesterol which emphasizes a point ob- served in earlier (unpublished) work, that slight or transient in- creases of fat may bring about an increase of lecithin without any change in cholesterol. In order to show more plainly the features characteristic of lipemia and to compare lipemia of different origin, data from this paper and from various other sources are collected together into Table III. Values for plasma alone are given since differences from the normal are most plainly seen there and the values chosen for comparison are those taken at the height of the lipemia. The value for the constituent in the normal condition, or in case this has not been determined, the normal value for the species, is taken as 1 and the values found are reported as multiples or fractions of 1. Discussion of Table III—A study of the data presented in Table III brings out the following regarding lipemia. (a) In all the types of lipemia there is always an increase of lecithin and cholesterol as well as fat. (6) At the height of the hpemia the greatest increases are to be noted in the fat, the next in the ipemia, L *78'0 pus 7e ‘eT vrutedr] Jo Suruursaq ye "G20 OF STTPF GT] Suruursoq yu "PIT 04 sosta *[BULIOU 04 STTBy je) T 8) T ) T “G9°0 04 STTRJ lIORIOHRIO «° 8°0 0} AT -SNONUI}ZUOD ST[Vj ET BIuedI] Jo Sutuulsaq 4B 'T Lt OE IT 6°% OT VG wT SVT OT 6'T wT £°% 90°T Vel €6°0 LS 6°0 (ane e¢é 8'é OTT FP eae * ) Gig Bee EI (GTaLoly) 0's eae “ae tse (( ag ) GuG Ce ON Ha( an ) COL Coto: O-ODeck . ( [319,7) Ge eee eee eee ( 5 ) 0°81 eee ee . ( 3 ) OIL 2 10 sake) is) » . ( [S19,7) ”? ”? *[BNI}SUI Jy ” ” "S997, ”? ‘uo1ye10dg 216 “uvUIny, a ee ee 8¢'0 62 6% SiGe ae et CF) ‘H » 80 Z'9 GT 8g eo a) = SO 88°0 vt 0% i ca aa Cai) == eis; 9F'0 6'9 1% he Pen ee oe ( » ) ‘uyog ,, g'0 9'8 0g OLOG ae ge a (100Tg) OSes eee eee ee eee. ae eal Ole oe a ‘Oryoqviq “uvuUINnyyT ea ee eee ee ee e [019489104 *[019989] P , uryyoeyT =| -O4D Saag e eae ee et ae ee ‘(7 = JowionN) vuuodvy fo yybray yw sprodvyT fo sanyo, aarnpaay ‘TH ATA VL “sy leuleyy “IoyyNV pues ‘aseo ‘erutedry jo dA J, cnt Be al Le $9} BOIpul 9 N UIY}L99] Ts ; : fe) : GZ a ee ‘ pl et eae 0% | sop aaryeqg | 7S | 'C » =) “IIL » ROD ee 808 marten ; 02 We Ns 3 80°T ‘sarp yeq | 908 | °C») WHI ” pu ye 67904 sprey2 | G0°1 $3 oe Uae ae Seti At "POIP 901J-78T A 8 e 9'F oer |" ( ) vq xe) 8 0 a 9 "yoIp qBT a G ” ” jaa) se ; @) , SZ eae i 2 0ey =dorp step ett oS WE! sory ooapqug | 2 SEL Co») al n aa 10 “9 ; 9°¢ ge Pee = -YSNOIY ON[BA SUIUUISeq OY} MOTO SUIvUIOI T 160 OT “qoIp FBT ue COnITOEL) VLA nemiedar ‘qe “sABp G UT 18°0 OF sty 5 ST wT 61 Dy el [ncaa rae Nelo atie) ” "G8°0 04 stv 2 0'T sas Tal BGA le eet C959) » “(JOM T) 109R] GZ" T 07 Sosta = 40'T rT rT 0S hala Cm) » 40'T 0% 0% ie oa alll CT) “GPT Way 4B = S1 61 £3 Eg po ([S1eqT) ‘stsopnoreqny, i sR ‘an[vA Suluutseq O . . : . eee: @s 6 a= . bs 18) SouIl} ¢°Z ST OUIT} SIYY 9B sSoposndi0o ul T a se'l ST ot (1001) ¥S 290 a ea ‘30q ee ee *jnOYySno1Y} MOT SUTvUTEI = ¢0 qh 0'F COR = ““([8req) = “Bd vourny TT CIEE “Sid voutnry oe ee 9°0 61 an LPG Ae alas Re) “Hon 8 G0 oP | ere o'¢ ‘2a As ee) OD » 5 “YPTA 4.1848 07 MOT ATOA [UNTTE STY} UT = oT ST LY OOo os Cea ms aes = 4 *£IDAODAI [IJUN SST 9UIODE OF SoNUT}ZUOD : 10 8g 8°S Gea oS 2 (LOD TES) "Vv WqqRy “UIVSE SESvOIDUL -) qnq skep Z 3xoU 94} Ul GO 0F Sosvarloop 'T sé OF 0’? OEg Mite ee ae god, 8G ” “SUIVUIOI PUB [BULIOU 0} SUIN}OI ? 69'0 0°9 GT 000 (4810) «| queurtredxy a ee ee ,[019389[0G) *[0.19489] urgqy10e'T (9) See eh nn ne ee ee ee *pepn79UuoO—-ill AIAV.L “‘SHIBUIOY “uIq}100T “4B *‘1oyyne pus ‘aseo ‘eturadry jo dA], 218 W. R. Bloor 219 cholesterol, and the least in the lecithin. As a result of the greater increase of cholesterol the ratio eS is generally below cholesterol the normal value. In most of those cases where the ratio is above the normal value it sinks to normal or below as the lipemia continues. The outstanding characteristics of persistent lipemia of whatever origin are then: increases of lecithin and choles- terol along with the fat; and a greater increase of the cholesterol Aectthin ratio below the above that of lecithin resulting in a cholesterol normal. The study of lipemia from the time it begins until it subsides (Tables I and II) brings out in addition the fact that after fat, lecithin is the first lipoid to increase in most cases, and cholesterol the last; which makes it possible to correlate persistent or chronic lipemia with the temporary lipemia produced in normal animals by feeding fat, since in this type of lipemia the sequence of changes in the blood lipoids is also fat, lecithin, cholesterol. However, as pointed out above, there are many exceptions; for example, increases of cholesterol in alimentary lipemia have never been demonstrated in this laboratory, possibly Ege ise the experi- ments have been too short. The Origin of the Fat in Persistent Det —Of the three possible sources, fat of the food, fat stored in the depots, and fat synthesized from other constituents of the food, the last may be eliminated from the discussion since we are entirely ignorant as to its amount or the conditions of its formation. With regard to the other two, both may be shown to be sources . of the fat of lipemia and the importance of each varies with circumstances. In the examples of diabetic lipemia given in Table I, of the first and most striking it was known that the sub- ject had consumed large amounts of fat just before the appearance of the lipemia and frequent previous examinations of his blood made it probable that there was no lipemia up to that time. Allen (2) has furnished evidence to show that in dogs made experimentally diabetic persistent lipemia may be produced by overfeeding with fat. The possibility that the fat in diabetic lipemia may also come from the fat stores cannot be denied but in view of the fact that severe diabetics are thin and have little 740 aa Lipemia stored fat such an origin does not seem important. The older argument (12, 16) that the fat in diabetic lipemia must originate by cellular destruction, since the blood contains not only fat but lecithin and cholesterol, falls before the later observations that these lipoids increase with the fat in alimentary lipemia. The origin of the fat in diabetic lipemia appears to be mainly the fat of the food. With regard to hemorrhagic lipemia, Horiuchi found that it may be produced on a diet practically free from fat and, therefore, the increased fat originated mainly in the fat depots, in which he is borne out by the observations of Boggs and Morris (13) that rabbits so treated become emaciated, and of Sakai (4) that in addition there was a marked fatty liver (which, however, is pro- duced whenever there is excess fat in the blood from any source). The extra fat in the blood in hemorrhagic lipemia may thus origi- nate mainly if not entirely in the fat stores. However, the work of both Horiuchi and Sakai make it plain that food fat when present may be equally effective in producing the lipemia.” The facts available thus point to the conclusion, which perhaps might have been expected, that persistent lipemia may be pro- duced by fat from either the fat depots or the food. The Origin of the Lecithin—Because of its close chemical relationship to fat and the fact that it appears in the blood in alimentary lipemia there is little doubt that it is synthesized from fat and phosphoric acid. The Source of Cholesterol—No stores of cholesterol are known in the organism which would account for the increase in the blood in lipemia. Both cholesterol and lecithin are found in food fats but ordinarily in traces only, so that it seems likely that the cholesterol is also synthesized. The Cause of the Lipemia.—An accumulation of fat in the blood could be brought about in only one way, an inflow greater than the outflow; and it must be decided in each case whether the 2 One endogenous source of the extra fat in hemorrhagic lipemia which might be emphasized in this connection is the marrow of the long bones which in youth is largely a blood-forming organ (red bone marrow) but which in the adult consists mainly of fat. With the stimulus of the severe hemorrhage new blood-forming tissue would be formed, pushing the fatty tissue out into the circulation. W. R. Bloor 221 inflow is abnormally great or the outflow abnormally small. Sakai, who has discussed the lipemia produced by hemorrhage in rabbits, assumed that in these cases it is the outflow which is less than normal and produces evidence to explain why the outflow is diminished. The method which he uses to prove his point—change in surface tension as measured by the stalagmo- metric method in a solution of tributyrin and blood serum—is open to criticism when used for the purpose on the following points: 1. The presence in normal blood of amounts of lipase capable of splitting more than traces of fat in the period of time that fat remains in the blood in alimentary lipemia (8 to 14 hours) has never been satisfactorily demonstrated (19). 2. Tributyrin is so much more easily hydrolyzed than ordinary food fat that its use as a measure of lipase activity is open to question. 3. The assumption of a splitting as shown by changes in surface tension, although possibly correct, cannot safely be accepted without further evidence of hydrolysis—increase of free fatty acids, which Sakai was unable to obtain. Further- more, as pointed out by Sakai himself, his most marked changes were obtained before any lipemia appeared, while when the lipemia was at its height, values obtained were in many cases but little below normal. There was no constant relation between the degree of lipemia and the drop number. But quite aside from the acceptability of his evidence the ques- tion may be raised whether his assumption of a diminished out- flow in hemorrhagic lipemia is correct. Turning to Horiuchi’s tables, which give the best picture of the lipemia, in Table IB on the 9th day of the bleeding the total fatty acids in the plasma (which may be taken as the measure of the lipemia) jumps from 0.32 to 3.07, remains high for 3 days and then falls on the next day to 0.51, the hemoglobin and, therefore, the red blood cor- puscle percentage remaining practically constant. In Tables II B and III B the same sudden rise and fall, but much less striking may be noted. In Table IV after a single large bleeding the plasma fatty acid rises from 0.25 to 0.66, is further increased by a second large bleeding to 1.24, and still further by a feeding of fat (sunflower seeds) to 2.15, but the day following it has fallen 222 Lipemia to a third of this value (0.69) with a change in hemoglobin of only two points. In Rabbit A in Table II above the same sudden falling off, 640 to 350 to 140 mg. on successive days with cor- puscle percentage of 19.5 to 20 to 21 may be noted. In the case of the results from the other rabbits of this table data on this. interesting point are not available. The examples chosen are purposely those in which the rabbits were on a fat-free diet so that the clearness of the results is not interfered with by the pre- sence of food fat, but the same rapid falling off, although not so marked, may be noted in his fat-fed rabbits where the data are sufficiently full. For example in his Table III A on the 4 successive days from the height of the lipemia the figures are 5.19 to 3.05 to 1.30 to 0.73.. In the second part of this table the fall is from 5.35 to 2.25 in 3 days and from 2.25 to 0.38 in the next 3 days. The rapid fall in the blood lipoids does not bear out the assumption that the ability to remove fat from the blood is appreciably diminished in these animals and therefore renders gratuitous the explanation that the blood lipase is diminished by the bleeding. The fact pointed out by Sakai that an ali- mentary lipemia may be produced in rabbits which have been bled and not in normal rabbits, and which is supported by the data in Horiuchi’s Table IV, may be explained by the probability that the fat from the food added to the large amount already in the blood exceeds the quantity which even a normal animal may dispose of. The rabbit normally is not a fat-eating animal and its ability to handle fat, being therefore probably small, may be relatively easily overstepped. In this connection the data on the dog are especially useful. The dog can normally eat and quickly dispose of large amounts of fat and it is therefore signifi- cant that it is not possible to produce an unmistakable lipemia in him even by a hemorrhage by which the corpuscles are re- duced from 45 to 13 per cent as in Table II. Of further signifi- cance is the fact that in another of the dogs in which it was im- possible to produce lipemia by bleeding, when the corpuscle percentage was at its lowest (reduced from 32 to 13 per cent), a feeding of 50 gm. of fat produced a normal reaction, plasma milky at 6 hours after feeding and clear the next day.$ 3’ Unpublished experiment. On W. R. Bloor 223 It appears incorrect, therefore, to assume that the lipemia produced by hemorrhage in rabbits is caused by a decreased outflow of fat from the blood since fat disappears from the blood of these animals with considerable rapidity and probably at not far from the normal rate. The lipemia may be equally well ex- plained as due to the sudden discharge into the blood of quantities of fat larger than the normal mechanism can dispose of at once. Not knowing the normal rate of removal of fat from the blood it is, of course, impossible to exclude the possibility that there may be some slowing of outflow due to inanition of the tissues, etc., but in the lack of further data it is unnecessary to assume that the lipemia of hemorrhage in rabbits is due to anything more than larger influx of fat than the organism can take care of at once. As to the cause of the sudden mobilization nothing definite can be said. There is undoubtedly more or less cellular inanition which might cause a movement of stored food material, and the displacement of fat in the long bones by blood-forming tissue is an interesting possibility. Turning now to the lipemia of diabetes, in the first example given in Table I the high beginning values (which were known to be due to a single intake of foot fat) persist unchanged for 3 days (the apparent increase in value being probably due mainly to concentration of the plasma), in 5 days more are only reduced to about half the highest value and milkiness of the plasma persists for 15 or 16 days longer. In the second example the high values and the milkiness persist for about 17 days and in the last ex- ample for about 14 days. Other examples of the slow disappear- ance of the lipemia in human diabetes are to be found in the literature (20). Allen (21) has shown that in suitably pre- | pared diabetic dogs it is easy to produce a persistent lipemia || while in the normal animal it is difficult or impossible. That lipemia is common in diabetes has been known for over a cen- tury (3) and it is the only disease in which it is at all frequent. Taking all the evidence together the assumption seems justified that in the diabetic the ability to remove fat from the blood is below normal and that herein is to be found the explanation of the lipemia of diabetes. The commonly accepted theory of the cause of diabetes is the lack of a hormone normally found in the ternal secretion of the 224 Lipemia pancreas, and of which the function is the removal of excess sugar from the blood and its utilization or storing in the tissues. When this substance is lacking the excess sugar is lost in the urine. An extension of this theory to postulate the presence in the internal secretion of another hormone of which the function is to aid in the removal of fat from the blood has already been made (21) and has the following in its favor. First, the observa- tion that persistent lipemia has always been commonest among diabetics; second, that lipemia may be readily produced in dogs made diabetic by removal of a suitable proportion of the pan- creas, while its production in normal dogs is practically im- possible; and third, that the blood lipoids in diabetics are almost always above normal, which may be regarded as evidence of diminished power to remove them. On the other hand, the deficiency of the hormone is probably never as great as that of the corresponding carbohydrate hormone since severe diabetics whose tolerance for carbohydrate is very low, do not ordinarily become lipemic, even with relatively large amounts of fat in their diet; in fact, they must depend largely on fat for their sustenance. Also, if it be taken for granted, as seems reasonable, that the appearance of persistent lipemia is equivalent physio- logically to the appearance of sugar in the urine, the lack of the fat hormone in suitable amounts is relatively infrequent. The factor of overwork must probably also be taken into consideration in examining into the cause of diabetic lipemia and possibly of other forms of lipemia. It is a biologic rule that a mechanism when worked within or up to the limit of its powers tends to become stronger, but when worked beyond its powers tends to fail. This principle is recognized in the treatment of diabetes by limiting the carbohydrate intake to an amount within the tolerance of the individual, in the expectation that by so doing the tolerance, or the amount which the individual may consume without waste, may be increased. It is well known that if this amount be exceeded the tolerance of the individual is likely to be lowered. There is very little evidence available to show that such is the case with regard to fat, but it seems probable. The most marked example of diabetic li- pemia presented in this paper was an individual who both before and after the lipemia was able to deal with considerable amounts W. R. Bloor 225 of fat in his diet without lipemia, but when a large amount was taken, the mechanism for utilizing fat to all appearances broke down and did not recover for a considerable time. It appears then that two types of lipemia are represented in the data given in this paper. One in which the ability to remove fat from the blood is little if any affected and in which the cause of the lipemia is to be sought rather in a flooding of the blood with amounts of fat greater than the normal mechanism can take care of at once. In this group are included the hemorrhagic lipemia of rabbits and probably also that of human beings, al- though enough data are not available to determine definitely about the latter. It is characterized by a rapid recovery as soon as the cause (the abnormal inflow of fat) isremoved. In the other type, represented by the lipemia of diabetes, the cause of the lipemia is to be sought primarily in a diminished power to remove fat from the blood, referable to an inadequate supply of the pancreatic hormone, and secondarily to a temporary dis- ablement of the fat burning mechanism due to overwork, which is recovered from rather slowly. The behavior of lecithin and cholesterol in lipemia calls for some comment. It is probable that they have to do with the metabolism of the fat, since whenever the fat of the blood is increased they are also increased, and conversely, when the fat of the blood is low they are also low. In alimentary lipemia the changes follow the changes in the fat within a few hours. In persistent lipemia, although the increase of lecithin and choles- terol follows rather closely (within 2 days) the increase of fat, the values remain high after the fat has decreased to nearly normal and fall to near the normal value only after several days. Because of its close chemical relationship to the fats the change in lecithin calls for little comment, but why cholesterol, whose relationship to fat metabolism is rather remote (formation of esters with the fatty acids), should increase more or less parallel to the fat and lecithin is not readily apparent. Lecithin and cholesterol are antagonistic in several important reactions in the living body. Their opposite effect on tumor growth and possibly also on bodily growth and development in mice has been pointed out by Robert- son and Burnett (22) and their antagonism in the matter of the stability of the red blood cells toward hemolysis seems well THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 1 226 Lipemia established. It is possible that the increase of cholesterol in lipemia, which is in most cases not only equal to but generally lecithin cholesterol below normal), is to be considered a protective mechanism to counteract the effects of the increased lecithin. The writer is, however, of the opinion that this is only one and probably a less important function of cholesterol, and that others, particular in relation to fat metabolism, will become apparent when more i known of the behavior of this mysterious substance in the or ganism. The slower rate of removal of lecithin and cholesterél as compared with fats points to a different mechanism for the removal of the two groups of lipoids. . greater than that of lecithin (the ratio of is generally SUMMARY. The characteristics of persistent lipemia as shown by the evidence so far available are as follows: 1. In lipemia of whatever origin all three blood lipoids (fat, lecithin, and cholesterol) are increased, fat generally showing the greatest ultimate increase, and cholesterol next. 2. There is perceptible, in most cases, a sequence in the appear~ ance and disappearance of the three lipoids, fat being the first to increase, lecithin next, and cholesterol last, while during the clearing up of the lipemia the fat diminishes first and the choles; terol last. High values for lecithin and cholesterol often per- sist for some time after the fat has reached approximately normal values. lecithin cholesterol *” markedly below normal, due to the greater increase of choles- terol over lecithin. 4. The fat which produces the lipemia may be of endogenous or exogenous origin or both, but the phenomena of the lipemia are the same in either case. 5. The cause of the lipemia being regarded as a disturbance of the balance between inflow and outflow of fat in the blood, the immediate causative factor in the hemorrhagic lipemia is probably an abnormally large inflow of fat while in diabetic lipemia it is an abnormally slow outflow. 3. In most instanees the values for the ratio e W. R. Bloor 227 BIBLIOGRAPHY. . Bloor, W. R., J. Biol. Chem., 1916, xxvi, 417. . Allen, F.M., Am. J. Med. Sc., 1919, elviii, 307. ° . Fischer, B., Virchows Arch. path. Anat., 1903, clxxii, 30, 218. . Sakai, S., Biochem. Z., 1914, Ixii, 387. (These findings have been cor- roborated by several experiments in this laboratory.) . Bang, I., Biochem. Z., 1918, xci, 104. . Terroine, E-F., J. physiol. et path. gén., 1914, xvi, 386. . Bloor, W. R., J. Biol. Chem., 1915, xxiii, 317.- . Bloor, W. R., J. Biol. Chem., 1916, xxiv, 447. . Iscovesco, H., Compt. rend. Soc. biol., 1912, lxxii, 920. . Bang, I., Biochem. Z., 1918, xci, 111. . Hueck, W., and Wacker, L., Biochem. Z., 1919, c, 84. . Klemperer, G., and Umber, H., Z. klin. Med., 1908, Ixv, 340. . Boggs, T. R., and Morris, R.S., J. Exp. Med., 1909, xi, 553. . Horiuchi, Y., J. Biol. Chem., 1920, xliv, 363. j ea A cEtachon, Z.; 1921, exv, 63. . Klemperer, G., and Umber, H., Z. klin. Med., 1907, Ixi, 145. . Imrie, C. G., J. Biol. Chem., 1915, xx, 87. . Bloor, W. R., J. Biol. Chem., 1920-21, xlv, 171. . Thiele, F. H., Biochem. J., 1913, vii, 275. . Gumprecht, Deutsch. med. Woch., 1894, xx, 756. . Allen, F.M., Am. J. Med. Sc., 1917, cliii, 313. - . Robertson, T. B., and Burnett, T. C., J. Exp. Med., 1913, xvii, 344. Robertson, T. B., J. Biol. Chem., 1916, xxv, 635, 647. STUDIES OF LIVER FUNCTION. BENZOATE ADMINISTRATION AND HIPPURIC ACID SYNTHESIS. By G. D. DELPRAT anp G. H. WHIPPLE. (From The George Williams Hooper Foundation for Medical Research, Uni- versity of California Medical School, San Francisco.) (Received for publication, September 27, 1921.) Much work has been done in an effort to understand the com- plex function of the liver and considerable information of value has been obtained. It is fair to say, however, that the estimation of functional capacity of the liver cannot as yet be made with any convincing accuracy. We have reason to believe that the liver has a very large factor of safety—that it can tolerate an extensive " injury and yet carry on its essential body functions. Any satis- factory functional test must include, therefore, some factor of strain or load which can measure the upper limits as well as the lower limits of liver function. It is certainly possible that some of the many functions of the liver may. be seriously impaired while others, perhaps of more fundamental significance, are not seriously disturbed. This may apply to acute as well as chronic disease conditions. Probably the liver function test which is of most value in phys- iological experiments is given by the use of phenoltetrachlor- phthalein as described by Whipple, Peightal, and Clark (1). This drug is given intravenously and eliminated promptly in the bile. It can readily be recovered from the feces and accur- ately estimated. This functional test shows a great impair- ment after extensive liver injury by chloroform or phosphorus, but a serious objection is at once obvious if any abnormalities of bile secretion are present. With complete biliary obstruction obviously the test is of little value and with chronic inflammation of the bile passages, we find a marked lowering in output of the phenoltetrachlorphthalein—for example in long standing biliary fistulas. For clinical use, therefore, this test has very serious 229 230 Studies of Liver Function limitations. We cannot rely in clinical diagnosis on any liver test which postulates a normal bile flow in normal bile passages. The ideal liver functional test would consist in the introduction intravenously of some non-toxic substance which would test by a synthetic demand the functional reserve of the liver (normal or abnormal). The product formed by this synthesis should be obtainable from the blood rather than from the urine, as renal abnormality would not confuse the issue. This ideal functional test is not at hand but it appears to be a possibility and all attempts along these lines are worth while and should give information of value. We submit these experiments, which were undertaken as a part of this experimental program to study liver function. Recognizing the importance of the liver in many synthetic endogenous processes, we felt it would be desirable to extend the study of hippuric acid synthesis under experimental conditions. It has been claimed by some that hippuric acid synthesis is a function of the kidney, by others a part of liver activity—in a * word, the question is open to debate. We feel that our experi- ments furnish new data which will aid in the final solution of this complex physiological equation. A critical review of the literature touching the influence of benzoic acid and benzoates upon the general metabolism of the experimental animal reveals the fact that the action of this drug is still far from being completely understood. The earlier investi- gators and particularly Schmiedeburg sought to establish the synthesis of hippuric acid from benzoic acid as a function of kidney activity, using principally what was considered at that time (1876) a very reliable method of transfusion of excised organs. These experiments have stood unquestioned until the last few years, when considerable evidence was adduced to show that the kidney was not alone responsible for hippuric acid production. The work of Kingsbury and Bell (2) with nephrectomized dogs and dogs in which an experimental nephritis had been produced shows little change in the hippuric acid synthesis from the pre- viously normal condition and that in the dog at least the kidney is not the site of the reaction. In rabbits the evidence still appears to be conflicting. On the other hand, the known réle of the liver in certain processes of metabolism and the influence of ben- zoic acid on metabolism as shown by our. experiments suggests G. D. Delprat and G. H. Whipple 231 that the liver might play a definite part in the conjugation of benzoic acid with glycocoll. An analysis of the parenchyma of various organs, after the injection of benzoic acid, shows that there is a higher content of hippuric acid in the liver than in other organs (Kingsbury and Bell). Further evidence that the liver may be involved in the production of hippuric acid is offered by Lackner, Levinson, and Morse (3) who show a definite de- crease of the hippuric acid elimination after a liver necrosis, which had been produced by hydrazine sulfate. Dogs were used and benzoate (0.5 to 2.0 gm.) was given by mouth. The actual mechanism whereby the benzoic acid radicle is neutralized in the body is not understood. The fact that the amount of glycocoll eliminated as hippuric acid is much larger than the amount of glycocoll in the food has led to various explana- tions which need only to be mentioned in passing. Parker and Lusk (4) suggested’ the possibility of a glycocoll reserve in the body which can be ‘‘ washed out” by successive doses of benzoate. The massive breaking down of body proteins is advocated by Ringer (5) and others. The synthesis of glycocoll from simple substances and simple amino-acids may occur (Epstein and Book- man, 6). Lewis maintains that products of metabolism, which might otherwise go to the formation of urea, may be diverted to the production of glycocoll (7) and this hypothesis is confirmed by the work of McCollum and Hoagland (8) who find a decrease in the urea nitrogen elimination of pigs on a carbohydrate diet following the administration of benzoates. Lastly the hypoth- esis of Umber (9), objected to by Abderhalden (10), that benzoic acid abstracts glycocoll from the globulin molecule of the tissues, thereby converting it to an albumin molecule; and that the action of benzoic acid in the body is merely to change the normal al- bumin-globulin ratio. That so many theories exist is evidence that the process is either not understood completely, or that no one explanation is sufficient. Since all hypotheses of internal metabolism are deduced from the study of the urinary nitrogen, and since the conflicting theories advanced are based on conflicting results of urinary analysis, an attempt will be made to show that dzffer- ent results may be referable to differences in animals used and to differences in dosage. 232 Studies of Liver Function Methods. For the following experiments female dogs were kept in standard metab- olism cages and the urine collected at a fixed hour each day, The 24 hour specimen included the cage collection, cage washings, bladder urine, and bladder washings made up to a standard dilution. The dogs were fasted for 3 or 4 days previous to the experiment to allow the urinary nitrogen excretion to reach a constant base line and food was withheld during the entire experiment. Water was supplied in the cage at all times and it was noted that after the injection of sodium benzoate the dogs showed a much greater thirst. The collected urine was subjected to the following analy- sis: Total urinary nitrogen by the Kjeldahl method; urea and ammonia nitrogen by Marshall’s method; free benzoic acid by the method of Raiziss and Dubin (11); hippuric acid by the method of Folin and Flanders (12); all determinations were made in duplicate with suitable controls. A slight modification was made in both the method for the determination of the free benzoic acid and in the method for the determination of the hippuric acid. Determination of Free Benzoic Acid.—100 cc. of urine are acidified by the addition of 1 ec. of strong nitric acid and 50 gm. of ammonium sulfate are added. To this are added, in a 250 ec. Erlenmeyer flask, 50 ee. of freshly distilled chloroform; the whole is then shaken thoroughly for some time and allowed to stand from3 to24 hours. At the end of this time, the chloroform is separated and is transferred into a second Erlenmeyer flask. The res- idue is washed into the first flask with 25 ec. of fresh chloroform and treated in the same way as the first extraction. The extract is washed by shaking with 100 ec. of saturated common salt solution and allowed to stand over night. On the next day the chloroform is separated in a clean dry funnel and is titrated with a standardized tenth normal alcoholic solution of sodium ethylate (phenolphthalein indicator). It was found convenient to keep the various chloroform fractions separate for the reason that, as was often found, the urine may be free from benzoic acid, a fact which can be noted on the first titration and further extraction with the chloroform rendered unnecessary. In each set of benzoic acid determinations (two daily for each dog) a complete duplicate was carried out with distilled water to insure no acid from the first acidification of the urine being re- tained in the extraction, and to show that the washing process had been sufficient. Determination of Hippuric Acid.—100 ce. of urine are rendered alkaline with 10 ce. of 5 per cent sodium hydroxide and are evaporated to dryness on a water bath. The residue is dissolved in 25 ec. of water, and is then de- canted into a Kjeldahl flask, and 25 cc. of strong nitric acid are poured into the same flask. A few crystals of copper sulfate are then added and the Kjeldahl flask is fitted with a long reflux condenser; an apparatus may be set up so that four or five determinations may be made at the same time with an extra flask for a control. The material in the flask is now boiled over a small flame for several (3 or 4) hours until the solution is a clear blue. When this stage is reached the condenser is allowed to cool and the inside G. D. Delprat and G. H. Whipple 233 is washed down twice with 25 cc. of distilled water. To the material in the flask, the volume of which is now about 100 cc., 50 cc. of freshly distilled chloroform are added and the whole is vigorously agitated for several min- utes and is then allowed to stand for several hours. The chloroform is sepa- rated in the same way as in the method for the estimation of free benzoic acid and is thus similarly washed and titrated. As in the former method, it is advantageous to keep the extracts separate and to continue with ex- tracting repeatedly until all the benzoic acid has been removed. It is hardly necessary to outline the rationale of the procedure at this time— reference has been made to the sources in which these methods were origi- nally described, where additional and further information will be found. Raiziss and Dubin (11) advise the use of toluene for the ben- zoic acid extraction, but in this work chloroform was found more convenient and it was determined experimentally that under the same conditions both solvents extract approximately the same amount of acid in the same time. It will be seen that the extraction is made from a solution which is very strongly acid, and that if for some reason the chloroform is impure some of the nitric acid will be retained in the chloroform and the titration with the sodium ethylate will be rendered ex- tremely inaccurate. It was found that small quantities of alco- hol in the chloroform caused this error, and on this account it was necessary to wash the chloroform repeatedly with salt solu- tion. When it was desired to reutilize chloroform which had been used in a previous extraction, it was rendered strongly alkaline and washed repeatedly. Since the used chloroform contained phenolphthalein, the washing in an alkaline solution was continued until the pink color had disappeared, and the chloroform was then considered ready for the first distillation. The distillation was carried out in the usual way, in a water bath, and the distillate washed by allowing it to stand for 24 hours over saturated salt solution, after which it was distilled a sec- ond time. Even the purest commercially obtainable chloroform was treated in this way, and it was found that such treatment was essential to avoid large and unexpected errors. With the methods outlined above, it was possible in control experiments to recover from an aqueous solution of sodium ben- zoate an average of 99.8 per cent with a fluctuation of 10 per cent between extremes. The hippuric acid shows a recovery of 92.2 per cent with a fluctuation of 7 per cent between extremes. 234 Studies of Liver Function It was found, in control experiments, that when bile, blood, or fecal material was added to the urine, and the urine was treated in the usual way for the determination of hippuric acid, that substances of an acid nature appeared in the chloroform, which were capable of neutralizing the sodium ethylate in the titration. The chemical nature of these substances was not satisfactorily determined. If the urine collected under the cage was found TABLE I. The Recovery of Hippuric Acid Before and After Anesthesia. Dog 20-81. Brown spaniel, mongrel, adult. , DOneeee, Date. [7 mere ee Hippuric acid. pie etal Weight. Per nia N. 4 Total. pound: Jan. 24 || Dog isolated and fasting begun. 1921 gm. gm. gm. gm. per cent gm. gm. lbs. Jan. 26 0) 0 0.470 | 0.485 — 1.970 | 2.408 | 24.5 te | 0) 0) 0.425 | 0.440 — | 2.040 | 2.632 | 24.0 mbes 0 0 0.510 | 0.510 — 1.860 | 2.520 |} 23.5 29 0 0 0.542 | 0.510 = 1.880 | 2.380 | 23.3 Bl +f 0.174 | 4.170 | 4.070 | -82.5 | 2.550 | 3.690 | 238.0 TE vail 4 0.176 | 4.850 | 5.140 | 102.0 | 2.340 | 3.660} 22.8 Feb. 1 f 0.178 —_ — — | 2.460 | 3.860} 22.5 eae 1D 0 0 — — a 2.460 | 3.470 21.8 Feb. 2 Chloroform anesthesia—60 minutes. Feb. 3 4 0.187 = = = — | 5.040 |) 26 ore 4 0.190 | 5.140 | 4.400 | 93.2 — | 5.150] 21.0 mS 4 0.200 | 4.985 | 4.700 | 98.6 | 3.030 | 4.200 | 20.0 to be contaminated by blood or bile or if the fecal contamination was in the form of a diarrhea which could not readily be filtered out, the urine sample of that day was discarded for obvious reasons. Except where specifically stated to the contrary, the benzoate was administered in the form of sodium benzoate, and injected into the jugular vein of the dog in a 5 per cent aqueous solution at the rate of about 20 cc. per minute. In the experiments marked “by mouth” the drug was given through a stomach tube, which was washed out with water. G. D. Delprat and G. H. Whipple 235 To determine the effect of chloroform anesthesia upon the syn- thesis of hippuric acid several experiments were conducted, of which Tables I to III inclusive may be taken as typical examples. It has been shown by Davis and Whipple (13) that after 3 days fasting, 60 minutes of light chloroform anesthesia will produce a central necrosis involving one-half of the liver parenchyma in TABLE Il. The Recovery of Hippuric Acid in the First 5 Hours after Benzoate Injection Intravenously. Dog 20-25. Fox terrier, adult. Date. Benne Time. Hippuric acid. Weight. 1920 gm. hrs. | gm. gm. per cent lbs. Oct. 30 3 1-5 1.940 1.750 55.4 22.0 Nov. 3 3 1-5 1.520 1.700 48.0 21.0 Ks 4 3 1-5 1.820 1.860 Soo 215 Gig 5 3 1-5 1.895 2.070 59.7 PALS ae chy = 1-5 1.615 1.700 50.0 21.0 lee). § 3 1-5 2.340 2.080 66.5 22.0 Soa se 3 1-5 2.380 2.520 74.0 22.0 Dec. 1 Chloroform anesthesia—50 minutes. Dec. 2 3 1-5 1.265 1.265 38.2 20.5 a 4 3 1-5 1235 1.270 37.6 19.5 s 6 3 1-5 0.925 0.945 27.9 19.0 Dec. 21 Dog dead from distemper—recovery from chloroform poisoning. Diet of bread and milk throughout experiment except on the 2 days pre- ceding chloroform anesthesia. most cases, usually followed by recovery and liver regeneration. We must refer to the paper of Davis and Whipple (13) for the data establishing the constancy of liver injury under uniform conditions. From the examination of the percentage amounts of hippuric acid recovered (Table I), it will be seen that the recovery on 2 successive days average 92.2 per cent. After 60 minutes of chloro- form anesthesia, the average recovery on 2 successive days is 95.5 per cent per 24 hours. The total recovery, therefore, is 236 Studies of Liver Function TABLE II-A. The Recovery of Hippuric Acid in the First 5 Hours after Benzoate Injection. Fatal Liver Injury. Dog 20-21. Airedale, adult. Date. Bersons Time Hippuric acid. Weight 1920 gm. hrs. gm. gm. per cent lbs Nov.6 3 1-5 2.150 | 2.00 63.1 PALS: Nov. 24 Chloroform anesthesia—60 minutes. Nov. 25 3 | 15 | 110 | — | a1 | as Nov. 25 Dog died, acute chloroform poisoning—extreme liver necrosis. not influenced by severe liver injury. From this and other ex- periments of a similar nature, it may be concluded that extensive liver necrosis due to chloroform has very little effect upon the total amount of hippuric acid synthesized during the whole 24 hours. The possibility suggested itself that there might be a delay in this synthetic reaction. With this in mind, the 24 hour period TABLE III. The Recovery of Hippuric Acid from the 5th to the 24th Hour after Benzoate Injection Intravenously. Dog 20-17. Small black mongrel, adult. Date. Benzonte Time. ' Hippuric acid. Weight. 1920 gm. hrs. gm. gm, per cent lbs. Dec. 23 3 |. 5-24 DeO25 2.180 63.70 21.8 ee. 3 5-24 2.010 — 60.60 22-9 ae BOA 3 5-24 1.980 1.690 55.20 22.0 Dec. 29 Chloroform anesthesia—60 minutes. Dee. 31 3 5-24 2.400 (1.850) 64.05 2120 1921 ayia 3 5-24 2.480 2.380 73.35 20.0 s 2 3 5-24 2.010 _— 60.60 20.0 oY 3 3 5-24 | 2.080 2.090 62.80 19.0 Complete recovery. G. D. Delprat and G. H. Whipple 237 was arbitrarily split so that the first collection was made 5 hours after the injection of benzoate and the second collection com- prised the specimen of the subsequent 19 hours. From an examination of Table IT, it will be seen that the amount of hippuric acid synthesized and excreted during the first 5 hours after the injection of sodium benzoate averages for a period of 7 days 58.4 per cent of the amount injected; while after 50 min- utes of chloroform anesthesia the average for a 3 day period drops to 34.5 per cent. In this experiment, after a period of 2 days fast, chloroform anesthesia was given for 50 minutes. Except for this interval the dog was fed bread and milk. The dog died 20 days later from distemper after recovery from the chloroform poisoning. Similarly in Table IJ—A, the recovery during the first 5 hours after the injection of sodium benzoate in the control period aver- aged 63.1 per cent, while after 60 minutes of chloroform anes- thesia it averaged 33.1 per cent. In this experiment, although the bread and milk on which this dog was fed was withheld for only 2 days previous to the chloroform anesthesia, an extreme liver necrosis was produced and the dog died the following day. His- tological examination of liver sections shows practically a com- plete liver cell destruction with only a few small groups of liver cells surviving about the portal tissues. A liver necrosis was produced in a dog (Table III) by allowing the dog to fast for 2 days and then giving 60 minutes of chloro- form anesthesia. Although the lesions thus produced may have comprised one-half or more of the liver parenchyma, the dog recovered from the effect of the intoxication. The effect on the hippuric acid output during the interval between the 5th and the 24th hour is much less marked than is seen in Table II. Before the period of anesthesia in an average of 3 days’ determinations, the dog was excreting 59.8 per cent of the dose between the 5th and the 24th hour. After the period of anesthesia, the dog ex- creted 65.2 per cent during the same time. The dog was fed bread and milk except on the 2 days preceding the chloroform anesthesia. On the basis of the tabulated and other experiments it would seem proper to conclude that a liver necrosis involving more than one-half of the liver parenchyma produces a definite delay in the 238 Studies of Liver Function hippuric acid conjugation and excretion, so that during the first part of the 24 hours the excretion is relatively less than during | the later part of the 24 hours. The total amount of hippuric acid synthesized seems to be little altered if the whole 24 hour period is considered. It will be a matter of dispute whether on the basis of these experiments a further deduction can be made that the synthesis of hippuric acid is a function of the liver. The absence of demonstrable lesions or pathological conditions else- where in the body following chloroform anesthesia make it seem probable that the lesions in the liver are the cause of these changes noted in the excretion of the hippuric acid. Control experiments show that dogs injured by chloroform do not present an impair- ment of renal function to explain the delay in excretion of hip- puric acid. In these experiments it will be seen that the amount of hip- puric acid recovered approaches 100 per cent of the amount in- jected. This, however, was not a constant finding. Dogs of the same weight and under the same conditions, show considerable variation in the amount of hippuric acid eliminated. No cor- relation could be established between the size of the dog, the size of the dose, and the amount eliminated. In some cases the amount recovered was only 70 per cent of the dose administered, and yet no free benzoic acid was recovered from the urine. These abnormalities deserve further study. The Rise in Urinary Nitrogen Following Sodium Benzoate Injection. A number of experiments (nine) were carried out to observe the effect of sodium benzoate in -varying doses on the urinary nitrogen, and as the results are very similar, only a few typical examples need be given. Table IV shows the typical rise in urinary nitrogen following the injection of 3.5 gm. of sodium ben- zoate in a 5 per cent solution into the jugular vein of a 15 pound dog. In this case the total urinary nitrogen excreted averages 1.596 before the injection of the benzoate, and rises to an average of 2.968 gm. after the injection. The greatest part of this rise (excluding the fraction contained in the hippuric acid) is made up of urea, which rises to an average of 2.008 gm. from a normal average of 1.005 gm. The ammonia shows a definite rise over the normal period. : | G. D. Delprat and G. H. Whipple 239 TABLE IV. The Rise in Ammonia, Urea, and Total Urinary Nitrogen Following Benzoate Intravenous Injection. Dog 20-96. Small black mongrel, senile. | Benzoate intravenously. | | Date. 2S SUERTE Tf Hippuric acid. Total. es. 1920 eee | oa | om. lao cenit ge oe ne June 24 0 0 | 0.236) 0.286) — 1.085} 0.185) 1.568) 15.00 fel Oh 0 0 | 0.498) 0.398) — | 0.995) 0.106) 1.652) 14.75 S26 0 0 | 0.286; — es 0.935) 0.184) 1.568] 14.50 7 f 3.5 | 0.185] 3.980) 3.680) 95.2 | 1.780} 0.268) 2.688] 14.25 4.070 June 28 3.5 | 0.188} 4.070) 3.750) 96.5 | 1.880] 0.204) 2.884! 14.00 sear 2G 3.5 | 0.191) 3.640) 3.300) 94.0 | 2.350] 0.224) 3.332) 13.75 = 30 0 0 — — — | 0.558) 0.157) 1.064] 13.50 TABLE V. The Effect of Gradually Increased Doses of Benzoate on the Urinary Nitrogen. Dog 20-77. Small fox terrier, brown, adult. ‘ Bensoete: Date. Lupita Hippuric acid. ae — Weight. Total seas ; 1920 gm. gm. gm. gm. |per cent} gm. gm. gm. lbs. Apr. 3 0 QO | 0.045) — — 0.789) 0.282) 1.663) 11.00 od 4 0 eo 10) — — pied 0.590} 1.624! 10.75 ae 5 0 QO | 0.066) — — | 0.850) 0.269) 1.512) 10.50 oe 6 0 QO | 0.117) — — | 0.810) 0.380} 1.456) 10.00 Ti 0 0 | 0.088) — — | 0.950) 0.095} 1.512) 10.00 ig 8 1.25 | 0.128] 1.380) 1.305] 94.6 | 1.120) 0.176} 1.680) 9.75 = 9 1.50 | 0.158} 1.230) 1.116] 70.0 | 1.410) 0.213} 2.016) 9.50 Mae ei 2.00 | 0.222) 1.845) 1.775 .6 | 1.170} 0.240) 2.167} 9.00 SS ik 0 0 |/0:279) — — | 0.950} 0.112} 1.512} 8.50 sD 0 0). } 0.235), — — | 0.660) 0.045) 1.036) 9.50 PAY 0 0 | 0.252; — — | 0.760) 0.067} 1.148) 9.50 sae aA | 0 OQ | 0.244, — — | 0.747) 0.101} 1.2382) 9.25 ee to 0 Or |s05219|, -— — | 0.803} 0.112} 1.204) 9.25 ot ee -3 2.50 | 0.278] 2.100} 1.800} 67.5 | 1.422) 0.224) 2.142) 9.25 4 4. 3.00 | 0.334) 2.360) 2.1380) 64.2 | 1.544) 0.246) 2.240) 9.00 ceed oO 4.00 | 0.445, Lethal dose. 240 Studies of Liver Function Table V shows the effect of gradually increasing the dose of the sodium benzoate. As the dose is increased from 1.25 to 4 em., it will be seen that there is a progressive increase in the total urinary nitrogen. The increase in the urea nitrogen corresponds somewhat to the increase in total urinary nitrogen but the rela- tion of the percentage of urea nitrogen to the total urinary nitro- gen is very variable. In this experiment doses varying from 0.128 TABLE VI. Benzoate Given by Mouth and Intravenously. Dog 20-101. Airedale, pup, weight 20 pounds. , Benzoate. Rae Date. | re Hippuric acid. yee monia Fafa Remarks. Total. unt N. 1920 gm. gm gm. gm pai” gm. gm. gm. May 8 0 0 {0.295} —]| — /1.530/0.153/2.380 ee 9 0 0 |0.3888] — | — }1.395/0.140)2.296 oO 0 0 0.838] — | — |1.260)0.134)2.212 eeomlel: 0 0 — | —] — }1.345)0.112/2.128 Seem 0 0 |0.3846] — | — |1.275)0.129)2.184 nis 3.0 | 0.113)/3.320/2.610) 79.0)1.715|0.258|2.800) Intravenous. “14 3.5 | 0.138)/3.880)3.420) 88.0/1.760)0.258|2 .940 e SRELD 4.0 | 0.152)4.425)3.830) 87.0)1.955)0.244/3. 136 ue to NGG} 0 0 | 1.155)|0.145/2.156 ays 0 0 — }| -—]| — ]1.382)0.168/2.100 eels 3.0 | 0.113)2.500)2.150 65. 0/1. 125)0.134/2.016 By mouth. sf 8 9 3.5 | 0.135)3.430\38.100) 79.8)1.170\0.370\2.268) “ 8 <0 4.0 | 0.154/4.023/3. 750) 85.5/1.585)0.291/2.660) <“ ” APAl 0 0 |0.512) — — |l 3800/0. 179)2.080 to 0.222 gm. per pound dog fail to produce a definite rise in am- monia, while larger doses seem definitely to increase the ammonia. During the first part of the experiment, the variation in the am- monia is within the physiological variation, but after the 19th, a lower normal base line is maintained which is definitely exceeded after the injection of larger doses. A lethal dose was reached with 0.445 gm. per pound body weight. To determine whether the method of administration of the drug exerts any influence on its metabolism through possible variations in rate of absorption, injections of the same amount of G. D. Delprat and G. H. Whipple 241 sodium benzoate were given intravenously and later by mouth to a dog under as nearly as possible identical conditions. The results of this experiment may be seen in Table VI. In this experiment the usual marked rise in total urinary nitro- gen, urea nitrogen, and ammonia nitrogen follow the intravenous injection. The first dose by mouth was partly vomited, which accounts for the low value of the recovered hippuric acid. On the 19th and 20th, the full dose was retained. It will be seen that the response to the ingestion is very much less marked than TABLE VII. Composite Table Showing the Total Urinary Nitrogen with Doses of Sodium Benzoate Varying from 0.079 to 0.445 Gm. per Pound. ee Total urinary nitrogen after | Dose of sodium benzoate Mar Noel nittasen, injection of benzoate. given per pound. Dog 3 days’ ee eee eee Orin average. | Istday | 2nd day | 8rd day | Ist day | 2nd day | 8rd day gm. gm. gm. lbs. 20-101 | 2.306 | 2.352 | 2.464 | 2.3872 | 0.079 |} 0.090 | 0.102 | 20.00 20-101 |- 2.240 | 2.800 | 2.940 | 3.136 | 0.113 | 0.188 | 0.152 | 20.00 20-77 1.553 1.680 | 2.016 | 2.167 | 0.128 | 0.158 | 0.222 9.70 18-88 2.760- | 4.440 | 3.860 | 3.330 | 0.146 | 0.150 | 0.150 | 19.00 20-81 2.485 | 3.690 | 3.660 | 3.860 | 0.174 | 0.176 | 0.178 | 22.00 20-96 1.596 | 2.668 | 2.884 | 3.332 0.185 | 0.188 | 0.191 14.30 20-20 1.344 | 2.016 | 2.240 |-2.240 | 0.200 | 0.200 | 0.200 12.10 18-388 2.760 | 3.466 | 3.819 | 2.979 | 0.216 | 0.216 | 0.220 19.00 20-20 1.222 | 3.024 |} 3.416 | 3.248 | 0.286 | 0.292 | 0.298 12.10 20-77 1.155 2.142 | 2.240 | Lethal.|} 0.278 | 0.334.) 0.445 9.70 the response to the intravenous injection, the urea and total urinary nitrogen being almost within the limits of physiological variability and the ammonia alone showing a very slight rise. Table VII shows that below a dosage of 0.140 gm. of sodium benzoate per pound of body weight, the change in the total urinary nitrogen is not significant. Doses larger than this, however, pro- duce a definite and constant rise in urea, ammonia, and total urin- ary nitrogen. The greatest responses noted were between the doses of 0.180 and 0.200 gm. per pound. In three dogs with increasing dosage the rise in total urinary nitrogen was progressive to a certain point after which there was a less marked rise; this is shown in Dogs 20-81 and 18-38, and to a lesser degree in Dog 242 Studies of Liver Function 20-20. A sufficient number of observations are not available to determine whether this is a constant finding. The fasting dog presents a constant base line of nitrogen ex- cretion and the sudden rise following the injection of sodium ben- TABLE VIII. The Rise of Urinary Nitrogen Following Benzoate Injection Prevented by Administration of Dextrin. Dog 20-20. Fox terrier, adult. Benzoate. — Date. Dex: aero Hippuric acid. te ame, Teva Weight. Total. gaunt: 1920 gm gm. gm gm gm Lap gm gm gm Ibs May 8 0 0 0 (0.210; — |} — |} 0.715) 0.075} 1.288} 13.00 Loe9 0 0 0 |0.236) — | — | 0.890} 0.067} 1.232] 13.00 HK) 0 0 0 0.270; — | — | 1.240} 0.112) 1.624) 12.75 eo lit 0 0 0 10.442) — — | 0.990) 0.177) 1.512) 12.50 i 2 0 0 0 |0.362; — | — | 0.920) 0.190} 1.456) 12.50 May 13 0 | 3.5 | 0.286/3.510/3.180} 88.5) 2.260) 0.277] 3.024] 12.25 “14 | 0 | 3.5 | 0.292/3.720/8.390| 96.0] 2.390] 0.274| 3.416] 12.00 “15 | 0 | 3.5 | 0.298|3.540/3.210 90.5] 2.240) 0.230] 3.248) 11.75 May 16 | 0 | 0 | 0 {0.440 —| — | 1.540] 0.151] 2.172] 11.25 June 3 0 0 0 {0.218} — | — |} 0.985] 0.089] 1.148} 12.25 oan AA: 0 0 0 |0.236; — | — } 0.975) 0.106) 1.148} 12.00 Tt 59 0 0 0 |0.210; — | — |} 0.965} 0.095} 1.262} 11.75 a ae Cy 0 0 0 |0.270; — | — |} 0.965) 0.078) 1.204) 11.50 eee 0 0 0 |0.285} — | — | 1.020] 0.095) 1.260) 12.00 June 8 30 0 0 |0.220; — | — | 0.780} 0.095} 1.064) 12.00 vemos) 30 0 _0 |0.255) — | — | 0.492) 0.218) 1.148} 12.00 ee 0 30 0 0 10.264 — | — | 0.589] 0.078) 0.800) 11.75 June 11 380 | 3.5 | 0.806/3.440)3.350) 86.5} 0.962) 0.185) 1.344) 11.50 oe oli 30 3.5 | 0.306/3.640/3.390) 87.5) 0.995) 0.525) 1.624) 11.50 aS 30 3.5 | 0.313/4.040/3.770| 97.0) 0.975) 0.224) 1.652) 11.25 zoate must clearly be due to the breaking down of body proteins. This breaking down of body proteins was prevented by Epstein and Bookman (6) by the stimulating administration of dextrin, an observation we were able to confirm, as indicated in Table VIII. G. D. Delprat and G. H. Whipple 243 A dog fasted for 7 days, during the last 4 of which his normal ammonia, urea, and total urinary nitrogen were determined. A dose of 3.5 gm. of sodium benzoate was injected which practi- cally doubled the urea and total urinary nitrogen. After a period of rest, the same dog again fasted 7 days, during the last 4 of which the same determinations were made and a new baseline established. 30 gm. of dextrin were given by mouth, which caused a slight drop in the nitrogen excretion. This was accompanied after 3 days by the same dose of benzoate which had previously caused such a marked rise in nitrogen elimination and little change in the nitrogen output occurred, showing that the dextrin had suc- ceeded in preventing the reaction observed in the control period. The Effect of Sodium Benzoate Injections on the Blood Serum Proteins. Several experiments were carried out to determine the effect of benzoate injections on the blood proteins. Dog 20-101 is a characteristic example. Before . After injection. injection. per cent per cent LINES? Rs Genes 6 ee rn 2.15 | 2.15 NBME IUCR re PI NN en a ss aia Rak eens 3.58 3.71 MIP Pes Rt Noe Ts ss sss ao oduslemre 2 1.69 1.42 FET ADLPET TRIE ce A i a 5.22 5.138 In this experiment, 4 gm. of benzoate were injected and the second determination was made 2 hours after the injection into a 20 pound dog. Determinations were made by the refractometric method of Robertson (14). If we assume the blood plasma volume of the dog to be 5 per cent of his body weight, the serum globulin 1.7 per cent of his blood plasma volume, and the glycocoll 3.5 per cent of the serum globulin, the dog should have 0.270 gm. of glycocoll combined in the form of serum globulins. On the other hand, 4 gm. of sodium benzoate combine with 1.85 gm. of glycocoll to form hippuric acid, which is actually seven times the amount present in all the blood serum globulins. It is interesting to note that the globulin- albumin ratio remains the same while this demand for glycocoll 244 Studies of Liver Function is being met. The injection of benzoate does not influence the non-protein nitrogen circulating in the blood. DISCUSSION. Our experiments show that a very severe and even fatal liver necrosis due to chloroform may not diminish the total 24 hour amount of synthesized hippuric acid following administration of a unit dose of sodium benzoate. This may suggest that the liver has nothing to do with hippuric acid synthesis, but we believe this con- clusion is not justified. Further study shows (Table II) that under conditions of liver injury, there is a delay in the synthesis and excretion of hippuric acid in the 5 hour period following the ad- ministration of the sodium benzoate. There is no decrease in renal function during this period as indicated by phthalein func- tional tests. The liver inquiry, therefore, delays this reaction which is so prompt in the normal dog. Our belief is that the liver is much concerned in the normal synthesis of hippuric acid in the body but after injury this function may be taken over in large measure by other cell protoplasm. This indicates too that other organs and tissues may take part in this synthetic reaction even under normal conditions. Our experiments show a definite rise in ammonia, urea, and total urinary nitrogen following the injection of sodium benzoate into the circulation of the dog, whenever a certain dosage is ex- ceeded. If these conditions be applicable to other animals, it would seem possible that the failure of some observers to note a rise in urinary nitrogen after administration of benzoate may be explained by the use of small doses. The rise in urea nitrogen is not in agreement with.the observations of Ringer, and McCollum and Hoagland, who found that large doses of benzoate decrease the urea nitrogen excreted. It must be remembered, however, that McCollum and Hoagland worked with pigs, and that the caloric requirements of the animals were supplied, whereas in our experiments, the dogs were fasting. In both instances, however, the nitrogen of the urine was wholly endogenous. Our doses per unit body weight exceed those of McCollum and Hoagland and this may explain the differences. How may we explain this rise in urinary nitrogen following the intravenous injection of sodium benzoate? Perhaps a tentative G. D. Delprat and G. H. Whipple —-245 explanation may be advanced as follows: When the drug is given intravenously, the demand for conjugation is very acute and lacking the available glycocoll, there is a breakdown of body protein to supply a part of this emergency requirement. We may imagine that certain elements (glycocoll) are removed from the large protein molecule which as a result disintegrates, at least in part. The less acute demand made by administration of the drug by mouth can be met by the usual body mechanism (glycocoll synthesis) unless very large doses are given and there isnoemergency breakdown of body protein. The death of the an- imal we may attribute to this destructive action which may break down body protein in such fashion that certain poisonous split pro- ducts are formed in sufficient amount to cause fatal intoxication. The interesting suggestion that glycocoll may be obtained from the globulins of the body finds no support in our experiments. It is known that under certain conditions there may be a rather prompt shift in the albumin-globulin ratio in the blood plasma. Therefore, this suggestion that glycocoll might. be furnished in this way by a change of globulin to albumin was worthy of serious consideration. It may be claimed that our experiments do not rule out this possibility but at least we can say that the plasma albumin-globulin ratio does not change after a benzoate injec- tion. If we wish to cling to this hypothesis, we must postulate an effect which is limited to the tiésswe globulin-albumin ratio— this would seem to be indeed a venture into realms of conjecture. SUMMARY. The synthesis of hippuric acid in the body following benzoate administration is not prevented by an extensive chloroform liver necrosis. A severe liver injury, however, will cause a distinct delay in the synthesis and excretion of hippuric acid. This indi- cates that the liver normally takes part in this synthesis but that other cell pee of the body may be concerned in this con- jugation and may in an emergency take over a greater part of the hippuric acid synthesis. This may apply particularly to the intravenous administration of the benzoate. Our experiments show distinct increases in ammonia, urea, and total urinary nitrogen wherever dosages of benzoate are given intravenously, exceeding a certain amount per pound body 246 Studies of Liver Function weight. The question of dosage may explain many discrepancies noted. in the literature. Under certain conditions benzoate injection causes a consid- erable protein breakdown due probably to the acute need for glycocoll which is taken from the body protein molecule. The suggestion that under emergency conditions the glycocoll may be obtained from the globulins finds no support in our experiments. The serum albumin-globulin ratio is not changed by administra- tion of large doses of benzoate. BIBLIOGRAPHY. 1. Whipple, G. H., Peightal, T. C., and Clark, A. H., Bull. Johns Hopkins Hosp., 1913, xxiv, 348. 2. Kingsbury, F. B., and Bell, E. T., J. Biol. Chem., 1915, xx, 73. 3. Lackner, E., Levinson, A., and Morse, W., Biochem. J., 1918, xii, 184; J. Biol. Chem., 1918, xxxili, p. xvi. 4, Parker, W. H., and Lusk, G., Am. J. Physiol., 1899-1900, iii, 472. 5. Ringer, A. I., J. Biol. Chem., 1911-12, x, 327. 6. Epstein, A. A., and Bookman, S., J. Biol. Chem., 1911-12, x, 353. 7. Lewis, H. B., and Karr, W.G., J. Biol. Chem., 1916, xxv, 18. 8. McCollum, E. V., and Hoagland, D.R., J. Biol. Chem., 1913-14, xvi, 321. 9. Umber, F., Berl. klin. Woch., 1908, xl, 885. 0. Abderhalden, E., Bergell, P., and Dérpinghaus, T., Z. physiol. Chem., 1904, xli, 153. 11. Raiziss, G. W., and Dubin, H., J. Biol. Chem., 1915, xx, 125. 12. Folin, O., and Flanders, F. F., J. Biol. Chem., 1912, xi, 257 13. Davis, N. C., and Whipple, G. H., Arch. Int. Med., 1919, xxiii, 612. 14. Robertson, T. B., J. Biol. Chem., 1915, xxii, 233. THE BASAL METABOLISM OF UNDERWEIGHT CHILDREN. By KATHARINE BLUNT, ALTA NELSON, anp HARRIET CURRY OLESON. (From the Department of Home Economics, University of Chicago, Chicago.) (Received for publication, September 12, 1921.) Benedict and Talbot’s monograph (1) has given a large amount of material on the basal metabolism of children with which sub- sequent data can well be compared. It includes not only their own extensive series of observations but also a summary and criti- cism of earlier work. Their own observations are chiefly on normal children—children in good general héalth and approxi- mately of the average weight for their height. Very little has been published on the basal metabolism of under- weight children. Some of the infants observed by Benedict and Talbot (2) were underweight, and in general showed a higher basal metabolism per kilo or per square meter than well nourished children. This difference was assigned by the investigators to a smaller proportion of fat and a larger proportion of active pro- toplasmic tissue in the underweight infants. Murlin and Hoobler (3), who in 1915 published observations on 10 infants varying in age from 2 to 12 months, also found that the underweight infants had a higher metabolism per kilo and per square meter than the normal or fat ones. Murlin and Hoobler compare their own se- ries with one by Howland, and with Benedict and Talbot’s; and they find that all but three of the forty-eight underweight and atrophic infants had a higher basal metabolism than the eighteen normal infants. Of the five overweight infants all but one had a lower basal metabolism than the normal ones. On the other hand, a group of young men who were studied by Benedict and others (4) during a prolonged period of underfeeding showed a basal metabolism progressively lower as their under- feeding continued. It was diminished whether expressed as total calories, calories per kilo, or calories per square meter. The 247 248 Metabolism of Underweight Children relative proportion of their active protoplasmic tissue may have gone up as they lost fat, but their “stimulus to cellular activity” had gone down with extensive loss of nitrogen from their bodies. These young men lived on a lower metabolic plane than normal, while the very thin infants lived on a higher metabolic plane. A group of underweight college women recently studied at the University of Chicago,! young women in fairly good health sup- posedly eating according to their appetites, showed a normal metabolism—unlike either the underweight infants or the under- fed young men. Only a few isolated observations on the basal metabolism of underweight older children have been found in the literature— a neurasthenic youth observed by Magnus-Levy (cited by Lusk, 5) who partially starved himself for a year or more and whose basal metabolism per square meter was 33 per cent below normal (similar to Benedict’s young men); and the thin brother of Rub- ner’s pair (also cited by Lusk) with a very high metabolism. How about the basal metabolism of the type of child commonly considered undernourished—the child apparently normal except for marked underweight? Will such a child show the high metab- olism of the underweight infant or the low metabolism of the underfed adult? EXPERIMENTAL PART. Children Studied.—This paper gives the results of basal metab- olism observations on two groups of children, most of them under- weight. The first group was made up of fourteen children attend- ing a Health School for underweight children held during the summer of 1920 in the Home Economics Department of the Uni- versity of Chicago”. The second group of 14 was from the Uni- versity Elementary School.’ The standard for underweight employed with both these groups was the commonly accepted one used by various child welfare organizations, and printed in convenient form by the Elizabeth ! Unpublished data. * A full report of this experimental school by the director, Lydia J. Roberts, is in preparation. * Our thanks are due to Miss Roberts of the Health School and to Prin- cipal Gillett and the teachers in the Elementary School for their helpful cooperation, Blunt, Nelson, and Oleson 249 McCormick Memorial Fund, Chicago. The figures for the children of our ages in these charts are from Boas, Burk, Bowditch, and Smedley based on measurements of school children.4 They give standards for age, height, and weight, but the calculations of underweight in this paper are from height-weight relation- ships only, without age considerations. While it is not believed that by any means the whole question of normal heights and weights of children can be settled by reference to these figures, TABLE I. Data for Health School Children. Name. Age. Height. Weight, nude. bony ves os yr. mo. in. cm. lb. kg. 1b. per cent sq.m. Cleland..... 12 0 | 53.7 | 186.5) 66.6 | 30.3 | 68.0 + 1.09 {Git aaa S| 4Gro, | 125.8) 51.0 | 23-27) SiG 10 0.91 Donald. .... 5) \Pacea | 156-2) 615 27-94) (Gse3 1 til 1.05 eA See 2: iP Se roteS 9130.9) 53.1 (24.2 54.1 11 0.94 Therina.....| 10 10 | 58.0 | 141-3) 74.1 | 33.7 75.8 12 1.18 Pinared | op ele o2eoe| 152.9) 56.7 | 25.8 | 57.2 12 0.98 Cambs. .:... 10 10 | 54.1 | 187.4) 60.7 | 27.5 | 62.4 12 1.04 arolds.-. .':.. 10 47] 53.9 | 136.7] 60.8 | 27.6 | 62.1 12 1.04 Elsie Ss. 9 7 | 49.1 | 124.8) 47.9 | 21.7] 48.8 13 0.88 1350) 9 3] 53.7 | 136.5) 60.6 | 27.5 | 62.6 14 1.04 Edmund 10. 11 | 55.2 | 142.8) 64.0 |} 29.1 64.4 I 15 1-10 Knox OF id Peer 122. 2):46.1 | 20:9.) 47.3 16 0.85 Helen...:....... OFS pie Stee |) 139.4) 59).1 | 2628] 6909 18 1.05 Elsie Sa.....| 9 3 | 48.0 | 122.0) 42.0 | 19.0} 43.0 20 0.84 it is true that the relationship serves as the most useful single criterion which we have for selecting “undernourished”’ children, and it is as such that they are used here. An underweight child is arbitrarily considered one more than 7 per cent below standard. The data as reported by Miss Roberts for the age, height, weight (with and without clothes), and percentage underweight of the fourteen children from the Health School at the time of entering school are given in Table I. Thirteen of these children were underweight according to the standard used (varying from 10 to 20 per cent below) and one, counted as normal, was 4 per cent under. They ranged in age from 9 to 12 years except one little 4See Baldwin, Physical growth and school progress, U. S. Bureau of Education, Bull. 10, 1914. 250 Metabolism of Underweight Children boy of only 8 years. All of the thirteen underweight children had more or less marked circles under the eyes, ‘‘winged”’ shoul- ders, and other features which characterize the undernourished child. The normal weight boy was normal in other ways. A medical examination of every child, made by Dr. Walter H. O. Hoffman, instructor of internal medicine (pediatrics) in Rush Medical School, revealed no clear symptoms of hyperthyroidism in any of the children. One boy, Harold, was diagnosed as having active tuberculosis of the lungs. Detailed case histories of the children will be given later in Miss Robert’s report. The second group was composed of ten underweight, two over- weight, and two approximately normal weight children from the University Elementary School of about the same age as the Health School children. They were observed from March to May, 1921. The ten underweights ranged from 9 to 27 per cent under. The two heavy children were 40 and 12 per cent over- weight. These children were weighed nude only, for the metab- olism experiment, and to give figures to compare with the stand- ard height-weight chart which is made up from averages of clothed children, 1 pound was added to the nude weight of the girls and 1.5 pounds to the nude weight of the boys, figures obtained by averaging the weight of the clothes of the Health School children. The general data for these children are given in Table II. Method of Determining Basal Metabolism.—In determining the basal metabolism the Benedict portable respiration apparatus was used with the general procedure described by Blunt and Dye (6). The child came without breakfast, lay quietly on the couch for half an hour, and then was connected with the apparatus by a mouthpiece cut down in size so as to be comfortable. The observations of oxygen consumption were made usually for at least two 10 minute periods with a moment or more rest in be- tween. Occasionally when the child grew tired or restless the time was shortened a little, and in a very few cases in the Health School and more in the Elementary School group, one 10 minute period instead of two was accepted for study. Two stop-watches were used for each period so that the observer had a check on her own reading. Of course the main problem was to keep the children quiet. Both groups of children, those from the Health School especially, were used to various “tests” and a very friendly relationship Blunt, Nelson, and Oleson 251 existed between the children and the observer. The Health School group organized a ‘‘basal metabolism club”—so called by the children—and especial prestige was attached to belonging to it; also prizes of various kinds were given. During the rest and observation periods the children were read aloud to, and so helped to keep quiet and relaxed. A simple marking device was used to keep count of any movements the child might make in each one of the 10 minutes, a — being recorded if no observ- TABLE II. Data for Elementary School Children. Weight, Weight | Surface ae Age. Height. | Weight nude. clothed. | variation.| Du Bois. yr. mo. in. cm. lb. kg. lb. per cent sq.m. Charles.....| 10 6 | 61.2 | 154.4/138.6 | 63.0 | 140.1 +40 1.60 Margaret...| 10 1 | 55.2 | 138.0} 83.0 | 37.7 | 84.0 +12 1.18 Monert.’.:. >: 10 O| 49.7 | 124.4) 55.0 | 25.0] 56.5 —3 0.94 GAM eivers os HOmeeioO2e9) 152.0) 62.55/)28°44 |) Ga.0 — 5 1.02 Beets. 2 <0, 7 7 46.3 | 115.9) 44.0 | 20.0 | 45.0 —9 0.80 John. Gaetlnioseen 1oa-2) 60.5 (27.4 | Gls —ll1 1.00 Saraune. so. S) os 00.2) | 125.5) 50.8 | 23.1 | 51.8 —12 0.90 Edward.....| 11 11 | 60.4 | 151.0} 84.3 | 38.3 | 85.8 —12 1.28 Paniee...). 8 8 | 58.4 | 133.5) 57.6 | 26.2 | 58.6 —14 0.98 omer... ; 12 0} 59.4 | 148.5} 76.8 | 34.9 | 78.3 —16 1.20 122100 beens 1) 45) Sie + 145.3| 67.3)| 30-6" | | 68.8 —18 1.12 Elizabeth...|} 8 7 | 55.4 | 138.5} 58.3 | 26.5 | 59.3 —22 1.01 Helen...:..:. ieee Glee) 153.0) 78.5) | song || 7970 —23 1.18 Bryant, ../!'. 9 115627 | 141.6] 58.3 | 26.57) 59.8 —27 1.04 able motion was made, a single + for a very slight movement of the hand or arm, several + signs for a larger movement. More than a very few + signs caused the observation to be discarded. Asa standard, the day’s duplicates of oxygen consumed per min- ute were considered satisfactory only when they agreed within 10 ec. of oxygen and in reality the agreement was usually very much closer than this. For example, the duplicates for Elsie Ss., the Health School child with the most markedly abnormal metabolism were 180 and 181 cc. of oxygen per minute the first day observed, 180 and 170 cc. a month later, and 165 and 168 ce. still later. With a few of the more difficult children several efforts had to be made before a satisfactory determination was obtained and a very few children who were attempted never were sufficiently 252 Metabolism of Underweight Children quiet to complete the experiment. Most, however, behaved surprisingly well, especially after the strangeness of the first time was past. ‘ Eleven of the Health School children were observed success- fully on 3 or more different days; one boy, Roy, only once be- cause he left school and another, Edmund, only once because it was only toward the end of the summer, after he had been taught how to lie still during the school rest periods that he succeeded in doing the experiment successfully. This training in lying quiet which was enforced upon all the Health School children by their daily 1 hour rest period was of great assistance in the metabolism work. In the Elementary School group three children were ob- served on 3 days and the rest on 2 days. DISCUSSION. The results of the metabolism observations are given in Tables III and IV. The observations, calculated for 24 hours, are expressed in the usual three ways—total calories, calories per kilo, and calories per square meter. Figures are also included in these tables for children of the same weights as ours, read from the smoothed curves given by Benedict and Talbot (1) to represent the trend or roughly the average of their metabolism observation. Lastly, there is given the variation of the metabolism of our children from that of Benedict and Talbot’s. Health School Children.—For the Health School group it may be noted that the metabolism of all thirteen of our underweight children is markedly higher than the standard and that of the normal boy close to it. Elsie Ss. is the most extreme case of this excessively high metabolism. Though not the most under- weight (only 13 per cent), she was one of the children the most below par in general frailty and nervousness. She produced heat (basal) at the rate of 1,188 calories per day, while the standard child of the same weight produces 860 calories or 39 per cent less. Calculated per kilo, her heat production is 53.1 calories and per square meter 1,350 calories, or 39 and 41 per cent higher than the standard curves. She is not only markedly higher than her point on the curve but much higher than even the highest of the several different children of her weight from whose obser- vations the curve was drawn. That is, every kilo of her body seems to be living at a higher rate than that of the normal child. Blunt, Nelson, and Oleson 253 This little girl, while showing most markedly the high metab- olism, indicates the tendency of the rest. Twelve of the thir- teen underweight children in the Health School have a basal metabolism 11 per cent or more higher than the standard curve, as shown by all three ways of calculating, and seven a metab- lism 20 per cent or more higher. Also, like Elsie Ss., all but one (Carl) of the thirteen children show a higher metabolism not only than the average curve, but than any of the children of similar weight making the curve. The average excesses above the curve for all thirteen Health School children are 22, 25, and 24 per cent; that is, the basal metabolism averages roughly a quarter higher than that of the average child of the same weight. There is no close relationship between the amount underweight of any child and the amount of excess metabolism. Elementary School Children.—Among the Elementary School group (Table IV) the connection between underweight and high metabolism is not quite so marked as in the Health School, but is still noteworthy. Metabolism not far from normal is shown by the two overweights, one of the normals, and two of the moderate underweights. Moderately excessive metabolism (over 10 per cent and under 20 per cent excess) was observed in five of the underweights, and high excess (20 per cent or over) in the other three. The average excesses for the ten underweight children over Benedict’s of the same weight are 14 per cent for the total calories, 16 per cent for the calories per kilo, and 18 per cent for the calories per square meter. These excesses, while not so high; are closely in line with the results with the other group. However, among the Elementary School subjects there was one boy, Robert, of approximately normal weight and yet with very high metabolism. Robert was a difficult subject, prob- ably the most difficult included in this paper, so that it is possible that in spite of fairly close agreement between his dupli- cate determinations, a certain amount of bodily tenseness during the observations may account for part of his excess. An inter- esting point about several of the children with very high metabo- lism is that they are what the school principal calls “problem cases” —children peculiarly difficult to manage in the school. Comparison of Our Children with the Standard for Same Surface and Same Age.—There are two other sets of comparisons which may well be made with Benedict and Talbot’s numerous curves— cl Ck6 910‘ T a 61 0&6 S21 T ST 9% 000'T | T9%‘T 8z SI ezo't | etz‘'t | 6r 8 Gs6—s|«6SO'T 9 quao sad quad sad Metabolism of Underweight Children *‘sSa0xqy | “JoIpoueg |‘poArosqg| ‘ssooxnt “ur ‘bs red satiojep ges 0° LE g'9€ 0 OP 0'SE T Le 6°SF 9°9F 9 LP “O[Ly red sart0yeg Il OPI‘ T SI 006 &Z G10‘ T cl 096 0 OST‘ T quao sad *SSOOXHT “SoT1O] BO [BIO J, ‘THI GTAV.L “qolpouog |" poaresqg “UILP]YY) 100YIY YIDAFT ‘uouyonpotg qoIE] Zt “OSVIOAY G GE sy LP 7 ‘sny Vv €& Fe oun Lz ‘OSVIDAY L FG 6I 5 1 ¥% gl Sny LG 9g 9une F 8% ‘IBVIODAY Z 62 6 “Sny G 8S gz A[ne 9° LZ oo GUN G &% 8g ount o'Te ‘OSVIOAY ris g “Sny PE € » 6° 0€ ct Ang € 0€ Lo eune “Oy 0s6l “WOI}eU -1ULIa}ap “HOIPBUTUTIOZAp jo Aep UISI[OG BOUT ‘Qysiom jo 930q] opnn at II II OT v quad sad “qolpouag |"poaresqg *BUIIOY ‘stuBe “prsuod “uyor *PUs[z[D *QUIGNT 259 Blunt, Nelson, and Oleson T€ ly 61 Or 86 OF6 S61‘ T Oot‘ T COZ T SE 6& 1G OT VG oTé 0°8E 0° LE 0° LE gos Lig Teg 3° tP L0¥ 6° €P LI 020‘ T ee OF0'T 6g 098 ST 090‘ T iE 020‘ T (a6 026 gz “Sny og oun ‘OSVIOAY 6 “sny LE » cz ounr ‘OSVIOAY pices p ‘sny ‘OBVIOAY jal ” al ‘Sny eI Arne "ISVIDAY SI ” eI “Sny 9z oun CT a! &I ai ai ol *punurpiy *£OY “SS SOrsT “plore yy ETO) Pathan q o Me) 1=| {2 oO = on 4 B H o co q 2) Gey (o) S mM = (o) a) 3S ars) 3 256 0€ 026 yeaa G¢ ¢ 07 se 4 €& 9G OF6 L81'T 66 ¢ Ss 8 SP 1K6 1g ceo'r | $98'T Ié G Ty € VS LG quand sad quao sad quaa sad “sseoxny | ‘golpauag |"paAsosqg | ‘“ssooxa, | “Jolpeueg |*paaresqQ | “ssooxny OSL 086 006 ‘ul ‘bs sad saiojep ‘OTL tod sars10[eg “SOTIO[BO [BIO], 1F0‘T F 61 €20' T 102 120‘ T 161 0g0'T 0°61 LEZ‘ T GLE 26% 1 Tse acm t 19% POS I 69% ISt'T ZB te FOL T 212 ZL0‘T aa 91Z' T 6 02 “Dy “qgorIpousg *padAsosqdg “uol}eU jo Awp “qQustIom epnn ‘IBVIIAY ¢ ‘sny FG ” 6T oun “OBVIOAY g “sny 9% Aine &S oun ‘OSVIOAY O}S ” FI ” 6 Ane 0661 -1uliajep | UoOTyBUTUIIaZap ursijoq Byeul JO 078 0% “BS oIs[ ST “u9[9H OL ‘xouy quao sad “*qUydIoM . BY eg BTeN SS ee ‘papn7IUoO—iIl W1aVL Blunt, Nelson, and Oleson Lin = Oe with their children of the same body surface as ours, and with children of the same ages as ours. In Table V we give the per- centage variations resulting from these two comparisons for the Health School children as well as from the weight comparisons repeated from Table III. Our children as compared with Bene- dict and Talbot’s children of the same surface show in most cases as would be expected, excesses closely like those for children of the same weight—on the average the total calories are 21 per cent higher and the calories per square meter 24 per cent higher. The age comparison brings out somewhat different points. Three of the Health School children were considerably taller than the average height for their age; Therina 4.5 inches, Roy 3.3 inches, and Helen 4.9 inches, and, therefore, the total heat pro- duction expected from them would be that for an older child; or, to express the same thing differently, they would be expected to show an excess when compared with children of their own age greater than the excess when compared with children of their own weight or surface. This is true for Therina and Helen, but not for Roy. On the other hand, three of the children were markedly below the average height for their age; Janis 2.8 inches, Knox’"2.0 inches, and Elsie Sa. 2.1 inches. These children would be expected to show a lower excess in total calories when compared ‘with children of their own age than by any other comparisons, which is markedly the case with all three. In making all these comparisons it must be remembered that Benedict and Talbot expressly state that their curves do not represent mathematical averages, merely show the trend of their observations, and that, of course, considerable variation in metab- olism must be expected from child to child. No such thing as an accurate “standard” is presented by them. Our figures would, of course, have little significance if they merely showed irregular variations from the curves. It is the marked excess over Benedict and Talbot’s observations, occurring almost with- out exception in the untlerweight children that makes our results significant. From all points of view it seems safe to draw the conclusion that all of the thirteen underweight children from the Health School were metabolizing at an abnormally high rate, and all but two of the ten underweights from the Elementary School. 258 “SOlIO]BO [BIO], “Udlp]YyY) J00YIY hivjpuamayy ‘uowyonpold 192 ]] ‘AI GTEVL _- emma — es ‘WOT}ElIVA | “JoIpIusg |" pearasqO |WOlZeIIwBA | “yoIpausg | poasresqy |‘uolyelIeA | ‘yorpousg |*paasosqg ee, : SSS SS SS SSS “ZY stom aqncy “UOIPBLIBA ‘gue ‘mi ‘bs rad sers0yeg “O[LY Jad sar10yeg epnNn JYSIOM - 259 Blunt, Nelson, and Oleson L- be+ 6 + Te+ 3+ ‘GIGI ‘646 ON ‘uopBurysy 4 '78Uuz aBausnyg “Dy * VT “QoIpouog puv “y *f¢ ‘STLIvT] UT SopTqvy oY} WOT poYV[NIVy) , Sh6 066 OL6 St6 096 096 O10‘ T 290‘ T 628‘ T ZIL'T P80‘ T Eh0'T Lett 260‘ 1 oI+ se-+ Cl Gs: 8I+ 6° SE 0°98 0° CE 0'8¢ € LE TOP 9°8P POF 8 FE 6 FP 6°68 Ler vet (ee Fae 61+ Gort: Ogi T 096 G8Z' T G18 0901 90‘ T G10‘ T 090‘ T 68h T SSP T O6F T pee’ T 978 T CPS T Z90' T £90‘ T 190‘ T D 6 TE 6 8& Pp LG ‘ODTIOAYV a ey OT “IBIN oo— ‘ODCIOAY LG ” gg “ad SI- ‘OSVIOAW LG ” 9% ‘Idy 9T— ‘ODVIDAV re! ” py “ABI i ‘OSVIOAV ‘ODVIOAY tI ” OL ABI ZI— ‘ODVINOAV ie ess ZI ady t= “yoquary “OUTLOIBS) Metabolism of Underweight Children 260 LT = ne ee oI+ O10O'T | SILT 5 0'8s L'&P Ti+ ceo‘ Tt | oor‘ T ‘OBBIOAV POL T 6s G6L‘ T G' 9% Zl ‘Ady Lo= “que dre Cc ORG op Tale OT | o-ee || O'S | ot+..| Ogn'T-| 79 1 ‘OBUIOAY ose’ T 6G » Z0F' I LZ idy oes‘ T LSS LZ IB Ss “Udo FT quad wad quaod Jad quaa sad “By 1é6I quaa sad UOIPBUIBA | “JOIpoudg |*poArosqy |"UolzerIeA | “yorpouag |*peA.1osqQ smorenre, “qoIpsuog | poArosqd 5 a | ION . eg “HOIYBIIBA ‘oueyy ‘ur ‘bs 10d sor10[eg *O[Ly rod sat10jeD “SOL1O[BO [BIOT, opnN WIM “‘papnjoUuog—Al ATAV.L Blunt, Nelson, and Oleson 261 Day by Day Variation in Metabolism.—The discussion above has been based upon the average of the determinations for several different days for most of the children. Two questions arise in connection with this procedure: Were the results for the different days in close agreement; and was there any regular change in the children’s metabolism in the Health School as the summer pro- gressed and the children improved in general condition? TABLE V. Variation in Heat Production from Benedict and Talbot’s Figures, Health School Children. Compared with a a Compared with children chilean of 8 children of same weight. [seme auehice: a4 of the same age. Name. $ 1 be be 1 be ies ' he be 2/2 )8 |8 |2 1/8 | #e/3 |B | 8 = 2 g og =! Sq | +2 2 3 oq S jag les |e tl} ag let] sO lag} es| ee eee | ore | Sot |) Sr | oror | omer (cue | ce tia oes s oF an = a on | Sa oA on | ae aa P B 1) é) & iS) q & 6) @) per per per per per per per per per per " cent | cent | cent | cent | cent | cent | cent | cent | cent | cent AC IOMANG 2.25 52,5. 4 0 6 8 6 8 |—0.7| —4 i 8 elit aa 10 15 19 18 16 19 |+0.6} 15] 21 iY Wornalds ces... .- 11 23 | 28) 26 | 25| 27 |—0.9) 15 BD | 28 ANS eee sites oe! 11 18 18 19 15 | .19 |-—2.8)- 5 | 27 18 PRRerin ae.) ats". 12 11 11 15 12 17 |+4.5}) 27 6 12 MWinldmeds access) 12) lee?) | 24 | 28.) 23) | (27 |-F0.8]) 225) 922) 125 (Chand 4S ae sae 12 7 10 10 9 11 |+0.8} 12 15 10 TOL sees atiscccal) . 12 sn 2b 19 19} 21 |/41.5) 13 24 18 IBIBIE/ OSs. 2 5.x. «- Istieoo | 39\| 41 37 | 41 |—0.7| 25) 48] 40 ERO. oss. 3 WSR | oo | + ob 31 34 |+3.3} 34] 36] 30 Edmund......... 15 17 16 15 13 16 |+1.7| 10 14 14 HSO Ky. eit) soy. 16 PCI) BT 31 28} 31 |—2.0) 10| 45 33 IGG Anas eee 18 27} 29| 26] 23] 26 |4+4.9) 34] 24] 22 BYSIG\OAr.do.: ss 20 | 33 35 30] 24] 27 |-2.1) Il 49 | 27 Average of all but (CLS Cy: a A 22) 251). 24) | sais Ze 18} 28 With the Elementary School, where most duplicates were obtained on consecutive or almost consecutive days, the agree- ment between the different days was decidedly close; for ten of the fourteen children the maximum difference was 5 per cent or less, and for the other four it was between 6 and 8 per cent. 262 Metabolism of Underweight Children In the Health School the agreement between different days is not so close, but also the intervals covered were very much longer, and the number of observations on most children 3 or 4 instead of 2 or 3. Only two of the children show as little as 5 per cent difference between duplicates, one shows as high as 20 per cent, and the average of all differences is 11 per cent. This is of interest in comparison with the rather wide day by day variation found by Blunt and Dye (6) for their normal women. There is no regular change in the children’s metabolism as the summer progressed, in spite of general improvement in health. Apparently the time was not sufficiently long and the gains in weight sufficiently great for their metabolism to go down to that of normal children. SUMMARY. Metabolism determinations made on twenty-eight children, mostly underweight, showed that the basal metabolism of under- weight children tends to be higher than that of the normal child. The excess metabolism was in some cases as high as 40 per cent above that read from curves given by Benedict and Talbot, and in most cases the metabolism was not only higher than the curve but higher than the highest observation of the child of the same weight from which the curve was drawn. The average percentage excesses for the underweight children in the Health School com- pared with Benedict and Talbot’s of the same weight were 22 for the total calories, 25 for the calories per kilo, and 24 for the calories per square meter. In the Elementary School the corres- ponding excesses were somewhat less—14, 16, and 18 per cent. No close relation was observed between the percentage under- weight and the excess metabolism. BIBLIOGRAPHY. 1. Benedict, F.G., and Talbot, F. B., Carnegie Inst. Washington, Pub. 302, 1921. 2. Benedict, F. G., and Talbot, F. B., Am. J. Dis. Child., 1914, viii, 1. 3. Murlin, J. R., and Hoobler, B. R., Am. J. Dis. Child., 1915, ix, 81. 4. Benedict, F. G., Miles, W. R., Roth, P., and Smith, H. M., Carnegie Inst. Washington, Pub. 280, 1919. 5. Lusk, G., Elements of the science of nutrition, Philadelphia and London, 3rd edition, 1917. 6. Blunt, K., and Dye, M., J. Biol. Chem., 1921, xlvii, 69. e > Ee RELATION BETWEEN THE CHLORIDE CONTENT OF THE BLOOD AND ITS VOLUME PER CENT OF CELLS. By A. NORGAARD anp H. C. GRAM. (From the Medical Clinic of the University of Copenhagen, Copenhagen, Denmark.) (Received for publication, December 28, 1920.) Analyses have shown that plasma contains a higher chloride percentage than is generally found in whole blood (McLean and Van Slyke, Van Slyke and Donleavy, McLean, and Austin and Van Slyke). The experiments reported in this paper were under- taken to determine the relation between the chloride content of the blood and its cell volume percentage. The examination of the blood was based on a micro determination of the chlorides and a determination of the volume of the blood cells. The following technique was used. 0.5 cc. of 3 per cent sodium citrate solution is measured off in each of two graduated, cylindri- cal 5 ec. centrifuge tubes (Oluf Thomsen’s tubes for the blood platelet count). By venous puncture approximately 4.5 ce. of blood are added to each of the tubes, and the blood and citrate mixed by corking and inverting. Any possible variation from the exact measure is read off on the scale graduated to 0.1 ce. One of the tubes is reserved for the determination of the volume per cent of the blood cells, while the other is used for the micro- chloride determination on citrated blood and citrated plasma in the manner described below.! The chloride content of the blood is calculated as sodium chloride and is given in per cent of 100 gravimetric parts of whole blood or plasma. 1 In regard to the exchange of chlorides between corpuscles and plasma due to differences in the CO, tension (Fridericia), our constant results tend to show that we must have studied blood with a nearly constant CO, tension. The tubes were corked immediately after the taking of the blood and were only opened for a moment to withdraw the blood or plasma for analysis. 263 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 2 264 Chloride Content of Blood The micro determination of Ivar Bang has been used. This method is based on extraction of about 100 mg. of blood absorbed on filter paper with 92 per cent alcohol for 5 hours. The alcoholic extract plus an equal volume of alcohol, used for washing out the test-tube containing the blood-imbibed paper, are titrated with a 0.01 n AgNO; solution. The chloride values obtained by this technique are somewhat lower than those by the former methods. Bang found that the extraction of the chlorides is complete after 5 hours. However, it seems that the process tends toward a balance of distribution between the coagulum and the alcohol. Thus the extraction cannot be complete with one por- tion of alcohol and even a second extraction does not absorb all the chlorides from the blood, though the succeeding extrac- tions after the second will contain only barely measurable quan- tities of chloride. In our preparations for this work 400 specimens of blood, serum, ascitic fluid, edema fluid, pleural exudate, and cerebrospinal fluid were examined by successive extractions. The average of the chloride percentage calculated from the first extraction was 0.476 per cent; the second, 0.047 per cent; the third, 0.007 per cent; the fourth, 0.0003 per cent; the fifth, 0; the sixth,-0.0005 per cent; and the seventh, 0. For practical purposes it will be sufficient to extract twice with alcohol, since the second portion plus the alcohol used for washing out the tube after this extraction contains about 10 per cent of the amount determined by the first extraction and since the third and following extractions will contain only unimportant traces of chloride. The results in the -present series of observations confirmed this inasmuch as the average content of chlorides found in the second alcoholic extraction was 8.7 per cent of the average amount found in the first. This phenomenon is not unknown from other similar processes and is a natural consequence of the laws of balance for the adsorp- tive distribution between the chlorides extracted in the alcohol ° and the amounts left in the coagulum. Bang titrated with a 0.01 n AgNO; solution. We have used the solution of Mohr diluted 20 times; 7.e., 1.4521 gm. of AgNO; and 1,000 ec. of water. The titration has taken place in diffuse we wee A. Norgaard and H. C. Gram 265 daylight with a micro-burette and potassium chromate as in- dicator. A standard color for the titration has been made in the manner described below, which is a very important point, if exact de- terminations are to be made. The procedure, in brief, is as follows: To a test-tube, Gi, containing 15 cc. of the alcoholic extract, add, drop by drop, enough (nec.) AgNO; solution to give a dis- tinct brownish color. The glass is kept in the dark for 15 minutes, which makes it slightly less colored. To another test-tube, G2, with the same amount of alcohol, add n cc. of AgNO; solution. A small addition of silver nitrate solution to G, makes the color of both mixtures identical. This tint does not alter appreciably during the titration if the light is not too strong; in which case it becomes somewhat darker. The value n is subtracted from the total number of cc. added to G,; before the calculation. Each ce. of the AgNO; solution used in excess of n means 0.5 mg. of sodium chloride. The titer of the AgNO; solution was tried before titration with a measured quantity of a standard solution of NaCl. The duration of the first extraction with alcohol has always been 24 hours, and the second at least 24 hours. The chloride values given are the average of two or three determinations which very rarely diverged appreciably. The calculation from the percentage in citrated blood and ci- trated plasma to uncitrated blood and uncitrated plasma has been made as stated under the description of the determination of the cell volume. The contents of chloride found in citrated blood and plasma must be reduced to the contents in uncitrated blood and plasma. The reduction of results from citrated blood to those for un- citrated blood is simple. If 0.5 cc. of citrate is mixed with 4.5 ec. of blood, then 100 parts of citrated blood will contain 90 parts of uncitrated blood. If, on the other hand, the same quantity of citrate is mixed with 4.7 ec. of blood, then the cor- responding volumes are 100 parts of citrated blood and 90.4 parts of uncitrated blood. The correction necessary to reduce citrated plasma to uncitrated plasma is a little more complicated, since one must know: (a) 266 Chloride Content of Blood the osmotic properties of the citrate solution; (b) the quantities of citrate and blood mixed; and (c) the volume per cent of the blood cells. It is evident that if the citrate solution is not isotonic, water or salts will enter or leave the corpuscles, thus altering the con- centration of the citrated plasma. We have chosen the 3 per cent citrate solution because it does not alter the cell volume, as will be seen from cell volume determinations (Table I) made on defibrinated blood. Table I shows clearly that the admixture of a 3 per cent sodium citrate solution to the blood in the ratio 1:9 does not alter the cell volume. TABLE I. Cell volume of 4.5 ce. of defibrinated blood Cell volume of pure defibrinated blood. + 0.5 cc. of 3 per cent citrate.* per cent per cent 40 40 43 43 54f 55 71t 70 687 68 * Values corrected for admixture of citrate. { The last three specimens of defibrinated blood were concentrated by drawing off some serum in order to exaggerate and make visible any slight variation. On the other hand, a determination of the cell volume on fourteen double specimens, of which one was mixed with a 10 per cent sodium citrate solution showed that in this way there is a considerable shrinkage amounting to 16 per cent of the shrunken cell volume in the citrated blood. The technique used for the determination of the cell volume was very simple. The tube set aside for this purpose was cen- trifuged for 90 minutes at 3,000 revolutions a minute. The cell precipitate, including the white cell layer, was read off and the cell volume per 100 ec. of uncitrated blood, volume per cent, calculated as shown in the following example: 4.5 ec. of blood are mixed with 0.5 ec. of 3 per cent citrate and give a cell precipitate of 1.80 ec. The cell volume percentage is then et = 40 volumes per cent. A. Norgaard and H. C. Gram 267 Of course it has to be proved that centrifuging in the way described does give the absolute cell volume. If this is the case three precautions must by observed: (1) The centrifuge must be kept at an even high speed and slow down very gradually; (2) the tubes must fit exactly into the centrifuge so that they do not wobble; and (3) in no case must the time of the centrifuging be cut down, even though 90 minutes allows a rather wide margin. In order to test our technique, it has been compared with a determination of the cell volume by the hematocrit method, the capillary tubes of the hematocrit being filled from the tube TABLE II. Cell volume. No. Hematocrit. Centrifuge. Difference. Average of two efettsenriniedararis One determination. per cent per cent per cent 1 43 43 0 2 30 31 5 3 39.5 40 +0.5 4 43.5 43 —0.5 5 37 38 +1 6 28 29 at 7 64.5 66 +1.5 8 63.5 64 +0.5 9 44.5 45 +0.5 10 33.5 33 —0.5 11 31 31 0 Average....... 41.5 42 of citrated blood. The volume of the blood absorbed in this way is inconsiderable and cannot be read off on the scale. The hematocrit ‘employed made about 9,000 revolutions a minute and the blood was centrifuged with an interval every 15th minute, stopping when the blood column had been constant in length thrice and transparent throughout. The results are given in Table II. The variation between the results is inconsiderable, even if the hematocrit results are slightly lower. Even the largest varia- tion, in a case of polycythemia, No. 7, will not materially affect 268 Chloride Content of Blood our calculations. The manner in which the chloride content of citrated plasma is reduced to that for uncitrated plasma is indicated in the following example: 0.5 ec. of citrate is mixed with 4.5 cc. of blood and gives on centri- fuging a cell precipitate of 1.80 cc. The cell volume is then 40 per cent. If the other specimen contains the same proportions of citrate and blood, then 100 volumes of citrated blood contain 64 volumes of citrated plasma, or 54 volumes of uncitrated plasma. Thus 100 volumes of citrated plasma contain $4.4 volumes of uncitrated plasma. If the chloride percentage in citrated plasma is found to be 0.48 then the real plasma percentage is 0.57, as the citrate solution used proved to be absolutely free from chloride contamination. In giving the results of our determinations of the chloride per- centage in plasma (and blood) we have divided the cases examined into groups. Cases with Normal Blood.—As such, have been tabulated cases in which the cell volume varied between 40 and 50 per cent of the blood,? and where the individual examined did not show symptoms which might influence the composition of the blood. The material consists of fifteen cases (five men and ten women) showing a cell volume within the above limits. The chloride percentage in plasma calculated as sodium chloride varied be- tween 0.59 and 0.63 per cent, the mean value being 0.61 per cent. We may in this connection mention that this value, 0.61 per cent, is the same that we have found as an average result in all the thirty specimens of edema fluid which we have examined. Table III, as well as the following tables, gives the name, sex, age, diagnosis, true cell volume, and sodium chloride per cent in the blood and in the plasma. The percentage in citrated plasma or blood is not given in these tables. The variation in the chloride content of the plasma is so small that it may be considered constant. The plasma percentage of chlorides is the same in both sexes, though men on the average have larger cell volumes (and hemoglobin values). Diseases in Which the Salt Metabolism Might Possibly Be Affected—In the four cases belonging to this group the cell 240 per cent is not the lowest normal cell volume, since down to 36 per cent may be found in perfectly sound women. For our purposes, which deal only with the purely mechanical proportions between plasma and the cell volume, this is of no importance. 4 q a A. Norgaard and H. C. Gram 269 TABLE III. Normal Cases. NaCl | NaCl No. | Name. Sex. Age. Diagnosis. Cell 2 t volume. | 41664. |plasma. yrs. per cent |per cent sor bak 1 | A.N. | Male. 35 | Normal. 50 0.46 | 0.62 Zo aver: Ae 24 | Sequela poliomyelitis.| 45 0.45 | 0.59 urls “ 21 | Normal. 44 0.47 | 0.63 Ai GocS: 6 36 | Neurasthenia. 42 0.47 | 0.60 ST Pi D ie 12% se 47 | Disseminated sclero-| 41 0.48 | 0.62 sis. 6 | C. C. | Female. | 42 | Encephalitis. 48 0.47 | 0.60 Fi F. os 33 | Normal. 44 0.50 | 0.61 So leRoN. «e 17 | Struma. 43 0.50 | 0.59 SFO: EB: se 35 | Goiter. 41 0.48 | 0.62 10) |G. F. = 19 | Pregnant. 40 0.48 | 0.60 ee AW “ 31 | Normal. 40 0.49 | 0.63 HORN Ri: « 48 a 40 0.53 | 0.62 LG) ies ea << 44 | Sciatica. 40 0.47 | 0.60 14 | L. A. “ 57 | Encephalitis. 40 0.47 | 0.59 1550 fe a OF “ 17 | Rheumatism. 40 0.49 | 0.62 EERE OS ois SARIS Eee RII 28 peo 42.5 | 0.481) 0.609 TABLE IV. Cases with Possibilities of Disturbed Salt Metabolism. ~g |= |] 834 No.| Name. Sex. |Age.| Diagnosis. Date. | 82/98 /|5 & | Remarks. S| 25 | 23 per per per cent cent cent 1| EH. P. | Male. 15 | Edema Apr. 22} 48 |0.46 |0.61 | Before fugax ~ attack. (Quincke). May 5| 47 (0.48 |0.61 | During attack. 2/|B. J. | Female. | 54 | Pneumonia.| June 7| 42 /|0.46 {0.58 3 | C. C. | Male. 17 | Albumin- “< -8} 40 . 10.45 0.58 uria. AM BS: ss 60 | Chronic Apr. 23] 40 {0.50 |0.62 nephritis. 270 Chloride Content of Blood volume varied within the limits arbitrarily decided to be normal (40 to 50 per cent). The sodium chloride percentage in the plasma varies between 0.58 and 0.62 per cent. Even if the value 0.58 is a little lower than any found in our normal cases, the variation is so small that it may be considered as normal. Polycythemia.—By a coincidence, the cell volume was the same in the three cases observed, being 64 per cent. In these cases the sodium chloride percentage in the ola varied between 0.57 and 0.62 per cent, being nearly the same as in the cases already mentioned, though the amount of plasma per unit volume of blood was considerably reduced in comparison with the normal cases. TABLE V. Polycythemia. Ooo ad aE/=8]-8 No Name Sex Age Diagnosis Date. Se O06 | Da .| 2° | 2a yrs cents) leet |aaee S| Mie re | Maile: 59 Polycythemia. | Apr. 21} 64 |0.46 |0.60 2 |J. C.| Female.| 44 &¢ May 3] 64 |0.42 |0.57 3 |M. L. | Male. 65 . “< 6| «64 10.44 |0.62 ANCL AG CN 5 os Heese IRE rel Sis SO ate leas sites 64 |0.440/0.597 Simple Anemia.—As such, were tabulated all cases examined where the cell volume was less than 40 per cent and the color index of the blood less than 1. The leucemias, however, were relegated to a special group. The examination of the blood in these ten cases of which three have been examined twice, shows a variation of the cell volume between 37 and 13 per cent. Nevertheless, we find in the plasma a sodium chloride percentage between 0.58 and 0.63 that is nearly within the same limits and with the same average, as we have found not only in normal blood but also in polycythemia, the opposite condition of anemia. There is in these cases no relation between the cell volume and the chloride content of the plasma. The highest plasma-chloride value is found with a cell volume of 30 per cent, the lowest with a cell volume of 37 per cent. 7 ~~ A. Norgaard and H. C. Gram 271 TABLE VI. Simple Anemia. Sa Ne ica No Name Sex. Age. Diagnosis. Date ce ate oi oO | Sa |] SS > |Z Ze rs. cent | cone | cent 1 |G. V. | Female 24 | Pregnant. Apr. 17| 37 |0.50 |0.60 2 Gael oe 8 | Splenic anemia. | May 22/ 37 |0.48 |0.59 June 1] 35 (0.52 |0.61 3 |J.R. | Female.| 30 | Goiter. Apr. 24] 37 |0.50 |0.60 4 J. Vis oc 32 | Anemia. June 5) 36 |0.51 |0.61 5 |H. J. | Male 18 | Gastric ulcer. May 29] 35 |0.48 |0.59 June 9] 37 /|0.48 |0.58 6 |S. 8. | Male. 48 | Goiter. May 27/ 31 |0.52 |0.61 7 |A.P. | Female.| 37 | Gastric ulcer. Apr. 29} 30 |0.53 |0.63 5. 1-0: i 34 | Simple anemia. | June 11! 27 |0.53 (0.61 9 | J. K. | Male. 56 f re “ -4) 24 «(10.57 10.62 10 | N.M. es 46 a: s Apr. 22! 13 |0.58 |0.60 | ‘ | May 28] 16 {0.55 |0.59 ———, JEDI D2. tas Sesto hs aac en -ovactamaner 2 ge 30.4/0.519|0.603 TABLE VII. Pernicious Anemia. | Soe ee No.| Name. Sex. Age Diagnosis. Date: =|>'5 | OS os Remarks. Os| sa | 3a Se 4 Ze yrs . dant li conigll cent i iAEas Female. | 28 | Pernicious| May 27/27 |0.53 |0.63 anemia. 2|F. U. | Male. 36 = “| June 10/22 (0.57 |0.64 |More ex- ¢ actly, 0.636. 3 |V.N. | Male. 55 | Pernicious} Apr. 27/22 |0.56 |0.63 anemia. | May 31/10 |0.57 |0.62 June 7} 5 /0.59 |0.63 A. F. | Female. | 49 | Pernicious} May 1/18 /0.55 |0.61 4 anemia. 5 | E. P. | Male. 64 “| June 14/17 |0.55 |0.62 6|J. A. | Female. | 39 ce “| Apr. 27/16 |0.54 |0.61 May 513 (0.54 |0.60 16.7)0.556/0.621 272 Chloride Content of Blood Three other specimens with a cell volume of 37 per cent, how- ever, show a chloride percentage of 0.60, 0.59, and 0.60 in the plasma, and in the two cases with the lowest cell volume the chlor- ide percentage in plasma is 0.59 and 0.60. Pernicious Anemia.—This group comprises six cases, of which one has been examined three times and another twice. The cell volume in these cases varies between 27 and 5 per cent, so that all cases are pronounced anemias and some even among the most severe. The sodium chloride in plasma varies between 0.60 and 0.636 per cent with an average of 0.62 per cent. Even if the average values found lie a little higher than those found in the other groups, the variations are too slight to allow any conclusions to be drawn. We do not doubt, however, that TABLE VIII. Malignant Tumors. Or ihe ag = 3 No.| Name. | Sex. | Age. Diagnosis. Date. | A o/s 5 Os | aa |] BS Bo] pcs eis —— ee Oe ee eo 1 | 0. P. | Male. | 41 | Carcinoma oesophagi. | Apr. 20} 41 |0.49 |0.57 2|V.M. se 62 « hepatis. June 4/31 |0.49 |0.57 3 Bee “| 23 | Hodgkin’s disease. Apr. 24| 17 {0.53 |0.58 PAVICTEP C5, 5 5:0: \ave\leyetsss, 5s okeiere) ora apsueie a eRe MMR et asaeiets chageisjare asters 29.7|0.503)0.573 the plasma in these cases differs from normal plasma or the plasma found in simple anemia, but the difference is too small to permit one to say whether one single specimen comes from a patient with simple or pernicious anemia. Therefore, we shall, in view of the much larger differences in the chloride percentage of whole blood, consider the chloride percentage of plasma from pernicious anemia as being the same as in the plasma from the other patients. Malignant Tumors—Among the patients examined we find two cases of cancer and one case of Hodgkin’s disease, all verified post mortem. ‘These three cases, only two of which were anemic, are put in a special group. : Though the cell volume in these cases varies from 41 to 17 per cent, the plasma contains about the same amount of chlorides : ; A. Norgaard and H. C. Gram 273 in all three cases. The chloridedn plasma varies between 0.57 and 0.58 per cent, that is about the lower limit found in the normal cases. It is of course not permissible to draw any far reaching conclu- sions from material consisting of only three cases, where the di- vergence consists in that the mean value is lower than usual. We shall, as we did in the cases of pernicious anemia, consider these values as lying within the limits 0.57 to 0.64 per cent which we put forward as the boundaries of the normal chloride per- centage of the plasma. Leucemia.—Finally we studied two cases of lymphatic leucemia. Both cases were anemic and the volume determination gives the total volume of red and white cells; the partial volume of the latter, however, was not very considerable. TABLE IX. Lymphatic Leucemia. | a | onl (Ben. |e No.| Name Sex = Diagnosis | Date E:5 ag 58 og oo | Pee 23/62/92 | Se mH 8 >| Z Ze —— | ——e |e —| ——_ ——————————— SF — | - I LC COO per | per er yrs. | cent | cent cae nee 1|J. B. | Male. | 68} Lymphatic leucemia.| June 2) 7I |37 (0.51 |0.63 meer... - 63 _ aS May 25] 52 |22 (0.55 |0.63 June 11} 47 |24 {0.55 |0.62 The cell volume varied between 37 and 22 per cent, but the sodium chloride percentages in plasma were nearly the same in all three determinations. These values, 0.62 to 0.63 per cent, lie rather near the upper limit, but do not pass beyond it. From Tables III to IX the following conclusions may be drawn. 1. In normal persons we have found a sodium chloride concen- tration in the plasma which varies but slightly and whose average value may be put at 0.61 per cent, which is the same amount that we have found in edema fluid. 2. In some diseases (angioneurotic edema, pneumonia, al- buminuria, and nephritis) where abnormalities of the chloride metabolism take place or might be supposed to take place, we find chloride percentages in the plasma not differing from those found in normal cases. ——E———————<—=Vas lr, el 274 Chloride Content of Blood 3. Sodium chloride percentages between 0.57 and 0.63 in the plasma have been found in cases of polycythemia and simple anemia. Even the most extreme forms of these diseases (cell volume 13 to 64 per cent) fall within these boundaries which we consider the boundaries of the normal plasma-chloride concen- tration. 4. In the six cases of pernicious anemia and the two cases of lymphatic leucemia examined we found a chloride percentage in the plasma within the limits found for the diseases mentioned above. The mean values, however, lie somewhat above those found in the other cases. 5. In three cases of verified malignant neoplasia the sodium chloride percentage in the plasma also were found to lie within the physiological limits, though the mean value was placed near the lower limit. 6. The average of the sodium chloride content in the plasma is 0.61 per cent, but the amounts vary between 0.57 and 0.64 per cent. The changes in the chloride percentage in the blood will ap- pear from our previous tables. In Table X we have arranged the results according to the cell volumes found. In this way the effects of a change in the cell volume are most clearly demon- strated, and for the sake of comparison the corresponding chlo- ride contents per 100 gm. of plasma are given. Table X includes all our material, and cell volumes between 64 and 5 per cent are represented. The chlorides in 100 gm. of plasma vary between 0.59 and 0.63 per cent, while the chlorides in 100 gm. of blood vary between 0.44 and 0.59 per cent. The table emphasizes what we have already shown, that the chloride percentage in the plasma is not influenced in any way by variations in cell volume. On the other hand, it shows most clearly that the chloride percentage in the total blood increases when the cell volume drops. The variation is rather large, from 0.44 to 0.59 per cent. That the opposite condition, polycythemia, will show a blood chloride percentage lower than the normal, has to our knowledge never been observed, but appears with absolute certainty from our results. A. Norgaard and H. C. Gram 215. As our analyses have shown that the plasma percentage of chlorides is nearly constant, while the blood percentage of chlor- ides will increase when the cell volume decreases, the explanation of this phenomenon must be sought in the fact that the blood TABLE X. Cell volume. NaCl in blood. NaCl in plasma. per cent per cent per cent 64-63 0.44 0.60 62-61 60-59 58-57 56-55 54-53 52-51 50+49 0.46 0.62 4847 0.47 0.61 46-45 0.45 0.59 44-43 0.49 0.61 42-41 0.48 0.60 40-39 0.49 0.61 38-37 0.49 0.60 36-35 0.50 0.60 34-33 . 32-31 0.51 0.59 30-29 0.53 0.63 28-27 0.53 0.62 26-25 , 24-23 0.56 0.62 22-21 0.56 0.63 20-19 18-17 0.54 0.60 16-15 0.55 0.60 14-13 0.56 0.60 12-11 10-9 0.57 0.62 8- 7 6- 5 0.59 0.63 corpuscles contain a smaller percentage of chlorides than the plasma. If the blood is rich in corpuscles, polycythemia, it will be poor in plasma, thus making the total amount of chlorides low, 0.44 276 Chloride Content of Blood per cent; while on the other hand, a blood which is poor in cor- puscles, anemia, will contain more plasma, thus causing an in- crease of the chloride percentage of the whole blood up to 0.59 per cent. 5 We have not directly analyzed the content of chlorides in the corpuscles, but from our analyses of blood and plasma we may calculate this value in the following way: If the chloride concentration in the blood is called Cy, the chloride concentration in the plasma C>», and the chloride concentration in the corpuscles X, then the last value may be calculated from the formula: 100 - Cy = Cp* (100 — volume per cent) + (X - volume per cent) where all values excepting X are known. If in this equation we introduce the mean values for C,, C,, and volume per cent found in Table III of our fifteen normal cases, it will read as follows: 100 - 0.481 = 0.609 - (100 — 42.5) + X - 42.5 X = 0.31 per cent The content of chlorides in normal corpuscles then should be about 0.31 per cent. If in the above equation we introduce in succession the mean values from Tables IV to IX we get in this order the following chloride percentages in the corpuscles: 0.30, disturbed salt meta- bolism (?); 0.35, polycythemia; 0.33, simple anemia; 0.23, per- nicious anemia; 0.30, malignant tumors; and 0.30, leucemia. The only very pronounced variation in the salt content of the corpuscles is found in the cases of pernicious anemia. If in the equation used above we call X, C, (corpuscle-chloride concentration) it will read thus: 100 - Cy = Cp* (100 — volume per cent) + C,* volume per cent (1) In this formula one may, according to our experience, call Cy and Ce constant,’ introducing instead the values 0.61 and 0.31 per cent which gives the formula: _ 61—(0.3 + volume per cent) C, 100 (2) It is seen from equation (2) that there is a simple relation be- tween the cell volume and chloride content of the blood and this § Except, however, the cases of pernicious anemia where C, varies. } 5 A. Norgaard and H. C. Gram 277 relation will persist even if the blood is diluted with edema fluid. One cannot find a higher chloride percentage in hydremia than that which may be calculated from formula (2). The maximal chloride content in the blood will be obtained by putting volume per cent =9, in which case one finds: a which means that the chloride content of the plasma is the limit. A chloride content of the blood higher than the plasma per- centage could conceivably occur only under two conditions: (1) The blood is diluted with a liquid richer in sodium chloride than edema fluid. Such an occurrence in the body is improba- ble. (2) The blood corpuscles could store a larger amount of chlorides. We have never found this to be the case, the corpus- cle-chloride concentration being lower than the plasma-chloride concentration in all cases observed. ,The earlier determinations of the chloride content of whole blood have been a disappointment for the reason that the plasma concentration of the chlorides is nearly constant and the con- centration of chlorides in whole blood is a function of the cell volume. The introduction of the cell volume percentage of the blood in these calculations is of the utmost importance. This value plays an important réle in the estimation of such substances as are not evenly distributed between plasma and corpuscles. Of course this is even more essential as regards substances which are found only in plasma. SUMMARY. 1. A 3 per cent (isotonic) citrate solution is used to obtain plasma for analysis. 2. The cell volume percentage in the blood is determined and shown to be a factor to be considered when the. concentration of a substance is determined on whole blood. 3. In 52 cases of various types we find that the content of sodium chloride in the plasma is nearly constant, about 0.61 per cent. 4, The corresponding chloride determinations on whole blood show that these values vary greatly. It is shown that the chloride 278 Chloride Content of Blood percentage in blood increases when the cell volume percentage (and hemoglobin) drops, and vice versa; these changes following certain laws, which are formulated. 5. The chloride content of the blood corpuscles is nearly con- stant, about 0.31 per cent, the only serious divergence being found in pernicious anemia, tee the average content is calculated to be 0.23 per cent. 6. An analysis of the blood-chloride percentage generally does not give more information thana simple cell volume determination. Only blood-chloride percentages which vary distinctly from the values calculated by formula (2) may be considered pathological. BIBLIOGRAPHY. Austin, J. H., and Van Slyke, D. D., J. Biol. Chem., 1920, xli, 345. Bang, I., Mikromethoden, Wiesbaden, 1920. Fridericia, L. 8., J. Biol. Chem., 1920, xlii, 245. McLean, F. C., J. Exp. Med., 1915, xxii, 212, 366. McLean, F. C., and Van Slyke, D. D., J. Biol. Chem., 1915, xxi, 362. * Van Slyke, D. D., and Donleavy, J. J., J. Biol. Chem., 1919, xxxvii, 551. A NEW METHOD FOR THE DETERMINATION OF THE FIBRIN PERCENTAGE IN BLOOD AND PLASMA. By H. C. GRAM. (From the Medical Clinic of the University of Copenhagen, Copenhagen, Denmark.) (Received for publication, April 1, 1921.) The variations in the fibrin (fibrinogen) content of the blood have interested several observers, though the technique for these determinations has never been very good or convenient. The spontaneously coagulable matter of the blood may be deter- mined in two ways, by precipitating either the fibrinogen as such or as fibrin by the natural process of clotting. The last named method is the oldest and has generally been carried out so that the fibrin has been whipped out of a measured or weighed amount of blood (Hoppe-Seyler, 1; Samuely, 2), or+the blood has been shaken with small rods and the fibrin adhering to these removed (Dastre, 3). 86°0-61°0 OF 0-010 “pool ur 3 “Usy “"“(p) [eipuy a “sS0(] 5 ; = ge ae ae eee a IOP a “UL0qG-MO9 NT ie ms ule os we sein @ eWeek sien (0%) IosnIyy p “UdIpIIyD Joe. ee Oo Se ess eecvens eo we (Sz) unIss10g ” ” ” eo (F@) Jerpoy puv oronboog * *s ew ape es Cewia. “e. 1c. wees (82) TOV ule 5 7 Tite li Sore es Si." (se 60. Le ual (ZZ) Lleprouyag i : peas |anats Shc arene roman (tz) 1eu0my ‘ “ eb Nets tay Fe ee ae (og) snqary a ‘ Sere | ae ees “ore es + <0) Hegaig . 2 = Seen nt Ge Teele (81) youory 7 5 ‘Surddry Sangh) 6) a) eh eee, eee a) ie 8's) salts’ “e tahe (LT) osung 1e0> ULIqha : ieag @ag: |" cutee: i en ae (9) sagrorg OF 0-080 » 3 ee et (OT) [eqYso0g pur ‘uosepy ‘opddry AA ‘OP 0 Jnoqy i wo 7 eae | ee th eae ea (FI) opddry Sh‘ 0-620 iy BORIC ET 4 LOL EAD GOO yma ay |S Mae Me a tpternnc exer (€1) JoMopeaq LF'0 , “SzLqqey *s oy soy apd \6. 0) site, alia O%a) bec au a, alld ie. 6 e060 els (ST) L9] [NIN It'0 “UsBOULIGI “OTHBO "4x04 999 £29 SEOe SUL eTRL ee, 6. AS 0° vires Bb cece (11) ay Aoyy “suse d ut “uliqy 10 ‘[wuIay ‘onbruyoa y, “Ioqyny 938 }U00I0 7 aSBymoo10g uesouliqriT S]DUULY PUD Uayy JoULLON fo DUSDIG puD poorg Ur sobnjuarig waboursgry Pun Uwguy ‘I ATAVL 282 Fibrin Percentage in Blood I shall now describe the. technique for determining the fibrin content followed by me and afterwards the experiments, which have proved its reliability. About 4.5 ec. of venous blood are run into a graduated 5 ec. centrifuge tube (Oluf Thomsen) divided into one-tenth of a ce. and containing 0.5 ce. of 3 per cent sodium citrate. The citrated bloodis shaken and the blood adhering to the cork and upper part of the glass is wiped off. After the specimen has stood for some time the corpuscles have sedi- mented sufficiently to draw off 0.025 ec. of plasma for the platelet count (Thomsen, 28; Gram, 29, 30) and 0.4 cc. for the determination of the coagulation time (Gram, 31). The velocity of this sedimentation depends upon the cell volume percentage and the fibrinogen percentage in the plasma (Gram, 32). The specimen is then centrifuged for 90 minutes at 3,000 revolutions per minute and stopping very slowly. The tubes must be securely fixed in the corresponding holes of the centrifuge. The amount of citrated blood and precipitate is noted, the cell volume being calculated by the equation P-100 Volume percentage = B Ie B Precipitate in cc. Blood in ce. The clear cell-free citrated plasma is drawn off with a pipette, and 2 cc. are transferred into a 50 mm. wide cylindrical vessel, whose bottom is slightly rounded off against the sides. 9 cc. of 0.9 per cent sodium chloride and 2 cc. of 1 per cent CaCl, - 6H:0 solution are then added and the vessel is placed in the thermostat at 35° C. for 13 hours. When the glass is taken out the contents will always be found clotted, except possibly in very severe cases of hemophilia or melena neonatorum. By inclining and turning the glass around the clot will always loosen completely, not even leaving traces on the vessel. It is thrown out on several layers of filter paper on which it forms a jelly-like cake. The water is absorbed very quickly by the paper, leaving a round shining membrane, which may easily be detached with a glass rod when the top layer of paper is thrown into a jar of water. The detached membrane is kept in distilled water for 15 minutes and is then transferred into absolute alcohol for 5 minutes and finally into ether for another 5 minutes to dehydrate and extract lipoids. H. C. Gram 283 The hardened, dehydrated fibrin which resembles a piece of paper is gripped by small wire pincers of known weight and hung in the thermostat for some hours or in an oven for a shorter period. When the weight is constant the fibrin is weighed either on the analytical balance or on a fine torsion balance. The balance generally used only gave an accuracy of + mg. In the cylindrical glass vessel there is always left a small amount of liquid (diluted serum), which may be used as a control, since clotting either spontaneously or by addition of a little serum shows that the precipitation of the fibrinogen has not been complete. This control test is very fine, giving positive results with quantities of fibrin too small to influence the weight; 7. e., less than + mg. When the recalcified plasma was kept in the thermostat for the usual length of time the control never was found positive even in hemophilia. In order to calculate the fibrin percentage in plasma (Fp) and blood (Fp) a knowledge of the following values is necessary: Wfe = Weight in gm. of fibrin in 2 ee. of citrated plasma. Cb = Citrated blood (0.5 co.). C = Citrate (0.5 cc.). P = Cell precipitate in ce. The formula for calculating the percentage in pure plasma is: 7, — (Co = P)‘Wfe' 100 p= “(Cb = C —P)2 The corresponding formula for the percentage in pure blood is: ma (Cb — P): Wf2* 100 aS (Cb =C)-2 The technique in the first instance has been based upon a study of the coagulation time of citrated plasma on recalcification which has been published elsewhere (Gram, 31). The exactitude of the cell volume determinations has been dealt with in another publication (Norgaard and Gram, 33). In this paper we shall, therefore, only present the experiments concerning: (1) the proper recalcification of the citrated plasma; (2) the possibilities of fibrinolysis and fibrinogenolysis; (3) the accuracy of the method and the mean error on it; and (4) com- 284 Fibrin Percentage in Blood . parisons between my technique and the results gained by simple defibrination of blood and by determination of the fibrinogen after Whipple’s method of heat coagulation. Proper Recalcification—1 drop of 1 per cent CaCl6H,O solution to 0.1 ce. of plasma equals 1 cc. of the solution to 2 ce. of plasma, since the pipette used in the coagulation experiments gave 20 drops to the cc. It was found that the optimal recalcification of citrated plasma from a mixture of 0.5 ce. of 3 per cent citrate + 4.5 cc. of blood was TABLE II. Effects of Various Recalcification of Plasma Kept in the Thermostat for 13 Hours After Recalcification. peed I II III IV Vv Hecale | ibrin Fibrin Fibrin Fibrin Fibrin ees ce: peniaee noni centage (oR peceie ons pentage eae cantage poss pe tee plas Beery piesa, laste ce. 0.5 0.32 _ 0.31 +- 0.51 — 0.39 — 1.0 0.33 _ Or32 _ 0.34 — 0.51 _ 0.37 _ 15} 0.33 _ 0.32 _ 0.34 - 0.49 _ 0.39 _ 220 0.33 — 0.82 _ 0.34 — 0.49 _ 0.39 — 265 0.33 _ 0.32 -- Oso _ 0.51 — 0.37 _ 310) 0.32 _ 0.32 a 0.34 _ 0.51 _— 0.39 _ 344 0.33 _ 0.32 — 0.34 -— 0.49 _ 0.37 — 4.0 0.32 _ 0.32 -- 0.34 -—- 0.51 — 0.33 + 4.5 0.33 _ 0.32 _ 5.0 0.32 — 1 to 4 ce. of 1 per cent CaCl: 6H2O per 2 cc. The effects of a varying recalcification are found in Table II. In two cases the coagulation is not complete after 13 hours, the recalcification in these specimens being respectively 0.5 and 4 cc. per 2 cc. of citrated plasma. With recalcifications between 1 and 3.5 cc. of 1 per cent CaCl: 6H2O the result is always the same except for the experimental error. A recalcification of 2 ce. of 1 per cent CaCl, - 6H20 per 2 cc. of citrated plasma must be considered safe, which also has been proved in practice. H. C. Gram 285 TABLE III. Fibrin Percentages in Equally Recalcified'S pecimens from the Same Individual Left for Various Periods in the Thermostat. Time after recalcification (200. of lper cent CaCle. 6H:0). Fibrin percentage in plasma. Control. 5 No coagulation. aaa 10 “cc “ ++ 15 0.09 a 20 0.32 -- 25 0.32 _ 30 0.33 — 35 0.32 - 60 0.32 — TABLE IV. Double Specimens, of Which No. I Is Left for 13 Hours, and No. II for 24 Hours in the Thermostat after Recalcification. Fibrin percentage = Diagnosis. in plasma. Differ- No. ence I II 1 | Convalescence after pneumonia....... 0.39 0.39 0 PP MCUMIACIC LEVER i. nso. essen sceeas 0.55 0.55 0 a iecarcinoma of the liver. ..............4! 0.72 0.72 0 <2) 0 1E SITES Sa en 0.32 0.32 0 So ymipnabieleucemia.................2| 0:82 0.32 0 Pe eer CYLUCTNIG | oc case eet ee ee ed a oOo 0.32 0 7 0 ee eee: 0.33 0.33 0 8 ST Ge ia Ee ea 0.25 0.25 0 ERECTING eo cn acs «ciclo ce copes nema cte 0.55 0.55 0 10 bal SheNOSIS*PYLOF........<.2 2002 =: 0.82 0.82 0 Bee Bromo MOLVATCNTICIS ...........0262555 0.65 0.65 0 a eeaeriogig.er the liver............<--.-| 0.18 0.18 0 Soe ermMcIous ANCMIA:........5...+-.000+: 0.18 0.18 0 ce Le rr ae 0.25 0.25 0 15 | Fatty degeneration of the liver....... 0.08 0.08 0 iy yee Ghyewinemian.: yf i:. o.02...5.52.00..%2) ) 0:80 0.30 0 u7, | Chroni¢ constipation. ...:...).......+-. 0.29 0.31 | +0.02 Se INGUPASGNENID 6). 66 as ct tines nites ee eels 0.28 0.27 | —0.01 Se EGRIG MEDNTIGIS. 0.2 520k dss as ee ene 0.49 0.47 | —0.02 20 | Splenectomy (postoperative)........... 0.43 0.44 | +0.01 Mie ANCCHOSIS PYlOT . ¢...0.025.5 0. 0.202 cce sf OBE 0.39 | +0.02 Memreherginny SYPHILMA. 6 3 fs.(is2 624/52 8s 0.66 0.63 | —0.03 23 | Fatty degeneration of the liver....... 0.12 |,0.138 | +0.01 Pie | ESeudoleucemia:.../.<)sr. - ssa lee eee) (0239 0.41 | +0.02 ame Carcmomsa of the liver... ......../0<-. 0.07 0.07 0 26 NOUITA Doe cas oS re Mate cee eee one 0.27 0. 26 —0.01 286 Fibrin Percentage in Blood If we consider how coagulation progresses in a series of identical specimens of citrated plasma recalcified in the same way, we find the results shown in Table III. The coagulation is complete after 25 minutes, and the control is fine enough to show a residue of fibrin, which is too small to influence the weight of the precipitated fibrin. Possibilities of Fibrinolysis and Fibrinogenolysis by an Analysis of Double Specimens from the Same Person—The first question TABLE V. Analysis of Double Specimens of Which No. I Is Centrifuged at Once and No. IT after Standing 24 Hours at Room Temperature. Cell vol Fibrin | Speci- ell volume. | percentage in ae men Diagnosis. plasma. pen No ete 2 & i I II I Il eee ee Oe ee ee Ls) Normal 7 eee seer eee elle: 39) | 0:27 | 0:27 0 2 Chita sence 41 41 | 0.23 |,0.3 0 3° || Pebriculas(cancerty) sous 40 40 | 0.54 | 0.54 0 4 Se cia: SB etch cate | tn SOC 43 42 | 0.62 | 0.62 0 5 | Convalescense after rheumatic FAVORS tae cisciscy: Ss ees 38 38 | 0.32) | 0%3z2 0 6 || Pernicioussanemial. 4p eee eo 29 | 0.22 |- 0.22 0 |) Nepbritis:*anenia: 3. eer eee 23 23 | 0.66 | 0.66 0 Se | Binberbis:. asc oc ote eee ae 40 41 0.28 | 0.28 0 9 | Permicious anemia... see. 27 27 | 0.24} 0.24 0 NOV | interitis. sec 12% 2 nae 43 44 | 0.33 | 0.33 0 | aINiOTema aye) Saree as skvctc.0/ ee eae een ne cL 46 | 0.35 | 0.36 |+0.01 12 Abdominal carcinoma..........| 44 45- | 0.52 | 0.54 |4+0.02 IS. |ORenalviim Ore. pee See eee lO 26 | 0.62 | 0.60 |—0.02 a4” || Constipation «x -\6t EN Worse, ae) Oo 39 | 0.28 | 0.29 |4+0.01 Lier] INOrM Slee aoctoy aerate eae ee 42 | 0.28 | 0.30 |+0.02 was settled by letting Specimen I stand for 14 hours and No. II for 24 hours in the thermostat after recalcification (Table IV). This shows that an appreciable fibrinolysis does not take place in cell-free plasma within 24 hours. Specimens 12, 13, 15, 23, and 25 show that the low fibrin percentage found in some diseases of the liver is not due to fibrinolysis. In the same way a series of double specimens (Table V) showed that a dissolytion of the fibrinogen did not take place, when a specimen of citrated blood was left at room temperature for 24 hours before being centrifuged. H. C. Gram 287 Error of the Method.—First, a series of double specimens of citrated blood were analyzed and showed very concordant results (Table VI). TABLE VI. Analysis of 29 Simple Double Specimens. Cell volume. Diagnosis. I iS Gitte at kA g Ade pane ee ec) erOcOlins eerie: 6. .--2. ANG 5s. NGS teas 9 a aa q roy a) 9T'0 | STO |S9T 0 st LE | 8& | OF | 68 | 8E | £00 | 62 0 * 420 | ¥2°0 | 42°0 my IL ty ” OT » € » {46 UVP 0S ,, G aD) . SD 8 6L°0 | 810 | 02°0 | 12°0 | 02°0 5 88 | 68 | OF | 88} ZE | 200 | TE'0 | O£'O | €€°0 | FEO | 2E'O sy |b ap ” Ls. S55 Vo» ss 0G: 9 P (>) Es SOF ZEON Ol 0) 16Lc0n "6E 0 Ay Gy | GV | PY | SP | IP | 200 | T€ 0 | 62:0} O€ 0 | €€0| EEO = ep ” Glens 8» T°99 [0G », € qo ‘& 11°0 | 81°0 | 210 | 8T'0 S SP | 67 | ZF | 8P} 10'0 €e°0 | s€'0 | €€°0 | Se'O GL» Qe Sse NE yell RGus a G 1Z'0 | 020 | 61'0 | co'0 | 12 °0 Gr | €V | SV | S&F | tr | 200 | ZE'O |SPE'O | SEO | 8E'0| 8E'0 ‘IA [ JO [VAIOJUT (ZT “Geaq | ¢ “Gey 6Z “ULL (ZZ “ULL I vfs OP Dal Le 7a is GN espa Mm ae rat I A AI Il ra I uloly ee a EE ESSE N ns 3B od utl ie e =Gb “poo Uuls uov01 I I a =a SENOS TPO es aaaera ul se easel seen ed "(UdWO Mf ALY Jsay ay) ‘Uap a4W 8 PUY | SON) SUOSsad JDWAON 8 Ut Uugiy sof sishjoup paywaday TIX WIAVL H. C. Gram 293 It has also to be settled whether the large variations in the normal fibrin percentage occur in the same person or whether each normal person has a certain level of fibrin content, which is kept fairly constant. The last seems to be the case according to the results given in Table XII. The fibrin was determined on six women (Table XIII) at differ- ent times of the day in order to see whether any variations could be traced or whether the digestive leucocvtosis did influence the fibrin content. This does not seem to be the case, the variations being always very small and independent of the digestive leucocytosis, which is very pronounced in the cases of Nos. 3, 5, and 6. TABLE XIIl. Fibrin Percentage in Plasma, Cell Volume Percentage, and Number of Leuco- cytes at Different Times of the Day. Fibrin percentage in | Cell volume. Leucocytes per c.mm. 10a.m.|2p.m./6p.m.|10a.m.|/2p.m.|6p.m. |10a.m.|2p.m./6p.m, SS 8 eS ee OO ee OOO 1 0.33 | 0.33 39 38 * 7,500) 7,100 2 0.32 | 0.32 42 41 5,900} 7,700 3 | 0.33 | 0.33 38 40 6,500) 11,700 + 0.29 | 0.31 42 42 8,200) 9,900 5 0.35 | 0.36 | 0.37 44 43 43 6,500) 10,900) 11,800 6 0.21 | 0.22 | 0.21 42 40 41 5,400) 5,700} 9,100 SUMMARY. 1. A technique is described by which the fibrin percentage in blood and plasma may be determined by recalcification of 2 ce. of clear cell-free plasma obtained from 5 ce. of citrated blood (0.5 ce. of 3 per cent citrate + 4.5 cc. of blood). This includes a determi- nation of the cell volume. 2. On the same specimen one may determine the platelet count by Thomsen’s method and the coagulation time by a method indicated by the author. 3. The technique described is shown to fulfill the conditions exacted for a trustworthy method. 294 Fibrin Percentage in Blood 4. The results by this technique in the blood of normal men and women are put forward. When the stress is laid upon the fibrin percentage per 100 cc. of plasma, this is caused by the results found in diseases of the blood which together with other pathologi- cal results have been published in brief elsewhere. BIBLIOGRAPHY. . Hoppe-Seyler, F., Handbuch der physiologisch- und pathologisch- chemischen Analyse fiir Aertze und Studirende, Berlin, 8th edition, 1909. . Samuely, in Abderhalden, E., Handbuch der biochemischen Arbeits- methoden, Berlin and Vienna, 1910, ii, 375. . Dastre, quoted by Richet, C., Dictionnaire de physiologie, Paris, 1904, vi, 411. . Andral, Essai d’hématologie pathologique, Paris, 1845. . Doyon, M., Morel, A., and Kareff, N., Compt.rend. Soc. biol., 1906, lviii, 681. . Pfeiffer, I., Z. klin. Med., 1897, xxxili, 215. . Cullen, G. E., and Van Slyke, D. D., Proc. Soc. Exp. Biol. and Med., 1915-16, xiii, 197. . Bang, I., Mikromethoden, Wiesbaden, 1920. . Green, J. R., J. physiol., 1887, viii, 373. . Goodpasture, E. W., Bull. Johns Hopkins Hosp., 1914, xxv, 330. . Reyhe, Nachweis und Bestimmung des Fibrinogens, Inaugural disser- tation, Strassburg, 1898. . Doyon, M., Morel, A., and Péju, G., Compt. rend. Soc. biol., 1905, lvii, 657. . Frédérick, Récherches sur la coagulation du sang, Gand, Paris, and Leipsic, 1878. . Whipple, G. H., Am. J. Physiol., 1914, xxxiii, 50. . Miller, P. T., Beitr. chem. Physiol. u. Path., 1905, vi, 454. . Whipple, G. H., Mason, V. R., and Peightal, T. C., Bull. Johns Hopkins Hosp., 1913, xxiv, 207. . Bunge, quoted by Sahli, H., Z. klin. Med., 1905, lvi, 264. . Richet, C., Dictionnaire de physiologie. Fibrine, Paris, 1904, vi, 410. . Erben, F., Z. klin. Med., 1902, xlvii, 302. . Arthus, Précis de chimie physiologique, Paris, 1909. . Arronet, Quantitative Analyse des Menschenblutes, ete., Dissertation, Dorpat, 1887, 65. . Schneider, Die Zusammensetzung des Blutes, ete., Dissertation, Dor- pat, 1891, 22. . Biernacki, E., Z. klin. Med., 1897, xxxi, 279. . Becquerel and Rodier, Recherches sur la composition du sang, Paris, 1844, . Berggriin, Arch. Kinderh., 1895, xviii, 178. H. C. Gram 295 . Kriiger, Uber das Verhalten des ipizin Blutes, etc., Dissertation, Dorpat, 1886, 37. . Meek, W. J., Am. J. Physiol., 1912, xxx, 161. . Thomsen, O., Acta med. Scandinav., 1920, liii, 507. . Gram, H. C., Arch. Int. Med., 1920, xxv, 325. . Gram, H. C., Acta med. Scandinav., 1920-21, liv, 1. . Gram, H. C., Bull. Johns Hopkins Hosp., 1920, xxxi, 364 (see corrections in following issues). . Gram, H. C., Arch. Int. Med., 1921, xxviii, 312. . Norgaard, A., and Gram, H. C., J. Biol. Chem., 1921, xlix, 263. THE JOURNAL OF BIOLOGICAL CHEMISTY, VOL. XLIX, NO 2 A FURTHER STUDY OF THE RESPIRATORY PROCESSES IN MYA ARENARIA AND OTHER MARINE MOLLUSCA.* By J. B. COLLIP. (From the Marine Biological Laboratory, Woods Hole, and the Marine Biological Station, Saint Andrews, Canada.) (Received for publication, October 21, 1921.) INTRODUCTION. In a preliminary paper (1) the conclusion was tentatively advanced that Mya arenaria is a facultative anaerobic organism. This conclusion was reached as a result of repeated demonstra- tion of the fact that this pelecypod continues to produce carbon dioxide in considerable amounts and for long periods when kept under anaerobic conditions. Further investigation has shown that various other bivalved forms and also certain of the gas- tropods continue to excrete, or to store and excrete carbon diox- ide when the supply of oxygen in the medium is at a minimum. The object of the present investigation was to study quantita- tively this apparent manifestation of anaerobiosis in such highly organized forms. It was thought that quantitative data relating to rates of oxygen absorption under normal or experimental con- ditions, to the hydrogen ion concentration maintained in the celo- mic fluid under various circumstances, and to rates of carbon dioxide excretion under anaerobic conditions should throw some light on the problem. Berkeley (2) has recently confirmed Collip’s observations on carbon dioxide production in Mya arenaria under anaerobic conditions, and has demonstrated the same phenomenon in the two species Saxidomus gigantea and Paphia staminea. He has endeavored to relate the process to the decomposition of gly- cogen with the production of methane and carbon dioxide. His results, however, showed that glycogen did not disappear at a * This work was carried out under tenure of a Fellowship granted by The Rockefeller Foundation. : 297 298 Respiratory Processes in Mollusca more rapid rate in the anaerobic specimens than in the aerated controls except in the case of Saxidomus. He concluded, there- fore, that a disappearance of glycogen invariably accompanies anaerobiosis in Saxidomus, and that no disappearance of glycogen accompanies anaerobiosis in Mya and Paphia. It would seem most probable that the source of oxygen during exposure of a clam to anaerobic conditions is the same in one species as another. It would follow then from Berkeley’s results that glycogen is not likely to be the source of all the oxygen which goes to form carbon dioxide when the medium is oxygen-free. In view of the possible significance attached to glycogen as relating to the process, determinations of this tissue constituent were made in certain instances. Methods. The methods employed were as follows: Determination of the pH of Celomic Fluid. The pH of the celomic fluid was determined colorimetrically. Phenol red was the indicator used. The celomic fluid was drawn off without loss of carbon dioxide and delivered under paraffin oil into distilled water containing the adequate amount of phenol red. A pipette of narrow bore and containing a little paraffin oil was thrust deep into the pericardium, after an opening had been made just lateral to the hinge. The celomic fluid was aspirated into the pipette and 1 ce. delivered into 15 cc. of distilled water. The method of Cullen was then followed.! Oxygen Consumption. The rate of oxygen absorption of clams was determined by the ‘well known Winkler method. Carbon Dioxide Production. Carbon dioxide production was determined by the use of the Van Slyke (3) apparatus for determination of blood plasma car- bon dioxide. ‘The specimens to be studied were placed in fresh 1T am indebted to Dr. Cullen for furnishing me with the details of his method before publication. ee ee J. B. Collip 299 "sea water in museum jars of known capacity and then sealed under water. After a definite period Of time had elapsed, the jar was opened, the sea water was drained off, and the clams were removed to an open dish, and allowed to drain further. The water which was thus collected was added to the larger fraction and the total volume determined. The clams were at once opened up and the interior of the shells freed of tissue. The whole contents of the shells were then drained, gentle pressure being used to expel the liquid portion. The fluid which was obtained in this manner was a mixture of true celomic fluid and of sea water trapped in the mantle cavities. The carbon dioxide content of the sea water used in the experiment was determined before and after, also the carbon dioxide content of the composite fluid expressed from the clam tissue was determined before on control specimens and after on the fluid obtained as described. The drained clam tissue was weighed. The volume of water displaced by the shells was deter- mined and the volume displaced by the clams as a whole was ob- tained by subtracting from the volume capacity of the container used in the experiment the volume of water whtch was drained off from the clams when the experiment was concluded. Sufficient data were thus available to enable one to’ calculate the carbon dioxide production per 100 gm. of wet drained tissue per hour, a factor which was in every case determined and used for compara- tive purposes. The method just described is not without error, but it gives, nevertheless, approximate results. As was previously shown (1), the carbon dioxide produced by a clam when it is re- moved from its natural habitat may be in large part retained in the celomic fluid, becoming fixed as calcium bicarbonate. When one determines, therefore, the carbon dioxide content of the sea water or celomic fluid in an experiment, one is dealing with carbon dioxide fixed as bicarbonate for the most part. The carbon diox- ide production is obtained by taking 50 per cent of that present in the noted bicarbonate increase. As the amount of free carbon dioxide is relatively very small in comparison with that held as bicarbonate no attempt was made to measure it directly. If the free carbon dioxide were determined also, the results throughout would be slightly higher. The above method for determining carbon dioxide production - was checked up by determining the carbon dioxide excretion 300 Respiratory Processes in Mollusea when clams were acrated with carbon dioxide-free. nitrogen and * the gas produced collectedin standard baryta which was afterwards titrated to phenolphthalein with standard acid.’ The results checked satisfactorily. ‘Results. The pH of the Celomic Fluid Under Anaerobic Conditions. It was found that no appreciable change in the pH of the celo- mic fluid of Mya or Venus occurred when they were kept for days immersed in paraffin oil, in a nitrogen atmosphere, in sea water in a sealed container, or simply exposed to atmospheric air. The pH for both normal and experimental specimens as determined by the method outlined was 7.8 to 7.9. Under all the experimental conditions imposed, the total carbon dioxide of the celomic fluid increased several fold. Rate of Oxygen Absorption. Considerable variation was found in the rate of oxygen con- sumption at constant temperature. Some forms, such as Venus mercenaria, Which are capable of closing the valves completely, seem at times to use little or no oxygen, but one can be assured that oxidation is going on in the specimen, due to the fact that one can demonstrate a constant production of carbon dioxide, which is under these circumstances retained in the celomic fluid. Mitchell (4) found that the oxygen absorption of Venus and the oyster was very low unless the valves were opened slightly and a current of water maintained throughout the mantle cavities. In the case of Mya, the rate of oxygen absorption, the tempera- ture being kept constant, is largely determined by the position of the siphon and the rate of flow of water through it. It is quite possible for the form to drawits siphon in completely and when this is done the rate of oxygen absorption falls almost to zero. It is, however, only under exceptional circumstances that specimens of Mya do not ventilate to a fair degree, while quite the reverse is true of Venus. It was not uncommon to find the carbon dioxide content of the celomic fluid of specimens of Venus mercenaria kept in running sea water in the laboratory for a few days over 30 volumes per cent, whereas the carbon dioxide content of these J. B. Collip 301 forms when they are perfectly fresh is 6.5 volumes per cent. Mya arenaria, on the contrary, may be Kept in the laboratory in running sea water for days without the carbon dioxide increasing in amount in the celomic fluid. Thisis due solely to the fact that the former type closes its valves and thereby prevents a free exchange of gases through the membranes of the gills and mantle, while the latter adapts itself to the new conditions, and ventilates freely. It was found by Mitchell (4) that the rate of oxygen absorption by oysters and clams varied directly as the temperature. This has been confirmed for Mya. It is difficult, however, to relate the results obtained directly to temperature changes. For example, the oxygen absorption for six specimens of Mya at 14°C. was found to be 1.40 cc. per 100 gm. drained clam tissue per hour. The same clams were then placed.in fresh sea water at 26°C., and the oxygen absorption rate was found to be 11.92 cc. per 100 gm. per hour. On being replaced in fresh sea water at 14°C., the rate of oxygen absorption fell to 7.93 ec. The estimation was for periods of 1 hour in each case. In this instance, the siphons were not ex- tended at first, but were well extended at the higher temperature, and fairly well in the last determination. The highest rate of oxygen consumption observed for fresh specimens of Mya at 14°C. was 5.78 cc. per 100 gm. per hour, and the lowest 1.40 ce. per 100 gm. per hour. The usual result, however, for specimens with siphons moderately extended was between 3 and 4 cc. per 100 gm. per hour. The oxygen absorption by specimens of Venus mercenaria under laboratory conditions was in all instances almost negligible. This was due, as will be shown later, to the valves being kept tightly closed. Effect of Cleansing the Shells upon the Rate of Oxygen Absorption. Owing to the possibility of organisms attached to the shells using up oxygen, the rate of oxygen absorption for specimens of Mya arenaria, which had had the shells sterilized by sponging with an alcoholic solution of bichloride of mercury, was deter- mined. For some hours following the cleansing of the shells the ‘oxygen absorption was very low, being from 0.2 to 0.4 cc. per 100 gm. per hour. That the result was not due to the killing of organ- isms on the shell was shown by the fact that the rate of oxygen 302 Respiratory Processes in Mollusca absorption came back to normal in 2 days time. The specimens for a time immediately following the cleansing operation drew in their siphons as completely as possible, and as they did not ven- tilate the mantle cavity and the gills they did not absorb oxygen. The presence of traces of mercury in the water would no doubt be a factor in causing the siphons to be withdrawn for a long period. This would appear to be the case, because cleansing the shells by means of sand paper caused a comparatively slight fall in the oxygen consumption. In one experiment, the rate of oxygen consumption immediately following sand paper cleansing was 1.9 ce. per 100 gm. per hour, as compared with 3.17 ce. per 100 gm. per hour in the controls. Oxygen Fixation by the Shell. Mitchell (4) found that a certain amount of oxygen was fixed by the shells of clams and oysters. The amount of oxygen which was removed from fresh sea water by the uncleaned shells of specimens of Mya studied was found to be very small. Shells displacing 100 cc. of water would cause the oxygen of fresh sea water to disappear at the rate of 0.5 to 0.9 cc. per hour. For very detailed work on oxygen consumption, the effect of the shell must be determined and the results corrected accordingly, as Mitchell has pointed out. Effect of Breaking the Shell upon the Rate of Oxygen Absorption. If the valves of specimens of Venus mercenaria be carefully cracked along the margin, the rate of oxygen absorption rises immediately. The rafe of oxygen absorption in one instance for specimens of Venus was 0.051 ce. per 100 gm. per hour, at 20°C., while the rate for specimens with shells cracked was 1.78 ce. per 100 gm. per hour. Similarly, specimens of Mya with a low rate of oxygen absorption, as a result of bichloride cleansing, can be made to absorb oxygen much faster by cracking the shells. In one such experiment, the rate of oxygen absorption in the bichloride cleansed specimen was 0.198 cc. per 100gm. per hour. After crack- ing the shells and allowing ventilation of the mantle cavity, the rate rose to 0.822 cc. per 100 gm. per hour. J. B. Collip 303 Rate of Carbon Dioxide Production under Anaerobic Conditions. _ A large number of experiments Were carried out to determine the rate of carbon dioxide production by specimens of Mya arenaria which were deprived of dissolved oxygen in the medium. These experiments were carried out under widely different conditions. At first, attempts were made to use oxygen-free sea water, but it was early noted that the sea water in which specimens were sealed up became oxygen-free within a very short period, espe- TABLE I. CO2 Production by Mya arenaria. Temperature....... 31°C. 20°C. 14°C. 8°C. aration of CO: per | Duration} COz2 per | Duration} COz per | Duration | CO: per re ese |e eae | aa eee hrs. ce, hrs. ce. hrs. ce. hrs, ce, 4.50 3.60 21.50 2.44 46 1.95 18 1.26 6.50 3.86 18 298 70 2.52 21 1.12 5.00 5.59 17 2.30 122 1.53 191 0.87 4.40 4.37 19 2.34 124 1.60 192 0.84 43 2.58 146 pie te) 192 0.86 13 2205 44 1.41 192 0.99 13 2.44 45 1.18 18 3.54 44.30'| 0.70 15 2.47 45 0.89 1 2.08 45 1.09 43 1.68 69 1.42 21 2.31 25 0.57 20 1.92 24 0.74 7 2.43 Average..... 4.35 2.43 1.29 0.99 cially if the volume of water was small. As many experiments were carried on for days and as the oxygen available in the medium would be removed within the first few hours, the error introduced by the use of this method was slight. The results of this series . of experiments are shown in Table I. While there is a certain amount of variation in the rate of carbon dioxide production as determined by this method, yet this is not surprising, since the periods over which the anaerobic condition was maintained varied from a few hours to several days. Also the temperature could not a gr Mis 4 304 Respiratory Processes in Mollusca be maintained absolutely constant, there being a variation of at least 1 +°C. The volume of water was, in some instances, large relative to the volume of clam tissue, while in other experiments the specimen jar was packed full of clams, and the water simply filled the interstices. Effect of Temperature upon the Rate of Carbon Dioxide Production under Anaerobic Conditions. The rate of carbon dioxide production under anaerobic condi- tions was determined at the following approximate temperatures: 8°C., the temperature of the ice box; 14°C., the temperature of © running sea water in the laboratory at Saint Andrews; 20°C., the temperature of running sea water in the laboratory at Woods Hole; 31°C., the temperature of an improvised water thermostat. The metabolic rate, as is shown by the results recorded in Table I, is a direct function of the temperature. Effect of Temperature upon Survival Time under Anaerobic Conditions. | A general correlation exists between the temperature and the length of time clams may be kept in good condition in an anae- ~ robic medium. At 31°C., the survival time is about 24 hours; at 14°C., about 8 days; and at very low temperatures, the speci- mens may be kept for weeks. Effect of Potassium Cyanide upon Oxygen Absorption and Carbon Dioxide Production. It was found that the rate of oxygen absorption was depressed by the presence of a small amount of potassium cyanide in the medium. The rate of carbon dioxide production under anaerobic conditions was likewise depressed by potassium cyanide. A con- centration of 1 part of KCN in 2,500 of sea water caused the CO» production to fall from 2.08 ec. per 100 gm. per hour, to 1.02 ce. Greater concentration depressed the CO. production still more, while lower concentrations had less effect. A point of interest in the effect of potassium cyanide upon the clam is that it tends to lower the difference in the concentration of CQO, in the celomic fluid and that of the outside water. Unless Ba A 3 Collip 305 an experiment were run for several days, it was found that the concentration of CO, in the celémic fluid was approximately 10 volumes per cent higher than that in the water of the container. In the presence of potassium cyanide, the rate of CO. production falls, and there is a tendency for the concentration of COs. in the medium to approach that maintained in the celomic fluid. Effect of Submersion in -Distilled Water upon the Rate of Carbon Dioxide Production under Anaerobic Conditions. As was previously shown (1) clams may be kept in distilled water for varying periods and under these circumstances, they yield their salt to the water very slowly. The rate of carbon dioxide pro- duction by specimens of Mya kept in distilled water, was deter- mined and it was found that the rate was greatly depressed. In one instance, a result of 1.02 cc. of CO. per 100 gm. per hour was obtained as compared with 2.44 ec. per 100 gm. per hour in the control. In Sycotypus the large whelk, the rate of CO, production fell in distilled water to 0.39 ec. per 100 gm. per hour from a normal of 1.10 ce. , = Pome sleet a eee * eg ants. satire = ti I IS Rn eS Rate of Oxygen Absorption Following Exposure to Anaerobic } Conditions. The rate of oxygen absorption by specimens of Mya which had been kept for some days under anaerobic conditions was deter- mined. It was found that immediately the specimens were transferred from the oxygen-free water to fresh sea water that the rate of oxygen absorption far exceeded the normal. Also the rate of absorption while far above normal at first, gradually fell until if it reached the normal level. It required 3 days in some instances before the normal rate was regained. The results illustrating this effect are shown in Table II. The rate of oxygen absorption following an anaerobic period varied directly with the temperature. Thus in one instance, after 4 days of oxygen want, the rate of oxygen absorption at 5°C. was 7.06 cc. per 100 gm. per hour; at 14°C. it was 15.10 cc. per 100 gm. per hour; while at mae: it was 21.20 ec. per 100 gm. per hour. 306 Respiratory Processes in Mollusea TABLE II. Oxygen Absorption Following Period of Anaerobic Condition. Beaper |, “Duration |» |[Rate ot eet) | cee 1 6 days at 14°C. 14.62 2 leet tome dikes 7.24 3 ne) ek AES 12.25 I ole i eae OR a 19.80 5 Pele oe amet eB 21.24 6 Ace Call Aci ce 6.99 At. 52G@: set) 146 7h 6 days at 14°C. 17.60 8.46 Aerated 24 hrs. 8 3 days at 14°C. 19.61 11.50 Aerated 1 hr. 11.26 a 2 hrs. 9 | 4 days at 14°C. 14.24 10 Bie oe a les Tails} Aerated 3 hrs. 3.20 ec 200 ess} a 44 “ 11 9 days at 8°C. 14.40 12.84 Aerated 1 hr. a20 ss 18 hrs. 7.38 o 2A 3.24 ce AS. acs Glycogen Content Before and After Periods of Oxygen Deficiency. It was found that the glycogen content of specimens of Mya arenaria gradually fell after they were placed in the laboratory tanks. The glycogen content of perfectly fresh specimens cal- culated as per cent of wet drained tissue may run over 11 per cent. In specimens left in the laboratory tanks for 2 weeks, this was not observed to fall below 2 per cent. Specimens with a glycogen content of 2 per cent sealed up for a period of 6 days at a tempera- ture of 18°C. were found to have only 0.25 to 0.30 per cent glycogen. Berkeley (2) has shown that the glycogen content of clams falls under anaerobic as well as aerobic conditions in the laboratory. DISCUSSION. The fact that no appreciable change in the hydrogen ion con- centration of the celomic fluid could be detected even after several J. B. Collip 307 days exposure of specimens of Mya arenaria and Venus mer- cenaria to abnormal conditions shows that these forms are capable of regulating the reaction of the body fluids quite as well as the higher forms. As was previously shown (1) the great increase in the carbon dioxide content of molluscan celomic fluid which is so readily brought about is due for the most part to increase in calcium bicarbonate. The source of the calcium is probably twofold; namely, the reserves in the liver and the shell itself. As a result of such forms being able to adjust their acid-base balance within very wide limits, a very simple method for determining metabolic rates under varying conditions is available. It was thought probable that the cause of death in many marine forms on exposure to anaerobic conditions might be duenotsomuch to oxygen deficiency, as to a change in body fluid reaction due to accumulation .of acid end-products. Attempts were therefore made to prolong the survival time of many forms by adding pre- cipitated calcium carbonate to the sea water in which the speci- mens were sealed up. Little success was attained however. The only forms which could be kept sealed up in oxygen-free sea water for long periods were the various calcareous shelled mollusks of both the pelecypod and gastropod types. As calcareous shelled mollusks such as Mya continue to pro- duce carbon dioxide at a fairly uniform rate under anaerobic con- ditions, and as this rate is directly dependent upon temperature and is depressed by the presence of potassium cyanide it is evident that oxidation still proceeds in the tissues even though the supply of oxygen from without is completely cut off. The question there- fore arises: What is the source of this oxygen? Does it come from the breaking down of glycogen or some other food principle with the production of CO, and a paraffin compound such as meth- une, or is there some compound in the tissue of the nature of an organic peroxide which is capable of supplying oxygen as it is re- quired, and which is itself reformed by the taking up of dissolved oxygen from the medium. The results which are herein reported, taken in conjunction with the negative findings of Berkeley as re- lating to a glycogen source for the oxygen would lend a measure of support to the latter hypothesis. It is difficult to determine with great accuracy the carbon dioxide production by specimens of Mya for very short periods of time, whereas oxygen absorption 308 Respiratory Processes in Mollusca can be measured with a fair degree of accuracy over periods of only a few minutes duration. There was no indication, however, in the experiments in which carbon dioxide production immediately following a period of oxygen want was determined, cf a great increase in the rate of CO, production parallel with the greatly increased rate of oxygen absorption. This point was also tested in another manner. Three specimens of Mya were aerated with pure nitrogen gas for a period of 30 hours, the carbon dioxide which was excreted, being collected in standard baryta. Air rendered CQO.-free, was then passed through the system. The CO, production for the anaerobic periods was 2.60 cc. per 100 gm. per hour, while it increased to 4.27 ce. when the oxygen was re- stored. Unfortunately, due to lack of nitrogen gas, this method could not at this time be pursued further. As the normal rate of oxygen consumption was higher than the average rate of CO: production during periods of oxygen want it is probable that the metabolic rate may be slightly less in the latter case than in the former, and that a slight increase in the metabolic rate might therefore be expected, following the renewal of the oxygen supply. However, the rate of oxygen absorption for some time following renewal of the oxygen supply after anaerobic periods was far in excess of any observed increase in CO, production. Further data on the CO, production must be obtained to make this point abso- lutely definite. This phenomenon points, in our opinion, to the re- building of a peroxide in the tissue or to the recuperation of some substance which is the source of the oxygen during anaerobic periods. It would appear then, while from a superficial view-point Mya arenaria and allied niolluscan forms are facultative anaerobic organisms, that they are not anaerobic in the strict sense of the term. They can live for varying periods of time in the absence of outside oxygen, but their survival time while long, is neverthe- less limited, and is directly proportional to the temperature. A possible explanation of the phenomenon of apparent anaerobiosis is that the tissues of these forms are capable of storing a large amount of oxygen which is readily available for the metabolic needs. The time of death under anaerobic conditions one would expect to be coincident with the disappearance of the oxygen store. ( } { | “ae J. B. Collip 309 No evidence was obtained which would indicate the exact nature of the substance which takes up Oxygen and holds it for the use of the tissues. That it is not in the celomic fluid or excreted tothe water in the container was shown by delivering celomic fluid and water used in the experiment into water of known oxygen con- tent and noting if any oxygen was fixed by the fluid added. The Winkler method was used and no appreciable change in oxygen content was demonstrated. It is therefore probable that the oxygen holding compound is present in the tissue only. While the results so far obtained lead us to suggest the above explanation of the phenomenon, sufficient quantitative data have not yet been brought forward to justify one in saying that this is the only way in which the oxygen used by the clam tissue during periods of imposed anaerobic conditions is supplied. CONCLUSIONS. Calcareous shelled mollusks by virtue of being able to use the calcium reserves of the liver and the shell exhibit the power to regulate their acid-base equilibrium with such precision that no change in the pH index is brought about even when they are sub- jected to most abnormal conditions. ; Mya arenaria, when placed under anaerobic conditions will sur- vive for a period of time which is dependent upon the temperature. During this anaerobie period, carbon dioxide is produced at a uniform rate and this rate is accelerated by raising the tempera- ture and depressed by lowering the temperature. The graph illustrating the temperature effect approximates to a straight line. Glycogen disappears from the tissues during anaerobic periods. Potassium cyanide depresses the rate of carbon dioxide production under anaerobic conditions as well as depressing the oxygen con- sumption under normal circumstances. This points to the carbon dioxide production under anaerobic conditions as being due to a series of oxidations. The rate of oxygen absorption immediately following imposed anaerobic periods is much higher than normal. The normal rate is regained gradually. It is suggested that these forms have a store of available oxygen in their tissue which suffices to supply the necessary oxygen when the outside supply is cut off. 310 Respiratory Processes in Mollusca These forms are therefore unique in two respects: they can re- tain if necessary a large part of the carbon dioxide produced under aerobic or anaerobic conditions by virtue of their ability to adjust their acid-base equilibrium at widely varying levels; and they have a source of oxygen which is available in the tissue for meta- bolic needs during long periods of oxygen insufficiency in the enveloping medium. My thanks are due to the Biological Board of Canada for af- fording facilities for the carrying out of a part of the work herein reported at the Marine Biological Station, Saint Andrews, Canada. BIBLIOGRAPHY. . Collip, J. B., J. Biol. Chem., 1920-21, xlv, 23. . Berkeley, C., J. Biol. Chem., 1921, xlvi, 579. . Van Slyke, D. D., J. Biol. Chem., 1917, xxx, 347. . Mitchell, C., Rep. U. S. Bureau Fisheries, 1912, 207. mm CO ND SULFATES IN BLOOD. By W. DENIS. (From the Laboratory of Physiological Chemistry, Tulane University Medical School, New Orleans.) (Received for publication, October 21, 1921.) Sulfates exist in blood in extremely minute amounts, and as the sulfate ion is not known to possess any striking physiologi- eal properties but little experimental work has been carried out on this phase of blood chemistry. Data concerning the sulfate content of blood, obtained by an analysis of the ash, include the sulfur present in combination in the protein molecule and have therefore but a limited usefulness in a study of the inorganic sulfur of this fluid. In fact, the only figures for inorganic sul- fates of blood which I have been able to find after a search of the literature are those published by De Boer.! This investigator furnishes figures for horse blood, the SO, content of which he finds to be 0.02123 per cent, ‘“‘an amount which is likely to vary in the case of different animals and more particularly in different periods of feeding.”” De Boer’s results were obtained by means of a sedimentation method proposed by Hamburger? in which the volume of barium sulfate resulting from the treatment of the sulfate-containing liquid with hydrochloric acid and barium chlo- ride is measured after centrifuging in a special calibrated tube. Protein-free filtrates were obtained by filtration through a cel- loidin filter under a pressure of five atmospheres. The deter- mination of small amounts of sulfate by means of the turbidi- meter has been in use among technical chemists for many years, and I have found it possible after a considerable amount of ex- perimental work to apply the same analytical principle in the determination of sulfates in blood and serum, although on ac- count of the exceedingly small amount of material to be deter- 1 De Boer, S., J. Physiol., 1917, li, 211. 2? Hamburger, H. J., Biochem. Z., 1916, lxxvii, 168. 311 THE JOURNAL OF BIOLOGICAL CHEMISITY, VOL. XLIX, NO. 2 BL2 Sulfates in Blood mined it has been necessary to substitute the nephelometer for the turbidimeter. Briefly stated the procedure consists in the removal of protein by means of a solution of mercuric chloride and hydrochloric acid, the formation of colloidal barium sulfate in the filtrate, and the measurement of the density of this precipitate by means of the nephelometer. My preliminary experiments were carried out on beef blood, but when after a satisfactory technique had been attained I attempted to apply this method to human blood I was unable to obtain results, because the sulfate content of human blood is but a fraction of that contained in the blood of any species of animal examined. The method as applied successfully to the blood of a variety of animals (see Table IT) is as follows: The blood is collected by venous puncture and coagulation prevented by means of powdered sodium citrate (80 mg. per 10 ce. of blood) in the manner usually followed in the case of specimens intended for chemical examination. To 10 ce. of blood or plasma contained in a 200 cc. Erlenmeyer flask are added an equal volume of 0.02 N hydrochloric acid and after an interval of 5 minutes 30 ce. of a 5 per cent solution of mercuric chloride containing 5 ec. of concentrated hydrochloric acid (sp. gr. 1.178) per liter. After vigorous shaking the mixture is allowed to stand for 1 hour and is then filtered through a dry 11 ee. filter paper. Filtration is fairly rapid and the filtrate should be absolutely clear. For this filtration the use of a high grade of ‘‘ashless”’ filter paper is essential; I have used What- man’s No. 44, which has been found free from all traces of sulfate. For the determination of total inorganic sulfates 10 ce. of the clear filtrate (equivalent to 2 ec. of whole blood or plasma) are pipetted into a 100 ee. beaker, and to this are added 5 ee. of a 1.0 per cent solution of ammonium nitrate, and with stirring 5 ce. of a 1.0 per cent solution of barium chloride containing 5 cc. of concentrated hydrochloric acid per liter. After a period of 10 minutes the colloidal suspension of barium chloride is compared to a standard which has been prepared simultaneously with the unknown in the following manner: To 10 ee. of a standard solu- tion of potassium sulfate (equivalent to 0.10 mg. of sulfur) are W. Denis a3 added 10 cc. of the acid mercuric chloride solution, 10 cc. of 1.0 per cent ammonium nitrate, and 10 ce. of 5 per cent barium chloride. To calculate the results divide the reading of the standard (usually 20) by the reading of the unknown and multiply the dividend by 50. This will give the result expressed as milligrams of sulfur per 100 ce. of blood. To determine inorganic sulfates in normal human blood the procedure is as follows: 5 cc. of oxalated blood or plasma are treated with 5 cc. of 0.1 n hydrochloric acid, 5 ce. of 5 per cent mercuric chloride solution, and 0.3 gm. of finely powdered mer- euric chloride. The mixture is then shaken vigorously for 5 minutes, and at intervals for 1 hour, and is then poured on a small dry ‘‘ashless”’ filter. 5 cc. of this filtrate, which should of course be absolutely clear, are treated with 1 ce. of a 1.0 per cent solution of ammonium nitrate and 1 cc. of the acidified barium chloride solution described above, and the turbidity so produced is compared after an interval of 10 minutes with a standard prepared by adding to 10 cc. of a standard solution of potassium sulfate 10 cc. of 5 per cent mercuric chloride solution, 4 cc. of 1.0 per cent ammonium nitrate, and 4 cc. of 1.0 per cent barium | chloride solution. While the technique described above gives excellent results with normal material it will be found that with pathological specimens the amount of sulfate present is sometimes so much increased that it is impossible to obtain a colloidal suspension of barium sulfate which will not precipitate. In such eases it is desirable if a sufficient amount of material is available to use the technique described for use with animal blood. As the sulfate concentration of human blood shows greater variations than does the blood of animals it is usually desirable to make use of three standard solutions of potassium sulfate of which 10 ce. are equivalent to 0.1, 0.05, and 0.03 mg. of sulfur respectively. Ammonium nitrate has been used ssdeate it facilitates the formation of colloidal suspensions of barium sulfate. Mercuric chloride was chosen after a trial of the more commonly used pro- tein precipitants as it appears to be the only one of these reagents which does not in some way interfere with subsequent precipita- tions of barium sulfate. 314 Sulfates in Blood All chemicals used should of course be tested for the presence of sulfates. All specimens of hydrochloric acid and ammonium nitrate examined have proved free from this impurity, but sev- eral samples of c. Pp. mercuric chloride have been found to con- tain appreciable amounts. Mercuric chloride can easily be purified so that it no longer gives precipitates with acidified barium chloride solutions by several (from three to five) recrys- tallizations from hot water. The standard solutions of potassium sulfate were made from the recrystallized salt. It is most convenient to prepare a stock solution containing 5.4870 gm. of KsSO, per liter, 1 ce. of which is equivalent to 1 mg. of sulfur, and from this strong solution to prepare by suitable dilution the three weaker standards. The procedure described above is believed to furnish a method of measuring the inorganic sulfates of the blood, with an accuracy of approximately 95 per cent. It would seem probable that other forms of sulfate analogous to the ethereal sulfate and neu- tral sulfur of urine exist in blood.’ I have carried out a number of experiments to obtain evidence of the presence of these forms of sulfur in blood but have invariably obtained negative results. In an attempt to demonstrate the existence of ethereal sul- fates in blood 10 ce. portions of the blood filtrate deproteinized by means of mercuric chloride were heated (after the addition of 1 ec. of concentrated hydrochloric acid) in a boiling water bath for periods of from } to 2 hours. After cooling, an amount of alkali just sufficient to neutralize the acid was added and the mixture then treated with an acid solution of barium chloride. This procedure has been carried out on a considerable number of samples of both human and animal blood, but in no case have I been able to obtain evidence of the existence of conjugated compounds of sulfuric acid, as shown by an increase in the sul- fates after the above acid treatment. In my search for the presence of sulfur compounds existing in the form of ‘‘neutral non-protein sulfur’? I have oxidized portions of the mercuric chloride filtrate with potassium chlorate and with nitric acid, without, however, obtaining evidence of any increase in the amount of barium sulfate precipitate obtained after this oxidation. Kahn, M., Proc. Soc. Exp. Biol. and Med., 1918-19, xvi, 139. W. Denis 315 While the above results can scarcely be taken to furnish abso- lute and final proof of the non-existence of conjugated forms of TABLE I. Recovery of Potassium Sulfate Added to Blood. S per 100 ce. Kind of blood. ; ae pele Sulfate added. Theory. Found. mg mg. mg. mg. Renae Bed iia, ein os 1.88 0.5 2.38 2.35 NS eee ee 1.88 1.0 2.88 2.97 7S hd 1.88 2.0 3.00 3.92 “SAGA See 1.88 2.0 3.88 3.90 Th at serene 1.88 4.0 5.88 5.81 Human 0.9 Ors 1.20 1.18 Rr hors oS Scie 0.9 0.5 1.40 1.38 Gime ene hos. Socien 0.9 1.0 1.90 1.90 ce cae, oh 0.9 4.0 4.90 5.00 TABLE II. Inorganic Sulfates in the Blood of Animals. Animal. s Bee Remarks. ™g. Bes: «ois ace oo ois 2s 2.6 Average of 18 specimens. a 1.8 Minimum “ 18 a SL ed oe fe hoe a od 8 0 mys see Maximum “ 18 3 TSE eed ca pea eee 1.8 Beg aEe eit crass ssa oso. 20) ok Oe a5 OD oe See ee 4.0 oS 2.8 oO rat | V2 hod) eee 22 TRI ood 3.2 OS) 2.0 2 LS 2.5 0 il ros 2.1 SD eee ee 3.9 sulfur in blood, they at least indicate that if sulfur compounds of this type do exist in this fluid the amounts present must be exceed- ingly minute. 316 Sulfates in Blood TABLE II. Inorganic Sulfates in Human Blood. Inorganic SOu. Non-protein No. Diagnosis. S per 100 cc. nitrogen per of blood. 100 ec. of blood. mg. mg. 1 Normal. O25. 2 ss 0.5 3 ss 0.6 4 €é 0.7 5 Kg 0.7 6 “ 0.9 7 es 1.0 8 af 1.0 9 sé 10) 10 se ite 11 Pregnancy. 133 30 12 og 1.0 28 { 13 Nephritis. 2 41 14 Pellagra. 0.7 25 15 Hyperthyroidism. 0.6 32 16 Cardiorenal disease. 0.7 32 17 Nephritis. 0.6 35 18 J 1.6 35 19 Leucemia. 7.0 68 20 Nephritis. 3.5 55 21 Epilepsy. 0.5 39 , Pipe Intestinal obstruction. 320 45 AB Diabetes. 120 24 Nephritis. 3.7 75 25 € 2.0 38 26 " 3.0 55 27 ef PATE 60 28 ‘‘ Be 65 29 bas 3.8 62 30 Cardiorenal disease. 12 40 31 bd oe 0.7 33 32 “ “ i) 3 33 33 . si 0.7 28 34 re tg 1.0 35 ao Uremia. 5.0 174 36 ef 4.6 80 37 ae 6.2 75 38 4g 8.0 142 39 re 123 187 40 = 12.0 210 16.0 200 W. Denis S17 In Table I are shown the zesults obtained when various amounts of potassium sulfate are added to beef and to human blood. ‘Table II contains figures obtained on the blood of vari- ous animals and in Table III are collected the results of the examination of human blood, both normal and _ pathological. SUMMARY. The inorganic sulfates of blood, as determined by a new and simple method, a description of which is given is found to amount to from 1.8 to 4.0 mg. of sulfur per 100 ee. of blood in the case of various species of animals and from 1.0 to 0.5 mg. in normal human blood. In nephritics with nitrogen retention there is also found a retention of inorganic sulfate, figures as high as 12 and 16 mg. having been obtained. Experiments made with a view to obtaining evidence of the existence in blood of conjugated compounds of sulfuric acid, and of bodies analogous to the neutral sulfur fraction of urine, have given negative results, indicating the probable non-existence of this class of bodies or the fact that they are present in amounts so minute that their presence cannot be detected by the methods now in use, ; SOME OBSERVATIONS ON CREATINE FORMATION IN A CASE OF PROGRESSIVE PSEUDOHYPERTROPHIC MUSCULAR DYSTROPHY. By R. B. GIBSON anp FRANCES T. MARTIN. (From the Chemical Research Laboratory of the Department of Theory and Practice of Medicine and Clinical Medicine, in Cooperation with the Department of Home Economics and the Graduate College, the State University of Iowa, Iowa City.) (Received for publication, October 3, 1921.) In the course of a study of a series of myopathies, we found that creatine administered by mouth was in large part or com- pletely eliminated in advanced, progressive pseudohypertrophic muscular dystrophy cases. This confirms the earlier observa- tions of Levene and Kristeller! who estimated the creatine con- tent of beef given such patients. Recovery of considerable amounts of ingested creatine in established creatinurias has been reported by Krause? for normal children, by Powis and Raper? in myotonia congenita, and by Gamble and Goldschmidt‘ for normal infants. Our muscular dystrophy patients have shown a much diminished creatinine coefficient and a high degree of creatinuria; appar- ently the power to convert creatine, both preformed and produced in intermediate metabolism, is markedly impaired. We shall present else- where evidence that the abnormalities of carbohydrate metabolism in muscular dystrophy (Janney, Goodhart, and Isaacson ;> McCrudden®) are not responsible for the creatinuria in the sense of the carbohydrate defi- ciency of Mendel and Rose.’ We are convinced that such cases afford a unique opportunity to study the problems of creatine formation in man. 1 Levene, P. A., and Kristeller, L., Am. J. Physiol., 1909, xxiv, 45. 2 Krause, R. A., Quart. J. Exp. Physiol., 1914, vii, 87. 3 Powis, F., and Raper, H. S., Biochem. J., 1916, x, 363. 4 Gamble, J. L., and Goldschmidt, S., J. Biol. Chem., 1919, xl, 199, 215. 5 Janney, N. W., Goodhart, 8. P., and Isaacson, V. I., Arch. Int. Med., 1918, xxi, 188. 6 McCrudden, F. H., Arch. Int. Med., 1918, xxi, 256; J. Am. Med. Assn., 1918, lxx, 1216. 7 Mendel, L. B., and Rose, W. C., J. Biol. Chem., 1911-12, x, 213. 319 320 Creatine Formation in Dystrophy The controversy as to whether creatine formation is primarily exogen- ous or endogenous has yielded the following facts: Creatinuria has been more easily established in normal adults (older children, women, adult males after creatine administration) on a high . protein diet than on a low protein intake (Folin and Denis’; Denis and Minot’); however, negative results are reported (Rose!°; Rose, Dimmitt, and Bartlett!!; Stearns and Lewis!”). In established creatinuria, the creatine excretion is increased in going from a low protein to a high protein diet (Levene and Kristeller!; Talbot and Gamble!*; Denis!*; Denis and Kramer“; McCollum and Steenbock'*; Steenbock and Gross!7; Harding and Young!8). In established creatinuria, ingested creatine is largely or completely recovered in the urine, in part as creatinine (Levene and Kristeller; Krause; Powis and Raper; Gamble and Goldschmidt). Preformed creatine may account for part of the creatine elimination in supposed creatine-free diets, © and must be taken into consideration (Gamble and Goldschmidt). It is indicated, therefore, that creatine may have an exogenous as well ‘as an endogenous origin. That is, it is a product of the catabolism of tissue, reserve, and food proteins. In the normal adult, creatine under- _ goes further change. It is not completely destroyed under conditions of carbohydrate deficiency (Mendel and Rose) or muscular insufficiency '(Shaffer!), but is excreted into the urine as such and in smaller part as creatinine. Attempts to find a precursorof creatine among the protein cleavage pro- ducts have centered on the amino-acid arginine and related bases. Nega- tive results are reported by von Hoogenhuyze and Verploegh?° following the ingestion of casein and gelatin, proteins relatively rich in arginine. Jaffé?4 and also Baumann”’,?* injected arginine into animals without finding an in- 8 Folin, O., and Denis, W., J. Biol. Chem., 1912, xi, 253. 9 Denis, W., and Minot, A. S., J. Biol. Chem., 1917, xxxi, 561. 10 Rose, M.S., J. Biol. Chem., 1917, xxxii, 1. 11 Rose, W. C., Dimmitt, J. S.,.and Bartlett, H. L., J. Biol. Chem., 1918, xxxiv, 601. A 12 Stearns, G., and Lewis, H. B., Am. J. Physiol., 1921, lvi, 60. 13 Talbot, F. B., and Gamble, J. L., Am. J. Dis. Child., 1916, xii, 333. 144 Denis, W., J. Biol. Chem., 1917, xxix, 447; xxx, 47. 15 Denis, W., and Kramer, J. G., J. Biol. Chem., 1917, xxx, 189. 16 WMeCollum, E. V., and Steenbock, H., J.Biol. Chem., 1912-138, xiii, 209. 17 Steenbock, H., and Gross, E. G., J. Biol. Chem., 1918, xxxvi, 265. 18 Harding, V. J., and Young, E. G., J. Biol. Chem., 1920, xli, p. xxxv. 19 Shaffer, P. A., Am. J. Physiol., 1908-09, xxiii, 1. 20 von Hoogenhuyze, C. J. C., and Verploegh, H., Z. physiol. Chem., 1908, lvii, 161. 41 Jaffé, M., Z. physiol. Chem., 1906, xlviii, 430. 2? Baumann, L., and Marker, J., J. Biol. Chem., 1915, xxii, 49. 45 Baumann, L., and Hines, H. M., J. Biol. Chem., 1918, xxxv, 75. R. B. Gibson and F. T. Martin 321 crease in creatine formation. Thompson*‘ has reported numerous experi- ments in the course of which arginine was injected into rabbits, dogs, and ducks; he was able to observe an increase in creatine excretion and the muscle content in creatine in most instances. Myers and Fine* obtained a higher muscle content in creatine for rats fed on the arginine-rich protein edestin. Inouye?* observed some formation of creatine from arginine added to liver extracts, and in liver perfusion experiments. Jaffé showed that glycocyamine might be converted into creatinine and creatine in experi- ments with rabbits; Doerner®’ confirmed this, finding also that glycocy- amidine was changed to creatinine. Mellanby** obtained no effect when glycocyamine was fed to rabbits and fowls. Baumann and Hines? found no evidence of the methylation of glycocyamine in the perfused liver in situ, though subcutaneous injection resulted in increased creatine excretion. Riesser*® augmented the creatine content of rabbit muscle by injecting (with urea) choline and betaine; the administration of sarcosine with urea led to creatine formation in a number of experiments. Baumann and Hines*! failed to obtain decisive results in perfusion experiments with sarcosine, betaine, choline, and methyl guanidine. Thomas and Goerne” obtained no creatinine formation in experiments with eguanido caproic acid and y-methyl aminobutyric acid. Harding and Young** have reported that cystine administration intensified the creatinuria in young dogs. Recently, Gross and Steenbock* report that arginine given to pigs caused an increased creatine excretion; positive results obtained with cystine are explained by a resulting acidosis. From a study of the creatine elimi- nation when sheep’s thyroid was fed to pigs, they suggest that creatine formation is primarily dependent upon the balance that obtains between the arginase and the oxidative systems whereby arginine is destroyed; the thyroid principle accelerates the latter at the expense of the former.* Before presenting the results obtained, we wish to explain that variations in the daily creatinine output in muscular dystrophy cases may be considerable, even with constant diet and care; Levene and Kristeller report a similar experience. The constancy 24 Thompson, W. H., J. Physiol., 1917, li, 111, 347. 25 Myers, V. C., and Fine, M. S., J. Biol. Chem., 1915, xxi, 389. 26 Inouye, K., Z. physiol. Chem., 1912, 1xxxi, 71. 27 Doerner, G., Z. physiol. Chem., 1907, lii, 225. bd 28 Mellanby, E., J. Physiol., 1908, xxxvi, 447. 29 Baumann, L., and Hines, H. M., J. Biol. Chem., 1917, xxxi, 549. 30 Riesser, O., Z. physiol. Chem., 1913, Ixxxvi, 415; 1914, xc, 221. 31 Baumann, L., and Hines, H. M., J. Biol. Chem., 1918, xxxv, 75. 8 Thomas, K., and Goerne, M. G. H., Z. physiol. Chem., 1918-19, civ, 73. 33 Harding, V. J., and Young, E. G., J. Biol. Chem., 1920, xli, p xxxvi. 34 Gross, E. G., and Steenbock, H., J. Biol. Chem., 1921, xlvii, 33. 35 Gross, E. G., and Steenbock, H., J. Biol. Chem., 1921, xlvii, 45. 322 Creatine Formation in Dystrophy of the creatinine excretion cannot be the commonly accepted check on the collection of the day’s specimen (cf. McCrudden and Sargent**) for such cases. EXPERIMENTAL. The experimental data which we present were obtained on one of our series of muscular atrophy cases (Case 7). We had hoped te confirm and extend these observations, but owing to more pressing clinical problems, it has been impossible to do so. A preliminary report of our findings has been made.*” A brief de- scription of the case follows: W. Stil., white, male, 12 years of age, weight 30.5 kilos, entered the State University of Iowa Hospital on Sept. 5, 1919. The patient had not walked for 2 years, but the condition had been noted 2 years before that time. The diagnosis was progressive pseudohypertrophie muscular dystrophy. There was general muscular weakness and marked atrophy, especially of the deltoids. There were contractures of the hamstrings, toes, and hips. The calf muscles showed the characteristic pseudo- hypertrophy. The patient was observed for 3 months in the metabolism unit. At the time of his discharge from the hospital, he weighed 39 kilos. The metabolic findings are given in Table I. Three low protein periods are given, days 2 to 8 inclusive, 26 to 34 inclusive, and 76 to 79 inclusive; the protein intake for the first two periods was 31.6 gm. and for the third 40 gm.; the energy value was 1,600 calories. Throughout the observations, the diet was non-purine and non-creatine. The diet for days 2 to 8 consisted of bread and honey with 250 gm. of milk. On the low protein diet, the total nitrogen elimination was about 3 gm., the creatinine from 0.12 to 0.15 gm., and the creatine a little over 0.3 gm. Ingested creatine is completely and promptly eliminated (days 5 and 28). On a high protein régime (75 gm. of protein), the creatinine excretion was increased so that figures of 0.16 to 0.175 are the rule; the creatine elimination at times was over 0.5 gm. per day. The total nitrogen varied from 6.25 to 8.67 gm. The figures are for the periods on days 16 to 22 and 40 to 47. The creatine curve follows the urea nitrogen rather than the total nitrogen excre- 36 McCrudden, F. H., and Sargent, C.8., Arch. Int. Med., 1916, xvii, 465. 47 Gibson, R. B., and Martin, F. T., J. Biol. Chem., 1920, xli, p. xxxvi. 50 52 R. B. Gibson and F. T. Martin 323 TABLE I. Creatine. Protein intake. “J Id Nn Or or Or Or Or Or Saeie oe ole, (= i= Bl hl | Remarks. Creatine, 0.5 gm. by mouth. Diet included 63.8 gm. of edestin. Diet included 40.0 gm. of edestin. Creatine, 0.6 gm. by mouth. Glucose, 55 gm. by mouth. Diet included 63.5 gm. of edestin. Diet included 50.3 gm. of gelatin. Urine lost; diet as on day 50. Diet included 55 gm. of gelatin. 324 Creatine Formation in Dystrophy TABLE I—Concluded. Day. gi vee eit au Creatine. Hae Remarks. gm. gm. gm. gm. gm. 53 | 8.86 | 7.24 | 0.179 | 0.615 | 75.0 | Diet as on day 52. 54 | 7.04 | 5.63 | 0.162 | 0.560 | 75.0 55 | 7.88 | 6.55 | 0.174 | 0.546 | 74.0 56 | 7.69 | 5.94 | 0.168 | 0.525 | 75.0 | Sarcosine, 1 gm. by mouth. 57 | 8.27 | 7-21 | 0.170.) 0.534) Faw 58 | 7.46 | 6.06 | 0.168 | 0.552 | 75.0 | Asparagine, 1 gm. by mouth. 59 | 7.60 | 6.20 | 0.162 | 0.560 | 75.0 60 | 7.41 | 5.85 | 0.161 | 0.757 | 75.0 | Glycocyamine, 0.5 gm. by mouth. 61 | 7.55 | 6.07 | 0.162 | 0.638 | 75.0 62 | 9.60 | 7.46 | 0.162 | 0.610 |186.0 | Hordein, 61 gm. added to diet. 63 | 7.64 | 6.07 | 0.162 | 0.475 | 76.0 66 | 8.58 | 6.13 | 0.155 | 0.478 | 75.0 67 | 7.51 | 6.24 | 0.152 | 0.514 | 75.0 68 {11.00 | 8.82 | 0.190 | 0.689 | 75.0 . 69 | 9.80 | 7.56 | 0.197 | 0.687 | 75.0 10 18.19" °5267 | Ono I OPA a aon 71 | 8.40 | 6.99 | 0.148 | 0.447 | 76.0 4 | 5.46 | 3.87 | 0.147 | 0.872 | 40.0 5 | 5.18 | 4.11 | 0.159 | 0.376 | 40.0 6 | 4.69 | 3.32 | 0.153 | 0.298 | 40.0 0.148 | 0.313 | 40.0 | Sodium benzoate, 4 gm. 78 | 4.38 | 3.08 | 0.147 | 0.313 | 40.0 0.147 | 0.306 | 40.0 80 0.142 | 0.272 | 40.0 | Histidine, 1.2 gm. by mouth. tion. Replacement of 63.5 gm. of the protein of the diet by the arginine-rich protein edestin (days 19 and 46) failed to increase the creatine output over figures for some of the control days. In order to increase the catabolism of the ingested protein, and try out the effect of feeding the incomplete protein gelatin, 50 to 55 gm. of gelatin were substituted for corresponding amounts of the mixed protein of the diet (days 49 to 53). Gelatin is practically creatine-free, commercial samples analyzed by Miss Booher in this laboratory contained 25 to 32 mg. of total creati- nine per 100 gm. of gelatin of which only 1 to 6 mg. represent creatine. An increased elimination of total nitrogen (10.25 gm.), — —s ——— a = a oe R. B. Gibson and F. T. Martin 325 urea nitrogen (8.15 gm.), creatinine (0.180 gm.), and creatine (0.654 gm.) resulted. Unfortumately, the specimen for day 51 was lost. The increased catabolism is to be regarded as exog- enous, since Murlin has shown that nitrogen equilibrium can be maintained on a low level when two-thirds of the protein of the diet are replaced by gelatin. The progressive nitrogen reten- tion (days 52 and 53) probably indicates an adjustment of the metabolism to utilize part of the amino-acids of the gelatin for growth purposes. Some days later (days 67 to 72), the patient for some unknown reason catabolized all of the ingested protein on 1 day and the metabolism was 2 more days in returning to the previous level. The metabolic picture is much like that ob- tained for the gelatin substitution experiment just presented. In a further attempt to associate characteristically constituted proteins with creatine formation, hordein was given on 1 day; this protein yields no lysine on cleavage and is rich in proline. On account of possible deficient digestion and absorption of the pro- lamine protein preparations, 61 gm. of hordein were given in ad- dition to the 75 gm. of the protein of the diet. In spite of an increase in the total and urea nitrogens, the creatine fell on this day and reached a low figure on the following day. The ex- periment is not satisfactory, however, as the creatine had not reached its customary level on account of the glycocyamine ob- servation. Sarcosine (day 56) and asparagine (day 58) were each without effect when given the patient. At least 36 per cent of 0.5 gm. of glycocyamine was converted into creatine without affecting the creatinine figure (day 60). Correction was made for the con- version of glycocyamine into glycocyamidine in the creatine anal- yses by a control determination after adding 50 mg. of gly- cocyamine to a tenth of the urine of a later day; 20 per cent of the glycocyamine was so converted. Assuming that glycocya- mine might be a stage in creatine synthesis in the body from glycine and the guanidine nucleus, 4 gm. of sodium benzoate were given to divert part of the available glycine as hippuric acid (day 78). The results were negative. 1.2 gm. of histidine was given the patient on the last day that he was in the hospital. There was no effect on the creatinine or creatine excretion. 326 Creatine Formation in Dystrophy Confirmatory evidence as to the effect of high protein intakes on creatinine and creatine excretion has been obtained on other progressive pseudohypertrophic muscular dystrophy cases in our series. The substitution (or addition to the diet) of egg white protein has not indicated that the high cystine content is more effective in inducing creatine elimination. The administration of cystine has given negative results also. CONCLUSIONS. The following conclusions may be drawn from the data pre- sented: Ingested creatine was promptly and completely eliminated chiefly as creatine, in part as creatinine, in an advanced pro- gressive pseudohypertrophic muscular dystrophy case. The creatine and to a lesser extent the creatinine excretion was increased as the result of a greater protein intake. This increase is obtained only from the protein that is catab- olized, including gelatin, and not from that retained for growth purposes. Preformed creatine in the diet was not an important factorto be considered in interpreting the results. The substitution of the arginine-rich protein edestin for 0.8 of the protein of the diet failed to increase the creatine excretion. . Hordein added to the diet increased the total nitrogen and urea elimination, but probably was without effect on the creatine; this observation is indicative only. Ingested sarcosine and asparagine did not lead to an increase in the creatine excretion. Glycocyamine was converted in part (at least 36 per cent) into creatine. It is probably not a stage in ordinary creatine formation. Experiments with cystine have been negative. THE ACTION OF NITROUS ACID ON CASEIN.* By MAX S. DUNN wirs HOWARD B. LEWIS.; (From the Laboratory of Physiological Chemistry, University of Illinois, Urbana.) (Received for publication, October 11, 1921.) INTRODUCTION. The current theory of the nature of the free amino groups of the protein molecule holds that these consist largely of the epsilon amino groups of lysine. This assumption is based upon the evi- dence obtained by Van Slyke and Birchard (1) and others (2) who found that the nitrogen liberated from native proteins by the action of nitrous acid for 30 minutes was approximately one- half of the lysine nitrogen present. However, it was the view of Kossel and Gawrilow (3) that the existence of a quantitative re- lationship between the free amino nitrogen and the lysine content of a protein was improbable. Evidence in support of this view has recently been ‘obtained by Edlbacher (4) and by Felix (5) who believe that there are free amino goups in the protein mole- cule other than those of lysine. These investigators observed that the lysine-free proteins, clupeine and salmine, methylated as easily as lysine-containing proteins and that the free amino nitro- gen content of histones, sturines, gelatin, and glycinin was greatly in excess of one-half of the lysine nitrogen of these proteins. It has also been shown by Herzig (6) that deaminized gelatin is methylated as easily as gelatin itself. These observations would seem to establish the fact that there are groups in the protein * An abstract of a thesis submitted by Max S. Dunn in partial fulfil- ment of the requirements for the Degree of Doctor of Philosophy in the Graduate School of the University of Illinois. t We wish to acknowledge our indebtedness to Professor Lafayette B. Mendel, of Yale University, who some years ago called the attention of the senior author (L) to certain aspects of Skraup’s work. 327 328 Deaminized Casein molecule, other than the free amino groups of lysine, which are capable of methylation but it does not follow of necessity that all of the groups which have been methylated are free amino in character. In view of the importance of this problem it seemed desirable to undertake a careful study of the product formed by the treat- ment of proteins with nitrous acid. Inasmuch as casein is easily obtained pure and its deaminaton product is insoluble, this protein was chosen for a study of the effect of deamination on the properties and composition of the protein molecule. The Preparation of Deaminized Casein. Pure casein was prepared according to the procedure of Van Slyke and Bosworth (7) omitting the treatment with ammonium oxalate. Nitrous acid was first used as a deaminizing agent for proteins in 1885 by Loew (8). This investigator found that one-third of the nitrogen of peptones was liberated by the action of nitrous acid. In 1896 Paal (9) used silver nitrite and hydrochloric acid to deaminize gelatin peptones, while in the same year Schiff (10) obtained a straw-yellow compound by the interaction of nitrous acid and egg albumin. 2 years later Schrétter (11) observed the formation of a similar substance from peptones. - In 1908 Treves and Salomone (12) allowed nitrous acid to react with egg albumin and obtained a yellow product which they believed was diazo albumin. Deaminized gelatin was obtained by Blasel and Matula (13) in 1914. Skraup (14) and his pupils have prepared and studied the deaminized products of the following proteins: casein (15); gela- tin (16); albumin (17); serum globulin (18); and edestin (19). Deaminized albumin, gelatin, and casein (20) and deaminized gliadin and vitellin (21) have been prepared and investigated by Levites. Levites and Skraup have been largely responsible for perfecting the methods used in preparing deaminized protein products. In the first method of Levites (20) a paste was made from the protein and sodium nitrite, warmed on the water bath, treated with dilute acetic acid, and the resulting olive-green product dried in vacuo. In his second (21) procedure Levites produced an emulsion of the M. 8. Dunn and H. B. Lewis 329 protein by agitating it vigorously in a shaking machine with 10 per cent acetic acid. This emulsion was warmed gently on the water bath with a 10 per cent solution of sodium nitrite and the resulting yellow product filtered, washed with alcohol and ether, and dried in vacuo. In the method used by Skraup (15) an acid solution of the protein, prepared by adding glacial acetic acid to a uniform suspension of the protein in water, was warmed gently with sodium nitrite on the water bath and the yellow product drained off on linen, desiccated, and dried in air. In the present series of experiments deaminized casein was pre- pared according to the methods outlined by Levites and Skraup. It was difficult to obtain a product of uniform color and appearance while the yield of 70 per cent reported by Skraup was rarely ex- ceeded. There are several objections to the use of a shaking machine as employed by Levites. It requires several hours to emulsify small amounts of protein and the emulsified product almost invariably contains gelatinous lumps which must be ground up in a mortar to avoid contamination of the deaminized product with unchanged protein. Most of the methods used for the prep- aration of deaminized proteins employ heat up to 40°C. to effect complete deamination of the protein. It is possible that this application of heat may cause a slight hydrolysis and other change in the protein. It is believed that the following procedure overcomes the ob- jections cited above. 100 gm. of casein were added to 2 liters’of distilled water contained in a 5 liter Pyrex flask. After stirring vigorously for 30 minutes with a mechanical stirrer a uniform suspension of the protein resulted. To this suspension 140 cc. of glacial acetic acid were added dropwise, with continued stirring, during the course of 1.5 hours. At the expiration of 20 minutes a good emulsion was formed, while at the end of the period, solution was effected. To this solution, 500 ec. of a solution of sodium nitrite containing 80 gm. to the liter were added dropwise, with continued stirring, during the period of 1.5 hours. After 150 ce. of this solution had been added a deep yellow precipitate rose to the top of the liquid as a yellow layer which, after standing for 18 hours, was filtered on a Buchner funnel using suction and a hardened filter paper. After triturating this substance fifteen times with hot water to the disappearance of an acid reaction to 330 Deaminized Casein litmus, it was granular and light yellow in color while the aqueous filtrate was similarly eolored. The yellow precipitate obtained by triturating four times with 95 per cent alcohol was thoroughly desiccated by triturating three times with dry ether, drying in air for 30 minutes, and in the oven at 80°C. for an equal length of time. Although the alcoholic filtrate was highly colored the pre- cipitate appeared to have lost but little of its yellow color in the washing process. a The deaminized product was of a uniform color and appearance and it was possible to secure practically a complete transforma- tion into the deaminized form. From three 100 gm. samples of the original casein yields of 90, 95, and 97.5 gm. of the oven-dried products were obtained. These deaminized products, designated subsequently as deaminized caseins A-64, A-66, and A-68 were very fine and powdery when passed through an 80 mesh sieve. With the second method of Levites, 17 gm. of deaminized casein A-18 were prepared from 25 gm. of casein. However, instead of permitting a complete deamination to occur the white precipitate formed upon the addition of the sodium nitrite solu- tion was filtered on a Buchner funnel as soon as possible. This precipitate, which was triturated fourteen times with hot water, three times with 95 per cent alcohol, and twice with dry ether, } was only faintly colored yellow. The Properties of Deaminized Proteins. Color.—Without exception, the deaminized protein products cited in the literature are yellow in color. The products prepared in the present research were colored light yellow when first pre- pared, but surfaces exposed to the light for a time became light brown. ‘Treves and Salomone (12) believed that these substances were diazo derivatives since they responded to the reactions given by diazo compounds. It is known, however, that primary aliphatic amines do not react with nitrous acid under ordinary conditions to give stable diazo derivatives. Furthermore were this a reaction with the amide groups which are present acids would be produced (22) and not diazo products. Therefore the assumption of these authors seems untenable. It is possible that deaminized proteins are colored because of the formation of nitroso compounds. There are numerous possi- M. 8S. Dunn and H. B. Lewis Bal bilities for nitrosation in the protein molecule. Histidine, tryp- tophane, and proline, each have one imino nitrogen, while there are two such nitrogens in the guanidine group of arginine. A nitroso group might enter tyrosine in the’ position ortho to the hydroxyl group in the benzene ring. It is unlikely that a nitrosation of the imide nitrogen making up the peptide linkage has taken place al- though it has recently been shown (23) that the nitroso derivative oi methyl phthalimidine is easily formed by treatment with nitrous acid in water solution. Since the carbonyl imide linkage present in methyl phthalimidine is the same as that found in peptides a possible nitrosation of the peptide linkage is suggested. However, this cannot have taken place to any marked extent because deam- inized proteins have been shown to contain less nitrogen than the original proteins. Solubility—Deaminized proteins are reported to be insoluble in water and insoluble or only slightly soluble in alkalies. Skraup noted the formation of a jelly-like substance when deaminized proteins were brought into contact with strong alkali. It was found in the present investigation that deaminized casein dis- solved in 0.5 per cent sodium hydroxide after standing for 2 days with the formation of a red solution and a small amount of un- dissolved residuum. With 1.5 per cent sodium hydroxide solution a red solution was formed in 24 hours while with concentrated alkali an orange to brown jelly was formed in a few minutes. Color Reactions —Deaminized casein prepared according to the method described above gave positive tests with Hopkins-Cole, Millon’s, and biuret reagents. Millon’s test was unquestionably positive although the color was less intense than with casein but the biuret reaction with deaminized casein was not characteristic ranging from a pink to a reddish purple. : Levites (20) is the only investigator to report a positive Millon’s reaction with deaminized proteins while the biuret reaction was found to be positive by Levites (20) and by Treves and Salomone (12). If it be true that deaminized proteins give a positive biuret reaction this would in- dicate that the grouping which is responsible for the color is not attacked or at least is only partially destroyed by the action of nitrous acid. Composition.—The elementary composition of native proteins appears to be but little altered in the deamination process. Skraup 332 Deaminized Casein (14) found a slight diminution in the phosphorous content of deaminized casein and a constant increase, with one exception, in the oxygen content of all of the deaminized proteins studied but neither observation was considered to be of particular signifi- eanece. Itis striking, however, that the nitrogen content of deam- inized proteins is lower than that of the original substance. Schiff (10) reported a reduction of 1.0 per cent in the values for nitrogen while the figures quoted by Skraup (14) range from 0.51 to 1.24 per cent lower in nitrogen than those of the original pro- teins. In the case of edestin (19) for some unaccountable reason the nitrogen of the deaminized product was found to be higher than that of the original protein. It will be noted from Table I that the nitrogen content of the deaminized casein prepared in this research ranges for the various samples from 0.22 to 0.68 per cent lower than the figures obtained for the original casein. TABLE I, Nitrogen Sampie. calculated on an ash-free basis. per cent Waser ds esis Aerie eRe ee eRe rene ee eee 14.56 Deaminized casemtA-18) suse eee ete eee eee 13.91 sé SO SAS G47 6 abies Ree: Eich mervache cones 14.34 s £6 ge ASO Oss oct ore CORE ee ere ees 13.88 cf Pit UNA Gy a arordie Ss Sa PVA EE See Meee ie eins 14.01 Free Amino Nitrogen.—In the present study of deaminized casein, free amino nitrogen was determined according tothe method — outlined by Van Slyke (1) with the use of the micro apparatus. To a suspension of 3 gm. of casein in distilled water was added a solution containing 0.375 gm. of sodium carbonate. The casein, which was usually in complete solution within an hour, was trams- ferred to a 100 ec. volumetric flask, diluted to the mark, and 2 ce. were taken for amino nitrogen analysis. This solution is neutral to litmus and as Van Slyke (1) has shown, an inappreciable hydrolysis takes place even after standing at room temperature for 48 hours. In the deamination of casein a foam inhibitor was found to be indispensable. For this purpose caprylie aleohol was found M. S. Dunn and H. B. Lewis 333 to be more effective than diphenyl ether although the blank re- sulting from the former substance was, as has been reported (24), considerably higher than that given by diphenyl ether. Even without the caprylic alcohol the blank from the sodium nitrite was in general slightly higher than that reported by Van Slyke for good samples of this substance. Casein was found to precipitate from solution immediately upon contact with the acid solution in the deaminizing chamber and to gradually change from a pure white to a yellow. Because of the fact that casein must undergo TABLE II. The Free Amino Nitrogen Content of Casein and Deaminized Casein. Total nitrogen Samples. as free | amino nitrogen. per cent Beemer niOriEins ee) 6). fe. Pct ow oe Oe eee 5.61 fe se ater Hammearsben: 2. 3 A COMPARATIVE STUDY OF THE HYDROLYSIS OF CASEIN AND DEAMINIZED CASEIN BY PROTEOLYTIC ENZYMES.* By MAX S. DUNN wits HOWARD B. LEWIS. (From the Laboratory of Physiological Chemistry, University of Illinois, Urbana.) (Received for publication, October 11, 1921.) The observations reported in the literature in regard to the digestibility of deaminized proteins are conflicting. Treves. and Salomone (1) reported that deaminized proteins are not digested _ by artificial gastric or pancreatic juice while Schiff (2) found them to be completely digested by dog’s gastric juice although the rate of digestion was much slower than with the original pro- teins. Levites (3) found digestion to be complete with dog’s gastric juice. In a review of the methods used for studying digestion in vitro the statement is made by Frankel (4) that “the best index of the extent to which a protein has been disintegrated is the ratio of the amino nitrogen at a given time to the total amino nitrogen ob- tained after complete hydrolysis.”’ Of the various methods by which the amino nitrogen of proteins may be estimated Frankel chose the Van Slyke procedure as ‘“‘best suited to the problem in hand.” After the examination of various methods for the study of proteolytic action Sherman and Neun (5) concluded that “the quantitative determination of . . . . the amino nitrogen of the digestion products . . . . appears to be more delicate as a means of detecting proteolysis than either the biuret or the nin- hydrin reaction and more delicate, accurate, and generally ap- plicable as a means for its measurement than any of the other quantitative methods here studied.”” These authors used the * An abstract of a thesis submitted by Max S. Dunn in partial fulfilment of the requirements for the Degree of Doctor of Philosophy in the Graduate School of the University of Illinois. 343 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 2 344 Deaminized Casein Van Slyke method for estimating amino nitrogen but assumed that results with the Sérenson method would run parallel. In the proteolytic studies of the present investigation the action of pepsin, trypsin, and erepsin alone and in series was studied. The samples of pepsin and trypsin used in these experiments were commercial preparations known to be active while erepsin was prepared from the intestinal mucosa of a dog according to Frankel’s (4) modification of the method outlined by Rice (6). The erepsin preparation was considered to be trypsin-free since it was without action upon fibrin. For the determination of the total amino nitrogen available on complete hydrolysis, samples of casein and deaminized casein were hydrolyzed according to the method of Henriques and Gjaldbaik (7), but the total amino nitrogen was determined by the nitrous acid method of Van Slyke instead of by formol titration. From this figure, the free amino nitrogen of the intact protein molecule was subtracted and the resulting figure was considered to represent the amino nitrogen in peptide linkage; 7.e., the maximum amount of amino nitrogen actually available for liberation by enzymes. The liberation of amino nitrogen during digestion was followed by means of the Van Slyke micro apparatus. Frankel ran controls “with all reagents and ferments in the same quantities except that no protein was added”’ to correct for the amino nitrogen present in the reagents. In the present study control experiments were carried out with the same amounts of protein and reagents, but using boiled enzyme solutions. By this technique corrections are made not only for the amino nitrogen present in the reagents and enzyme added but also for the free amino nitrogen content of the proteins and for that which may have been liberated by the hydrolytic action of the reagents. It is believed, therefore, that the corrected values for the free amino nitrogen are an accurate measure of the amino groups actually liberated by the digestive action of the enzymes employed. The results given in Table I and Chart 1 were obtained by the simultaneous digestion of 5.0 gm. samples of casein 1 and deamin- ized casein, A-64, the preparations of which were described in the preceding paper. These proteins were suspended uniformly in 250 ce. of 0.2 per cent hydrochloric acid and 20 ce. of an aqueous solution, containing 0.2 gm. of pepsin, added. Controls contain- M. 8. Dunn and H. B. Lewis 345 TABLET. The Peptic, Tryptic, and Ereptic Digestion of Casein 1 and Deaminized Casein A-64. Total available amino nitrogen liberated. Enzyme. Hours. Casein 1. Casein A-64. per cent per cent Pepsin ee 0.0 0.0 0.0 Be SASS ctevhe ee 14.5 9.6 2.6 AP ere atone 38.5 cs ie | 3.3 a SY, iy ERs Ae gen 86.5 Liege 3.3 eee RR 110.5 11.8 3.3 Trypsin added at the end of 110.5 hours. oT See ee 12225 53.2 25.8 ete 146.5 | 68.3 30.9 SEMA eis cise Secured oe a. 170.5 79.9 33.4 SEAS? Ss esata, os 242.5 78.7 33.5 Erepsin added at the end of 242.5 hours. BPR 32 cle sods 6's. = = 260.0 91.7 56.2 ah 284.0 92.1 65.4 RS Se rn = 308.0 | 95.8 65.8 * 332.0 95.7 65.4 THE LIBERATION CF AMINO NITROGEN BY ENZYMES fo 7s 120 160 200 2t°o 2Bo 320 360 Time 1n Hours CuHart 1. 346 Deaminized Casein ing the protein, all reagents, and boiled enzyme solutions were run. After thoroughly mixing the contents of each flask 5 ce. of toluene were added as a preservative and incubation at 38°C. was begun., At the end of 3 hours incubation the casein sample in which the active enzyme was present had gone entirely into solution, but the deaminized casein and the controls had settled out. The supernatant liquid of the deaminized casein sample was colored yellow indicating that a partial digestion had taken place since the supernatant liquid of the controls was colorless. At the expiration of the 3 hour period uniform samples from each flask were taken for analysis of amino nitrogen in the Van Slyke micro apparatus. To 5 cc. portions from each flask was added 0.5 ec. of N sodium hydroxide solution to stop digestion. The resulting solution was diluted to 10 cc. on a volumetric flask and 2 ce. were taken for analysis. At indicated intervals in the di- gestion subsequent amino nitrogen determinations were made. At the end of 110 hours digestion 200 cc. aliquots from each peptic digest were taken, 20 cc. of 15 per cent sodium carbonate solution and 0.4 gm. of trypsin added, and incubation at 38°C. was begun. Boiled trypsin solution was added to the controls. After several hours the deaminized casein sample went into solution giving a brown but transparent liquid but there appeared to be no change in the controls. At the end of 132 hours digestion with trypsin, 150 ce. aliquot portions were removed from each flask, 20 ce. of erepsin solution added, and incubation was continued for 89.5 hours. Boiled solutions of the enzyme were added to the controls. For the determination of amino nitrogen in the tryptic and erep- tic solutions 5 cc. aliquot portions were used. After arresting digestion by the addition of 0.5 cc. of glacial acetic acid these samples were diluted to 10 ce. in a volumetric flask and 2 ee. used for the analysis of amino nitrogen. Similar experiments were carried out with trypsin (Table II) and with erepsin (Table III) to determine whether these enzymes could attack deaminized casein without the preliminary action of pepsin. 2 gm. samples of casein 1 and deaminized casein A-64 were dissolved in 120 ce. of 0.5 per cent sodium carbonate solution. Casein 1 gave a solution of medium opalescence while casein A-64 formed a deep red solution in which gelatinous particles were suspended. 20 cc. of a solution containing 0.2 gm. of trypsin M. S. Dunn and H. B. Lewis 347 dissolved in a 0.5 per cent solution of sodium carbonate or 20 ce. of the erepsin preparation were added to each flask. 10 cc. of an alcoholic solution of thymol were added in each case as a preservative. After incubation at 38°C. for 22.5 hours a 5 ce. sample from each flask was taken, 0.5 cc. of glacial acetic acid added to arrest digestion, and the resulting mixture diluted to 10 ce. in a volumetric flask. A clear but brown liquid in the case of casein A-64 indicated that some digestion had taken place in each instance. The appearance of the proteins in the control flasks was not altered. The results obtained from the enzymatic hydrolyses of casein (Table I) agreed closely with the observations of Frankel (4). Pepsin liberated 11 per cent of the total available amino nitrogen of casein in 87 hours; trypsin superimposed upon the pepsin digest set free 79 per cent in 60 hours; and by the further action of erepsin for 66 hours 95 per cent was liberated. These values are maximum for these enzymes since the amino nitrogen was found to remain approximately constant over a period of 25 or more hours of additional digestion. The digestion of deaminized casein proceeded in every case at a slower rate than that of casein and the total cleavage was considerably less. Only 3 per cent of the total available amino nitrogen of deaminized casein was liberated after 110 hours of peptic digestion in contrast to 11 per cent with casein. Tryptic digestion for 132 additional hours set free only 33 per cent of the total available amino nitrogen as compared with 79 per cent for casein while the further action of erepsin liberated only 65 per cent as contrasted with 95 per cent for casein. It is possible that difference in digestive action between casein and deaminized casein may be due to decreased solubility of the latter and its subsequent less intimate contact with the enzymes. There is also the possibility that reactions incidental to the proc- ess of deamination may have taken place to alter the peptide linkage in such a way that its cleavage by enzymes became more difficult. It is evident from the results reported in Table II that trypsin will digest deaminized casein without the preliminary action of pepsin. However, the rate of digestion by trypsin alone is slower than that by trypsin after the preliminary action of pepsin, and the percentage of hydrolysis is less. The digestion of deaminized 348 Deaminized Casein TABLE II. The Tryptic Digestion of Casein 1 and Deaminized Casein A-64. Total available amino nitrogen liberated. Hours. Casein 1. Casein A-64. per cent per cent 0.0 0.0 0.0 22.5 41.3 25.3 45.0 50.4 27.9 68.5 D250 27.8 92.5 52.4 28.7 TABLE III. The Ereptic Digestion of Casein 1 and Deaminized Casein A-64. a Total available amino nitrogen liberated. Hours. Casein 1. Casein A-64. per cent per cent 0.0 0.0 0.0 21.5 8.0 0.0 45.5 8.0 0.0 93.5 11.6 0.0 TABLE IV. Date. Weight. Total nitrogen, Urea and ammonia nitrogen. 1931 kg. - gm. gm. per cent 4 ave if “ 8 ‘“ 9 HE) 9. 90- ore Bi 9.90 1.355 0.841 62 Ze ey 9.88 1.342 0.910 67 oe ais 9.88 1-251 0.757 60 So! 9.86 1.427 0.872 61 foarlat 9.82 1.955 ~ 1.622 7 = 216 9.75 1.363 0.946 69 de oe IY: 9.75 1.246 0.960 77 a LG 9.75 1.299 0.968 74 "£9 9.73 * 10 gm. of deaminized casein A-64 were fed in addition to the standard dict. M. S. Dunn and H. B. Lewis 349 casein by trypsin is less extensive and proceeds at a slower rate than that of casein. Since it is known that casein is attacked by erepsin without the preliminary action of other enzymes it was of interest to test the action of this enzyme towards deaminized casein. It was found (see Table III) that 11 per cent of the total available amino nitrogen of casein was liberated after 93 hours of ereptic digestion but no amino nitrogen was liberated from deaminized casein. It would appear, therefore, that erepsin does not attack deaminized casein easily. Since in the present investigation digestion experiments in vitro indicated that deaminized casein was digested, although at a slower rate than casein, it was desirable to study the behavior of deaminized casein in the animal organism. A female dog, weighing about 10 kilos, was maintained upon a uniform diet for 10 days to permit a constant level of nitrogen excretion to be reached. At the expiration of the period 10 gm. of deaminized casein A-64 were added to the standard diet and the elimination of extra nitrogen in the urine determined. The standard diet consisting of 400 cc. of water, 10 gm. of bone ash, 25 gm. of starch, 25 gm. of lard, 35 gm. of sucrose, and 50 gm. of beef heart was considered to be calorifically adequate for a dog of the size used. The urine was collected daily by cathe- terization. As shown in Table IV about 43 per cent of the added nitrogen was eliminated in the urine. -Since there was a corre- sponding or slightly greater increase in urea nitrogen elimination the assumption that utilization of the deaminized casein in the animal organism has occurred seems well founded. Further experiments were planned to include a study of the relative efficiency of casein and deaminized casein in the main- tenance of nitrogenous equilibrium in the dog, but it was found that repeated administration of deaminized casein resulted in vomiting and loss of appetite. For this reason the experiments along this line were not carried further. SUMMARY. Casein and deaminized casein were digested in vitro by pepsin and trypsin. Erepsin digested casein readily but attacked deaminized casein only after the preliminary action of pepsin or 350 Deaminized Casein trypsin. In every case the digestion of deaminized casein pro- ceeded at a slower rate than the digestion of casein. In one experiment after the feeding of deaminized casein to a dog the increased elimination of total and urea nitrogen indicated that the deaminized product was digested and metabolized in the animal body. BIBLIOGRAPHY. . Treves, Z., and Salomone, G., Biochem. Z., 1908, vii, 11. Schiff, H., Ber. chem. Ges., 1896, xxix, 1354. . Levites, S. J., Z. physiol. Chem., 1904, xliii, 202. . Frankel, E. M., J. Biol. Chem., 1916, xxvi, 31. Sherman, H. C., and Neun, D. E., J. Am. Chem. Soc., 1916, xxxvili, 2199. . Rice, F. E., J. Am. Chem. Soc., 1915, xxxvii, 1319. . Henriques, V., and Gjaldbak, J. K., Z. physiol. Chem., 1910, Ixvii, 8. NOOR WDE NOTE ON THE DETERMINATION OF B-HYDROXY- BUTYRIC ACID.* By ROGER S. HUBBARD. (From the Laboratory of Biological Chemistry, Washington University School of Medicine, Saint Louis.) (Received for publication, October 4, 1921.) In studying the determination of 6-hydroxybutyric acid three _ methods have been most used. These are: first, the isolation of the acid, with its subsequent determination by the. polariscope; second, the dehydration of the acid to give a-crotonic acid; and third, the oxidation of the acid to give acetone. In Hurtley’s (1915-16) paper, in which the discovery and subsequent investigation of the properties of this acid are dis- cussed exhaustively, it is shown that all of these methods are closely connected with the first studies made by Kulz (1884), Stadelmann (1883), and Minkowski (1884) into the properties of the organic acid isolated by them from urines which contained an excess of base greater than that theoretically needed to neu- tralize the inorganic acids found. Kulz (1884) first fermented diabetic urine and determined the rotation, thus measuring the amount of the acid present. Wolpe (1886) modified this method. He extracted the acidified urine with ether, took up the extracted acid with water, and deter- mined the acid in the extract by the polariscope. Magnus-Levy (1901), Bergell (1901), Geelmuyden, and Black (1908-09) im- proved this extraction method further. Ohlsson (1916) has recommended ethyl acetate instead of ether for the extraction. The second method was suggested by Darmstaedter (1903). He converted 6-hydroxybutyric acid into a-crotonic acid by the action of concentrated sulfuric acid, and determined the unsatu- * This paper formed a part of a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy at Washington University, Saint Louis, in June, 1921. 351 352 6-Hydroxybutyric Acid rated acid by titration after distillation—a reaction noted by Kulz (1884) and Stadelmann (1883). This method has been modified by Pribram (1911-12) and others, but has not been found to be very satisfactory (Shaffer, 1908-09; Shindo, 1907). The third method for the determination of 6-hydroxybutyric acid was proposed by Shaffer (1908-09) based on a reaction mentioned by Minkowski (1884). 6-Hydroxybutyric acid when oxidized with sulfuric acid and potassium dichromate forms acetone and carbon dioxide, probably giving acetoacetic acid as an inter- mediary product. In his first paper, Shaffer (1908-09) stated that this oxidation was quantitative. Embden and Schmitz (1910) in Abderhalden’s Handbuch stated their belief that the results by this method were low. In 1913, Shaffer and Marriott (1913-14) studied the reaction, using pure 6-hydroxybutyrie acid - prepared from calcium zine $-hydroxybutyrate. They found that their recovery was from 5 to 10 per cent low. The method ‘described is a slow one, as 3 to 4 hours are required for maximum oxidation with the amounts of sulfuric acid and potassium dichro- mate recommended. Marriott (1914), Folin and Denis (1914), and Kennaway (1914) have used this oxidation of 6-hydroxy- butyric acid in connection with different methods for the determination of acetone. Lately Van Slyke (1917) has studied the reaction carefully and investigated the causes that lead to incomplete oxidation. At the time the experiments reported below were performed, only the preliminary report of this last paper was available. : For determining $-hydroxybutyric acid in urine the method of oxidizing with acid and potassium dichromate was chosen as most suitable. Two main problems presented themselves for solution in using this method. Was it possible to find some way in which the determination could be made exact? That is to say, could conditions of the determination be found that would lead to a recovery of the theoretical amount of acetone from a given amount of B-hydroxybutyric acid? Could the time of oxidation be shortened? Many different oxidizing mixtures with widely varying Amounts of sulfuric acid and potassium dichromate were tried, and large and small amounts of 6-hydroxybutyric acid were used in the determinations, but a recovery of the theoretical amount of acetone was not accomplished. In attempting to R. 8. Hubbard 353 solve the second problem, an investigation of conditions govern- ing the oxidation of 6-hydroxybutyric acid by acid dichromate solutions was undertaken, and it was found that the rate and completeness of oxidation depend on the relative concentrations of acid and dichromate. Van Slyke (1917) has published a series of experiments establishing this fact, and the results found do not differ from those described by him. Shaffer and Marriott (1913-14) give the following description of the method (p. 271): “the contents of the distilling flask containing the oxybutyric acid was diluted to about 600 ce., 30 ec. of sulphuric acid (sp. gr. 1.59) added, and a total of about 0.5 -gram of K.Cr2O; in very dilute solution dropped in during the distillation which was continued about three and one-half hours.” Using this method as a basis, two methods were worked out for the determination of 6-hydroxybutyric acid in a shorter period of time. The first, as applied to large amounts of the acid, has been described in Folin’s Manual (Folin, 1916), and by Shaffer and Hubbard (1916). In this method 6-hydroxybutyric acid was oxidized in the presence of 9 to 10 nN sulfuric acid, and the amount of potassium dichromate was adjusted so that the reaction was complete in 15 minutes. Continued use of this method showed that aes were two objections to it; first, the acid was so strong that the tin tubes of the ordinary Kjeldahl still were quickly corroded; and second, when the technique was extended to very small amounts of acetone it was necessary to use a correspondingly decreased amount of potassium dichromate. In working with conditions in which less sulfuric acid was used, a great many different procedures were tried out to find which one would give the maximum recovery in a short time. It was found impossible to get complete oxidation in 15 minutes, and the time was lengthened to half an hour, and potassium dichromate was added at intervals instead of by drops. If all the reagent was added at once, yields were lower. After many trials of different amounts of acid and potassium dichromate, and of different methods of adding the dichromate, the following was selected: to the 8-hydroxybutyric acid con- tained in 100 ce. of solution, heated to boiling in a Kjeldahl flask attached to a water-cooled condenser, 30 ec. of sulfurie acid 354 s-Hydroxybutyric Acid (concentrated sulfuric acid diluted with an equal volume of water) and 20 cc. of a potassium dichromate solution, 0.1 to 0.2 per cent, were added through a dropping funnel. The burners were regu- lated so that about 50 ce. distilled in 10 minutes. After 10 minutes 50 ec. of 0.1 to 0.2 per cent potassium dichromate were added ‘and the distillation was continued; 10 minutes later 50 cc. more of 0.1 to 0.2 per cent potassium dichromate were added and distil- lation was continued another 10 minutes. The boiling was not interrupted while the additions were made. The total distillate TABLET. Determination of Pure Solutions of B-Hydroxybutyric Acid. Oxidized for } hour as described (results are expressed in terms of acetone): - Present. Found. Percentage. mg. mg. per cent 0.1202 0.1036 86 0.1202 0.1050 88 0.698 0.609 j 87 0.698 0.605 87 0.698 0.610 87 1.396 1.215 87 1.396 1.215 87 5 26492 2.405 86 2.792 2.395 86 6.01 5.05 84 6.98 5.90 84.5 12.02 10.25 85 24.04 20.90 87 39.53* 38.2 86 * Carried out by oxidizing for 23 hours. was -collected in a second Kjeldahl flask, and redistilled from sodium peroxide for 10 minutes into an Erlenmeyer flask. In both. distillations a little water was present in the receiving flask, and the delivery tube dipped below the surface. Acetone was determined in the contents of the Erlenmeyer flask by the method described in an earlier paper (Hubbard, 1920). There was a blank amounting to 0.01 mg. of acetone. Table I shows the results obtained by the method described. In these experiments pure solutions of calcium zine 6-hydroxy- butyrate were weighed out, and aliquots were used for each R. S. Hubbard ODD determination. The ‘amount present’? is expressed as milli- grams of acetone as calculated“from the amount of this salt present, and the ‘‘amount found” represents the results of the titration corrected for the blank on the reagents. The recovery . ranged from 84 to 88 per cent, and is approximately the same given by other methods in which sulfuric acid and potassium dichromate are used. The average of the figures is 86 per cent. Table I shows that the method as described is applicable for amounts of 6-hydroxybutyric acid varying from 0.1 to 25 mg. SUMMARY. A method is described for the determination of 6-hydroxy- butyric acid when that compound is present in widely varying amounts. The oxidation requires only half an hour, and the final determination is by 1odometric titration. BIBLIOGRAPHY. Bergell, P., Z. physiol. Chem., 1901, xxxiii, 310. Black, O. F., J. Biol. Chem., 1908-09, v, 207. Darmstaedter, E., Z. physiol. Chem., 1903, xxxvii, 355. Embden, F., and Schmitz, E., in Abderhalden, E., Handbuch der bio- chemischen Arbeitsmethoden, Berlin, 1910, iii, 934. Folin, O., Laboratory manual of biological chemistry, New York and London, 1916. Folin, O., and Denis, W., J. Biol. Chem., 1914, xviii, 263. Geelmuyden, H. C., Lahf. Forth. Ips., Hammarsten Memorial Number, 11. Hubbard, R. S., J. Biol. Chem., 1920, xliii, 43. Hurtley, W. H., Quart. J. Med., 1915-16, ix, 301. Kennaway, E. L., Biochem. J., 1914, viii, 230. Kulz, E., Arch. Biol., 1884, xx, 165. Marriott, W. McK., J. Biol. Chem., 1913-14, xvi, 281. Magnus-Levy, A., Arch. exp. Path. u. Pharmacol., 1901, xlv, 389. Minkowski, O., Arch. exp. Path. u. Pharmacol., 1884, xviii, 35. Ohlsson, E., Biochem. Z., 1916, Ixxvii, 232. Pribram, B. O., Z. exp. Path. u. Therap., 1911-12, x, 279. Shaffer, P. A., J. Biol. Chem., 1908-09, v, 214. Shaffer, P. A., and Hubbard, R. S., J. Biol. Chem., 1916, xxiv, p. xxvii. Shaffer, P. A., and Marriott, W. McK., J. Biol. Chem., 1913-14, xvi, 265. Shindo, S., Inaugural dissertation, Munich, 1907. Stadelmann, E., Arch. exp. Path. u. Pharmacol., 1883, xvii, 419. Van Slyke, D. D., J. Biol. Chem., 1917, xxxii, 455. Wolpe, H., Arch. exp. Path. u. Pharmacol., 1886, xxi, 138. ——“—--— i ii al i Y a DETERMINATION OF THE ACETONE BODIES IN URINE.* By ROGER 8S. HUBBARD. (From the Laboratory of Biological Chemistry, Washington University School of Medicine, Saint Louis, and the Laboratories of The Clifton Springs Sanitarium, Clifton Springs, New York.) (Received for publication, October 4, 1921.) The determination of the acetone bodies in urine has been the subject of much investigation. Most of the methods for the determination of acetone described previously (Hubbard, 1920) and for the determination of 6-hydroxybutyric acid described in the preceding note (Hubbard, 1921) have been used to determine the acetone bodies in urine, and, in many instances, were pri- marily developed for that purpose. Besides the references given in these two papers, other articles may be found in the bibliog- raphies in papers by Shaffer (1908-09), Shaffer and Marriott (1913-14), Hurtley (1915-16), Van Slyke (1917); and Engfeldt (1920). The present article contains the description of a method which has been found convenient for the determination of acetone from preformed acetone plus acetoacetic acid and from 8-hydroxy- butyric acid on the same sample of urine, even when they are present in yery small-amounts—— Preliminary Treatment. In analyzing urine for the acetone bodies, particularly for B-hydroxybutyric acid, by any method, there are various inter- fering substances which must be removed. Shaffer (1908-09) removed these in two ways: first, by a preliminary precipitation; and second, by redistillation to remove compounds other than * The work reported in this paper formed a part of a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philoso- phy at Washington University, Saint Louis, in June, 1921. oor 358 Acetone Bodies in Urine acetone which react with alkaline iodine solutions. As a pre- liminary treatment before analysis he precipitated sugar and other interfering substances from urine with basic lead acetate and an excess of ammonium hydroxide. When this method was tried it was found that the filtrate often contained lead which was pre- cipitated by subsequent treatment with sulfuric acid. Following a suggestion of Plimmer and Skelton (1914), sodium hydroxide was substituted for ammonium hydroxide, and it was found that glucose could be removed in concentrations up to about 5 per cent, and that lead could be completely precipitated at the same time, if the quantity of alkali was adjusted so that it was approxi- mately equivalent to the lead present. During the study reported here a paper by Van Slyke (1917) appeared in which the use of copper sulfate followed by calcium hydroxide was recommended as a preliminary treatment before analysis. This treatment removed not only sugar, but other interfering compounds as well and was found necessary even when normal urines were analyzed by the technique described by him. An older method for remoy- ing sugar from urine was described by Salkowski (1879), in which copper sulfate and sodium hydroxide were used. The following method was found to give almost complete removal of glucose and other easily oxidized compounds. 10 ee. of urine were measured into a 250 cc. graduate shaking eylinder, and diluted to 100 or 150 ce. 10 ec. of Goulard’s extract! and 10 ce. of 20 per cent copper sulfate were added, followed by sodium hydroxide in not too great excess. (Usually 10 cc. of 2.N concen- tration were found to be the correct amount for the purpose.) This solution was diluted to 250 ec. and filtered after standing about half an hour. This combination of the lead and copper precipitation methods appeared, in some cases at least, to remove interfering (easily oxidizable) compounds more completely than did either procedure when used alone. There is a relationship between the amount of glucose present in urine, and the amount of sodium hydroxide necessary to insure its removal by this technique. If 10 cc. of the Goulard’s extract . and 10 ce. of 20 per cent copper sulfate solution are added to 10 ee. of normal urine diluted with 100 cc. of distilled water, 5 ce. of twice normal sodium hydroxide will not precipitate all the lead 1 Pb,O(CH; * COO)2, 290 gm. to 1,000 gm. of solution, prepared accord- ing to U. S. P., 1916, ix, 249. R. 8. Hubbard 359 from the solution. Under the same conditions, 15 ce. will dis- solve a part of the lead, while if 7 or 12 ec. are used no lead will be found after filtering. If 10 per cent of glucose is present in the urine, 15 ce. of 2 N sodium hydroxide will not cause the appear- ance of lead in the filtrate. 10 cc. of the alkali described above TABLE I. Precipitation of Urine. Found after precipitation. Glucose. NaOH Glucose. Lead. Copper. per cent ce, 0.0 5 Trace. 0 0.0 7 ) 0 0.0 12 0 0 0.0 15 Trace. 0 0.0 20 - 0 lat 12 0 0 0 15) 15 0 Trace. 0 2.5 7 Very faint trace. 0) 0 5 7 Trace. 0 0 5 10 0) 0 0 7.5 10 Trace. 0 0 10 10 0 0 10 12 Faint trace. 0 0 10 15 0 0 0 15 15 | Faint trace. 0 0 20 12 Sta 8) 0 20 15 Trace. Trace. Trace. 20 20 55 ate at Glucose was added to normal urine to give the percentage listed. 10 ce. of each were measured into 250 cc. shaking cylinders and diluted to about 100 ce. 10 cc. of Goulard’s reagent and 10 ec. of 20 per cent copper sulfate followed by different amounts of 2N sodium hydroxide were added, and the solution was diluted to the mark and filtered at once. The filtrate was tested for glucose with Benedict’s solution, for lead with an excess of sulfuric acid, and for copper with ammonia. will remove glucose up to a concentration of about 5 per cent. while larger amounts—15 cc. of 2 N—must be used if the concen- tration is 10 per cent. Higher concentrations than this were not removed by the treatment, and urines containing more than 10 per cent of glucose must be correspondingly diluted before treatment. The facts discussed above are shown in Table I. — SE ns 360 Acetone Bodies in Urine For purifying acetone distilled from acetone bodies present in urine Shaffer (1908-09) used different methods for the different fractions which he determined. He added sodium hydroxide to the distillate containing acetone from acetoacetic acid, and redis- tilled before titrating to remove volatile acids. To the fraction corresponding to 6-hydroxybutyric acid he added hydrogen perox- ide as well as alkali before redistilling to oxidize acetaldehyde and related compounds to the corresponding acids. Folin and Denis (1914) used sodium peroxide instead of hydrogen peroxide for this purpose. These methods are satisfactory for urines in which the acetone bodies are increased to any considerable extent, but for normal urines further treatment for the removal of inter- fering compounds was found to be necessary if the final deter- mination was to be carried out with dilute alkaline iodine solutions. This further purification was accomplished by redistilling first from a solution of acid plus potassium permanganate, and then by distilling again from sodium peroxide. The following directions describe the method used for the analysis of the filtrate from the copper and lead precipitation for acetone plus acetoacetic acid and for B-hydroxybutyric acid. Determination of Acetone Plus Acetoacetic Acid. Measure 150 cc. of the filtrate from urine precipitated as de- scribed into a 300 ec. Kjeldahl flask, add 10 ec. of sulfuric acid (1 part concentrated acid diluted with 1 part water), insert a two-holed rubber stopper, with a dropping funnel in one hole and a bent distilling tube in the other, connect with a condenser, and distil at such a rate that about 50 ce. of distillate come over in 10 minutes. Collect the distillate in a 500 ee. flask containing a little water with the end of the delivery tube below the surface, as it should be in all cases in which acetone solutions are distilled, and make the distillate up to a volume of about 150 ee. Add to the contents of this receiving flask 5 ec. of strong sulfuric acid (1 part concentrated acid plus 1 part water), 0.2 gm. of potas- sium permanganate, and distil, collecting the distillate in a second 500 ec. flask; continue distillation 10 minutes or more, obtaining a final volume of about 100 cc. and taking care that none of the permanganate solution boils over. Add to the contents of the ee a R. S. Hubbard 361 second distilling flask about 0.5 gm. of sodium peroxide, and distil 10 minutes into an Erlenmeyer flask containing a little water, collecting 50 to 100 cc. If care is not taken at the start, the solution will foam over. Cork stoppers should be used for this distillation. This technique insures maximum oxidation of interfering compounds and does’ not oxidize acetone (Hubbard, 1920). When more than mere traces of acetone are present, that is, when the urine gives a distinctly positive test with ferric chloride or with sodium nitroprusside and alkali, the purification by the successive redistillations is unnecessary; for these urines the single redistillation from alkali as recommended by Shaffer (1908-09) is most satisfactory (see Table II). This technique is to be pre- ferred under these conditions, not only because it consumes less time and takes less apparatus, but also because it reduces chances of loss through the vaporization of acetone. . Determine acetone in the final distillate as follows (Hubbard, 1920): add 10 to 25 ce. of a solution of iodine in potassium iodide of such a strength that 1 ec. is equivalent to 1, 0.1, or 0.2 mg. of acetone (the concentration of the iodine to be used is indicated by preliminary qualitative tests on the urine); add 2 ce. of a con- centrated solution of alkali (200 gm. of electrolytic sodium hydrox- ide dissolved in 300 cc. of distilled water), mix thoroughly, and allow to stand for 10 minutes or more; acidify with sulfuric acid, and titrate after about 5 minutes with sodium thiosulfate of a concentration equivalent to that of the iodine used; add a small amount of starch before the end-point is reached to serve as indicator. Control titrations of the thiosulfate against iodine treated successively with alkali and acid should be run daily, as the alkali uses up some of the iodine, and the strength of the latter reagent varies somewhat from day to day. The difference between this control titration and the titration found after distillation measures the acetone present in the sample taken, equivalent to 6 cc. of urine. In cases where there is very little acetone present it is sometimes necessary to correct for a blank given by the reagents after distillation. The question of this correction is discussed later. The stock iodine solution is prepared by weighing out 13.13 gm. of iodine, dissolving with the help of 25 gm. of potassium iodide, and diluting tolliter. Dilute solutions are prepared from this by diluting with 2.5 per 362 Acetone Bodies in Urine cent potassium iodide to the desired iodine concentration. These dilute solutions change their strength slowly. The stock thiosulfate solution of an equivalent strength is made by dis- solving 25.65 gm. of the pure salt in distilled water. This is standardized after 24 hours against an equivalent solution of potassium biiodate con- taining 3.362 gm. per liter, and protected from the action of carbon dioxide with soda lime; so protected, the solution will keep its strength unchanged formonths. The dilute solutions are not as stable, and are prepared from this stock solution daily. In Table II are given the values obtained for acetone from acetone plus acetoacetic acid as found after distilling a sample of normal urine, and of the filtrate from the copper and lead treat- ment of the same urine, from various oxidizing reagents. The TABLE II. Effect of Successive Distillation from Different Reagents on Urine Acetone. ————— eee Straight Precipitated ae : ‘ urine in urine in Distilled successively from reagents given. : ; mg mg. my. mg. ToS Oattsto 5.2 Sos cl rede ce a ee ee oe 0.76 {12.7 |0.6 10.0 H.S0,; NaOH Fe CII TCT cert per ee Se oe Qt 1.8 |0.10 ea H2S0O,; NasO. Me Seia yas cist Sub ah (Se RRM eRe ete cae eer = 0.17 2.8 10.16 227 FSO. BssO7 4 KMinOg. cee ee ee es 0.14 2.3 10.11 nD: H2SO.; HeSO4 + KMnQ,; NazOo............... 0.05 0.8 |0.03 0.5 H2SO.; H2SO, + KMnQ,; NasOo............... 0.0462*| 0.8*/0.0496*| 0.8* * Carried out on a different sample of urine and titrated with a weaker solution of thiosulfate. All results are expressed in terms of acetone. repeated distillation is shown to be necessary when the amount of acetone is very small, but not necessary when it is increased as, In the latter case, the very small difference lies within the limits of experimental error. The table also shows agreement between figures on untreated urine and on filtrates from the pre- cipitation with copper and lead when the distillate is purified by successive redistillations. Determination of B-Hydroxybutyric Acid. To determine the 8-hydroxybutyric acid in the urine, treat the contents of the first distilling flask (urine filtrate plus sulfuric acid) by the technique described in the preceding note (Hubbard, 1921). EE R. S. Hubbard 363 To the boiling solution add, through the dropping funnel, 20 ce. of the strong sulfuric acid (1 part concentrated acid plus 1 part water), 30 cc. of 0.1 to 0.2 per cent potassium dichromate, and continue the determination as described for solutions of pure B-hydroxybutyrie acid. Redistil the acetone obtained as in the case of the first fraction from sulfuric acid plus potassium perman- ganate and from sodium peroxide, and carry out the determination on the final distillate in the manner described for the fraction from preformed acetone plus acetoaceti@acid. This determina- TABLE III. Duplicates on Urines. Acetone + acetoacetic | 8-Hydroxybutyric i acid. Redistillations. Meine pa Sample. | Per 100cc.| Sample. | Per 100 cc. mg. mg. mg. mg. iron Wits")... 2. Sis .05- 1 0.067 0.7 | 0.180 1.8 1 0.0830 | 0.8 | 0.290 | 29 1 0.077 0.8 0.200 2.0 1 0.035 0.35 | 0.183 Ls 1 | 0.045 0.45 | From H.SO, + KMn0O,7 and from NazOz...... 3g 0.078: | |>-aingy Ol ae 23 2 0.084 1.4 0.130 Ps 2 0.07 bee ot SOE LEO 1.8 Z 0.096f 1.6f 0.1102 1.8f 2 esyier * Samples of 10 cc. each used for these determinations. 7 Samples of 6 cc. each used for these determinations. t Glucose added to the urine to give a concentration of 5 per cent. tion gives the acetone formed from the oxidation of the 6-hydroxy- butyric acid present. A correction of 15 per cent must be added to the result to make up for the incomplete recovery of acetone. If much acetone is present the redistillation from sulfuric acid plus potassium permanganate may be omitted. Table III shows the agreement between duplicates obtained by this technique on both fractions of acetone from normal urine, as contrasted with the agreement when the redistillation from acid plus potassium permanganate was omitted. 364 Acetone Bodies in Urine Recovery of Substances Added to Urine. Table IV shows the recovery of acetone, acetoacetic acid, and 6-hydroxybutyric acid by this method. The acetone used was puri- fied by repeated redistillation until the boiling point was constant. The acetoacetic acid was prepared by hydrolyzing acetoacetic ethyl ester with sodium hydroxide, aerating to remove acetone, and acidifying. The product was then analyzed by distilling from sulfuric acid, andgredistilling from sulfuric acid plus potas- sium permanganate and from sodium peroxide to remove alcohol, and titrating the acetone formed by the usual Messinger method. The 6-hydroxybutyric acid was prepared from calcium zinc B-hydroxybutyrate which was shown, by its optical activity, to be 99 per cent pure. A solution of this salt was acidified, set in plaster, and extracted with ether for about 10 hours. The ether was distilled off, and the 6-hydroxybutyric acid taken up with water. The solution was boiled with a little bone-black, filtered, made up to 100 cc. with distilled water, and read in a _polariscope. The reading in a 2.2 dm. tube was 0.865. This corresponds to a concentration of 1.630 gm. in 100 cc. (two determinations). Bee ee 1.630 2.2 X 24.12 Two dilute solutions were prepared from this by diluting 1 to 10. A 5 cc. portion of each was analyzed by the technique described in the preceding note (Hubbard, 1921), and it was found that the acetone recovered was 85 per cent of the theoretical amount, the usual percentage recovered by this oxidation. The recovery of these substances when added to normal urine was satisfactory (Table IV). Blanks on Reagents. In all determinations in which distillation precedes the final analysis with dilute iodine solutions there is a blank. Its value is small when measured in terms of milligrams of acetone, but may amount to a high percentage of the total determinations when very small quantities of acetone are present. A large num- ber of experiments were carried out to determine the source of this blank, and to find out conditions which should reduce it to R. 8S. Hubbard 365 aminimum. Many of the precautions which were used to obtain very low blanks have been since found described by Widmark (1919) in his paper on the determination of acetone in blood. TABLE IV. Recovery of Substances Added to Urine. feteate Taken. le per|~ ti Found. ek as Sera Per cent. 100 ec. Filtrate. | Urine. Acetone. mg. cc. ce. mg. mg. mg. 2.6 150 6 0.088 0.146 0.144 101 5.8 150 6 0.178 0.296 0.289 102 14.8 150 6 0.424 0.707 0.739 96 30.4 150 6 1.00 1.67 1.52 110 69.0 150 6 1.88 Salo 3.00 104 149.0 150 6 4.32 7.20 7.45 97 600 150 6 Zs 29.2 29 .98 97.5 Acetoacetic acid. 2.6 150 6 0.0656 0.109 0.108 100 "25.8 150 6 0.834 1.39 1.29 108 60.6 150 6 2.07 3.46 oe 104 iy 150 6 3.82 6.37 6.41 100 132 150 6 3.87 6.47 6.64 98 320 150 6 9.55 15.92 16.01 100 §38 150 6 19.67 32.70 31.97 102.5 684 150 6 18.23 30.4 34.2 90 6-Hydroxybutyrie acid. 4.5 150 6 0.230 | 0.382 0.454 85 9.1 150 6 0.470 0.785 0.909 86 18.2 150 6 0.975 1.62 1.82 89 45 .45 150 6 2.20 3.90 4.545 84 90.9 150 6 4.69 7.81 9.09 86 181.8 150 6 9.54 15.9 18.2 87.5 454.5 150 6 22 24: ay) 16 45 .45 82 All results and figures are given in terms of acetone. One source of the acetone value found in blank determinations seems to be the presence of a very small amount of some impurity in the reagents, possibly in the lead subacetate used in precipi- 366 * Acetone Bodies in Urine tating the urine, but this forms only a small percentage of the total value. The larger part of the blank is present when water alone is distilled successively from the different reagents, and seems to come largely from the last distillation from sodium peroxide. Cork stoppers must be used to connect the distilling flask with the condenser, unless, as suggested by Widmark, an all glass still is available. It is necessary to boil water through this still before each determination, and it was found best to boil a solution containing the same amount of sodium peroxide in the flasks before they were used for this final distillation. This pro- cedure again resembles that recommended by Widmark. The still and flask used in the distillation with acid and potassium permanganate were similarly boiled out each day to remove TABLE V. Date. Acetone + acetoacetic acid. 8-Hydroxybutyric acid. ~ 1920 my. mg. Sept. 11 0.0125 0.0185 fy ils 0.0125 0.0165 e 11 0.0138 0.0131 Be AD 0.0147 0:0190"%) Tae Fy 12 0.0262 0.0189 ps iP 0.0151 0.0135 All results are expressed in terms of the acetone equivalent of the blank. These blanks were obtained by redistilling the first distillate from H.SOs + KMn0O, and from NazOz successively. grease which might yield, on oxidation, substances reacting with alkaline iodine solutions. Another source of error which should . be avoided is the presence in the air of the laboratory of ammonia, formaldehyde, reducing gases, and other fumes which react with alkaline iodine solutions. In working with dilute reagents such as are used compounds of this nature may cause serious complica- tions. If, however, the analyses are carried out on urines in which the acetone content is only slightly increased, these pre- cautions may be omitted. In Table V a number of determina- tions of blanks on urine reagents treated as in the determination are given; these blanks show good agreement with each other. It is noticeable that the values of the blanks are so small that they are of importance only when there is very little acetone present. —s- ~~ vee R. 8S. Hubbard 367 In Table VI the results obtained by titrating the distillates finally obtained from normal ufines are compared with results - obtained by the use of the reagent described by Scott-Wilson (1911). Acetone gives a turbidity with this reagent and the reaction is a very delicate one. The turbidity obtained from the acetone was matched against that produced by a known amount of acetone freshly distilled into the reagent. In some cases the turbidities were matched in Nessler tubes; determinations could be made to an accuracy of about 0.005 mg. by this technique. In other cases the solutions were read in the nephelometer (Marriott, 1913-14), and when the amounts of acetone were comparatively large, the colorimeter as described by Folin and TABLE VI. Comparison of Urine Acetone by Iodine and by Scott-Wilson Reagent. Acetone + acetoacetic acid. 8-Hydroxybutyric acid. | { | Scott-Wilson reagent.| Iodine titration. |Scott-Wilson reagent.| Iodine titration. No. | Concen- | Concen- | Concen- Concen- 6 ce. tration 6 ce. tration 6 ce. tration 6 cc. tration per 100 ce. per 100 cc. per 100 ce. per 100 ce. a mg. mg. mg. mg. mg. mg. mg. mq. ii 0.033 0.55 0.033 0.55 0.090 15, |. O4S 1.9 2 0.025 | 0.4 0.023 0.4 0.083 1.4 0.075 £25 3 02027 |) 0245 0.032 0.6 2.2 0.120 | 2.0 0.130 | Threenormalurinesused. All results are expressed in terms of acetone. Denis (1914) was used to measure the turbidity. Good agree- ment was found by the two methods, and it seems certain that the ‘results must represent acetone. Whether the source of that ace- tone in the determinations of 8-hydroxybutyric acid was actually that acid present in very small amounts in the normal urines analyzed is a question which the results do not establish. Table VII contains results obtained by the use of the method described on a number of urines. The cases included range from normal subjects to diabetics showing a moderately advanced degree of acetonuria. A few facts are noticeable. First, the values found for the normals are very low, in the vicinity of about 2 mg. per 100 cc. of urine, for acetone from all acetone bodies. These results correspond with the lowest values included in the literature. 368 Acetone Bodies in Urine TABLE VII. Urine Determinations. © Pref 1 ee 3 £ Z, Sex Date. acetone + aceto- b-H Ae nae ae ; Remarks. ow mg. per\jam. per\ma. per|gm. per 100 cc. | 24 hrs. | 100 cc. | 24 hrs. 1 | Male. is 2.4 Normal mixed. 2 oe 0.5 | 0.007} 2.5 | 0.0380 os case 1. 3 g 0.7 | 0.009} 1.6 | 0.021 H esa 4 | Female. 0.4] 0.001) 1.4 | 0.010 oo hulled 5 | Male. 1.6 | 0.003) 2.3 | 0.009 as #6 6 ce 0.8 | 0.002} 1.6 | 0.003 “f * 7 | Female. ; Mar. 13, 1921) 1.9 | 0.016] 1.2 | 0.010) Arthritic. 8 - Feb. 17, 1921} 0.2 | 0.004) 0.5 | 0.009 “ case 3. 9 “ Mar. 11, 1921) 3.2 | 0.026} 2.0] 0.016 e beers: | 10 = Nov. 10, 1919) 1.1 | 0.006} 6.1 | 0.040} Overweight. 11 - Dec. 7, 1920] 19.6 | 0.192} 19.4 | 0.190) Diabetic. 12 | Male. Aug. 15, 1919] 0.4 1.3 es SS 1OM1919) 2S TOF024 eos ROFOL 7 13 | Male. “ 17, 1919} 0.5 | 0.009} 1.2 | 0.020) Diabetic. <7 19) 1919 eon RORO25 sl aalnOLd2s 14|Female.| “ 17, 1919} 11.2 | 0.182] 15.0 | 0.244) Diabetic. “18, 1919] 9.2 | 0.156] 17.8 | 0.302 “19, 1919} 11 -2'| 0.063) 22.9 | 0.128 15|Female.| “ 18, 1919} 15.9 | 0.312} 20.5 | 0.403] Diabetic. “19, 1919) 7.2 | 0.163) 4.9 | 0.110 “20, 1919} 3.3 | 0.092} 4.0 | 0.011 16|Female.| “ 18,1919} 1.6 | 0.021) 4.3 | 0.055} Diabetic. #9 49,1919]. ES: \SOPOL7) 834 05088 “20, 1919] 2.5 | 0.024) 5.3 | 0.052 17 | Male. “15, 1919} 6.4 | 0.204] 6.3 | 0.203} Diabetic. ge 1919) 691 8.8 ass eal ke) 6.24) 0.176 1) 19,1919) 3.04 0.400) 1s 70 18 | Female. * 15, 1919} 0.4 | 0.007} 2.1 | 0.041] Diabetic. ray dials 1919) 1.3 | 0.058) 3.4 | 0.058 oe “Sie ad 4.4 0.085) 4.3 | 0.079 ec BR | 0. 2.5 | 0.048! 19, tB9) R. 8. Hubbard 369 TABLE Vil—Concluded. oe. Preformed £5) Sex. Date. acetone + aceto-\ sap aap 4 — Remarks. Dp acetic acid. AES EMSS: mg. per|gm. per\mg. perlgm. per 100 cc. | 24 hrs. | 100 ce. | 24 hrs. 19 | Female. | Aug. 15, 1919} 1.3 | 0.037; 2.2 | 0.065) Diabetic. SEZ ILO), 8 5>|' 0. 052| oem Ino. Uoz « 18, 1919} 2.7 | 0.041] 3.2 | 0.069 “19, 1919) 0.5 | 0.012) 1.2 | 0.027 20 | Male. Nov. 30, 1920} 24.3 | 0.467) 20.8 | 0.400} Diabetic case 5. Dec. 1, 1920) 34.3 | 0.686) 54.3 | 1.28 ae 2, 1920} 49.4 | 0.673)137 1.86 3 3, 1920) 56.1 | 0.822)137 2.00 a 5, 1920} 61.0 | 1.12 |150 2.77 3 7, 1920) 75.5 | 1.32 |199 3.48 ¥ 9, 1920} 53.0 2.49 1.08 107 All results are expressed in terms of acetone. Second, there is, in most normal cases more acetone from 8-hydroxybutyric acid than from acetone plus acetoacetic acid. In cases in which the excretion of the acetone bodies is slightly in- creased these two fractions are nearly equal, and in some of these cases there is an excess of the fraction from preformed acetone plus acetoacetic acid over that from 6-hydroxybutyric acid. In the cases in which larger amounts of these substances were ex- creted the fraction from $-hydroxybutyric acid was again found to be the larger, as has been repeatedly found in cases of diabetes in which there was a marked degree of acetonuria. These facts militate against the theory formerly held that there is any definite ratio between the amounts of the acetone bodies formed in the body which can be studied from the proportions excreted (see also Hurtley, 1915-16). In the two following experiments it is shown that these same relationships occur in the same individual under conditions which lead to a gradual development of ace- tonuria. In Table VIII? are recorded data obtained on a normal subject while he was living on a diet the fat content of which was increased. 2 This experiment was carried out in the metabolic ward of Barnes Hospital, Saint Louis, through the courtesy of Dr. William H. Olmstead. ‘940700 JO StUI0} UT passordxe o1B SOTpOd 91090" OY} JO SUOIZVUIULIOJOP JO S}TNSOY Oo ee eee ee it tin Eee |: ee a Se o. #20'0 | 0'F 700'0 | ¥'T 06F'0 | F6'6 | O19 Shr ‘SZ | OFS ost | 2°29 Cie Lae rad Go. a 920'0 | FS ez 0 S-6 299'0 | 9°OT | OT8 Shr ‘Z | 9% ost | 2°89 cA mame aa 1 3s a) 6z0°0 | pe - | 9200 | SF 029'0 | SZ°6 | 098 StF ‘Z| 9 FZ ost | 2°89 (1 aon a 4 Gh 0G a 5 1900 | 8’ 201'0 | g:OL | 6F8'0 | € OT | 0F0‘T gtr‘2|9 FZ | OST | 2°€9 oH amy a CL ot ae = zee'0 | 6 Se | F4Z'0 | 0'8% | F960 | FET | 086 str‘2|9 FZ | OST | 289 CH imme 4 CL BE hey Q tet | O'ZIT | 2080 | 0°89 | F160) Sat OST Sh Te Ses LZ 0¢ F°28 Ga TOL 89 LE; nes z9g'0 | 0°69 | £2h'0'| 0°89 | S180 | 9 FT, ST8 Tare eed 0g v 28 yeaa tO) 89 oe e) 9020 | 8'8t | P2480 | ors | 2820} 8 Al G60 TE) -TLeeee|- S22, 0S ¥ 28 Lee > VOL 89 TAS, aa ce0'0 | g's G90°0 | 2 Z 9110 | 9°9T | G06 e1¢ ‘Z | ST Oot | §°69 102 GT 96 PT. Pay cB) zr0'0 | 8S 8F0'0 | FF SFr 0"| O 2E | 000 1 | E29 24) S°S1 oot | $69 102 at 96 el. § 020°0 | a2 900°0 | 2°0 SFL'0 | S°2r | $86 e1¢'% | 2ST oot | $69 10Z cat 96 ZI ‘ARIN = “wb a oor 774) pe oor “mb “ub oF) quao sad “ub quao wad “m6 quaa sad "wb LIGI ed So cd we ee 2 “poe ‘plow : ; : q . : oudgnqAxorpAy-g |oreoerp + au0qooy NHN |'N T870.L]0UrnToA | Sele ayerpAyoqiey) 4B ulojO1g ee Sie ee Se eae SS SS et ae “oy8q] *SInoY FZ sod uli) “qo1d THA GTAVL 370 Ts 3 R. S$. Hubbard 371 The subject of the experiment (the author), was a man 5 feet, 103 inches in height, weighing 165 pounds, who was doing light laboratory work at the time. The amounts of fat eaten were as follows: 200 gm. of fat during the first 3 days of the experiment; 250 gm. during a second period of the same duration; and 175 gm. during an after period of 5 days. The carbohydrate was also varied during the different periods as shown. There was a slight increase in the acetone excretion during the first period, a marked increase during the second period, and a return to practically normal values at the end of the experiment. Since the appearance of a paper by Shaffer (1921) on the relationship of glucose and fat to each other in diets which show acetonuria, the percentage of the total calories fed as fat, as protein, and as carbohydrate have been calculated, and are given in the table. It is noticeable that the first diet taken, which caused a slight increase in the excretion of acetone, was markedly higher in actual and available carbohydrate than the one which he has described as the border- line diet for acetonuria; that is, one containing the foods in the ratio of 10 per cent in the form of carbohydrate, and 80 per cent in the form of fat. The excretion of acetone on this diet was increased only very slightly, however, and can, perhaps, be prop- erly attributed to variations in the mixtures of fat and carbo- hydrate burned at different times during 24 hours; with diets at or on the border-line of producing acetone it is evident that such variations may be important in causing slight increases in the excretion of acetone. The relationship between the two fractions of acetone referred to above is shown in this experiment. At the start there was an excess of acetone from $6-hydroxybutyric acid over that from ace- tone plus acetoacetic acid, a condition often, although not invari- ably, found in normal urine. During the first part of the experi- ment the fraction from acetone was in excess, but when the diet was markedly high in fat, and the total excretion of acetone consequently increased, the relationship of the substances was the same as that found at the start of the experiment; that is, there was an excess of acetone from 6-hydroxybutyric acid similar to that usually described in diabetic urines. When the diet became more nearly normal, and the excretion of acetone began to de- crease, there was again a period in which there was more acetone 372 Acetone Bodies in Urine from preformed acetone plus acetoacetic acid than there was from 6-hydroxybutyrie acid. The total nitrogen (determined by the Kjeldahl method) shows a slight negative balance during the first part of the experi- ment, but this negative balance was not sufficiently pronounced to allow conclusions to be drawn from the data. The ammonia nitrogen excretion was increased, and roughly paralleled the increased acetone excretion. The ammonia was determined by the method of Folin and Macallam (1912). Table IX, which contains results from another normal subject observed during a short fast, shows somewhat the same picture TABLE IX. Acetone Bodies Excreted during a Short Fast. Way of fast) oo. ede ek onc/sshid os cee eee eee Before.*) First. |Second.| Third. a. Urine |p apes) Scared ear bas 1920 ce, mg. mg. mg. ~ mg. Nov. 13 150 0.0147 0.0049 0.0262 0.0087 a 1S 100 0.0112 0.0056 0.0182 0.0091 ce) 13 50 0.0082 0.0082 0.0107 0.0107 orem IS; 50 0.0082 0.0082 0.0097 0.0097 cee le 50 0.0092 0.0092 0.0137 0.0137 ee ES 50 0.0067 0.0067 0.0072 0.0072 42 alr 150 0.0205 0.0068 0.0165 0.0055 ee le 100 0.0155 0.0077 0.0235 0.0117 ame ret/ 50 0.0065 0.0065 0.0135 0.0135 eg 50 0.0065 0.0065 0.0110 0.0110 elf 50 0.0125 0.0125 0.0095 0.0095 Hoa Wf 50 0.0070 0.0070 0.0225 0.0225* All results are expressed in terms of the acetone equivalent of the blank. When determinations were carried out in the routine manner on 50 ce. of filtrate (equivalent to 2.5 ec. of blood) the difference between the most widely varying blanks from preformed acetone plus acetoacetic acid is equivalent to a difference of 0.1 mg. in 100 ec. of blood; under the same conditions the maximum variation for the blank from the $-hydroxy- butyric acid fraction is 0.6 mg. in 100 cc.; if the starred (*) value is omitted, this difference of 0.1 and 0.2 mg. is based on eight and on seven separate determinations, respectively. These blanks were obtained by redistilling the first distillate from NaOH, H2SO,; + KMnO,, and Na.Os, successively. It was thought that possibly small amounts of acetoacetic acid might be precipitated by the preliminary treatment of blood. It was possible that substances in blood might be lost in this way, although, when larger amouuts were added, such added amounts might be recovered. To test this possibility a series of bloods ranging from an acetone content of 0.05 mg. per 100 ec. to a content of 2.56 mg. per 100 cc. was tested by the following technique: 380 Acetone Bodies in Blood TABLE IT. Analysis of B-Hydroxybutyric Acid in Samples of Defibrinated Beef Blood. f - ene : { Blank Acetone Shey eres ae distillate. Difference. Acetone. equivalent, Con ee sre rea cc. thiosulfate | cc. thiosulfate mg. mg. acetone mg. 9.80 8.90 0.90 0.090 0.050 9.80 8.95 0.85 : 0.085 0.04 0.045 24.60 PeBSALD 0.85 0.085 0.045 24.60 23 .65 0.95 0.095 Pe: 0.055 TABLE IV. B-Hydroxybutyric Acid in 100 Cc. Defibrinated Beef Blood. Corrected for Sample taken for analysis. Acetone found. 80 per cent recovery. ce. mg. 5 3.6 4.5 5 3.6 4.5 5 3.6 4.5 5 3.7 4.6 1 3.6 4.5 The sample of blood gave 0.5 mg. of acetone from preformed acetone and acetoacetic acid. TABLE V. Effect of Precipitation on the Determination of Acetone in Blood. | Taken. Blood Acetone found. Acetone found. aken. = Pi oe Filtrate. Blood. ce mg 10 a GES ce mg "100 ae 5 0.0605 12, 50 $ 0.0240 LEG 5 Olga ee. 7 50 : 0.0695 | 2.7 5) 0.0953 1.9 50 3 0.0417 I bey / 5 0.0133 0.3 50 3 0.00944 0.4 5 0.142 2.8 100 5 0.138 2.8 5 0.0020 0.04 100 5 0.0025 0.05 5 0.0565 ol 100 5 0.0497 1.0 Results are acetone from preformed acetone plus acetoacetie acid expressed in terms of acetone. Each pair of determinations was carried out in different bloods. All were purified by redistillation as described. R. S. Hubbard 381 5 ec. of blood were measured into an 800 ce. Kjeldahl flask, about 200 cc. of water and 10 cc. of sulfuric acid (1 part of the concen- trated acid diluted with 1 part of water) added, and the mixture was distilled for 20 minutes. The distillate was then purified by successive redistillation from sodium hydroxide, sulfuric acid plus potassium permanganate, and sodium peroxide. Duplicate samples of the same bloods were analyzed by the technique described in this paper; that is, they were precipitated and aliquots of the filtrate were distilled and redistilled as described. Table V contains the results obtained, and shows that there is good agree- ment between the values from the treated and untreated blood. In view of the recovery of acetone and of acetoacetic acid when added to blood before precipitation, and of the agreement shown TABLE VI. Comparison of Blood Acetone by Iodine and by Scott-Wilson Reagent. Taken. No. SS ee Scott-Wilson reagent. Iodine titration. Filtrate. Blood. C63 ce: mg. 700 Sad mg. 00 3 il 50 24 0.004 0.2 0.005 0.2 ye 50 23 0.008 Oe 0.0094 0.3 Acetone from preformed acetone plus acetoacetic acid determined on two normal bloods. between results on blood and the filtrate from the precipitation, it is certain that these compounds are not removed even in small amounts by the precipitation with colloidal iron, basic lead acetate, and sodium hydroxide. The acetone plus acetoacetic acid from two bloods was deter- mined by both iodine titration and precipitation with Scott- Wilson’s reagent. The results are given in Table VI, and show good agreement between the values obtained. The analysis with Scott-Wilson’s reagent was carried out as described in the pre- ceding paper. ‘Table VII presents the results obtained on a series of bloods from different types of cases. It contains, in addition to the values of the acetone bodies, the values for blood sugar as deter- mined by the method of Benedict (1918) and of carbon dioxide- 382 pies Sew 1 Male. 1 “ 1 “ 1 “ y “ 2 “ 3 “ 4 “ec 5 “cc 6 Female ‘71 “ 8 4 8* “ 9 “c 10 Male 11 “ 12 Female 13 “cc 14 “ce 15 “ce 16 “ce 17 “ 17 “cc 18 Male. 18 “ 19 Female. 19 “cc 20 e “cc 21 “ 22 Male. 23 Mh: 24 Female. 25 = 26 42 Acetone Bodies in Blood TABLE VII. Blood Determinations. Acetone aceto- | B-Hy- Date. | ace |tesy- | Sugar. express-| acid. ed as acetone. "100 ec. | 100 cc. | Per cent Nov. 19, 1919 |} 0.3 | 0.8 Dec. 8, 1919 | 0.1 | 0.1 | 0.098 Jane ot, LOZ0NROoL 0.4 Nov. 11, 1920 0.4 | 0.098 Dec. 17, 1919 | 0.1 Oks) Osta Nov. 18, 1920 | 0.7 | 0.4 | 0.185 Dec. 14, 1919 |} 0.38 | 0.4 | 0.125 Nov. 27, 1920 | 0.3 | 0.4 | 0.102 Dec. 3, 1920 | 0.2 | 0.3 | 0.125 Nov. 5, 1919 | 0.8 0.118 Dec. LORISISS ORG On |202 iT Feb. 17, 1921 | 0.9 | 0.02 | 0.139 “20; 19215 SSP e229" | ORS Nov. 24, 1920 | 0.6 | 0.4 | 0.128 Dec. 18, 1920 | 1.4 | 1.6 | 0.167 Nov. 6, 1920 | 1.1 0.8 Dec. 11, 1919 | 0.7 | 2.5 | 0.098 * 10; 1919) 0.03 | 0.2 |} 0-102 Nov. 29, 1919 | 0.1 0.6 | 0.109 Dec. 7, 1919 | 0.2 O23) + O.125 Nov. 24, 1920 |} 0.8 |.0.9 | 0.222 Dec. 17, 1919 | 0.8 | 0.6 | 0.190 Jan. 9, 1920]1.2 | 1.4 | 0.125 Dec. 4, 1920/6.5 | 8.9 | 0.222 S18; 19200 eo amealnone 0.154 10, 1920) | eG oe 1 hea 18, 19201 1.2, 13.1) | © 190 “719907, 170 SI oe PO MIeT Nov. 11, 1919 | 0.0 0.4 0.266 ee La 1919 O89 0.0 0.144 Jans 10; 1920) 228enleona 0.185 Aug. 18, 1919 | 0.7 | 2.0 | 0.128 1851919.) 15 Deo 0.092 Dec. 5, 1920| 0.8 | 0.7 | 0.156 * After diet markedly high in fat fed for 4 days. Results for acetone, acetoacetie acid, and B-hydroxybutyrie acid are expressed as acetone. Alkali reserve is measured as the CO.-combining capacity of the plasma. vol. per cent 65.3 74.0 0 68.3 76.8 78.0 61.7 61.5 56.8 58.7 51.0 50.4 46 .2 62.4 56.0 66.0 78.7 ie 50.0 67.2 42.8 Remarks. Normal. Normal. Obese. Arthritic. Nephritic. “ Gastroin- testinal. Thyroid. “ce “ “ Diabetic. “ Diabetic. Diabetic. Diabetic. R. S. Hubbard . 383 combining power of the plasma as determined by the method of Van Slyke and Cullen (Van Slyke, 1917, a; Van Slyke and Cullen, 1917). The bloods analyzed were kept from clotting with potas- sium oxalate, and were analyzed on the same day that they were taken. In most cases the bloods were drawn before breakfast. Inspection of the table shows that the values for normal bloods range from 0.1 to about 1.0 for acetone from all three acetone bodies. Results on cases of diabetes sometimes fall into the same range, and are sometimes much higher. Neither in normal nor in pathological specimens is there any relationship between the amount of acetone from acetone plus acetoacetic acid and that from 8-hydroxybutyric acid except in the case of diabetic bloods in which both values are high; in them the acetone from 6- hydroxybutyric acid is in excess. CONCLUSION. A method for the determination of the acetone bodies in blood has been described which gives a high and constant percentage of - recovery for added acetone bodies, and which gives good agree- ment between duplicate determinations. Agreement is also found between the values of the acetone as determined by two different methods—a fact which renders it probable that the substance so determined is acetone. The results obtained by this method on blood from normal subjects are low. The accuracy of the determination is about 0.1 mg. per 100 ce. of blood. My thanks are due to various members of the Staff of The Clifton Springs Sanitarium for the pathological specimens of blood analyzed. BIBLIOGRAPHY. Benedict, 8S. R., J. Biol. Chem., 1918, xxxiv, 203. Hubbard, R. S., J. Biol. Chem., 1920, xliii, 48. Hubbard, R.S., J. Biol. Chem., 1921, xlix, 357. Hubbard, R.S., and Wright, F. R., J. Biol. Chem., 1921, xlvi, p. xili. Kennaway, E. L., Biochem. J., 1914, viii, 230. Kennaway, E. L., Biochem. J., 1918, xii, 120. Ljungdahl, M., Biochem. Z., 1917, lxxxiii, 103. 384 Acetone Bodies in Blood Ljungdahl, M., Biochem. Z., 1919, XClil, 325. Marriott, W. McK., J. Biol. Chem., 1913-14, a, xvi, 281. Marriott, W. McK., J. Biol. Chem., 1913-14, b, xvi, 289. Marriott, W. McK., J. Biol. Chem., 1914, a, xviii, 241. Marriott, W. McK., J. Biol. Chem., 1914, b, xviii, 507. Scott-Wilson, H., J. Physiol., 1911, xlii, 444. Short, J. J., J. Biol. Chem., 1920, xli, 503. Van Slyke, D. D., J. Biol. Chem., 1917, a, xxx, 347. Van Slyke, D. D., J. Biol. Chem., 1917, b, xxxii, 455. Van Slyke, D. D., and Cullen, G. E., J. Biol. Chem., 1917, xxx, 289. Van Slyke, D. D., and Fitz, R., J. Biol. Chem., 1917, xxxii, 495. Van Slyke, D. D., and Fitz, R., J. Biol. Chem., 1920, xxix, 23. 2 Widmark, E. M. P., Biochem. J., 1919, xiii, 430. BLOOD ACETONE BODIES AFTER THE INJECTION OF SMALL AMOUNTS OF ADRENALIN CHLORIDE.* By ROGER 8S. HUBBARD anp FLOYD R. WRIGHT. (From The Clifton Springs Sanitarium, Clifton Springs, New York.) (Received for publication, October 4, 1921.) In an earlier paper (Hubbard, 1921) results were reported on the study of normal subjects under conditions which caused a slightly increased excretion of the acetone bodies. The results showed that under such conditions there was an amount of aceto- acetic acid excreted during the development of acetonuria which was in excess of the 6-hydroxybutyric acid simultaneously ex- ereted. Conclusions from these findings were uncertain, as the results were complicated by differences in the kidney thresholds of the different compounds, and these prevented certain inter- pretation of conditions within the organism. For this reason it seemed desirable to investigate conditions other than dietary changes which might give rise to increased production of the ace- tone bodies, and in which the increase and return to normal values would take place within a comparatively short period of time. The effect of the injection of adrenalin chloride was selected for study. Peters and Geyelin (1917) have reported experiments which showed that after the injection of adrenalin chloride solution there were changes in the carbon dioxide-combining capacity of the plasma as well as in the blood sugar content and blood pressure. These experiments indicate that there are extensive changes in the chemistry of the blood brought about by the presence of large amounts of this substance, and it was thought that changes might be found in the acetone bodies. Eiselt (1910) has reported an increase in urine acetone as the result of the injection of adrenalin into a patient suffering from Addison’s disease. * A preliminary report of the work described below was given before the American Society of Biological Chemists, in December, 1920 (Hubbard and Wright, 1921). 385. 386 Blood Acetone Bodies A series of seven experiments was run on normal men. Each subject was fed a standard simple breakfast, and an hour after- wards received an injection of 0.5 or 1 ec. solution of adrenalin chloride in a one to one thousand dilution. A sample of blood was taken before the injection was given, and other samples were taken at various intervals after the injection. Each sample was analyzed for acetone from preformed acetone plus acetoacetic acid and for acetone from $-hydroxybutyrie acid by the tech- nique described in the preceding paper (Hubbard, 1921), for sugar by the technique described by Benedict (1918), and for the car- bon dioxide-combining capacity of the plasma by the method of Van Slyke and Cullen (Van Slyke, 1917; and Van Slyke and Cullen, 1917). Changes in the systolic and diastolic blood pres- sure and the pulse rate were also recorded. These last showed no anomalies, except the expected variations with the larger dose of adrenalin chloride, and a detailed description of them has accord- ingly been omitted from this paper. The results obtained from the chemical analysis of the various samples of blood on the different patients are listed in Table I. Three of the experiments show a distinct rise of the acetone bodies, with a subsequent return to the values found preceding the injection of the adrenalin. In two of these cases both of the frac- tions show the increase. The experiments which showed the most marked variations are experiments in which the subject received a dose of 1 ec. of the drug. There was one experiment in which a dose of 1 ec. was given, and in which there was no change in the acetone bodies. The subject of this experiment showed a high blood sugar value in the sample taken before the injection and perhaps should not be classed as normal but there is no other reason for making such an assumption. There is no constant relationship in the degree of response of the different acetone bodies, and the results in the cases where there is a rise noted do not seem to bear any relationship to the changes in blood sugar nor in the carbon dioxide-combining power of the plasma. The magnitude of the rise observed in some cases and the subsequent return to normal values indicate that the changes are real changes in the substances present in the blood, induced by the adrenalin chloride administered. R. 8S. Hubbard and F. R. Wright 387 TABLE I. ee ee igs B-Hy- oe plus jd = asma — Sex. Date moe Time ae butyric Sugar doa acid per!’ 100 cc os ce hrs mg mg. | percent ee 1 |Male.| Dec. 8, 1919 3 Before. | 0.1 | 0.1 | 0.098} 65.3 3 0.3 0.3 0.156 13 0.05 | 0.2 0.160 72.4 23 0.100) 71.0 1 as Nov. 7, 1920 1 Before. 0.25 | 0.098) 74.0 ' 3 0.6 Eek 0.167 1 Lips: 1-7 0.225) 53.8 2 0.5 0.2 0.136) 51.0 2 od Dec. 17, 1919 3 | Before 0.1 0.3 0.121) 71.0 $ 0215) Ges Ot 1Ssi ceo 13 | 0.25 | 0.4 0.222) 69.1 23 | 0.1 0.4 0.160) 71.0 2 | “ | Nov.18, 1920 | 1 | Before. | 0.135 68.3 3 0.3 0.157| 59.8 1 0.2 0.0. | 0.215] 56.6 Z 0.2 0y2 0.194) 55.1 5 ¢ «14, 1919 3 Before. | 0.3 0.4 0.125) 76.8 4 0.5 1 key 0.154! 76.8 13 0.6 0.5 0.128) 67.3 23 0.25 | 0.35 | 0.122) 76.8 5 se Dec. 3, 1920 4 Before. | 0.2 URS: 0.125) 61.7 3 0.3 0:3 0.154) 62.6 1 0.6 0.1 0.200) 62.1 2 O23 0.3 0.122) 59.8 4 ot Nov. 27, 1920 1 Before. | 0.3 | 0.4 | 0.102) 65.3 3 0.8 O25 0.236) 59.5 1 0.8 0.9 0.266) 55.7 0.3 0.3 0.166) 61.4 Results for all three acetone bodies are expressed as acetone. Alkaline reserve is measured as the CO.-combining capacity of the plasma. 388 Blood Acetone Bodies The most satisfactory explanation for the results reported is in the probable local restrictions of the blood supply induced by the drug, which lead to a local production—or failure of combus- tion—of the acetone bodies. Such production could occur under these conditions in spite of the increased glucose content of the blood. The production of these acetone bodies certainly cannot be looked upon as responsible in any degree for the marked lowering of the alkaline reserve observed. The experiments do not afford any information concerning the question of the order of the production of the acetone bodies in the organism. Our thanks are due to the members of the Staff of The Clifton Springs Sanitarium who served as subjects for these experiments, and particularly to the late Dr. Malcolm 8. Woodbury, superin- tendent of the Sanitarium, for the continued encouragement which he extended to us during our experiments. BIBLIOGRAPHY. Benedict, 8. R., J. Biol. Chem., 1918, xxxiv, 203. Eiselt, R., Z. klin. Med., 1910, lxix, 398. Hubbard, R. S., J. Biol. Chem., 1921, xlix, 375. Hubbard, R. S., and Wright, F. R., J. Biol. Chem., 1921, xlvi, p. xiii. Peters, J. P., Jr., and Geyelin, H. R., J. Biol. Chem., 1917, xxx, 47), Van Slyke, D. D., J. Biol. Chem., 1917, xxx, 347. Van Slyke, D. D., and Cullen, G. E., J. Biol. Chem., 1917, xxx, 289. a SOME NUTRITIVE PROPERTIES OF NUTS. II. THE PECAN NUT AS A SOURCE OF ADEQUATE PROTEIN.* By ¥. A. CAJORI. (From the Department of Chemistry, Stanford University, Palo Alto.) (Received for publication, October 13, 1921.) It has been recently observed that young rats would grow at a normal rate and attain adult size on diets in which the essential source of the protein of the ration was derived from various nuts.! From these results it was concluded that the proteins of these nuts furnished suitable amounts of those amino-acids necessary for growth and that they can be regarded as ‘“‘complete”’ from the point of view of nutrition. With the exception of the pecan nut successful feeding trials resulted with all the nuts investigated, a list that included many of our important protein-rich nuts. The rations containing the pecan nut as the source of protein were complete in every other known dietary essential. Two causes may account for the failure of rats to grow at a normal rate on such diets. The proteins of this nut may yield insufficient amounts of those amino-acids that determine the nutritive value of a protein; or the pecan nut may contain some substance which renders rations of which it is an important component distasteful or injurious to rats. At the time of these observations there were on record neither detailed studies of the type of protein existing in the pecan nut nor studies of their chemical make-up. In view of the importance that this nut is assuming as a food crop in the United States, * The results reported in this paper were presented before the Pacific Coast Division of the Society of Experimental Biology and Medicine at the meeting of October 15, 1921. 1 Cajori, F. A., J. Biol. Chem., 1920, -xliii, 583. 2 In 1909 the production of pecan nuts in the United States was 9,890,769 pounds. In 1919, 45,619,000 pounds, a gain in 10 years of 400 per cent. These figures were taken from the 1910 United States Census Reports and the United States Bureau of Crop Estimation report, 1919. 389 390 Nutritive Properties of Nuts. II it seemed desirable to study the chemical character of the pro- teins of the pecan nut, and with such a basis for evaluation of | their nutritive properties, to repeat feeding experiments on animals. Chemical Experiments. Methods for the isolation and purification of plant proteins have been developed by Osborne*® and more recently by Johns and his coworkers‘ in their investigations of seed proteins. We have followed these methods in our work on the proteins of the pecan nut. The Van Slyke method? was used for protein analysis. Preliminary Experiments. Several kilos of shelled pecan nuts (1.5 per cent nitrogen) were passed through a meat grinder and the finely divided mass put in a laboratory press. This process, followed by repeated extraction with petroleum ether, removed all but traces of oil. The oil-free residue was ground in a mortar; the resulting product was a fine gray powder containing 4.3 per cent nitrogen. Small samples of the pecan meal were extracted over night with sodium chloride solutions of different strengths, and aliquots of the clear filtrate analyzed for nitrogen. Table I shows the amount of protein extracted from 2.5 gm. samples of the meal. Fractional Precipitation with Ammonium Sulfate.—Solid, finely powdered ammonium sulfate was added to 10 ce. of a 9 per cent sodium chloride extract of pecan meal. The salt was added in amounts calculated to gradually saturate the solution, each addition increasing the degree of saturation by 0.5 per cent. After each addition of ammonium sulfate the solution was shaken and the salt completely dissolved before the next quota of salt was added. The solution was examined for precipitated ° protein at the various stages of saturation. When the solution was 0.2 saturated a slight cloud appeared. At 0.4 saturation the solution became turbid and at 0.45 saturation a flocculent precipitate formed which was removed by filtration. Further addition of ammonium sulfate to the filtrate caused no change until 0.8 saturation was reached when the solution again became slightly turbid. This turbid- ity persisted, unchanged, when the solution was completely saturated. Examination of the saturated solution after it had stgod over night showed that a small flocculent precipitate had formed. 3 Osborne, T. B., The vegetable proteins, London, New York, Bombay, and Caleutta, 1909. 4 Johns, C. O., and Waterman, H. C., J. Biol. Chem., 1920, xlii, 59. Johns, C. O., and Gersdorff, C. E. F., J. Biol. Chem., 1920-21, xlv, 57. 5 Van Slyke, D. D., J. Biol. Chem., 1911-12, x, 15; 1915, xxii, 281. F, A. Cajori 391 To 5cce. of the 9 per cent sodium chloride extract, 1.86 gm. of ammonium sulfate were added so as to make the solution 0.5 saturated. A heavy precipitate formed. After 2 minutes this was filtered and 0.74 gm. of ammonium sulfate was added, making the solution 0.7 saturated. A very slight turbidity was noticed. To completely saturate the solution 1.16 gm. of the salt were added. A marked turbidity resulted. Temperature of Coagulation.—5 ec. of the 9 per cent sodium chloride solution were acidified with a drop of very dilute acetic acid. A thermom- eter was inserted in the test-tube containing the acidified solution and the tube heated in a double jacketed water bath. The temperature was so regulated that there was a rise of not over 1° in 2 minutes. At 55°C. a slight turbidity was noted but no flocculent coagulum ap- peared until the temperature of 70°C. was reached. The solution was kept at 71°C. for 1 hour. After cooling, the precipitate was removed by filtration and the clear filtrate again heated. At 70°C. the solution became slightly cloudy, and at 79-82°C. a precipitate began to form. A definite TABLE I. Extraction of Pecan Proteins with Sodium Chloride Solutions. NaCl N extracted. Protein extracted (N X 6.25). per cent mg. per cent Ont 12.4 31 1.0 ad 3.1 3.0 26.4 6.9 5.0 35.1 oa! 7.0 45.8 10.6 9.0 50.9 11.9 11.0 58.0 14.4 coagulum formed at 86°C. and the solution was kept at this temperature for 1 hour. The heavy coagulum present at the end of this time was re- moved by filtering, and the heating continued. At 90°C. a slight cloudi- ness appeared. Long heating at 97-100°C. caused a slight coagulation to form. The preliminary experiments indicate that pecan meal con- tains, as its principal protein, a globulin, salted out by 0.5 satura- tion with ammonium sulfate and, coagulated at 79-86°C. There is evidence of a trace of an albumin, which starts to coagulate at 55-60°C. and is precipitated by ammonium sulfate only at a point of complete saturation. The fact that at the temperature of boiling water a slight coagulum forms, in addition to that caused by lower temperatures, is evidence, perhaps, of a second globulin, present in very small amounts. 392 Nutritive Properties of Nuts. II Preparation of a Pecan Globulin. Several hundred grams of pecan meal were extracted with seven times their weight of 10 per cent sodium chloride solution. The undissolved residue was separated from the solution by squeezing the mass through cheese-cloth. A solution free from suspended particles was obtained by filtering it several times through thick layers of paper pulp on a Buchner filter. This slightly opalescent solution was made 0.5 saturated with ammonium sulfate by the addition of the calculated quantity of the salt and allowed to stand over night. As much as possible of the clear super- natant solution was syphoned off from the precipitated globulin before it was transferred to a filter. The protein was removed from the filter paper and redissolved with distilled water and a little 10 per cent sodium chloride. It was again filtered clear and then dialyzed in parchment paper bags against running water for 120 hours. After that time it gave no more marked a test for sulfates and chlorides than did the tap water against which it was dialyzed. The contents of the bag were transferred to a Buchner filter and washed with distilled water and alcohol. The globulin was dehydrated by suspending it in absolute alcohol over night. It was filtered on a hardened filter paper, washed with anhydrous alcohol and ether, dried in a Freas oven at 110°C., and finally placed in a vacuum desiccator over concentrated sulfuric acid. The pecan globulin, prepared in this way, was a light gray powder, with no evidence of crystalline structure, containing 15.76 per cent nitrogen and 0.83 per cent sulfur, calculated on a moisture- and ash-free basis. It gave the usual protein tests including a strongly positive test for tryptophane with the Hop- kins-Cole reagent. With a-naphthol a very faint color developed due undoubtedly to slight contamination of the preparation with traces of filter paper. The distribution of the nitrogen in this globulin as determined by the Van Slyke method, after complete hydrolysis with 20 per cent hydrochloric acid, is shown in Table II. In general this analysis agrees fairly well with the recently published results of Dowell and Menaul. These authors deter- mined the nitrogen distribution of the mixed proteins extracted from pecan meal by barium hydroxide and sodium hydroxide. The fact that our analyses are similar would indicate that the globulin constitutes the large part of the proteins of the pecan nut. It may be noted that there is no evidence either in our 6 Dowell, C. T., and Menaul, P., J. Biol. Chem., 1921, xlvi, 437. i ie F. A. Cajori 393 results or in those of Dowell and Menaul of the unusually high content of histidine or a low arginine content in pecan proteins, reported by Nollau,’ in an analysis published some years ago. In considering the nutritive value of the pecan nut, the large amounts of basic amimo-acids yielded by the globulin are significant. TABLE II. Distribution of Nitrogen in Pecan Globulin. After hydrolyzing 2.1727 gm. of the protein, the solution contained 341.9 mg. of N. mg. per cent JATTDTKG ECE GS BRA Cea eae BAe Le eRe 8 Be 30-0 9.8 idlvenaviay INCE es ee ae ie e eee ar 12.3 3.6 pNMNCTERE AI nr terre ayo, s, sini a. . Ker. a 408 Studies on Experimental Rickets. IX | a il ce chelate ea eneray CnHart 4, = N\ EL ee NK NS i a Eee eae Shipley, McCollum, and Simmonds 409 These several groups of growth curves of rats are presented in order to illustrate how effectively a diet of purified food substances can be sup- plemented by means of additions which do not contain demonstrable amounts of the antiscorbutic substance. They form a complete demon- stration of the fact that the bone changes induced in those rats which were selectively fasted for water-soluble B only were due to lack of this dietary essential. It has been pointed out by Vedder (9), Hess (10), and others, that there are similar nervous manifestations and pathological changes in the heart in human eases of scurvy and beri-beri. Unfortunately, one cannot place much confidence in the description of these diseases as they occur in man, for an examination of the diets on which people develop either of these diseases makes it practically certain that beri-beri scarcely ever occurs except in an individual who is a border-line case of scurvy. Scurvy doubt- less has, however, occurred many times without complication with beri- beri. We have no assurance that any description of histological studies of any human tissues were made on subjects suffering from one of these diseases uncomplicated with the other. Nevertheless there is an added interest, in the repeated statements of others that there are lesions in beri-beri and scurvy which are common to the two diseases. Our observations show that the histological changes in the bones of rats suffering from uncomplicated beri-beri are apparently identical with those seen in the guinea pig suffering from uncomplicated scurvy. BIBLIOGRAPHY. 1. McCarrison, R., The pathogenesis of deficiency disease. II. The effects of deprivation of ‘B’ accessory food factors, Indian J. Med. Research, 1919, vi, 550. . Findlay, G. M., An experimental study of avian beri-beri, J. Path. and Bact., 1921, xxiv, 175. 3. McCollum, E. V., and Simmonds, N., The nursing mother as a factor of safety in the nutrition of the young, Am. J. Physiol., 1918, xlvi, 205. 4, Parsons, H. T., The antiscorbutic content of certain body tissues of the rat. The persistence of the antiscorbutic substance in the liver of the rat after long intervals on a scorbutic diet, J. Biol. Chem., 1920, xliv, 587. 5. McCollum, E. V., and Parsons, H. T., The antiscorbutic requirement of the prairie dog, J. Biol. Chem., 1920, xliv, 603. 6. McCollum, E. V., Simmonds, N., and Pitz, W., The nature of the dietary deficiencies of the oat kernel, J. Biol. Chem., 1917, xxix, 341. 7. Tozer, F. M., On the histological diagnosis of experimental scurvy, Biochem. J., 1918, xii, 445. bo 410 Studies on Experimental Rickets. IX 8. Shipley, P. G., Park, E. A., McCollum, E. V., and Simmonds, N., Studies on experimental rickets. III. A pathological condition bearing fundamental resemblances to rickets of the human being resulting from diets low in phosphorus and fat-soluble A, The phosphate ion in its prevention, Bull. Johns Hopkins Hosp., 1921, xxxli, 160. 9. Vedder, E. B., Dietary deficiency as the etiological factor in pellagra. Third report of the Thompson Pellagra Commission of the New York Post-Graduate Medical School and Hospital, Arch. Int. Med., 1916, xviii, 137. 10. Hess, A. F., Subacute and latent infantile scurvy. The cardio- respiratory syndrome, J. Am. Med. Assn., 1917, Ixviii, 235. EXPLANATION OF PLATES. PLATE 1. Fics. land 2. This rat was in a condition of polyneuritis induced by feeding Diet 3106 for 70 days. Fig. 1 shows the animal rising to his feet from the spread-eagle posture assumed during rest in a ventral position (Fig. 2). Note the spasticity and the extreme extension and abduction of all four extremities, the extension of the digits, and the roughness of the coat. PLATE 2. Fic. 3. Section of a long bone from a guinea pig with acute scurvy. The bone is well calcified throughout but is extremely osteoporotic. The marrow cavity shows numerous hemorrhages. H = areas of hemorrhage. Leitz microsummar. 35 mm. objective. No ocular. Fic. 4. Long bone of a rat killed during an attack of polyneuritis. This bone was osteoporotic and the hematopoietic marrow was almost entirely replaced by hemorrhage. H = areas of hemorrhage. Same mag- nification as Fig. 3. PLATE 3. Fie. 5. High power picture of the marrow cavity of the bone shown in Fig. 3 to show the hemorrhagic marrow. H = areas of hemorrhage. Leitz objective No. 6. No ocular. Fic. 6. To show the replacement of the bone marrow by hemorrhages during polyneuritis. H = hemorrhage. Magnification as in Fig. 5. PLATE 4. Fic. 7. Bone marrow of a polyneuritic animal during the congestion stage before hemorrhage has occurred. The marrow consists only of reticular tissue supporting widely dilated congested blood vessels. M = marrow. Same magnification as in Fig. 5. Fic. 8. To show the bone of a rat which was fed on the diet which we used to induce scurvy in our guinea pigs. This bone is quite normal. pe | \ THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX. PLATE 1. Hire=e2: (Shipley, McCollum, and Simmonds: Studies on experimental rickets. IX.) THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX. PLATE 2. (Shipley, McCollum, and Simmonds: Studies on experimental] rickets. IX.) VOL. XLIX. PLATE 3, (Shipley, McCollum, and Simmonds: Studies on experimental rickets. IX,) ie + fe ~ + leek Te : ex ig’ * a. 4g Soc) Oh * E ‘ q ; werd ae 2) ae > ans (Shipley, McCollum, and Simmonds: Studie S On experimental Tickets. IX.) AMMONIA EXCRETION FOLLOWING EXPERIMENTAL ADMINISTRATION OF ACIDS VIA THE STOMACH AND PERIPHERAL VEIN. By ROBERT W. KEETON. (From the Otho S. A. Sprague Memorial Institute, Laboratory of Clinical Research, Rush Medical College, Chicago.) (Received for publication, October 1, 1921.) The type of acidosis which attains its highest expression in diabetes is, as is well known, very frequently attended by a marked increase in the excretion of ammonia in the urine. The acids in this case (6-hydroxybutyrie and acetoacetic) are in part neutral- ized by the organic basic ammonium. Yet the phenomenon of neutralization by ammonium is not constant in the other types of acidosis in man. It appears to play relatively a small part in the acidosis of uremia (1), even when this attains an extreme degree, and it is missed in some of the other types of spontaneous acidosis, which have been observed in man. On the other hand, the feed- ing of dilute mineral acids (2, 3) to man, if in sufficient quantity, leads consistently to increased excretions of ammonia in the urine comparable to those seen in the diabetic type of acidosis. How- ever, if a man, who is excreting increased quantities of ammonia, either as an incident in the course of a spontaneous diabetic acidosis or as the result of an alimentary administration of an acid such as hydrochloric, be given a sufficient quantity of bicarbonate of sodium (4), the increased excretion of ammonia is depressed or wholly annulled. The administration of the stronger base results in the displacement of the excess of ammonia from the urine, and it would not seem irrational to compare the phenomenon With that which occurs in the test-tube, when sodium bicarbonate, magnesium oxide, or calcium oxide, are added to a watery solu- tion of ammonium chloride, acetoacetate, 6-hydroxybutyrate, or the like with the liberation of free ammonia and the formation of a new salt of the base employed. In the body such a release 411 w 412. Ammonia Excretion after Acids of ammonia from its salts (or the prevention of its union with acids) by the action of a stronger base could be conceived as leading simply to some other disposition of the ammonium radicles displaced or spared from combination with acids. They might then, for example, enter into the synthesis of urea. Although the phenomenon of neutralization of acids by ammonia has long attracted interest and has been spoken of by some writers (5) as one of the important compensatory mechanisms by which the body may protect itself against inroads of acid, it is a noteworthy fact, that it is not a mechanism which is called into play consist- ently by the human organism under all conditions, in which it might be supposed in reason that every available means of pro- tection should be employed. A patient may die of acid intoxi- cation developing in the course of a pyelonephritis without ever having called on this resource to a noticeable degree. In man the classical examples of neutralization of acids by ammonia are afforded in acidosis of the diabetic type. It has been reported in clinical methyl alcohol poisoning (6) which, however, under experimental conditions (7) may not always be associated with an acidosis. It is also said to develop in the acidosis of cholera (1) and it has been shown to occur in connection with certain of the diarrheal diseases of children, which in some cases, resemble closely the acidosis of diabetes. The two chief conditions, how- ever, in which the phenomenon is known to occur regularly are those mentioned. Moreover, this type of acidosis associated with increased ammonia output is not seen in all types of animals but occurs most readily in man and the carnivorous animals, and only to a lesser extent in the herbivorous animals. These considerations suggest that the neutralization of acids by ammo- nium is nota mechanism of universal significance in the whole problem of neutrality regulation by living cells, but that it is rather a special phenomenon which only occurs under certain conditions. It has been pointed out by others (8) that, in discussing observed variations in the excretion of acetone bodies in the urine, it is pertinent to consider the question of where the acetone bodies are formed. Well known evidence afforded by liver perfusion experi- ments, favors the view, that the liver, of all the organs, may be regarded as the largest single producer of the acetone bodies. The R. W. Keeton 413 liver is also commonly recognized as an organ in which the total traffic in NH2 groups is greater than in any single tissue, even though the function of deamination and the synthesis of urea were performed to some extent elsewhere as well.1 With its deaminizing and urea-forming function it must be considered that the liver is preeminent in respect to the total quantity of ammonia equivalents that it normally handles in the course of the day. It is generally accepted that the ammonia which in acidosis appears in the urine in the form of salts of acids represents basic nitrogenous material, which normally would be excreted for the most part in the form of urea. Also acids, which are formed and liberated in this organ which conducts the heaviest traffic in ammonia or its equivalents, are generated in a part of the body where the chances for neutralization by ammonia should be exceptionally favorable. If it be granted that the phenomenon of neutralization of acids by ammonium has been observed with the greatest consistency and in the highest degree in two conditions: (a) a type of acidosis in which the acids are largely hepatogenous in origin, and (b) a type following the administration of acids—but that the phenome- non is found to be less marked or entirely absent in other types of acidosis the question would arise as to why the ingestion of acid should produce an effect like that of one type of acidosis and unlike that of the other types. Could this mean that in order to be neutralized to the highest degree by ammonium the free acids involved must be formed within or be introduced into the liver? In reviewing the literature one is struck by the fact that those who have observed increased excretion of ammonia following the administration of acids have almost without exception given the acid by the alimentary route. Winterberg (10) injected rabbits subcutaneously (two experiments) with 0.25 N acid and found the ammonia, sodium, and potassium content of the urine greatly 1 Stadie and Van Slyke (9) in their study of a case of acute yellow atrophy have recently concluded that the réle of the liver in these processes cannot be assumed in its entirety by the rest of the body. Only in patho- logical conditions, which are associated with almost complete destruction of the liver tissue, does the organism fail to metabolize in a normal fashion the amino-acids even though they be presented in high concentrations. br 4 7 ' 3 414 Ammonia Excretion after Acids increased. The total nitrogen was not correspondingly increased. But one of these animals died at the end of 24 hours and with the other the experiment was discontinued in about the same time. These two isolated experiments would scarcely seem sufficient to settle the point. One would have to consider that the local de- structive effect on the tissues of an amount of 0.25 N acid sufficient to cause death in 24 hours is not to be neglected. Whereas, the literature contains other reports (11, 12, 18) concerning the injec- tion of acids by vein we have not found any in which the behavior of the urinary ammonia was recorded. The experiments to be described were undertaken for the specific purpose of comparing the relative effects on the ammonia excretion in the urine of acid administration by the alimen- tary route and by the peripheral vein. Alimentary adminis- trations should subject the liver to the effects of a relatively large . quantity of free acid, while the latter would afford opportunity for the neutralization of more of the acid in the blood or tissues outside of the liver. EXPERIMENTAL. Methods. Female dogs of approximately 16 kilos body weight were used. They were fed a constant diet, consisting of milk 100 to 200 ce., bread 50 to 100 gm., sodium acetate 0.5 gm. twice daily at 6.30 am. and p.m. ‘To the morning feeding 100 cc. of physiological salt solution were added, which served to keep the volume of liquids the same in the experimental and the control periods. The sodium acetate was given to promote uniform elimination. Col- lections of urine were made twice daily by catheterization and irri- gation of the bladder. The cages were scrubbed and disinfected every 12 hours to prevent decomposition of occasional specimens passed spontaneously. This procedure gave consistent ammonia values. The urines were placed on ice without preservatives, the ammonia and total nitrogen being estimated shortly after the closure of the 24 hour period. The ammonia was estimated by means of formalin titration (Ronchese and Malfatti) as described by Wiechowski (14). The urine was well diluted and the titration was carried to the first R. W. Keeton — ALB permanent pink. This method was checked with Folin’s (15) aeration method. The practical agreement of the methods for present purposes is shown in the following protocol. Formalin titration. Folin method. 0.30 0.2 0.25 0.25 0.37 0.37 180 cc. of 0.1 nN HCl by stomach. 0.41 0.41 0.46 0.44 0.39 0.39 0.37 0.38 0.42 Not estimated. 250 ec. of 0.1 N HCl by vein. 0.46 0.45 0.51 0.51 0.52 Not estimated. 0.33 0.33 0.49 0.48 0.44 0.42 The total nitrogen was estimated by the Kjeldahl method, the creatinine according to Folin (15). On the morning of the experiment the salt solution was omitted from the diet. The acid was then administered either by stomach tube or into the leg vein. For the latter a cannula was inserted into the vein under local anesthesia, and the acid was introduced by gravity throughout the period of 1 hour. Experiments.—Five animals were used in the studies, and seven complete (gastric and intravenous administration) experiments made, but the protocols of only two animals are reported in detail. Dog 2 died during the course of the intravenous injection, the time of which was shortened from 1 hour to 30 minutes. Dog 3 was a young, highly irritable animal, which never gave consistent results during the control period. However, the results in this case were in keeping with those reported. The dose of acid in- jected into Dog 4 proved to be too toxic, so that at the end of 24 416 | Ammonia Excretion after Acids hours food was refused, the nose was hot, and the quantity of urine was much reduced. However, the output of ammonia and total nitrogen during the first 24 hours was the same as for the day preceding the experiment, in agreement with the findings to be reported in detail. On the succeeding days the total nitrogen Fiaal u- 2 SaNCanBB BE SG | aes me Et el ihe aa | | ~ a oe a a OVE EO CaRC anne ee eoaeee by vein | Wea Nittogen | | CI ae, | sPanneUBienfecsecsicati ECE PEE POEM ais is ict iak tet 1 baal deta vein ae | —2b0ce oH? bu Cuarti. Changes in nitrogen excretion after acid administration in Dog 1. In the ordinates of Charts 1 to 4 are expressed the values of total and ammonia nitrogen in grams, the latter multiplied by 10. In the abscissxe two spaces represent 12 hours. Record covers 48 hours before and after experiment. 200 cc.Holil] by stomach . C by vein Cuart 2. Changes in nitrogen excretion after acid administration in Dog 1. fell slightly below normal, but the ammonia increased. The results of the experiments on Dogs 1 and 5 are presented in the charts. In Charts 1 to 4 the quantitative output of ammonia nitrogen and the total nitrogen, expressed in grams, are plotted for 48 hours before and after the experimental periods. Chart 5 R. W. Keeton 417 shows the complete record of Dog 1 through two consecutive experiments. The nitrogen values here are expressed in 24 hour intervals. The figures from which the charts were compiled are to be found in Tables I to V, Chart 1 corresponding to Table I, ete. ae — 28, ea (Se Fe eS |_| | | cad US RF Pe Ge (EY i ie PS Fe To De ati b Lsieiaiat Foster ee eatele cat ares] Fi t 1 — Aokeon HERR RC CELE EEE EERE Er Eee etl |) Sere | Jn tas? oe asses See See - See Ee TEED AER val_|_| [a | | | iH Bae roger _| leh 1H | EE Eee eae: Bala | 180 cc. Yo FIC TL hy stomach 180cc.H0oH{C/ by vein Cuarr 3. Changes in nitrogen excretion after acid administration_in Dog 5. |_| [eal PAE | } 250 cc. Po LIC: by stomach L50cC Ho TIC: by vey Cuart 4. Changes in nitrogen excretion after acid administration in Dog 5. ha a st Mane a1 pase SEEHH Cuart 5. Changes in nitrogen excretion following successive acid administrations in Dog 1. In the ordinates of Chart 5 are expressed the values of total and ammonia nitrogen in grams, the latter multiplied by 10. In the abscisse two spaces represent 24 hours. All of the time (11 days) which elapsed in the course of the experiment is represented in the chart. Ammonia Excretion Following Administration of Acid by Stom- ach.—These graphs show clearly an absolute increase in the ammonia nitrogen without changes in the total nitrogen. In ——— = 418 Ammonia Excretion after Acids other words, some of the nitrogen which normally appeared in the urine in other forms (e.g., urea) has been diverted to the neutralization of the acid and now appears as ammonia in the urine. This simply confirms commonly accepted facts. Ammonia Excretion Following Administration of Acid Intra- venously.—The curve of total nitrogen gradually rises. The sud- denness of the rise appears to depend within limits on the quantity TABLE I. Changes in Nitrogen Excretion after Acid Administration in Dog 1. Nitrogen excretion per 12 hrs. Date. Total. In ammonia. 1921 gm. gm. Jan. 18, p.m. 2.86 0.17 “19, a.m. 2.43 0.17 <' 19, p.m. 2.41 0.14 fe, 20,4... 2.35 0.16 A | Gastric administration 125 cc. of 0.1 N HCl. | 20.3 |+-2.7 | 8.1 | O20la 3 8 “(Nia sailiG) tea gel P29 noe OLS 2a al ORO =e: Zh 0.015 Gly collic-acid,..7.6 gm... 42526 = 16.8 | 0.94 | 19.1 |41.5 | 9 0.013 - “ 7. “ (Nasalt)..| 16.8 | 0.88 | 18.5 |4+0.9 fai 0.015 FIGS prety Boor ich ya seer 0 | 0.84] 19.2 |+1.6| 7.5 | 0.015 Glycine, 9.5 gm. + NaHCOs, LORY OMe ceaerrtia it ss sc eee 20.1 | 0.84 | 22.9 |+5.3 | 10 0.013 (GINICOSE POSTINI Atee . e eee 217.8 1 OL "227 32)- Fae 7 0 0.014 - 50 “ + lactic acid, fo Lid DOs ot Le Seer SA ERAS Soe 217.0 | 1.04 | 22.2 |44.6 | 11 0.014 Glucose, 50 gm. + acetic acid, SOOT: ae caeunrres pesca ecte Te: ores. heck 198.25) 0.97 | 24.9 |+-7.2 | 12 0.013 The administration of these small quantities of materials to a dog is not easy, and the limitation of the method employed ap- pears to me to have been reached in this research. Rubner (20), speaking in the Prussian Academy of Sciences in 1913, said, “Die Wanderung der Stoffe bis zur Zelle zu verfolgen, sie quantitativ zu messen und experimentell zu variieren, das gehért, bis heute wenigstens, zu den unlésbaren Aufgaben. Es ist auch kaum zu G. Lusk 469 erwarten, dass sich in Biilde die entgegenstehenden Schwierigkeiten iiberwinden lassen.” It is realized that the methods employed here are relatively crude and the results often merely suggestive. XIII. SUMMARY. 1. The ingestion of 400 cc. of a broth containing 2.5 gm. of Liebig’s extract of beef increased the heat production of a dog from a basal level of 16.1 calories by 0.5 calory per hour. 2. The addition of 8 gm. of sodium bicarbonate to the broth caused no further change in the metabolism. 3. The addition of 3 gm. of acetic acid to the broth caused an increase of 3.1 calories per hour. The material was evidently rapidly consumed. 4. Lactie acid, 4.8 gm. given as before, raised the basal metab- olism by 2.7 calories. With 10 gm. of sodium lactate the heat production increased only 1.4 calories per hour, possibly because the alkali favored its transformation into glycogen. 5. Glycollic acid, 7.6 gm., with twice the number of potential H ions contained in it than were present in 3 gm. of aceticacid, in- creased the metabolism by 1.5 calories or less than half that ef- fected by acetic acid. A like quantity given as glycollate of sodium increased the metabolism 0.88 calory. 6. Hydrochloric acid, 1.8 gm., caused the metabolism to in- crease 1.6 calories per hour. 7. Glycine, 9.55 gm., containing 20 calories and neutralized with sodium bicarbonate, caused the heat production to rise 5.3 calories per hour. Neutralization therefore did not avail to reduce the activity of the product. 8. Glucose, 58 gm., and glucose, 50 gm., plus lactic acid, 8 gm., manifest exactly the same increases above the basal metabolism, 4.7 and 4.6 caloriesrespectively. Lactic acid therefore behaves like an intermediary metabolite of glucose and not like alanine which would have shown a summation effect. 9. Glucose, 50 gm., plus acetic acid, 3 gm., shows an increase of 7.23 calories per hour. This indicates a summation of the in- fluences of the two factors involved; it is identical with the be- havior of glucose and fat when given together; it supports the con- ception that acetic acid is an intermediary metabolism product 470 Animal Calorimetry TABLE Xll—General Summa@ ; Calories. } Experi- Date. ment Time. CO2 Oz R.Q. H:20 Be | No. . Non- : — if Protein. protein: Indirect.| Dire | 1920 gm. gm. gm. Jan. 12 4 |11.52-12.52| 7.02 | 6.86 | 0.74 | 8.05 3.50 | 18.96 | 22.46 | 21.38 52— 1.52) 6.524) 6509 | 0.78") 72 3.50 | 16.59 | 20.09 | 21.19% 42.55 | 42.52 ane (9 6 |11.15-12.15) 5.94 | 4.94 | 0.87 | 6.74 3.50 | 13.16 | 16.66 | 17.62%) 6.62 3.50 | 15.43 | 18.93 | 19.92 35.59 | 37.54 « 46 | 7 |11.11-12.11] 6.05 | 5.46 | 0.81 | 6.09 3.50 | 14.62 | 18.12 | 17.83 12.11- 1.11] 5.98 | 5.60 | 0.77 | 6.08 3.50 | 14.91 | 18.41 | 16.5 36.53 | 34.38m i Of | 10. |11:28-12.28) 7.3885 41) 1 64 |) 808 3.50 | 17.99 | 21.49 | 19. 12.98- 1.286. 24 leyesgac0 S24) 7 oT 3.50 | 14.92 | 18.42 | 19. 39.91 | 38.70 #97), 11. 111, 30-1230) 752aee aS 5] 10.86 | oni 3.50 | 17.92 | 21.42 | 19.94 12.30- 1.30] 6.65 | 5.92 | 0.82 | 7.20 3.50 | 16.22 | 19.72 | 20.4 41.14 | 40.34 : : : 0.83 | 8.53 3.50 | 16.80 | 20.30 | 22. 12.44- 1.44) 6.57 | 6.40 | 0.75 | 8.33 3.50 | 17.45 | 20.95 | 20.62 41.25 | 42.77 44, )0.81 |. 8.90 3.50 | 17.95 | 21.45 | 19. .87 | 0.74 | 8.09 3.50 | 15.68 | 19.18 | 19.139 “ 28.| 14 |11.23-12.23 6.51 5.87 | 0.81 | 8.20 3.50 | 16.00 | 19.50 | 20.6 12.23- 1.23] 6.49 | 5.85 | 0.81 | 7.98 3.50 | 15.93 | 19.43 | 22. Feb. 5 | 19 /11.47-12.47| 5.91 | 4.92 | 0.87 | 8.49 3.50 | 13.09 | 16.59 | 16.87 12.47- 1.47] 5.52 | 4.99 | 0.80 | 7.76 3.50 | 13.03 | 16.53 | 16.65 io 20 {11.30-12.30) 7.37 | 6.09 | 0.88 | 9.49 3.50 | 17.13 | 20.63 | 14.46 12.30— 1.30} 7.18 | 5.33 | 0.98 | 8.59 3.50 | 14.88 | 18.38 | 17.08 g XIX. yee =e ydy temperature. End. |Difference. 38.06 37.76 ) 37.90 L 37.96 3 38.21 5 38.14 3 38.12 5 38.45 ‘1 38.10 7 37.73 —0.20 —0.19 —0.30 +0.05 +0.08 —0.11 —0.11 +0.30 —0.11 —0.34 Morning weight. kg. 10.60 10.57 10.57 10.57 10.55 10.52 10.32 10.35 10.27 10.30 G. Lusk Behavior of dog. Quiet. Single movements. Quiet. Slight movements. Moving 1 minute. Quiet. Slight movement. Occasional movement. Slight movement. Single movements. Quiet. Moving 3 minutes. 3 slight movements. Quiet. Slight movement. Many slightmovements. Quiet. “ Quiet. “ IE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XLIX, NO. 2 471 Food. Se eS Acetic acid, 3 gm.; meat extract, 2.5 gm.; water, 400 cc. at 10.40 a.m. Basal metabolism. Meat extract, 2.5 gm.; water, 350 ec. at 10.30 a.m. Lactic acid, 4.8 gm.; meat extract, 2.5 gm.; water, 400 ce. at 10.17 a.m. Lactic acid, 4.8 gm.; meat extract, 2.5 gm.; water, 400 cc. at 10.40 a.m. Acetic acid, 3 gm.; meat extract, 2.5 gm.; water, 400 cc. at 10.21 a.m. Acetic acid, 3 gm.; meat extract, 2.5 gm.; water, 400 ce. at 12.35 p.m. Water, 400 cc.; meat extract 2.5 gm. at 10.34 a.m. Basal metabolism. Glycollic acid, 7.6 gm.; meat ex- tract, 2.6 gm,; water, 400 cc. at 10.35 p.m. oy ” 6 ce ABI ; ZFI— 1°99 | O'TE | $°6¢ | 00S ake Saw ates » 8 02 *9°q bs L2°o+ is. + 0'FI— 8°e¢ | 8's¢ | 0'6F | 4°89 | 00S ‘ud g “prow 01908] pa 0z “390 6I6I 5 GI+ 9°242— | O'FF | OIF | FIP | FGF | F'8E | OOF "ud g*y ‘plow o1[[OOATH) 9 02 “494 = T'e+ Lt+ T°8¢ | L°8¢ | I°2¢ | 00F ms § ” G g ‘uve : O¢6T 0 T'e+ G°8¢ | ¢°8¢ | 00 ws G°Z ‘plow o1W90V if to 90 = Lte+ t+ 0°19 | T°09 | 6’6¢ | OST » 0S » g I “AON S Lot G'Z— G’6S | OLE | $78 | 00 "ud Og ‘esoonty z Fo ~ tah. =r ‘Vinge re fi ‘ or 7 i o? tw i + 44) blair 0 paw 21 17 ae 4. 4 woh “SS soiecsiiae AY me, ’ hee ’ veh a . Be hh ¥ * 7. THE FATE OF SULFIDES IN THE BLOOD.* By HOWARD W. HAGGARD. WITH THE COLLABORATION OF THOMAS J. CHARLTON. (Investigations Performed for the U. S. Bureau of Mines in the Laboratory of Applied Physiology, Yale University, New Haven.) (Received for publication, October 18, 1921.) The investigation embodied in this paper was primarily under- taken in an endeavor to determine in what manner H,S is trans- ported in the blood after inhalation of this gas. That it is trans- ported in the blood in a loosely combined and readily dissociable manner is evident from the fact that, when HS is injected into any of the body cavities, it is almost at once detectable in the expired air (1). On the question in what form hydrogen sulfide combines with and is carried by the blood there is diversity of opinion, however. Two modes of transportation have been canenicde That the hydrogen sulfide combines with the hemoglobin to form a sulf- hemoglobin compound; and that the gas in its capacity as an acid unites with the alkali of the blood to form sodium sulfide. Since both of these combinations are dissociable both views neces- sarily assume a trace of H.S in simple solution in the plasma. Paradoxically, although the first of these views is definitely disproved by evidence contained in the literature, the belief that inhalation of HS is followed by the formation of a sulfur hemo- globin compound is still generally prevalent. The persistence of this idea is no doubt due in part at least to the need for an explanation of the rare disease, sulfhemoglobinemia, one of the endogenous cyanoses. * This is one of a series of papers dealing with hydrogen sulfide poisoning. Other papers of this series will probably appear in the American Journal of Physiology, the Journal of Pharmacology and Experimental Therapeu- tics, and the Journal of Industrial Hygiene. 519 520 Sulfides in Blood The reaction of hemoglobin to H2S has been the subject of much study. Hoppe-Seyler (2) in 1863 and later Araki (3), by passing a stream of H.S through oxygenated blood, obtained a dirty greenish pigment. It showed an absorption spectrum with two bands in the red, one quite close to C and the other midway between C and D. This compound was designated sulfmethemo- globin. The green discoloration seen in cadaveric decomposition was assigned to the postmortem formation of sulfmethemo- globin. Gamgee (4) in 1898 branded the sulfmethemoglobin of Hoppe-Seyler and Araki as a mixture of decomposition products. Harnack (5), a year later, while also emphatically denying the exis- tence of this substance, obtained in solution by the action of HS upon reduced hemoglobin, a compound which he designated as sulfhemoglobin. Its absorption spectrum gave but a single band between the Fraunhofer lines C and D and extending from d.610 to d.625. - In 1907, Clarke and West (6) verified Harnack’s work and attempted to isolate the compound, but were unsuccessful. They noted that complete reduction of the hemoglobin was essential to the formation of this substance. They found further, that very high concentrations of HS were necessary to force a union between hemoglobin and this gas. These concentrations were far higher than those which would be instantly fatal if mspired. The presence of powerful reducing agents, such as phenylhydra- zine, however, greatly facilitated the reaction and rendered a partial combination possible at relatively low concentrations of H.S and even in the presence of oxyhemoglobin. It is on the basis of. this last observation that the occurrence, during life, of blood giving the spectrum of sulfhemoglobin—the characteristic of the disease known as sulfhemoglobinemia—is explicable. Wallis (7) found that the formation of sulfhemo- globin in the blood of patients suffering from this disease, is due to the presence in their blood of a strong reducing agent, a hydrox- ylamine derivative, which presumably comes from the splitting of protein by a nitrosobacillus which inhabits the buccal cavities. A mere trace of H»S is necessary under these conditions to form sulfhemoglobin and this might be derived from the intestinal tract. — a > ie H. W. Haggard 521 From these facts it is apparent that the formation of sufhemo- globin within the living body is primarily dependent upon an abnormal condition of the blood, rather than upon any function of the hemoglobin as the transporting agent of H.S. Within the body of a healthy living individual—one whose mouth does not contain this nitrosobacillus—complete reduction of the blood would be essential before any combination of the gas with hemo- globin could occur. As a postmortem change, however, sulfhemoglobin may be rapidly formed through bacterial action. This change, we may presume, consists first in a reduction of the blood and then a combination of the hemoglobin with the H2S liberated during the process of decomposition. It is a highly significant fact, that the blood taken very soon after death from H.S poisoning does not show the spectrum charac- teristic of sulfhemoglobin and is abnormal, if at all, only in the degree of its reduction (7, 8). A fortiori hemoglobin is not the normal transporter of HS. Diakonow (9) was apparently the first to point out that, through its properties as an acid, H.S should act upon the bicarbonates of the blood plasma to form sodium sulfide. He demonstrated this experimentally upon bicarbonate solutions. Following this lead Pohl (10) came to the conclusion that H.S must be transported in the blood in this form alone. In support of this view he pointed out the remarkable similarity between poisoning with H»S gas and that induced by injection of sodium sulfide. It seems plausible that, to some extent at least, sodium sulfide would be formed within the plasma after inhalation of this gas. For this reason the following experiments were undertaken. In the main, however, they have led to quite another conclusion. EXPERIMENTAL. The following experiments which were repeatedly performed, show a marked difference between the reactions toward H.S of sodium bicarbonate solution on the one hand and of plasma on the other. The former combines with the gas to some extent, pre- sumably as NasS and even when washed free of dissolved H.S gives a persistent test for sulfide. The plasma, however, retains 522 Sulfides in Blood no sulfide, detectable as such, in any form other than the dissolved HS gas. Evidently plasma after exposure to a moderate amount of H.S does not contain NaS. Experiment 1. The Action of H2S upon Sodium Bicarbonate Solution.— 5 ec. of 0.2 per cent sodium bicarbonate solution were shaken in a flask con- taining an atmosphere of 0.5 per cent H.S in air. The flask was then opened and air passed through until the odor of H.S was no longer detected. The solution was then tested, and gave a sulfide reaction both with lead acetate and with ammoniacal sodium nitroprusside. The passage of pure air, 5.5 per cent COs: in air, or oxygen through the liquid for 3 hours did not render it incapable of giving a sulfide test. The addition of dilute HCl to the fluid was attended with the evolution of H.S. Experiment 2. The Action of H2S upon Plasma.—Dog’s plasma was shaken in a flask with 0.5 per cent H.S and then aerated. The plasma failed to give any of the above tests for sulfide nor was H2S evolved upon the addition of dilute, HCl. This experiment with plasma (Experiment 2) affords a clear-cut and decisive negative on the question to which it is primarily directed. The striking difference from the result with bicar- bonate solution (Experiment 1) prompts the further question as to just what does happen when H.S is brought into contact with plasma. To investigate this question, further experiments were there- fore performed. Plasma was exposed to a definite volume of H.S in air, nitrogen, COs, or oxygen and then washed free of dissolved H.S with the same atmosphere.. The hydrogen sulfide recoverable from the gas was estimated quantitatively. In some cases the COs-combining power of the plasma before and after treatment with H.S was determined as a means of following any change in the sodium bicarbonate of the plasma. Experiment 3. The Reaction between H,S and Plasma or Blood.—The CO:;combining power of a sample of normal plasma from dog’s blood was determined at 40 mm. partial pressure CO2. Some of the plasma was then evacuated of gas by means of a suction pump, and 5 ec. samples were pipetted into 1 liter flasks containing, in successive tests, atmospheres of air, oxygen, nitrogen, and 40 mm. CO, in air. To all of these atmospheres, 0.5 per cent H,S had been added. Each flask was rotated for 1 minute. The H.S which could be recovered was then determined by aerating the flask with the same atmosphere with which it was filled, minus the H.S, and drawing the gas through a bead tower containing 0.01 N iodine (11). H. W. Haggard 523 In each case at the end of the aeration the plasma was tested for sulfide; but only negative results were obtained. Samples of whole blood were treated in the same manner except that the initial evacuation was omitted. Table I embodies the results obtained. From Experiment 3 it appears that not only does H.S fail to form sodium sulfide when acting upon blood or plasma but that a portion of the gas is actually destroyed. The destruction of ’ TABLE I. CO2-combining power of plasma at 40 mm. HS Atmosphere in flask. parka Deere: Before | After expos- | expos- ure. ure. Reaction between plasma and H2S. Peis pete ce cc per cent J SER EDTIG WEA OSI 0) ial 8 Io re lo 54 40 | 3.10 | 2.10 | 40 eee ANG ee ES, jo) fe ss 54 46 | 2.941 1.66] 36 “os SSN CU & Ce ee ee 54 At “ha252"|"2. 088] vas “ with CO, at 40 mm. and 5.7 ce. of H»S| 54 45 | 3.42] 2.28} 40 AGO s x s — . ee TOBN 2s a] (3) a 3 1S) ABN a a a ® A © paTtoy ‘jeoTY on Oo- $s 08 7 3 \Oel Za ! él ay |00¢ PN | SA” 2 : 22 a aenlie ove THO T9ATT poo Oma TEP aoe poo ONHEN g*t. ie OHEN TORN ee cag Nee ({punoz)| xe6q9 0°ot 4 suveg f£aen g°8 Es 1—staq] G6 S40 PeTTOH| o°6 | OTH g°6 | OZ Tull al 3°61 qeousy 6°0e SUPT yey , | | aaa See cb es : 10 McCollum, Simmonds, Shipley, and Park 11 Cod liver oil, we have repeatedly observed with different diets, protects growing rats against the injurious effects of lack of calcium and enables them to grow and appear well nourished for a considerable period, where they would fail to grow and would be very inferior with even much greater amounts of butter fat instead of the cod liver oil. It is especially remarkable that whereas 3 per cent of butter fat suffices when added to food mixtures of the type under discussion to meet all the requirements of the rat for fat-soluble A and for the specific calcium-depos- iting factor, if there be one, when the diet contains favorable concentra- tions of calcium, 10 to 20 per cent of butter fat fails to shield them in a manner at all comparable with 1 per cent of cod liver oil. Again, it is remarkable that animals specifically fasted for calcium are protected as effectively by 1 per cent of cod liver oil as by3 per cent or higher planes of intake. We do not believe that our samples of butter fat could differ in the concentration of the calcium-depositing substance, since, as will be shown in later charts, when the diet contains calcium in amounts ranging from 0.1 to 0.5 per cent of the carbonate, butter fat becomes adequate to supply the nutritive needs of the body and the differences in the effects of cod liver oil and of butter fat in the diet tend to disappear. Cuart 3. Lot 2821 was fed a diet essentially comparable to those dis- cussed in the preceding charts except that its content of 20 per cent of casein raised the phosphorus content of the food mixture. Like the preced- ing diets it was very poorin calcium. 5 percent of butter fat did not suffice to protect the animals against lack of calcium. This chart should be com-— pared with Chart 4, Lot 2822, which differed significantly only in containing 2 per cent of cod liver oil and 3 per cent of butter fat. Lot 2821 was very poorly nourished and became badly deformed. The coats of these rats were rough and thin (see Fig. 1) and they aged very early. Cuart 4. Lot 2822 had a diet exactly similar to that of Lot 2821 (Chart 3) except that 2 per cent of cod liver oil replaced 2 per cent of butter fat. This modification of the diet made a remarkable difference in their growth and well being (see Fig. 2). The former were stunted, infertile, and short lived. The latter grew to full normal size, presented a well nourished appearance, and were fairly fertile, and succeeded in rearing a considerable number of their young to the weaning age. The young were puny, pot- bellied, almost completely stunted in growth, and died early. The mothers declined rapidly after nursing two or three litters. The males, while well nourished for an interval following the completion of growth, soon pre- sented a poorly nourished appearance and aged early. The effect of the cod liver oil was to make the animals in some degree immune for a time to the injurious effects of lack of calcium. Even ten times as much butter fat could not do this. If a small addition of caleium were made to this diet the butter fat would supply sufficient of some organic oe es EL XII Studies on Experimental Rickets. 12 13 McCollum, Simmonds, Shipley, and Park 14 Studies on Experimental Rickets. XII substance intimately associated with the development of the osseous system, and the animals would develop normally. In order to differentiate between the nutritive value of cod liver oil and of butter fat in a qualitative way it must be administered with a diet poor in calcium, for with an ade- quate supply of calcium and phosphorus in the food either serves equally well to supplement such diets as are here discussed. Cuart 5. Lot 2766. This and succeeding groups of experimental animals illustrate the comparative value of cod liver oil and of butter fat in the presence of suboptimal amounts of calcium, where the phosphorus content of the diet is near the optimum. A small amount of calcium (0.045 gm.) was added in the 5 per cent of milk powder but this was too small to con- tribute much to the well being of the animals. These records should be: compared with those of Chart 6 (Lot 2765), whose diet was almost identical except that it contained 1 per cent of cod liver oil instead of 10 per cent of butter fat. The rats were protected in a remarkable way against the effects of lack of calcium by this small amount of cod liver oil. 10 per cent of butter fat with this diet afforded some pro- tection, but although this amount is at least three times that required to meet all the needs of the rat for protection against ophthalmia and to enable it to develop a normal skeleton when the diet contains the optimal amount of calcium, it fails to supply enough of some substance intimately concerned with bone formation when the calcium intake is low. These results point to the existence of a specific calcium-depositing substance distinct from fat-soluble A (antixerophthalmic substance). On this diet the animals grew slowly but never attained the full adult size. The females were capable of producing several litters each but the infant mortality was high and they early developed signs of senility. The second generation confined to this food supply was greatly stunted and inferior. Their forms were very short and stocky. They had large depos- its of body fat. Cuart 6. Lot 2765 should be contrasted with Lot 2766 (Chart 5). The significant difference in the composition of the diets of these two groups was in the nature of the fats which they contained. Both diets were far below the optimal in their content of calcium, but were otherwise well constituted. 1 per cent of cod liver oil very effectively protected these animals against the harmful effects of calcium starvation because of its content of some organic substance which appears to be distinct from that substance (fat- soluble A) which is essential for growth and is a specific agent in preventing ophthalmia of dietary origin. The protection afforded by the cod liver oil is not complete. It consists in enabling growth to proceed and causes the animals to appear externally to be well nourished. This is very apparent when we contrast the animals described in Chart 5, which had the same diet with 10 per cent of butter fat, with those in Chart 6, which had 1 per cent of cod liver oil. This contrast we have repeatedly observed in rats fed other diets low in calcium where fat-soluble A was in one case supplied by cod liver oil and in the other by butter fat. 15 McCollum, Simmonds, Shipley, and Park tO vd fe ie a Va fl O2el O JdA TORN i =) OU UT es Ey suveqg faeyy! | seed S40 pet low | eoTH| & 7B Ou Reel G9L2 JOT *9 | DEVO | McCollum, Simmonds, Shipley, and Park 17 That the animals receiving cod liver oil were in a state of nutritional instability notwithstanding their good external appearance and fertility, is shown by the tendency of the females to collapse and die while nursing asecond or third litter of young. Their skeletons were very poorly calcified and when the animals were boiled in water it was impossible to separate even the bones of the pelvis. The femur remained intact but was fre- quently deformed in second or third generation animals. The skull bones disintegrated in this treatment. Succeeding generations tend to fail on this diet, but not as rapidly as rats on similar diets containing butter fat even in large amounts (ten to twenty times as much fat as in the cod liver oil diets). Thesmall amount of calcium added in the 5 per cent of milk powder exerted an observable effect in improving the well being of these rats. This is easily seen by comparing Lots 2732 and 2733 (Chart 2) with Lot 2765 (Chart 6). Cuart 7. Lots 2957 and 2958. These groups, and those described in later charts, illustrate the beneficial effects of adding small amounts of calcium to the standard diet employed in most of the experiments here de- scribed. A comparison of Chart 1 with Chart 7 shows how marked is the effect in promoting the vitality of rats by the addition of even 0.1 to 0.2 per cent of calcium carbonate to the diet. Lot 2957 failed to grow as well with 0.1 per cent as did Lot 2958 with 0.2 per cent addition. The contrast is much greater in the vigor and capacity to grow of the young produced by these two groups. 0.2 per cent of calcium carbonate made it possible for the second generation to develop fairly well and produce young, whereas the addition of but 0.1 per cent did not enable the young appreciably to grow or to extend their lives beyond about 60 days after weaning. Both groups aged early and had abnormal forms. They were short and stocky. Even 1 per cent of cod liver oil makes rats on this diet develop long, lithe forms, although they fail early. Cuarts 8 and 9. Lots 2952 (Chart 8) and 2953 (Chart 9) should be com- pared with Chart 7. The diets were the same except that in Chart 7 the fat addition consisted of 8 per cent of butter fat, whereas in Charts 8 and 9 it was 2 per cent of cod liver oil. Lot 2952 had 0.1 and Lot 2953 had 0.2 per cent of calcium carbonate added. The great superiority of cod liver oil over five to ten times as much butter fat is easily seen. The general appearance of these animals was superior to the animals fed butter fat. Fertility was high and the mortality was low, although the nursing period was frequently prolonged beyond the normal time. The young did not look sleek and well nourished while depending on the mother, but later when placed on the family diet greatly improved in appearance. Cuarts 10 and 11. Lot 2959 (Chart 10) should be contrasted with Lot 2954 (Chart 11). These show the growth and fertility of rats fed the basal diet discussed in preceding charts to which 0.3 per cent of calcium carbonate was added. In the former, 8 per cent of butter fat was added, whereas 2 per cent of cod liver oil was added to the latter. Both groups grew well, oro, os Se Saar oe Ee 4 eee ce, . a ws ee oo —— ee LG6 é :HOT3 By LOT *4 GYVHO ss o NVRadadad ¢ ice Rolled oats it . 0989 : £OoHEN ees £aey | | eet | | Bi 22 ee Ne a ee | ee mi | aN aN7 oe | | el JL a FECES PR PLEASE NOCD GA0ccoqQgnnMNsaD s ee se ee aao00qCceo AO® | acd an aa Eh at qssso suseq f£asyy | seed 8480 ee McCollum, Simmonds, Shipley, and Park 23 but it is easy to see that the group receiving the cod liver oil was superior to those getting butter fat. It is apparent from these and preceding charts that as higher additions of calcium are made to this otherwise satisfactory diet, the difference between the dietary qualities of butter fat and cod liver oil tend to disappear. This is further emphasized in Charts 12 to 15, in which higher planes of calcium intake were furnished. Fertility was high in both these groups and rearing of the young was the rule. In these respects the cod liver oil group was more successful than the butter fat group. -There was a corresponding difference in the smart- ness of their appearance. CuHarts 12 and 13. Lots 2960 and 2955 correspond in every way to the two groups described in Charts 10 and 11, except that 0.4 per cent of calcium carbonate was added in each case. One diet contained 8 per cent of butter fat and the other 2 per cent of cod liver oil. The well being of these two groups presented less contrast than did those in Charts 10 and 11, but it was still possible to detect that those receiving cod liver oil were better nourished than those getting butter fat. This was shown especially in the behavior of the second generation, those from the cod liver oil achieving greater size and presenting a smarter appearance than those from the butter fat group. The animals described in Charts 10 to 16 inclusive showed essen- tially the same degree of solicitude in caring for their immature young. Cuarts 14 and 15. Lots 2961 and 2956 continue the series now under discussion. They received 0.5 per cent of calcium carbonate in each case, and Lot 2961 had 8 per cent of butter fat, whereas Lot 2956 had 2 per cent of cod liver oil. With this moderate addition of caleium, which represents diets containing less than half the optimal content of this element, the differences in the supplementary value of butter fat and cod liver oil practically disappear. Both groups appeared to be nearly equally capable of growth and fertility and succeeded about equally well in rearing thgir young. The succeeding generations, including the third, showed but slight differences in vitality. There was, however, a slight advantage in favor of the cod liver oil. : Cuart 16. Lot 2839 completes the argument we are presenting, to the effect that there are differences in the dietary properties of cod liver oil and of butter fat, which we can explain only on the assumption that we are dealing with two uncharacterized dietary factors in butter fat and cod liver oil. One of these is the antixerophthalmic substance, fat-soluble A. The other we suggest is a substance which plays a more important réle in influencing the anatomic elements in the osseous system and may bedes- ignated as a calcium-depositing or phosphorus-mobilizing factor. Butter fat is richer in fat-soluble A than in the calcium-depositing factor. Cod liver oil is exceptionally rich in both substances. It has only been found possible up to the present time to demonstrate the differences between these fats by using diets poor in calcium, for with aAiIoos et SYS S| SAAS Gul ele / we 0 B60 (ie aes me 3 | a | Sele: na oo aed a pee Tl A ee BSA 26 ooaqaoo ownsd e* @« ee @e ez- @« oo aogm anc acd nOo°eoda = °o e LfO IS8ATL po 008 “ODHB Toe upess suseq AaBn BBeg 8460 poftoy e0Ty 02 Tey 7BouM HOT By bG62 LOI *ST| LYVHO AN, | le: et ACE eae Btu CH;—C= CH—C = O——> CH;—CH—CH.—C=0 7 HG CH; CH; Acetone Ethanal 2 mols | -H:0 Isovaleric aldehyde H de CH;—C = CH—CHg= aa = CH—C=0 da 3 CH + 2H H Ke, CH;—C = CH—CH.—CH,—C = CH—C =O du. Citral \CH; * ie ~ xX 41 \+ 2H ie ae N H | CH,—C = CH—CH.—CH.2—C = CH—CH,—OH ee H;—C= CH-—CH.-—CH.—CH—CH.—C = 0 CH; CH; H Geraniol a : | O-H Citronellal CH,—C = CH—CH.—CH,—C—CH = CH: | CH, H; | | Linalool oie fox | +2H | | Bi ol | ——— \/—O /ep | -| I| | Tinea ae POR oO halal ine ey Terpineol Isopulegol Pulegone eee a, O18 |—O-H is i | — #20 | + 5:0 Cineol y, | Bk + H202 ky ~N @ eo (ae Sor Ge Sa 0 WO ve ( —OH | Methyl-1-cyclo- | \ i as hexanone 3 aN } nthone Ss Limonene ek — OH | — 220 4 + | | | | O VOL 4\—OH J\=0 VAN I (ey fh oft Oper aA Acetone Dat YN UN aN Dihydro- ’Carveol Carvone Limonene carveol Peppermint (Mentha piperita) Spearmint (Mentha spicata) Compounds known to occur in the oils are underlined 33 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 34 Biogenesis of Oil of Peppermint This scheme or picture of how the oils of spearmint and pepper- mint may be elaborated offers, in spite of its tentative nature, numerous suggestions for speculation and experiment. Thus the striking parallelism in the relations of analogous compounds, for example of menthol and pulegone to isopulegol in the one mint and of dihydrocarveoland carvone to carveol in the other unmistakably suggests a common mechanism of formation. But most inter- esting is the thought that the Mendelian ‘‘factor,” if such it is, which governs the formation of the carvone group in spearmint and of the menthone group in peppermint, lies in the conditions affecting the reduction of citral. For, as outlined, the two groups are separated at that point by a common type of reaction. And it is easy to conceive that the conditions which caused the reduc- tion of the aldehyde group in spearmint were modified in cross- breeding to bring about the reduction of a carbon to carbon double bond in peppermint. CONCLUSIONS. Although it was found that the oils of American and Japanese peppermints are not strictly comparable in a_ botanical sense, still some interesting results are evident as a result of studying the problems originally suggested by that idea. The constituents of the volatile oils of Mentha pztperita, var. officinalis and Mentha spicata Hudson have been listed and com- pared. A possible scheme of the biogenesis of the most important constituents has been drawn up. It is evident that a great similarity exists in the reactions by which the menthol group on the one hand and the carvone group on the other are elaborated. _ The two groups may have a common precursor, and are each derived from it by a reaction involving the addition of two atoms of hydrogen—the difference being only in the point of reduction. The urgent need of a reliable examination of the volatile oil of Mentha aquatica becomes evident. It should finally be emphasized that these data are presented only as a preliminary survey of what is known and as a glimpse of what may be. Also it is desired to emphasize the opportunity existing in the study of these common plants for the elucidation of phenomena which may be hereditary in character. The work is being continued. THE CHEMISTRY OF THE OXIDATION OF SULFUR BY MICROORGANISMS TO SULFURIC ACID AND TRANSFORMATION OF INSOLUBLE PHOSPHATES INTO SOLU- BLE FORMS.* By SELMAN A. WAKSMAN anv JACOB 8S. JOFFE. (From the Department of Soil Chemistry and Bacteriology, New Jersey Agricultural Experiment Station, New Brunswick.) (Received for publication, October 13, 1921.) Oxidation of Sulfur by Microorganisms. When sulfur is added to unsterile soil, it is slowly oxidized to sulfuric acid; when the soil is previously sterilized, oxidation of the sulfur takes place only to a very limited extent depending upon the other chemical substances present. But when a sulfur- oxidizing organism is introduced, the sulfur is rapidly oxidized to sulfuric acid. This acid acts upon insoluble soil constituents such as calcium and magnesium carbonates, calcium silicates, and tricalcium phosphate, and brings them into solution. This process has been utilized by Lipman and associates (1916) for the transformation of the insoluble tricalcium phosphate into soluble forms by composting rock phosphate, sulfur, and soil to which the sulfur-oxidizing bacteria have been added. A few principles involved in these transformations, both by crude and pure cul- tures of the sulfur-oxidizing organisms, are set forth in this paper. As a result of a series of studies, several organisms have been isolated, which are able to oxidize sulfur under various condi- tions. The oxidation of sulfur under acid and alkaline condi- tions seems to be affected by different groups of microorganisms. A detailed study of occurrence, morpholagy, and physiology of * Technical Paper No. 54 of the New Jersey Agricultural Experiment Station, Department of Soil Chemistry and Bacteriology. 30 36 Oxidation of Sulfur these organisms is found elsewhere (Waksman and Joffe, 1920— 22). One of the most interesting organisms isolated by the authors is an aerobic, autotrophic, minute bacterium, Thiobacillus thioovidans Waksman and Joffe, which is able to oxidize sulfur to such an extent as to reduce the hydrogen ion concentration of the medium to a pH of less than 1.0, even in the presence of buffering materials. It derives its energy from the oxidation of the sulfur and the carbon from the CO, of the atmosphere. The nitrogen can be supplied in the form of inorganic or organic materials. In taking up the chemistry of the sulfur oxidation, attention must be called to the aerobic nature of the phenomenon. 28 -- 2H:O a 305 = 2H SO, 64 96 196 Thus, for 64 units of sulfur, 96 units of oxygen are required to produce 196 units of sulfuric acid. The effect of oxygen is, therefore, of prime importance in the oxidation of sulfur. Experiments with crude cultures of the organism reported elsewhere (Joffe, 1922) substantiate the theoretical suppositions based on the empirical equations involved in the chemistry of sulfuric acid. Mixtures of sulfur, rock phosphate, and soil inocu- lated with crude cultures of sulfur-oxidizing organisms were prepared; one set was aerated and the other left unaerated. The amount of phosphates brought into solution and the change in the hydrogen ion concentration, as expressed by the exponent pH of Sdérensen, were used as criteria. The aerated mixtures were leading and, after 100 days, the percentage increase of sol- uble phosphates in the aerated over the non-aerated was 6 per cent, with a similar correlation in the increase of the hydrogen ion con- centration. It is interesting to record here the fact that this biological process follows the laws of inorganic reactions. Accord- ing to the mass law, the velocity of any reaction depends on the mass of the active ingredients involved and is at any moment proportional to the molecular concentration of the reacting com- ponents and a constant, which is characteristic of the chemical nature of the reacting substances. Whatever transformations the oxygen undergoes in the metabolism of the organism, the end-product is sulfuric acid; an increase in oxygen tension increases S. A. Waksman and J. S. Joffe ; 37 the mechanism of oxidation of sulfur by the organisms. It is also possible that the oxygen from the air is not the only source; as pointed out above this particular organism derives its energy not from carbohydrates but from the oxidation of sulfur and is autotrophic in nature. Like green plants, the autotrophic organ- isms use the carbon dioxide from the air for structural purposes, but, unlike plants, these organisms accomplish it without the intervention of the photochemical reactions. The process of assimilation of carbon dioxide is accompanied by the splitting off of oxygen, which may also be used by the sulfur organisms in the process of oxidation. Oxidation of Sulfur in the Ordinary Cultivated Soil. Several typical experiments will be reported here to illustrate the mechanism of sulfur oxidation in the soil, both in the absence and in the presence of small and large amounts of rock phosphate. The sulfur and phosphate were added to the soil and well mixed. A crude well developed culture was used for inoculation. The moisture content of the soil was kept at an optimum by the addi- tion of water at weekly intervals. The cultures were incubated at 25-27°C. The pH values were determined colorimetrically, according to the method of Clark and Lubs (1917); the phos- phates and sulfates according to the method of the Official Agri- cultural Chemists (1916). The results tabulated in Table I represent the oxidation of small amounts of sulfur in the soil. In this case 22.5 mg. of sulfur and 90 mg. of rock phosphate were added to 600 gm. of soil. The results tabulated in Table II represent the oxidation of large amounts of sulfur when introduced with large amounts of rock phosphate into the soil. In this case 30 gm. of sulfur and 90 gm. of rock phosphate were mixed with 480 gm. of soil in small pots. When the course of change in reaction due to the sulfur oxida- tion, in the presence of tricalcium phosphate is studied, we find that the curve is regular till the pH reaches 2.8, then it becomes flat. This is a crucial point and, as long as there will be any phosphate left undissolved, the reaction will not go down very much, since at that point all the acid formed from the oxidation of the sulfur is used not in increasing the reaction of the medium, but in transforming the rock phosphate into soluble form. Once bee | 38 Oxidation of Sulfur the phosphates have been made soluble, the acidity begins to increase. This will be made clearer in the discussion of the sulfur oxidation by pure cultures in solution. TABLE I. The Oxidation of Small Amounts of Sulfur in the Soil. Period of incubation. pH value. days 0 6.2 3 6.2 9 6.2 15 6.0 22 5.8 29 5.6 39 5.6 56 ae 70 502 102 one TABLE II. The Oxidation of Large Amounts of Sulfur in the Soil in the Presence of Large Amounts of Rock Phosphate. : Citrate-soluble Period of incubation. pH value. ieee orn in a phe days mg. of SOs mg.of P 0 6.2 0.95 2.83 3 6.2 0.96 2.83 9 oO 15 3.4 3.60 4.28 ae, ee 20.80 7.13 29 3.0 39 3.0 30.80 14.76 56 252 Soo 20.67 85 2.0 102 1.8 19.58 Oxidation of Sulfur in Solution by Thiobacillus thiooxidans. When a proper medium is used, with sulfur as the only source of energy, the pure culture of the organism rapidly oxidizes the sulfur to sulfuric acid. To prevent a rapid change in reaction S. A. Waksman and J. 8. Joffe 39 buffering substances are used. The course of reaction depends chiefly upon the nature of the buffering agents. When soluble phosphates are used, the curve is more or less continuous; when insoluble phosphates are used the curve has a definite flat portion at a pH of 2.6 to 2.8, the point at which the insoluble phosphates become soluble, and, only after all the phosphate has gone into solution, the curve rises again. When more insoluble phosphate is added at this point, the curve reaction will be kept at the pH of 2.8 to 2.6, till all the insoluble phosphate has disappeared. ‘The medium used for this experiment consisted of sulfur, 10 gm.; (NH4)2 SO,, 2 gm.; MegsS0Os,, 0.5 £mM.; FeSO, 0.01 gm.; KH2PO,, TABLE III. The Oxidation of Sulfur by Pure Culture of Thiobacillus thiooxidans. Period of incubation. No Caa(POs)e. 0.5 per cent Ca3(POs)2. Gr See of days pH pH pH 0 4.4 5.0 4.4 3 4.4 5.0 4.2 7 3.2 4.4 Bas 11 2.2 2.8 2.8 17 t6 2.6 (3 20 1.6 2.4 eee 33 1.4 Zi 3.0 52 12 1.8 2.8 -*0.5 gm. of Ca3(PO,4)2 has been added per 100 cc. cf medium. 5 gm.; distilled water to make 1,000 cc. When Ca;(POx,)s is not used, 0.25 gm. of CaCl, has been added per 1,000 cc. of medium. The medium was placed, in 100 ce. portions, in 250 ec. Erlenmeyer flasks and sterilized on 3 consecutive days in flowing steam. The flasks were then inoculated with a pure culture of Thiobacillus thiooxidans by means of a loop, and incubated at 25-27°C. The results are presented in Table III. It has been pointed out elsewhere (Waksman and Joffe, 1921) that the optimum reaction for the activities of Thiobacillus thiooxidans lies at a pH of 3.0t04.0. If the reaction of the medium is less acid, the reaction changes in the beginning only very slowly, but, once the optimum is attained, the curve rises rapidly. 40 Oxidation of Sulfur A detailed experiment is next reported which will show definitely the relation between the sulfur oxidation, as demonstrated by change in reaction, accumulation of sulfates, and the transforma- tion of phosphates, as shown by the amounts of soluble phosphates and calcium. A medium containing 2 gm. of (NH4)oSO,, 0.5 gm. of MgSOx,, 5 gm. of KH»,PO,, and a trace of FeSO, per liter, was placed, in 400 ce. portions into fifteen 1 liter Erlenmeyer flasks. 3 gm. of Cas3(POx.)2 and 4 gm. of powdered sulfur were added to each flask. The flasks were inoculated with 1 cc. portions of a pure culture of TABLE IV. Course of Sulfur Oxidation as Indicated by Change in Reaction and Amount of Soluble Sulfates, Phosphates, and Calcium in the Culture Solution. Agsoteuiture. | pa | PRSpRstEe jn 100] Sulla ts mg. of P mg. of SOs mg. of Ca Control. 6.0 123 230 17.4 20 hours. 6.0 230.4 17.44 TQ) 5.4 88) 3/4 4.9 25226 248.0 24.74 Li K0) S65 a220 260.15 26.85 ia 3.0 200. 06 S222 Sle0 6.5 days. 26) 171.64 366.4 64.2 Sema 2A6 210. 04. 498.8 118.8 1ORSE sts 255.46 511.4 104.9 ISG. 253 350.00 450.6 101.4 195 Fs 2a 81.6 34 s ie! Thiobacillus thiooxidans and incubated at 25-27°C. At various intervals, small amounts of the liquid were taken out from four to six flasks, and determinations made of the pH value, of the sulfates, phosphates, and calcium in solution. The results based on the average of four to six determinations, are tabulated in Table IV. The course of sulfur oxidation is best followed by the change in the hydrogen ion concentration of the medium (pH value). Of course, where there are large amounts of buffering agents or insol- uble carbonates or insoluble calcium phosphate, a much larger eS. are, S. A. Waksman and J. 8. Joffe 41 amount of sulfur will have to be oxidized to bring about a definite change in the pH value. The amount of sulfur oxidized has been reported in Table IV as sulfates; it may be pointed out here that in practically all cases the sulfur oxidized, as indicated by the amount of residual sulfur in solution, has been almost quantita- tively transformed into sulfates. The phosphates and calcium columns in Table IV will be discussed below. Transformation of Insoluble Phosphate into Soluble Forms. The reactions involved in the conversion of rock phosphate (insoluble tricalcium phosphate) into soluble forms (di- and monocalcium phosphate and phosphoric acid) by means of acids belong to the type of reactions of heterogeneous systems. The rock phosphate minerals have no definite composition and the products formed are not always definite. In such heterogeneous systems the speed of the reaction is a function of a greater number of variables than in the case of a homogeneous system. According to Kazakov (1913), there are some factors which are common to both systems: and these are: (1) concentration of the reacting mass; (2) temperature of the reacting medium; (3) the amount of contact of the reacting substances; (4) the speed of diffusion of the reacting substances; and (5) catalytic agents. Besides these factors we have others in a heterogeneous system where solid solution phases occur. These are: (1) the size of contact surface;! (2) chemical composition of the solid phase; (3) the physical properties of the solid phase; and (4) the influence of formation of a solid phase as a result of the reactions. The factors; chemical composition of the solid phase, and the physical properties of the solid phase; have a tremendous influence on the speed of the reaction and they are the least known, since the chemical make-up of the rock phosphate is still obscure. 1 The size of the particles of the rock in the manufacture of acid phos- phate has an important influence. Theoretically, all other conditions being equal, the speed of solution of a solid in a liquid is proportional to the contact surface and in circular bodies (as we would suppose in finely powdered rock phosphate) the surface is proportional to the square of the radius; then, particles with a radius of 0.1 mm. will dissolve twenty- five times faster than particles with a radius of 0.5 mm. Soe . oe ee eee. «= 42 Oxidation of Sulfur The process on the transformation of insoluble phosphates has been the subject of study by a number of investigators in this country and in Europe. We may merely refer to the work of Cameron and Bell (1907) of the Bureau of Soils, of Schucht (1909), Meyer (1905), Stoklasa (1911), Kazakov (1913), and others. According to Kazakov (1913), the scheme of reactions involved in the formation of soluble phosphates are: Resultants obtained. No. H2S04 Liquid phases. Solid phases. 1 When added in ex-| H3;PO, + H.SO,4-+ sulfates | CaSO, cess. of Ca, Al, and Fe. 2 | Close to optimum.|! H;PO, +- sulfates of Ca, Al, | CaSO, and Fe. 3 | Optimum. H;PO,+ sulfate of Ca + | CaSO, phosphates of Aland Fe. 4 | Not enough acid. | H;PO,+ sulfate of Ca-+ | CaSO,+ part of phosphates of Ca, Al,and| undissolved Fe. phosphate. Before we go into a discussion of the scheme, we shall take up the experimental results of the transformation of the tricalcium phosphate into soluble phosphate through the oxidation of sul- fur by Thiobacillus thiooxidans. The culture medium given above has been used, with a slight modification: the KH.PO, was reduced to 1 gm. per liter and 1 gm. of Ca3(PO,)2 was added to each flask containing 100 ce. of medium. The medium was sterilized in flowing steam on 3 con- secutive days, 30 minutes each day, then the flasks were inocu- lated with Thiobacillus thiooxidans and incubated at 27°C. Only the pH values and water-soluble phosphates (in solution) are reported in Table V and Fig. 1. The results are based on aver- ages of four to six flasks. The column of soluble sulfates is of extreme interest. Ata pH of 2.6, a sudden rise in the amount of soluble phosphates takes place after the soluble sulfates have reached a maximum. ‘This is in accordance with the scheme suggested by Kazakoy. Up tc the point of pH = 2.6 to 2.7, the liquid phase consists of mono- calcium phosphate and gypsum, we therefore have a large amount S. A. Waksman and J. S. Joffe 43 of soluble sulfates. However, as soon as more sulfuric acid is formed through the oxidation of sulfur, the monocalcium phos- phate is attacked first, since in any reaction the liquid phase comes in first; the products formed are phosphoric acid and gypsum. CaH4(POs,)2 H.SO, -- 9H,.O=CaS0O,°2H2O0 a 2H;PO,4 The CaS0,-2H,0 is soluble to a certain extent in phosphoric acid, but is forced out from solution because of the fact that the phos- phoric acid reacts with the remaining tricalcium phosphate, form- ing again monocalcium phosphate until the reaction comes to an equilibrium forming gypsum and phosphoric acid. With the TABLE V. Course of Sulfur Oxidation and Transformation of Insoluble Phosphates. Aeeot culture. pH Solana in Soluble aeeouae in mg. of S mg. of P Control. 5.4 68.39 45.57 1 5.4 67.64 42.61 2 5.3 69.70 47.20 +t 4.6 73.79 55.00 6 3.5 8 2.6 152.53 ~ 103.56 10 2.6 109.7 93.00 12 2.6 78.54 186.30 15 2.4 87.6 207.28 accumulation of the phosphoric acid, more gypsum goes in solution and the soluble sulfates increase again. The continuous increase of the insoluble sulfates after all of the tricalcium phosphate goes in solution is then due to the further oxidation of sulfurie acid. The column of soluble phosphates also proves the mechanism of the process suggested by Kazakov. Here also we find a gradual increase of the soluble phosphates, since the amounts of sulfuric acid in the early part of the incubation period is small. As soon, however, as the pH reaches 2.6 to 2.7, which is the crucial point of the reaction, the soluble phosphates increase rapidly. Prae- tically all of the tricalcium phosphate goes into solution in 2 days after the crucial point is reached. | ——— 44 Oxidation of Sulfur | The course of conversion of insoluble phosphates in composts of rock phosphate and sulfur has been taken up by Joffe (1922). P 4 P| | lo | jk RE |p curve | | | | | in1oocc 4 of P setuble | |) ESS (Ptasdras durke| 16 gooce gh dE || a eT | ee H Se ™@ TMI EImEee > 120 He 3.4 100 sae eal Nea otal 70 Lal 54. 60 eA Days 0 150 acunae \ a Mo ee Incubation period Vig. 1. Change in reaction and increase of soluble sulfates and phos- phates by pure culture of Thiobacillus thiooxidans. In this case the reactions are not so apparent, since the indefinite chemical make-up of the raw phosphates introduces a great num- ber of side reactions. S. A. Waksman and J. S. Joffe 45 SUMMARY. 1. The curve of sulfur oxidation both in the soil and in solu- tion by pure and impure cultures of Thiobacillus thiooxidans is a growth curve. 2. The mechanism of sulfur oxidation to sulfuric acid by Thio- bacillus thiooxidans obeys the laws of inorganic catalysis. 3. The transformation of insoluble rock phosphate to soluble phosphates by the sulfuric acid formed from the oxidation of sulfur by Thiobacillus thiooxidans is similar to the process taking place in inorganic reactions. BIBLIOGRAPHY, Cameron, F. K., and Bell, J. M., U. S..Dept. Agric., Bureau of Soils, Bull. 41, 1907. Clark, W. M., and Lubs, H. A., J. Bact., 1917, ii, 1. Joffe, J. S., Soil Sc., 1922, xiii, in press. Kazakov, A. V., Moskau Inst. Agron., 1918, ix, 21-45, 57-68. Lipman, J. G., McLean, H. C., and Lint, C., Soil Sc., 1916, ii, 499. Meyer, T., Z. angew. Chem., 1905, xviii, 1382. Official Agricultural Chemists, J. Assn. Official Agric. Chem., 1916, i, 1. Schucht, L., Die Fabrikation des Superphosphats, 1909, 1-460. Stoklasa, J., Biochemischer Kreislauf des Phosphat-Ions im Boden, Jena, 1911. 5 Waksman, S. A., and Joffe, J. S., Proc. Soc. Exp. Biol. and Med., 1920-21, Xviii, 1; Science, 1921, liii, 216; J. Bact., 1922, in press. CHANGES IN THE REFRACTIVE INDEX OF THE BLOOD SERUM OF THE ALBINO RAT WITH TEMPERATURE. By F: 8S. HAMMETT anp IDA TELLER. (From The Wistar Institute of Anatomy and Biology, Philadelphia.) (Received for publication, November 3, 1921.) It is a well known fact that changes in the temperature of a solution induce changes in its refractive index (1). In as far-as dilute solutions of proteins are concerned it is apparent from the studies of Robertson (2) that these changes are due to changes of the refractive index of the solvent, usually water. This effect on the solvent must be taken into account when measurements of the refractive index are made at different temperatures. Non- recognition of this fact invalidates not a few papers dealing with studies of the refractive index of blood serum. Observations of the temperature variation of this laboratory for over a year have shown an extreme range of 15°C. In view of this fact it became necessary to determine the temperature correction factor for blood serum of the albino rat for use in connection with some refractometric studies now under way. The sera for examination were obtained from mature albino rats. The animals were etherized and the heart was exposed. A cut was made in the ventricular wall and blood from the beat- ing heart was collected in small test-tubes. These were immedi- ately tightly corked and allowed to stand for 2 or 3 minutes when coagulation was complete. The clot was broken up with a fine glass rod and the serum was separated from the fibrin and corpuscles by centrifuging for hour. At the end of this period the supernatant serum was poured into another small test-tube and again centrifuged for a like period. During the process of centrifuging the tubes were tightly corked to prevent loss of water by evaporation. There resulted from this procedure a clear serum which was transferred to the cell of the refractometer by means of a pipette. After mixing, a portion was taken for the 47 ae ee ee 48 Refractive Index of Blood Serum determination of the water content. The cap of the cell was then lowered into place and the cell was sealed with paraffin having a melting point of 58-60°. This prevented undue loss of water from the sample under examination. The instrument used for these tests was a Pulfrich refractometer made by Carl Zeiss at Jena. Readings were made to tenths of a minute. The instru- ment was connected with the temperature regulating apparatus described by Reiss (3). Since much of the work reported here was carried on in the winter months the water coming in from the outside contained considerable dissolved air which was liberated on warming and accumulated in the bend of rubber tubing con- necting the warming chamber of the cell with the outflow. This interrupted the even flow of water through the instrument and resulted in distressing irregularities of temperature. These were eliminated by inserting into the system a small bottle with a three-hole rubber stopper. Into the central hole there was placed a long glass open tube extending above the level of the tank from which the water is supplied. The other two holes served to connect the bottle with the system. The bottle was placed in the part of the system between the warming coil and the refrac- tometer. With this addition to the apparatus most of the bubbles of air were caught before getting into the refractometer and es- caped through the glass tube. When the cell had been sealed a reading was taken of the angle of refraction and of the tempera- ture—usually around 17°—of the enclosed serum. The tempera- ture of the cell and contents was then increased by warming the water circulating through the apparatus. The attempt was made to raise the temperature by steps of 1°. When the new tempera- ture level had been reached and maintained for at least 1 minute another reading was made. This procedure was repeated until the desired maximum temperature had been reached, usually around 35°. The apparatus was then rapidly cooled to approxi- mately the temperature at the beginning of the examination and a final reading was taken. This gave the effect of the heating — on the refractive index of the serum constituents. The cell was then opened, a sample of serum removed, and its water content determined. Thus any loss of water from the serum by vapori- zation during the examination was determined. F:. S. Hammett and I. Teller 49 As Hatai (4) has shown, the value of the refractive index of the serum of the albino rat varies with the age and to a less de- gree with the size of the animal. The animals in this series varied somewhat in size although they were all of about the same age. These differences in size resulted in differences in the ini- tial refractive index of the serum. In order that the values might be brought to a common basis for purposes of study the percentage difference between the refractive index of the serum and that of water at the initial temperature of observation was determined. The subsequent observed indices for the serum at the different temperatures were multiplied by this factor, thus making the curve of the change in refractive. index with temperature of the serum comparable with the temperature curve of water, the solvent. Any changes in the refractive in- dex due to the influence of temperature on the serum constituents other than water would then be shown by a deviation of the curve for the serum from that for water. The results of the observations reported in this paper were obtained from seven- teen sera. When the calculation noted above had been made it was seen that the sera fell into two groups with respect to their accommo- dation to the water curve. In the first group there were eight sera. These had been obtained from rats early in the winter. In the second group there were nine series of observations. These had been made during the early spring. The averages of the calculated refractive indices for each degree of temperature were determined for each group and are plotted in Charts 1 and 2, using the curve of the refractive index of water on temperature as the norm. Chart 1 shows conclusively that in this group (1) of sera the changes in the refractive index with temperature are solely due to the changes in the refractive index of the solvent water. This holds up to a temperature of about 29°. Above this point there is an indication of a tendency for the refractive index of the serum to increase more than does that of water. This curve tends to support the contention of Robertson (2) that when the solvent change is considered the influence of proteins in solution on the refractive index is independent of the temperature between 20 and 40°. In Group 2, however, the curve of which is plotted in 50 Refractive Index of Blood Serum Chart 2, it is seen that the refractive index tends to fall regularly away from that of water with the rise in temperature. The reaction of this group fails to support Robertson’s belief mentioned Water curve Se Cuart 1. Showing the refractive index of blood serum of the albino rat with increasing temperature. Group 1, winter rats.... Serum values. Water curve. above. This discrepancy may possibly be explained by the fact that the concentration of protein material in blood serum is considerably above: the concentrations used by Robertson in his F. S. Hammett and I. Teller 51 temperature studies. This greater concentration may bring into relief differences which in smaller concentrations would be within ~ 1.33335 ged: 2) Water qurve eae Serum qurve — 1.33515 Serun values _@8 @80@6 1.33295 as 1.53155 1.55135 Temp. 15° 20° eae 500 Org = Cuart 2, Showing the refractive index of the blood serum of the albino rat with increasing temperature. Group 2, spring rats.... Serum values. woeen----- Serum curve. ————— Water curve. the limit of error of the method of observation and hence unnotice- able. The difference in the temperature effect on the two groups is statistically valid as will be shown presently. =e 52 Refractive Index of Blood Serum In Table I there is given the mean, standard deviation, and probable error of the mean of the body length, body weight, and water content of the sera in both groups before and after the refractometric examination. The table shows that none of these are factors in the difference in behavior of the sera towards changes in temperature. Nor can the difference be attributed to a general difference in the state of digestion and absorption, since at least 18 hours regularly intervened between the last feeding and the time of the taking of the blood. The only fac- tor at present indicated is a possible seasonal variation, since the determinations of the first group were made on winter rats and those of the second group on spring rats. Although we would not care to stress this point, when the fact that in the spring there is an increased sexual activity in the rat is correlated with the TABLE I. Group 1. Group 2. Standard|Probable Standard|Probable Mean. | devia- | errorof | Mean. | devia- | error of tion. mean. tion. mean. Body length, mm........... 188 12.5 3.4 | 184 15.1 3.6 Body weight, gm...........| 188 26.6 tas) e180 50.1 | fees Water, before, per cent.....| 92.2} 0.51] 0.18] 92.2] 0.25 | 0.06 Water, after, per cent....... 92.0 | 0.45 | O-11 | 92:0 | (0 0.10 observation of Hatai (4) that irregularities in the course of the curve of the refractive index of the serum of the albino rat on age occur around the period of puberty, it would appear as if an interpretation based on such a seasonal modification is strengthened. Whatever the determining factor may be, it is evident that in some sera certain constituents other than the solvent water are so influenced by temperature changes that they give rise to changes in the refractive index entirely apart from those produced by the solvent. Robertson (5) has stated that the change in the refractive in- dex of a solvent is a function of the size of the molecules of the solute. This, of course, refers to conditions at uniform tempera- ture. It is not improbable that the state of the colloidal equilib- F. S. Hammett and I. Teller 53 rium existing in those sera showing regular deviations from the water curve with rising temperature may be relatively unstable and that as a result of the changes in temperature, changes in the degree of dispersion of the colloidal constituents take place. Such an assumption is supported by Robertson’s (6) observation that as protein solutions approach the point of coagulation there occurs a decrease in the refractive index. It should be noted in this connection that this change is not rapidly reversible since when the serum is cooled to the point where the initial reading was taken, the value of the refractive index is usually somewhat greater than it was at the beginning of the examination. Turning now to the practical application of these results, there is given in Table II, the mean, standard deviation, and probable error of the mean of the change in reading of the angle of refrac- TABLE II. Statistical Values of the Angles in Minutes of Refraction of the Two Groups. Group 1. Group 2. minutes minutes RULED SO SIRE ie ee RS ae ea 1.08 1.41 RUE rOL eVIAbIOUl ss. nee seis Ss os fe ec 0.23 0.36 ENO ple error Of MCAD Ss, Sees.) oc aa es keene 0.02 0.03 tion for 1° of temperature for the two groups. It is seen that in the first group each rise of 1° is accompanied by a corresponding increase of approximately 1 minute in the angle of refraction. In the second group the increase of the angle of refraction for each degree of temperature is 1.4 minutes, a somewhat greater value. This difference between the two groups is statistically valid and substantiates the curve in Chart 2. Notwithstanding the difference, an inspection of the index of variability shows that for all practical purposes the mean of the change in the angle of refraction of the two groups—1:25 min- utes—for each degree of temperature is an acceptable factor and gives results accurate within plus or minus half a minute of re- fraction.. This can be used for correction of the observed read- ings. Taking 20° as the standard temperature, when a refrac- tometric reading is made of blood serum at a temperature above 0 ee 54 Refractive Index of Blood Serum this value, 1.25 times the difference between 20° and the observed temperature should be subtracted from the observed reading. If the temperature is between 17.5 and 20° the difference times the factor is added. The following formulas are simple expres- sions of this relation; J =7—1.25(¢—20°) and I =7+1.25(20°—?) where J is the corrected angle of refraction; 7 the observed angle of refraction; and t the observed temperature. SUMMARY AND CONCLUSIONS. A study of the changes in the refractive index of the blood serum of the albino rat with rising temperature showed that two types can be distinguished according to the nature of the response. In the first type the changes in the refractive index coincide with those of the solvent water and can be attributed to this serum constituent. In the second type the curve of the change of refraction with rising temperature falls away from that of water. This demonstrates a participation in the response of serum con- stituents other than the solvent water. The causes of this dif- ference are unknown, although there is a possibility that a seasonal variation may be a determinant. It is certain that in this series the factors of body length, body weight, age, and water content of the serum both before and after the experiment, and previous. state of digestion and absorption are not the causes of the differ-. ence between the two groups. The correction for the reduction of the observed angle of re- fraction to the common base at 20° when readings are taken at different temperatures is obtained by use of formulas given in the text. These formulas hold for temperatures between 17.5 and 35°C. BIBLIOGRAPHY. 1. Biehringer, J., in Abderhalden, E., Handbuch der biochemischen Arbeitsmethoden, Berlin, 1910, i, 567. 2. Robertson, T. B., J. Phys. Chem., 1909, xiii, 469. 3. Reiss, E., in Abderhalden, E., Handbuch der biochemischen Arbeits- methoden, Berlin, 1915, viii, 84. . Hatai, S., J. Biol. Chem., 1918, xxxv, 527. . Robertson, T. B., J. Biol. Chem., 1909-10, vii, 359. . Robertson, T. B., J. Biol. Chem., 1912, xi, 179. ao COLORIMETRIC DETERMINATION OF URIC ACID. ESTIMATION OF 0.03 TO 0.5 MG. QUANTITIES BY A NEW METHOD. By J. LUCIEN MORRIS anp A. GARRARD MACLEOD. (From the Biochemistry Laboratory of the School of Medicine, Western Reserve University, Cleveland.) (Received for publication, October 27, 1921.) Determinations of uric acid, which depend upon the direct weighing of the substance as such or upon an estimation of the nitrogen content of an insoluble salt, have been largely replaced — by methods which quantitatively measure its oxidation. Any of these procedures, whether volumetric or colorimetric, are use- ful in proportion to their success in (a) separating the uric acid from other substances which might give a value in the oxidation reaction and in (b) making the actual conditions of the reaction as specific as possible for uric acid. Volumetric methods, in most cases, have not been sufficiently sensitive for satisfactory application to less uric acid than is found in 100 cc. of urine. Also the means for preliminary pre- cipitation of uric acid made use of in these methods were crude. One of us! determined conditions for a complete precipitation of uric acid in any amount with zinc salts, and found conditions under which permanganate oxidation could be used successfully on such quantities of uric acid as are found in 5 cc. of urine. The much smaller amount of uric acid found in blood is below the limit of accuracy which it is possible to attain with the zinc pre- _ cipitation-permanganate oxidation procedure. Folin and Macallum? recognized the need of separation of uric acid from interfering substances and attempted the removal of polyphenols from urine residues before oxidizing uric acid with 1 Morris, J. L., J. Biol. Chem., 1916, xxv, 205. 2 Morris, J. L., J. Biol. Chem., 1919, xxxvii, 231. 3 Folin, O., and Macallum, A. B., Jr., J. Biol. Chem., 1912-18, xiii, 363 55 56 Colorimetric Estimation of Urie Acid alkali phosphotungstate. Folin and Denis‘ improved this separa- tion by their adaptation of the Salkowski’ precipitation of uric acid to precede the phosphotungstate oxidation. The method described in this paper makes use of the zine salt separation of uric acid which has proved very satisfactory in our hands and determines the uric acid so separated by a new colorimetric pro- cedure which possesses the double advantage of greater speci- ficity and greater color obtainable for unit weight of uric acid. With slight variations it is equally applicable to urine and blood. In the latter case we have removed the proteins by the tungstic acid precipitation of Folin and Wué and found the procedure satisfactory for our purpose. The results which we have obtained with the new method are, we believe, more dependable and uniform than the usual ones obtained with any of the procedures based upon silver precipi- tation. One cause for irregularities in the results obtained by means of the silver methods is inherent in the metal used. Sil- ver solutions, even when protected from light, soon become clouded with, and eventually -precipitate, a form of “reduced silver.’ The acid silver lactate solution used by Folin and Wu,* is noticeably cloudy soon after preparation and develops a heavy precipitate on standing. The ammoniacal silver magnesium solution of Benedict and Hitchcock’ also forms a scum around the neck of bottles, spouts of dropping bottles, ete., which occa- sionally contaminates the reagent added and thus gives erroneous results. This “reduced silver,” when taken up in sodium cyanide and the mixture made alkaline with sodium carbonate and phos- photungstic acid reagent added, gives a deep blue color even when taken in small amounts. (The color can also be developed by reversing the order of these reagents.) This possibility of the presence of such a substance is sufficient to cast doubt upon any determination which makes use of a silver precipitation. It is evident that the urine procedure of Folin and Wu! is particularly 4 Folin, O., and Denis, W., J. Biol. Chem., 1912-18, xiii, 469. 5 Salkowski, E., Virchows Arch. path. Anat., 1870, lii, 58. Ludwig, E., Wien. med. Jahrb., 1884. ; 6 Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. 7 Benedict, 8. R., and Hitchcock, E. H., J. Biol. Chem., 1915, xx, 619. § Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 459. — ~—— - J. L. Morris and A. G. Macleod 57 open to criticism on this point for the reason that the entire pre- cipitate is dissolved in sodium eyanide previous to the develop- ment of color. The extraction step of their blood procedure to a large extent eliminates this error. Myers’? suggestion in regard to this method that cyanide be added before the second centri- fuging can hardly be accepted as an improvement when con- sidered in the light of these facts. We have found that the re- liability of these methods is greatly increased when care is taken to use only perfectly clear silver reagents. Another possible source of error in the Folin and Wu proce- dure is a common impurity found in many of the best grades of sodium sulfite we could obtain. Three out of four lots of this salt, from as many sources, gave a reaction with the uric acid reagents. The color which developed in each instance was suf- ficient to introduce a considerable error if the sulfite had been used in the amount prescribed for the determination. Neither silver nor sulfite is used in the new method described in this paper. The new method requires the use of no metal which may exist in a reduced form or reagent which is likely to contain any interfering impurity. Fortunately, in addition to their adaptability to the purpose, there is the additional advan- tage that all the reagents used are relatively inexpensive. The use of potassium cyanide suggested by Benedict and Hitchcock’ for tying up the silver greatly improved the earlier Folin-Denis‘ method. We successfully applied cyanide for the purpose of forming a double radical with zine and found it use- ful as well in two other particulars. Used in larger amounts its alkalinity is sufficient for the complete and rapid develop- ment of the color when the actual molecular concentration of alkali is still less than one-third that used in the Benedict-Hitch- cock procedure. The second advantage in the use of cyanide is the very marked increase of color which is obtained from a given amount of uric acid. That this increase is due to the uric acid reduction of phospho- tungstate, while the cyanide only accelerates the reaction, is shown by results of the kind typified in Table I. The negative result of Flask 10 shows that the cyanide will not act directly on the phosphotungstate. The 18 per cent increase of color in ®* Myers, V. C., J. Lab. and Clin. Med., 1919-20, v, 499. 58 Colorimetric Estimation of Urie Acid TABLE I. Effect of Sodium Cyanide upon Phosphotungstate Oxidation of Uric Acid. Flask No. Usioacid. | sogt eet cert at, |e eee Value found. mg ce. mm mg. 1 0.25 None. 30.0 1.00 X 0.25 2 0.25 0.05 Ape 1.18 XX: 0225 3 0.25 0.25 PATE 1420225 4 0.25 0.50 20.4 L.A. Xe0s25 5 0.25 1.00 19.2 1:56 X 0.25 6 0.25 2.00 ih ayeab 1.86 X 0.25 7 0.25 4.00 14.1 2.13) 0s 8 OR25 8.00 1225 2.40 X 0.25 9 0.25 12.00 1B! 2.46 X 0.25 10 None 10.00 No color None Color developed in each case by addition of 1 ce. of phosphotungstic acid reagent and 10 ec. of 20 per cent sodium carbonate. All made up to 50 ce. volume after 10 minutes and compared with No. 1. TABLE II. Effect of Sodium Cyanide upon Arsenotungstate Oxidation of Uric Acid. Flask No. Uric acid. soto nide. peer Value found. mg. ce. mm. mg. 1 0.4 10 148.0 0.054 2 0.4 15 52.0 0.156 3 0.4 WAY 34.4 0.236 4 0.4 PA 30.0 0.266 5 0.4 3.0 Palle 0.287 6 0.4 OL 25.9 0.309° Uf 0.4 4.0 24.7 0.324 8 0.4 4.5 23.5 0.340 9 0.4 55.(0) PAL oe 0 376 10 0.4 eo 20.2 0 396 Til: 0.4 10.0 20.0 , 0 400 MP2 0.4 15.0 20.0 0 400 13 None 10.0 No color. None. Color developed by addition of 2 cc. of arsenotungstic uric acid reagent and the amount of cyanide indicated for each flask. * No. 11 was used as a standard and the others compared with it. J. L. Morris and A. G. Macleod 59 Flask 2, due to 1 drop of cyanide, recalls an increase of the same amount which Benedict and Hitchcock observed under similar conditions (2 drops of cyanide and 15 cc. of carbonate in a 50 cc. flask). Flasks 8 and 9, with 50 per cent more cyanide in the latter, showed nearly a constant value with two and a half times as much color as Flask 1. Obviously the cyanide causes dis- placement of the equilibrium so as to approach the maximum amount of the blue compound for these conditions. Attempts to determine the full extent of this effect met with little success at first for the decrease of carbonate and increase of cyanide was required to accomplish further deepening of the color, and this change in the alkalies caused precipitation before the time for development of color had passed. Also we found that a certain amount of carbonate was necessary to prevent the blue color which otherwise develops when phosphotungstic acid reagent is made alkaline with sodium cyanide. We had previously made many conjugated tungstic acids, sub- stituting analogous acids for phosphoric, concerning which we expect to make a more extended report soon. Upon trial we found that one of these, arseno-18-tungstic acid, in addition to other desirable properties, gave absolutely no color with sodium cyanide even when no other alkali than the cyanide was present, and did not precipitate in the presence of large amounts of cyanide. By its use we made further observations of the effect of cyanide upon the oxidation of uric acid by tungsten compounds and se- cured the results recorded in Table II. It will be noted that the amount of uric acid used in each flask was 0.4 mg., for our ex- perience had shown that the color obtained from that quantity was of a desirable depth for colorimetric comparisons. It is also apparent that the color in Flasks 1 to 10 increased with additional quantities of cyanide, the increments being progres- sively smaller. In Flasks 11 and 12 there was practically no increment, though the quantities of cyanide were, respectively, 133 and 200 per cent of that in Flask 10. Evidently the condi- tions had been reached by which the oxidation of uric acid by a tungsten compound was complete. The determination of relations between the amount of color obtainable by the new arsenotungstate-cyanide method and the former phosphotungstate-carbonate procedures could not be abso- 60 Colorimetric Estimation of Urie Acid lute, owing to the great difference in the concentrations of the reaction liquids. Approximate comparisons show that the new uric acid method gives 3.3 times the color of the Folin-Macallum- Denis method, 2.8 times that of the Benedict-Hitchcock pro- cedure, and 2.5 times that of the Folin-Wu method. There is, of course, a great mechanical advantage in being able to get three times as much color from the very limited amount of uric acid in blood and other body fluids. In addition, there is greater satis- faction in the realization that the deeper color made possible by the new method represents a truer value of the uric acid present in that it is the result of a complete chemical action and not sub- ject to the disturbing variations which may occur in procedures which depend upon artificially maintained equilibria. Finally, the use of sodium cyanide in the manner described in this paper possesses another advantage in the selective applica- tion of its driving power. How exclusive this selective action of cyanide, whether there are other substances present in body fluids upon which it will act, and what the nature of the driving power may be are questions now being investigated further in this laboratory. Method. Reagents. Preparation of Arseno-18-tungstic Acid Solution. —Boil a mixture of 100 gm. of hydrated sodium tungstate (Naz WO,: 2H: O), 125 gm. of arsenic acid anhydride (As: O;), and 650 ec. of water for 2 to 4 hours in a flask. If the reagent so formed has a blue or green color after it has boiled the required time, it should be decolorized by boiling with sufficient bromine water to make the color a clear yellow or yellowish brown.!° After boiling off any excess bromine add distilled water to make the volume 1 liter. The arseno- tungstic acid reagent so prepared is a somewhat lighter color than the phos- photungstic acid reagent. Other Reagents Required.—2.5 per cent zine chloride solution; 10 per cent sodium carbonate solution (if monohydrated sodium carbonate is used, allowance must be made for the water of crystallization) ; 10 per cent hydro- chlorie acid solution; 10 per cent sodium cyanide solution; standard uric acid solution (phosphate solution of Benedict-Hitcheock).7 For Removal of Proteins..—10 per cent sodium tungstate; 3 N sulfuric acid, within 5 per cent by titration; solid potassium oxalate. 1° Decolorization in this way is desirable for any conjugated tungstic acid which is to be used for colorimetric work. A dark blue or green rea- gent (either phospho- or arsenotungstic acid) introduces a very noticeable error when used where the color to be read is light. J. L. Morris and A. G. Macleod 61 Determination. The method is essentially the same when used in uric acid solu- tions of such different concentration as urine and blood. Con- venient quantities of reagents and choice of volumetric flasks which facilitate colorimetric comparison are the principal points of difference in the procedures described. Procedure as Used in Urine.—Pipette 1 cc. of urine into a 50 ce. centri- fuge tube and dilute with distilled water to about 40 cc. Add 1 ce. of 2.5 per cent zine chloride and mix with a stirring rod. Add1.0cc. of 10 per cent sodium carbonate which should make the solution alkaline to litmus and stir thoroughly. Centrifuge for about 2 minutes, drain off, and discard the supernatant liquid. Dissolve the residue, with stirring, in 3 or 4 drops of 10 per cent hydrochloric acid, dilute with 5 ec. of water, add 10 ce. of 10 per cent sodium cyanide, and transfer quantitatively to a 100 ec. volumetric flask, and dilute to about 60 cc. If 1 ec. of urine contains more than 0.5 mg. of uric acid the amount of cyanide should be doubled (20 cc.) and a 200 ce. flask used. In this case dilute to about 120 ce. To prepare a stand-, ard containing 0.2 mg. in 50 cc. pipette 1 cc. of the phosphate standard solu- tion into a 50 ec. volumetric flask and 25 to 30 ce. of distilled water and 5 ec. of 10 per cent sodium cyanide. Develop the color in both by addition of the arseno-18-tungstic acid reagent, 1 cc. to the standard (50 cc. flask), 2 ec. to the unknown if in 100 cc. flask or 4 cc. if in the 200 ce. flask. Shake, dilute to volume, let stand 2 or 3 minutes, and compare in the colorimeter. The color develops with such rapidity that the time interval indicated is sufficient if the standard and the unknown are made simultaneously. If, for any reason, they are not so prepared, it is best to allow 10 minutes to elapse before making the color comparison. Procedure as Used in Blood —Collect oxalated blood in the usual man- ner, drawing the blood from a vein into a weighed flask containing 2 mg. of potassium oxalate for each cubic centimeter of the sample taken. After determining the amount of blood by weight, pour it into seven times its volume of distilled water, add 1 volume of 10 per cent sodium tungs- tate solution and then, while shaking, run in slowly 1 volume of 3 N sulfuric acid. Shake for several minutes and filter (precipitation method of Folin and Wu‘). Pipette 25 cc. of the clear filtrate (corresponding to 2.5 ec. of blood) into a 50 ce. centrifuge tube and dilute with distilled water to about 40 cc. Add 1 ce. of 2.5 per cent zine chloride and mix with a stirring rod. Add 1.0cc. of 10 per cent sodium carbonate to make just alkaline to litmus and stir thoroughly. Centrifuge for about 2 minutes, drain off and discard the supernatant liquid. Dissolve the residue with stirring in 3 or 4 drops of 10 per cent hydrochloric acid, dilute with 5 ce. of water, and add 2.5 ec. of 10 per cent sodium cyanide and transfer quan- titatively to a 25 cc. volumetric flask. Prepare two standards containing 0.1 and 0.2 mg. in 50 ce., by pipetting 0.5 and 1 ec. of the phosphate stand- ard solution into two 50 cc. volumetric flasks. Add about 30 cc. of distilled 62 Colorimetric Estimation of Urie Acid water and 5 ce. of 10 per cent sodium cyanide to each. Develop the color by the addition of the arseno-18-tungstic acid reagent, 0.5 and 1 ec. respective- ly into the unknown and standards. If the color has been developed simu- taneously, shake, dilute to volume, let stand a minute or two, and compare in the colorimeter; if not, the same lapse of time should be allowed as indi- cated in the case of urine. Both procedures are adapted to the quantities of uric acid found in the largest number of urine and blood samples analyzed by us. In a very few cases we found it advantageous to choose volumetric flasks of a larger or smaller size to contain the un- known. This may be done with good results if the concentra- tions of arseno-18-tungstic acid and sodium cyanide are kept comparable. For this purpose the following simple rule must be observed: 100 cc. flask contains 10 ce. of 10 per cent sodium cyanide and 2 ce. of arseno-18-tungstic acid reagent; 50 ce. flask contains 5 cc. of 10 per cent sodium cyanide and1 ce. of arseno-18-tungstic acid reagent; 25 ce. flask contains 2.5 ec. of 10 per cent sodium cyanide and 0.5 ec. of arseno-18- tungstic acid reagent. : TABLE III. Comparative Estimations of Uric Acidin Urine. Urine. Benedict Hines Folin-Wu method. New method. mg. mg. mg. 1 390 375 388 2 752 625 702 3 514 492 533 4 547 498 508 Table III presents comparative uric acid results for several urine specimens. In their analyses we used the method described here, the Folin-Wu method, and the Benedict-Hitchcock proce- dure. Precaution to use only the clearest possible, silver reagents had the effect of practically eliminating irregularities due to reduced silver. In spite of similar precautions in work with blood specimens there were marked irregularities. Upon under- taking the work of their explanation we were led into various problems connected with the chemistry of the methods and the chemical nature of the uric acid present in blood. Some of the results obtained appear in the following paper." 1! Morris, J. L., and Macleod, A. G., J. Biol. Chem., 1922, 1, 65. J. L. Morris and A. G. Macleod 63 SUMMARY. Combination of zine precipitation with a new colorimetric method has made possible the estimation of very small quanti- ties of uric acid. Arseno-18-tungstate proves a great improve- ment over phospho-18-tungstate of earlier methods. Sodium cyanide is used as the only alkali for development of color and serves to bring about the complete oxidation of uric acid. In comparison with the amount of color obtained in the methods which depend upon oxidation to-a point of equilibrium, the new conditions of complete oxidation permit the development of three times as much color per unit weight of uric acid. The same conditions that bring about completion of the reaction are also responsible for greater speed in reaching the maximum color and in a very marked permanency of the color. The use of cyanide as alkali practically eliminates the precipitation of various com- pounds in the colored liquid which, next to fading, was the most serious difficulty accompanying the use of carbonate. Precipitation of uric acid with zinc salts lends itself just as well to the subsequent formation of a double radical with cyanide as in the case with silver methods. In addition there is excluded all possibility of a reduced metal giving erroneous Be in the later oxidation reaction. Finally, the number of reagents required is small, they are easily prepared and the use of inexpensive zinc chloride instead of ex- pensive silver salts makes the determination much more de- sirable, especially where many analyses are run, as in medical classes and extended research work on purines. STUDIES ON THE URIC ACID OF HUMAN BLOOD. By J. LUCIEN MORRIS anp A. GARRARD MACLEOD. (From the Biochemistry Laboratory of the School of Medicine, Western Reserve University, Cleveland.) (Received for publication, October 27, 1921.) Uric acid estimations have probably been subject to more irregu- larities and losses than attend the analysis of other well known biological products. Explanations of these irregularities have usually been based upon its very slight solubility, the unusual ease and variety of its oxidation reactions or the more vague property of different forms in which it has been supposed to exist in body fluids. The new method of uric acid analysis described by the authors in the foregoing paper,! while characterized by extraordinary agreement between successive determinations in urine and blood, differed in quite an irregular manner when its values for blood were compared with those obtained by the method of Folin and Wu.2 The present paper sets forth some of the results of our efforts to find an explanation for the apparent dis- erepancies. In addition to evidence which casts considerable doubt upon the accuracy of uric acid data obtained through the use of earlier methods, we have made observations which can be explained only by the existence of uric acid in more than one form in human blood. Differences in form of uric acid, or more correctly urates, have been used as a basis for several hypotheses concerned with the physiology and pathology of purine metabolism. One of the more definite ideas of these differences is that which pictures uric acid present in various stages of change according to the effect of blood conditions upon its property of keto-enol isomerism. Gudzent first discovered that uric acid forms two series of primary 1 Morris, J. L., and Macleod, A. G., J. Biol. Chem., 1922, 1, 55. 2 Folin, O. and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. 3’ Gudzent, G., Z. physiol. Chem., 1909, Ix, 38. 65 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 66 Uric Acid in Blood urates, differing in stability and solubility. They were named the lactam and lactim forms. Shifting the point of equilibrium between these tautomeric forms might conceivably take place under changing blood conditions. The resulting change of point of saturation has been urged as an explanation of many clinical symptoms, such as the deposition of urates in cartilages of gout patients, etc. Another idea of the difference in form of uric acid is that some part of the whole amount is combined with another substance or substances. Minkowski? advanced the hypothesis that uric acid may be combined with nucleic acid in the body tissues and suggested that this combined uric acid might regulate the chemical relationships of free uric acid. The basis for this conclusion was the observation that the precipitation of a uric acid solution by acetic acid or ammoniacal silver magnesium solution is pre- vented by the addition of nucleic acid. No experimental evi- dence has been advanced that combined uric acid exists in any part of the human body. Benedict® several years ago found com- bined uric acid in‘mixed beef blood, in quantity many times greater than the free uric acidin the same blood. A year later Bene- dict® made the following statement, ‘‘ with the exception of man, all mammals probably have two forms of uric acid in the blood. In the case of human blood the data so far available are not conclusive. It is quite probable that here, too, uric acid exists in the blood in at least two forms but they are quite unlike the forms present in ox blood.’”? Davis and Benedict’ recently re- ported the isolation of a crystalline substance from beef blood which they identified as a ribose-uric acid compound. The data presented later in this paper furnish the first positive experi- mental proof that human blood contains uric acid in at least two forms. We analyzed many samples of blood and serum by the Folin-Wu method and by the new method in order to compare the values 4 Minkowski, O., Die Gicht, in Nothnagel, H., Specielle Pathologie und Therapie, Vienna, 1903, vii, pt. 2, 189-190. 5 Benedict, S. R., J. Biol. Chem., 1915, xx, 633. 6 Benedict, S. R., J. Lab. and Clin. Med., 1916-17, ii, 1. 7 Davis, A. R., and Benedict, S. R., J. Biol. Chem., 1921, xlvi, p. v. J. L. Morris and A. G. Macleod 67 so obtained. In all bloods the Folin-Wu method of protein precipitation was used, the filtrate in each case serving as acommon source of the samples taken for both methods of analysis. Recog- nizing the danger of abnormal results in the Folin-Wu method due to the possible presence of ‘‘reduced silver” every pre- caution was taken to assure the use of clear silver lactate solution. TABLE I. Uric Acid Estimation in Blood. Comparison of Methods. : Folin-Wu New Specimen. method. method. Group 1. Examples of blood and serum which show marked irregularities. mg. mg. = Riri s pa Pe. ities Le he as 119 Somers wet sushi aoe 4.1 4.9 20S AS SC Se A, ee ene ee 2 3.4 ER See Se Ae er ee 1.8 Zo EMT eR Sn. RRB ic ciate Me nc pia aw kwh nm 4.4 Sas oS wl) Leo ep 9-2 ile Sinn ale Sg ibs Aa erred, Paae nin 9) 5h) Eee S30. 2. STs es | 4 22% arene dette ete Rs nr coset eos wins. clapsaiael Boa ae Bed 3.4 OT Ca RRR” en ta a ee Dd. 4.8 PS ELD SSL Se Cea re 1.9 3.0 MOM EAES CUUITIA AC for. os, 2 53:2) ach PME ere oF aie % SieLd sage = 1.4 Zul 7 OD gh RRR ht 2 peat, CAML Ee eta s 1.6 2.5 Group 2. Examples of blood which show little or no irregularity. Blood 1, G. Mo... cee cece t ttt Pe 7.9 ay UE cet ta dion: eee 3.5 3.2 © ESTES (uBR th rh Coe I 3.4 3.4 ne ee ss hee 3 1.6 eee ay te eet Ove an 3.2 3.2 Semememanra Sarno fi. 6c .. 20, Se. 2i3 2 RUE: (233 cbs] io). oc \).w id aod ies sabes re 2.1 In addition, suspicious variations between duplicates, were immediately followed by check determinations. In this way the Folin-Wu values which are recorded in this paper represent the highest accuracy which can be obtained by the method. Even after these unusual precautions there were puzzling irregularities in the agreement of the two series of results. Typical examples of the comparative blood analyses are presented in Table I. 68 Urie Acid in Blood It is apparent from the figures tabulated that the new method gives values for blood which range from those which are essen- tially the same (Group 2, Bloods 1 to 7) to those (Group 1, Bloods 1 to 9) which are higher than the corresponding Folin-Wu values. The largest increase in value shown in the bloods is 75 per cent. In the case of the mixed serums, A and B, the results are similar but the amounts of increase range from 60 to 120 per cent. This high range of increase is characteristic of all mixed serums we have had under observation. We made repeated attempts to lower the results of the new method to the level of the Folin-Wu method. We could neither accomplish such a decrease nor find evidence of the presence of any substance which gave an added color value. The specific nature of the new cyanide-arsenotung- stic reaction (in its driving effect upon uric acid to the exclusion of any other substance so far investigated) was good evidence that the higher value is due to a more complete measurement of the uric acid present. Attempts were then made to increase the Folin-Wu values. None of these was successful until we added potassium oxalate. Though smaller amounts of oxalate have the effect of elevating the value obtained with the Folin-Wu method, we found it desirable to use an excess sufficient for all possible variations in blood specimens. Therefore, we added 100 mg. of potassium oxalate (measured as 10 cc. of a 1 per cent solution) to each 20 ec. quantity of filtrate before beginning -the determination. The presence of the oxalate increased the results of the determination as shown in Table IT. The recorded results obviously fall again into two groups. The values of the Folin-Wu determinations are markedly increased in the first group of seven blood samples and two mixed serums while there is no increase or a very slight increase in the second group of six bloods. Evidently the value found by the new method determines not only the question of whether an increase will result from the use of potassium oxalate but constitutes the approximate limit of the increase when present. With minor differences of the order of variation between duplicates, the figures secured by means of the modified Folin-Wu method (potassium oxalate preceding the Folin-Wu procedure) are the same as those obtained by the new method. This observation is peculiarly significant when considered in connection with the J. L. Morris and A. G. Macleod 69 fact that the presence of oxalate had little or no effect upon the analysis of those blood specimens (entered as Nos. 1 to 6 of Group 2, Table II) for which both methods gave essentially the same values. The explanation of the irregularities was thus shown to be less a matter of method weakness and more a matter of character of the content of individual bloods. TABLE II. Effect of Potassium Oxalate upon Blood Uric Acid Values Obtained by the Folin-Wu Method. ee . Folin-Wu metho New Specimen. after P method. addition of method. K2C204. Group 1. Examples of blood and serum which show a marked increase, mg. mg. mg. istibavave | Ik Chal 313 [ne eae rte eet he th nee aie 4.1 4 4.9 RS MES oe As rs 0 20g 51d SE STE RE Dae Al, ah haere 2G 3.5 3.4 Rey ACN ee Fe a 5.2 genes Sa, Arete 1.8 Die 4 335 SRI OE INA I cc's do Hecate fs kev 0) ods 4.4 ii Al ae MME SSO AT acy a ss wos s EBT Dee he ete ee Daft DES “T(J 1 Dec che Soca ye eZ, ZED, IPT 2 “7G OS 1831 0 Ie Pedy te gee ee Ree 2 Bee 34 IN Disayel See haa ae Se 1.4 4.3 oe ee TEN ol Bae RE he. A a 1.6 33) PS Blood 1, G. Mu... cece cece rae 8.0 7.9 7 Mog «SU en 3.2 3.3 3.2 ie A eo ES a 3.4 327 3.4 AE ORAM: “oC, (a a 1.8 1.6 1.6 Seer Pehle meee eee ae 3.2 3.6 3.2 nS TE Te a ee re 2.4 2.3 2.2 We undertook an investigation of different blood specimens in the hope of finding what chemical difference exists that deter- mines for each specimen whether the values given by the silver and zinc methods are to agree or disagree and a corresponding agreement or disagreement between the values by the former method in its original form and as modified by the addition of oxalate. The very limited quantity of blood in each case hindered 70 Uric Acid in Blood ° progress of the investigation as did also the impossibility of judging before analysis whether each new specimen would show agreement or disagreement between the methods. The use of other than human blood was inadvisable in view of the radical difference in form and quantity recognized by Benedict as charac- terizing the blood uric acid of different species. Analysis of mixed serums (from many blood specimens drawn for routine Wassermann tests and found negative) consistently showed the new method value higher than the Folin-Wu value. Further- more, the percentage increase was about twice as great as in the case of whole blood specimens. Evidently serum was by nature and quantity availability the best material for use in identifying the character of the substance responsible for the divergent uric acid values. Serum was saved over a period of weeks. The proteins were precipitated from each day’s quantity of serum by the tungstic acid method. The filtrates of successive days were poured together until there was a volume of 5 liters. Several such lots were investigated after each was analyzed by the three procedures. The two sets of results presented in the foregoing tables are typical analyses of mixed serum. As a result of many observations on the mixed serums we were convinced that we were dealing with more than one form of uric acid. We attempted the separation of uric acid as such for evidence supporting the analytical data. 1,000 ec. of mixed serum were precipitated in 50 ec. portions by the zine procedure. After centrifuging and pouring off the mother liquid the com- bined precipitates were dissolved in 10 per cent acetic acid, then 100 ec. of water and about 0.5 gm. of bismuth carbonate added. Hydrogen sulfide was bubbled through the solution until it was saturated; it was then heated to the boiling point and filtered. The filtrate was evaporated to small quantity and then heated to dryness on the water bath. After dissolving in 25 ce. of hot water and transferring to a 50 ec. centrifuge tube, the uric acid was again precipitated by the addition of 4 cc. of the ammoni- acal silver magnesium reagent of Benedict and Hitchcock and separated by centrifuging.’ To the precipitate were added 5 ce. 8 The second precipitation (as a silver compound) and subsequent sepa- ration of uric acid as such substantially follows the method used by Benedict in identifying uric acid in beef blood (Benedict, 8. R., J. Biol. Chem., 1915, xx, 637). J. L. Morris and A. G. Macleod 71 of water and 5 cc. of saturated hydrogen sulfide solution (freshly prepared). After mixing thoroughly the silver sulfide was re- . moved by filtration and the hydrogen sulfide from the filtrate by evaporation to dryness on a water bath. The residue was taken up in successive small portions of boiling water and transferred to a weighed 15 cc. centrifuge tube. The volume was then about 5 ec. lee. of glacial acetic acid was added and the tube set aside. 5 days later the mother liquor was poured off and the characteris- tic crystalline precipitate of uric acid washed first with water and then with alcohol. After drying, the tube was weighed. In addition to the weight.and crystalline form further evidence of the nature of the precipitate was obtained by colorimetric analysis. The characteristic acid precipitation was checked by the data so secured. ‘The results are presented in Table III and with them are recorded, for comparison, the values obtained in the similar treatment of a standard uric acid solution, which contained 3 mg. and was diluted to a liter before precipitation. Reference to the table shows in the case of each serum that, after the laborious process of removing uric acid from solution twice by formation of two different salts, the amount found present by analysis of the final crystalline product is greater than the Folin-Wu method originally indicated. (Serum A shows 2.2 mg. against 1.4 mg. and Serum B shows 2.0 mg. against 1.6 mg.) On the other hand, the amount of uric acid which could be simi- larly recovered from the standard solution is much less than was originally present (1.2 mg. from 3.0 mg.). The latter yield is as large as might be expected after the tedious steps of the double precipitation, the subsequent removal of the metallic sulfides, filtrations, evaporations, etc. If the loss in the case of the serum filtrates was actually comparable to that of the uric acid solu- tion, and we should not expect it to be otherwise, the original amount of uric acid in the mixed serums must have been about 4 mg. Analysis by the new method and the Folin-Wu method after the addition of oxalate gave values of 3.2 and 4.3 mg. for Serum A, 3.2 and 3.5 mg. for Serum B. Similar treatment of other mixed serums resulted in a yield of separated uric acid which invariably exceeded the Folin-Wu figure. Repeated attempts to obtain larger amounts of uric acid from standard solutions containing 3 mg. in a liter, never resulted in a yield higher than 50 per cent. 72 Uric Acid in Blood It cannot be supposed that the significance of these facts is only a demonstration that the Folin-Wu method gives low results. Such bloods as those of Group 2, Tables I and II, disprove such an idea. There the range of uric acid amount is from 1.6 to 7.9 mg. and the two methods agree. The explanation can only be that TABLE III. Identification of Uric Acid Removed as Such from Blood Serum and Uric Acid Solution. Uric acid * 4 * . Weight of Col t lysis of s content of | ~~ Seid separated.” | aaa mg. mg. mg. Holin=Wiesscao.se se 1.4 Twice washed enrystals.n.... 2. 2.2 Folin-Wu follow- Mother liquor. ....0.0 3.1 Serum A. : ing oxalate..... 4.3 New method...... 3 Wash water and aleono]-aees eee 0.02 Folin-Wu......... 1.6 Twice washed Crystals.s bee ses 2.0 Folin-Wu follow- Mother liquor. ....0.38 per Serum B. : ing oxalate...... 3.5 New method...... 30 Wash water and CHCOHIKO) Hard Guan er 0.07 Twice washed crystalstescten te. 12 Uric acid standard........... 3.0 Mother liquor. ....0.23 2.2 Wash water and alcohol anes 0.02 uric acid is present in the mixed serums and some bloods (Group 1, Tables I and II) in more than one form. In other bloods (Group 2, Tables I and II) there is but one form, or traces only of the second form. In consideration of the fact that the Folin-Wu method and the new method give quite comparable results when J. L. Morris and A. G. Macleod a applied to standard urie acid solutions, it must be the second form of uric acid which the Folin-Wu method fails to include while the new method includes it. That it is some form of uric acid rather than any other substance which reacts colorimetrically follows of necessity from the facts here presented that: (a) it carries successively through the precipitations with zinc salt and silver magnesium mixture, which are chemically different but equally characteristic; (b) it then precipitates quantitatively upon acidification of its solution in the form of crystals which cannot be differentiated from those of uric acid; (c) it is changed quantitatively at room temperature in contact with potassium oxalate to a form readily precipitated and extracted by the usual Folin-Wu procedure; and (d) the new method gives a value for this second form, as well as the first, in spite of the ex- clusion of all substances so far tried from the multiplying effect of the cyanide upon the color. Further work to determine the chemical nature of the second form of uric acid is now under way in this laboratory. As observed above, there is apparently a greater relative amount of the second form of uric acid in serum than in whole blood. We secured freshly drawn samples of blood of sufficiently large volume to allow three sets of analyses by the original Folin- Wu method, Folin-Wu following oxalate method, and the new method. Three determinations were run upon a filtrate from the whole blood, three more upon a filtrate from a serum portion, and another three upon a filtrate from a corpuscle portion. Ap- proximate separation of the bloods was effected by means of the centrifuge. While analysis of the serum portion and corpuscle portion does not furnish strictly quantitative data on the uric acid content of either serum or corpuscles uncontaminated by the presence of small amounts of the other, nevertheless the results unmistakably indicate the order of uric acid distribution between the corpuscles and serum. The figures of Table IV, which repre- sent that distribution in a typical blood specimen would, by more complete separation of the formed elements, be changed so as to further emphasize the fact that the uric acid content of serum is from one and a half to nearly twice that of the corpuscles. Values by the Folin-Wu following oxalate method are 5.7 mg. against 3.0 mg. and the corresponding new method figures are 4.1 mg. a — 74 Urie Acid in Blood against 2.3 mg. The results obtained when using the original Folin-Wu method are to be considered very approximate since the color due to the small amount of uric acid present was too small for accurate estimation. The second form uric acid, apparent in the table as the difference between the new method and the Folin-Wu method values, is present in the serum portion to an extent three times as great as the first form, 3.2 mg. against 0.9 mg. and in the corpuscle portion is twice as great, 1.5 mg. against 0.8 mg. Such direct observations of the relatively smaller second form uric acid content of corpuscles substantiate the relatively large second form uric acid content of serum previously mentioned. Also we have interpreted these observations, that the added uric acid value is unevenly distributed between cor- TABLE IV. Distribution of Uric Acid in Blood. ; Folin-Wu method. | following | New method. K2C201. Whole blood 75.22.58 -2iss eee 2.2 3.9 3 five SELUMPPOLbLOMI,. 2.6, 2)f: ay. eee 0.9 0.4 | 5.7 3.0 | 4.1 2.2 Corpusele portion, :.:.;../:2.sseapere 0.8 0.3 | 3.0 1.4} 2.3 1.3 The figures recorded in the first column for each method represent quantities in 100 cc. of serum or corpuscles. The figures in the second column represent quantities in 100 cc. of blood. puscles and serum, as further evidence supporting the existence of the second form of blood uric acid. Finally, it should be noted that the distribution of uric acid just described for human blood is in marked contrast with that observed by Benedict® for mixed ox blood. In the latter Benedict found all the uric acid (free and combined) in the corpuscles, none in the serum. In the former we find the uric acid in both, but in much greater quantities in the serum. This different distribution suggests that the second form of uric acid in human blood is probably different from the “com- bined” uric acid of ox blood. Whether this is the case is only one of the many important questions, which we hope may be attacked by means of the new zine precipitation-arsenotungstate cyanide J. L. Morris and A. G. Macleod 75 method. We are already studying some of these problems and expect to investigate others as rapidly as new facts concerning the forms of uric acid can be ascertained. Until there is thus de- veloped a more complete understanding of blood uric acid of different species it probably is not desirable to theorize on the physiological and pathological significance of the second form of uric acid in human blood. Our thanks are due to Dr. H. W. Gauchat of the Cleveland City Hospital for enthusiastic cooperation in securing suitable blood specimens for the particular requirements of this investiga- tion. We gratefully acknowledge the technical assistance ren- dered by Mr. H. W. Hottenstein. We are also indebted to Dr, E. E. Ecker for making available large quantities of serum from the Wassermann laboratories of Lakeside Hospital and the Cleveland Board of Health. mm rid SERRE) / y F - F aE Sean wih cAly ue a r y + <9 ’ We mta t a iy ; oe eee Pe . Me ys Mab) Sa of + ae ’ lex | . § y 2 ™~ 30 ba Ree ie ; * : : ; . ‘ 7 » “ = , A 2 F re A P 4 hae . ‘ e : ral . i { | ‘ ; ‘ “i a EXPERIMENTAL RICKETS IN RATS. III. THE PREVENTION OF RICKETS IN RATS BY EXPOSURE TO SUNLIGHT.* By ALFRED F. HESS, L. J. UNGER, anp A. M. PAPPENHEIMER. (From the Department of Pathology, College of Physicians and Surgeons, Columbia University, New York City.) PLATES 2 AND 8. (Received for publication, October 21, 1921.) In recent papers it was shown by Hess and Unger that rickets in infants could be cured by frequent short exposures to the sun’s rays (1,2). By this means and without any alteration what- soever of the dietary, the characteristic signs of this disorder markedly diminish in 3 to 4 weeks, as noted by clinical examina- tion and by the x-ray. As a result of favorable experiences of this nature it was concluded in a study of ‘‘the seasonal incidence of rickets’ (3) that ‘‘hygienic factors, especially sunlight, and not dietetic factors, play the dominant réle in the marked seasonal variations of this disorder.’’ It seems probable that the ultra- violet rays play a large part in this curative power of the sun, judging from the work of Huldschinsky (4) and others (5, 6, 7) who recently have shown that infantile rickets can be cured by means of the rays produced by the mercury-vapor lamp. In 1918 we tried the curative effect of rays from this source, but, lacking the aid of x-ray examinations, could not convince ourselves of their efficacy; since then we have succeeded in curing rickets by this means. Having found sunlight efficacious in the rickets of infants, we proceeded to test its value in the prevention of rickets in rats. To this end a series of white rats was placed on the diet (No. 84) described by Sherman and Pappenheimer (8) consisting of patent flour 95.0 per cent, calcium lactate 2.97 per cent, sodium * Read in abstract before the Society of Experimental Biology and Medicine, October 19, 1921. (ei 78 Experimental Rickets in Rats. III chloride 2.0 per cent, and ferric citrate 0.1 per cent. It has been the experience of the investigators in this laboratory that such a diet invariably leads to the development in rats of lesions which are anatomically identical with those of infantile rickets. In carrying out experiments on rats our practice had been to keep the colony in a semidark room, the yellow shades being drawn at all times. In testing the effect of sunlight, the rats (weighing at the outset about 40 gm.) were kept in absolute darkness, one series being taken out of the room and exposed to the direct sunlight for a period of 15 or 30 minutes. There was no difference whatsoever in the diets of these two groups. After a period of about 3 weeks the animals were radiographed in order to observe early lesions of the epiphysis, and after 30 to 40 days were killed and autopsied. These experiments were begun in April, when the weather permitted four to five exposures a week. It was found for the first time in our experience that Diet 84, the ‘‘rachitic dietary,’ did not lead to rickets—that the rats which received sun treatment did not show signs of rickets either by x-ray or by histological examination of the bones. It is un- necessary to discuss in detail the histological criteria which we consider characteristic of rickets, as this question has been fully considered in a previous paper (9). It may be stated briefly that they consist of increased width and irregularity of the prolifera- tive cartilage, absence of calcium deposition, and great excess of osteoid in the region of the metaphysis and along the shafts of the bones. It will be seen from Figs. 1 and 3 that the rats which were kept at all times in the dark showed these lesions, whereas the bones of those exposed to the sun did not show them (Figs. 2 and 4). In the paper previously referred to it was shown that the intro- duction of 0.4 per cent of secondary potassium phosphate (IK,HPOs,) in place of an equal weight (replacing about one-seventh of the calcium lactate contained in the rickets-producing diet) completely prevented the development of rachitic lesions; this constitutes an addition of 75 mg. of phosphorus per 100 gm. of the diet. In order to test the counterbalancing effect of phos- phate and darkness, a series of tests was carried out in the dark with additions of small and increasing amounts of potassium Hess, Unger, and Pappenheimer 79 TABLE I, 3 Alan Mi i Diet fe eal nr. «| Berean } Darkness. days No. 84 | 34 | 246 | Rickets. | Rickets. (Gr Tad a SS nia aA eas eee 23) |, 247 = 22 | 248 a Bette = eel lant cer vate | 30 | 438 i Rickets. IONS DOING Po esckcecldact abies 39 262 ae f 39 263 “f a 39 | 264 és ‘s 28 443 “ 28 | 444 of Rickets (slight). 28 445 re OES TOs 38 | 121 Negative. | Negative. 38 122 3 ef 38 123 es a, Sunlight. No. 84 34 | 249 | Negative. | Negative. (S13) TO aay Bie Nee eee ane 32) 250 - o 35 251 oe < 33 | 439 x f 33 440 “ “ : 33 441 “ “ce [ 33 4492 ‘“ “ Mo: 4 + 26 me. Po o.oo. ee eeee 39 | 259 « ee \ 39 260 c os 39 261 e < INOS EDS Sly ayh 10 ied Eee eee Seeder e 38 | 124 iM 3 38 125 3 = 80 Experimental Rickets in Rats. III phosphate to the standard dietary (No. 84); to one series 25 mg. were added, to another 75 mg. (constituting Dietary 85). The rats on these diets were kept in the dark but, to serve as control, half of each series was exposed to sunlight for 30 minutes daily when this was possible. As was to be expected in view of our previous experience and the fact that phosphate tends to pro- tect against rickets, none of the rats which were treated with sun- light developed rachitic lesions. Among the group, however, which was kept at all times in the dark, active rickets developed in spite of an addition of 25 mg. of phosphorus. The addition of 75 mg. was found to be sufficient to prevent the development of this disorder. This amount constituted the minimum pro- tective supplement to Diet 84, which in itself contains about 86 mg. of phosphorus. Thus it will be noted that a short exposure to sunlight was equivalent to almost doubling the protective dose of phosphate. If the phosphate content of the diet is adequate, rats do not develop rickets in spite of being kept in the dark throughout the experiment. The effect of sunlight with other dietaries was also studied, and is being continued. Without entering at this time into a detailed discussion of their influence, it may be of interest to record the observation that in one series of animals where 10 per cent of egg albumin was substituted for an equivalent amount of flour, rickets developed in some of the rats in spite of the sunlight treat- ment; whether this is to be attributed to a reduction of phos- phate incidental to diminishing the percentage of flour, or to the injurious effect of the egg albumin itself, will be determined by experiments which are in progress. Possibly the increased rate of growth of these animals with the accompanying increased phosphorus requirement may have been a factor of moment. DISCUSSION. As sunlight has a marked effect on the bony development of rats, it is evident that in future in similar nutritional investiga- tions, the light factor will have to be controlled and standardized. It seems probable that some of the irregularities and lack of con- formity observed by investigators in this ‘field may be attributed to keeping the experimental animals under dissimilar intensities Hess, Unger, and Pappenheimer 81 of light. The most interesting aspect of the question, however, is the phenomenon that the sun’s rays are able to stimulate a deposition of inorganic salts where these are lacking. The dam- aging effect of darkness emphasizes the fact that sunlight is of great importance, not merely for the vegetable world but also for the higher animals. Furthermore, the fact that sunlight is efficacious in the rickets of both human beings and rats, serves to show the similarity of this disorder in these two species. These results indicate that in the prevention and causation of rickets at least one hygienic factor plays an important rdle which will have to be carefully considered in future studies of this disorder. CONCLUSIONS. Rachitic lesions which develop regularly in rats upon a diet adequate in calcium but low in phosphorus, may be prevented by short exposures to direct sunlight. This protection is equivalent to the addition of at least 75 mg. of phosphorus to the diet in the form of basic potassium phosphate. BIBLIOGRAPHY. . Hess, A. F., and Unger, L. J., Proc. Soc. Exp. Biol. and Med., 1920-21, XV, 298. . Hess, A. F., and Unger, L. J., J. Am. Med. Assn., 1921, Ixxvii, 39. . Hess, A. F., and Unger, L. J., Am. J. Dis. Child., 1921, xxii, 186. . Huldschinsky, K., Z. orthop. Chir., 1920, 1xxxix, 426. Putzig, H., Therap. Halbmonatsch., 1920, xxxiv, 234. Riedel, G., Miinch. med. Woch., 1920, xxix, 838. . Erlacher, P., Wien. klin. Woch., 1921, xxxiv, 241. . Sherman, H. C., and Pappenheimer, A. M., J. Exp. Med., 1921, xxxiv, 189. . Hess, A. F., McCann, G. F., and Pappenheimer, A. M., J. Biol. Chem., 1921, xlvii, 395. — 82 Experimental Rickets in Rats. III EXPLANATION OF PLATES. PLATE 2. Fig. 1. Rat 246. 34 days on Diet 84. Darkness. Chondrocostal junc- tion showing advanced rickets. (Silver nitrate—Van Gieson stain.) Fic. 2. Rat 249. Same litter as Rat 246. 34 days on Diet 84. Sunlight. Chondrocostal junction showing no rickets. (Silver nitrate—Van Gieson stain.) PLATE 3. Fig. 3. Rat 263. 39 days on Diet 84 plus 25 mg. of P added as K2HPQO,. Darkness. Radiograph showing rachitiec changes at knee-joint. Fic. 4. Rat 261. Same litter as Rat 263. 39 days on identical diet. Sunlight. No rickets. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L. PLATE 2. ies: Fra. 2. (Hess, Unger, and Pappenheimer: Experimental rickets in rats. III.) THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L. PLATE 3 (Hess, Unger, and Pappenheimer: Experimental rickets in rats. II1.) SOME HUMAN DIGESTION EXPERIMENTS WITH RAW WHITE OF EGG. By MARY SWARTZ ROSE anp GRACE MacLEOD. (From the Department of Nutrition, Teachers College, Columbia University, New York.) (Received for publication, November 16, 1921.) The behavior of raw white of egg in the alimentary tract of the dog has been studied extensively by Bateman! but there does not appear to be in the literature any corresponding investigation in regard to the digestibility of white of egg in this state by the human subject. Falta? reports two experiments in which dried egg albumin (presumably raw) was added to a basal ration, and the amount of nitrogen in the feces determined. TABLE I. Per cent of intake Total N in feces. lostanikeccnt N intake With egg Per cent of | Experiment ay on basal Aa RAL N from egg ration. albumin. | Oy pasal | With egg | On basal | With egg ration. added. ration. added. gm gm. gm gm i! 16.40 27 .68 40.0 1.22 1.81 7.0 6.5 2 19.20 36.12 47.0 6.05 6.25 31.0 17.3 From these figures it appears that the addition of considerable amounts of dried egg albumin (80 gm. in the first experiment and 120 gm. in the second) instead of depressing the coefficient of digestibility actually raised it when the larger amount of egg was taken, the coefficient on the basal ration in the second experiment being 69 per cent and on the egg ration 83 per cent. Falta followed the nitrogen elimination in the urine and found that the highest point was reached somewhat later for egg albumin 1 Bateman, W.G., J. Biol. Chem., 1916, xxvi, 263. 2 Falta, W., Deutsch. Arch. klin. Med., 1906, xviii, 517. 83 84 Digestion of Raw White of Egg than for gelatin or casein. It seems clear that native egg white offers some resistance to the speedy action of the digestive enzymes, but as Bayliss* has shown, though trypsin acts more quickly at first on cooked egg albumin, it will if sufficient time be allowed digest the uncooked as completely as the cooked. Wolf and Osterberg? studying primarily the urinary nitrogen and sulfur on diets in which various protein foods were in turn added to a simple mixed ration, determined nitrogen in food and feces and found in one case with a total intake of 23 gm. of nitrogen 70 per cent of which was derived from raw egg white, the loss of nitrogen in the feces was 41 per cent of the intake; but in another period, with total intake of 14 gm. and 51 per cent of the nitrogen from the egg white, the loss was only 15 per cent of the total intake or about the same as on the basal ration alone. It would scarcely seem fair to draw conclusions from these two conflicting experiments. The authors have accordingly conducted experiments on ten ~ subjects,® all healthy young women, who took daily from ten to twelve whites of eggs as a part of a simple mixed diet, first cooked, in a 3 day period, then raw for the same length of time. The diet was uniform throughout the experiment and furnished 67 gm. of protein, to which the egg whites contributed 48 gm. or 70 per cent of the total. The experiments were divided into three groups, one in which the raw egg whites were taken thoroughly beaten, one in which they were taken in their natural state, and a third in which half were beaten and half unbeaten. In no case was there any sign of indigestion, such as discomfort or diarrhea, though one or two subjects found them slightly laxative. The cooked eggs were never subjected to a temperature or method of cooking (such as frying) which would render them tough or other- wise interfere with ease of digestion. Coefficients of digestibility have been calculated for the total protein of the diet, which seems the fairest way to judge experi- ments of this sort; and also in the conventional way for the egg 5 Bayliss, W. M., The nature of enzyme action, London, 1908, 148. 4 Wolf, C.G.L., and Osterberg, E., Biochem. Z., 1912, xl, 234. 5 Some of these experiments have been reported in a preliminary paper (Rose, M.8., and MacLeod, G., Proc. Soc. Exp. Biol. and Med., 1919-20, Xvil, 119). M. S. Rose and G. MacLeod 85 protein alone, making average allowance for the loss of the pro- tein of the other foods in the diet. The cooked eggs were uniformly well digested, coefficients ranging from 83 to 91 per cent with an average of 86 per cent for the diet as a whole; or from 82 to 93 per cent with an average of 86 per cent for the egg white alone. On the whole the raw whites were well utilized, the average difference between the cooked and raw being only 4 per cent for the protein of the whole ration or 5.5 per cent for the egg white protein alone, in favor of the cooked. The differences between the cooked and the raw whites varied with the mode of preparation, those beaten light being the best utilized, and those taken in the natural state least well absorbed, as shown by Table II. ' TABLE II. Difference between Coefficients of Digestibility of Egg Whites in Favor of Cooked Whites. Per cent of difference. Group No. Mode of preparation. SS For whole |For egg protein ration. alone I Unbeaten. +6.8 +9.6 II Half beaten and half unbeaten. +3.3 +4.6 Ill Beaten light. 1.7 +3.0 EXPERIMENTAL. The daily ration consisted of the same foods in all ten cases, but the proportions of the individual foods differed slightly, as indicated below. Analyses for total nitrogen were made in the laboratory and protein calculated as N X 6.25. The feces were marked off by carmine, and analyzed for total nitrogen in 3 day periods. The output of each subject is given in Table IV. The coefficients of digestibility calculated from the foregoing are given in Table VI. In estimating the coefficients for the egg white alone, the following arbitrary allowances have been made for loss in digestion of the protein of the other items in the diet. 86 Digestion of Raw White of Egg TABLE III. Daily Intake of Food. For Subjects G. S., M. K., L. For Subjects G. B., E. B., S.,M.F., M MEE, ek; ’ . ” ” Food materials. Weight of food. Protein. Weight of food. Protein. Heo whites eee oles: 372 47 .66 375 47 .63 Rae ab enero on ae aaa 85 6.80 85 8.75 reams: Beer ak 123 Dale, 79 1.98 SHUG ab ga s0 asec anion 28 3.05 35 4.68 Merce eae etecac ae 85 1.49 85 1.00 Rant; JUICes cee 600 5.06 600 3.07 IBUthere crise oases 39 0.39 39 0.39 Olivievorlkynauce ones 33 33 SIE akeon Daeee eB ee 50 50 Elnontisil es See ete! es ote 67.17 67.47 TABLE IV. Daily Output of Nitrogen in Feces Calculated to Protein. Nitrogen in feces X 6.25. Group No. Subject. | = eee Raw egg Cooked diet. egg diet. gm. gm. I (Raw whites, unbeaten). ME: 11.24 16.35 E. B. 10.00 8.13 M. E. 15.25 8.438 II (Raw whites, half beaten, halfunbeaten). | G. s. 9.59 11.70 L. 5. 9,29 11.78 . M. R. 8.19 11.56 RRs 11.26 12°22 III (Raw whites, beaten light). M. K 11.31 10.83 G. B 12.56 10.26 Der. 7.94 6.30 MeTuUGGA .fa.0. ces Gres ee ae ee S6.0 6 6 cles e ae: ote s & 616 @ 816 Ue, 6 6 «ale « ola ©. 6 0.0 Cid 0 0/6 6 0.2 0m 6 6 a8 6 oo me 86) 8 4 Cel Ot 6.0 @ a ele e 6b) « S. Rose and G. MacLeod 87 TABLE V. Allowances for loss in digestion of protein. Food material, per cent ABLE VI. Coefficients of Digestibility for Raw and Cooked Egg Whites. Group. I (Raw whites, un- beaten). Average for Group I. II (Raw whites, partly beaten and partly un- beaten). Average for Group II. III (Raw whites, beaten light). Average for Group III. Average for all cases. B For total protein of ration. For ee egg whites Cooked] Raw Difference |aooked| Raw Difference in favor in favor Subject.| ege | ege | 1" ege | egg . ia of cooked 2 5 of cooked white. | white. eae “white! white. | white. egg white. per cent|per cent| per cent {per cent|per cent| per cent MPS (83-82 )575-7 |e 7 - Oe Sieon|) FEZ |, 10s BE. BL, |-87295),85.3)| -eo2ar SS.) 84.8.) a9 Mi Ee. |) 87.5 077.4.) -RROP TOS Eel fec8 |) lees 86.2 | 79.4 |. + 6.8! 86.2 | 76.6 | + 9.6 Ges. | 85.7 1.82.6 | - seb es 4) elo} -- 424 LS. | 86.2 | 82.5 | +2327 | 86-0 80:8 | -- 5:2 MoE. | 87:9 | 82.9 | =Sean0)\788.8°/Si22 | -- 7.0 Pee, (83.3 | SL.9 | = Bea ers. | 79-8) 280 85.8 | 82.5] -- 3.31 85.3 | 80.7 | + 4.6 MITES} So224| SacOe ea Ol ial se | gl 2 | — Ip GB...) 84.8 |°Sk.4) = 3:4 | 84.2 | 79.4 | + 4:8 DE...) 9Os6i Seems hoe 92.5 8900 |, 4-4 86.0 | 82.0} + 4.0] 85.9 | 80.4) + 5.5 rn 88 Digestion of Raw White of Egg CONCLUSIONS. Raw whites of eggs, in as large amounts as ten to twelve whites daily, are well utilized in the human subject, the average coeffi- cient of digestibility calculated for the raw egg white alone being 80 per cent as compared with 86 per cent for cooked whites in the same diet. The absorption varies with the method of prepara- tion, being less for raw egg whites taken in their natural state than when beaten light. A mixture of whites partly beaten and partly unbeaten gave an intermediate value. The quantities con- sumed are regarded as maximal in dietary practice, and it seems unnecessary to emphasize the difference between raw and cooked eggs if the raw eggs are beaten. A MODIFICATION OF FOLIN’S COLORIMETRIC METHOD FOR THE DETERMINATION OF URIC ACID.* By HENRY JACKSON, Jr., ann WALTER W. PALMER. (From the Chemical Division, Medical Clinic, the Johns Hopkins University and Hospital, Baltimore.) (Received for publication, October 19, 1921.) INTRODUCTION. In 1912, Folin and Denis (1) introduced phosphotungstic acid as a color reagent for the detection of uric acid, and from Folin’s laboratory, within a year, appeared methods for the quantitative estimation of uric acid in urine (2, 3) and blood (4). The method was greatly improved by Benedict and Hitchcock’s (5) discovery that cyanide very considerably increased the intensity of the color developed and retarded the fading, and recently, the method! has been still further improved by Folin and Wu (6). Volu-, metric methods have been proposed by Curtman and Lehrman (7) and Morris (8), but the colorimetric procedure has been the method most commonly used. * A preliminary report of this work appears in Proc. Soc. Exp. Biol. and Med., 1920-21, xvili, 126. 1 Transfer a measured volume of blood into a flask having a capacity of fifteen to twenty times that of the volume taken. Dilute the blood with 7 volumes of water and mix. With an appropriate pipette, add 1 volume of 10 per cent solution of sodium tungstate (Na2W0O.2H:;0), and mix. With another pipette, add to the contents of the flask (with shaking), 1 volume of 2 normal sulfuric acid. Close the mouth of the flask with a rubber stopper and give a few vigorous shakes. Much oxalate or citrate interferes with the coagulation and later with the uric acid determination. 20 mg. of potassium oxalate are ample for 10 cc. of blood. When the blood is properly coagulated, the color of the coagulum changes from pink to dark brown. If this change does not occur, the coagulation is incomplete, but the sample may be saved by adding 2 N sulfuric acid, drop by drop, shaking vigorously after each addition, and allowing the mixture to stand for a few minutes before adding more, until the coagulation is complete. Pour the mixture on a filter paper large enough to hold the entire contents of the flask and cover with a watch-glass. The filtrate should be water- clear. 89 en, 90 Folin’s Colorimetric Method The method is simple and accurate and is admirably adapted to the tungstic acid protein-free filtrate so widely used in blood analysis. There are, however, two disadvantages which in routine prac- tice are of some importance; first, the relatively slight intensity of the color developed with such amounts of uric acid as are in normal blood; and secondly, the troublesome crystalline precipi- tate which often appears in the colored solution, thereby rendering reading impossible without filtration, a procedure which some- what diminishes the color. We have endeavored to overcome these difficulties and believe we have succeeded. In an effort to find the cause for the troublesome precipitate developing in the colored solution, we found that if Folin’s uric acid reagent were dialyzed in heavy parchment membranes against large amounts of tap water, until all the free acid was gone, and the solution so dialyzed was evaporated to dry- ness, a reagent was obtained. which, in the presence of uric acid and an excess of NaCN, gave a very intense color and a more or less dense flocculent precipitate. The latter did not alter in amount or character over a period of 24 hours or more. If, on the other hand, Folin’s uric acid reagent were boiled cautiously to dryness without dialysis, a reagent was obtained, which, in the presence of uric acid and an excess of sodium cyanide gave the same intense color, but also a dense crystalline precipitate in the course of 3 to 5 minutes. This last reagent, which we call sodium phosphotungstate ‘‘B’” when mixed with the dialyzed sodium phosphotungstate ‘‘D’”’ will cause dissolution of the flocculent precipitate and at the same time, no crystalline precipitate will develop unless too much ‘‘B” is added. The proper mixture of these two-salts results in a reagent which under conditions of the determination, gives a color nearly five times as intense as that given by Folin’s procedure and no precipitate results. Preparation of Phosphotungstic Reagents. Preparation of Phosphotungstate ‘‘B’’—1,000 cc. of Folin’s reagent” in a large porcelain casserole are boiled rapidly over a * Sodium tungstate (100 gm.), 85 per cent phosphoric acid (80 ce.), plus water (700 cc.) boiled with reflux condenser for 24 hours. Cooled and diluted to 1 liter. H. Jackson, Jr., and W. W. Palmer 91 free flame until the volume is about 400 cc. The size of the flame is then reduced and boiling gently takes place, until the whole is the consistency of pea soup. If the solution turns green, de- colorize with bromine water. At no time should the temperature of the fluid rise above 110°C. Cool in the ice box or ice solution until the solution is about 10°C. A mass of heavy yellow crystals will separate out. Allow them to settle. Pour the supernatant syrupy liquid through a Buchner suction funnel and filter the crystals off in the same funnel. Suck as dry as possible with tamping and strong suction, continuing suction for 3 to 4 hours. Then dry on filter paper over night in an incubator (37°C.). The crystals are sodium phosphotungstate ‘‘B’’, contaminated with a small amount of sodium phosphate. They should be perfectly dry. Yield 90 gm. Preparation of Phosphotungstate ‘‘D’’.—1,000 cc. of Folin’s solution are placed in a sac of ‘‘special” parchment paper, capable of holding 4 liters, and dialyzed against 10 liters of tap water for 5 days. The water should be changed once a day. Otherwise, the procedure needs no attention. The solution in the sac will increase to about 3,000 ce. At the end of 4 or 5 days, titrate 5 ce. of the solution with 0.1 n NaOH, using phenolphthalein as the indicator. 5 ec. of a 20 per cent solution of ‘‘D” neutralizes about 15 cc. of 0.10 N NaOH.. To be sure that all the free phos- phoric acid has gone from the dialyzed fluid, it is necessary to titrate. The total titration value of the solution should be 1,000 to 1,600 ce. If more alkali is needed, dialyze another day to get rid of the remaining phosphoric acid. If not, proceed as below. Transfer the solution in the sac to a large casserole and evaporate over a free flame. If the solution turns green, decolorize with bromine water as often as necessary. As the amount of solution approaches 400 cc. the flame is lowered, and the boiling takes place cautiously, until the amount of solution is about 200 ce. The solution is now transferred to the steam bath, and evapora- tion continued until solid material begins to separate out. Do not allow the solid mass to become hard, so that it cannot be broken easily with a stirring rod. When almost dry, that is, when no free liquid is seen, break up into small lumps and com- plete the drying in the air with occasional breaking up of the -—————- 92 Folin’s Colorimetric Method lumps and occasional heating in the steam bath. If dried com- pletely without breaking up, the mass becomes stony hard and is very difficult to get out of the casserole. This salt is sodium phosphotungstate ‘‘D”. Its properties will depend on the dialyzing paper used in its preparation. The heaviest grade of dialyzing paper made by the Reeve-Angel Company, 7 Spruce Street, New York City, is satisfactory. In this laboratory we have used a very heavy Belgian parchment paper, purchased before the war. It has given far better results than any other paper, but we have been unable to duplicate it. Celloidin saes are useless, as the reagent passes rapidly through such membranes. Ordinary dialyzing papers yield only *‘B’, or “B” and “D” mixed. Papers otherwise unsatisfactory had been made better by coating the inner surface of the paper with a thick celloidin membrane, but we have not employed this procedure sufficiently to speak with assurance as to its value. The longer the dialysis the more intense are the specific properties of ‘“‘D’’, as distin- guished from those of ‘‘B’’, but also the smaller the yield. 5 days dialysis is the average time necessary for 1,000 cc. of Folin’s solution. Smaller quantities take a shorter time and vice versa. In case of doubt, it is better to dialyze another day. If any free phosphoric acid is left in the solution, the reagent is spoiled in evaporation. Yield about 85 gm. Preparation of the Reagent Mixture of “B” and “D”. Make a 20 per cent solution of ‘“B” water. Also make a 20 per cent watery solution of ‘‘D’’. Decolorize with bromine water, if not clear yellow. Boil off the excess of bromine. ‘‘B” will dissolve in water completely giving a perfectly clear solution. There is usually an insoluble residue in ‘‘D’’, and the solution should be warmed and filtered. If the ‘‘B” salt alone be used in analysis, a crystalline precipitate will form in the final colored solution on standing. Should the ““D” salt alone be used, no crystalline precipitate will form, but a more or less dense flocculent precipitate will form immediately. When the ‘‘D”’ salt is pure this precipitate will remain unaltered for 24 hours or more. This flocculent precipitate undergoes dissolution giving a clear solu- tion if some of the ‘‘B” be added. If too little ‘‘B” be added, the flocculent precipitate will not disappear; if too much, it will H. Jackson, Jr., and W. W. Palmer 93 disappear rapidly and the erystalline precipitate will subsequently form. The crystalline precipitate will form early, in case much ‘‘B’”’ has been added, later, if less. There is a point, how- ever, where a solution can be made that will give a clear solution in which the crystalline precipitate will not develop for at least 48 hours, and there is a considerable range on either side of this point where solutions quite satisfactory for ordinary use can be made, in which the precipitate will not develop for more than 12 hours under the conditions of the analysis. Asa rule, from { to 1 part of ‘‘B” to 1 part of ‘‘D”’ is satisfactory. Mix the “B” 20 per cent solution in definite proportions, as follows: 5 ec. “B” + 20 ce. “D” = 1:4. 5 6 eR — 15 “ “p” = 1:3. 5 eeRy =e 10 “ “D” = 1:2. 5 «6 Ry + 5 “t—p”? = 4°14. In each of four small Erlenmeyer flasks put 1 ec. Benedict’s standard, measured roughly. 3 “ 5 per cent sodium cyanide. 5 “ distilled water. 2 “ 10 per cent NaCl in 0.1 n HCl. To each flask add 1 cc. of the above mixture of ‘‘B” and “‘D”. A flocculent precipitate should develop in each flask. Let the flasks stand and watch carefully for the development of a crystal- line precipitate. Choose for the final solution that proportion of “B” and “D” which remains clear for 4 hour or more. If all of them precipitate inside that time, the ‘‘D” salt has been im- properly prepared. If the solutions are not clear of the flocculent precipitate a still greater proportion of ‘‘B’” should be used. Should all the test flasks precipitate, the dialyzed reagent may, of course, be used without any addition of ““B”. 5 cc. of this ‘special reagent’ are ample to produce the full color with 1 mg. of uric acid under the conditions of the determination. Use of Reagents for Determination of Uric Acid in Blood. The essential points of difference between our modification and ’ the original method are as follows: 94 Folin’s Colorimetric Method 1. No sodium carbonate is used, sodium cyanide se the | requisite alkalinity. 2. The solution in which the color is to be acmiaiea must be diluted to a definite volume. 3. The solutions must be diluted at accurate intervals after the reagent is added. 4. A specially prepared uric acid reagent must be used. A protein-free filtrate is obtained with tungstic acid according to the method proposed by Folin and Wu (6). For normal bloods, 20 ec. of filtrate, equivalent to 2 ec. of blood, are used, and for bloods suspected of having over 10 mg. of uric acid per 100 cc., 10 ec. or even 5 ec. may be used, if economy of blood be desirable. To the protein-free filtrate in a 35 ce. centrifuge tube, add 5 ee. of 5 per cent silver lactate in 5 per centlacticacid. Stir thoroughly with a fine glass rod. Wash the rod into the tube with a few cc. of water, and centrifuge 2 to 3 minutes. The supernatant liquid should be clear. Add a few drops of silver lactate. Ifa precipitate forms, add 2 ee. more of silver lactate, stir and centri- fuge again. Decant the supernatant liquid and drain as com- pletely as possible. Add 2 ce. of 10 per cent NaCl in 0.1 n HCl to the precipitate. Let the solution run into the middle of the precipitate, and not down the side of the tube, a procedure which tends to make the precipitate creep. Break up the precipitate very thoroughly with a fine glass rod. So far the procedure is that of Folin and Wu. Now add 4.5 ce. of water as accurately as possible with a Folin.pipette, and stir again very thoroughly. Wash the rod with 0.5 ec. of water from the Folin pipette. Centrifuge rapidly for 5 minutes. Now trans- fer the supernatant liquid quantitatively to a 25 ce. flask. This is best accomplished by holding the flask and a fine glass rod, whose tip just touches the side of the flask neck, in the left hand and pouring with the right. When the bulk of the liquid has drained out, touch and retouch the lip of the tube to the rod, until no liquid adheres to the rod as the tube is taken away on several trials. With the rod still in position, run in (down the rod to wash it) 3 ec. of 5 per cent NaCN. This should be accurate to 0.1 cc. Drain the rod and take it out. - ae H. Jackson, Jr., and W. W. Palmer 95 Now prepare two standards in 50 ce. flasks. To one flask, add 0.5 ec. of Benedict’s standard,? to another 1.0 cc. Now add 9.5 and 9.0 cc. of water, respectively, and to each flask 4 cc. of 10 per cent NaCl in 0.1 n HCl, and exactly 6.0 ec. of 5 per cent NaCN. To the unknown, add 1.5 ce., and to the standards 3 cc. of the special reagent described below. Rotate each flask briskly to insure complete mixture. The additions of the reagent should be made as nearly simultaneously as possible. It is perhaps better to add the reagent at minute intervals and dilute at corre- sponding times. Let the flasks stand 10 minutes by the clock and dilute to the mark in the same order as the reagent was added. The colorimetric estimations may be made immediately after dilution. The calculation is made according to the formula: ISK GSW mace where R is the reading of the standard, a the figure by which it is necessary to multiply the sample taken to make 100 cc. of blood (50 in case 20 ce. of filtrate are taken), b the amount of uric acid in the standard, expressed in milligrams, and FR, the reading of the unknown. The equation must be divided by 2, since the un- known is in a 25 ce. flask, while the standard is in a 50 cc. flask. The color develops slowly, the maximum not being reached in 10 minutes, but the color is sharply proportional to the amount of uric acid present. The depth of color is dependent on the con- centration of the cyanide in the final solution. Cyanide alone in this concentration gives no color until 8 to 12 hours later, when a 3 Benedict’s standard uric acid solution is prepared as follows: 9 gm. of pure crystallized disodium hydrogen phosphate, together with 1 gm. of crystallized sodium dihydrogen phosphate, are dissolved in 200 ec. of hot water and the solution is filtered, if not perfectly clear. The filtrate is made up to a total volume of about 500 cc. with hot water, and this hot solution is poured upon exactly 200 mg. of pure uric acid, suspended in a few cc. of water in a liter volumetric flask. The mixture is agitated for a moment or two, until the uric acid completely dissolves, and then cooled. Exactly 1.4 ec. of glacial acetic acid are added, and the flask is diluted to the mark and mixed. 5 ce. of chloroform are then added to prevent the growth of bacteria or molds. 5 cc. of this solution contain exactly 1 mg. of uric acid. This solution keeps perfectly well for at least a month. 96 Folin’s Colorimetric Method faint tinge of blue develops. A much greater concentration of cyanide will develop a blue color without uric acid. The color continues to increase very gradually over a period of many hours. The increase is proportional in the standard and the unknown. No carbonate is used. Folin’s uric acid standard cannot be used since the presence of sulfite prevents the development of the deep color If serwm, rather than whole blood, be used, the determination may be made after the plan for urine (see below). We have been unable thus far to find polyphenols in serum. That portion of the precipitate from silver lactate which remains behind after liberation of the uric acid with HCl has never given any color with the reagent when serum was used. Apparently, the polyphenols are confined to the cells. Use of Reagents for the Determination of Uric Acid in Urine. The same principles of modification apply to the analysis of urine as to the analysis of blood. Very careful control of dilution and accurate time intervals are required and again cyanide is the only alkali used. Add to 1 or 2 ce. of urine according to concentration in a 35 ce. centrifuge tube 6 cc. of water. Then add 5 ce. of silver lactate, stir and centrifuge. Decant the supernatant liquid. To the precipitate add exactly 4 ec. of 5 per cent NaCN and exactly 10 ce. of water. Stir until all the precipitate is dissolved. Prepare in two 100 ee. flasks suitable standards. To one flask add 2 ec. of standard, to another 4 ec. of standard and 8 and 6 cc. of water, respectively. To each standard add 4 ce. of eyanide. Then with time intervals as with blood, add 5 ce. of special reagent to each flask, and to the unknown in the tube. Mix thoroughly. Allow each to stand exactly 10 minutes, and dilute to the mark. The unknown can be poured into the 100 ce. flask just before the 10 minutes are up and the washings serve to dilute it at the right time. Using 2 ce. of urine in a 100 ce. flask, and 2 and 4 ce. of stand- ard in a 100 ce. flask, a range of from 0 to 60 mg. of uric acid per 100 ce. is covered. With 1 cc. of urine and the same standards a range of from 60 to 120 mg. per 100 cc. is covered. If there is albumin in the urine, it must be removed by heat and acetic acid, or tungstie acid, as even very small amounts of albu- min prevent the proper precipitation of uric acid by silver lactate. H. Jackson, Jr., and W. W. Palmer 97 EXPERIMENTAL. When phosphotungstate ‘‘B” is used alone as the reagent for uric acid a crystalline precipitate develops very quickly; a floccu- lent precipitate develops when phosphotungstate ‘‘D’”’ is used alone. On the other hand, the effect on the formation of a precipitate of combination of phosphotungstate ‘‘B’” and ‘‘D” is quite striking for it has been possible to find a proportion between the two salts which keeps the solution clear (Table I). The proportion of the two salts which gives me best results may vary with the different lots. The depth of color is dependent upon the concentration of the cyanide radical. The tendency to precipitate is determined by TABLE I. Effect of Varying Amounts of ‘‘B’’ and “D’ on Time of Appearance of Precipitate. Proportion ‘‘B”’ to ““D.” Precipitate appeared. 34 minutes. 16 “ce 7) an age Se 60 “c No precipitate. Always cloudy. Nocrystalline precipitate. a Oe BPW Wee oO the OH ion concentration. From these facts it might be inferred that Folin’s original solution could be used with NaCN as the only alkali. But while increased color and freedom from precipi- tate can be obtained in this way, such a large experimental error is introduced that the method is quite worthless,-except when the readings of the standard and the unknown are very close together. This is shown in Table II. Method 1 indicates the method described in this paper. Method 2 indicates the use of Folin’s reagent with NaCN as the only alkali. Method 3 indicates the regular method of Folin and Wu. In the above experiments, the same blood filtrate was used throughout and varying amounts were taken for analysis in order to vary the readings of the unknown in comparison with that of the standard. It will be seen that unless the readings are close THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 Po eee .98 Folin’s Colorimetric Method together very large errors are introduced when Folin’s solution is" used with cyanide alone. TABLE I. Comparison of Methods Using NaCN as the Only Alkali. Readiue(o ibis Amount of uric acid. Method. Reading of unknown’ mg. per 100 cc. au 3.0 1 29 ‘ 2, . 20 3.0 22 : 3 20 3.0 22 ; 1 alk 2.9 33.9 ; 2 Eas 74 44.0 F 20 : 13.4 so 2 aE 4.2 9.5 ; TABLE III. Rate at Which the Color Develops. Solution made. Diluted. Interval. Reading. a.m. - am. min. 10.45 ; 10.50 a 20.0 10.47 10.53 6 20.0 10.49 10.56 ef 20.0 10.50 11.00 10 20.0 11.45 11.50 5 20.2 Development of the Color. The color develops slowly and flasks should stand before dilu- tion for 10 minutes. The rate of increase of color has by this .; H. Jackson, Jr., and W. W. Palmer 99 time fallen off so much that a considerable error in time elapsed before dilution does not introduce an appreciable variation in the reading, as may be seen in Table III. Standards and unknowns made up in 1 hour are comparable with one another. Solution made. Reading. 12.15 20.0 12.45 20.0 1.00 19.9 2.15 21.6 The intensity of the color is, as we have said, dependent on the concentration of the cyanide radical. A considerable variation in cyanide concentration, however, is consistent with accurate readings. Flasks containing 1 ec. of standard, 5 cc. of water, and 1.5 ce. of reagent were made alkaline with varying amounts of cyanide. See Table IV. TABLE Iv. Effect of Varying Amounts of Cyanide. Cyanide. Reading. ce 3.0 20.0 3311 19.9 ee 20.0 3.0 20.0 4.5 18.9 From the foregoing table, it is evident that when the method calls for 3 cc. of NaCN, a variation of + 0.5 cc. does not alter the result. However, in our work we measure the cyanide solution to within + 0.1 ce. _Sodium carbonate or other strong alkali prevents the develop- ment of the deep color, and sodium sulfite lessens the color even more, so that Folin’s sulfite standard cannot be used in our method. The color develops slowly and does not reach a maxi- mum for an hour or more, but there is a sharp proportionality between the amount of uric acid and the depth of color. 100 Folin’s Colorimetric Method Accuracy. We have taken as our standard of accuracy, Folin’s method. The method we propose, we believe is quite as accurate as Folin’s. The analyses in Table V, taken from a large series of determina- tions, serve as examples. TABLE V. Folin’s Method Compared with That of the Authors, Milligrams of Uric Acid in 100 Ce. of Blood. Folin’s method. Authors’ method. PBS 3} 231.2 15.0 15.0 15) 74 15.0 3.0 3.0 3.4 3.0 | Specificity. The reagent, prepared as we have suggested, reacts towards organic reducing substances in alkaline solution exactly as does Folin’s solution. Composition of Phosphotungstates. We have been unable to make satisfactory analyses of the two salts used in preparing the reagent. Our attempts in this direc- tion have convinced us that neither is a pure chemical substance. Both give indications of being mixtures of two or more compounds. Sodium phosphotungstate ‘‘B’’ can be easily recrystallized from water or alcohol. Sodium phosphotungstate ‘‘D” is very difficult to purify and seems to crystallize fractionally as if it were composed of several compounds. We have prepared the free acid from sodium phosphotungstate ‘‘B’’. Furthermore, we have prepared in a comparatively pure state, the blue reduced tung- state. It seems to be stable in air and is blue-black in color. This was prepared by evaporating a solution of the salt in the presence of metallic zine in an evacuated desiccator containing a strong solution of pyrogallic acid in strong NaOH. H. Jackson, Jr., and W. W. Palmer 101 SUMMARY. A modification of Folin’s method for the determination of uric acid in blood and urine is described. In our hands, this modifica- tion eliminates the two disadvantages, the faint color, and the precipitate in the final solution for colorimetric estimation, of the original procedure. ; The color developed by our method is nearly five times as great as that developed on Folin’s methad, and does not fade over a period of several hours. In fact the color increases gradually and proportionally in both standard and unknown, so that when standard and unknown are made up at the same time, they may be read at any time during the next 2 hours or more. No crys- talline precipitate develops in the colored solution. BIBLIOGRAPHY. . Folin, O., and Denis, W., J. Biol. Chem., 1912, xii, 239. . Folin, O., and Macallum, A. B., Jr., J. Biol. Chem., 1912-13, xiii, 363. . Folin, O., and Denis, W., J. Biol. Chem., 1913, xiv, 95. Folin, O., and Denis, W., J. Biol. Chem., 1912-138, xiii, 469; xiv, 29. . Benedict, S. R., and Hitchcock, E. H., J. Biol. Chem., 1915, xx, 619. . Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81, 459. : eerana: L. J., and leehenisncs A., J. Biol. Chem., 1918, xxxvi, — Morris, J. L:, re Biol. Chem.., 1919, EXXVIN, 230, DANOnwrwWnN re AMINO-ACIDS IN NUTRITION. IV. A MODIFIED BIOLOGICAL METHOD OF STUDYING AMINO- ACID DEFICIENCIES IN PROTEINS. CYSTINE AS A GROWTH- LIMITING FACTOR IN THE PROTEINS OF THE GEORGIA VELVET BEAN (STIZOLOBIUM DEERINGIANUM). By BARNETT SURE. (From the Laboratory of Agricultural Chemistry, University of Arkansas, Fayetteville.) (Received for publication, October 19, 1921.) Employing essentially the technique of McCollum and Davis (1), using the seed as the only source of protein, and satisfying all the other dietary essentials in the rations, the author attempted to make amino-acid additions to the pea (Vicia sativa) as a part of his general scheme of studying the capacity of the animal organism to synthesize the pyrrolidine nucleus of the protein molecule (2), but in that study has met with no positive results. This paper will show, by a modified method of procedure, the practicability of making amino-acid additions to such complex substances as seeds. Waterman and Jones (3), in a recent communication, state that amino-acid deficiencies in the Chinese and Georgia velvet beans cannot account for the failure to promote growth; for analy- ses (4, 5) have shown the above mentioned proteins to be adequate ‘except possibly with respect to cystine; and no improvement resulted from the addition of this amino-acid. It will be shown in the following pages by the modified pro- cedure adopted by the author that cystine is unquestionably a growth-limiting factor in the proteins of the Georgia velvet bean, which could not have been found by employing previous methods of making straight amino-acid additions. The results of the experiments are given in the following charts. 103 104. Amino-Acids in Nutrition lee — - ——— — Cuart I, Lor LV. This chart indicates that when Georgia velvet beans are fed as the only source of protein at a level of 40 per cent, very little growth takes place, and that arachin, one of the globulins and the main protein from the peanut, does not furnish DENS acids to supplement those deficient in the velvet beans. At point “‘a’”’ 0.4 per cent of the ration was added in the. form of cystine, but no response was obtained. sate ca Salts] #32 peal eee aii aes eae Dextr Cuart II, Lor XCV. This experiment demonstrates that, in the presence of gliadin, there is a definite response to cystine addition to the velvet bean proteins, which is not, however, very marked 2 weeks after this amino-acid addition. 105 Barnett Sure ‘ulezZ pue UIPVITS sv suroqyoid yuotoyop yons Aq porddns sv poysiz¥s o1B Splov-OUTU 19Y}O 10458 Ajuo yus1edde sautooeq Yor ‘UB JOATOA VIBIOOD oY} Jo suteyord oY} Ul 10z0vJ SUI}IUAT][-4YI MOIS B ST oUTySAD 4BY JsoBsns A[Zu013s ‘AQX JOT ‘Burpoooid oy} pus quowliodxe sIyy, *.,q,, purod 4% oursho oy} 07 ouvydoydéay JO UOTJIPPV oY} UO YMOIS JO TOJOVIVYO OY} Ul OSVOIOUT TOYZANJ OU YIM YjMOId Ul yUOWOAOCIdUIT ayuyop nq MOS B ynoge yyZno1q ,,B,, yuIod 4v ouTysXd Jo UOTZIPpE oY], ‘poprezor ATQvioptsuoo SI YMOIS JO OANYVU OY} SYOOM Q 10}jv yVYy} poo oq ]]IM 4 “Gnq “YZMOoIs jo sporied Jelp1ve ayy UI UBOG JOATAA VISIOIH OY} Jo suIezo1d oT} qguowe[ddns 03 sues UlIaZ “[OX LOT ‘J]]T LuvHH CR fsbo peddle edad “outa pA wimz bq4 uT| poppe sua ueT4Zea Jo quep wd : . L [frend see nrfsaef mein of ozo afenfrnefoy fame fom x || A) SSSR ReS eee eee ee | ——_— f, Te i ele oe Ll pa Va | ee oT oT Be Oe OFT OP 09 08 OOT 106 Amino-Acids in Nutrition ‘| velvpt Beaps 40 Gelatin 9 CuartIV, LorLVIII. The behavior of the animals on this ration certainly indicates that gelatin does not furnish the necessary missing links deficient in the proteins of the velvet bean. Se Meas Spee CERES re AmB resi i A Es = e sed i form af Cuart V, Lor CXIX. This is a duplicate of the preceding experiment, Lot LVIII. The experiment was repeated using animals a little heavier, so that they could better maintain themselves on the poor velvet bean- gelatin ration, and in order to have an opportunity to make a cystine addi- tion. After the adding of 0.4 per cent of the ration in the form of cystine, Be Salad Barnett Sure 107 the writer found, to his great surprise, that all the animals, although they have previously failed to make any growth (Chart IV, Lot LVIII) and pro- duced only maintenance curves on the same ration in this experiment, have begun to grow in a very marked manner, and have made excellent growth for a period of 10 weeks after the amino-acid addition, after which time the experiment was discontinued. This experiment, then, furnishes conclusive evidence that, providing other amino-acids in the form of the deficient protein (gelatin) are supplied, cystine shows itself up remarkably as one of the determining growth-limiting factors in the proteins of the Georgia velvet bean. Eile allak. ea SNR aes) LEVER ec 2eWwResir = aes | =] hy ZEEE GEen i B NN af) a v8 ala co oe JER SERRE el r mie ata by HABRS | Cuart VI, Lor CIX. This ration illustrates that even when the Geor- gia velvet bean proteins are fed at as high a level as 60 per cent, which would furnish 16.5 per cent protein, very little growth takes place. At point ‘‘a” 0.2 per cent of the ration was added in the form of tryptophane, but without any response. At point “‘b’’, after the addition of cystine to the extent of , 3 [. SE eee 128 Sugar Excretion in Severe Diabetes added and thereafter 25 cc. more each 4th day until six additions had been made, then 25 cc. cream every other day. In this case each fresh addition to the diet tended to cause a slight increase of the sugar excretion for that day followed by a return to the former level on the following day. These fluctuations were most marked early in the experiment and grew less as time passed. The average excretion was very constant at 600 to 700 mg. While no change took place in the total quantity of sugar eliminated per day (500 to 1,000 mg.), during the time that the glucose equivalent of the diet rose from 49 gm. to 70.3 G and the calories from 830 to 1,270, there was a sharp break in the sugar eliminated with the next addition of 25 ce. cream. With the diet aggregating 1,270 calories with G = 70.3 gm. the excretion ran 942 and 1,420 mg. on 2 successive days. The next addition was followed by excretions of 1,504 and 1,906 mg. Another addition was ® Glucose 43 24 CHART 7. then made and the excretion rose to 2,650, 5,568, and 6,300 mg. In this case it will be noted that when the total glucose equivalent of the diet was 68 gm. the excretion was still only 825 and 529 mg. on 2 days. The subse- quent addition of food equivalent to 6.3 gm. glucose led to the excretion of approximately this quantity of glucose over and above the former average. There was, in short, a virtually complete excretion of all the glucose supplied in excess of a certain limit. CONCLUSION. Study of the curves obtained leads to the conclusion that indi- viduals with severe diabetes, when brought into the non-diabetic status (or, as it is sometimes called, the ‘sugar-free’ state) by H. Felsher 129 fasting or other more suitable adjustments of the diet, may then excrete small quantities of sugar not greater than those excreted by normal individuals under parallel conditions. The quantities in this series averaged between 10 and 15 mg. per kilo per day. As the diet is gradually increased stepwise at 1 to 4 day intervals, there is at first little or no permanent increase of the sugar excre- tion. The sugar excreted has remained entirely unaffected; or it has shown a definite but temporary acceleration with each new addition to the diet to be followed by a restoration of the former level; or it has shown a slight rising tendency from the start. But in any case the total permanent increase of the sugar excre- - tion has remained slight or even unrecognizable until the total glucose equivalent of the diet has risen above a certain limit (which varies with the individual). Once this limit has been passed, further additions to the diet lead to rapid—even sudden— accelerations of the sugar excretions, out of proportion to any which have occurred before. The curve may then bend rapidly upward or show a true critical break. This observation is in harmony with the well known concep- tion of a clearly definable “‘tolerance limit’’ for glucose in dia- betes; and that an ‘‘abnormal” sugar excretion may develop with critical suddenness when this limit is overstepped. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 STUDIES ON PROTEINOGENOUS AMINES. XII. THE PRODUCTION OF HISTAMINE AND OTHER IMIDAZOLES FROM HISTIDINE BY THE ACTION OF MICROORGANISMS. By MILTON T. HANKE anp KARL K. KOESSLER. (From the Otho S. A. Sprague Memorial Institute and the Department of Pathology of the University of Chicago, Chicago. ) (Received for publication, October 21, 1921.) CONTENTS. Part I. The products formed from histidine by the action of Bacillus coli—in the narrower sense—which includes Bacillus coli communis, Bacillus coli communior, Bacillus lactis aero- genes, and Bacilweaeids lachiet:, .. cscs se cen Genel « to sate 138 Part II. The products formed from histidine by the action of other members of the colon typhoid group; Bacillus enteritidis, Bacillus typhosus, Bacillus paratyphosus A, Bacillus para- typhosus B, Bacillus dysenteriae Flexner, Morgan, and Shiga, Bacillus fecatis alealtigenes I and Ill...7.........-.....- 155 Part III. The products formed from histidine by the action of other microorganisms; Bacillus mucosus capsulatus, Bacillus bifidus communis, Bacillus influenze, Bacillus proteus vul- garis, Bacillus cloace I, Streptococcus hemolyticus, Pneu- mococci I, II, III, and IV, Staphylococcus, and Bacillus HARES o to 5 6 ACEO SOOD PAGOOID CR Cr DOCOLaE DESEO AUaooen 165 Part IV. The production of histamine from histidine when other amino-acids or peptones are added to the medium........ 177 INTRODUCTION. In a series of papers published in 1919! we communicated a method for the microchemical colorimetric estimation of imid- azole derivatives and for the quantitative separation of histamine from histidine. These methods enabled us to study the metab- olism of histidine under various conditions. Experiments 1 Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1919, xxxix, 497, 521, 539. 131 ns 132 Studies on Proteinogenous Amines. XII with a colon bacillus, isolated from a case of cystitis, gave results that led us to the following conclusions: 1. When the colon bacillus is allowed to metabolize histidine, either alone or in the presence of nitrates or ammonium salts, histamine is not formed. 2. In a medium containing histidine and glycerol, but no nitrates or ammonium salts, histamine is not formed. In this case imidazole propi- onic acid appears to be formed; but only when the bacillus is forced to grow anaerobically. 3. In a medium containing histidine, glycerol, or glucose and a source of nitrogen, either KNO;, NH.Cl, or both, about 50 per cent of the histidine is converted into histamine in the course of 2 weeks when oxygen is present. In the absence of atmospheric oxygen, this and all the other metabolic ac- tivities of the bacillus are greatly reduced, probably because the colon bacillus is an aerobic organism by preference. 4. The production of histamine is always coincident with the production of a medium that is distinctly acid. We believe that the histamine is formed by the bacillus to neutralize the excess of acidity that is simul- taneously produced from the glycerol. 5. Contrary to the statement sometimes given in text-books and in the literature that carbohydrates prevent the formation of histamine from histidine, we have found that histamine is never formed except in the presence of an easily available source of carbon such as glycerol or glucose. Having established the above facts for one particular colon bacillus, we were then led to a consideration of the following queries: 1. Are all strains of colon bacilli capable of decarboxylating histidine in our standard medium? 2. Are other organisms capable of converting histidine into histamine under identical conditions? 3. How does the addition of amino-acids or peptones to our standard medium containing histidine, influence the production of histamine? Procedure In every case the same number of microorganisms—nine billion*—was introduced into 200 cc. of an autoclaved medium * For a detailed description of the method see Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1919, xxxix, 539-555. ’ The method employed in counting the organisms was that of Breed and Brew (Breed, R. S., and Brew, J. D., New York Agric. Exp. Station, Techn. Bull. 49, 1916). aa | lc M. T. Hanke and K. K. Koessler 133 having the following composition: 0.2000 gm. histidine dichloride, 0.2000 gm. ammonium chloride, 0.1000 gm. potassium nitrate, 0.4000 gm. potassium dihydrogen phosphate, 0.800 gm. sodium chloride, 0.0200 gm. sodium sulfate (anhydrous), 0.4000 gm. sodium bicarbonate, 0.0100 gm. calcium chloride (anhydrous), and 4.00 cc. glycerol dissolved in sufficient distilled water to give a final volume of 200 cc. The inoculated flasks were then incubated at 37° for 14 days, unless otherwise specified, after which the mixture was forced through a Berkefeld filter. The hydrogen ion concentration of the filtrate was determined colorimetrically (see below for de- tails). The filter was then carefully washed with at least 200 ce. of water. Concentrated sulfuric acid (1.0 cc.) was added to the combined filtrate which was then freed from water by evapora- tion in a glass dish on the water bath. The syrupy residue was transferred, with distilled water, to a 25 cc. precision cylinder and diluted to exactly 25 cc. Of this test liquid 10 cc. were trans- ferred to a 35 cc. glass-stoppered bottle, treated with 3 gm. of solid sodium hydroxide, and extracted six times with amyl] alco- hol using 20 ce. for each extraction. This divides the material into two fractions, the amyl alcohol extract, which may contain histamine and methyl imidazole and which we refer to as the histamine fraction, and the alkaline aqueous liquid which con- tains histidine and may contain imidazole acetic, propionic, lactic, and acrylic acids. The combined amy] alcohol extracts were extracted with normal sulfuric acid, which removes the imidazoles. The acid extracts were nearly neutralized with 5 n NaOH and the resulting liquid was diluted to exactly 100 cc. The amount of histamine pres- ent was then determined colorimetrically by means of the well known reaction that occurs between p-phenyldiazonium §sul- fonate and imidazole derivatives in a solution rendered alkaline with sodium carbonate.!. When the presence of histamine was indicated by the colorimetric determination, an amino nitrogen determination was also carried out on this fraction. The values obtained by these methods check closely when histidine is the only amino-acid present in the original medium. In other cases the amino nitrogen values are invariably high. To be certain that histamine was the only imidazole present, a methyl imidazole 134 Studies on Proteinogenous Amines. XII determination was also carried out. We have never encountered methyl imidazole as a product of the bacterial decomposition of histidine. The presence of histamine was, moreover, qualita- tively verified by means of physiological methods. The histidine fraction was transferred to a 25 cc. graduated precision cylinder with water and 7 cc. of 37 per cent HCl. The cooled acid liquid was then diluted to 25 cc. Portions of this liquid were than tested for histidine, colorimetrically and by means of a Van Slyke amino nitrogen determination. When the values obtained check closely, histidine is probably the only imidazole present. If the color obtained is too red, and the colorimetric determination indicates the presence of considerably more imidazole than can be accounted for as histidine by the amino nitrogen method, the excess color is probably due to im- idazole acetic, propionic, lactic, or acrylic acids. Which of these acids is present cannot be determined without an isolation ex- periment. We have made no effort to isolate these acids but have calculated the excess imidazole value as imidazole propionic acid. When the amino nitrogen determination indicates the presence of considerably more histidine than can be accounted for colorimetrically, we are confronted with two possibilities: 1. Some of the introduced ammonia may have been converted into a carboxylated, alkali-stable, primary amino compound— possibly an amino-acid—which would give off nitrogen with nitrous acid; or 2. Some of the histidine may have suffered a rupture of the imidazole ring with the liberation of free amino groups. Up to the present time we have brought no absolute proof that either of these is the correct explanation for the facts. At present we are inclined to believe that the second of these possi- bilities is the most probable and our reason for this belief is out- lined in the following pages. Determination of the Hydrogen Ion Concentration. In our earlier work we employed the set of standard phenol- sulfonephthalein tubes furnished by Hynson, Westcott and Dun- ning to determine the pH of our media after incubation. This set is inadequate for work of this kind because it is useful only ee eee M. T. Hanke and K. K. Koessler 135 within the narrow limits of pH 6.6 to 8.6. Most of our final media were too strongly acid to fall within the range of this in- dicator. We have, therefore, prepared a series of standard tubes with a lower pH value of 1.2 by following the method of Clark and Lubs.t The indicators selected were phenol red, range 8.6 to 6.6; brom-cresol purple, range 6.8 to 5.2; methyl red, range 6.0 to 4.4; brom-phenol blue, range 4.6 to 3.0; and thymol blue, range 2.8 to 1.2. We have found that the buffer solutions containing the sul- fonephthalein derivatives can be kept for at least 2 years, and hence probably indefinitely, in a sealed Pyrex glass tube after preservation with thymol. It is possible, therefore, to prepare a permanent set of tubes, similar to those distributed by various con- cerns, for the phenol red range, for all of the sulfonephthalein derivatives. Methyl red, which is not a sulfonephthalein deriv- ative, deteriorates so rapidly that a set of permanent tubes containing this indicator cannot be prepared in the usual way. We have found, however, that the colors produced with methyl red can be imitated by means of an aqueous solution containing mixtures of Congo red and methyl orange. These thymol-pre- served aqueous solutions containing Congo red and methyl orange can then be sealed up and kept in Pyrex glass tubes similar to those used for the other standard sets. They seem to keep indefinitely. The determinations were carried out as follows. A drop of the filtrate to be tested was transferred to a porcelain test plate and mixed with 1 drop of indicator. This procedure was repeated until a color was obtained that was within the range of one of the indicators. Then 1.0 cc. of the filtrate to be tested was mixed with 0.10 cc. of the proper indicator in a Pyrex test-tube and the color compared with that of the standard tubes. The inherent color of the filtrates was never sufficiently intense to have any effect upon the accuracy of the determination. Method and Table for the Estimation of Small Amounts of Imidazole Lactic Acid. In our previous papers! we reported tables by means of which colorimetric readings could be converted into milligrams of his- 4 Clark, W.M., and Lubs, H.A., J. Bact., 1917, ii, 1. 136 Studies on Proteinogenous Amines. XII tidine, histamine, methyl imidazole, and imidazole acetic and propionic acids. It seemed desirable to have a similar table for the estimation of imidazole lactic acid. This substance was, therefore, prepared using the method described by Frinkel.® The perfectly white solid obtained by this method, after two re- crystallizations from water, had the following properties. . Melting point 217°. . Chlorine—none. 4 . Residue on ignition—none. . Ammonia—none. . Amino nitrogen (Van Slyke method)—none. 6. The solid—0.1000 gm.—was dissolved in 10 ce. of 0.10 Nn NaOH and allowed to react for 1 hour at room temperature. The excess of alkali was determined by titration with 0.10 n HCl, using phenolphthalein as indicator. The first change in the indicator was obtained when 4.3 ec. of the 0.10 nN acid had been added, 4.5 cc. of the acid being required to give a colorless solution. The indefinite end-point obtained is exactly what one would expect of a substance having a fairly strong acid group and a feebly basic group. The 5.7 cc. of 0.1 n NaOH used for 0.1000 gm. of substance agrees very well with the 5.74 cc. demanded by theory for CeHsN203'H2O. We therefore considered the substance to be 100 per cent pure. Cte Whe A stock solution was prepared by dissolving 0.5000 gm. of the solid in 28.7 ce. of 0.10 N HCl and diluting with water to 50 ce. From this the standard test solution was prepared by diluting 1 cc. to 100 cc. in a volumetric flask. The tabular values were then obtained by mixing different amounts of this standard solu- tion with the alkaline p-phenyldiazonium sulfonate reagent as previously described.!. The color produced was then compared in a Duboseq colorimetér with a standard solution of Congo red.° The color produced matches that of the Congo red solution perfectly. A color of maximum intensity is obtained within 3 to 5 minutes and it is stable for from 5 to 10 minutes during which time an accurate comparison can easily be made. With § Frinkel, S., Monatsh. Chem., 1903, xxiv, 229. ° To prepare the Congo red solution, vacuum-dried Grubler’s Congo red (2.5000 gm.) is mixed with 50 cc. of absolute alcohol in a 500 ce. volume flask. Water is then added to the mark. This is the stock solution which keeps indefinitely. From it the standard indicator solution is prepared by diluting 1.00 ce. with distilled water to 500 cc. in a volume flask. M. Hanke and K. K. Koessler 137 TABLE I. Estimation of Small Amount of Imidazole Lactic Acid. Imidazole lactic acid (CsHsN203-H2O) in the test cylinder. (Total volume 8 cc.) Test cylinder set at 20 mm. Depth of indicator solution (CR) required to match the color in the test cylinder. mm, gm. 1.0 0.000001 2.0 0.000002 3.0 0.000003 4.1 0.000004 5.1 0.000005 6.1 0.000006 7s 0.000007 if 8.1 0.000008 9.2 0.000009 10.2 0.000010 ; ? 11.2 0.000011 ' 12.2 0.000012 13.2 0.000013 | 14.3 0.000014 15.3 0.000015 16.3 0.000016 . ee 0.000017 18.3 0.000018 19.4 0.000019 ‘ 20.4 0.000020 | 21.4 0.000021 22.4 0.000022 23.4 0.000023 24.5 0.000024 25.5 0.000025 26.5 0.000026 27.5 0.000027 28.5 0.000028 » 29.6 0.000029 30.6 0.000030 31.6 0.000031 32.6 0.000032 33.6 0.000033 34:7 0.000034 35.7 0.000035 36.7 0.000036 a7 7 0.000037 i 138 Studies on Proteinogenous Amines. XII | TABLE I—Concluded. Imidazole lactic acid (CeHsN203.H2O) in the test cylinder. (Total volume 8 cc.) Test cylinder set at 20 mm. Depth of indicator solution (GR) required to match the color in the test cylinder. mm. gm. 38.7 0.000038 39.8 0.000039 40.8 0.000040 41.8 0.000041 42.8 0.000042 43.8 0.000043 44.9 0.000044 45.9 0.000045 46.9 0.000046 47.9 0.000047 48.9 0.000048 50.0 0.000049 51.0 0.000050 quantities of this acid in excess of 0.000025 gm., the color pro- duced is too intense to enable one to make an accurate comparison. In such cases the test cylinder was set at 10 instead of at 20. The reading obtained was then multiplied by two before subtracting the correction blank, which is 0.30 mm. as in the case of the other imidazoles.’ PART I. On The Products Formed From Histidine by the Action of Bacillus colt. These four organisms (Bacillus coli communis, Bacillus coli communior, Bacillus lactis aerogenes, and Bacillus acidi lactict) com- prise the colon group (in the narrower sense) according to the classification of the American Public Health Association. This differentiation is based upon the different behavior of these four organisms toward dextrose, lactose, dulcitol, and saccharose and is represented as follows: 7 This correction blank represents the amount of color that is autoge- nously produced by the reagent even when imidazoles are absent. See foot-note 1. M. T. Hanke and K. K. Koessler — 139 Bacillus coli Group. Dextrose + Lactose ++ | | Dulcitol + Dulcitol — B. communior B. aerogenes B. communis B. acidi lactici | r | Tae. Saccharose + Saccharose — Saccharose + Saeccharose — B. communior B. communis B. aerogenes B. acidi lactict The behavior of the organisms investigated by us is summar- ized in Table IJ. We have included a large number of tests that were not required for differentiation purposes because we hoped to find a correlation between these reactions and the pro- duction of histamine. Such a relationship seems, however, not to exist. The organisms Coli K, H, B, 88, 51, 80, 84, 90, 52, 74, and Schwartz were isolated from human feces and investigated as soon as they had been obtained in pure culture. The other strains were stock cultures obtained from a variety of sources. Some of these strains have been growing on artificial media for years. Bacillus coli communior. Of the organisms investigated, eight belonged to this group. Their behavior on our synthetic medium is summarized in Table III. ; Coli P-3-19.—This organism did not rupture the imidazole ring to an appreciable extent. Of the histidine originally introduced, 29.7 per cent was converted into histamine and 59 per cent was recovered unchanged. Imidazoles other than histamine were not formed. Of the ammonia originally introduced, 42 per cent was removed by the micro- organisms. Coli K (red).—There is a sufficient discrepancy between the amino nitrogen and color values for both the histidine and the histamine fractions to suggest a slight rupture of the imidazole ring with the production of primary amino groups. Of the histidine originally introduced, 37.6 per cent was converted into histamine and 50 per cent was recovered un- changed. Imidazoles other than histamine were not formed. Of the ammonia originally introduced, 9.7 per cent was removed by the bacilli. TABLE II. Classification. Bacillus communior. Bacillus communis. ‘uoonpoid aurmeysiyT | + ++ | | | | “roneysorg-ses0, | | ‘ajopay | + ae ‘unryeg | | | ‘uynuy | | | Tong, [FAR eeebee ae aso | + asOuIqeVly | + ‘asouuwyy | + ‘jonuueyw | + + ‘osougey | + + ‘asoye my | + ‘esomaay | + “aSOJOR[B | + vasormqoong | + + +++ +++ sopet| ++++++++ esonxog | ++ ++++++ (Jo woryenseoo) IW | +++4+++++ ears Miata +)+|+]+]+]+)+|+ + 4 Sa, ine” WR exe oy BSoC iia vo = ‘opua uo AuOlO_ BD" SH uaa Sats Soe eal Yee = ures | | | | | it it it-i “ergo |~ 4 A) ete aa Name of strain. “ “ce Ve ln eee ae ‘s SEL OUINI NOL Ore cesta tn ee. (0) hig! nas ot Io Fe ls eo a ee we he” te ue, Aa et a ee ee . _ ~ - ~ - < +e a 9 Sel) a ~~ S MS) See al 4 “WYoD) IpIow snypong false ++ 4. a +- + ao 4. aa +|+ +|+ +|+ +|+ +|+ +/+ + ++" 1 | ae ee a FFE EH FEE E+ p+++++ FHFEHEEEHE EH FHF HE HTH HH leas ls eG ee ala tHE EEH EHS | | | PE ee oie ha een es raroor OD Ts sis (ql6)(@) (ore! 0) aire) (ele, 0, @ joke -e!te) s)(4) af ete ce! (away o. Olea ees Sie) oe) 6.96 eves (6 (sl6la) entice le leis) s'e re st re Name of strain. Coli P-3-19. Coli K (red). Coli Y, 14 days. 30 days. Coli Jd. Coli Lac. Aer., 7 days. 14 days. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. = 9.6mm. 0220'S =1952 Match perfect. 96% 0.10 ec. = 10.8 mm. 0:20:-**" =. 21°65 55 Match perfect. 108% 0.10 ec. = 10.0 mm. 0.20. * = 2020. Match perfect. 100% O=10 ce =) 7-4anme: 0:20°% = 1408>"5 Match perfect. 74% 0.10 ee. = 10.1 mm. 0.20. © = 20.2% Match perfect. 101% 0.10 cc. = 8.7mm. O20 4 = 1740 0* Match perfect. 87% 0.10 cc. = 5.2mm. 0,20“ =10; 4°" Match perfect. 52% Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. = 5.9mm. 0:20) “ssl Sms Match good. 59% 0.10 ce. 0:20 ‘—= 10.0 -* Match perfect. 50% 0.10 cc. = 9.0mm. 0.20 “ =—18.0 “ Match perfect. 0.10 cc. = 6.6mm. 0.20) 12 ae Match good. 66% 0.10 cc. = 8.6mm. 0.20 See Match perfect. 0.10 cc. = 7.5 Of20 oo — alos Oe Match perfect. 75% 0.10 cc. = 4.5mm 0320) 3° = Son0nes Match good. 5.0 mm. TABLE WI—Bacillu Unchanged histidir (Van Slyke method) 5 cc. of test solution. | | .18 cc. Nz at 22° ang 745 mm. 0.1327 gm. histidins dichloride. | 66.3% — 1.14 cc. Ne at 26° an 750 mm. 0.1264 gm. histidin dichloride. 63.2% | 1.63 cc. Nz at 27.6) and 744 mm. | 0.1795 gm. histidin dichloride. 89.7% 2.46 cc. Ne at 24° an 748 mm. 0.275 gm. histidin dichloride. | 137.5% | \ i i _ .55 ec. Ne at 22° an 743 mm. 0.1738 gm. histidir dichloride. 86.9% .37 cc. Ne at 16° an 75020m. 0.1597 gm. histidir dichloride. 79.8% =" .81 ec. Ng at 23° an 748 mm. 0.2032 gm. histidir dichloride. ; 101.6% — * Colors matched against the (CR-MO) standard. 142 into histamine (colorimetric deter- mination). _ Color value of istamine fraction.* tamine in en- olor develops like that of histamine. lution. 37.6% of hista- mine present. 50 cc. 4.8mm. .00 “ 9 : 6 “ olor develops like 0.00128 gm. of histamine in entire testso- None. Histidine converted H 0.048 gm. of his-| 0.46 ec. Ne at tire test solu- | 0.042 gm. his- entire test so- | 0.0785 gm. his- ‘that of histamine. lution. 1.98% 20 cc. = 4.0mm. | 0.00668 gm. of mn =. 8:0.“ histamine di- olor develops like chloride in that of histamine. entire test so- lution. 4.1% istidine converted into histamine 0.10 N HCl neutralized (Van Slyke method) by NH; from with 5 ec. of test entire test solution. solution. 210 Hence the = of 15 ce. of 0.1 n NH; used 22° and748 mm. that of histamine. tion. tamine di-| by microorgan- 29.7% of hista- chloride. isms. mine present. 26% 05 cc. = 9.1mm. | 0.0607 gm. of | 0.9cc. N2at30° 32.0 m “=18.2 “ histamine in and 748 mm. | Hence the = of 3.5 ec. of 0.1 Nn NH3 tamine di-| used by the micro- chloride. organisms. 48.6% 20.0 Hence the = of 16 ce. of 0.1 n NH; used by microor- ganisms. Reaction, Before | After incu- | incu- bation. | bation. None. Hence the of 36 ce. of 0.1 n NH; used by microor- ganisms. ~ = 28.6 Hence the = of 8 cc. of 0.1 n NH3 used None. None. by microorgan- isms, 36.7 Hence the = of 0.7 ec. of 0.1 n NH; was produced by microorganisms. 38.0 Hence the = of 2 ce. of 0.1 n NH; was produced by the microorganisms. 143 pH pH 7.3 | -539 mE 7.3) 620 7.3 | 4.6 7.3] 5.6 me ay 144 Studies on Proteinogenous Amines. Name of strain. Coli 88. Coli 51. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ec. = 10.0 mm. 0.20 “ = 20.0 “ Match perfect. 100% 0.10 cc. = 10.0 mm. 02200 — 220) Ores Match perfect. 100% Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10'ce. = 8-3 mm: 0.20 “t= 36/0522 Match perfect. 83% 0.10 cc. = 8.2mm. 0320") 16:45 Match perfect. 82% 0.1775 gm. histidin XII Unchanged histidine (Van Slyke method) with 5 ec. of test solution. | i 1.62 ce. N2 at 27° ane 746 mm. 0.1775 gm. histidin dichloride 88.8% 1.62 ec. Ne at 27° ane 746 mm. dichloride. 88.8% M. T. Hanke and K. K. Koessler 145 oye ae Histidine converted Reaction. atantealoct Hisndine eee into histamine 0.10 N HCl neutralized istamine fraction.* (colorimetric deter- ee ke method) ant en ahem Before | After incu- | incu- bation. | bation. mination). solution. 37.0 Hence the = of 1 ce. of 0.1 n NH3 was| 7.4] 5.4 produced by the microorganisms. eo SO ee | | 33.75 Hence the = of 2.25 ce. of 0.1 n NHs was removed by the microorgan- isms. 146 Studies on Proteinogenous Amines. XII Coli Y.—During the first 2 weeks of incubation, 8 per cent of the intro- duced histidine and 44 per cent of the introduced ammonia were removed by the microorganisms. Of the histidine originally introduced, about 2 per cent was converted into histamine. There is no evidence that the imidazole ring was ruptured because the check between the histidine values by amino nitrogen and colorimetric determinations is perfect. During the next 16 days of incubation an additional 2 per cent of his- tamine was produced. The liquid was free from ammonia; hence 100 per cent of the ammonia originally introduced was removed by the organ- isms. The most striking observation, however, was the discrepancy between the histidine values obtained by the colorimetric and the amino nitrogen methods. In a previous paper we called attention to the fact that his- tidine might be decomposed by microorganisms according to the following structure: ic Oaal H C-— Nv C — NH; | cH N C — NH: O | | Vi a Saas eee OH COOH COOH The triamino compound formed, because of its carboxyl group, would form a sodium salt, in the presence of a strong alkali, that should be soluble in water and difficultly soluble in amyl alcohol. In the course of our standard procedure this triamino compound should appear together with histidine in the histidine fraction. Although histidine contains but one primary amino group, the triamino compound contains three such groups; hence each mole- cule of triamino compound would give three times as much nitro- gen, by the Van Slyke method, as a molecule of histidine. Going back to the table we find that 90 per cent of the introduced histi- dine was left unchanged at the end of 14days. After 30 days of incubation, 66 per cent of histidine was present and 2 additional per cent of histamine had been produced. In all, then, 68 of the 90 per cent of histidine left after 14 days of incubation can be accounted for colorimetrically. What became of the 22 per cent of histidine that cannot be accounted for colori- metrically? If we assume that all of this histidine was converted into triamino compound, an amount of Ne equivalent to 22 times 3 or 66 per cent of histidine would be evolved in an amino nitrogen determination. M. T. Hanke and K. K. Koessler 147 If to this we add the nitrogen evolved by the 66 per cent of histidine present in the same liquid, one would expect to obtain an amount of nitrogen equivalent to 132 per cent of histidine which compares very well with the 137.5 per cent actually obtained. We realize, of course, that an excess of amino nitrogen may not necessarily indicate the presence of a histidine disruption prod- uct. It is possible that non-volatile, carboxylated amino com- pounds, possibly amino-acids, might be synthesized from ammonia and glycerol, and if these were present, the amino nitrogen figure would be high. At present, however, we are inclined to believe that a triamino compound is responsible for the excess amino nitrogen because a quantitative relationship similar to the one given above has been found to hold in five other cases. A quanti- tative agreement might be obtained once or twice by accident; but it seems hardly reasonable to assume that the accident should occur six times. The formula of the triamino compound would lead one to be- lieve that the compound might have some physiological activity. A search of the literature revealed the fact that diamino acety- lene derivatives have not yet been prepared. We hope to make the preparation and properties of this triamino eoooud the subject of a subsequent paper. Finally we can raise the question, why do certain microorganisms rupture the imidazole ring with the liberation of free amino groups? Two reasons suggest themselves. This type of nuclear disrup- tion is the most certain way to expose for future use all of the nitrogen and carbon of the molecule. This is also a decomposi- tion that converts a feebly basic substance into one that is strongly basic. Coli Jd.—This organism did not rupture the imidazole ring. Of the histidine originally introduced, 86 per cent was recovered unchanged. Histamine and other imidazoles were not formed. Of the ammonia origin- _ ally introduced 22 per cent was removed by the microorganisms. Coli Lac. Aer.—During the first 7 days of incubation, this rapidly growing organism reduced the histidine concentration of the solution to 75 per cent of its initial value. There is practically no indication that a nuclear rupture occurred. The concentration of ammonia in the solution was greater after 7 days of incubation than it was at the outset of the experiment. Some of this ammonia must have been derived either from the disrupted histidine, or from the potassium nitrate. 148 Studies on Proteinogenous Amines. XII The second 7 day period of incubation reduced the histidine concen- tration to 45 per cent of its initial value. Some of this histidine was ap- parently converted into triamino compound and some into ammonia. Although there is no doubt, in this case, that a carboxylated amino com- pound was formed from the histidine, quantitative proof for the formation of a triamino compound is lacking because the decomposition continued beyond this stage with the production of ammonia. Histamine and other imidazoles were not formed. The increase in the ammonia figure might suggest that this organism was unable to utilize ammonia as a source of nitrogen, and that the imid- azole ring was ruptured to render nitrogen available. This was, however, not the case because this organism grows splendidly on a medium that contains only inorganic salts, glycerol, and NH,Cl. On this histidine- free medium the ammonia value drops from 36 to 25 cc. of 0.1 N HCl in the course of 2 weeks. .The rupture of the imidazole ring might, therefore, have been resorted to in an attempt to lower the hydrogen ion concen- tration of the cell protoplasm. Coli 88.—This organism did not rupture the eandeeels ring with the production of appreciable quantities of a carboxylated amino compound; but some of the histidine nitrogen must have been converted into ammonia because the concentration of ammonia at the end of the 14 day incubation period was greater than the initial concentration. Of the histidine origin- ally introduced, 83 per cent was recovered hiieuanics Histamine and other imidazoles were not formed. Coli 51.—The results obtained with this organism were so nearly like those obtained with Coli 88 that a detailed discussion seems superfluous. In this case, however, the ammonia consumption was more rapid than its production so that the initial ammonia value was reduced by 6.2 per cent. Bacillus coli communis. Of the organisms investigated, five belong to this group. Their behavior on our synthetic medium is summarized in Table IV. Coli cystitis.—This is the organism that was used in our earlier work. Since its behavior has been described in detail in our previous articles, it seems unnecessary to do more here than to call attention to the fact that it has lost none of its power of decarboxylation. In this experiment 57 per cent of the histidine originally introduced was converted into histamine as compared to a 50 per cent conversion obtained by us some 2 years ago. Coli Wk.—This organism ruptured the imidazole ring to a considerable extent, apparently with the formation of a carboxylated amino compound. Quantitative proof for the formation of a triamino compound is, however, lacking in this case. Of the histidine originally introduced 76 per cent was recovered. Histamine and other imidazoles were not formed. Of the M. T. Hanke and K. K. Koessler 149 ammonia originally introduced, 8.3 per cent was removed by the micro- organisms. Coli Hm—This organism seems not to have attacked the histidine at all because 96 per cent of this amino-acid was recovered. The nitrogen requirements were obviously supplied by the ammonia whose final con- centration was only 75 per cent of that originally introduced. Coli Cs—This organism did not rupture the imidazole ring with the formation of a carboxylated amino compound. Of the histidine originally introduced, 86 per cent was recovered. Histamine or other imidazoles were not formed. Of the ammonia originally introduced, 25 per cent was removed by the microorganisms. Coli K (White)—This organism ruptured the imidazole ring to the extent of 8 per cent, apparently with the formation of a triamino compound (see under Coli Y). Of the histidine originally introduced, 92 per cent was recovered. If we assume that the 8 per cent of histidine that dis- appeared in the course of 2 weeks incubation, was quantitatively converted if into triamino compound, the total amino nitrogen value of the liquid | should have been eight times 3 plus 92 equals 116 per cent, which compares very well with the 113.4 per cent actually obtained. Histamine and other imidazoles were not formed. A small amount of ammonia was produced either from the histidine or from the potassium nitrate. Bacillus lactis aerogenes. Of the organisms investigated, five belong to this group. Their behavior on our synthetic medium is summarized in Table V. Coli bovis 3—During the first 2 weeks of incubation, this organism reduced the histidine concentration of the solution to 92 per cent of its initial value and removed 17 per cent of the ammonia that was originally introduced. There is no indication that a nuclear rupture occurred. His- tamine and other imidazoles were not produced. After 30 days of incubation, the histidine concentration was reduced to 84 per cent of its initial value. During this period, 2.4 per cent of histidine was converted into histamine. In all, then, 86.4 per cent of histidine can be accounted for colorimetrically, after 30 days of incubation. After 14 days of incubation, 92 per cent of histidine was recovered. If we assume that the 5.6 per cent of histidine that disappeared during the second 16 day period was converted quantitatively into triamino compound, an amino nitrogen value of 5.6 times 3 plus 84 equals 100.8 per cent should have been obtained for the histidine fraction which compares very well with the 103.5 per cent actually obtained. This organism seems, there- fore, to have converted 5.6 per cent of histidine into triamino compound. | Of the ammonia originally introduced, 28.5 per cent was removed by the microorganisms. a — 1 ae 150 Studies on Proteinogenous Amines. Name of strain. Coli cystitis. Coli Wk. Coli Hm. Coli Cs. Coli K (white). Total color value of test solution as histidine dichloride. = (0.20 gm. = 100%.) 0.10 cc. = 11.2 mm. 0°20.“ = "22548 Color developed very rapidly. Zo 0.10 cc. = 8.4mm 0.20" “ =916-97* Match perfect. 84% 0.10 cc. = 10.38 mm. 0.20) “ = 2056 3% Match perfect. 103% 0.10 ce. = 10.0 mm. 0.20 (= 2070)" % Match perfect. 100% 0.10 cc. = 10.0 mm. 02208" |= 20°04 Match perfect. 100% Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. OF20 eT Sn Onan lI Match perfect. 40% 0. 10:ce: = 7-6 mm. OF 20° “a 152s Match good. 76% 0.10 cc. = 9.6mm. 0.20 = 19:2>" Match perfect. 96% 0.10 cc. = 8.6mm. 02201 =e Match perfect. 86% 0.10 ce. = 0.20. “ =18.3 % Match perfect. oF 92% * Colors matched against the (CR-MO) standard. 4.0mm. 9.2 mm. XII 4 | \ TABLE Iv—Bacill Unchanged histidine | (Van Slyke method) wit} 5 cc. of test solution. 0.85 cc. Nz at 20° ar 750 mm. 0.097 gm. _histidir dichloride. 48.5% 1.95 ee. Ne at 22° an! 748 mm. 0.220 gm. histidin dichloride. 110% .80 ec. Nz at 20° all 750 mm. | 0.206 gm. histidin dichloride. | 103% 1.57 ec. Ne at 23° an 750 mm. 0.1768 gm. histidin dichloride. 88.4% 2.07 ce. Ne at 28° an 750 mm. 0.2268 gm. histidin n dichloride. 113.4% — > -_ ae Pa ee | ; Histidine converted into histamine (colorimetric deter- |. Color value of ' histamine fraction.* mination). 105 cc. = 13.8mm. | 0.092 gm. of DRS a a histamine di- olor develops like chloride in that of histamine. the entire test solution. 57% of hista- mine present. i None. | None. None. M. T. Hanke and K. K. Koessler Histidine converted into histamine 0.10 n HCI neutralized (Van Slyke method) by NH: from with 5 cc. of test entire test solution. solution. 1.04 ec. Ne at 27 20° and 754]! Hence the = of 9 ce. mm, of 0.1 Nn NH; used 0.0964 gm. his-| by the microorgan- tamine di- isms. chloride. 59.8% 151 é communis. Reaction. Before incu- After incu- bation.| bation. 7.3 5.4 33 Hence the = of 3 cc. of 0.1 n NH; used by the microorgan- isms. 27 Hence the = of 9 ce. of 0.1 Nn NH; used by the microorgan- isms. 27 Hence the = of 9 cc. of 0.1 n NH3 used by the microorgan- isms. 37 Hence the = of 1 ce. of 0.1 n NH; was produced by the microorganisms. 7.3 7.3 5.2 5.8 152 Studies on Proteinogenous Amines. Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) Name of strain. 0.10 ce. = 10.6 mm. | 0.10 ce. = 9.2mm. OR20 Se —a2 oa ROB 2 Ope or ere Colibovis3, Match perfect. Match perfect. 14 days. 106% 92% 0.10 cc. = 9.0mm, | 0.10 cc. = 8.4mm. 0.202 s—el Ce Omen OF20) Gee ss Match perfect. Match perfect. 30 days. 90% 84% 0.10 ce. = 10.3 mm. |°0.10 cc. = 10.0 mm. 0220) —20) 6 mee 032000 Coli bovis 4. Match perfect. Match perfect. 108% 100% 0.10 cc. = 9.0mm. } 0:10 cc. = 8.0mm. 0620 = 185 0m 0:20) 16.0, Coli 80. Match perfect. Match perfect. 90% 80% 0.10 ce. = 9.3mm. | 0.10 cc. = 8.6mm. 020° <= 1826) 0.20°" =i Coli 84. Match perfect. Match perfect. 93% 86% 0.10 cc. = 10.0mm. | 0.10 ce. = 8.4mm. 0.20 “ = 20.0 “ 0.20 “=16.8 “ Coli 90. Match perfect. Match perfect. 100% 84% * Colors matched against the (CR-MO) standard. XII % ‘fe TABLE V—Baciilus Unchanged histidine — (Van Slyke method) with 5 ec. of test solution. 1.65 cc. N2 at 27° and 749 mm. 0.182 gm. histidine dichloride. 91% 1.85 ec. Ne at 21° ane 738 mm. : 0.207 gm. histidine dichloride. . 103.5% 1.93 ce. Ne at 26° and 745 mm. 0.212 gm. histidine dichloride. 106% 1.97 ec. Nz at 30° and 747 mm. 0.213 gm. histidine dichloride. ; 106.5% 1.79 cc. Nz at 25° and 751 mm. 0.2000 gm. histidine dichloride. 100% 1.51 ec. Nz at 21° and 752 mm. 0.172 gm. histidine dichloride. 86% M. T. Hanke and K. K. Koessler 153 See Histidine con- Reaction. Color value of ee eee verted ae ae ae b aC penealeer as 4 3 ~ mine (Van Slyke vy 3 from histamine fraction.* lonn method) with 5 ce. entire test solution. = pier ony of test solution. ation eee ce pH pH 29.7 Hence the = of 6.3 ec. of 0.1 n NH3| 7.3} 5.4 used by microor- ganisms. 0.40 ec. = 4.7mm. | 0.0039 gm. of 25.7 eo = 8.4 “ histamine di- Hence the = of 10.3 Color develops like chloride in ec. of 0.10 n NH; that of histamine. the entire used by the micro-| 7.3] 5.2 test solution. organisms. 2.4% of hista- mine present. 36.0 Hence appreciable ae quantities of NH; e seem not to have { been removed by the microorganisms. 74) Ot 29 Hence the = of 7 ce. of 0.1 Nn NH; used} 7.4] 5.4 by the microorgan- isms. 37 - Hence the = of 1 ce. f None. of 0.1 Nn NH; was| 7.4] 5.6 produced by the microorganisms. 37.5 Hence the = of 1.5 ce. None. of 0.1 n NH; was| 7.4] 5.6 produced by the microorganisms. 154 Studies on Proteinogenous Amines. XII Coli bovis 4.—This organism did not grow as well as the other coli and it seems not to have attacked appreciable quantities of either the histidine or the ammonia. Coli 80.—This colon bacillus ruptured the imidazole ring with the for- mation of carboxylated amino compounds. Of the histidine originally introduced, 80 per cent was recovered. Histamine and other imidazoles were not formed. The ammonia concentration was reduced by 19.5 per cent. Coli 84.—A nuclear rupture with the formation of carboxylated amino compounds is again indicated in this case. Of the histidine originally introduced, 86 per cent was recovered. Some ammonia was produced either from the disrupted histidine or from the potassium nitrate. Hista- mine or other imidazoles were not formed. Coli 90.—The liquid finally obtained in this case was free from carboxylated amino compounds other than histidine. Of this amino-acid, 84 per cent was recovered unchanged. Some ammonia was produced either from the disrupted histidine or from the potassium nitrate. His- tamine or other imidazoles were not formed. Bacillus acidi lactict. Of the organisms investigated, twelve belong to this group. Their behavior on our synthetic medium is summarized in Table VI. Ten of these organisms gave results that are qualitatively iden- tical; so they can be discussed collectively. Only Coli P-2-19 and P—5-19 require special consideration. All of the ten similar organisms ruptured the imidazole ring with the production of carboxylated. amino compounds. The disruption did not proceed quantitatively to the formation of triamino compound in any of these cases. Histamine and other imidazoles were not formed. Some of the organisms—Coli H, Schwartz, 52, 51 (white), and 74—produced ammonia, and others —Coli P-1-19, P-4—19, P-6-19, and I (K)—removed some from the solution. Coli P-2-19——The formation of a carboxylated triamino compound is again quantitatively indicated in this case. Of the histidine originally introduced, 95 per cent was recovered. If we assume that the 5 per cent of histidine that disappeared was quantitatively converted into a triamino compound, an amino nitrogen value of five times 3 plus 95 or 110 per cent should have been obtained which agrees almost exactly with the 110.2 per cent actually obtained. Histamine and other imidazoles were not formed. Of the ammonia originally introduced, 57 per cent was removed by the microorganisms. M. T. Hanke and K. K. Koessler 155 Coli P-5-19.—This interesting organism converted 14.5 per cent of the histidine originally introduced, into histamine. Since 79 per cent of histidine was recovered unchanged, 93.5 per cent of the original histi- dine can be accounted for colorimetrically. The amino nitrogen deter- mination on the histidine fraction gave a value of 100.5 per cent calcu- lated as histidine. Of this amino nitrogen only 79 per cent could have been derived from histidine. If we assume that the remaining 21.5 per cent of amino nitrogen was derived from the triamino compound, this would account for 7.17 per cent of histidine; 7.e., 21.5 divided by 3. Sum- ming up, then, we have ; 79.0 per cent as histidine. i445. “ <.-® histanme. 7.17 “ “ “© triamino compound. , 100.67 “ “ total recovery. which is a truly remarkable agreement. PARE I. The Products Formed from Histidine by the Action of Other Members of the Colen Typhoid Group. The organisms investigated were Bacillus enteritidis, Bacillus typhosus, Bacillus paratyphosus A (3 strains), Bacillus dysenterice Flexner, Bacillus dysenterie Morgan, Bacillus dysenterie Shiga, Bacillus fecalis alcaligenes J, and Bacillus fecalis alcaligenes III. The behavior of these organisms on our synthetic medium is summarized in Table VII. B. enteritidis 228 —This organism grew very well on the liquid medium. There is no evidence of a rupture of the imidazole ring after 7 days of incubation although 20 per cent of the introduced histidine disappeared during this time interval. Of the ammonia originally introduced 32 per cent was removed by the microorganisms. Histamine and other imidazoles were not formed. The attack on the remaining histidine was so intense, during the second 7 day period, that 40 per cent of that amino-acid disappeared. A nuclear rupture occurred with the formation of some carboxylated amino compound and considerable ammonia. Histamine and other imidazoles were not formed. B. typhosus —This organism grew poorly on our medium. Histamine and other imidazoles were not formed. Very little acid was produced. Of the ammonia originally introduced, 80 per cent was recovered. The results on the histidine fraction suggest that a small amount of histidine was deaminized because the colorimetric value for histidine ran higher than the amino nitrogen value. The conversion was too slight, however, to be of particular significance. 156 Studies on Proteinogenous Amines. Name of strain. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. O80 © S200 & Coli B. Match perfect. 100% 0.10 cc. = 10.0 mm. 0.20 120 nO vere Coli H. Match perfect. 100% 0.10 ce. = 10.0 mm. 0220. = )2020 Fr Coli P-1-19. Match perfect. 100% 0.10 cc. = 10.0 mm. 0:20 = 2002 Coli P-2-19. Match perfect. 100% 0.10 ce. = 9.0 mm. 0220) <2 = 1870) = Coli P-4-19. Match perfect. 90% 0.10 cc. = 9.8mm. 0:20 “-= 19.4,“ Coli P-5-19. Match perfect. 98% 0.10 cc. = 9.4mm. 0.20 “ =18.9 “ Coli P-6-19. Match perfect. 10.0 mm., Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. = 9.2mm. O20 TS eaaen Match perfect. 0.10 ce. = 8.0mm. OF20F GRO Match perfect. 0.10 ce. = 9.1mm. 0.20: Sees Match perfect. 91% 0.10 cc. = 9.5mm. O20 ee — 98 Oe Match perfect. 95% 0.10 ce. = 8.4mm. 0:20 S =167 Match perfect. 84% 0.10 ees = 7. 9anm-e 0.20 “ = 15.8 “ Match perfect. 94% 0.10 cc. = 9.4mm. 0.20 “ = 18;89— Match perfect. 94% * Colors matched against the (CR-MO) standard. XII 1.80 cc. Ne at 30° and 1.70 ce. Nz at 21° and 1.99 ec. Ne at 25° and 1.85 cc. Nz at 22° and . TABLE VI—Bacillus Unchanged histidine (Van Slyke method) with | 5 cc. of test solution. 1.90 cc. Nz at 28° and 747 mm. 0.208 gm. histidine dichloride. 104% _ 747 mm. 0.1945 gm. histidine dichloride. 97.2% 744 mm. 0.192 gm. histidine dichloride. 96% 745 mm. iy 0.2204 gm. histidine | dichloride. | 110.2% 747 mm. 0.209 gm. histidine dichloride. 104.5% 1.77 ec. Nz at 20° and 745 mm. i 0.201 gm. histidine” dichloride. 100.5% 1.81 ec. Nz at 21° and © 746 mm. ; 0.205 gm. histidine © dichloride. 102.5% M. T. Hanke and K. K. Koessler 157 t lactici. sys: Histidine converted Reaction. Histidine converted into histamine 0.1 N HCI neutralized Color value of into histamine wv. 2 oh: ; : : . an Slyke method) by NH: from histamine fraction.* (colorimetric deter- ab Ga (it tank Peter Grlaton. mination). Sinan Before | After incu- | incu- bation.| bation. ce. pH pH None. Te ORS 38 Hence the = of 2 ce. None. of 0.1 N NH; was| 7.4 produced by the } microorganisms. 31 Hence the = of 5 ce. None. of 0.1 n NH3 was| 7.4 used by the micro- organisms. 15.5 if Hence the = of 20.5 cee Ot OLN Ns) fa4 | oan was used by the microorganisms. 32.5 Hence the = of 3.5 ce. of 0.1 Nn NH3} 7.4] 5.9 was used by the microorganisms. or for) or ~J 10 cc. = 7.0mm. | 0.0235 of hista-| 0.35 cc. Ne at 35.5 mo “ = 14.0 “ mine dichlo-| 22° and 744| Hence the = of 0.5 Jolor develops like rideintheen-| mm. ec. of 0.1 n NH; that of histamine. tire test solu- | 0.0317 gm. his-| was used by the} 7.4] 6.2 tion. tamine di-| microorganisms. 14.5% of hista- chloride. mine present. 19.6% 27 Hence the = of 9 ce. None. of 0.1 n NH3 was| 7.4] 6.0 used by the micro- organisms. nn , ‘ Name of strain. 158 Studies on Proteinogenous Amines. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) Coli 1 (K). 0.10 ce. = 9.0mm. 0:20) 1S se Match perfect. 90% Coli Schwartz. Coli 52. Coli 51 (white). Coli 74. 0.10 ec. = 10.0 mm. (OO) YN) Match perfect. 100% 0.10 cc. = 10.0 mm. 0:20). 822070 = Match perfect. 100% 0.10 cc. = 10.0 mm. 05205 -9— 720505. Match perfect. 100% 0.10 cc. = 10.0 mm. 0.20 “ = 20.0 “ Match perfect. 100% Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) Ost0Kce: 0:20) 4; =16.08 "4 Match perfect. OMOices—. SP 9mm, F200 s=17..8" Match perfect. 89% 0.10 cc. = 9.0mm. 0:20: i =418.0) Match perfect. 0.10 ce: = 9.3 mm. 0320 —sIS6a Match perfect. 93% 0.10 cc. = 8.8mm. 05205 += sO nee Match perfect. 88% XIT 8.0 mm. Unchanged histidine (Van Slyke method) 5 cc. of test solution. 1.71 ce. Ne at 22. and 745 mm. 0.1925 gm. histidi dichloride. 96.2% 1.89 cc. Ne at 28° an 749 mm. 0.207 gm. histidi dichloride. 103.5% .87 cc. Nz at 27° ani 746 mm. 0.205 gm. histidin dichloride. 102.5% .89 cc. Nz at 28° ani 749 mm. 0.207 gm. _ histidin dichloride. 103.5% .83 cc. Ne at 28° an 747 mm. 0.200 gm. histidin. dichloride. 100% M. T. Hanke and K. K. Koessler 159 cluded. d ee converted Ape ee Reaction. Color value of into histamine into histamine as NeuLranze i - : . : (Van Slyke method) by NH: from i aniine fraction.” (colorimetric deter- with Ff ce. of test entire test solution. mination). solution. Histidine converte Before | After incu- | incu- bation. | bation. ccs pH pH a2. 9 Hence the = of 3.5 None. ce. of 0.1 n NH3| 7.3] 5.5 ’ was used by the microorganisms. ee 4 38.5 . Hence the = of 2.5 + ce. of 0.1 n NH3 . None. was produced by 1) the microorgan- 7.4} 5.6 39 Hence the = of 3 cc. None. of 0.1 Nn NH; was| 7.4| 5.4 produced by the microorganisms. 38 Hence the = of 2 ce. None. of 0.1 N NH; was| 7.4] 7.0 produced by the | microorganisms. 37 Hence the = of 1 ce. None. of 0.1 n NH3 was| 7.4} 5.5 produced by the i microorganisms. \ ante 160 Studies on Proteinogenous Amines. Name of strain. Total color value of test solution of histidine dichloride.* (0.20 gm. = 100%.) 0.10 cc. = 8.6mm. Bacillus enteritidis | 0.20 “ =17.1 “ 228, Match perfect. 7 days. 86% 0.10 cc. = 4.0mm. O52 0a Oe 14 days. Match poor. Color too yellow. 40% 0.10 cc. = 9.6mm. Bacillus typhosus. Bacillus paratyphosus A 3. Bacillus paratyphosus A 4, 7 days. 14 days. Bacillus paratyphosus A (K). Oe SS sei & Match perfect. 96% 0.10 ce. = 9.3mm. O20 a Sa Gmc Match perfect. 93% 0.10 ce. = 8.2mm. 0-20) =16 Ae Match perfect. Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. OFZ0)R-- = Toso). = ’ Match perfect. 80% OF10ce. = 4.0mm: A)? () eee 1) ee Match good. 40% 0210ce-= 8.5mm. OVA Sra © Match perfect. 85% 0.10 ec. = 0:20 “ = 16.7 °% Match perfect. 83% OMOkce—2 S720 O320 7h — sll Ge aes Match perfect. 82% 0.10 cc. = 4.0mm. 0220 — Th Oise Mateh good. 40% 0.10 ce. = 10.0 mm. 0.20) = =.2007 = Color too pink and de-| Color too pink and de- velops too rapidly for histamine. <~/0 0.10 cc. = 4.0mm. Of 20 ie — es One Match good. 40% 0.10 ec. = 10.0 mm. 0.20.“ =20:0 = velops too rapidly for histamine. 100% * Colors matched against the (CR-MO) standard. XII 8.0 mm. 8.3 mm. TABLE ViI—Col Unchanged histidine (Van Slyke method) 5 ec. of test solution. | 1.46 cc. N2 at 29° a 746 mm. 0.1584 gm. histidi dichloride. 79.2% .25 ec. Ne at 238° a 745mm. 0.140 gm. histidi dichloride. .35 ec. Ne at 21° a dichloride. 16% 1.47 ec. Na at 20° 738 mm. 0.1655 gm. histidine dichloride. 82.7% .73 cc. Nz at 28° an 746 mm. 0.1886 gm. histidi dichloride. 94.3% 1.95 ce. Ne at 24° an 753 mm. 0.2195 gm. histidi dichloride. 109.7% 0.95 ec. Ny at 24° am 753 mm. 0.107 gm. histidin dichloride. 53.5% M. T. Hanke and K. K. Koessler “T6k hoid Group a - “verted ae Reaction. olor value istamine . : < : 0.1.N HCl neutralized by NH; ae ypionic agi pa to Bhi ee pins from entire test solution. Before After ie 5 cc. of test incuba- incuba- solution. tion. tion. ee ee | Ce en lor pH 24.5 a Hence the = of 11.5 ec. of 0.1 N ase. a NH; used by the microorgan- i Ae isms. : 39.5 . Hence the = of 15 cc. of 0.1 N NH; | None. None. None. was produced during thesecond} 7.3 5.2 7 day period. : 29 . Hence the = of 7 ce. of 0.1 N NH; None. None. was removed by the microor- ie 6.6 | ganisms. i Ce ee z 31.5 ; , P Hence the = of 4.5cc. of 0.1N None. None. None. NH; wasremoved by themicro-| 7.3 “a organisms. ee Ee Sa) a a Se Ee ed ee . 39 Hence the = of 3 cc. of 0.1N NH; None. | None. None. was produced by the microor- 7.3 6.4 ganisms. 38 Hence the = of 1 cc. of 0.1 N NH; Rika. Nata was removed during the second 73 5.4 7 day period. S025 Hence the = of 5.5 ce. of 0.10 N ace. ane NH; was removed by the micro- 73 6.8 organisms. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 162 Studies on Proteinogenous Amines. Name of strain. Bacillus dysenterie Flexner. Bacillus dysenterie Morgan. Bacillus dysenteriae Shiga. Bacillus faecalis alcaligenes a wl, Bacillus fecalis alcaligenes III, 3 days. 14 days. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10 cc. = 10.0 mm. 02205 = 205055 Color too pink and de- velops too rapidly for histamine. 100% 0.10 ce. = 9.5mm. UFO 2 =i19.0 © Match perfect, color develops rapidly. 95% 0.10 cc. = 11.8 mm. 05200 s— 222 0mm Color too pink and de- velops too rapidly for histamine. 118% 0.10 ce. = 9.7mm. 0220, = 1974, Color slightly too pink. 97% 0.10 ce. = 9.0mm 0220. as n0 Match good. 90% 0.10 ce. = 9.2mm. 0.205 -=18e0) << Match perfect. 92% Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 cc. = 9.6mm. O720/ °° = TORR Ss Color too pink and de- velops too rapidly for histamine. 96% OMOIces — 7.6 mm: OF200 ae—s1 550) Match perfect, color develops rapidly. 78% 0.10 ce. = 9.3mm. O220 F = 1ss08 Color too pink and de- velops too rapidly for histamine. 93% 0.10 ce. = 0:20) = 14.0 Color slightly too pink for histidine. 70% 0.10 cc. = 7.0mm. 0.20 “ =.14:0 Match perfect. 7.0mm. 70% 0.10 cc. = 7.0mm. 0.20) = 142 0= Match perfect. 70% XII TABLE V. Unchanged histidine | (Van Slyke method) wit 5 ec. of test solution. | 0.91 cc. Nz at 22° a 745 mm. 0.1025 gm. histidi dichloride. 51.2% 1.15 ce. Ng at 22° a: 743 mm. 0.129 gm. _histidiy dichloride. 65% 1.24 ec. Ne at 19° an 748 mm. 0.1423 gm. histidine dichloride. * 71.1% 0.80 ce. Nz at 23° aj 749 mm. 0.090 gm. histidir dichloride. 45% 1.30 cc. Ne at 23° a 750 mm. 0.1464 gm. histidir dichloride. 73.2% 1.17 cc. Ne at 28° an 749 mm. 0.1317 gm. histidir dichloride. 66% cluded. i Histidine con- (by differ- ence). ).033 gm. { || verted into | imidazole ropionic acid ie 27.1% | Color value of histamine fraction.* None. None. M. T. Hanke and K. K. Koessler 163 * Histidine con- ; verted into Reaction. istamine : 0.1N HCl tralized by NH needa from entire test solution. Before After 5 cc. of test incuba- incuba- solution. tion. tion. ce pH pH 34.5 Hence the = of 1.5 ce. of 0.1 N None. NH; was removed by the micro-| 7.3 el organisms. 35.5 Hence the = of 0.5 cc. of 0.1 N None NH; was removed by themicro-| 7.3 7.0 organisms. 33 Hence the = of 3 ce. of 0.1 N NH; was removed by the microor- 7.3 Coal None. . gantsms. Beas Hence the = of 3.5 ce. of 0.1 N None NH; was removed by the TiS 7.4 microorganisms. 32.5 Hence the = of 3.5 ce. of 0.1 N None NH; was removed by the mi- Y fee 7.4 croorganisms. 33 Hence the = of 3 cc. of 0.1 N NH; None was removed by the microor- (ec Take ganisms. 164 Studies on Proteinogenous Amines. XII B. paratyphosus A 3.—This organism grew very poorly on our medium, apparently because an available carbon source was lacking. Both the glycerol and the histidine remained largely unattacked. Histamine, other imidazoles, and acid were not formed. Of the ammonia originally intro- duced, 13 per cent was removed by the bacilli. B. paratyphosus A 4.—This organism grew far better than B. para- typhosus A 3. Some of the glycerol was metabolized with the production of acid. The imidazole ring was ruptured to a considerable extent with the formation of carboxylated amino compounds and ammonia. Hista- mine and other imidazoles were not formed. B. paratyphosus A (K).—This organism deaminized histidine with the formation of imidazole acetic, propionic, lactic, or acrylic acids because the color produced by the histidine fraction was far too red for histidine and the amino nitrogen value was much lower than the colorimetric value. If we assume that the amino nitrogen was derived exclusively from histidine, the presence of a maximum of 53.5 per cent of that substance is indicated. Colorimetrically this would give a reading of 5.835 mm. (CR — MO) for 0.10 cc. of the diluted test solution. The reading obtained was 10.0 mm. (CR — MO) for 0.10 ce. The difference between these two readings— 4.65 mm. (CR — MO) or5.12 mm. (CR)&’—must have been due to either imidazole acetic, propionic, lactic, or acrylic acids. Which of these acids was present was not determined; but the excess color value was calculated as imidazole propionic acid using the table that was previously published for that substance. The calculations show that 0.034 gm. of imidazole propionic acid was present; hence 27.7 per cent of the histidine originally introduced was deaminized. This organism grew poorly, used glycerol to only a small extent, pro- duced very little acid, and removed the equivalent of 16 per cent of the ammonia originally introduced. Histamine was not formed. B. dysenteriae Flexner, Morgan, and Shiga.—The three classes of dysen- tery bacilli can most advantageously be discussed together because, although the results are quantitatively somewhat different, they are qualitatively identical. In every case the color obtained with the histi- dine fraction was too red to have been due only to histidine. The color- imetric values were higher than the amino nitrogen values on this fraction; hence histidine was not the only imidazole present. When the discrep- ancies between the values obtained colorimetrically and by the amino nitrogen method are calculated as imidazole propionic acid, the presence of 0.033 gm. (27.1 per cent), 0.01 gm. (8.1 per cent), and 0.016 gm. (13 per cent) of this substance is indicated for Flexner, Morgan, and Shiga respectively. These organisms grew poorly, used glycerol only to a small extent, produced practically no acid, and removed very little ammonia from the solution. Histamine was not formed. § The (CR) standard is 19 as intense as the (CR — MO) standard. M. T. Hanke and K. K. Koessler 165 B. fecalis alcaligenes I—The color obtained with the histidine fraction was too red for histidine and the amino nitrogen value was 25 per cent lower than the colorimetric value. This suggests the presence of imidazole acetic, propionic, lactic, or acrylic acids. When the discrepancy between the values obtained colorimetrically and by the amino nitrogen method are calculated as imidazole propionic acid, the presence of 0.018 gm. (14 = per cent) of this substance is indicated. This organism grew so poorly that the medium was almost clear after 14 days of incubation. Some of the bacilli were, nevertheless, alive at the end of this period. Little or no use was made of the glycerol, and acid was not produced. Of the ammonia originally introduced, 90 per cent wasrecovered. Histamine was not formed. B. fecalis alcaligenes III—This organism grew about as poorly as B. fecalis alcaligenes I. During the first 3 days, 30 per cent of the histidine was removed. After that there was practically no change in the histidine content of the liquid. Imidazole propionic acid and histamine were not formed. Of the ammonia originally introduced, 90 per cent was recovered. The slight production of alkalinity during the las: 11 days of incubation may have been caused by autolytic changes because apparently the active life of the organisms came to an end after 3 days of meager growth. PART Il. On the Products Formed from Histidine by the Action of Organisms Other than Those Belonging to the Colon Typhoid Groups. The organisms investigated were Bacillus mucosus capsulatus (2 strains), Bacillus bifidus, Bacillus influenze, Bacillus proteus vulgaris (2 strains), Bacillus cloacew, Streptococcus hemolyticus, (2 strains), Pneumococcus Types I, IJ, III, and IV, and Bacillus tuberculosis (5 strains). The behavior of these organisms on our synthetic medium is summarized in Table VIII. B. mucosus capsulatus—An excellent growth was obtained. This organism attacked the histidine immediately and so effectively that 75 per cent of it was destroyed in14 days. The histidine seems to have been attacked because of its nitrogen. The following facts led us to this con- clusion: 1. The ammonia concentration remained practically unchanged through- out the entire course of the experiment in spite of the fact that an unusu- ally abundant growth was obtained. The nitrogen requirements of the organism must, therefore, have been obtained from the histidine. 2. Although the histidine was largely removed from the solution, none of its nitrogen appeared as ammonia and only a very small amount of it appeared in the amino condition. During the first few days, all of the nitrogen derived from the disrupted histidine was consumed. Toward the 166 Studies on Proteinogenous Amines. - TL “* 4 Name of strain. “‘mucosus S., Bacillus capsulatus LL. 3 days. 14 days. Bacillus MUCOSUS capsulatus L. S. plus 0.10 gm. leu- cine, 14 days. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) * 0.10 cc. = 8.2mm. 0320) “Ss IGE4 ee Match perfect. 82% 0:10 ce> =" 228 mm: 0220) oro Match fair. 28% 0.10 cc. = 8.0mm. 0.20 * =16i04% Match good. 80% 0.10 cc. = 8.4mm. Bacillus mucosus)| 0.20) = NG.8 capsulatus 27 (S), | Match perfect. 3 days. 84% 0.20 ce. = 5.8mm. O40 = Ee Gr a 5 days. Match good. 297% | 0.50 cc. = 2.4mm. Color brownish yel- 8 days. low. Match poor. 4.8% 0.10 ec. = 10.0 mm. Bacillus bijfidus, 14) 020°“ = 2070" “ days. | | { | 1 | Match perfect. 100% *Colors matched against the Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. O20 = 1455. Match perfect. 73% 0.20 ec. = 0.40 “ =10.2 “ Match good. peu 0.10 ce. = 7.5mm. OF 20 a5 0a Match good. 15% OF10rees = 7) mam 0220) a1 5" Ome Match perfect. 15% 0.20 ce. = 4.9 mm. OO kt 9) See Match good. 23% 0. 50:cc,, = 2.0 mm: L003 a) er 40 Match good. 4.0% 0.10 cc. = 8.7 mm. 0.20) oy eee Match perfect. 87% (CR-MO) standard. 7.3 mm. 5.1mm. Unchanged histidine (Van Slyke method) with 5 ec. of test solution. 1.33 cc. Ne at 18° and 750 mm. 0.1535 gm. hisudaag dichloride. 76.8% 0.75 ec. Nz at 25° and 745 mm. 0.083 gm. hiseeiaa dichloride. 41.5% 1.46 ce. Ne at 21° and 742 mm. 0.1655 gm. histidine dichloride. ~ 82.7% 0.46 ce. N, 752 mm. 0.052 gm. histidine dichloride. 26% cl a at 23° and 0.80 cc. Nz at 21° and 737 mm. 0.0894 gm. histidine dichloride. ’ 44.7% ‘ 0.75 ce. Nz at 21° and) 737 mm. 0.0888 gm. histidingj dichloride. 41.9% 1.55 ec. Nz at 23° and 744 mm. 0.173 gm. dichloride. 86.5% histidine M. T. Hanke and K. K. Koessler 167 laneus Organisms. ee Histidine con- . Histidine con- verfadianto Reaction. verted into 5 - Pre Color value histamine - dazol 5 : 0.1N HCl tralized by NH propionic acid | faction” | method) with fromentize testsolution. | Bo) after Pics) 5 cc. of test incuba- incuba- : solution. tion. tion. cc si. ee 36.7 Hence the = of 0.7 cc. of 0.1 N None. None. None. NH; was produced. 7.3 6.2 36.7 An ammonia equilibrium was None. None. None. established after 7 days of (oe 6.8 growth. 34 Hence the = of 2 cc. of 0.1 N NH; None. None. None. was removed by the microor- ipo ganisms. 0.0365 gm. 38.5 Hence the = of 2.5 ce. of 0.1 N 29.7% None. None. NH; was produced by the es microorganisms. 44 Hence the = of 8 ce. of 0.1 N NH; None. None. was produced by the microor- 7.3 ganisms. 49 Hence the = of 13 cc. of 0.1 N None. None. NH; was produced by the 7.3 microorganisms. 32.5 Hence the = of 3.5 cc. of 0.1 N None. None. NH; was removed by themicro-| 7.3 organisms. 168 Studies on Proteinogenous Amines. XII TABLE VIII— Total color value of Color value of histidine : test solution as histidine fraction as histidine wen ene pista | Name of strain. dichloride.* dichloride. * ae ae 8 ate i a) wi ‘ (0.20 gm. = 100%.) (0.20 gm. = 100%.) 0. OS 8 ee eee 0.10 cc. = 10.0mm. | 0.10 cc. = 9.0mm. | 1.55 cc. Nz at 32° and Bacill P OR20) = 2000 OF208 = 1870s 746 mm. Beeline yp Uyiuence, Match perfect. Match perfect. 0.165 gm. histidine 14 days. ; dichloride. 100% 90% 82.5% 0.10 cc. = 7.6mm. | 0.10 cc. = 6.9mm. |} 1.13 cc. No at 21° and Pilins lnroieus vat C20 © 2 aE da? 42 754 mm, | Noes Match good. Match good. 0.1294 gm. histidine | garis 186, dichinua ies. ichloride. 76% 69% 64.7% 0.20 cc. = 9.0mm. | 0.20 cc. = 9.0mm. | 0.78 cc. Ne at 21° and: OF30 Soa ieee 02305 = 3rbe 752 mm. 14 days. Color too yellow for| Match fair. - Color| 0.089 gm. histidine — histidine. slightly too yellow. dichloride. 45% 45% 44.5% 0.10 cc. = 8.1mm. | 0.10 cc. = 7.4mm. | 2.20 cc. Ne at 20° and | Bacillus proteus vul-| 0.20 “ =16.2 “ 0.20: * = 14.7 “ 738 mm, garis A. I. K., 14] Match:perfect. Match perfect. 0.2475 gm. histidine days. dichloride. 74% 123.7% 0.10 cc. = 10.0mm. | 1.85 cc. Ne at 26° and 0520) — 200 as 745 mm. Match perfect. 0.2036 gm. histidine dichloride. 100% 101.8% 0.10 cc. = 10.0 mm. 0.20. = 2020-" Match perfect. Bacillus cloace I, 14 days. 100% 0.10 cc. = 10.0 mm. | 0.10 ce. = 8.5mm. | 1.51 cc. Na at 22° andy 02200 e200) O20 — 10 748 mm. Streptococcus hemo- aise ak \T4 dave: Match perfect. 0.1705 gm. histidine Match perfect. dicho 100% 85% 85.2% 0.10 cc. = 10.0mm. | 0.10 cc. = 8.4mm. | 1.65 cc. Na at 28° and Streptococcus hemo-| 0.20 “ = 20.0 “ 0.20°:" = 16-358 746 mm. lyticus II KR, 14] Match perfect. days. Match perfect. 0.180 gm. histidine dichloride. 84% 90% 100° M. T. Hanke and K. K. Koessler 169 lontinued. Histidine con- F - verted into Reaction. Color value | histamine 0.1 HCl neutralized by NHs eee eee eee f histamin an Slyke : : ; propionic acid| ° earned Bis 4) in th from entire test solution. Before “After _ (by differ- 5 cc. of test incuba- incuba- ence). solution. tion. tion. cc. pH pH 31.5 Hence the = of 4.5 ce. of 0.1 N NH; wasremoved by themicro-| 7.3 7.3 organisms. None. 35 Hence the = of 1cc. of 0.1 N NH; was removed by the microor- 7.3 leae ganisms. None. None. Hence the = of 4 cc. of 0.1 N NH; was produced by the microor-| 7.3 (ie: ganisms. None. None. ee SS 28 Hence the = of 8cc.of 0.1N NH; was removed by the microor- to 5.2 ganisms. None. None. 30 Hence the = of 6 ce. of 0.1 N NH; was removed by the microor- 7.3 ganisms. = on None. None. 32.5 Hence the = of 3.5 cc. of 0.1 N NH; was removed by the micro-| 7.3 organisms. None. None. “J ow EE OO | —_<_<__ | —___ 33 Hence the = of 3 cc. of 0.1 N NH; was removed by the microor- 7.3 7.4 ganisms. None. None. 170 Studies on Proteinogenous Amines. XII TABLE VitI— Name of strain. Pneumococcus Type I, 14 days. Pneumococcus Type I plus 0.10 gm. leucine, 14 days. Pneumococcus Type II, 14 days. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ec. = 10.0 mm. O=20 <2 00 eee Match perfect. 100% 0.10 ee. = 10.0 mm. 0:20) = 20500a Match perfect. 100% 0.10 ce. = 0:20. “*=<19708 Match perfect. 95% Pneumococcus Type II plus 0.10 gm. leucine, 14 0.10 cc. = 10.0 mm. 0220) ** = 20809 ¢ Match perfect. Penumococcus Type III, 14 days. Pneumococcus Type III plus 0.10 gm. leucine, days. 100% 0.10 cc. = 9.5mm. OF20) S19 uO Match perfect. 95% 0.10 cc. = 10.0 mm. 0220" “= 2070! -* Match perfect. 9.5 mm. Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 cc. = 02209" = 16:8 Match perfect. 84% 0.10 cc. = 8.6mm. OBZ 0 Re — elie an Match perfect. 86% 0.10 cc. = 7.8 mm. OF2 0 —al role Match perfect. 78% 0.10 cc. = 8.9mm. OF20% Ss. = 78 * Match perfect. 89% 0.10 ce. = 7.9mm. (LAO) ere cmealisyeesya Match perfect. 79% 0.10 cc. = OL200SS = 170) Match perfect. Pneumococcus Type IV, 14 days. Match perfect. 97% 14 days. 100% 85% 0.10 cc. = 9.7 mm. 0.10 cc. = 8.0mm. OP20 ie eal Olean a 0820) 5 = GOs Match perfect. 80% 8.4mm. 8.5 mm. q Unchanged histidine 4 (Van Slyke method) with 5 ec. of test solution. 1.48 ec. No at 21° and 750 mm. 0.1684 gm. histidine dichloride. 84.2% 1.54 cc. Ns at 29° and 746 mm. 0.167 gm. histidine dichloride, 84% 1.38 ec. Nz at 32° and 748 mm. 0.1475 gm. histidine dichloride. TA% 2.05 ce. Ne at 32° and 746 mm. 0.2185 gm. histidine dichloride. 109.2% 1.65 cc. Nz at 32° and 748 mm. 0.1765 gm. histidine dichloride. 88.2% 1-81 cc. Ne at 32° and © 748 mm. 0.1935 gm. histidine dichloride. 96.8% 1.47 cc. Nz at 31° and 748 mm. 0.158 gm. histidine dichloride. 79% M. T. Hanke and K. K. Koessler eel Histidine con- un - soe into olor value istamine : - . 0.1 N HCI neutralized by NHs eee Bi eee from entire test solution. 5 cc. of test solution. Histidine con- verted into imidazole propionic acid (by differ- ence). cc. 33 Hence the = of 3 cc. of 0.1 N NH; was removed by the microor- ganisms. F 33 Hence the = of 3 ce. of 0.1N NH; was removed by the microor- ganisms. 33 Hence the = of 3 cc. of 0.1N NH; was removed by the microor- ganisms. 32 Hence the = of 4cc. of 0.1 N NH; was removed by the microor- ganisms. 32 Hence the = of 4c. of 0.1 N NH; was removed by the microor- ganisms. 32 Hence the = of 4ce. of 0.1 N NH; was removed by the microor- ganisms. : 33 Hence the = of 3 cc. of 0.1 N NH; was removed by the microor- ganisms. 172 Studies on Proteinogenous Amines. Name of strain. Pneumococcus Type IV plus 0.10 gm. leucine, 14 days. Bacillus tuberculosis H. Sp., 45 days. Bacillus tuberculosis 1305, 45 days. Bacillus tuberculosis 3161, 45 days. Bacillus tuberculosis O. H., 45 days. Bacillus tuberculosis (Novy) Ki, 5 days. 45 days. Total color value of test solution as histidine dichloride.* (0.20 gm. = 100%.) O10%ce: 9.7mm. OF20) << 1954S Match perfect. 97% 0.10 ce. = 10.0 mm. OR20 e200 eee Color too red for his- tidine. 100% 0.10 ec. = 10.0 mm. O20 5 — 2070 hae Color too red for his- tidine. 100% 0.10 ce. 9.6mm. OF20) a — ORS Color too red for his- tidine. 96% 0.10 ec. = 10.0 mm. 0:20 = 2070 Color too red for his- tidine. 100% 0.10 cc. = 5.8mm. 0.203 = 41-5 Match fair. Color too yellow. 58% 1.00 ce. = 15.0 mm. Color brown. Match poor. Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. = Pe OMinis rOs20IT* = 16.0725 Match perfect. 79% 0.10 ce. = 9.0mm. ORZ0 ee — 1820) se Color too red for his- tidine. 90% 0.10 dec. = 9.5mm. OF20 — 19208 Color too red for his- tidine. 0.10 cc. = 8.1mm. O20 e— Gee Color too red for his- tidine. 81% 0.10 cc. = 9.5mm. 0220 —19 0S Color too red for his- tidine. + 0:20.“ .= 9.3 Match fair. too yellow. 47% 1.00 cc. = 12.0 mm. Color brown. Match poor. 907 12% XII 1.23 ec. Ne at 22° and 0.95 cc. Nz at 20° and TABLE VIII— Unchanged histidine (Van Slyke method) “a | 5 cc. of test solution. 1.60 cc. Nz at 31° and 748 mm. 0.172 gm. histidine dichloride. 86% 756 mm. 0.140 gm. histidine dichloride. 70% 750 mm, 0.1085 gm. histidine dichloride. 54.2% 0.95 cc. Ne at 20° and 752 mm. 0.109 gm. histidine dichloride. 54.5% 1.09 ec. Ne at 21° and 761 mm. 0.126 gm. histidine dichloride. 63% 1.27 cc. Ne at 18° and 740 mm. 0.1446 gm. histidine dichloride. 72.38% 0.81 cc. Ne at 21° and 762 mm. 0.0936 gm. histidine dichloride. 46.8% M. T. Hanke and K. K. Koessler tis Histidine con- | : Reaction i verted into verted into : | imidazole helenae ee, 0.1 HCl neutralized by NHs ee Se ERTS ; ae eae faaton:* method) with from entire test solution. Before After ene) 5 ce. of test incuba- | incuba- solution. tion. tion. ce. pH pH 32 Hence the = of 4cc. of 0.1 N NH; None. None. was removed by the microor- Cee 7.0 ganisms. 27 Hence the = of 9cc. of 0.1 N NH; None. None. was removed by the microor- 7.3 ao ganisms. 28.5 Hence the = of 7.5 ce. of 0.1 N None. None. NH; wasremoved by themicro-| 7.3 tol organisms. 31 Hence the = of 5-ce. of 0.1 N NH; None. None. was removed by the microor- hee Ho? ganisms. 32.5 Hence the = of 3.5 cc. of 0.1 N None. None. NH; was removed by themicro-| 7.3 6.9 organisms. 38.5 Hence the = of 2.5 cc. of 0.1 N None. None. NH; was produced by the mi- bo Ula” croorganisms. 37.5 Hence the = of 1 ce. of 0.1 N NH; None. None. was removed during the 40 day (hoc 6.8 period. 174 Studies on Proteinogenous Amines. XII end of the 2 week period non-volatile amino nitrogen was liberated slightly faster than it was being consumed. 3. When leucine was added to the medium, along with the histidine, the growth of the organisms was augmented; but it was the leucine nitrogen that was removed from the solution, the histidine being largely unat- tacked. That the leucine nitrogen was removed is proved by the fact that the amino nitrogen value on the histidine fraction ran only slightly higher than the colorimetric value for histidine. One might be led to assume that this organism cannot use ammonia to satisfy its nitrogen requirements. Thisis, however, not the case because a very good growth is obtained in a medium containing only glycerol and ammonium chloride together with the usual inorganic salts. In this case considerable ammonia is removed by the microorganisms. It seems that this organism used amino nitrogen in preference to ammonia nitrogen. Histamine and other imidazoles were not formed. In addition to the above stock culture, two strains of Bacillus mucosus capsulatus were isolated in this laboratory, from the feces of two cases of colitis. They were quantitatively almost identical in their action. The horizontal Columns 4, 5, and 6 contain a summary of the behavior of one of them. An excellent growth was obtained. The amino nitrogen value of the histidine fraction, obtained after 3 days of incubation, shows that a maxi- mum of only 26 per cent of unchanged histidine could have been present. The color obtained with this fraction was too red to have been exclusively due to histidine, and the color value was far too high to agree with the amino nitrogen value; hence the presence of imidazole acetic, propionic, lactic, or acrylic acids is indicated. This discrepancy between the color value and the amino nitrogen value is equal to 0.0365 gm. (29.7 per cent) calculated as imidazole propionic acid. This organism therefore attacks the histidine in two distinct ways to obtain nitrogen. Some of the mole- cules are deaminized, others suffer a nuclear disruption. Some, but not much, ammonia is generated during the first 3 days. During the next 2 days the imidazole propionic (?) acid formed at first was destroyed with the liberation of some non-volatile primary amine and considerable ammonia. The color obtained with the histidine fraction in this case was just like that obtained with histidine. We have assumed, therefore, that the color was at least largely due to histidine. If this con- clusion is correct, the attack during this 2 day period was directed almost exclusively against the imidazole propionic acid formed during the first 3 days. During the next 2 days the remaining histidine was destroyed almost completely with the liberation of ammonia. Histamine was not formed at any time. The three strains of Bacillus mucosus capsulatus studied by us satisfied their nitrogen requirements at the expense of the amino-acid histidine in preference to ammonia. Two of the strains deaminized histidine and ruptured the nucleus simultaneously. The other strain merely ruptured the nucleus. Glycerol was, in every case, the chief source of carbon. M. T. Hanke and K. K. Koessler 175 B. bifidus —This organism grew very poorly on our medium, apparently because an available carbon source was lacking. The glycerol originally introduced seemed not to have been attacked. Of the histidine originally introduced, 87 per cent was recovered. There is no evidence that the nucleus wasruptured. The equivalent of 90 per cent of the initial ammonia was recovered. Histamine was not formed. When leucine was added to the medium the organisms grew for a few days during which time all the leucine nitrogen disappeared and some acid was produced. Histamine was not produced in this case. B. influenze.—This organism grew very poorly on our medium. There is slight evidence for the formation of imidazole propionic (etc.) acids. Of the ammonia originally introduced, 91 per cent was recovered. Acid was not produced; glycerol was not utilized. When leucine (0.10 gm.) was added to the“medium, the organisms grew for a few days. They apparently found, in leucine, an easily available source of carbon and nitrogen because none of the leucine nitrogen could be accounted for at the end of 14 days. In this case 95 per cent of the histidine originally introduced was recovered. Histamine was not formed in any case. B. proteus vulgaris 186.—This organism grew poorly on our medium apparently because an easily available source of carbon was lacking. Acid was not produced; glycerol was not attacked. The histidine was pro- gressively destroyed, the excess nitrogen appearing as ammonia. The organisms must, therefore, have destroyed the histidine to obtain carbon. Histamine and other imidazoles were not formed. B. proteus vulgaris A. I. K.—This organism grew very well on our medium. Glycerol was consumed and acids were produced. The great discrepancy between the color and amino nitrogen values for histidine is good evidence that a carboxylated amino compound was produced from the histidine. Since this organism was able to use ammonia as a source of nitrogen, the excessive amine production could hardly have been re- sorted to because of the nitrogen thus rendered available. The amines were more probably produced to lower the hydrogen ion concentration of the cell protoplasm. The equivalent of 77 per cent of the ammonia originally introduced was recovered. Histamine and other imidazoles were not formed. B. cloace I.—As far as we could tell, this organism died shortly after it was introduced into our medium. The histidine and the glycerol were not attacked. Streptococcus hemolyticus 4A and II R.—Neither of these organisms grew perceptibly on our medium although they were still alive at the end of 14 days. Glycerol was not consumed; acid was not produced. About 85 per cent of the histidine originally introduced was recovered in each cease. The initial ammonia concentration was reduced by only 10 per cent. Histamine and other imidazoles were not formed. The results obtained with a strain of Streptococcus viridans were quanti- tatively almost identical with those obtained above. 176 Studies on Proteinogenous Amines. XII Pneumococci, Types I, II, III, and IV.—These* organisms grew poorly on ourmedium. They removed 15 to 20 per cent of the histidine originally introduced in such a way that the nitrogen cannot be accounted for either as NH; oras NH». There is no evidence that a nuclear rupture occurred. The constancy of the ammonia value indicates strongly that these organ- isms do not use it as a source of nitrogen. Histamine and other imidazoles were not formed. When leucine (0.10 gm.) was added to the medium the organisms multi- plied rapidly for a few days. The addition of leucine had no effect upon the NH; and pH values and the histidine was removed to about the same extent as in the leucine-free medium. The leucine, however, was almost completely destroyed in every case. This would seem to indicate that leucine is a good source of nitrogen and carbon for the pneumococcus while histidine, ammonia, and glycerol are poor sources of these elements. B. tuberculosis H.Sp., 1305, 3161, and O.H.—These four strains of tuber- , cle bacilli gave results that were qualitatively identical though quantita- tively somewhat different. In every case there is good evidence that imidazole acetic, propionic, lactic, or acrylic acids were present. The color obtained with the histidine fraction was too red for histidine. The values obtained for histidine by the amino nitrogen method were always decidedly lower than those obtained colorimetrically. The discrepancies between the color values and the amino nitrogen values are equal to 0.015 gm. (12.2 per cent) for H.Sp., 0.080 gm. (24.5 per cent) for No. 1305, 0.02 gm. (16.3 per cent) for No. 3161, and 0.028 gm. (18.7 per cent) for O.H.., calculated as imidazole propionic acid. The organisms grew very well on the surface of our medium. An excess of acid was not produced during this time interval although the glycerol had been largely destroyed. The ammonia consumption was distinctly different in the four cases, varying from 40 per cent in the case of H.Sp. to only 10 per cent in the case of O.H. Histamine was not produced in any case. : A striking contrast to these four organisms was that of a fifth strain which is widely known as K;. This organism has been grown on synthetic media for so long—about 40 years—that it has lost its virulence entirely. It grows excellently on our medium, the growth not being confined to the surface. Chemically it did not behave like a typical tubercle bacillus as can be seen by examining the last two horizontal columns of Table VIII. This organism attacked the histidine immediately and so effectively that 53 per cent of it had been destroyed at the end of 5 days of incubation. Some of the histidine nitrogen was converted into ammonia; some appeared as a non-volatile, carboxylated amine. After 45 days of incubation only 12 per cent of the histidine originally introduced remained. An excess of acid was not formed although the glycerol had been destroyed. Imidazole propionic acid and histamine were not produced. In concluding this part of the work we wish to call special attention to the fact that these experiments have all been carried out in an artificial synthetic medium. Our conditions are, at M. T. Hanke and K. K. Koessler LF best, highly artificial and the composition of our synthetic medium is far simpler than that of living tissue or even than that of the usual composite culture media. We are well aware that results obtained in such a medium must be: applied carefully, if at all, to the general problem of amine production in the living organism. We feel, however, that our mode of procedure is justified by the fact that we have first to develop methods for estimating imida- zoles and phenols under simple conditions before we can hope to estimate these substances in complex mixtures. The final answer to the question, ‘‘which microorganisms are probable amine pro- ducers in the human organism” can only be given after a medium has been used that contains all of the essential ingredients of a tissue or tissue extract. How definitely the amine production is influenced by the composition of the medium is shown clearly in the Part IV of this report where other amino-acids or peptones were added to our synthetic medium. PART IV. The Production of Histamine by Bacillus coli cystitis when Other Amino-Acids Are Present in the Medium Together with Histidine. This investigation was undertaken to ascertain which amino- acids when added to our medium, augment the production of his- tamine and the growth of the microorganisms. The colon bacillus used in this work (Bacillus coli cystitis) was the strain employed by us in our original investigation.| We have found that this organism will always convert approximately 50 per cent of the histidine originally introduced into histamine when precautions are taken to have the initial pH of the medium 7.3 and to main- tain a uniform temperature of 37°. In the first experiment of this series we compared the carboxy- lase activity on our standard medium with that on a medium containing leucine and histidine in one case, and peptone and histidine in the other. The results obtained are summarized in Table IX. The results obtained in the medium containing histidine as the only amino-acid call for very little special comment because they closely resemble those obtained and reported 2 years ago. In 178 Studies on Proteinogenous Amines. XII TABLE IX—Effect of Leucine and Peptone on the Composition of the medium. Histidine dichloride, 0.2 gm. Potassium nitrate, Oya Ammonium chloride, 0.2 “ Glycerol, 4.0 ce. Other inorganic salts as per Nutritive Medium 3. Distilled water to 200 ce. Histidine dichloride, 0.2 gm. Leucine, Om < Potassium nitrate, Om Ammonium chloride, 0.2 “ Glycerol, 4.0 ce Other inorganic salts as per Nutritive Medium 3. Distilled water to 200 ee. Histidine dichloride, 0.2 gm. Peptone (Witte), 0.05 “ Potassium nitrate, Osis = Ammonium chloride, 0.2 “ Glycerol, 4.0 ce Other inorganic salts as per Nutritive Medium 8. Distilled water to 200 ce. Total color value of the test solution as histidine dichloride.* (0.20 gm. = 100%.) 0.10cce. = 11.2mm. 0:20 “ ='22.45) ** Match perfect. Color develops promptly. 112% 0.10 ce. = 10.3 mm. 0.20 “ =20.6°" Match perfect. Color develops promptly. 103% 0.10 cc. = 11.2mm. 0.20 “~ = 22.4 “ Match perfect. Color develops promptly. 112% * Colors matched against the (CR-MO) standard. Color value of histidine fraction as histidine dichloride.* (0.20 gm. = 100%.) 0.10 ce. = 4.0 mm. 0320 —s 87 Oe Match perfect. 40% 0.50 ce. = 4.5 mm. 1300) “<=:950. a5 Match poor. Color too yellow. 9.0% 0.20 ce. = 3.0 mm. 0:40. i= 640% <5 Match good. ¥ Unchanged histidine (Van Slyke method) with 5 cc. of test solution. 0.85 ce. Ne at 20° 0.72 cc. Ne at 21° and Y5e mm. 0.0822 gm. his- tidine dichlo- ride. 41.1% 0.65 ce. Ne at 22° and 754 M. T. Hanke and K. K. Koessler 179 mc ion of Histamine by Bacillus coli (cystitis).» Reaction. a eae converted into 0.10 n HCI neutralized er i 3 Abe t ; 2 S : ‘olor value of the histamine fraction. (Van Slyke method) corel nies aero = Z¢ 5 ce. of test solution. ; z= Se se | 22 ES ee a a ce. pH pH D5 ce. = 13.8 mm. 1.04 ce. Ne at 20° 74 ae me <= 275 “ and 754 mm. Hence the = of 9 ce. or developed like that of hista-| 0.0964. gm. histamine of 0.1 N NH; was | aoe dichloride. removed by the 092 gm. of histamine dichloride microorganisms. 7.3 | 5.4 ir entire test solution. 59.8% % of histidine converted into istamine. lor too intense to run directly;| 2.55 ec. Ne at 22° 25 hence 10 cc. were diluted to 100 cc. and 744 mm. Hence the = of 11 cc. Of this solution 0.231 gm. histamine of 0.1 n NH; was dichloride. removed by the em. microorganisms ce “ Ss i 15.2 143% ant oe lor develops like that of hista- ; im mine. 275 gm. of histamine dichloride n entire test solution. 9 of histidine converted into istamine. lor too intense to run directly;| 1.45cc. N2. at 23° and} rence 10 cc. were diluted to 100ce. 755 mm. If this solution 0.133 gm. histamine PS cc. = 9.3mm. eee me = 18.5 “ 82.4% 7.3/5.2 lor develops like that of hista- ; or 25 gm. of histamine dichloride n entire test solution. 4% of histidine converted into 4 | nine. listamine. 180 Studies on Proteinogenous Amines. XII this case 57 per cent of the histidine originally introduced was converted into histamine. The agreement between the values obtained colorimetrically and by the amino nitrogen method is good. é 2 Although the organisms multiplied rapidly in the above medium, they grew far better in a medium containing leucine or peptone together with the histidine. When leucine was present, 79 per cent of the histidine originally introduced was converted into histamine, as determined from the colorimetric reading an in- ° crease of 22 per cent over the leucine-free medium. The amino nitrogen value is obviously much too high to represent histamine alone. A portion of the leucine may have been converted into isoamylamine which would appear together with histamine in the amyl alcohol extract and thus raise the amino nitrogen value of the histamine fraction. In the presence of peptone, 77 per cent of the histidine orig- inally introduced was converted into histamine; hence leucine and peptone are about equivalent in their ability to promote histamine formation. In this case the check between the colori- metric and amino nitrogen values is sufficiently close to warrant the conclusion that other amines were not formed in appreciable . quantities. In this case the acidity was also neutralized less. perfectly than in the case of leucine where a large amount of amine production, other than histamine, was indicated. In short then, we would conclude that the presence of either leucine or peptone stimulated the production of histamine. We can draw no conclusions from this experiment, as to the rdle played by the leucine or peptone in the amine production. We might conclude, with Jacoby,° that leucine was an easily available source from which carboxylase enzymes could be synthesized; but our experiment seems to indicate that the chief factor is the enormous increase in the speed of growth of the microorganisms which would, of course, increase the rate of acid production and hence render the early formation of histamine necessary to neu- tralize the acid. We have grown all of the organisms-discussed in the first three sections of this paper on a medium containing both leucine and histidine and have found that: * Jacoby, M., Biochem. Z., 1917, 1xxxi, 332; Ixxxiii, 74; Ixxxiv, 358; 1918, Ixxxvi, 329. M. T. Hanke and K. K. Koessler 181 1. The addition of leucine always facilitates the growth of the organisms. 2. If the organisms produce no histamine in the absence of leu- cine, they will not produce histamine when leucine is present. 3. If*the organisms produce histamine when leucine is absent, they will always produce 20 to 25 per cent more of this amine, in 2 weeks of incubation, when leucine is present. The leucine augments a power that already exists; it does not call forth a new enzymatic activity. The second series was an enlargement upon the first for in this case all of the easily available amino-acids were employed; namely, glycine, alanine, cystine, leucine, arginine, glutamic acid, tyrosine, and tryptophane. In addition to the above amino-acids two flasks were prepared containing different makes of peptone; namely, Witte and Difco. The results obtained are summarized in Table X. : Table X shows that the histamine value on our standard medium was 20 per cent lower this time than it has been in any of our other experiments with this organism. The value on the leucine- histidine medium is also 20 per cent low. All of the flasks of this series were incubated at the same time in the same incubator. During the first few days the air in the incubator smelled strongly of hydrogen sulfide which was being evolved from the cystine- containing flask. We feel convinced that the retardation in speed of activity, and probably in rate of growth, was caused by the hydrogen sulfide. This conclusion is strengthened by the fact that only 2 per cent of histamine was formed in the cystine- histidine medium. The growth in this medium was also very meager. Since this hydrogen sulfide interference must have been approximately equivalent for all of the media, we believe that the values obtained are accurate in so far as they may be compared with one another. In discussing this series, a conversion of 31 per cent of the histidine originally introduced into histamine is considered to be the normal conversion on our standard medium. When more than 31 per cent of histamine was present, the amino- acid added can be considered to be an amine production stimula- tor, and vice versa. The amino nitrogen values on both the his- tidine and histamine fractions have, in every case, been calculated as histidine and as histamine, respectively, in spite of the fact 182 Studies on Proteinogenous Amines. XII Composition of the medium. Histidine dichlor- ide, 0 KNOs, 0. NHiCl, 0 Glycerol, 4 Usual in- organic salts. Distilled water to 200 ‘ Total color value calculated as % of histidine di- chloride. (0.20 gm. = 100%), per cent. Unchanged histi- dine colorimet- ric method. (0.20 gm. = 100%), per cent. Amino nitrogen value of histi- dine fraction calculated as % of histidine dichloride. (02205 cm 7= 100%), per cent. Histidine conver- ted into hista- mine (colori- metric method) per cent. Amino nitrogen value of hista- mine fraction calculated as % of histamine dichloride, per cent, te 60 31 TABLE X—Fffect of Amino-Acids and Peptone a Histidine dichlor- ide, 0.2 gm. Glycine, 0.065 ‘‘ KN@s,) (0:150 2" NEC,” 02a Glycerol, 4.0 ce. Usual in- organic salts. - Distilled water to 200 ‘ 108 43 102.5 37.2 49 Histidine NH.CI, 0.2 “ Glycerol, 4.0 cc. Usual in- organic salts. Distilled water to 200 ‘* 100 30 154 be | So “J Histidine dichlor- ide, 0.2 gm Leucine, 0.115 ‘ KNOs, “10:1e5e4 NHiCl,_.0:27 0" Glycerol,4.0 ce. Usual in- organic salts. Distilled water to 200 ‘* 112 sual in- organic salts. Distilled water to 96 87 147 2.06 M. T. Hanke and K. K. Koessler 183 oduction of Histamine by Bacillus coli (cystitis.) Saree Histidine Histidine Histidine ee I re re oe ~ carbon- Tyrosine, 0.16 "| TTYPto- Peptone (Difco), 0.05 “ ate, 0.2 * |KNOs, 0.1 “| ~Phane, 0.18 ‘* | _ (Witte),0.05 “ |uNG,” one KNOs, 0.1 ‘ NHCi, 02 “ Pl eee eet ao. « NHCi, 02 “ ae me * Cae 4.0 cc. Glycerol, 4.0 cc. Glycerol, 4.0 ce. ay esol 4.0 cc, anal in- ¢ organic Usual in- Usual in- organic organic salts CS ore salts salts. Distilled aed ee Distilled 7 . “oe = ‘ Pallet 200 * water to 200 waterto 200 ‘* waterto 200 ‘ water to 200 110 99 103 97 52 1S 60 40 50 if 109.8 ; 10.6 79.5 60.6 67.1 41 31.4 26.3 42 42 108.4 56.6 62.8 Composition of the medium, 0.10 n HCI neu- Histidine dichlor- ide, 0.2gm. KNOs;* 0} NH.CI, 0.2 ‘* Glycerol, 4.0 cc. Usual in- organic salts. Distilled water to 200 ‘* 20.5 tralized by NH;| Hence the = from the entire test solution, CC. pH after 14 days of incubation. The initial pH was 7.3 in all ’ cases, pH. Tyrosine conver- ted into vola- tile phenols (colorimetric method), per cent. Tyrosine conver- ted into aro- matic hydroxy- acids. Calcu- lated as oxy- phenyllactic acid (colori- metric method) per cent. Unchanged tyro- sine (colorimet- ric method), per cent. Tyrosine conver- ted into tyra- mine (colori- metric method) per cent. of 16 ce. of 0.1 N NH; was used by microor- ganisms. 5.5 Histidine Glycerol, 4.0 cc. Usual in- organic salts. Distilled water to 200 ‘* 16.5 Hence the = of 19.5 cc. ie (Ol ay _NH; was used by mi- croorgan- isms. 5.5 184 Histidine dichlor- ide, 0.2 gm. | Alanine, 0.16 * KNOs, O00 | NHuCI, 0.2 ** Glycerol, 4.0 cc. Usual in- organic salts. Distilled water to 200 ‘ 17 Hence the = of 19 cc. of 0.1 n NH; was used by microor- ganisms. 5.6 Histidine NH.Ci, 0.2 * Glycerol, 4.0 ce. Usual in- By 4 Hence the = of 12.3 ce. of, 0. ier NH; was used by mi- croorgan- isms. Hence the 9.2 of 26.8 of “O22 NH; used by croorgan}’ isms. | ef | proorgan- sms. Histidine NH.Cl, Glycerol, Usual in- organic salts. Distilled water to 200 ‘* 19.5 Hence the = of 16.5 ce. Of Oeil = N NH; was used by mi- croorgan- Histidine dichlor- ide, Tyrosine, 0.16 ‘' KNOs 7 Ot NH.Cl, - Glycerol, 4.0 ce. Usual in- organic salts. Distilled water to 200 ‘* ans Hence the = of 12.5 cc. of 0.1 N NH; was used by mi- croorgan- isms. 0.2 gm. Histidine dichlor- ide, 0.2 gm. Trypto- phane, 0.18 “* KNOs, 30.1 ‘* ING. © 0.25 Glycerol, 4.0 ce. Usual in- organic salts. Distilled water to 200 ‘ 20 Hence the = of 16 cc. of Of aN NH; was used by microorgan- isms. Histidine dichlor- ide, 0.2 gm. Peptone (Witte),0.05 ‘* WIN Os.2 8 OSE) 1c INEGCI 0:2 °** Glycerol, 4.0 ce. Usual in- organic salts. Distilled waterto 200 ‘ 24.5 Hence the = of if. 5 ce. 0.1 N NH; was used by microorgan- isms. Histidine di- chloride, 0.2 gm. Peptone (Difco), 0.05 ** KNOs, OTs NH.Cl, 0.29% Glycerol, 4.0 cc. Usual in- organic salts. Distilled water to 200 ‘* 22.2 Hence the = of 13.8- ce. of 0.1 NNH; was used by microorgan- | isms. None. 3.4 185 186 Studies on Proteinogenous Amines. XII that the values must obviously be high because of the presence of the other amino-acid or peptone. These amino nitrogen figures are of value because they give some information about the fate of the added amino-acid. Glycine and Histidine—The organisms grew distinctly better in a medium containing glycine and histidine than they did in a medium containing only histidine. The histamine production was increased by 6.2 per cent. Most of the glycine nitrogen remained in the histidine fraction in the primary amino condition which indicates that only a small amount of this amino-acid was utilized by the microorganisms. Arginine and Histidine —The microorganisms grew very much better in this medium than they did in the one containing glycine and histidine. The histamine production was increased by 10 per cent. The amino nitrogen figures indicate that over half of the arginine was converted into some product that passed into amyl alcohol from a strongly alkaline solution. The arginine may, therefore, also have been decarboxylated. Peptone and Histidine (Witte and Difco)—The presence of either of the above peptones is a great stimulus to the rate of growth of these bacilli, the growth being about equivalent to that ob- tained in the arginine-histidine medium. The histamine pro- duction was increased by 11 per cent in each case. Alanine and Histidine-——The organisms grew even better in a medium containing alanine than they didin one containing peptone. In this case the histamine production was increased by 20 per cent, the highest value obtained with any of the amino-acids. The alanine used in this work was a mixture of equal parts of the d- and I-varieties. Two equivalents of this alanine were used. It is interesting to note that one and a quarter equivalents re- mained, probably as alanine, in the histidine fraction and that a maximum of one-fifth of an equivalent passed into amyl alcohol, possibly as an amine. In short, then, over half of the introduced available nitrogen was consumed by the microorganisms. (We have assumed that /-alanine did not serve as a food.) Leucine and Histidine-—The growth obtained in this case was slightly better than that obtained on the histidine-alanine medium. The histamine production was increased by 18.6 per cent which is almost identical with that obtained on the alanine-histidine M. T. Hanke and K. K. Koessler 187 medium. Most of the leucine nitrogen appeared in the hista- mine fraction which indicates that the leucine was also decar- boxylated. Tryptophane and Histidine—The growth in this medium was about as dense as that obtained in the peptone-histidine medium; but the histamine production was decreased, not increased. The 26.3 per cent of histamine obtained in this case is about 5 per cent less than that obtained in a medium containing histidine as the only amino-acid. We anticipated some trouble with our colorimetric process in this case; but we were relieved to find that tryptophane, or its decomposition products, did not interfere in any way with the color production. Glutamic Acid and Histidine -—The growth on this medium was slightly less than that on the tryptophane-histidine medium. The presence of glutamic acid decreases the histamine production by about 9 per cent. The amino nitrogen figures indicate that most of the glutamic acid nitrogen was still attached to a car- boxylated molecule. Cystine and Histidine —The growth on this medium was very meager. The presence of cystine decreases the histamine pro- duction by 29 per cent so that the formation of this amine is almost nil. Tyrosine and Histidine-—The fact that tyrosine and its deriv- atives give a color with alkaline p-phenyldiazonium sulfonate, made it necessary for us to modify our usual procedure in this ease. The method employed was briefly as follows: 1. The filtered medium, after the usual treatment with 1 ce. of 95 per cent sulfuric acid, was subjected to a distillation, under ordinary pressure, the distillate being collected and examined colorimetrically!® for volatile phenols. Phenols were absent. 2. The contents of the distilling flask were transferred to a glass dish and evaporated on the water bath. The residue was transferred, with water, to a 25 cc. precision cylinder and diluted to 25 ce. This is the test liquid. 3. Of this acid test liquid 10 ce. were transferred to a 35 ce. extraction bottle and extracted ten times with ether, using 20 ec. for each extraction. The combined ether extracts were treated 10 Hanke, M. T., and Koessler, K. K., J. Biol. Chem., 1922, 1, 238. 188 Studies on Proteinogenous Amines. XII with 25 cc. of water, the ether being then removed by distillation at first under ordinary pressures and then in vacuo. The solution so obtained was transferred, with water, to a 100 cc. precision cylinder and diluted to 100 cc. It was tested, colorimetrically,!° for aromatic hydroxy-acids. In this way 3.4 per cent of the tyrosine originally introduced was found to have been converted into oxyphenyllactic acid. 4. The aqueous liquid dbtained above, that had been freed from aromatic hydroxy-acids by extraction with ether, was trans- ferred to a 250 cc. Pyrex flask with 90 cc. of water. Silver nitrate, 10 cc. of a 20 per cent solution, was added and the result- ing mixture treated with 12 gm. of barium hydroxide in 50 ec. of warm water. The dark brown mixture was filtered, the pre- cipitate being washed with a cold saturated solution of baryta.! This divides the material into two fractions; the silver precipi- tate, which will contain all of the histamine and which should con- tain most of the histidine, and the silver filtrate which should con- tain all of the tyrosine and tyramine. 5. The silver precipitate was suspended in water and treated with an excess of 37 per cent HCl and Na,SO, as has been previously described. The resulting mixture was filtered after 2 hours of digestion on the water bath and the filtrate evaporated on the water bath. The residue was then transferred to an extraction bottle with 10 cc. of water and treated just like any of the histi- dine test liquids. It was found to contain 11.5 per cent of his- tidine and 31.4 per cent of histamine. The recovery of imid- azoles was only 43 per cent. We have not checked up on the sil- ver precipitation method for histidine sufficiently to be certain that some histidine may not have remained unprecipitated. We are certain that the histamine figure is accurate because the method employed in this experiment was identical with that used in the quantitative precipitation of histamine in some of our earlier work. 6. The silver filtrate. The barium and silver ions were exactly removed with H»SO, and HCl, respectively. The filtrate from BaSO, and AgCl was neutralized with sodium hydroxide and evaporated on the water bath. The residue was transferred to 11 Hanke, M. T., and Koessler, K. K., J. Biol. Chem., 1920, xliii, 543. M. T. Hanke and K. K. Koessler 189 an extraction bottle with 10 cc. ef water. The subsequent sep- aration of tyramine from tyrosine and the colorimetric estima- tion of each substance in its respective fraction was then carried out as described in the following paper." Tyramine was found to be absent. Of the tyrosine originally introduced, 72.6 per cent was recovered. The histamine production on this medium proceeded at a rate identical with that on a tyrosine-free medium. Tyrosine neither augmented nor retarded the histamine formation. Tyrosine was not decarboxylated. It appears then, that the decarboxylation of histidine is influ- enced by the presence of other amino-acids in all of the three possible ways. Tyrosine is without effect. Leucine, alanine, arginine, and glycine increase the rate of decarboxylation. Of these, leucine and alanine are by far the most efficient. Cystine, glutamic acid, and tryptophane decrease the rate of decarboxylation. Of these, cystine is by far the most efficient. The rate of decarboxyla- tion is not entirely coincident with the rate of growth of the microorganisms because, with the exception of cystine, all of the amino-acids augmented the growth of the organisms; but tryp- tophane and glutamic acid decreased the histamine production. SUMMARY. 1. The behavior of a large number of microorganisms has been studied on a liquid medium consisting of histidine dichloride (0.2 gm.), ammonium chloride (0.2 gm.), potassium nitrate (0.1 em.), potassium dihydrogen phosphate (0.4 gm.), sodium chloride (0.8 gm.), sodium sulfate (0.02 gm.), sodium bicarbonate (0.4 gm.), calcium chloride (0.01 gm.), and glycerol (4.0 cc.), in a total aqueous volume of 200 ce. : 2. The organisms studied were Bacillus coli communior (7 strains), Bacillus coli communis (5 strains), Bacillus lactis aerogenes (5 strains), Bacillus acidi lactici (12 strains), Bacillus enteritidis, Bacillus typhosus, Bacillus paratyphosus R—C-+H+H+-CT 70 = che a + NH, OH NH, OH ae Ch 104 (0) —2(.) 3 HO-+Ch ~O—>Cco,+ 8.0 5 ar Ce = OH OH | 2. With a triply positive carboxyl group. ; H H 5 — -- | R—C+—-Ct70+2© +2) > R—C-+H+H+-CT7O + + = a | NH, OH NH, OH The formic acid would then give CO, and H.O, by oxidation, as in C above. The above equation tells nothing about the mechanism of the reaction. Following Nef, we might consider the above reaction to proceed as follows.’ H a _— R—C+-C{7O @ R—C+ +H+-CT7O0 + + - -- NH, OH | NH, OH ig Re = 120 ko “p + NH. NH; 3 Koessler, K. K., Proc. Inst. Med. Chicago, 1920, iii, 46. 196 Studies on Proteinogenous Amines. XIII H 72 R=C. ~oOny> Ro eee + + NH NH; For reasons that will be discussed later, the schematic possi- bility represented by equation (1, C) is hardly worthy of con- sideration because it involves the reduction of a quadruply positive carbon atom. As far as we now know, only the chloro- phyil-containing plants have this power. A consideration of Types I and II shows that the kind of reac- tion resorted to must depend upon the direction of the electrical field between the carboxyl group and its neighboring carbon atom. If we were certain that the carboxyl group in all amino-acids was negative with respect to its neighboring carbon atom as in formula (2), we could state with a fair degree of assurance that the decar- boxylation of amino-acids must be associated with an oxidation reduction process. Unfortunately, we are at present unable to make a definite statement as to the electrical conditions that prevail in any amino-acid (see under allyl chloride, pages 231 to 233). It seems indeed, as if the charge on the carboxyl group is not the same for all amino-acids. Although we are, therefore, unable at present to give an elec- tronic interpretation of the carboxylase activity as applied to amino-acids, the electronic point of view offers a fascinating ex- planation for a related type of carboxylase activity. Neuberg* and his coworkers have shown that yeasts and yeast extracts have the faculty of decarboxylating pyruvic and many other organic acids with the formation of carbon dioxide and aldehydes; e.g., CH3—CO—COOH—CO, + CH3;—CHO. Karezag and Breuer® have shown that many bacteria, although they ferment pyruvic acid with gas formation, produce no alde- hyde. The gas produced, instead of being pure CO, as in the case of the yeasts, consists of hydrogen to the extent of 90 per 4 Neuberg, C., and Hildesheimer, A., Biochem. Z., 1911, xxxi, 170. Neuberg, C., and Tir, L., Biochem. Z., 1911, xxxii, 328. Neuberg, C., and Karezag, L., Biochem. Z., 1911, xxxvi, 68, 76; xxxvii, 170. Neuberg, C., and Kerb, J., Biochem. Z., 1912, xlvii, 413, 405. ° Karezag, L., and Breuer, E., Biochem. Z., 1915, Ixx, 320. M. T. Hanke and K. K. Koessler 197 cent. Obviously, then, the actidn of yeasts on pyruvic acid is radically different from that of the investigated bacteria. The authors believe that some light is shed upon these facts and others to be presented later, by a consideration of the electronic formulas of these substances. Acetaldehyde and carbon dioxide could be obtained from py- ruvic acid by either of the three following processes.® 1. Hydrolysis. HC-+C-+C{—O0+H+-0H> HC-+C-+H+07 {C+ -08 fae + ++ + oO .\OH O OH L CO, + H.O 2. Reduction followed by oxidation. H.C-+C-+C01 7 042) +2) H.C-+C-+H+H+-CT 70 eee P + +-— ie On ‘on OH H+-Ct~O+ (0) #24) —~ OL TC4+-O0H +7 as OH a CO, + H,0 Process 2(a). H -- H.C -+C-+Ct 7042 +2©—> HC-+C+4-CT 7 ++ + + + Oa Or OH OF H -+- HC-+C+-Ct >O+H+-0H > H.C-~+CT704+H+-CTTO + + — reas OH OH ti OH H+—CPIO+#(O) +24 BS COs 10 OH 6 The proof of the electronic formulas for the compounds used in the equations throughout this paper is given in Part III. 198 Studies on Proteinogenous ‘Amines. XIIT The decomposition of pyruvic acid by yeast cannot occur according to process (2) or (2,a) for the following reasons. Process (2) demands the intermediate formation of formic acid by the reduction of a quadruply positive carbon atom. Such a re- duction would be equivalent to the direct conversion of CO, to formic acid. Up to the present time only the chlorophyll-contain- ing plants have been proved to be capable of reducing COs, and they can do so only in the presence of sunlight. This process is, therefore, highly improbable from the start. If we pass lightly over this first serious objection, we come to another that is Just as serious. The formulation demands that the formic acid formed at first must then be oxidized to CO... That bacteria have this power of oxidation is clearly proved by the work of Omelianski’ and Pakes and Jollyman.® In every case, however, hydrogen is evolved along with the CO.. Hydrogen is never evolved in the early stages of yeast fermentations. Although the evolution of CO, from pyruvic acid by yeast is extremely rapid so that the fermentation is practically over in 24 hours irrespective of the variety of yeast employed, Neuberg and Tirt have shown that certain of these yeasts will evolve no gas at all from sodium for- mate solutions and that none of them evolves more than traces of CO, in the course of 24 hours. An objection might be raised to these conclusions from the work of Neuberg and Tir because these authors used a pure solution of sodium formate in which no real fermentation could occur. It is conceivable that the oxidation of formic acid might take place only in the presence of sugar. Franzen and Steppuhn?® have shown conclusively that when sodium formate is mixed with a nutrient medium rich in sugar, most of the pure cultures of yeast employed, although they fermented the sugar vigorously, either did not ferment the added formic acid at all or did so only to a very small extent. In no case was the destruction of the formic acid rapid enough to conclude that this substance was an intermediate product in the fermentation of sugar. Formula- tion (2) is, therefore, highly improbable. 7 Omelianski, W., Centr. Bakt., 2te Abt., 1904, xi, 177, 256, 317. 3 Pakes, W. C. C., and Jollyman, W. H., J. Chem. Soc., 1901, Ixxix, 322, 386, 459. ® Franzen, H., and Steppuhn, O., Z. physiol. Chem., 1912, Ixxvii, 129. L M. T. Hanke and K. K. Koessler 199 Process (2, a) demands the intermediate formation of lactic acid. That the reduction of pyruvic acid to lactic acid might be resorted to by bacteria in case the pyruvic acid was being used by them as a source of carbon is very possible because as can be seen from the formula, this is an indirect but simple method for converting an unavailable quadruply positive carboxyl group into one that is triply positive. That yeasts do not resort to this proc- ess, in the case of pyruvic acid, is proved by the fact that lactic acid is not fermented at all by some yeasts and only to a very slight extent by most (see article by Neuberg and Tirt and Buch- ner and Meisenheimer’’). Even if lactic acid were fermented by yeast, the formic acid difficulty would still be insurmountable (see above under proc- ess (2)). Apparently then, formula (2, a) cannot represent what happens when pyruvic acid is fermented by yeast. , Process (1) is the only one left and its very simplicity is exactly what one would expect from a reaction that proceeds so smoothly and rapidly. The above considerations have convinced the authors that the fermentation of pyruvic acid by yeast ts ac- complished by a purely hydrolytic process. From the ordinary structural formula for lactic acid it is diffi- cult to see why a similar hydrolysis does not occur in this case with the formation of CO, and alcohol. The hydrolysis of lactic acid with dilute sulfuric acid, which takes place quite as readily as the similar hydrolysis of pyruvic acid, gives acetaldehyde and formic acid. The electronic formulas for these two compounds show clearly why CO, could not be obtained by the hydrolysis of lactic acid and they show moreover why yeast cannot hydrolyze lactic acid at all. Lactic acid contains a triply positive car- boxyl group while pyruvic acid contains one that is quadruply positive. Apparently then, yeast can hydrolyze only those acids containing quadruply positive carboxyl groups. With this idea in mind it is easy to see why oxalacetic acid should give acetaldehyde and COs, acetone dicarbonic acid should give acetone and COs, and in general why all of the a-ketonic acids should give an aldehyde and CO, as primary products because they all contain quadruply positive carboxyl groups. A striking confirmation of this statement is found in the fact that glyoxylic 10 Buchner, E., and Meisenheimer, J., Ber. chem. Ges., 1910, xliii, 1773. 200 Studies on Proteinogenous, Amines. XIII acid, the mother substance of the a-ketonic acids, zs practically not fermented by yeast.t This acid has been shown to contain a negative carboxyl group and it could not give COz on hydrolysis. No attempt has been made to explain the mechanism of hydroly- sis. Just why pyruvic acid should hydrolyze readily whereas acetic acid must be fused with an alkali to hydrolyze it is not ex- plained by the electronic formulas. The failure of yeast to fer- ment an acid does not prove that the acid contains a negative carboxyl group. On the other hand, if yeast does split CO: from an acid readily it is fairly safe to conclude that the acid contains a quadruply positive carboxyl group. Acetic acid is fermented very slowly by yeast. Neuberg and Tir make the state- ment that the gas evolved is only partially absorbed by sodium hydroxide. It would be of interest to ascertain if the unabsorbed gas was methane because if it was methane, the yeast would have proved itself capable of hydrolyzing even this very resistant acid. PART Ii. On the Beta Oxidation of Fatty Acids in the Animal Body. The work of Knoop, Dakin, and others! leaves little doubt that fatty acids undergo oxidation predominantly in the beta position in the animal body. Just why the living cells and hydro- gen peroxide should oxidize the beta carbon atom by preference while the halogens react only with the alpha hydrogen atoms of the fatty acids is a question that has been much discussed in the past. It occurred to the authors that this tendency toward beta oxidation in the body (and by hydrogen peroxide) might be closely associated with the electrical structure of the molecule. The carbon atom can occur in five stages of oxidation; namely, “= — + + 1 —-C— 2. —-C+ 3. —C+ 4, —C+ 5. +C+ - + + + Carbon dioxide is the simplest substance that contains carbon in the completely oxidized condition represented by form (5). Neither microorganisms nor the cells of the animal organism seem to have the faculty of reducing CO,. Only the chlorophyll- 1tThe references to the original literature are given in Dakin’s mono- graph (Dakin, H. D., Oxidation and reductions in the animal body, London, 1912). M. T. Hanke and K. K. Koessler 201 containing plants have been proved to be capable of reducing CO, to formaldehyde and they can do so only in the presence of sunlight. Forms (2), (3), and (4) seem to be oxidizable or reduc- ible by living matter depending upon the conditions of the experi- ment. At present it seems advisable to make no definite state- ment as to which of these forms is most readily oxidizable. The work of numerous investigators seems to indicate that form (1) is not readily oxidized by living matter. Thus Karezag and Breuer® showed that oxalacetic, acetone dicarbonic, and acetic acids were not changed by a large number of microorganisms. Of these acids, acetone dicarbonic and acetic acids contain only quadruply negative and quadruply positive carbon. Oxalacetic acid has one quadruply negative, two quadruply positive, and one triply positive carbon atom. From this it would appear that quadruply negative carbon was not readily oxidized by bacteria. Although acetic, acetoacetic, and malonic acids, all of which contain only quadruply negative and positive carbon, are com- pletely oxidized in the normal animal body, acetone is attacked with difficulty" and phenylacetic acid is not oxidized at all.'* Ace- tone has been proved to contain only quadruply positive and quad- ruply negative carbon and phenylacetic acid must have the formula H — CHe+-C-+CT 70 = ib + - a It seems therefore, that even the normal body cells frequently find it difficult to oxidize quadruply negative carbon. From these facts two conclusions can be drawn. 1. If a compound containing quadruply negative carbon to- gether with other more easily oxidizable varieties were fed to a normal animal one would expect a predominant and initial oxi- dation of the more easily oxidizable carbon atoms. In the case 12 For a critical review of this subject see Bayliss, W. M., Principles of general physiology, London, New York, Bombay, Calcutta, and Madras, 1915, 564. 13 Geelmuyden, H. C., Z. physiol. Chem., 1897, xxiii, 431. 14 Knoop, F., Beitr. chem. Physiol. u. Path., 1905, vi, 150. 202 Studies on Proteinogenous Amines. XIII of most compounds, these primary oxidation products would then undergo further oxidation, because of the vigorous oxidizing conditions that prevail in the normal organism, so that an excre- tion of more than traces of these partially oxidized products is hardly to be expected. 2. In certain diseases, notably diabetes, the oxidizing power of the body cells is greatly diminished. Under these conditions the difficultly oxidizable quadruply negative carbon atoms might escape oxidation, at least partially, while the more easily oxidizable carbon atoms in the same molecule might be oxidized. These partially oxidized products, with the quadruply negative carbon still intact, should then be excreted. It is a well known fact that acetone and acetoacetic acid are excreted in large quantities by diabetics. One would expect also that acetic acid, acetone di- carbonic acid, citric acid, and a large number of other substances containing quadruply negative carbon should be at least partially excreted unchanged by diabetics. Citric acid contains one triply positive carbon atom. It might, therefore, be eliminated +as acetone dicarbonic acid or if the carboxylase activity was too great it should surely appear as acetone. (The carboxylase of yeast readily decomposes acetone dicarbonic acid into carbon dioxide and acetone by a purely hydrolytic process as has been shown in Part I. A similar decomposition might occur in diabetics. ) _ With the idea in mind that quadruply negative carbon is not readily oxidized by the body cells, we can now consider the prob- able primary oxidation products of a few acids whose electronic formulas shall be repeated here for the sake of simplicity. H ot CHs+-C-+CP 70 - + + — H ., OH (Phenylacetic acid.) H 4H H 4H + + + + Ci +-C+-C-+C770 H+—-C+-C-+Cf 70 - = + - - .+ + + = + + - Hy) ee Soe H HH O8 (Phenylpropionie acid.) (Propionic acid.) M. T. Hanke and K. K. Koessler 203 H 4H Hives Hh? 5 cH + + so id gat BS ee, ea eee EC ie gent Sopa tir top TA gee Tis a H H OH Hy ee Et A OF (Phenylbutyric acid.) (Butyric acid.) Phenylacetic acid contains the perfectly stable quadruply positive carboxyl group, the very resistant benzene ring, and the presumably difficultly oxidizable quadruply negative alpha car- bon atom. This compound should, therefore, not be readily oxidized even by the normal body cells. As a matter of fact phenylacetic acid is not oxidized by the normal body cells. Phenylpropionie acid is partially oxidized only in the beta position. This triply negative carbon atom should be the most vulnerable position in the molecule. An oxidation at this point would lead finally to benzoylacetic acid which by one type of hydrolysis would give benzoic acid and acetic acid. Knoop found that phenylpropionic acid was converted into benzoic acid in the normal animal while Dakin was also able to isolate a small amount of benzoylacetic acid under similar conditions. Phenylbutyric acid 7s partially oxidized only in the beta position. This doubly positive carbon atom should be the most easily attacked position in the molecule. An oxidation at this point would lead to the formation of phenylacetoacetic acid which by one type of hydrolysis would give phenylacetic acid and acetie acid. Knoop found that phenylacetic acid was formed from phenylbutyric acid in the normal animal. In propionic acid zt 7s again the beta position that is partially oxidized and therefore, should be most readily still further oxidized. The final product of the oxidation would be malonic acid. Ma- lonic acid is, however, very readily oxidized by the normal body cells; hence the isolation of a definite end-product is hardly to be expected in the normal animal. As a matter of fact propionie acid is completely oxidized in the normal animal without leaving a clue as to the mode of its decomposition. In diabetics, malonie acid, a polymer of the half aldehyde of malonic acid, or a conden- sation product of this half aldehyde might possibly be found after feeding propionic acid; but these products seem not to have been sought for. : O 204 Studies on Proteinogenous Amines. XIII Butyric acid zs partially oxidized only in the beta position. Here too, this should be the most vulnerable position. A further oxida-~ tion at this point would lead finally to acetoacetic acid which by hydrolysis could lead either to acetone and carbon dioxide or to acetic acid. The increased excretion of acetoacetic acid and acetone after feeding butyric acid to diabetics is a well estab- lished fact. The cited examples would lead one to conclude that the estab- lished fact that fatty acids oxidize predominantly in the beta position in the animal body, besides having a chemical verification is the fact that the same type of oxidation occurs with hydrogen peroxide, is exactly what one would expect from the electronic formulas of the fatty acids and their derivatives. PART TELE, Proof of the Electronic Formulas of a Few Biologically Important ; Organic Compounds. The General Principles Used in the Determination of Electronic Formulas. Principle I. The Polarity of a Double or Triple Bond.*—Con- sider an olefine derivative having the following structural formula: EE EE A—C—C—B 16 Falk, K. G., and Nelson, J. M., J. Am. Chem. Soc., 1910, xxxii, 1637. Stieglitz has presented this matter in lecture form for the past 6 years, and has applied it to the derivation of the complete formulas for the vinyl derivatives, including acetaldehyde, and for the partial formulas of many organic compounds such as propylene, acrylic acid, crotonic acid, and cinnamic acid. The formula so derived for the vinyl derivatives is H4+—C_{TC+—X; for acetaldehyde is HsC — +C a O; for pro- = 35 ae H H H pylene is H;C —C vee C—+H; for crotonic acid is CH; —C a Cc a5 at + a H H H H — COOH; and for cinnamic acid is CsH; — C BRE C — COOH i+ | + | M. T. Hanke and K. K. Koessler 205 The following three electronic formulas are possible. H He) H H H H aot et = oo cis on se 5 Gee es A ee © B A OS ae B A 2 ee B Formula 1. Formula 2. Formula 3. Which of these is the correct formula can be ascertained by de- termining how the olefine absorbs compounds of accepted polarity such as H+—OH, H+-—Cl, H+-—Br, H+-—I, H+ —NHR, and H+—OR. A compound having formula (1) will absorb HBr, for example, to give —C-—B exclusively. — H A compound having formula (2) will absorb HBr to give a mix- ture of H H H H A—C+-—C—B and A—C—+C-B + - + H Br Br H A compound having formula (3) will absorb HBr to give H 4H A-—C—+ e —B exclusively. _- = H Br Principle II. The Determination of the Charge on a Carboxyl Group.—Three methods have been employed. 1. In a few cases ketenes are known that have the general for- mula R—C=C_O. They always have the electronic formula R—C=ZiCL=—O because they yield acids, anilides, amines, and esters when treated with water, aniline, ammonia, and alco- hol, respectively. This is an application of Principle I. In such cases the carboxyl group is quadruply positive." 2. Many carboxylated compounds lose CO. when heated. Since CO, must have the formula O = { C{ =O, such compounds contain a quadruply positive carboxyl group.” 16 Hanke, M. T., and Koessler, K. K., J. Am. Chem. Soc., 1918, xl, 1730. 17 Fry, H.S., J. Am. Chem. Soc., 1912, xxxiv, 664; 1914, xxxvi, 248. 206 Studies on Proteinogenous Amines. XIII 3. Some carboxylated compounds split off formic acid when they are heated to 100-150° with dilute sulfuric acid.!® Since formic acid must have the formula the carboxyl group from which it was derived must be triply positive and it must have been attached to the neighboring car- bon atom as follows: 4. The above carboxylated compounds give off carbon monox- ide when they are warmed gently with fuming sulfuric acid. Carbon monoxide must have the formula ~C==0O. The ecar- bon is identical, electrically, with the carbon in formic acid. Its evolution has a significance identical with that ascribed to formic acid in the preceding paragraph. A similar decomposition occurs with phosphoric acid. The results are less reliable, however, because much higher temperatures are required to start the decomposition. Principle III. Hydrolysis.—Frequent use has been made of the principle of hydrolysis which has been so thoroughly discussed by others!® that a repetition seems superfluous. Assumptions. 1. The charges on the carbon atoms are definitely polarized in at least the large majority of organic compounds. There has been a tendency in recent years to assume that organic compounds are non-polar.?? An entirely non-polar com- 18 Hanke, M. T., and Koessler, K. K., J. Am. Chem. Soc., 1918, xl, 1726. 19 Nelson, J. M., Beans, H. T., and Falk, K. G., J. Am. Chem. Soc., 1913, xxxv, 1810. Selivanow, Ber. chem. Ges., 1892, xxv, 3517. Stieglitz, J., Am. Chem. J., 1896, xviii, 756. Noyes, W. A., J. Am. Chem. Soc., 1913, xxxv, 769. *° For a criticism of this view see Falk, K. G., and Nelson, J. M., J. Am. Chem. Soc., 1914, xxxvi, 209. M. T. Hanke and K. K. Koessler 207 pound would be chemically inactive. It may, of course, be true that some or even most of the molecules are really non-polar but such molecules can be of no importance in initiating chemical reactions. Such inactive non-polar molecules must always be in equi- librium with a certain number of active polar molecules. These active molecules are responsible for the chemical activity of the compound and they are very definitely polar! The present paper is limited to a discussion of the active polar molecules. 2. The electrical charges hold their places until they are forced to shift because of a definite overwhelming strain. Such shifts are infrequent, belong to classes of compounds rather than to individuals, and must never be assumed to have occurred unless this can be undeniably proved.” References. It will be necessary to refer frequently to the carbon atoms and to the bonds between the carbon atoms. Instead of placing numbers on the carbon atoms in each of the formulas, which would lead to confusion, the general rule has been adopted of numbering the carbon atoms mentally from right to left in a horizontal formula and from the top down in a vertical formula. A concrete example will make this clear. Crotonic acid has the structural formula CH;—CH = CH—COOH. The carboxy] ear- bon is carbon atom 1, and the double bond is the (2-3) bond. The facts, upon which the electronic formulas are based, can be verified by referring to any large handbook of organic chemistry such as Beilstein’s Handbuch der Organischen Verbindungen and Richter’s Organic chemistry. A Consideration of the Formulas. Propiolic Acid (H-C=C—COOH). When the potassium salt of this acid is heated with water it decomposes smoothly according to the following equation. 21 Lewis, G. N., J. Am. Chem. Soc., 1913, xxxv, 1448. 22 Examples of such shifts can be found in the numerous papers by Stieglitz and his collaborators on the electronic interpretation of the Beckmann rearrangement. 208 Studies on Proteinogenous Amines. XIII 2HC — C — COOK + H,0 — 2C.H:2 + CO2 + K.CO; Since the carboxyl group is eliminated as COs, carbon atom 1 must be quadruply positive. When propiolic acid is treated with the halogen acids, only the B-halogenated acrylic acids are formed. The exclusive forma- tion of beta derivatives proves that the triple bond is unsymmetri- cally polar, the electrons being attached to carbon atom 2. The entire formula for propiolic acid, is, therefore, += Waa H+—C+—-C-+cCTf7o (1) + “ OH 6-Halogenated Propiolic Acids (X -C=C—COOH). The halogenated propiolic acids decompose readily into X -C= C—H and COs. The carboxyl group is, therefore, quadruply positive. The halogenated propiolic acids react with the halogen acids to give beta derivatives exclusively. Thus £-chloropropiolic acid gives 6-dichloroacrylic acid with HCl. This proves that the triple bond is unsymmetrically polar so that the electrons are attached to carbon atom 2. The entire formula for this class of compounds is == 4+ — X= CG Cae (2) +— + OH The carbon atoms in these acids are identical, electrically, with the carbon atoms in malonic acid. ° Allylene (CH;—C=C—H). Allylene is readily absorbed by the halogen acids, the products being CH;—CCl,—CH;, CH3;— CBre—CHs3, and CH3;— Cl,.—CHs. In every case the negative halogen attaches itself to the central carbon atom. This unsymmetrical addition of the halogen acids proves that the triple bond is unsymmetrically polar, the electrons being attached to carbon atom 1. M. T. Hanke and K. K. Koessler 209 When allylene is heated with dilute sulfuric acid, acetone is formed. Here again the negative oxygen attaches itself to the central carbon atom. Acetone has been proved™ to have the formula Both of the end-carbon atoms.are quadruply negative. The methyl group in methyl acetylene must, therefore, also be quad- ruply negative. The entire formula for methyl acetylene is H ae « = _ = H+—¢C—+ C=-—C—-+H (3) — — = — H The carbon atoms in allylene are identical, electrically, with the carbon atoms in acetone. Tetrolic Acid (CH;3—C=C—COOH) and Its Relation to Aceto- acetic Acid. . When tetrolic acid is heated to 210° it decomposes into allylene and carbon dioxide. +at-o6 — +—- Ls CH, -C -C-COOH—> H.C-+C+-C-+H+0 7707 + The complete electronic formula for tetrolic acid is, therefore, H a = = gre H+—C-+¢C+-c-4ct7o (4) ~ +— + o> a H OH 23 See foot-note 2. This paper also contains the proof of the formulas for acetic, acetoacetic, acetone dicarbonic, and citric acids. 210 Studies on Proteinogenous Amines. XIII Since the above decomposition occurs at 210° it alone would not suffice to establish the formula for tetrolic acid. That the above formula is actually correct is proved by the following reactions for tetrolic acid. It combines with the halogen acids to give $-halogenated te- trolic acids. When heated to 105° with a concentrated solution of potassium hydroxide, it decomposes completely with the formation of large quantities of acetone and carbon dioxide and a small amount of acetic acid. This strongly suggests the intermediate formation of acetoacetic acid. This is to be expected from the assigned formula, because the carbon atoms in acetoacetic and in tetrolic acids are electronically identical. HO =4 Ch 0 HOS + Ceo ar ai eee ate Ce Hp ¢ = a is (iar H Ha + +2 > eee caer Son CTO + + (Glale CH; The acetoacetic acid would then give acetone and COs, in pre- ponderance, and acetic acid in small amounts as is its habit. Acetylene Dicarboxylic Acid (HOOC —C=C—COOH). When this acid is warmed with water it decomposes into propi- olic acid and carbon dioxide. $V += a Seb b= tem +— HOOC ~C-C-COoH = OT tot 70 4H+-C+-C-4+CF70 += - OH Both carboxyl groups are quadruply positive. The formula for acetylene dicarboxylic acid is 2 +- OF 104 =6 = Germs (5) - +- +t HO OH M. T. Hanke and K. K. Koessler 211 Acrylic Acid (CHy = CH—COOH). The recorded reactions for this substance are: 1. It unites with HCl and HI to give 6-chloro- and 8-iodo- propionic acid, respectively. 2. It gives 8-oxypropionic acid, not lactic acid, when heated to 100° with an aqueous solution of sodium hydroxide. 3. When ethyl acrylate is heated with ethyl alcohol containing some sodium ethylate, 6-ethoxypropionic acid is formed. These three entirely one-sided reactions prove that the charges constituting the double bond are polarized so that the electrons are attached to carbon atom 2. ; When potassium acrylate is fused with potassium hydroxide, hydrogen, potassium formate, and potassium acetate are formed. This reaction is obviously an hydrolysis at the double bond in which two positive hydrogen atoms are united with the alpha and two negative hydroxyl radicals with the beta carbon atom in exactly the manner that would’ be expected from the charges assigned to the double bond. CH: — CH — COOK + H;0 — H.C—O + CH; — COOK The formaldehyde formed at first reacts in the customary man- ner with the potassium hydroxide to give hydrogen and potassium formate. H.C = O + KOH -—> H, + HCOOK The most important fact from our point of view, is the forma- tion of acetic acid. The formula for this substance has been proved”? to be H -- H+-C-+C{70 —- + + — H OH It contains a quadruply positive carboxyl group. The carboxyl group in acrylic acid must, therefore, also be quadruply positive. This leads to the following complete formula for acrylic acid. 212 Studies on Proteinogenous Amines. XIII Eis) a ee Hi Cl GG ieee (6) OH Cinnamic Acid (CsH;—CH = CH—COOH). When cinnamic acid is treated with HBr or HI, 6-halogenated phenylpropionic acids are formed exclusively. Hypochlorous acid is absorbed so that the negative hydroxyl group attaches itself to the beta carbon atom the product being phenyl a-chloro, B-lactic acid. The exclusive formation of beta derivatives proves that the double bond is unsymmetrically polar so that the electrons are attached to the alpha carbon atom. When cinnamic acid is fused with potassium hydroxide, ben- zoic acid and acetic acid are the products, the negative OH group being again, at this high temperature, attracted to the beta posi- tion. The formation of acetic acid proves the carboxyl group to be quadruply positive. The formula for cinnamic acid is CH—CT7c-+cP70 —TT + + + 5 H H OH The remaining unsolved bond is probably (+—), because cin- namic acid gives ortho and para derivatives with HNO; exactly as is the case with phenol, aniline, and the halogenated benzenes in which the substituting group is surely negative. The entire form- ula for cinnamic acid would then be Cn 0 ere aaa (7) + + = H H OH It is electronically identical with acrylic acid, a positive CsH; group having replaced the positive hydrogen. B-Chloro-, Brom-, and Oxyphenylpropionic Acid (CsH;—CHX — CH,—COOH). The formulas for these substances follow as a natural corollary from that of cinnamic acid as can be seen from the following equation. M. T. Hanke and K. K. Koessler 213 C.H; , C,H; -- H+—C+ H+=—C+-—x + f= + EA 2 = ie (8) H+—C¢c— 4H H+-C-—+H a +- ms +- He Ot 6 HO—+-Cy _O (8) is the formula for 8-chloro-, brom-, or oxyphenylpropionic acid. Crotonic Acid (CH;—CH—CH—COOR). Crotonic acid combines with hydrogen iodide to give 6-iodo- butyric acid, with hydrogen bromide to give 6-brombutyric acid and with sodium ethylate to give @-ethoxybutyric ester. In every case the negative radical attaches itself to the beta carbon atom. The double bond is, therefore, unsymmetrically polar, the electrons being attached to the alpha carbon atom. * When crotonic acid is fused with potassium hydroxide it is hydrolyzed at the double bond to give acetaldehyde and acetic acid. The acetaldehyde formed at first is not liberated as such but reacts with KOH in the customary manner to give acetic acid and hydrogen. CH; — CH — CH — COOK + KOH — CH; — CHO + CH; — COOK CH; — CHO + KOH — CH;— COOK + H:z At high temperatures the double bond is still unsymmetrically polar so that the negative OH groups attach themselves to the beta carbon atom. The important fact is the formation of 2 molecules of acetic acid because this proves the polarity of the (1-2) and the (38-4) bond. Acetic acid contains a quadruply positive carboxyl group; so the carboxyl group in crotonic acid must also be quadruply positive. The second molecule of acetic acid, coming as it does from the acetaldehyde formed at first, proves that the (8-4) bond must have an electrical polarity like that of acetaldehyde; namely, (C—+C). The entire formula for crotonic acid is 214 Studies on Proteinogenous Amines. XIII + + + H+—-C-+CtT7C-+cC770 (9) 4- =. H OH Crotonic acid contains the acrylic acid nucleus. B-Oxy- and B-Halogenated Butyric Acids (CH;—-CHX—CH,— COOH). The facts that are needed for the proof of these formulas have already been given under crotonic acid. They will be repeated here for the sake of clearness. When crotonic acid is heated with an aqueous solution of sodium hydroxide, 6-oxybutyric acid is formed. When crotonic acid is heated with hydrogen bromide or iodide, B-brom- and £-iodobutyric acids are formed, respectively. Elec- tronically these reactions proceed as follows. CH; CH; + elt Or 130 Dt Ons ee + os + = = = (10) H+-—-C-— +H H+—-C-—+H + -- .46+= ~.etr HO C20 HO-+C7 = Formula (10) is that of 6-oxy, 6-ethoxy, and $-halogenated bu- tyric acids. The carbon atoms are electrically identical with the carbon atoms in crotonic acid. It is not surprising, then, to find that they readily lose HX to give crotonic acid. Glycollic Acid (CH,OH —COOH). When glycollic acid is warmed with concentrated sulfuric acid it decomposes into carbon monoxide, water, and trioxymethylene. 3 CH,OH — COOH — 8CO + 3H:0 + (CHO); In this case the carboxyl group is split off as carbon monoxide and the formaldehyde formed at first polymerizes to trioxymethy- lene. The carboxyl group must, therefore, be triply positive. The complete formula for glycollic acid is M. T. Hanke and K. K. Koessler 215 H -- H4+-C+-C}7O0 (11) + + OH OH This formula is substantiated by the preparation of glycollic acid from formaldehyde and prussic acid. EL MEM eT ges tit cits H af i i =f ce : Biers a aioe H+-C+-O- + H+-C+-N—>H+-C+-—C+-N a i = = The nitrile then gives the acid on hydrolysis which can produce no change in the electrical conditions of the molecule. Malonic Acid (COOH —CH.—COOH). When the acid is heated above its melting point, 132°, it decom- poses smoothly into CO, and acetic acid. The evolution of CO, proves that at least one of the carboxyl groups is quadruply positive. Acetic acid has been proved to contain a positive carboxyl group; * hence the second carboxyl group in malonic acid must also be quadruply positive. The entire formula for this substance is, therefore, +—- HO= + Gay = © tone Hee eels (12) a HO-+C 770 Tartronic Acid (HOOC—CHOH—COOH). When tartronic acid is heated above its melting point, 184°, it decomposes smoothly into carbon dioxide and a polymer of gly- collic acid. The evolution of CO, proves that one of the car- boxyl groups is quadruply positive. Glycollic acid has been proved to have a negative carboxyl group. The formula for tartronic acid must then be 216 Studies on Proteinogenous Amines. XIII H = OTC es Come (13) ae + +t HO Gm OH The decomposition can be represented electronically as follows HO-+CT7O 2 H + - H+—-C+-O0H+18" > O>fct70+H+-C+-cfT7o0 + + + HO-+Ct 70 i) OH OH (2) (3) Mesoxalic Acid (HOOC—CO—COORH). When this acid is heated above its melting point it decomposes into carbon dioxide and glyoxylie acid. The evolution of CO, proves that one of the carboxyl groups is quadruply positive. The formula for glyoxylic acid cannot be proved from the recorded reactions of that substance; but the following indirect proof seems rather convincing. The formula for chloral must be Cl + CCS en me - + CSE because it gives chloroform and sodium formate when treated with aqueous alkalies. If it were possible to substitute OH groups for the three negative halogen atoms without separating the carbon atoms, glyoxylic acid would be formed. ‘The formula for glyoxylice acid should then be CeCe cenae - + HO H From this one would conclude that one of the carboxyl groups in mesoxalic acid was triply positive and that the entire formula for this substance is M. T. Hanke and K. K. Koessler oii HO—gct—o eB . 0) (14) HO—+ O a — ot ae ei 28 i One assumption has been made; namely, that the substitution of negative oxygen for negative chlorine in chloral has produced no change in the polarity of the charges binding the carbon atoms together. That we were justified in making the above assumption, and that the formula assigned to mesoxalic acid is really correct is proved in a striking manner by the fact that the acid decomposes on boiling with water into carbon monoxide and oxalic acid. Electronically this reaction would be represented as follows. 43 HO-+c770 cTto = t +) Tero a HO-+C} 70 HO-+C7—0 The carbon monoxide might arise from either the central or the triply positive end-carbon atom. In any case zt would be im- possible to obtain CO and oxalic acid from a molecule having any but the above electronic structure. Lactic Acid (CH;—CHOH —COORH). When lactic acid is treated at 60° with fuming sulfuric acid, the carboxyl group is eliminated as carbon monoxide. A similar decomposition occurs when lactic acid is heated to 130° with dilute sulfuric acid, the products being acetaldehyde and formic acid. CH, — CHOH — COOH > H.C-+CT70+H+-CTTO - - +. = H OH ae (1) 218 Studies on Proteinogenous Amines. XIII From the above equation the formula for lactic acid is easily seen to be Hi H + + H+—-C-+C+-CcT70O (15) = eee + — — H OH OH B-Oxypropionic Acid (CH,OH —CH,.—COOH). When acrylic acid is heated to 100° with an aqueous solution of sodium hydroxide, B-oxypropionie acid is formed. Electroni- cally, this reaction can be written as follows. Jel H + + H+-C+ — OH H+-—C+-—-—OH + + _ 2 _ (16) H+-C-— + +H H+-—C-—+H ae oa HO-+C07—0 HO-+C770 The carbon atoms in 6-oxypropionic acid are identical, elec- tronically, with those in acrylic acid. It is not surprising to find, then that B-oxypropionic acid readily loses water to give acrylic acid. To obtain acrylic acid from lactic acid, a far reaching electronic rearrangement is necessary and one would expect that to bring about this transformation, a more drastic treatment would be required. This is the case. The difference in ease of activity between the a- and 6-oxy (and halogenated) deriva- tives is readily explained, in this manner by the electronic formulas. Pyruvic Acid (CH;—CO—COOH). When pyruvic acid is heated to 150° with dilute sulfuric acid it is hydrolyzed, the products being acetaldehyde and carbon dioxide. CH; — CO- COOH + H.0 > H,C-+CT7O0+0> 7 Cf 70+H,0 + H (3) (2) (1) M. T. Hanke and K. K. Koessler 219 From the above partially electronic equation the entire formula for pyruvic acid is readily seen to be ist -e H+-C-+C-+cCT 70 (17) — ++ + + aoe — H O CH Both pyruvic and oxalic acids contain the electrical dyad a5 + AAe f : (¢ C—+C f). They are similarly hydrolyzed by sulfuric acid. Malic Acid (HOOC —CH,—CHOH —COOH). When malic acid is heated with dilute sulfuric acid, it decom- poses into COs:, formic acid, and acetaldehyde. When malic acid is warmed with fuming sulfuric acid it decomposes quanti- tatively according to the following equation. HOOC — CH: — CHOH — COOH —~ HOOC — CH, — CHO (A) + HCOOH t { C;H,0. — COOH (B) CO+H:20 Compound (B) is a polymer of (A). When (B) is boiled with dilute sulfuric acid it decomposes into CO, and acetaldehyde. Although compound (A) has never been isolated—it polymerizes too readily—its formation as an intermediate product in the decomposition of malic acid by sulfuric acid is proved conclu- sively by the character of the condensation products formed when malic acid, sulfuric acid, and phenols are heated together. This reaction proves the entire formula for malic acid. Since CO is evolved—or formic acid with dilute sulfuric acid—one of the carboxyl groups must be negative. The negative car- boxy] group is attached to carbon atom 1 (see the above equation). The evolution of CO. from the polymer of HOOC—CH:—CHO proves that the other carboxyl group is quadruply positive. The other product is acetaldehyde, a compound of known electronic constitution (CH;—-+CHO0). 24von Pechmann, H., and Welsh, W., Ber. chem. Ges., 1884, xvii, 929, 1649, 220 Studies on Proteinogenous Amines. XIII - The entire formula for malic acid is H OH ana — =f oe + — O C0 2 ea0 4 eee (18) HO H H OH This proof is rendered clear by expressing the decomposition electronically. H OH H a5 rf == —+ ee _ct- —-+o4-¢G-4ct-o0 4+cet- eS tan le C= ae C-—+C] 10 + 2@aaam -~- + + = - + + HO H H OH HO ial H 1 H Ey —+ot olfcyTrloO+H+- C- +cf7o = + H H Oxalacetic Acid (HOOC —CH,—CO—COOR). The formula for this substance is clearly proved by the follow- ing decompositions of its diethyl ester. When boiled with alka- lies it is hydrolyzed so that the products are oxalic acid, acetic acid, and aleohol. When boiled with dilute acids it is hydrolyzed so that carbon dioxide and pyruvic acid are formed. When heated under ordinary pressures it decomposes into carbon monoxide and malonic ethyl] ester. ; Only one formula can be written that readily explains all of these decompositions; namely, ) H ‘ z) | o> fc+-c-+c-4+cT 70 (19) + = ++ + HO it O ‘OH M. T. Hanke and K. K. Koessler 221 As ean be seen, an hydrolysis at bond (2-3) would give acetic and oxalic acids; at bond (3-4) would give carbon dioxide and pyruvie acid. The loss of carbon atom 2 would give carbon monoxide and malonic acid. Tartaric Acid (HOOC—CHOH —CHOH—COOBH). Tartaric acid reacts readily with concentrated sulfuric acid or phosphoric acid. The several observers seem to have obtained different products when operating under slightly different con- ditions. Thus Vangel®> found that when tartaric acid is heated to 150° with phosphoric acid, equal volumes of CO, and CO were formed. He did not examine the non-gaseous residue, nor did he measure the exact quantity of gas obtained per unit weight of tartaric acid. Bouchardat®® found that when tartaric acid is warmed to 40-50° with fuming sulfuric acid containing 80 per cent of SO:, a gas is formed that is composed of 4 parts of CO and 1 part of SO2, and which contains from 2 to 4 per cent of carbon dioxide. He states, moreover, that the carbon dioxide and SO, appeared toward the end of the reaction. An examination of the non-gaseous residue revealed the presence of some racemic tar- taric acid and small amounts of glycollic and pyruvic acids. He gives no data by means of which one could calculate how many of the carbon atoms of tartaric acid were evolved as carbon monox- ide. Since these data were necessary to establish the electronic formula for tartaric acid, the following experiment was carried out. Tartaric acid—1.5000 gm., M.P. 170°—was mixed in a 50 ce. round bottomed flask with 25 ec. of fuming sulfuric acid containing 18 per cent of SO3. The flask was connected to a 1,000 ce. narrow mouthed precision cylinder that had been inverted and arranged so that the evolved gases could be collected by displace- ment of water. The flask was heated to 65°. A gas was slowly and steadily evolved that was pure carbon monoxide at first. Not a trace of either SO, or COs was present until the gas vol- ume, corrected for temperature, pressure, and aqueous tension was 445 ec. At this point the evolution of gas becomes livelier. 25 Vangel, B., Ber. chem. Ges., 1880, xiii, 356. 26 Bouchardat, M. G., Bull. Soc. chim., 1880, xxxiv, series 2, 495. 922 Studies on Proteinogenous Amines. XIII Considerable SO. was present as evidenced by the white fumes that formed when the gas came in contact with the wet walls of the cylinder. The reaction was practically over when about 950 ec. of gas had been collected. The volume shrunk to 840 ee. on agitation with water. This removed all of the SO, and probably also some of the CO: A subsequent agitation with aqueous sodium hydroxide gave a final volume of 800 cc. of gas that proved to be pure carbon monoxide. The sulfuric acid was colored very pale yellow after the reac- tion was over so that no charring had occurred. An examination of this reaction shows the following. Carbon monoxide (445 ec.) was evolved from 1.5 gm. of tartaric acid before either SO, or CO, were formed. This amount of CO was liberated by the primary action of the sulfuric acid before an oxidation reaction had occurred. This quantity of tartaric acid should give about 225 ee. of CO per carbon atom or 900 ce. if all of the carbon atoms had been converted into CO. The 445 ce. actually obtained represent two carbon atoms, half of the tar- taric acid molecule. Since two of the four carbon atoms were evolved as CO before an oxidation reaction had occurred, it is fairly safe to conclude that both of the carboxyl groups were eliminated as CO according to the following equation (CHOH), (COOH): — 2CO-+ CHO — CHO + 2H;0 Both of the carboxyl groups must, therefore, have been triply positive. The assumption that glyoxal is an intermediate prod- uct in the decomposition of tartaric acid by fuming sulfuric acid is in perfect agreement with the observed fact that all of the car- bon atoms in tartaric acid are finally evolved as CO. A primary oxidation of glyoxal would yield glyoxylice acid which would then be converted largely into CO and to a small extent into oxalic acid, CO, and COs. CHO — CHO + H.S80, 1.{CHO — COOH \HCOOH A ae — COOH + H,S80, “'\COOH — COOH CHO — COOH + SO; + H.0 HCOOH + CO CO + H.0 COOH — COOH + SO, + H20 CO + COz Pele dean M. T. Hanke and K. K. Koessler 223 The evolution of so much CO and so little CO, could not be ex- plained at all if the assumption wete made that even one of the carboxyl groups in tartaric acid was quadruply positive. The complete electronic formula for tartaric acid and its primary decomposition into CO, glyoxal, and water can be represented as follows. tigits poor 7 =e H+-—C+-—OH + at =+ 4+ — = (0) => 2 Fo Oreo eee = pew a Peeler on.” ea Te i oe es + + + aces E> cel ee Se (3) 2) Propionic Acid (CH;—CH,—COOH),. That the carboxyl group in propionic acid is quadruply posi- tive is proved by the absorption reactions of methyl ketene, CH;:—CH—C_—O. Methyl ketene unites with water to give propionic acid, with alcohol to give ethyl propionate, and with ammonia to give propionamide. In every case the negative rad- ical attaches itself to carbon atom 1. This proves that the double bond is unsymmetrically polar; that the partial formula for me- thyl ketene is eat CHC" Ber so $V 4 and that propionic acid has the partial formula H + ~c-40t- ae fs — — H OH The polarity of the remaining unsolved bond cannot be determined directly. The following partially speculative proofs seem con- vincing to the authors. 224 Studies on Proteinogenous Amines. XIII Just two formulas are possible, namely, H H H H + + + + () H+-C+-C-4+CtT7O @ H+-C-4+O-4+C7 70 ent A dee =) eee lege strane: t+ + = H H OH H H OH It is, of course, possible that both of these formulas are correct and that the two electrons between carbon atoms 2 and 3 are vibrating so that they are attached first to one and then to the other carbon atom (or they may hold an electron in common). Although the possibility of such an electrometric equilibrium is undeniable, we have no right to assume its existence without proof. Most of the simple compounds examined so far have shown a rigid polarity. In the absence of proof to the contrary we will assume that the same rigidity exists in this compound. There are two proofs that formula (1) is representative of the électrical conditions existing in the molecule of propionic acid. 1. Acrylic acid has been proved to have the formula H + + Hy eee ne cae OH By reduction acrylic acid gives propionic acid. The reduction consists in the addition of two electrons to the beta carbon atom. The formula for propionic acid should then be adel H + + H+-C+-C-+4+C770 (21) - = + a + = H H OH One assumption must be made; namely, that an electronic shift does not take place within the molecule after the electrons and the hydrogen atoms have been added. 2. Propionic acid can be prepared from the sodium compound of acetoacetic or malonic ester and methyl iodide. The reactions can be represented electronically as follows. M. T. Hanke and K. K. Koessler 225 — H — a 5 Bo ¢ — + Cn; . — +Na +I-+C-+H-— My + H,0 se = o-t+e¢ C+ “ -+4 = H ie C-—-+H ai = H H-+—C— +H +¢H,—+ coon — CH; This formula is also obtained by the similar reaction of methyl iodide with the sodium compound of malonic ester and is identi- cal with formula (1). Butyric Acid (CH;—CH,—CH,—COORH).. The carboxy] group in butyric acid is quadruply positive. This can be proved from the absorption reactions of ethyl ketene (CH;—CH.—CH_C_O). Ethyl ketene absorbs water, al- cohol, and aniline so that the negative OH, OC.H;, and NH— C,H; groups attach themselves to carbon atom 1, the products being butyric acid, ethyl butyrate, and butyric acid anilide. These entirely one-sided reactions prove that the double bond in ethyl ketene is unsymmetrically polar so that the electrons are attached to carbon atom 2. The partial formula for ethyl ketene is, therefore, rom: ery ohamme ae) +¥ + aE H The partial formula for butyric acid must then be A a+ CH; —CH,—C-+CT 70 — + a — H OH THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 226 Studies on Proteinogenous Amines. XIII The remaining bonds can be solved only by speculation as follows. Crotonic acid has been proved to have the formula +-ciact- + + 3 H H OH By reduction, crotonic acid gives butyric acid. The reduction consists in the addition of two electrons to the beta carbon atom. The formula for butyric acid should then be Sareea G (22) + + H H H OH Although this must certainly be the formula at the moment of reduction, the objection might be raised that the electrical charges could shift after the hydrogen atoms and the electrons had been added. To assume that such a shift did occur would be unrea- sonable unless the molecule so formed contained electrical dyads of uncertain stability. On the other hand, if the certain stability of every dyad in the molecule could be proved, the above formula should be considered correct until some definite proof to the contrary was found. The undoubted stability of every dyad in this formula is proved by considering the electronic skeleton of the acrylic acid molecule. DT ase as - = > 3 4 BOR OO irri eal aa (Acrylic acid.) (Butyric acid.) The dyad +C-++-—C— appears once in the electrically very stable acrylic acid and twice in the assigned formula for butyric acid. The formula assigned to butyric acid is, therefore, not only possible but extremely probable. Phenylpropionic Acid (CjH;—CH,—CH2—COOH). The method used in proving the formula for phenylpropionic acid is almost identical with that just described for propionic M. T. Hanke and K. K. Koessler 227 acid. A detailed proof is withheld to avoid unnecessary repeti- tion. The formula is H 4H + + CHs+—-C+-—C-+CT 70 (23) if = air — = H H OH The charge on the benzene ring is given as positive because phenyl- propionic acid gives ortho and para derivatives just as is the case with all of the negatively mono-substituted derivatives of ben- zene. Phenylpropionic acid is electronically identical with propionic acid, a positive hydrogen atom having been replaced by a positive phenyl group. Two New Electronic Principles. Principle IV.—The polarity of the double bond in an unsaturated aliphatic compound is determined by the electrical charge on the first substituting group. An examination of the formulas just presented has led us to conclude that the polarity of the double bond in an unsaturated aliphatic compound is determined by the electrical charge on the first substituting group which we will call the directing group. There are two sharply defined types of unsaturated compounds, those belonging to the vinyl type |S OF ae A(t ae 4) 2 Sales and those belonging to the acrylic acid type Gh co 2488 2 ae a H When the first substituent is negative the compound belongs to the vinyl type. Such a compound will always give alpha deriva- tives by the absorption of a polar compound like HCl. When the first substituent is positive the compound belongs to the acrylic acid type; it gives only beta derivatives by the absorption of polar compounds like HCl. Beside these two simple types, one very stable and common mixed type is possible; namely, V— pO G—-h; A com- 228 Studies on Proteinogenous Amines. XIII pound belonging to this type is both a vinyl and an acrylic acid derivative. Such compounds are electrically very stable be- cause their polarity is fixed by two groups both of which are striving to produce the same electrical configuration. Of the compounds discussed so far, allylene belongs to the vinyl type, propiolic acid to the acrylic acid type, and crotonic, cinnamic, tetrolic, and the 6-halogenated propiolic acids belong to the mixed type. The chemical properties of many unsaturated compounds are too incompletely known to make it possible to prove their electronic formulas completely. In such cases Principle IV is frequently of service because it gives us a means of writing a fairly authentic formula on the basis of a few absorption reac- tions. Examples of such usage will be given presently. Principle V.—A carbon atom that is attached to an oxygen atom will become at least doubly positive, when this is at all possible. Glycollic acid is the simplest example of the truth of this prin- ciple. The electronic formula for acetic acid is H ate H+-C-+C770 - + + — H OH When one of the paraffin hydrogen atoms is replaced by an OH group, the compound formed, glycollic acid, has the formula’ H + H+—-C+-CT—0 + + OH OH The introduction of one OH group has caused carbon atom 2 to lose four electrons. From this it would seem that a triply nega- tive carbon atom is electrically unstable when it is combined with oxygen. The triply negative carbon atom will, when this is possible, lose two electrons to a neighboring carbon atom. Of the compounds discussed so far, citric, tartronic, lactic, malic, and tartaric acids can be cited as illustrations of the truth of this principle. All of these compounds are a-hydroxy-acids and 4 M. T. Hanke and K. K. Koessler 229 they all contain a negative carboxyl group. In all of these cases, the alpha carbon atom, which is united with the OH group, is at least doubly positive. The real value of Principles IV and V lies in the fact that they give us a means of accurately predicting the behavior of certain classes of organic compounds and of establishing the formulas for some compounds whose chemical properties are too incompletely known to make a complete proof of their electronic formulas possible. We will now consider a few such examples. Propylene (CH2_CH—CH;). In all of its absorption reactions, propylene gives only isopropyl derivatives. Thus HCl, HBr, and HI give isopropyl chloride, bromide, and iodide respectively. Sulfuric acid gives isopropyl sulfate which by hydrolysis gives isopropyl alcohol. This proves that the electrical charges constituting the double bond are un-. symmetrically polar, the electrons being attached to carbon atom 3. H+—C77C—CcH, aE + H H The polarity of the remaining unsolved bond cannot be deter- mined directly. That the CH; group carries a negative charge is rendered highly probable by the following line of reasoning. The polarity of the double bond in propylene is like that in the vinyl compounds. Propylene must, then, be a vinyl derivative 3. hence by Principle IV, the methyl group is negative. The com- plete formula for propylene is H+—C77C+-—cH, (24) ses ae Ho Set That an unopposed methyl group normally carries a negative charge is clearly shown by an examination of some aromatie and aliphatic compounds. Toluene gives ortho and para derivatives almost exclusively just as is the case with phenol and chloroben- zene. The methyl group, like the OH and C] radicals, functions negatively. Negative methyl] groups have also been encountered 230 Studies on Proteinogenous Amines. XIII in acetic acid, acetone, acetoacetic acid, tetrolic acid, allylene, acetone dicarbonic acid, and citric acid. Propioni¢ acid is the only compound that we have encountered up to date that may contain a positive methyl group. Allyl Alcohol (CH,_H—CH.OH). When allyl alcohol is heated with a saturated solution of po- tassium acid sulfite, CH,SO;H—CH,—CH,OH is formed, the negative SO3H group going to the beta carbon atom. When allyl alcohol is treated with hypochlorous acid, CH» OH —CH,Cl—CH:0H is formed, the negative OH group attach- ing itself to the beta carbon atom. This proves that the double bond in allyl alcohol is unsym- metrically polar so that the electrons are attached to carbon atom 2. The charge on the CH,OH group cannot be determined directly; but there is good evidence that it is positive and that the entire formula for allyl alcohol is H H H ec H+—C{7C-+C+0H (25) -- H The indirect proof of this statement is as follows. 1. Allyl alcohol gives only beta derivatives when it absorbs electrically polar compounds. In this respect it behaves exactly like acrylic acid, and it belongs, therefore, to the acrylic acid type, not to the vinyl type. The CH.OH group should, therefore, be positive like the carboxyl group in acrylic acid (Principle IV). 2. By Principle V, a carbon atom that is attached to oxygen will become at least doubly positive. The oxymethyl group in allyl aleohol should, therefore, be positive. 3. To assume that the oxymethyl group was negative would be unreasonable, because if that were true, the polarity of the double bond in allyl aleohol and in propylene should be identical. We can see no reason why an added oxygen atom should change the remote electrical charges on the molecule if it were incapable of changing the charge on the carbon atom to which it is attached. M. T. Hanke and K. K. Koessler Zar Allyl Chloride (CH,_H —CH.Cl). Allyl chloride combines with hydrogen chloride at 100° to give propylene chloride, CH;—CHCI—CH;Cl. Concentrated sulfu- ric acid is readily absorbed by allyl chloride at room tempera- tures. When the resulting acid sulfate is treated with water, the product is CH;—-CHOH—CH.Cl. The negative group attaches itself to carbon atom 2 in each case. These facts prove that for temperature ranges from 20 to 100° the charges consti- tuting the double bond are unsymmetrically polar, the electrons being attached to the beta carbon atom. The partial formula for allyl chloride is The charge on the CH.Cl group cannot be determined directly; but there is good evidence that it is negative and that the com- plete formula for allyl chloride is H. ee H+—-C_>7C+-cC+-Cl 26) + = H H The absorption reactions ‘of allyl chloride are identical with those of propylene and the viny! derivatives but the reverse of those of acrylic acid and allyl alcohol. Allyl chloride is, there- fore, a vinyl derivative; the charge on the CHCl group is nega- tive. The CH,OH group has just been shown to be positive. Although the negative oxygen atom will, when possible, repel two electrons from the carbon atom to which it is attached, this seems not to be the case with the negative chlorine atom. Stated differently, when a carbon atom is attached to an oxygen atom the carbon atom will be charged her in preference to Ore the latter form being rarely capable of existence. A carbon atom that is attached to chlorine is perfectly stable when it is charged 232 Studies on Proteinogenous Amines. XIII Si Gua This would indicate that triply negative carbon is not inherently unstable; it is the attached oxygen atom that renders it so. The negative charge assigned to the CH.Cl group is in perfect agreement with its directing force as manifested in the aromatic series. Benzyl chloride gives ortho and para derivatives on halogenation and nitration just as is the case with all of the nega- tively mono-substituted benzene derivatives. It might not be out of place at this point to call attention to the need of certain facts that would help greatly to establish the electronic formulas of some biologically very important compounds. 1. Since the CH.OH group in allyl alcohol is surely positive, it ought also to be positive in benzyl aicohol. Then benzyl al- cohol or any of the ethers derived from it should direct the second incoming substituent predominantly to the meta position. 2. It is very important for protein chemistry that we know how the NH» group affects the carbon atom to which it is attached. Triply negative carbon has been shown to be stable when attached to chlorine and unstable when attached to oxygen. Although CH,OH+—COOH is surely the formula for glycollic acid, CH;Cl— + COOH is probably, therefore, the formula for monochlo- roacetic acid. What is the formula for aminoacetic acid? This will be hard to prove directly because glycine is an extremely” stable compound; but the determination of the following facts would render an indirect proof possible. Does allyl amine give alpha or beta derivatives when it is treated with polar compounds? Does benzyl amine give ortho and para or meta derivatives? From these facts it would be possible to conclude whether the NH, group repelled electrons from the carbon atom to which it is attached as is the case with the OH group, or whether triply negative carbon is stable when attached to the NH» group as it is when attached to chlorine. 3. Does oxygen repel electrons from the carbon atom to which it is attached because of its strongly negative character? This could be determined by examining the properties of certain fluorine deriva- tives. Since fluorine is more decidedly negative than oxygen, it should be able to repel electrons from the carbon atom to which it is attached at least as certainly as oxygen 7f this repulsion was occasioned by the negative character of the fluorine atom. The oxy- M. T. Hanke and K. K. Koessler 233 gen atom may prove to be unique in this respect, which might throw some light upon the intra-atomic structure of this atom. The absorption reactions of allyl fluoride and the behavior of benzyl fluoride on nitration, etc., would settle this point. Acrolein (CH, _CH—CHO). When acrolein is treated at low temperatures with hydrogen chloride or hydrogen bromide, 8-chloro- and 8- brompropionalde- hyde are formed respectively. This proves that the electrical charges constituting the double bond are unsymmetrically polar, the electrons being attached to carbon atom 2. The partial formula for acrolein is H -- H+-C{7c-—ct To + + H H The charge on the aldehyde group cannot be determined directly; but there is good evidence that it is positive and thatthe complete formula for acrolein is H+-ct7c-+ct7o (27) + + + H H H 1. The absorption reactions of acrolein are identical with those of acrylic acid and allyl alcohol; but the reverse of those of the vinyl derivatives. Acrolein, therefore, belongs to the acrylic acid type; the charge on the aldehyde group is positive (Principle IV). 2. Since the charge on the carboxyl group of acrylic acid has been proved to be positive and that on the oxymethy] group of allyl alcohol was proved to be most probably positive, it would be strange indeed if the aldehyde group, which stands between the CH.OH and the COOH groups in degree of oxidation, should be negative. 3. Aldehyde groups are usually if not invariably positive. In glyoxylic acid, for example, where an aldehyde and a carboxy] group vie for the electron, it is the aldehyde group that is positive. In the aromatic series the aldehyde group directs the second in- coming substituent predominantly to the meta position as is the case with nitrobenzene. STUDIES ON PROTEINOGENOUS AMINES. XIV. A MICROCHEMICAL COLORIMETRIC METHOD FOR ESTI- MATING TYROSINE, TYRAMINE, AND OTHER PHENOLS. By MILTON T. HANKE anp KARL K. KOESSLER. (From the Otho S. A. Sprague Memorial Institute and the Department of Pathology, University of Chicago, Chicago.) (Received for publication, October 21, 1921.) INTRODUCTION. The discovery of the diazonium salts by Griess in 1866 opened up a most fruitful field for scientific investigation by giving the investigator a class of chemical compounds that are highly sus- ceptible to a large variety of chemical changes. Among these chemical reactions is one that is of particular value because it gives rise to colored compounds; namely, the ability of diazonium compounds to combine (couple) with imidazoles, phenols, and _amines in alkaline solutions. p-Phenyldiazonium sulfonate, usually called diazobenzenesulfonic acid, is particularly valuable for colorimetric determinations because it is comparatively stable in water solutions, it does not couple with itself to give a highly colored compound, it is easy to prepare, the dyes formed from it are sufficiently soluble in alkaline solutions to render the use of a colorimeter possible, and the colors produced, usually orange to red, are easily compared because they do not readily produce retinal fatigue. It is not surprising then, to find that p-phenyl- diazonium sulfonate has been used for some time as a qualita- tive test for imidazoles, phenols, and amines.2>*? Up to the present time two methods have been proposed for the use of this diazonium salt in the quantitative estimation of imidazoles;* ® 1Griess, P., Ann. chem. Pharm., 1866, exxxvii, 39. 2 Ehrlich, P., Z. klin. Med., 1882, v, 285; Char. Ann., 1883, viii, 140; Deutsch. med. Woch., 1883, ix, 549; 1884, x, 419. 3 Pauly, H., Z. physiol. Chem., 1904, xlii, 508; 1905, xliv, 159. 4 Weisz, M., and Ssobolew, N., Biochem. Z., 1913-14, lviii, 119. 5 Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1919, xxxix, 497. 235 236 Studies on Proteinogenous Amines. XIV but as far as is known to the authors this reagent has not been % used, heretofore, for the quantitative estimation of phenols. This paper contains a detailed description of methods for the micro- chemical colorimetric determination of phenol, 0-, m-, and p-cresol, p-oxyphenylacetic, p-oxyphenylpropionic, and p-oxyphenyllactic acids, tyrosine, and tyramine. : From the standpoint of color development, these phenols can be divided into three classes; namely, ; 1. Phenols in which the para position is not occupied by a second substituent. 2. Phenols in which the para position is occupied by a second substituent that does not contain an amino group. 3. Tyrosine and tyramine. When the position para to the phenol radical is not occupied, as in phenol, o-cresol, and m-cresol, the substitution with p- phenyldiazonium sulfonate occurs, at least predominantly, in this position. a H et H H a oo | aa SUNT LS i Ne re J ToS Tie. ag ae : HO¢ >H+HO-N=N-{_ >-S0:-ONa—HO¢ >-N=N-C_)-S0: ONa + HC a EL igo Ll H H He | The coupling is so rapid that about 95 per cent of the color has . appeared before a reading can be taken with a Dubosceq colori- meter. The color is fully developed in about 2 minutes and it does not change for about 3 to 5 minutes. In these cases the color is predominantly yellow with just a dash of red. When the position para to the phenol radical 7s occwpied, as in p-cresol and the aromatic hydroxy-acids, the substitution with p-phenyldiazonium sulfonate must occur in the ortho position. H H O H H O H H HY I Hes eo + HO-—N = N-—< >—-SO2 -ONa—> | = N= N-< >-S0:- ONa+H.0 a. / 7H HW ‘seme e CH: zt GHs The initial rate of coupling in this case, is somewhat slower than that of phenol; but a color of maximum intensity is,nevertheless, obtained in about 3 minutes. The colors produced are predomi- nantly red so that a neutral solution of Congo red can serve as a comparison standard. These colors are stable for from.3 to 5 ss Mu Hanke and K. K. Koessler 230 minutes. The stability of the color and the character of the change that the color undergoes after it has reached its maximum intensity is a fairly accurate index of the composition of the aromatic hydroxy-acid that is present, if only one of these acids is present. This is discussed in detail in the experimental section of the paper. Tyrosine and tyramine, in small concentrations, show an anoma- lous behavior toward p-phenyldiazonium sulfonate in alkaline solutions. When a solution of tyrosine or tyramine is first added to the alkaline reagent, a pink color begins to develop promptly as in the case of the aromatic hydroxy-acids. After about 30 seconds, however, the color changes sharply to yellow and fades. The intensities of the yellow colors so produced are not directly proportional to the amount of tyrosine or tyramine present. Obviously, then, the usual procedure is worthless for the estima- tion of either tyrosine or tyramine. The sharp change in color from pink to yellow suggested a chemical change, perhaps a tautomeric shift. In an effort to stabilize the pink color, or rather to revive it after the initial reaction had gone to completion, we added a small amount of a concentrated solution of sodium hydroxide. The color was in- tensified and some of the pink tint was restored; but again the color intensities were not directly proportional to the concentra- tion of the phenol. The phenomena noted were so similar to those that one would expect from a compound in which a tauto- meric equilibrium existed between a carbonyl form and an enol form that we thought to stabilize the carbonyl] derivative, which we believed to be most highly colored, by allowing it to react with hydroxylamine to form an oxime. The addition of hydroxylamine hydrochloride to the alkaline reagent did not give rise to an intensification of color; but when sodium hydroxide was added previous to the addition of hydroxylamine, a very intense bluish red color was produced which was directly proportional to the amount of tyrosine or tyramine added. Briefly then, the method for esti- mating either tyrosine or tyramine consists of 1. A reaction between tyrosine or tyramine and p-phenyldia- zonium, sulfonate in an alkaline—sodium carbonate—solution which gives rise to a primary yellow color. This reaction is allowed to proceed for 5 minutes. 238 Studies on Proteinogenous Amines. XIV 2. The addition of 2 cc. of 3.0 Nn NaOH, after the initial reaction period of 5 minutes, which intensifies the color, stops any further action between the phenol and the diazonium salt and converts the diazonium salt into a sodium diazotate which cannot couple with the hydroxylamine that is to be added later. The alkali is allowed to act for 1 minute. 3. The addition of 0.10 ce. of a 20 per cent solution of hydroxyl- amine hydrochloride, which gives rise to a very intense bluish red color that is stable for at least half an hour. An explanation of the chemical reactions involved must ac- count for the following facts: Tyrosine and tyramine are the only phenols, that we have examined, that give a color intensi- fication with sodium hydroxide and hydroxylamine. Since oxy- phenyllactic acid, which is structurally identical with tyrosine excepting that an OH group replaces the NH», group, does not give a color intensification with NaOH and NH.2OH, the NH» group must play a réle in the color development with these phenols. - A number of other compounds in which a tautomeric equi- librium exists between a carbonyl form and an enol form, for example acetaldehyde, acetone, and acetoacetic acid, also give an intense color under the conditions just described. The color produced is almost identical with that obtained with tyrosine or tyramine. We offer the following explanation tentatively for these color phenomena. Work is now under way to determine the correctness of the following formulations. Since the position para to the phenol group in tyrosine and tyramine is occupied by an alanyl side chain, the substitution with diazobenzenesulfonic acid must occur in the ortho position. H H H/\ :H+ HO: —N = N—CiHi-80r0Na Se Ge = N—C:H,-SOs: ONa Rae a 4 H, ae H, + H.0 f | ACNE i H, (A) COO Na r COO Na : f Compound A—the yellow phenol form—would be in tautomeric equilibrium with a small amount of the supposedly red quinoid form M. T. Hanke and K. K. Koessler 239 H O O HI H H/\—N = N—C,H.SO:-ONa H/\=N—N—-C.HSO;-ONa | W “ CH, — CH: | | HCNH, Bo (B) COO Na COO Na In a solution made alkaline with sodium carbonate, practi- cally all of the compound would exist in its phenol form A. The addition of a strong alkali, like sodium hydroxide, would give rise to the sodium phenate derivative of A which, because of its high degree of dissociation, would pass more easily into the quinoid derivative than the phenol itself. Na | O O \| H oy = N—C,H.SO:;:ONa Bann N—N-—C,H.802-ONa | }- | ang ia i ea +H.O = CH, 4+ NaOH | ole ae (B) COO Na Yellow. COO Na Red. The result would be an intensification of color with the influx of red. The amount of quinoid derivative formed under the influence of alkali must, nevertheless, be small because the color intensification is not very great. The addition of hydroxylamine hydrochloride to such a strongly alkaline solution would give rise to the following reactions. fe NH.OH-HCl1 + 2 NaOH — H. N—O Na+ NaCl+ H.20 O N-O Na | H | H H/\= N—N—CcH.SO2:O Na a3 N—N-C,H,SO:-O Na fr a 7 H Den Nal + H,N—ONa—-> ae + H.O0 HCNH, HCNH:2 | | COO Na COO Na 240 Studies on Proteinogenous Amines. XIV One would expect the quinoneoximehydrazone derivative (C) formed in this way to be dark red. Such o-quinoneoximehydra- zones seem not to have been studied; but the para derivatives have been known for some time. They all give deep red alkaline solutions. o-Quinonedioxime H ay ie H\ /=N-0H H which is a closely related compound, also gives a deep red alka- line solution.’ Just why tyrosine should give a highly colored quinoneoxi- mehydrazone while p-oxyphenyllactic acid does not give such a highly colored derivative, is a problem for future investigation. The réle played by the amino group in the side chain of tyrosine and tyramine is still obscure. The reactions that probably occur when certain aldehydes and ketones are treated with diazobenzenesulfonic acid, sodium hydroxide, and hydroxylamine can be represented by the follow- ing type equations. Ri Ri b=0 b=0 + H.0 R.—C :H, + 0: = N-N—C:H,SO,-0 Na— Re-C= N-N—CH.802-ONa Pale yellow. Ri Ri b-0 + HiN—O Na— ( = N—-ONa + H:0 eae N—N—G:H80;-ONa RC = N-N-CHSe We believe that the red compounds formed with acetaldehyde, acetone, and acetoacetic acid are the sodium salts of hydrazo- xime derivatives. Von Pechmann and Wehsarg® have prepared a hydrazoxime derivative of methyl glyoxal which is similar to the compound that we believe to be formed with acetone. 5 Borsche, W., Ann. chem. Pharm., 1907, ceclvii, 171. 7 Hantzsch, A., and Glover, W. H., Ber. chem. Ges., 1907, xl, 4344. 8 yon Pechmann, H., and Wehsarg, K., Ber. chem. Ges., 1888, xxi, 2996. M. T. Hanke and K. K. Koessler 241 H H 12 fa oe C = N—N—C,H,—SO: - OH C = N—NH-C,H; d =N-—-OH | |. CH; CH; von Pechmann and Wehsarg’: s Hydrazoxime formed from hydrazoxime. acetone. They give no data on the color of the compound in alkaline solu- tions although they found the alcoholic solution to be yellow and the sulfuric acid solution to be dark red. EXPERIMENTAL. Reagents Employed. Most of the reagents employed in these estimations are identi- cal with those employed by us in the estimation of imidazoles.° A detailed description is repeated here for the sake of completeness. Stock Sulfanilic Acid—Sulfanilic acid (4.5 gm.) is mixed with 45 ec. of 37 per cent hydrochloric acid (sp. gr. 1.19) in a 500 ce. volumetric flask. Water is then added to the mark. The solid dissolves slowly; but completely. Stock Sodium Nitrite-—25 gm. of 90 per cent odin nitrite are dissolved in water and diluted to 500 cc. in a volumetric flask. Sodium Carbonate-—Baker and Adamson’s anhydrous sodium carbonate (5.50 gm.) is dissolved in water and diluted to exactly 500 cc. We recommend the above grade because we have found it to give uniform results. The purity of this carbonate is a very important factor in the color development. Some grades of carbon- ate contain impurities that give yellow colors of inferior intensities. The finished carbonate solution must be preserved in a glass vessel that has little tendency to dissolve in alkali. Pyrex glass vessels have proved to be entirely satisfactory. Stock Methyl Orange-—Vacuum-dried Griibler’s methyl orange (0.5000 gm.) is dissolved in water and diluted to exactly 500 ce. This solution keeps indefinitely. Stock Congo Red.—Vacuum-dried Griibler’s Congo red (2.5000 gm.) is mixed with 50 ce. of absolute alcohol in a 500 cc. volumet- 242 Studies on Proteinogenous Amines. XIV ric flask. Water is then added to the mark. This solution keeps indefinitely. Stock Acid Fuchsin—Vacuum-dried Harmer’s acid fuchsin (2.5000 gm.) is dissolved in water and diluted to exactly 500 ce. in a volumetric flask. The thymol-preserved solution keeps indefinitely. Stock Phenol Red—Hynson, Westcott and Dunning’s phenol red (0.0500 gm.) is dissolved in water and diluted to exactly 500 ce. The thymol-preserved aqueous solution can be kept for at least 1 year. The alcoholic solution employed by many in- vestigators for pH determinations deteriorates quite rapidly and cannot be used in this work. Standard Indicator Solutions—For the estimation of phenol, a solution containing 10 ce. of stock phenol red in a total aqueous volume of 100 cc. is employed. Redistilled water should be used for dilutions and the volumetric flask should be thoroughly rinsed with distilled water before the phenol red solution is introduced. It is best to prepare a fresh standard every day. The color so obtained matches that produced by phenol perfectly. When this standard indicator solution has been used for comparisons the symbol (Ph—R) is suffixed to the reading obtained. For the estimation of o- and m-cresol, a solution containing 5 ec. of stock methyl orange in a total aqueous volume of 500 ce. is employed. The color so obtained matches that produced by o- and m-cresol almost perfectly. When this standard indi- eator solution has been used for comparisons the symbol (MO) is suffixed to the reading obtained. For the estimation of p-cresol, and p-oxyphenylacetic, p-oxryphenyl- propionic, and p-oxyphenyllactic acids, a solution containing 1.00 ec. of stock Congo red in a total aqueous volume of 500 ce. is employed. When this standard indicator solution has been used for comparisons the symbol (CR) is suffixed to the reading obtained. For the estimation of tyrosine or tyramine, a solution containing 1 ce. of stock acid fuchsin and 1.8 ec. of stock methyl orange in a total aqueous volume of 500 cc. is employed. When this stand- ard indicator solution has been used for comparisons the symbol (F— MO) is suffixed to the reading obtained. Cl OT M. T. Hanke and K. K. Koessler 243 Preparation of p-Diazobenzenesulfonic Acid Solution (The Reagent). 1.50 ce. each of the stock sulfanilic acid and sodium nitrite solutions are measured into a 50 cc. volumetric fiask. The flask ~ is then immersed in an ice bath for 5 minutes. Then 6.00 ce. more of the stock sodium nitrite solution are added and the well mixed solution is again allowed to lie in the ice bath for 5 minutes. Dis- tilled water is then added to the mark and-the flask returned to the ice bath where it is kept. This reagent must not be used for at least 15 minutes after diluting with water. We have found it to give perfect results after 24 hours. It is best, however, to prepare a fresh reagent every day. Preparation of the Phenols. Phenol.—Merck’s highest purity phenol was distilled 7m vacuo. The product collected boiled at 95° in a vacuum of 10 mm. of Hg. The colorless oil solidified readily. The colorless erystals so obtained had a melting point of 43°. The substance was as- sumed to be 100 per cent pure phenol. o-Cresol—A chemically pure product obtained from a local supply house was doubly distilled im vacuo. The colorless oil obtained boiled at 81° at a pressure of 10 mm. and at 188° under atmospheric pressure. The oil solidified on ‘standing. The colorless crystals so obtained had a melting point of 32°. It was assumed to be 100 per cent pure o-cresol. m-Cresol—A chemically pure product obtained from a local supply house was doubly distilled in vacuo. The colorless oil so obtained boiled at 104° under 10 mm. pressure and at 201° under atmospheric pressure. It did not solidify. It was assumed to be 100 per cent pure m-cresol. p-Cresol—The p-cresol was obtained as a by-product in the preparation of tyramine.’ The substance boiled at 90-91° at a pressure of 10 mm. and at 198°under atmospheric pressure. The oil solidified after standing for some time. The colorless crys- tals so obtained melted at 36°. It was assumed to be 100 per cent pure p-cresol. ® Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1919, xxxix, 585. 244 Studies on Proteinogenous Amines. XIV Oxyphenylacetic Acid.—This substance was obtained as a by- product in the preparation of tyramine.’ The colorless solid had the following properties. 1. It melted at 150°. 2. It left no residue on ignition. 3. The solid (0.1000 gm.) was dissolved in 10 cc. of water and titrated with 0.1 n NaOH using phenolphthalein as indicator. Exactly 6.59 cc. of the alkali were required to produce the first change in the indicator. The amount demanded by the formula is 6.58 cc. The substance was, there- fore, 100 per cent pure. Oxyphenylpropionic Acid.—This compound was _ prepared synthetically as follows. Cinnamic acid was reduced, with sodium amalgam, to hydro- cinnamic acid.'° p-Nitrophenylpropionic acid was then prepared by a modifi- cation of the method of Beilstein and Kuhlberg."! Stohr!” prepared p-aminophenylpropionic acid by reducing the ethyl ester of p-nitrophenylpropionic acid with zine and hydro- chloric acid. He isolated the zine double salt of the acid and treated this with acid and sodium nitrite to prepare p-oxyphenyl- propionic acid. We reduced p-nitrophenylpropionic acid (6 gm.) — directly with zinc (14 gm.) and hydrochloric acid (50 ce. of the 37 per cent acid) in alcoholic solution (70 cc. of 95 per cent alcohol) adding the acid slowly and keeping the temperature below 30°. The colorless liquid obtained, after the reaction had been allowed to proceed for 24 hours, was treated with 23.5 ce. of 95 per cent sulfuric acid and subjected to a distillation in vacuo at 60°. The pale yellow solid so obtained was dissolved in water (200 cc.) treated with sulfuric acid (20 ce. of the 95 per cent acid) and cooled in an ice bath. Sodium nitrite (60 cc. of a 5 per cent solution) was poured into the cold liquid. The resulting solution was allowed to remain in the ice bath for 1 hour after which it was treated with 250 ec. of water and heated to the boiling point. There was a copious evolution of nitrogen; but practically no tar formation. The aqueous solution was 10 Fischer, E., Anleitung zur Darstellung organischer Priparate, Bruns- wick, 8th edition, 1908, 39. 11 Beilstein, F., and Kuhlberg, A., Ann. chem. Pharm., 1872, elxiii, 132. 12 Stohr, C., Ann. chem. Pharm., 1884, eexxv, 60. M. T. Hanke and K. K. Koessler 245 cooled and extracted with ether. Removal of the ether, by distillation, left a pale yellow crystal cake weighing 4.5 gm. which was recrystallized first from water and then from hot benzene. The colorless solid finally obtained had the following properties. 1. It melted sharply at 130° (corrected). 2. It left no residue on ignition. 3. 0.166 gm. of the vacuum-dried solid was dissolved in water and the solution titrated with 0.20 n NaOH using phenolphthalein as indicator. Exactly 5.00 cc. of alkali were required. This is the amountedemanded by the formula. The solid was 100 per cent pure. p-Oxyphenyllactic Acid—This was prepared from tyrosine by the method of Kotake.“. The colorless solid obtained after drying to constant weight at 105° had the following properties. 1. It melted sharply at 172° (corrected). 2. It left no residue on ignition. 3. 0.25 gm. was dissolved in water and the solution titrated with 0.1 n NaOH using phenolphthalein as indicator. Exactly 13.70 ce. of alkali were required, which is almost exactly the amount (13.74 ec.) required by the formula. The solid was 100 per cent pure. Tyrosine.—The tyrosine used in this work was prepared from horn. The pure white, asbestos-like solid had the following properties. . It did not melt at a temperature of 290°C. . It was free from ammonia. . It left no residue on ignition. . It contained no cystine. 5. The ammonia obtained from 0.1000 gm. of the solid—Kjeldahl method —neutralized 5.50 cc. of 0.10 Nn HCl as compared with 5.52 cc. required by the formula. The substance was 100 per cent pure. wm CO be Tyramine Hydrochloride——The synthetic preparation and prop- erties of this compound have been previously described by us.° The product used was 98.14 per cent pure tyramine hydrochloride. The impurity consisted entirely of sodium chloride. 18 Kotake, Y., Z. phystol. Chem., 1910, Ixv, 398. 246 Studies on Proteinogenous Amines. XIV Procedure for Estimating Phenols Other than Tyrosine or Tyramine. Process I. The method used in developing the tables as well as the general procedure for estimating phenols other than tyrosine and tyra- mine is illustrated by the following example. (1-X) ee. of water and 5 ce. of the 1.1 per cent sodium carbon- ate solution are accurately measured into the right-hand cylinder of a Duboséq colorimeter. 2 cc. of reagent are measured into a 5 second delivery 2 ce. pipette, the time noted to the second, and the reagent allowed to flow into the alkali. The contents of the cylinder are then thoroughly mixed by allowing the liquid to flow repeatedly up the inclined tube as far as safety from loss will permit. The mixing should not take over 30 seconds. X cc. of the phenol solution is allowed to flow into the cylinder exactly 1 minute after the reagent began to mix with the alkali. The contents of the cylinder are mixed thoroughly as above.“ The test cylinder is then transferred to the colorimeter and set at 20mm. The left-hand cylinder, which should contain the appro- priate standard indicator solution is then adjusted constantly until a maximum reading has been obtained. The most accurate readings can be obtained by choosing such an amount of phenol solution that the standard indicator-con- taining cylinder has to be set at from 5 to 20 mm. The method described here can be used on quantities of the phenol-containing solution varying from 0.01 to 1 ec. The com- bined volume of water and phenol solution used should always be 1cc. Thus, if 0.10 cc. of the phenol solution is to be used, 0.90 cc. of water is added to the test cylinder. Then X equals 0.10 ce. and 1—X equals 0.90 cc. Estimation of Small Amounts of Phenol.. A stock 1.00 per cent phenol solution was prepared by dissolving 2.0000 gm. of the pure solid in water and diluting to exactly 200 cc. From this the standard solution was prepared by dilut- 14 Process II described on page 257 is the same as Process I up to this point. M. T. Hanke and K. K. Koessler 247 ing 1 ec. to 100 ce. Each ce. of the standard solution contained 0.0001 gm. of phenol. The color produced with phenol is very intense. It develops so rapidly that it has reached 95 per cent of its maximum value before a reading can be taken. A maximum color is obtained within 2 minutes of the time that the phenol is added to the alka- line reagent. The color remains of maximum intensity for about 5 minutes, then fades slowly without changing appreciably in tint. Most people find this greenish yellow color hard to com- pare, without practice, because of the speed with which it pro- duces retinal fatigue. Accurate readings can only be obtained by observing the colors briefly, setting the cylinder rapidly and making the fine adjustment after a brief interval of rest. When this is done the readings can be checked with an accuracy of about 1 per cent. The previously described (Ph-R) standard indicator solution was used for comparisons in the compilation of Table I. This table shows clearly that the color production is directly propor- tional to the amount of phenol used. Where only a few determinations are to be carried out it is usually simpler to compare the color produced by an unknown with that produced by a measured amount of a standard phenol solution, the two colors being prepared simultaneously. When the amount of standard phenol used is nearly equal to that in thé test liquid, very accurate results can be obtained by this method. Estimation of Small Amounts of o-Cresol. A stock 0.1 per cent o-cresol solution was prepared by dissolving 1.0000 gm. of the pure solid in water and diluting to exactly 1,000 ec. From this the standard solution was prepared by diluting 10 cc. to 100 ec. Each ce. of the standard solution contained 0.0001 gm. of o-cresol. The color produced with o-cresol is very intense. Its speed of development is like that of phenol. The color, which is predomi- nantly yellow but which contains slightly more pink than that produced by phenol, remains of maximum intensity and perma- nent tint for about 3 minutes; then it fades rapidly and acquires aredder tint. This color is easier to compare than that of phenol. 248 Studies on Proteinogenous Amines. XIV TABLE I. Estimation of Small Amounts of Phenol. Depth of indicator solution (Ph—R,) required | Phenol in the test cylinder (total volume 8 cc.) to match the color in the test cylinder. test cylinder set at 20 mm. mm. gm. . 1.2 0.000001 2.3 0.000002 3.9 0.000003 4.6 0.000004 5.8 0.000005 6.9 0.000006 8.1 0.000007 9.2 0.000008 10.4 0.000009 LIS 0.000010 eG 0.000011 13.8 0.000012 15.0 0.000013 16.1 ~ 0.000014 17.3 0.000015 18.4 0.000016 19.6 0.000017 20.7 0.000018 21.9 0.000019 23.0 0.000020 24.2 0.000021 PAN 0.000022 26.5 0.000023 27.6 0.000024 28.8 0.000025 29 9 , 0.000026 obel 0.000027 32.2 0.000028 33.4 0.000029 34.5 0.000030 The previously described (MO) standard indicator solution was used for comparisons in the compilation of Table II. This table’ shows clearly that the color produced is directly proportional to the amount of o-cresol used. M. T. Hanke and K. K. Koessler 249 TABLE II. Estimation of Small Amounts of o-Cresol. Depth of indicator solution (MO) required to o-Cresol in the test cylinder (total volume match the color in the test cylinder. 8 cc.) test cylinder set at 20 mm. mm. gm. 1.3 0.000001 2.4: 0.000002 3.6 0.000003 4.8 0.000004 6.0 0.000005 7.3 0.000006 8.5 0.000007 9.7 0.000008 : 10.9 0.000009 i 124 0.000010 13.3 0.000011 14.5 0.000012 ) 15.7 0.000013 | 16.9 0.000014 isi 0.000015 19.4 0.000016 20.6 0.000017 21:8 0.000018 23.0 0.000019 24.2 0.000020 25.4 0.000021 26.6 0.000022 27.8 0.000023 29.0 0.000024 30.2 0.000025 31.5 0.000026 3.7 0.000027 33.9 0.000028 35.1 0.000029 36.3 0.000030 Estimation of Small Amounts of m-Cresol. A stock 0.10 per cent solution of m-cresol was prepared by dissolving 1.000 gm. of the colorless oil in water and diluting to exactly 1,000 ce. From this the standard solution was prepared _ 250 Studies on Proteinogenous Amines. XIV ~ by diluting 10 ce. to 100 cc. Each cc. of the standard solution contained 0.0001 gm. of m-cresol. TABLE III. Estimation of Small Amounts of m-Cresol. Depth of indicator solution (MO) required to | m-Cresol in the test cylinder (total volume match the color in the test cylinder. 8 cc.). test cylinder set at 20 mm. mm. gm. 10 0.000001 1.9 0.000002 2.9 0.000003 3.9 0.000004: 4.9 0.000005 5.8 0.000006 6.8 0.000007 7.8 0.000008 8.7 0.000009 OR7 0.000010 10.7 0.000011 11.6 0.000012 12.6 0.000013 13.6 0.000014 14.6 0.000015 L525 0.000016 16.5 0.000017 Miata: 0.000018 18.4 0.000019 19.4 0.000020 : 20.4 0.000021 PAS) f 0.000022 22:53 0.000023 23.3 0.000024 24.3 0.000025 2or2 0.000026 26.2 0.000027 27.2 | 0.000028 eT | 0.000029 29.1 | 0.000030 The intense color produced with m-cresol is qualitatively iden- tical with that obtained with o-cresol. In this case, however, M. T. Hanke and K. K. Koessler 251 the time of complete development is about 5 minutes although . most of the color appears immediately. At first the color is slightly more yellow than the (MO) comparison standard. In the course of 3 minutes the match is perfect and a maximum of color is obtained in 5 minutes. The color fades slowly and becomes pinker. The previously described (MO) standard indi- cator solution was used for comparisons in the compilation of Table III. This table shows clearly that the. color production is directly proportional to the amount of m-cresol used. Estimation of Small Amounts of p-Cresol. A stock 0.10 per cent solution of p-cresol was prepared by dis- solving 1.000 gm. of the colorless solid in water and diluting to exactly 1,000 ce. From this the standard solution was prepared by diluting 10 cc. to 100 cc. Each cc. of the standard solution contained 0.0001 gm. of p-cresol. In this case the color, which is predominantly red, develops to its maximum intensity during the 30 seconds required to mix the contents of the test cylinder. The readings recorded are those obtained immediately after the test cylinder was transferred to the colorimeter. The comparison must be made within 2 min- utes during which time the color remains of constant and maxi- mum intensity. Then the test liquid, although it changes but little in intensity, rapidly acquires a cloudy appearance which renders further comparisons untrustworthy. This same cloudy appearance is also obtained when the test liquid contains a finely divided precipitate which leads us to believe that an insoluble compound appears, after 2 minutes, although none can be seen with the naked eye. The previously described (CR) standard indicator solution was used for comparisons in the compilation of Table IV. As can be seen from the table, the color production is directly pro- portional to the amount of p-cresol used. Estimation of Small Amounts of p-Oxyphenylacetic Acid. A stock 1.00 per cent solution of p-oxyphenylacetic acid was pre- pared by dissolving 1.0000 gm. of the colorless solid in water and diluting to 100 ce. This solution was preserved with chloroform. 252 Studies on Proteinogenous Amines. XIV m _From it, the standard solution was prepared by diluting 1 ec. to 100 ec. Each ee. of the standard solution contained 0.0001 gm. of p-oxyphenylacetic acid. TABLE IV. Estimation of Small Amounts of p-Cresol. Depth of indicator solution (CR) required to p-Cresol in the test cylinder (total volume match the color in the test cylinder. = 8 cc.) test cylinder set at 20 mm. mm. gm. 2.5 0.000005 3.0 0.000006 3.5 0.000007 4.0 0.000008 4.5 0.000009 o.0 0.000010 5.5 0.000011 6.0 0.000012 6.5 0.000013 7.0 0.000014 0.5 0.000015 8.0 0.000016 8.5 0.000017 9.0 0.000018 9.5 0.000019 10.0 0.000020 10.5 0.000021 11.0 0.000022 11.5 0.000023 12.0 0.000024 12.5 0.000025 13.0 0.000026 13.5 0.000027 14.0 0.000028 14.5 0.000029 15.0 0.000030 In this case the color, which is predominantly red, develops to its maximum intensity in about 2 minutes. This color of maxi- mum intensity is permanent for about 5 minutes after which it fades slowly and becomes yellow. ME: Hanke and K. K. Koessler 253 The previously described (CR), standard indicator solution was used for comparisons in the compilation of Table V. It has a color that is slightly more red than that produced by p-oxyphenyl- acetic acid; but the match is so close that accurate comparisons are easily made. As can be seen from Table V the color produc- tion is directly proportional to the amount of p-oxyphenylacetic acid used. TABLE V. Estimation of Small Amounts of p-Oxyphenylacetic Acid. Depth of indicator solution (CR) required p-Oxyphenylacetic acid in the test cylinder to match the color in the test cylinder. (total volume 8 cc.) test cylinder set at 20 mm. mm. gm. 3.6 0.000010 4.0 0.000011 4.3 0.000012 4.7 0.000013 5.0 0.000014 5.4 0.000015 5.7 0.000016 Gal 0.000017 6.5 0.000018 6.8 0.000019 (OP 0.000020 eo 0.000021 7.9 0.000022 8.3 0.000023 8.6 0.000024 9.0 0.000025 9.4 0.000026 9.7 0.000027 10.1 0.000028 10.4 0.000029 10.8 j 0.000030 Estimation of Small Amounts of p-Oxyphenylpropionic Acid. A stock 1.00 per cent solution of p-oxyphenylpropionic acid was prepared by dissolving 1.0000 gm. of the solid in 30 cc. of 0.20 N NaOH and diluting, with water, to exactly 100 cc. Chloro- form was added as a preservative. The standard solution was prepared from the stock solution by diluting 1.00 cc. of the latter 254 Studies on Proteinogenous Amines. XIV to 100 cc. Each cc. of the standard solution contained 0.0001 gm. of p-oxyphenylpropionic acid. The color, which is predomi- TABLE VI, Estimation of Small Amounts of p-Oxyphenylpropionic Acid. Depth of indicator solution (CR) required | p-Oxyphenylpropionic acid in the test cylinder to match the color in the test cylinder. (total volume 8 cc.) test cylinder set at 20 mm. mm. gm. 3.1 0.000010 3.4 0.000011 3.7 0.000012 4.0 0.000013 4.3 0.000014 4.6 0.000015 5.0 0.000016 5.3 0.000017 5.6 0.000018 5.9 0.000019 6:2 0.000020 6.5 0.000021 6.8 0.000022 oll 0.000023 7.4 0.000024 Unth 0.000025 8.1 0.000026 8.4 0.000027 8.7 0.000028 9.0 0.000029 9.3 0.000030 9.6 0.000031 9.9 0.000032 10.2 0.000033 10.5 0.000034 10.8 0.000035 a 0.000036 Lee 0 000037 11.8 0.000038 12.1 0.000089 12.4 0.000040 ——— eee EEE EES nantly red, develops to its maximum intensity in about 2 min- utes and undergoes no change for from 1 to 3 minutes depending M. T. Hanke and K. K. Koessler PATS upon the quantity of p-oxyphenylpropionic acid present and the room temperature. This period of color constancy is followed, rather sharply, by a rapid decline in color intensity, a cloudy ap- pearance and a change in tint. The values recorded in Table VI are those obtained just previous to this change. The previously described (CR) standard indicator solution was used for comparisons in the compilation of Table VI. This table shows that the color production is directly proportional to the amount of p-oxyphenylpropionic acid present. Estimation of Small Amounts of p-Oxyphenyllactic Acid. A stock 1.00 per cent solution of p-oxyphenyllactic acid was pre- pared by dissolving 1.0000 gm. of the pure solid in 56 ce. of 0.1 N NaOH and diluting with water to exactly 100 ce. Chloroform was added as a preservative. The standard solution was pre- pared by diluting 1.00 cc. of the stock solution to 100 cc. Each ec. of the standard solution contained 0.0001 gm. of p-oxyphenyl- lactic acid. The color, which is predominartly red, develops to its maximum intensity in about 4 minutes and undergoes no change for about 5 minutes. The (CR) standard, which was used for compari- sons in the compilation of Table VII, is slightly more red than the color produced by p-oxyphenyllactic acid; but the discrep- ancy in tint is too slight to interfere with the accuracy of the comparisons. The slowness of the color development, the slight yellow tint, and the stability of the color are characteristics for oxyphenyllactic acid. In this case there is no sharp color change and the liquid does not become cloudy as in the case of oxyphenyl- propionic acid. The tabular values have been carried out to the second place because these figures are obviously more correct than the actual readings which can, of course, be obtained with an accuracy of only one decimal place. Thus for 0.000019 gm. of p-oxyphenyl- lactic acid a reading of either 4.70 or 4.80 mm. is obtained; but the correct reading would be 4.75. This figure (4.75) therefore, appears in the table. As in the case of the other phenols, the color production is directly proportional to the amount of the phenol used. . TABLE VII. Estimation of Small Amounts of p-Oxyphenyllactic Acid. a Depth of indicator solution (CR) required p-Oxyphenyllactic acid in the test cylinder to match the color in the test cylinder. (total volume 8 cc.) test cylinder set at 20 mm. mm. gm. 2.5 0.000010 2.75 0.000011 3.0 0.000012 3.25 0.000013 3.5 0.000014 3.75 0.000015 4.0 0.000016 4.25 0.000017 4.5 0.000018 4.75 0.000019 5.0 0.000020 §.25 0.000021 5.5 0.000022 5.75 0.000023 6.0 0.000024 6.25 0.000025 6.5 0.000026 6.75 0.000027 7.0 0.000028 7.25 0.000029 7.5 0.000030 7.15 0.000031 8.0 0.000032 8.25 0.000033 8.5 0.000034 8.75 0.000035 9.0 0.000036 OF25 - 0.000037 9.5 0.000038 9.75 0.000039 10.0 0.000040 10.25 0.000041 10.5 0.000042 10.75 0.000048 1 0.000044 11.25 0.000045 re 0.000046 11.75 0.000047 12.0 0.000048 12.28 J 000049 jas ie 9 ONNDN5O 256 M. T. Hanke and K. K. Koessler 257 Procedure for the Estimation of Small Amounts of Tyrosine and Tyramine. Process II. As has already been pointed out in the introduction, tyrosine and tyramine show an anomalous behavior toward p-phenyldia- zonium sulfonate in alkaline—sodium carbonate—solutions. When a solution of tyrosine or tyramine is added to the alkaline reagent, a pink color develops promptly, as in the case of the aromatic hydroxy-acids. After about 30 seconds, however, the color changes sharply to yellow and fades. The pink color is too evanescent to compare and the intensities of the yellow colors ’ are not directly proportional to the amount of tyrosine or tyramine used. The procedure that has been found to give satisfactory results with the other phenols, and with imidazoles, cannot be used for the estimation of tyrosine or tyramine. The proceedure finally adopted is given below; the reasons for its adoption and an explanation of the chemical reactions involved have already been given in the introduction. Process I described for the other phenols (see page 246) is followed as far as foot-note 14. The test cylinder is allowed to stand for exactly 53 minutes from the time that the tyrosine or tyramine solution began to mix with the alkaline reagent. This gives rise to a primary color that is yellow to orange and whose intensity is not directly proportional to the amount of the phenol used. Sodium hydroxide (2.00 cc. of a 3.0 N solution) is added and the contents of the cylinder thoroughly mixed as before. This gives rise to a marked color intensification with a change of tint toward red, the color still being, however, predominantly yellow. Exactly 1 minute after the sodium hydroxide solution began to mix with the liquid in the test cylinder, 0.10 cc. of a 20 per cent solution of hydroxylamine hydrochloride is rapidly introduced and the contents of the cylinder are again thoroughly and rapidly mixed. At first there is no change in color. Then suddenly, after a latent period of from 5 to 10 seconds, an intense bluish red color 18 This was obtained from the Special Chemicals Co., Highland Park, Illinois. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 258 Studies on Proteinogenous Amines. XIV develops. This secondary color develops to its full intensity while the cylinder and contents are being agitated (30 seconds) and changes very little in tint or intensity in half an hour. It is best to introduce the hydroxylamine solution from a rapid delivery 0.10 ce. pipette and to have the liquid well mixed before the second- ary color begins to develop. The test cylinder is transferred to the right-hand side of the Duboscq colorimeter and set at 25 mm. The left-hand cylinder, which contains the previously described (F-MO) comparison standard, is then adjusted until the two halves of the field are identical in tint and intensity. The color is easily compared and the determinations are accurate to about 0.5 to 1.5 per cent. If the process is properly carried out, the correction blank (see later) is about 0.30 mm. (F—-MO). This amount must be subtracted from the actual reading to ob- tain the values recorded in Table VIII. We have found that a high laboratory temperature raises this correction blank from 0.10 to 0.20 mm. The hydroxylamine solution must not be intro- duced until the sodium hydroxide has been allowed to react with the alkaline reagent for at least 1 minute; otherwise a colored compound is produced with the hydroxylamine. | The tabular values are accurate for laboratory temperatures ranging from 18 to 25°. At lower temperatures the colors are very slightly less intense and at higher temperatures they are slightly more intense. It is always advisable, before carrying out a determination on an unknown, to run a few determinations on a standard tyrosine solution to be certain that the tabular values are accurate for the existing laboratory conditions. It is a curious coincidence that the tabular values are identical for both tyrosine and tyramine hydrochloride. The same table may, therefore, be used in the estimation of either. A stock 1.00 per cent solution of tyrosine was prepared by dis- solving 2.0000 gm. of the pure solid in 75 cc. of 1.0 n HCl and diluting with water to 200 cc. From this the standard solution was prepared by diluting 1 cc. to 100 cc. in a volumetric flask. Each ce. of the standard solution contained 0.0001 gm. of tyrosine. A stock 1.00 per cent solution of tyramine hydrochloride was pre- pared by dissolving 2.0379 gm. of the 98.14 per cent solid in water and diluting to exactly 200 ec. Chloroform was added as a preservative. The standard solution was prepared by diluting M. T. Hanke and K. K. Koessler 259 1.00 ce. of the stock solution to 100 cc. in a volumetric flask. Each ec. of the standard solution contained 0.0001 gm. of tyramine hydrochloride. TABLE VIII. Estimation of Small Amounts of Tyrosine and Tyramine Hydrochloride. Tyrosine or tyramine hydrochloride in the Perio mateh the color inthe test cylinder. | __t#8t cylinder (total volume 10.1 ce.) mm. gm. 4.0 0.000005 ae 0.000006 5.6 0.000007 6.4 0.000008 CS 0.000009 8.0 0.000010 8.8 0.000011 9.6 0.000012 10.4 0.000018 11.2 0.000014 12.0 0.000015 12.8 0.000016 13.6 0.000017 14.4 0.000018 15.2 0.000019 16.0 0.000020 16.8 0.000021 17.6 0.000022 18.4 0.000023 19.2 0.000024 20.0 0.000025 20.8 0.000026 21.6 0.000027 22.4 0.000028 23.2 0.000029 24.0 0.000030 The Correction Blank. When the reagent and alkali are mixed in the absence of a phenol derivative, a very pale yellow color is produced in the course of 5 minutes. This color has an intensity value equivalent to about 0.30 mm. of any of the standard indicator solutions used. This 260 Studies on Proteinogenous Amines. XIV same amount of color is also produced in the presence of phenol derivatives along with the color produced by the phenol; so the readings obtained are high by about 0.30 mm. in every case. It is necessary, therefore, to subtract 0.30 mm. from the readings obtained before comparison with the tables. Substances That Do and Do Not Interfere with the Quantitative Determination of Phenols. The tests for interference were carried out as follows, unless otherwise specified. Tyrosine (1.00 cc. of the stock 1 per cent solution) was mixed, in a 100 ce. graduated precision cylinder, with quantities of the interfering substances as given below. Water was added to give a total volume of 100 cc. Then 0.10 and 0.20 cc. portions of this liquid were taken for the colorimetric determinations. Interference with the determination of tyrosine was studied particularly because this determination involves a heretofore unstudied type of reaction. Phenols other than tyrosine and. tyramine are determined by a process identical with that used for imidazoles and the interference with that process has already been studied and reported. Sodium and potassium chloride, sulfate, phosphate, acetate, and citrate in 5 per cent concentrations do not interfere with the colorimetric determination of tyrosine or tyramine. Ammonium salts interfere very seriously with the determination. When the concentration of ammonium chloride is 5 per cent, so much of a greenish yellow color is produced that the presence of tyrosine would not be suspected. Needless to say, tyrosine could not be determined under these conditions. When the concentration of ammonium chloride is 1 per cent, the inter- ference is still considerable because 0.10 cc. of the solution had a color value equal to 9.0 mm. (F—MO) as compared to a normal value of 8.0 mm.; which is a positive error of 12.5 per cent. The color was quite yellow. When the concentration of ammonium chloride is reduced to 0.5 per cent results were obtained that are fairly satisfactory for now 0.10 ec. had a color value equal to 8.2 mm. (F—MO) which is 102.5 per cent of the correct reading. The color was exactly like that of the standard. Ammonium salts must not be present in concentrations exceeding 0.5 per cent in liquids that are to be examined for tyrosine colorimetrically. M. T. Hanke and K. K. Koessler 261 Amino-acids also interfere seriously with the colorimetric estimation of tyrosine as can be seen from the following data. Leucine.—A solution was prepared containing 5 ce. of a 1 per cent leucine solution, 0.20 cc. of a 1 per cent tyrosine solution, and sufficient water to give a total volume of 10 cc. Of this solution 0.05 ec. had a color value equivalent to 10.6 mm. (F — MO) and 0.10 “ec 3 “ “< “ “ce “c 13:7 “ (F —_ MO). The color contained far more yellow than the comparison stand- ard. Both of these values are high, the 0.05 cc. reading being 30 per cent high and the 0.10 cc. reading being 15 per cent high. This excess of color is due to a yellow compound that is formed when leucine reacts with p-phenyldiazonium sulfonate. In this case the ratio of leucine to tyrosine was 25 to 1. A second liquid was prepared containing 1.00 cc. of a 1 per cent leucine solution, 0.10 cc. of a 1 per cent tyrosine solution, and suffi- cient water to give a total volume of 10 cc. Of this solution 0.10 ec. had a color value poles to 8.4mm. (F —MO) and 0. 20 “ce “ “ce “ “cc “e 16 #3 “ (F = MO). The color obtained contained slightly more yellow fhe the com- parison standard; but the match was very good. These values are 5 per cent higher than they should be. The ratio of leucine to tyrosine in this case was 10 to 1. These experiments show that leucine interferes seriously with the colorimetic estimation of tyrosine when the ratio of leucine to tyrosine exceeds 10 to 1. Glycine —A solution was prepared containing 0.30 cc. of a 1 per cent glycine solution and 0.10 cc. of a 1 per cent tyrosine solution in a total aqueous volume of 10 cc. Of this solution 0.10 cc. had a color value equivalent to 8.0 mm. (F — MOQ) and 0.20 “ “ “ “ec “ “ce “e 16.0 “ec (F kaa MO). The color produced matched that of the standard perfectly. The values obtained are exactly what they would have been if glycine had not been present. When the ratio of glycine to ty- rosine is 3 to 1, the colorimetric estimation of tyrosine is not’ interfered with. 262 Studies on Proteinogenous Amines. XIV A second solution was prepared containing 0.50 cc. of a 1 per cent glycine solution and 0.10 cc. of a 1 per cent tyrosine solution in a total aqueous volume of 10 cc. Of this solution 0.10 cc. had a a yaane equivalent to 8.5 mm. (F — MO) and 0. 20 ce a9 ce ce “ abr 0 (t4 (F — MO). The color produced was distinctly yellow and not easy to com- pare with the (F-MO) comparison standard. The values are too high by 6 per cent. Even at this ratio (5 to 1), glycine inter- feres seriously with the colorimetric determination of tyrosine. With higher concentrations of glycine the interference becomes so pronounced that a determination of any kind is impossible. The color produced is predominantly yellow so that ue color due to tyrosine is masked almost completely. These experiments show that glycine interferes seriously with the colorimetric estimation of tyrosine when the ratio of glycine to tyrosine exceeds 5 to 1. From these experiments it is clear that the direct determination of tyrosine colorimetrically in the phos- photungstate filtrate fraction of a protein is impossible by means of this method as it now stands. Tyrosine, excepting for that part which can be separated by crystallization, is always associated with a high percentage of other amino-acids. ‘To estimate tyro- sine under these conditions we have either to separate tyrosine, or some derivative into which it can be quantitatively converted, from the bulk of the other amino-acids, or to remove the inter- fering NH» groups of these other amino-acids without destroy- ing the colorimetric properties of tyrosine. Attempts are now being made in this laboratory to modify this method so that it will be applicable to the estimation of tyrosine in proteins. Hydrogen Peroxide.—A solution was prepared containing 0.01 gm. of tyrosine and 0.20 ce. of a 3 per cent commercial hydrogen peroxide solution in a total aqueous volume of 100 ce. Of this solution 0.20 ee. had a color value equivalent to 10.0 mm. (F — MO) as compared to a normal value of 16.0 mm. The color con- tained far more yellow than the comparison standard. Hydro- gen peroxide is, therefore, a very potent interfering substance. M. T. Hanke and K. K. Koessler 263 In its presence extremely low and entirely unreliable values are obtained. Fortunately, hydrogen peroxide can be readily re- moved from an aqueous solution by treating it with platinized as- bestos. An example of such a treatment is given below under formaldehyde. Formaldehyde—A liquid was prepared containing 0.01 gm. of tyrosine and 5 cc. of commercial formaldehyde in a total aqueous volume of 100 cc. Of this solution 0.20 cc. had a color value equivalent to 9.2 mm. (F — MO). The reading should have been 16.0 mm. The color matched that of the comparison standard very well. A liquid was now prepared containing 0.01 gm. of tyrosine and 2 cc. of formaldehyde in a total aqueous volume of 100 ce. Of this solution 0.20 cc. had a color value equivalent to 14.4 mm. (F — MO). In this case the reading was low by 10 per cent. Finally a solution was prepared containing 0.01 gm. of tyrosine and 1.00 cc. of formaldehyde in a total aqueous volume of 100 cc. of which 0.20 ec. had a color value equivalent to 16.0 mm. (F — MO). This is the normal reading. In this concentration (1.0 per cent) formaldehyde does not interfere with the colorimetric estimation of tyrosine. It is of interest to note that low values are obtained when formaldehyde is present. Formaldehyde does not give a colored substance with p-phenyldiazonium sulfonate. In this respect this aldehyde is very different from acetaldehyde, acetone, and acetoacetic acid since the latter substances give intensely colored compounds with p-phenyldiazonium sulfonate in alkaline solu- tion after treatment with sodium hydroxide and hydroxylamine. Formaldehyde cannot form an olefine enol, it cannot give rise to hydrazoxime derivatives; the other three carbonyls can. This may account for their different behaviors. Since it is conceivable that tyrosine determinations might have to be carried out on liquids containing formaldehyde, we sought 264 Studies on Proteinogenous Amines. XIV for a method that would destroy formaldehyde without injuring tyrosine. Blank and Finkenbeiner' have described a method for the quantitative conversion of formaldehyde into sodium formate by means of hydrogen peroxide. We found that this method destroyed formaldehyde and did not destroy tyrosine; but the hydrogen peroxide left after the reaction had gone to completion made it impossible to determine the tyrosine colorimetrically. We then removed the hydrogen peroxide by means of platinized asbestos and found that the concentration of tyrosine had not changed. The experiment was conducted as follows. A solution was prepared containing 0.01 gm. of tyrosine, 7 cc. of water, 3 cc. of commercial formaldehyde, and 17 cc. of 3 N ~ NaOH. The substances were mixed in a 100 cc. graduated cylinder. After a reaction period of about 15 minutes, hydrogen peroxide (50 ec. of a 3 per cent commercial preparation) was slowly added to the contents of the cylinder without cooling. After the evolution of gas had ceased (15 minutes) 0.30 gm. of a 5 per cent platinized asbestos was added to the liquid. The evolution of oxygen was practically nil after 30 minutes. The liquid was neutralized with 5 N sulfuric acid (which took 3 ce.) and diluted to 100 cc. Of the clear colorless liquid obtained after filtration, 0.10 cc. had a color value equivalent to 8.0 mm. (F — MO) and 0.20 “ cc “c “ “ “ “ 15.9 “ (F arn MO) which is equivalent to 0.01 gm. of tyrosine for the entire test liquid, 100 per cent of the starting material. Formaldehyde and hydrogen peroxide do not destroy tyrosine under the conditions of the above experiment. Both of these interfering substances can be completely removed without interfering with the color- producing properties of tyrosine. Carbonyl-Enol Interference (Acetaldehyde, Acetone, and Acetoacetic Acid). These three substances have been classified as carbonyl-enols because they are capable of existing in two tautomeric forms; namely, 16 Blank, O., and Finkenbeiner, H., Ber. chem. Ges., 1898-99, xxxi, 2979. M. eeaeteand WevkKocaler = 265 Carbony] form. , Enol form. CH;—C:0O H.C = C—OH Ht 1H CH;—C—CH,; CH;—C = CH, 6 on H CH;—CO—CH,—COOH CH;—C= b_coon bu These compounds are particularly potent as interfering agents because they produce a color, under the conditions employed in the estimation of tyrosine, that closely resembles that produced by tyrosine. These colors are so intense that as little as 0.00005 gm. of any of the three substances gives sufficient color to be easily read in a colorimeter. When the carbonyl derivative is first added to the alkaline reagent a brilliant red color is produced that fades in a few seconds and becomes yellow. The addition of sodium hydroxide, after the initial 5 minute reaction period, enhances this color and changes its tint to bluish red. The subsequent addition of hy- droxylamine gives rise to an intense bluish red color that is al- most identical with that obtained with tyrosine. The fact that so simple a substance as acetaldehyde should give a color similar to that obtained with tyrosine would lead one to suspect that this carbonyl-enol tautomerism played a réle in the color production by tyrosine and tyramine (see introduction for the chemical reactions involved). Phenols can also occur in two tautomeric forms that would properly be called carbonyl and enol, respectively oO H || O C Cc fio® is HG ae ne cH a 8 Ho oe Vi a na | H R 266 Studies on Proteinogenous Amines. XIV Acetone is the only member of this group upon which we have carried out any extended experiments. A stock 5 per cent solu- ; tion of acetone was prepared by diluting 5 gm. of the pure prod- uct with water to 100 cc. A standard solution was then pre- pared by diluting 4 cc. of the stock solution to 100 cc. Each ec. of the standard solution contained 0.002 gm. of acetone. When 0.20 ec. of this standard solution (equivalent to 0.0004 gm. of acetone) was mixed with the alkaline reagent and subsequently treated with sodium hydroxide and hydroxylamine as in the tyrosine determination, a color was produced that was equivalent to about 8.5 mm. (F-MO), the color being slightly redder than that of the comparison standard. It was almost impossible to obtain exactly this value on repetition because the slightest change in the primary reaction time produced a large change in the read- ing finally obtained. We, therefore, tried several experiments (using 0.20 cc. of the standard acetone solution) in which this reaction time was elongated. With a reaction time of 15 minutes a color value of 29.6 mm. (F—MO) was obtained. With a reaction time of 25 minutes a color value of 31.6 mm. (F—MO) was ob- tained. Obviously then, the 5 minute reaction period that we found to be ample for the complete development of color in the case of tyrosine and tyramine is too short for acetone. We have made no attempt to change the conditions so that acetone would give a maximum constant color; but this could, no doubt, be done. We wish merely, to call attention to the fact that with 0.0004 gm. of acetone a color value of 29.6 mm. (F—MO) was ob- tained when the reaction period was 15 minutes; hence as little as 0.00004 gm. of acetone would give a distinctly perceptible color. This method might, therefore, be useful as a qualitative test for traces of acetone even though further experiments might prove that this method could not be made to give readings con- stant enough for quantitative work. It seems hardly necessary to say that these carbonyl-enols can never really interfere with the estimation of tyrosine or tyramine because the carbonyl compounds can always be removed by dis- tillation or evaporation. Glucose.—Glucose and the other sugars containing a free alde- hyde or ketone group would be expected to interfere with the colorimetric estimation of tyrosine because of the yellow color ; : | M. T. Hanke and K. K. Koessler 267 that these compounds give with sodium hydroxide. When 0.10 ec. of a 5 per cent glucose solution is subjected to the treatment used for estimating tyrosine, a very pale yellow primary color is produced which changes to brown with marked intensification on the addition of sodium hydroxide and then becomes redder when hydroxylamine is added. The color finally obtained is still far too yellow to make a comparison with the (F—MO) standard possible. Tyrosine (0.01 gm.) dissolved in 100 cc. of a 5 per cent glucose solution, produces its own color just as it does in the absence of glucose; but the color finally obtained is the sum of the tyrosine and glucose colors. Thus, of this solution 0.10 cc. had a color value equivalent to 10.8 mm. (F — MO) and 0.20 “ “ “ “c “ec “ “ce 19.3 “ (F eat MO). The colors obtained in each case contained more yellow than the comparison standard. If these readings are calculated as tyrosine, they are 20 and 13 per cent high, respectively. Although so high a concentration of glucose would interfere seriously with the colorimetric estimation of tyrosine, lower con- centrations are less potent. Thus a tyrosine solution containing 1 per cent of glucose gives a color whose tint and intensity are exactly like those of a solution containing no glucose. In 1 per cent concentration glucose does not interfere with the estimation of tyrosine. Alcohols—Methy1 and ethyl alcohol in 10 per cent concentra- tions give rise to a color with p-phenyldiazonium sulfonate that resembles that produced by tyrosine. The colors are very faint, however, so that a positive interference of about 20 per cent is obtained with 10 per cent concentrations of alcohol. We believe that the pure alcohols do noé interfere with the colorimetric process and that it is the acetone and acetaldehyde present in the alcohols that. produce the color. Ethyl alcohol, that has been freshly distilled over potassium hydroxide augments the color of a tyrosine solution very little even when the alcohol con- centration is 10 per cent. An aqueous tyrosine solution, that has been saturated with amyl alcohol, gives values that are about 5 per cent too high. A chloroform extraction of this liquid reduces the reading to about 268 Studies on Proteinogenous Amines. XIV 2 per cent above normal; but does not remove the amyl alcohol completely. Any of these alcohols can, of course, be completely removed by distillation or evaporation; so they can offer no per- manent difficulty. Chloroform, toluene, and ether that has been distilled over sodium, do not interfere with the colorimetric estimation of tyrosine. ; | Charcoal.—In our earlier work we found that animal or vege- table charcoal adsorbed appreciable quantities of imidazoles and we advised against the use of charcoal in any liquid that was to be tested quantitatively for imidazoles. The adsorption power of charcoal for phenols is far greater than for imidazoles as can be seen from the following data. 5 ec. each of the 1 per cent stock solutions of tyrosine, tyramine, and phenol were separately diluted to 100 cc. and treated with - 1 gm. of Pfanstiehl’s decolorizing charcoal. After 10 minutes -of agitation the liquids were filtered and colorimetric estimations made on the clear filtrates. Of the tyrosine solution, 0.10 ec. had a color value equivalent to 9.2 mm. (F — MO) which is equivalent to 0.0115 gm. of tyrosine; hence 0.0385 gm. of tyrosine was adsorbed by 1 gm. of charcoal. Of the tyramine solution, 0.10 cc. had a color value equivalent to 20.5 mm. (F — MO) which is equivalent to 0.0256 gm. of tyramine hydrochloride; hence 0.0244 em. of tyramine hydrochloride was adsorbed by 1 gm. of charcoal. Of the phenol solution, 1.0 ec. had a color value equivalent to 18.4 mm. (Ph — R) which is equivalent to 0.0016 gm. of phenol; hence 0.0484 gm. of phenol was adsorbed by 1 gm. of charcoal. Barium sulfate precipitates do not adsorb phenols from a neutral or acid solution. M. T. Hanke and K. K. Koessler 269 SUMMARY. 1. Methods have been devised for the quantitative colorimetric estimation of phenol, o-, m-, and p-cresol, p-oxyphenylacetic, p- oxyphenylpropionic, and p-oxyphenyllactic acids, tyrosine, and tyramine. 2. These methods are based upon the well known fact that phenols react with diazonium compounds in alkaline solutions to give colored derivatives. A freshly prepared solution of p-phenyl- diazonium sulfonate is mixed with a dilute solution of sodium carbonate. A dilute solution of the phenol whose concentra- tion is to be estimated is mixed with the alkaline reagent which gives rise to a primary color that is yellow to red depending upon the character of the phenol. 3. The phenols studied can be divided into three classes. A. Phenols in which the para position is not occupied by a second substituent. B. Phenols in which the para position is occupied by a sec- ond substituent that does not contain an amino group. C. Tyrosine and tyramine. Phenols belonging to Class A (phenol and o- and m-cresol), couple with great speed and give rise to yellow colors. Phenols belonging to Class B (p-cresol, p-oxyphenylacetic, p-oxyphenylpropionic, and p-oxyphenyllactic acids), couple more slowly than those belonging to Class A. The color produced is predominantly red. Tyrosine and tyramine show an anomalous behavior toward alkaline: (NasCO;) p-phenyldiazonium sulfonate. An evanescent pink color is produced at first, which fades in 30 seconds to a yellow of inconstant intensity. The simple process employed for the estimation of imidazoles and the other phenols cannot, therefore, be used for the estimation of tyrosine and tyramine. The primary yellow color produced by tyrosine or tyramine is enhanced somewhat by the addition of sodium hydroxide. The colors produced are not directly proportional to the amount of. phenol present. If this strongly alkaline liquid is now treated with a small amount of hydroxylamine hydrochloride, a very intense bluish red color is produced whose intensity is directly proportional to the amount of tyrosine or tyramine present (Process. IT). 270 Studies on Proteinogenous Amines. XIV 4. Tables are given for the direct determination of quantities of these phenols ranging from 0.000001 to 0.00005 gm. The amount of the phenol derivative in any quantity of liquid can then be determined, by multiplication, with an accuracy of from 0.5 to 3 per cent. 5. The alkali salts of the common organic and inorganic acids do not interfere with either of the above two colorimetric processes. Ammonium salts and amino-acids give an intense yellow color with the process used for the estimation of tyrosine and tyramine (Process II). High values are obtained if these nitrogen com- pounds are present in sufficient concentration. . Hydrogen peroxide and formaldehyde suppress the color pro- duction by tyrosine; hence in the presence of these compounds low values are obtained. Acetaldehyde, acetone, and acetoacetic acid give rise to a color that is qualitatively identical with that obtained with tyrosine and tyramine. The colors are so intense that the possibility of using this method in the estimation of these carbonyl derivatives sug- gests itself. The presence of the ordinary alcohols leads to high readings probably because of the presence of aldehydes or ketones in the alcohols. , 1 gm. of vegetable charcoal adsorbs 0.0385 gm. of tyrosine, 0.0244 gm. of tyramine, and 0.0484 gm. of phenol from 100 ce. of an aqueous solution that originally contained 0.05 gm. of the above phenols. STUDIES ON PROTEINOGENOUS AMINES. XV. A QUANTITATIVE METHOD FOR THE SEPARATION AND ESTIMATION OF PHENOLS INCLUDING PHENOL, o-, m-, AND p-CRESOL, p-OXYPHENYLACETIC, p-OXYPHENYL- PROPIONIC, AND p-OXYPHENYLLACTIC ACIDS, TYROSINE, AND TYRAMINE. By MILTON T. HANKE anp KARL K. KOESSLER. (From the Otho S. A. Sprague Memorial Institute and the Department of Pathology, University of Chicago, Chicago.) (Received for publication, October 21, 1921.) A study of the products formed from tyrosine by the action of living cells has always been less difficult than the study of the products formed from histidine because the phenols formed can be easily separated into several well defined groups. Experi- ments on tyrosine have almost invariably been conducted on a large scale, however, because the products formed have had to be identified and estimated by a process of isolation and puri- fication. A purely chemical method, applicable to small amounts of material, that would effect a quantitative separation into the maxi- mum. possible number of groups and that would permit an accurate determination of the constituents of each group without the necessity for an actual isolation of the constituents, would be superior to any previously described method. In the preceding communication,! two methods were described by means of which small amounts of phenols can be accurately estimated. In that paper, no claims were made for the applica- bility of the method to mixtures of phenols. It is the object of this report to show that the methods can also be applied to mixtures of phenols and that the exact quantity of phenol, o-, m-, and p-cresol, p-oxyphenylacetic, p-oxyphenylpropionic, and p-oxyphenyllactic acids, tyrosine, and tyramine can be rapidly and accurately determined. The method has been found to be 1 Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1922, 1, 235, 271 272 Studies on Proteinogenous Amines. XV applicable, without modification, to liquid media in which bacteria — have been allowed to metabolize tyrosine in the presence of salts and glycerol or glucose. In its present form the method is not directly applicable, in its entirety, to more complex liquids such as urine or blood. We hope to modify the method so that it ean be applied to such liquids. The underlying principles of the method are as follows: The simple phenols that are apt to occur associated with liv- ing matter can be divided into four groups. A. Those volatile with steam (phenol and p-cresol). B. Those that will pass into ether from an acidified aqueous — solution (p-oxyphenylacetic, p-oxyphenylpropionic, and p-oxy- phenyllactic acids). C. Tyramine, which can be extracted from an alkaline (sodium - carbonate) solution by means of amyl] alcohol. D. Tyrosine, which remains after all of the other phenols have been extracted and which can be determined colorimetrically if imidazoles and amino-acids are not present in high concentrations. The acidified mixture of phenols is first subjected to a distilla- tion under ordinary pressures. The volatile phenols, phenol or p-cresol,? pass over quantitatively into the distillate where they can be determined colorimetrically if only one of them is present. If these two phenols are present together in nearly equal amounts our method will not estimate them because phenol gives a yellow and p-cresol a red color with alkaline p-phenyldiazonium sulfo- nate; and these two colors cannot be estimated separately with a Duboscq colorimeter. - The residue obtained after the volatile phenols have been re- moved by distillation, is. transferred to a glass dish, concentrated to a syrupy consistency on the water bath, and diluted to exactly 25 ce. Of this solution 10 cc. are transferred to an extraction bottle and extracted ten times with ether using 20 cc. for each extraction. p-Oxyphenylacetic, p-oxyphenylpropionic, and p- oxyphenyllactic acids pass quantitatively into the ether. The com- bined ether extracts are treated with water (25 ec.) and phosphorie acid (5 drops of the 85 per cent acid) after which the ether is * o- and m-cresol are also quantitatively volatile with steam. We have not included them in our analytical process because they are not apt to occur in biological fluids under ordinary conditions. M. T. Hanke one K. K. Koessler 273 removed by distillation, at first under ordinary pressures and then in vacuo. The aqueous solttion is then diluted to 100 ce. and the concentration of aromatic hydroxy-acids determined colorimetrically. Our method does not differentiate between the members of the aromatic hydroxy-acid fraction if they are present together in the same solution. Usually, however, only one member of this group will be present under any given condition. It is possible to tell not only how much of a given acid is present but also which one of the*acids is present by the character of the color produced. The color produced by p-oxphenylacetic acid develops to its maxi- mum intensity within 2 minutes and is slightly brown. The color produced by p-oxyphenylpropionic acid is also fully developed within 2 minutes; but it contains no brown and it undergoes a sharp change after 2 to 3 minutes of color constancy, that is characteristic for p-oxyphenylpropionic acid. The color, which has matched that of the (CR) comparison standard per- fectly, suddenly takes on a cloudy appearance and a bluish tint that makes further comparisons impossible. The color produced by p-oxyphenyllactic acid comes up slowly so that 5 minutes are required to give a color of maximum intensity. The color finally obtained is slightly more yellow than that of the (CR) comparison standard. It is also quite stable: zt does not fade perceptibly for 5 minutes after it has reached its maximum in- tensity. This color does not undergo a change like that of p-oxyphenylpropionic acid. The acid-containing aqueous liquid, which has been freed from volatile phenols by distillation and from aromatic hydroxy-acids by ether extractions, is carefully treated with solid anhydrous sodium carbonate until the liquid stops effervescing. An excess of sodium carbonate (2 gm.) is then added and the alkaline aqueous liquid extracted six times with amy] alcohol, using 20 cc. for each extraction. The amyl alcohol extracts contain tyramine and the alkaline aqueous liquid contains tyrosine. A quintuple extraction of the amyl alcohol extracts with N H.SO, removes the tyramine completely from the amyl alcohol. Tyramine can then be estimated colorimetrically in the aqueous acid liquid after neutralization and dilution to 100 ce. 274 Studies on Proteinogenous Amines. XV The tyrosine fraction, which contains an excess of sodium car- bonate, is transferred to a glass dish, treated with an excess of 37 per cent HCl, and concentrated on the water bath. The crys- talline residue obtained is transferred with water to a 25 cc. eraduated cylinder and diluted to the mark. Tyrosine is then estimated colorimetrically, in this fraction. Sections I to IV of this report contain a detailed account of the results of experiments on known solutions of phenol deriva- tives by which the accuracy of the technique of the method for separating the phenols, described in Section V was experimentally established. EXPERIMENTAL PART. I. Tyrosine and Tyramine Not Destroyed by the Prolonged Action upon Them of Hot Hydrochloric Acid and Sodium Hydroxide. A. Hot 20 Per Cent Hydrochloric Acid. Tyrosine.—Tyrosine (0.1000 gm.) was mixed with 100 cc. of 20 per cent HCl in a 400 cc. round bottomed Pyrex flask. The solution was boiled for 24 hours over an electrically heated sand bath. The resulting pale yellow liquid was evaporated on the water bath in a glass dish. The residue was treated with water and sufficient hydrochloric acid to give a clear solution. This liquid was then transferred to a volumetric flask and diluted to 1,000 ce. Of this solution 0.10 cc. had a He value equivalent to 8.0 mm. (F—MO) and 0. 20 66) 66,2 166 “cc “ (73 16. 0 “ec (F- MO). The entire test liquid must, therefore, have contained 0.1000 gm. of tyrosine which is exactly the amount originally introduced. Continued boiling with 20 per cent hydrochloric acid does not destroy tyrosine to the slightest extent. An entirely similar experiment was carried out on 0.1000 gm. of tyramine hydrochloride. Of the solution finally obtained 0.10 ec. had a color felde equivalent to 8.0mm, (F—MO) and 0.20 cc 6 “ “ 16. 0 “ (F-— MO). M. T. Hanke and K. K. Koessler par hi This, by table? is equal to 0.1000 gm. of tyramine hydrochloride for the entire test solution which is 100 per cent of the amount. originally introduced. Tyramine is not injured to the slightest extent by continued boil- ing with 20 per cent hydrochloric acid. B. Hot 10 Per Cent Sodium Hydrowide. Tyrosine.—Tyrosine (0.10 gm., 10 ce. of a 1 per cent solution) was mixed with 10 cc. of a 20 per cent solution of sodium hydrox- ide. The resulting solution was then heated for 10 hours on the boiling water bath in a small, long necked, round bottomed flask. The resulting colorless liquid was transferred, with water, to a 1,000 ce. volumetric flask, neutralized to litmus paper with 5 Nn H.SOz:, and diluted to 1,000 ec. Of this solution : 0.10 ec. had a color ae equivalent to 8.0 mm. (F—MO) and 0. 20 “ cc “ ““ “ “ 16. 0 “ (F- MO), ' which for the entire solution represents 0.10 gm. of tyrosine, 100 per cent of the amount originally introduced. Tyrosine 1s not injured to the slightest extent ae it is heated for 10 hours with 10 per cent sodium hydroxide. Tyramine.—An entirely similar experiment was carried out on 0.1000 gm. of tyramine hydrochloride. Of the solution finally obtained 0.10 ce. had a color value equivalent to 8.0 mm. (F—MO) and 0.20 “ “ “ ce “ “ “ 16.0 cc (F—MO) which, by table, is equivalent to 0.10 gm. of tyramine hydrochlo- ride, 100 per cent of the amount originally introduced. Tyramine is not injured to the slightest extent when it is heated for 10 hours with 10 per cent sodium hydroxide. II, Phenol and o-, m-, and p-Cresol are Quantitatively Volatile with Steam. p-Cresol (0.01 gm., 10 ec. of the stock 0.1 per cent solution) was mixed with 140 ce. of water in a 500 ce. distilling flask. The 3 Tables for converting colorimetric readings into gm. of phenols are given in the preceding article (J. Biol. Chem., 1922, 1, 235). 276 Studies on Proteinogenous Amines. XV flask was heated with a gas burner through a small hole in an asbestos gauze. The hot vapors were condensed in a spiral glass condenser, the distillate being collected in a 100 cc. grad- uated cylinder. Collections. 1. Total volume of the distillate was 50 ec., of which 0.10 ec. had a color value equivalent to 8.0 mm. (CR). This, by table, is equivalent to 0.0080 gm. of p-cresol, 80 per cent of that originally introduced. 2. Total volume of the distillate was 50 ec. of which 0.40 ec. had a color value equivalent to 6.3 mm. (CR). This, by table, is equivalent to 0.001575 gm. of p-cresol, 15.75 per cent of the amount originally introduced. At this time, 100 ce. of distilled water were introduced into the distilling flask and the distillation was continued. 3. Total volume of the distillate was 60 ce. of which 1.00 ec. had a color value equivalent to 3.0 mm. (CR). This, by table, is equivalent to 0.00036 gm. of p-cresol, 3.6 per cent of the amount originally introduced. 4. The fourth 50 ce. of distillate had no color value. The total recovery of p-cresol in this case was First 50 ce. contained:...........-.as~.-:s»++>+-1 0.0080" ams Second 50 “ HE Ml, Jee a ch Ree eee oa eee 0.001575 “ Third 60 “ Soy 2 ccs ok athe lee. See 0.00036 “ 0.009935 “ which is 99.35 per cent of the amount originally introduced. A second experiment was now carried out to see if a large quan- tity of p-cresol could’ be recovered quantitatively. This was a duplicate of Experiment 1 excepting that 0.10 gm. of p-cresol was used instead of 0.01 gm. Collections. 1. Total volume of the distillate was 100 cc. The color produced with 0.10 ce. of this liquid was so intense that a direct comparison was impos- sible; hence 10 ec. of the distillate were diluted to 100 ce. Of this diluted solution 0.20 cc. had a color value equivalent to 9.4 mm. (CR) which, by table, is equivalent to 0.094 gm. of p-cresol for the entire dis- tillate, 94 per cent of the amount originally introduced. M. T. Hanke and K. K. Koessler BUG _ 2. Total volume of the distillate was 100 ec. of which 0.20 cc. had a color value equivalent to 6.3 mm. (CR). This, by table, is equal to 0.0063 gm. of p-cresol for the entire distillate, 6.3 per cent of the amount originally introduced. 3. Total volume of the distillate was 50 cc. of which 1.00 cc. had a color value equivalent to 7.2 mm. (CR). This, by table, is equal to 0.000072 gm. of p-cresol, 0.072 per cent of the amount originally introduced. 4. Total volume of the distillate was 50 cc. of which 1.00 cc. had a color value equivalent to 3.5 mm. (CR). This is equal to 0.000035 gm. of p-cresol, 0.035 per cent of the amount originally introduced. 5. The fifth 50 ce. of distillate had no color value. The total recovery in this case was frst >) t00ec.contained, <>... os 2. sess 0.094000 gm. Second 100 “ i is eel eee 0.0063 “ ward 50 * LE OR See er ie 0.000072 “ Fourth 50 “ Fa pills hae tate ita 0.000035 <“‘ 0.100407 “ of p-cresol which is 100.4 per cent of the amount originally introduced. Entirely similar experiments were conducted on solutions of phenol, and o- and m-cresol. In every case between 99 and 100.5 per cent of the phenol was recovered and accounted for, colori- metrically, in the distillate. Phenol, o-, m-, and p-cresol are completely volatile with steam and they can be estimated quantitatively in the distillates. III. The Aromatic Hydroxy-Acids Can Be Quantitatively Extracted from an Acidified Aqueous Solution with Ether. When an acidified aqueous solution containing any of the aro- matic hydroxy-acids is extracted ten times with redistilled ether, the aromatic hydroxy-acids pass quantitatively into the ether because the aqueous liquid no longer gives the slightest color with Pauly’s reagent. If the ether, which must surely contain all of the aromatic hydroxy-acids, is now removed by distillation, and the residue diluted with water to a definite volume, a colorimetric determination by the usual process either fails to reveal the presence of any phenol or gives values that are far below the theoretical. 4 Pauly, H., Z. physiol. Chem., 1904, xlii, 508; 1905, xliv, 159. 278 Studies on Proteinogenous Amines. XV We thought, at first, that the aromatic hydroxy-acids might be slightly volatile with steam or ether vapor; but we soon proved that this was not true. The other possibility was that the ether contained some imperfectly volatile substances that prevented the phenol from combining with p-phenyldiazonium sulfonate in alkaline solution. We surmised that the interfering substance was an oxidation product of the ether, perhaps peroxide in char- acter. To remove this, we agitated a sample of redistilled ether with alkaline permanganate until the permanganate had been decolorized, and redistilled the ether layer. An extraction of the aromatic hydroxy-acids with this freshly prepared ether was then carried out. Theoretical values were always obtained with these ether extracts. If, however, the ether was not used for a day or two, low values were again obtained. To avoid the ne- cessity of having to prepare a fresh supply of ether each day, we modified the usual colorimetric procedure as follows. ’ (1-X) ce. of water and X ec. of the aromatic hydroxy-acid-con- taining liquid were treated, for 2 minutes, with 2 cc. of the nitrous acid containing p-phenyldiazonium sulfonate reagent. The 1.1 per cent sodium carbonate solution (5 cc.) was then added. This inverse process, which is similar to the process usually used for the qualitative determination of phenols or imidazoles, gave theoretical values both with pure solutions of aromatic hydroxy- acids and with ethereal extracts even when the ether had not been previously treated with permanganate. Obviously, then, the free nitrous acid present in the reagent modified the interfering sub- stances so that they no longer prevented the combination between the phenol and the diazonium salt. Fortunately, the nitrous acid did not react with the aromatic hydroxy-acids in such a way as to prevent their coupling with the diazonium salt. This inverse process should always be used in the estimation of aromatic hydroxy-acids. The following experiment will illustrate the method used in the extraction and estimation of the aromatic hydroxy-acids. p-Oxyphenylpropionic acid (1.00 ce. of the 1 per cent stock solu- tion was mixed in a 35 cc. extraction bottle® with 0.2 cc. of 95 per ® Any 35 cc. narrow mouthed bottle with a carefully fitted glass stopper will answer the purpose. M. T. Hanke and K. K. Koessler 279 cent H.SO, and 9 cc. of water. 20 cc. of specially prepared ether,® measured by graduate, were introduced into the bottle, the glass stopper inserted and the liquids vigorously mixed for a few minutes. The bottle was then transferred to a centrifuge tube and centrifuged for from 1 to 2 minutes. This gave a sharp separation into two layers. Separation of Ether—The ether layer was separated from the aqueous layer by means of a device similar to that shown in a previous article.’ The ether, instead of being drawn into a Squibb funnel, is drawn into a 700 ec. round bottomed flask. Capillary F is, at first, carefully immersed just below the surface of the ether layer because the vaporization of the ether as it comes in contact with the large surface of the warm flask usually pro- duces enough pressure to eject some ether back into the extrac- tion bottle which would stir up the aqueous layer if capillary F was too deeply immersed. After the first momentary back pres- sure, however, capillary F can be gradually lowered until all but a thin film of ether has been drawn into the receiving flask. The ether not only extracts the aromatic hydroxy-acids but it also reduces the volume of the aqueous layer. It is necessary, therefore, to add sufficient water after each extraction to reestab- lish the initial volume of 10 cc. It is best to mark the extraction bottle with a carborundum pencil at a level corresponding to a volume of approximately 10 cc. The above process was repeated nine times so that a total of ten extractions was made. The aqueous acid liquid gave no Pauly reaction. The ether extracts were then treated with 25 cc. of water and 5 drops of 85 per cent phosphoric acid. The mixture was agitated and sub- jected to a distillation at first under atmospheric pressure and then zm vacuo until the ether had been removed entirely. An ebullition tube was not used and the distillation was not contin- § Commercial ether (800 ce.) is agitated with 50 cc. of alkaline perman- ganate, such as is used in amino nitrogen determinations by the Van Slyke process, and 50 cc. of water in a 2,000 cc. separatory funnel. After the permanganate has been reduced to MnO, the ether layer is poured off and distilled. The redistilled ether so obtained is then ready to use for extractions. 7 See Fig. 1, Studies on proteinogenous amines. III (Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 1919, xxxix, 526. 280 Studies on Proteinogenous Amines. XV ued until all of the water had passed over. The aqueous liquid left in the flask was carefully transferred with water to a 100 ce. graduated precision cylinder and diluted to 100 cc. Of this liquid, 0.10 cc. was mixed with 0.90 ce. of water and 2 ce. of reagent in the right-hand cylinder of the Duboscq colorimeter. After 2 minutes, 5 cc. of the 1.1 per cent sodium carbonate solution were added. The cylinder was transferred to the colorimeter set at 20 mm. and the color compared in the usual manner with that of the (CR) comparison standard. A reading of 3.1 mm. (CR) was obtained. A 0.20 ce. portion of the liquid had a color value of 6.2 mm. This is equal to 0.01 gm. of p-oxyphenylpro- pionic acid, 100 per cent of the amount originally introduced. Entirely similar experiments were carried out on solutions of p-oxyphenylacetic and p-oxyphenyllactic acids. 100 per cent recoveries were obtained in every case. The aromatic hydroxy-acids are quantitatively extracted from an. acidified .aqueous solution with ether. IV. Tyramine Quantitatively Separated from Tyrosine by Means of Amyl Alcohol. A. When a Small Amount of Tyrosine Is Mixed with a Large Amount of Tyramine. Tyrosine (0.20 ec. of the 1 per cent solution), tyramine (5 ce. of the 1 per cent solution), 0.20 ce. of 95 per cent H2SQ,, and 5 ce. of water were mixed in a 35 cc. extraction bottle. Anhydrous sodium carbonate was added carefully until the liquid no longer effervesced. Then 2.00 gm. of the carbonate were added and the liquid was agitated until the solid had dissolved. Redistilled amy] alcohol (20 cc., measured by graduate) was introduced into the bottle, the glass stopper inserted, and the liquids were vigorously mixed for a few minutes. The bottle was then transferred to a centrifuge tube and centrifuged for about 2 minutes. This gave. a sharp separation into two layers. , Separation of Amyl Alcohol.—The amy] aleohol layer was sepa- rated from the aqueous alkaline layer by means of the device and technique described in a.previous article.? The extraction was repeated five times so that a total of 120 ce. of amyl alcohol was used. As in the case of the previously M. T. Hanke and K. K. Koessler - 281 described ether extractions, thesamyl alcohol extracts not only the tyramine but it also markedly reduces the volume of the aqueous layer. It is necessary, therefore, to add sufficient water after each extraction to reestablish the initial volume of 10 ce. Removal of Tyramine from Amyl Alcohol—The combined amyl alcohol extracts were extracted five times, in the same Squibb funnel, with 1.0 n H.SO, using 20 cc. for the first and 10 ce. for each of the remaining four extracts. The sulfuric acid extracts were collected in a 100 cc. glass-stoppered precision cylinder and neutralized to litmus paper with 40 per cent sodium hydroxide. The solution was then rendered very faintly acid by adding a few drops of 1.0 N H.SO,, transferring to a glass dish, and evap- orating on the .water bath to remove the amyl alcohol. The crystalline residue was dissolved in water, transferred to a 100 ce. graduated precision cylinder, and diluted to 100 ce. Of the solu- tion so obtained 0.10 ec. had such a high color value that accurate comparisons were impossible; hence 10 cc. of it were diluted with water to 50 cc. Of this diluted solution 0.10 cc. had a color value of 8.0 mm. (F—MO) and 0.20 “ec “ “ “ “cc “ 16.0 “ce (F—MO). This, by table, is equal to 0.05 gm. of tyramine hydrochloride for the entire original test solution which is 100 per cent of the amount originally introduced. The Alkaline Aqueous Liquid (Tyrosine Fraction).—The alka- line aqueous liquid was transferred with water to a glass dish. The dish was covered with a watch-glass. An excess of 37 per cent HCl was added and the liquid concentrated. on the water bath. The crystalline residue was dissolved in water with the aid of a few drops of 37 per cent HCl, and diluted to exactly 100 ec. Of this solution 0.50 ec. had a color gage equivalent to 8.0 mm. (F—MO) and ° a 00 “ “ce “ “ “ “ce 16. 0 “ (F-— MO) which, by table, is equivalent to 0.002 gm. of tyrosine, 100 per cent of the amount originally introduced. Tyramine is quantitatively extracted from an alkaline (sodium carbonate) aqueous solution by amyl alcohol. Tyrosine, when present im small amounts, does not pass into amyl aleohol from such an alkaline aqueous solution. 282 Studies on Proteinogenous Amines. XV B. When a Large Amount of Tyrosine Is Mixed with a Small Amount of Tyramine. Tyrosine (5.00 cc. of the stock 1 per cent solution), tyramine (0.20 cc. of the stock 1 per cent solution), 0.20 cc. of 95 per cent sulfuric acid, and 5 cc. of water, were mixed in a 35 cc. extraction bottle. Anhydrous sodium carbonate was added carefully until the liquid no longer effervesced. Then 2.00 gm. of the carbonate were added and the mixture agitated until the solid had dissolved. This liquid was then extracted six times with amyl alcohol as described in Section IV, Part A. Alkaline Aqueous Liquid (Tyrosine Fraction M).—This was treated as described in Section IV, Part A. The solution was finally diluted to 500 cc. of which 0.10 cc. had a color eee paumyalert to 7.5mm. (F—MO) and 0. 20 “ce “ce <9 iT “ce 15. 0 “ (F-— MO) which, by table, is equivalent to 0.0469 gm. of tyrosine. Since © 0.0500 gm. of tyrosine had been originally introduced, 0.0031 gm. of tyrosine must have passed into the amyl alcohol. First Amyl Alcohol Extract—The combined amy] alcohol extracts were extracted with Nn H.SO, as described in Section IV, Part A. It was necessary in this case, as it will be in most cases, to remove the sulfuric acid with barium hydroxide to avoid the accumulation of a large amount of salts which would interfere with the subsequent treatment. The combined acid extracts were transferred to a 250 cc. Pyrex flask and heated on the water bath. Barium hydroxide (9.0 gm.) was dissolved in 50 ce. of hot water. The resulting solution was added slowly to the acid liquid. The faintly acid mixture so obtained was digested on the water bath for 2 hours and filtered through a hard folded filter. The paper and contents were thoroughly washed with hot water. The filtrate and washings were collected in a glass dish, exactly neutralized with sodium hydroxide, and evaporated on the water bath. The solid residue so obtained was transferred with the aid of a few drops of 5 N H.SO, and 10 cc. of water to a 35 ce. extraction bottle. The liquid was treated with 2.00 gm. of anhy- drous sodium carbonate. This alkaline solution was then extracted with amyl alcohol'in the usual manner which again divides the M. T. Hanke and K. K. Koessler 283 material into two fractions; the alkaline aqueous liquid II, which should contain the tyrosine that passed into amyl alcohol at the time of the first extraction, and the purified tyramine fraction which should be free from tyrosine. Alkaline Aqueous Liquid IIT (Tyrosine Fraction IT).—This was acidified, evaporated, and finally diluted to 100 ec. as in the case of the main tyrosine fraction. Of this solution, 0.50 ce. had a color ne oe to 11:2 mm. (F—MO) and ie 00 “ “ “ a9 ce 99 4A “cc (F-— —MO) which, by table, is equivalent to 0.0028 gm. of tyrosine. Since 0.0469 gm. of tyrosine was recovered in the first, main tyrosine fraction, a total of 0.0497 gm. of tyrosine was accounted for. This is 99.4 per cent of the amount originally introduced. Purified Tyramine Fraction—This was extracted with N H.SO, as previously described. The acid was nearly neutralized with 40 per cent sodium hydroxide. The resulting faintly acid solution was transferred to a glass dish and evaporated on the water bath. The residue was transferred, with water, to a pre- cision cylinder and diluted to 100 ce. Of this solution, 0.50 cc. had a color ae equivalent to 7.9 mm. (F—MO) and ‘ly! 00 6G) 166. Ce “c “ “cc 15. 8 “ (F-— —MO). This, by table, is equal to 0.001975 gm. of tyramine hydrochloride, 98.75 per cent of the amount originally introduced. — When the concentration of tyrosine is high, a small amount of i passes into amyl alcohol from an alkaline (sodium carbonate) solution. To free the tyramine fraction from this small admixture of tyro- sine it is necessary to conduct a second amyl alcohol extraction on the first tyramine fraction. This slightly longer process with its double amyl alcohol extraction is to be recommended because it makes the determination of tyramine certain and reliable. In most cases it is probably advisable to remove the excess of sul- furic acid from the final tyramine fraction with baryta because such solutions would then be practically free from salts and ready for physiological experiments. 284 Studies on Proteinogenous Amines. XV V. Separation of Phenols into Four Fractions: Volatile Phenols, Aromatic Hydroxy-Acids, Tyramine, and Tyrosine. The Accurate Determination of One Member of Each Fraction. The method outlined below is primarily intended to be used in bacterial metabolism studies on tyrosine. We have carried out a sufficient number of such metabolism experiments to be cer- tain that the method gives accurate results. We hope to report these experiments in the near future. Some of the steps in the following experiment were taken, not because they were necessary in this case, but because they were necessary in the metabolism experiments. A solution containing the following was prepared from the stock solutions; 10 cc. of 1 per cent tyrosine, 4 cc. of 1 per cent tyramine hydrochloride, 4 cc. of 1 per cent p-oxyphenyllactic acid, 2 ce. of 1 per cent phenol, 80 cc. of water, and 100 ce. of Nutritive Medium 3.8 Filtration.—The clear liquid was forced through a Mandler filter. The flask and filter were washed free from. phenols with 200 cc. of water. The filtrate was then treated with 0.50 cc. of 95 per cent H.SO,. Estimation of Volatile Phenols (Phenol).—The filtrate and wash- ings were transferred, with water, to a 1,000 cc. long necked, round bottomed, Pyrex flask. The flask was heated with a gas burner through a small hole in an asbestos gauze. The hot vapors were condensed in a spiral condenser, the distillate being collected in 8 Nutritive Medium 3 contains: INE ON se Stas Piclo sin vb he eR Oe a a ee eee 4.00 gm FON Og 3 ees wht ere die soins ws FER Salonen Fee 2.00 “ OPO ge Cmciih- ask halk cas hee anes ee oe nee 8: 00F IN AGE tS aths cio Wine 05,8 lo wil + avo Sak Oe hk en eee 16.00 “ NaSO, Sinha ax bi x Ow levy ee ws! fer aielase ys WiSbela elubeuntts e tabeieta ian ait yeaa OF OLS BaEICO iis hei Ss ss OE De ro 00 Ca sake dre Sige: age, slave W's .n Sg gd eee cee ae 0.20: CGI COTO) oid pions s nin st ano,s.0 drat siete ieichoe iw» 6 Gee ee 80.00 ce in a total aqueous volume of 2,000 ce. Koessler, K. K., and Hanke, M. T. J. Biol. Chem., 1919, xxxix, 579. °In bacterial metabolism studies, the hydrogen ion concentration is determined, colorimetrically, on 1 ce. of this filtrate before it is diluted with wash water. M. T. Hanke and K. K. Koessler. 285 a 250 ce. graduated cylinder. Exactly 200 cc. of distillate were collected of which 0.10 ee. had a color v: alee equivalent to 9.6 mm. (Ph—R) and 0. 20 “ “ “ “ “ “ 19. 2 ce (Ph— R). This, by table, is equivalent to 0.01666 gm. of phenol for the entire 200 cc. of test liquid. The distillation was continued until about 125 ce. of distillate had been collected. This was diluted to exactly 200 ec. Of this solution 0.20 ec. had a color value equivalent to 4.0 mm. (Ph—R) and 0.40 “ “cc “ “ it “ce ce 8.0 “ (Ph—R) which, by table, is equivalent to 0.0034 gm. of phenol for the en- tire test solution. In all, 0.02006 gm. of phenol was.obtained which is 100.3 per cent of the amount originally introduced. The ‘colors obtained were, in every case, exactly like that produced by pure phenol; hence none of the other phenols volatilized. Estimation of Aromatic Hydroxy-A cids (p-Oxyphenyllactic A cid).— The liquid left in the flask was carefully transferred, with water, to a glass dish and concentrated on the water bath. The pale yellow syrup was transferred, with water, to a 25 ce. precision cylinder and diluted to exactly 25 cc. We will refer to this as the test liquid. Of this acid test liquid, exactly 10 cc. (measured by pipette) were transferred to a 35 cc. extraction bottle and extracted ten times with specially prepared ether as described in Section III of this paper. The ether extracts were then treated as described in Section III. The solution finally obtained (volume 100 cc.) was examined colorimetrically for p-oxyphenyllactie acid. 0.10 ec. had a color vite equivalent to 4.0 mm. (CR) 0: 20 here Ke. 8766 “ “ ““ Re 0 “cc (CR) 0.30 “ce “ “ “ “ “ce it3 12.0 “ (CR). The colors were exactly like that obtained with a pure solution of p-oxyphenyllactic acid. This, by table, is equivalent to 0.0400 gm. of p-oxyphenyllactic acid for the entire 25 ce. of test liquid, which is 100 per cent of the amount originally introduced. 286 Studies on Proteinogenous Amines. XV. Separation of Tyramine from Tyrosine.—The acid liquid left in the extraction bottle, equivalent to 10 cc. of the test liquid, was carefully treated with anhydrous sodium carbonate until the liquid no longer effervesced. Then 2 gm. of the carbonate were added. The mixture was warmed and agitated until the solid had passed into solution. This solution was then extracted six times with amyl alcohol as described in Section IV of this paper. The Alkaline Aqueous Liquid (Tyresine Fraction M).—The alkaline liquid left in the extraction bottle was transferred to a glass dish with 100 cc. of water. The liquid was concentrated on the water bath to a volume of about 50 cc., which removed the ammonia completely. The liquid was then treated with 3.5 cc. of 37 per cent HCl, precautions being taken to prevent loss of the solution through spattermg. The strongly acid liquid was concentrated on the water bath. The crystalline residue was dissolved in water, with the aid of a few drops of 37 per cent HCl, transferred to a graduated precision cylinder, and diluted to 25 cc. This is the main tyrosine fraction M. Of this solution 5 ce. (measured by pipette) were diluted to 80 ee. Of this diluted solution 0.10 ce. had a color oe equivalent to 7.3 mm. (F—MO) and 0. 20 cc it “ cc “ cc 14. 6 “ (F-— MO) which, by table, is equivalent to 0.0912 gm. of tyrosine for the entire original test liquid, 91.2 per cent of the amount originally introduced. As one would expect from Experiment B, Section IV, some tyrosine passed into the amy] alcohol from which it will be recovered when the second extraction is carried out (see below). Amino Nitrogen Determination on Tyrosine Fraction M.— The above tyrosine fraction M (5 cc.) was subjected to an amino nitrogen determination by the Van Slyke method. 1.07 cc. of Ne were obtained at 25° and 747 mm. which is equal to 0.00729 gm. of nitrogen for the entire test liquid. Tyrosine always gives off about 2 per cent more gas by the Van Slyke process than it should theoretically. We have found this to be invariably true and Van Slyke’ gives figures that are in perfect agreement with © Van Slyke, D. D., J. Biol. Chem., 1911, ix, 193. M. T. Hanke and K. K. Koessler 287 this statement. This nitrogen figure must, therefore, be reduced by 2 per cent before the tyrosine value is calculated. The cor- rected nitrogen figure is 0.00715 gm.; which is equivalent to 0.00925 gm. of tyrosine, 92.5 per cent of the amount originally introduced. The check between the tyrosine values obtained by these two methods is good. We wish again to call attention to the fact that this fraction does not contain all of the tyrosine. A second fraction is obtained later when the first tyramine fraction is reextracted with amyl alcohol. The second tyrosine fraction so obtained usually con- tains about 0.003 gm. of tyrosine which is too little to estimate by the amino nitrogen method but which is easily determined colorimetrically. First Amyl Alcohol Extract (Tyramine Fraction I).—This frac- tion contains all the tyramine together with a small amount of tyrosine. A quantitative separation can be effected by carrying out a second extraction with amy! alcohol in alkaline solution. The amy] alcohol was, therefore, extracted with n H.SOx, the acid removed with baryta, the resulting iquid made alkaline with sodium carbonate and reextracted with amyl alcohol as described in Section IV, Part B. Second Alkaline Aqueous Liquid (Tyrosine Fraction IT).—This was acidified, evaporated, and finally diluted to 100 ec. as de- scribed in Section IV, Part B. In this case the ammonia was not removed by evaporation because an amino nitrogen determina- tion is not to be carried out. Of the solution so obtained 0.20 ec. had a color wees equivalent to 4.8 mm. (F—MO) and 0. 40 “ “ “ “ “ce cc 9. 6 “cc (F—MO) which, by table, is equal to 0.0075 gm. of tyrosine for the entire original test liquid. Since 0.0912 gm. of tyrosine was recovered in the main tyrosine fraction M, a total of 0.0987 gm. of tyrosine was accounted for, which is 98.7 per cent of the amount originally introduced. Purified Tyramine Fraction (Second Amyl Alcohol Extract).— This was extracted six times with nN H,SO, as previously described. The acid was nearly neutralized with baryta, the barium sulfate removed by filtration, the filtrate concentrated on the water bath in a glass dish, and the residue dissolved in water and diluted to 288 Studies on Proteinogenous Amines. XV exactly 100 ec. as described in Section IV, Part B.“ Of this solution 0.10 cc. had a color value equivalent to 12.8 mm. (F—MO) which, by table, is equal to 0.0400 gm. of tyramine hydrochloride, 100 per cent of the amount originally introduced. SUMMARY. This paper contains the description of a method by means of which volatile phenols, aromatic hydroxy-acids, tyramine, and tyrosine can be quantitatively separated and estimated. The phenols are determined by a colorimetric process described in the preceding paper. Volatile phenols—phenol, 0-, m-, and p- cresol—are distilled off and estimated in the distillate. Aro- matic hydroxy-acids—p-oxyphenylacetic, p-oxyphenylpropionic, and p-oxyphenyllactic acids—are extracted with ether from the acidified aqueous liquid which has been freed from volatile phenols by distillation. The aromatic hydroxy-acids are estimated in the ether extracts. The remaining liquid, which contains all of the tyramine and tyrosine, is made alkaline with sodium car- bonate and freed from tyramine by extraction with amyl] alcohol. Tyramine is then determined in the amyl alcohol extract; ty- rosine is determined in the alkaline aqueous liquid. The sepa- rations are quantitative and the colorimetric determinations are accurate to 0.5 to 1.5 per cent. 11 It is necessary to remove the excess of H.»SO, with baryta only if physi- ological or isolation experiments are to be carried out on this fraction. In other cases it is simpler to neutralize with 40 per cent NaOH as described in Section IV of this report. SOLUBILITY OF CARBON MONOXIDE IN SERUM AND PLASMA.* By H. R. O’BRIEN{ anp W. L. PARKER.t (From the Department of the Interior, Bureau of Mines, Pittsburgh.) (Received for publication, November 12, 1921.) INTRODUCTION. In the course of the work of the writers on a method of deter- mination of carbon monoxide in blood, the question arose as to how much of the total gas united with the hemoglobin, and how -_ much merely dissolved in the serum. With the known strong affinity of carbon monoxide for hemoglobin (220 to 300 times as strong as that of oxygen) (1) it would be expected that by far the largest percentage would enter into combination with the hemo- globin, but it seemed of value to investigate just what percentage could be accounted for as being in simple solution. This was especially the case with the Van Slyke method, where the total amount of carbon monoxide in the blood is measured, and not the amount of carbon monoxide hemoglobin. Van Slyke (2) passes over the subject in the case of carbon monoxide; but when dealing with oxygen in the blood (3) has a table of deductions and corrections. These are estimated on the basis of Bohr’s recommendation that the solubility of air in serum is roughly nine-tenths that in water at the corresponding temperature. On this subject of the solubility of gases in serum and blood, Bohr goes quite into detail (4). He states! that the absorp- tion coefficients of oxygen and carbon dioxide in whole blood and that of carbon dioxide in plasma cannot be obtained directly, * Published by permission of the Director of the United States Bureau of Mines, Washington. t Assistant Surgeon (Reserve), United States Public Health Service. t Junior Chemist, United States Bureau of Mines. ’ 1 Bohr (4), p. 62. 289 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 1 290 Carbon Monoxide in Serum and Plasma because there is also a chemical reaction taking place. But the percentage of deduction (or relation of the coefficient in plasma to that in water) of various gases, when they do not react, is about the same (5). Therefore, when he made determinations and found that the absorption coefficients of oxygen and nitrogen in plasma are 97.5 per cent of those in water, and the absorption coefficient of hydrogen in whole blood is 92 per cent of the figures in water, he felt safe in making the generalization. For later comparison his figures are given; they are calculated by the aid of his deductions. Coefficient of Absorption in Cc. of Gas (at 0° and 760 Mm.) Oz Ne CO2 15° 38° 15° 38° 15° 38° Blood plasma........}| 0.033 | 0.023 | 0.017 | 0.012} 0.994] 0.541 Whole blood.........| 0.031 | 0-022 0.016] 0.011 | 0.987 | Ov5iT Blood corpuscles.....| 0.025 | 0.019 | 0.014} 0.010] 0.825] 0.450 In the case of carbon monoxide, he says? simply that the plasma absorbs the gas physically, but proportionately to the tension, and in a slightly lesser amount than would be absorbed by the same volume of water. It seemed worth while, therefore, to make direct determinations instead of relying on estimates. During the progress of this work, conducted at the Pittsburgh station of the Bureau of Mines, some other data were obtained which resulted in certain important conclusions on the use of a table of deductions to correct for the amount of carbon monoxide in the blood, uncombined with the hemoglobin. Method of Obtaining Serum and Plasma. As a medium of investigation, beef serum was selected as the one most readily obtainable in quantity; furthermore, most of the related work was being done on beef blood. To supplement this the results were checked on sheep and human sera. Van Slyke and others have assumed that the solubility of carbon monox- ide is the same in serum and in plasma. This was also checked ? Bohr (5), p. 122. ete H. R. O’Brien and W. L. Parker 291 on beef plasma. The serum in each case was gotten by allowing the blood gathered in the slaughter house to clot quietly in sterile containers, pouring off the serum, and removing remaining cor- puscles with a high speed centrifuge running for 10 to 20 minutes. Beef serum has a golden tinge; that of sheep blood is grayer. With human blood, considerable trouble arose from a tendency on the part of the corpuscles to hemolyze; but a few specimens of good serum were obtained. Beef plasma was gotten from blood caught in a sterile container as it poured from the vessels of a freshly killed animal. This was at once poured into 100 ce. bottles containing as a preservative, 0.2 gm. of sodium oxalate, 0.3 gm. of fluoride, and 0.4 gm. of citrate, well mixed. The blood was centrifuged for 3 to 4 hours; the supernatant plasma then pipetted off and further purified by half an hour more in the cen- trifuge. The oxalated blood separates most readily; best results are gotten by centrifuging within 2 or 3 days after it is drawn from the animal. Introduction of Carbon Monowxide into the Liquid. The carbon monoxide used was made by dropping formic acid into concentrated sulfuric acid at 150°C. The evolved gas was washed through potassium hydroxide and stored in a gasometer over water, whence it was delivered as needed, by water displace- ment. The specimen of serum was allowed to reach thermal equilibrium within a thermostat adjusted to the desired tempera- ture. The carbon monoxide reached this same temperature by being forced through a glass spiral also placed within the ther- mostat. From here it passed through a bubbler into the serum. 15 minutes were thought sufficient for the saturation of the sample. Frothing was prevented within the specimen tube by the addition of a drop of caprylic alcohol. The effluent gas passed off through a tube into a hood; the serum was thus saturated at atmospheric pressure. The analysis of the gases dissolved in the serum was performed on the Van Slyke apparatus. All sera saturated below room temperature were kept in ice water, to prevent loss of gas, pending the time when analysis could be made. The method of analysis employed was a modification of that used by Van Slyke in his 992 Carbon Monoxide in Serum and Plasma determination of oxygen, hemoglobin, and carbon monoxide in blood. It consists in general of drawing off the gases under a vacuum produced by a mercury pump, and of analysis of the evolved substances. The exact technique used in the Bureau laboratory is as follows. Technique of the Analysis of Gases Dissolved in Blood Serum by the Van Slyke Apparatus. Before each analysis the apparatus is washed out first with a solution of concentrated ammonia diluted with 3 volumes of water and then twice with distilled water. 5 ec. of serum are measured by pipette into the cup. The end of the pipette is held below the surface of the liquid after starting the delivery, so as to reduce as much as possible exposure of the liquid surface to air. 2 drops of caprylic alcohol are added to the serum, the liquid is drawn into the burette, and the capillary above the stop-cock is sealed with mercury. The apparatus is evacuated and shaken for 2 minutes. The serum is drawn down into the lower bulb and the extracted gases are measured over mercury at atmospheric pressure as described by Van Slyke. The serum is run back into the extraction chamber and shaken further for 1 minute under the same conditions, then the gas volume is read as before. The extraction is repeated until the volume is con- stant. It is recorded as total gases. The volume contracts a little with standing only a few seconds, as COs, dissolved in the small amount of serum on top of the mereury. Oxygen, CO, and N» are completely given off by 1 minute shaking in vacuum; the CO: comes off much more slowly, requiring usually about 2 minutes. 0.5 cc. of 10 per cent KOH is added to the cup, and is carefully drawn into the pipette, the mercury in the leveling bulb being held slightly below that in the burette. The CO. is quickly absorbed. The volume of contraction is noted and the KOH is drawn down into the lower bulb with the serum. About 5 ce. of potassium pyrogallate® solution are next put into the cup and a drop of straw oil quickly added on top of the pyro *120 gm. of KOH are added to 80 ce. of water, and 50 gm. of pyrogallic acid to 150 ce. of water; 300 ec. of the alkali are then mixed with 40 ce. of the acid solution. H. R. O’Brien and W. L. Parker 293 to exclude the oxygen of the air. The “‘pyre” is drawn into the burette, but not the straw oil, 4s the latter interferes with gas absorption by solutions, especially CO by CueCl. Oxygen absorption is slow, but is hastened by working the leveling bulb down and up to insure complete contact of gas with the “pyro.” When the volume has become constant the “pyro” is drawn down into the liquid in the lower bulb. The volume of gas is read as before. The cup is rinsed out with distilled water, as the pyro- gallate left would form a precipitate with the next reagent to be used. A small dropping pipette has been found convenient for removing liquids from the cup. 0.5 ec. of ammoniacal cuprous chloride is now added and care- fully drawn through the remaining gases. COvis quickly absorbed and a constant volume obtained almost immediately. The remaining gas is probably nitrogen, though its volume is some- times higher than would be expected. A complete analysis is made by this method in less than 30 minutes. With practice, readings may easily be made to within 0.005 ce. In checking up the carbon monoxide determination by this method, the total gases extracted from untreated serum were analyzed as a blank. An analysis made on sheep serum will serve as an example of the results that were obtained. Volume of total gas. Volume after absorption by Ist 2nd 3rd extraction. extraction. extraction. KOH Pyro. CuCl ce cc cc cc cc ce 0.270 0.290 0.280 0.065 0.060 0.060 It was found that ammoniacal cuprous chloride does not give contraction in volume of the gases extracted from serum when CO gas is not present. 4400 gm. of cuprous chloride and 500 gm. of NH,Cl are dissolved in _ 1,500 ce. of water. For use this is mixed with NH,OH (sp. gr. 0.90) in ’ proportions of 3:1 (Winkler, L. W., Handbook of technical gas analysis, London, 2nd English edition, 1902, 73). as Sees TABLE I. Solubility of 100 Per Cent CO in Serum and Plasma in Cc. of Gas per Ce. of Serum. Temperature. Beef serum. Sheep serum. Human serum. Beef plasma. Pie i eS le oe | Sm Phenyluraminocystine « 99 | 0.139] 0.041) 0.091 31.1| 68.9 |, 2%10-30.m. per os re. “ 91 | 0.072] 0.048] 0.024] 66.7 | 33.3 | ae of ae pm. (= «“ 92 | 0.036] 0.028] 0.008] 77.8 | 22.2 ae «“ 93 | 0.037] 0.030] 0.007] 81.1 | 18.9 aminocystine. The excretion of such an amount of sulfate sulfur from this source should be accompanied by increased nitrogen elimination, which on the basis of a N : §S ratio in protein of 14 would amount to over 0.300 gm. However, no appreciable change in the nitrogen excretion was observed. It must be concluded that a partial oxidation of the sulfur has occurred. The total sulfur recovered corresponded to 73.1 per cent of the intake. The detailed figures for the other experiments are very similar to the ones just discussed. In the experiments recorded in Table II, ? a H. B. Lewis and L. E. Root 307 cystine was fed to afford a control experiment and to demonstrate that the sulfur of cystine administered appeared promptly in the urine as sulfate sulfur. It seemed possible that the slight oxida- tion of phenyluraminocystine which was observed might be the result of bacterial action in the intestine and that if absorption was facilitated the increase in sulfate sulfur might be lessened. To test out this point the phenyluraminocystine was administered in two doses in order to provide increased opportunity for absorp- tion (Table II). No differences in the degree of oxidation were noted. TABLE III. Rabbit C. Male, red. Weight 3.64 kilos. Daily diet: 40 gm. of oats and 150 ce. of milk. Date. N Total S. a te Remarks, 1920 gm. gm gm. gm Dec. 4 as 0.067 | 0.049 0.018 ss 5 1.44 0.082 0.060 | 0.022 ee 6 1.33 0.069 | 0.053 0.016 aoa | 1.32 0.078 | 0.062 | 0.016 |{1.0 gm. phenyluraminocyst- e 8 | 1.50 0.154 0.060; 0.094 ine subcutaneously (S = - 9 1.48 0.062 0.040 0.022 0.134 gm.). cr. 10 1.49 0.078 0.061 0.017 : gemma 4 | 1.52 0.098 0.080 0.018 fs0lsansdeyntiie aceon “ 12} 1.49 | 0.090/| 0.081 0.009 pea # eget 9-37" |) -9.009°)- G:08a' |. 0-016 [kB PEO Rea << ' 14 1 APY 0.107 0.086 0.021 15 1.49 0.107 0.092 0.015 When the phenyluraminocystine was introduced parenterally, however (Tables III and IV), no oxidation of the sulfur of the molecule occurred. All of the ‘‘extra’”’ sulfur eliminated appeared in the unoxidized sulfur fraction and no increase in sulfate sulfur was evident. The total amount of “‘extra’”’ sulfur recovered was, however, somewhat less than in the experiments in which the phenyluraminocystine was fed. Cystine when injected in amounts comparable to those of the phenyluraminocystine used did not give rise to any appreciable increase in the unoxidized - sulfur fraction. The animal body was able to oxidize completely the sulfur of injected cystine. 308 Oxidation of Cystine In one experiment (Table V) the phenyluraminocystine was fed as a suspension in milk. In this case also no oxidation of the compound took place, and all of the “extra” sulfur was eliminated as unoxidized sulfur. As was to be expected the rate of elimina- tion was slower than in those experiments in which the phenyl- TABLE IV. Rabbit D. Female, grey. Weight 1.98 kilos. Daily diet: 150 cc. of milk and 30 gm. of oats. Total Unondixed Remarks. Date. Total S. SoiS. “15 | 0.065 0.052 Ono Tt) « 161] 0.109 0.096 0.015 ete gm. cystine as hydrochloride subcutaneously (S = 0.134 gm.). subcutaneously as sodium salt “ 18 | 0.083 0.067 0.016 /{1.0 gm. phenyluraminocystine (S = 0.134 gm.). TABLE V. Rabbit E. Female, grey. Weight 1.68 kilos. Daily diet: 150 ce. of milk, 10 gm. of sugar, and 30 gm. of oats. Date. Total 8. aoe Gmoegved Remarks. 1921 gm. gm. gm. Jan. 22 | 0.037 0.027 0.010 i 28-0087 0.026 0.011 “24 | = 0.033 0.025 0.008 “ 25 |, 0.043 0.032 0.011 “ 26 | 0.102 0.029 0.073 me aa MUA rg 0.031 0.036 “ 28 | 0.043 0.029 0.014 «29 | 0.038 0.026 0.012 {1.0 gm. phenyluraminocystine \ suspended in milk. uraminocystine was fed as the soluble sodium salt. More “extra” sulfur appeared in the urine on the day following the administration than in the preceding experiments. The results appear to demonstrate conclusively that, when deamination of the cystine molecule was prevented, the oxidation of the sulfur of the molecule did not take place normally. It . H. B. Lewis and L. E. Root 309 seems probable that in the case of cystine as with the aromatic amino-acids, complete oxidation of the molecule is connected with the deamination process or the further oxidation of the prod- ucts of deamination. The reason for the slight degree of oxida- tion when the sodium salt of phenyluraminocystine is administered per os is not evident, but we believe this to be the result directly or indirectly of some bacterial action. The study of some other derivatives of cystine in which the amino group is “ protected” is in progress. SUMMARY. The sulfur of phenyluraminocystine when administered sub- cutaneously as the sodium salt was not oxidized in the organism of the rabbit, but was eliminated as ‘extra’? unoxidized sulfur. Cystine under the same experimental conditions did not increase the unoxidized sulfur content of the urine. When the sodium salt of phenyluraminocystine was fed to rabbits, a limited oxida- tion of the sulfur fraction of the molecule, resulting in a slight increase in the elimination of sulfate sulfur occurred, although the greater part of the sulfur administered was recovered in the unoxidized sulfur fraction. Since uramino-acids are not broken down in the organism, these results are believed to indicate that the oxidation of the sulfur of the cystine molecule is connected with the process of deamination or the oxidation of the deamina- tion products. BIBLIOGRAPHY. . Dakin, H. D., J. Biol. Chem., 1909, vi, 208. . Dakin, H. D., J. Biol. Chem., 1909, vi, 235. . Dakin, H. D., J. Biol. Chem., 1910, viii, 35. . Dakin, H. D., Oxidations and reductions in the animal body, London, 1912, 71. 5. Friedmann, E., Beitr. chem. Physiol. u. Path., 1903, 1. 6. Gibson, R. B., J. Biol. Chem., 1909, vi, p. xvi. 7. Foster, M. G., Hooper, C. W., and Whipple, G. H., J. Biol. Chem., 1919, xxxviii, 421. 8. von Bergmann, G., Beitr. chem. Physiol. u. Path., 1904, iv, 192. 9. Salkowski, E., Arch. path. Anat., 1873, lviii, 460. 0. Schmidt, C. L. A., von Adelung, E., and Watson, T., J. Biol. Chem., 1918, xxxiii, 501. Hm Wh Re 310 Oxidation of Cystine oi 12. 13. 14. 15. 16. Schmidt, C. L. A., and Allen, E. G., J. Biol. Chém., 1920, xlii, 55. Neuberg, C., and Ascher, E., Biochem. Z., 1907, v, 451. Salkowski, E., Z. physiol. Chem., 1880, iv, 100. Lewis, H. B., J. Biol. Chem., 1912-18, xili, 347. Rohde, A., J. Biol. Chem., 1918, xxxvi, 467. Patten, A. J., Z. physiol. Chem., 1903, Xxxix, 350. THE VITAMINE CONTENT OF MICROORGANISMS IN RELATION TO THE COMPOSITION OF THE CULTURE MEDIUM. By C. EIJKMAN, C. J. C. van HOOGENHUIJZE, anv T. J. G. DERKS. (From the Hygienic Institute of the University of Utrecht, Utrecht, Holland.) (Received for publication, June 3, 1921.) The Vitamine Content of Yeast. That yeast is a valuable source of the antineuritic factor was first demonstrated by Schaumann. It has also been claimed that the growth-promoting, water-soluble B substance is abun- dant in yeast, and it is believed by some workers that these factors are identical, a hypothesis to which we shall return later. Moreover, yeast and yeast extracts have already been introduced into therapeutics, and the latter are also used in ae (soup cubes and so forth). The following observation led us to examine the vitamine content of yeast in regard to the composition of the culture medium. In my attempts to isolate the antineuritic factor from an aqueous solution of extract of rice polishings, among other methods which IJ tried, I removed the sugars through the agency of yeast. But the unwished for result was that the medium had lost its antineuritic properties. This experience, in combination with the fact that bakers’ as well asbrewers’ yeast is obtained by culti- vating them in media originally containing vitamine, gave rise to the supposition that the yeast cell may not be able to synthesize the vitamine but may take it as such from the medium. First, we cultivated Saccharomyces, isolated from bakers’ yeast in vitamine-free media, namely in glucose-peptone broth as well as in a synthetic medium, containing only well known chemi- cal compounds. For this purpose we prepared a solution of 311 ri We Vitamine Content of Microorganisms 0.5 gm. of NaCl, 0.2 gm. of KH2PO,, 0.05 gm. of CaSO,, 0.02 gm. of MgSO,, 5 mg. of FeSO,, 1 mg. of MnCh, 1 mg. of ZnCl, 0.3 em. of NH,Cl, and 5 gm. of dextrose per 100 gm. of water. Henceforth we shall: indicate this solution as ‘‘synthetic wort.” It proved to be suitable for the growth of yeast. The yeast species which we cultivated at 27°C. in vitamine- free media, proved in experiments on polyneuritic fowls to fail in curative effect. In this connection an old experiment of one of ‘us may be called to mind, according to which polished rice, after its preparation, in the cooked state, into a sweet meat (‘‘tapej’’) by the addition of Chinese rice yeast (“ragi”), nevertheless remains deficient in the antineuritic factor. On the other hand, control experiments with the same species of bakers’ yeast, cultivated at 27°C. in aqueous solution of extract of rice polishings (sp. gr. 1.045), after washing with physiological salt solution in order to remove the adherent traces of the medium, gave a distinctly positive result. This aqueous extract had been previously divided into two portions, one of which was boiled for a short time only and then filtered and inoculated with the yeast, whereas the other portion was heated before filtering for 1 hour in the autoclave at 120°C. in order to destroy the antineu- ritic factor. Both of these yielded highly active yeast, but the liquids, separated from the yeast at the end of the fermentation, were found to be inactive. The same experiments were repeated with other materials, but the result was the same as before. This time we chose a Saccharomyces originating from beer yeast, and for media, beer- wort and also a “synthetic wort” so far different from the former that ash of beer-wort, in the same concentration as in beer-wort took the place of the artificial mineral salt mixture, while the reaction of the medium was made slightly acid by the addition of lactic acid. We took the beer-wort from the brewery in two respective stages of its preparation. The sample of the first stage was heated for a short time only at about 75°C., and proved in experi- ments on fowls to contain the antineuritic factor. The second sample was boiled for about 2 hours and was hopped. This sample was found to be practically devoid of antineuritic proper- ties. The pure culture from beer yeast which we used in these Eijkman, van Hoogenhuijze, and Derks 313 experiments, was a so called bottom yeast and the fermentation of the three liquids took place at 6-7°C. As has already been indicated, the “synthetic wort” yielded yeast without any marked curative power against polyneuritis of fowls whereas the two samples of beer-wort produced a highly, active amount of yeast, the liquids themselves on the contrary proving to be inactive at the end of the fermentation. It seems, therefore, that yeast not only takes eventually its antineuritic factor as such from the culture medium but that it is not even capable of synthesizing the vitamine unless the medium contains at least the products of decomposition of the vitamine by heating. In the light of the above mentioned hypothesis this conclusion does not agree with that of Nelson, Fulmer, and Cessna! who in experiments on young rats found that yeast can synthesize the growth-promoting, water-soluble B substance in a medium consisting merely of aqueous solution of mineral salts, NH,Cl, and cane-sugar. These apparently contradictory facts give us reason to doubt whether the antineuritic factor and the water- soluble B substance are really identical. As neither of them has thus far been isolated in an unquestionably pure state, their suggested identity is principally based on the presence or absence of both in the same foodstuffs as also on their conduct towards .- the same physical and chemical agents. Mitchell,? however, drawing attention to the fact that the correlation in all these respects is far from being without exceptions, concludes that in evaluating the data on the occurrence and properties of the two vitamines, there seems to be very good reason for doubltng their identity. Since our experimental results made it probable that the yeast cell may take its antineuritic factor as such from that of the cul- ture medium, although the conclusive proof that the antineuritic factor and water-soluble B are indeed the same has not yet been furnished, the question arose whether from its minute dimensions the yeast cell was able to absorb that factor in a similar way as was already known with regard to charcoal and fullers’ earth. In order to decide this question we took advantage of our experience 1 Nelson, V. E., Fulmer, E. I., and Cessna, R., J. Biol. Chem., 1921, xlvi, 77. 2 Mitchell, H. H., J. Biol. Chem., 1919, xl, 399. 314 Vitamine Content of Microorganisms that yeast, when cultivated in a “synthetic wort,” is devoid of the antineuritic factor. Therefore, such yeast was put into vitamine-containing beer-wort, and the mixture shaken for about half an hour at a low temperature. During this time there oc- curred no noticeable growth of yeast, as could be controlled by sedimenting tests with the hematocrit. After shaking, the yeast was separated from the medium by centrifuging and washing. In experiments on fowls, it did not show any curative effect. The conclusion may be drawn that the process by which the yeast cell takes its vitamine from the medium is not properly of a physicochemical nature; 7.e., an adsorption, but a relatively slow, biological one—a resorption. The Vitamine Content of Bacillus colt communis. According to some authors,’ various bacterial cultures contain water-soluble B. We did not succeed in confirming these experi- ments in regard to the antineuritic factor. A coli stock, isolated from the intestines of a fowl, was cultivated at an adequate temperature (42°C.) in a highly active aqueous extract of rice polishings, which had been cautiously sterilized by discontinuous heating at 75°C. After 3 days cultivation the amount of coli was collected on a bacterial filter and cleansed by - washing. Unlike yeast gained under similar conditions itshowed no marked antineuritic properties. SUMMARY. Though with some reserve, arising from our insufficient knowl- edge concerning the composition of the antineuritic factor, we may conclude as follows: 1. The yeast cell can take its antineuritic factor as such from the culture medium. ‘This is not merely an absorption process. 2. The yeast cell is not able to synthesize, in the strict sense of the word, the antineuritic factor, but only to regenerate it after it has been denatured by heating. 3. Bacillus coli communis, even after having been cultivated in a medium which contains the antineuritic factor, remains devoid of this vitamine. 4, The antineuritic factor and the growth-promoting, water- soluble B substance are not identical. ?Pacini, A. J. P., and Russell, D. W., J. Biol. Chem., 1918, xxxiv, 43. Bierry, H., Compt. rend. Soc. biol., 1919, Ixxxii, 307. A STUDY OF THE EFFECT PRODUCED ON THE COMPO- SITION OF MILK BY THE ADMINISTRATION OF CER- TAIN INORGANIC AND ORGANIC SUBSTANCES. By W. DENIS, WARREN R. SISSON, anp MARTHA ALDRICH. (From the Laboratory of Physiological Chemistry of the Tulane University Medical School, New Orleans, and the Department of Pediatrics of the Massachusetts General Hospital, Boston.) (Received for publication, November 21, 1921.) Attempts to modify the composition of milk by experimental means have been numerous and as a rule unsuccessful. To reopen the subject would therefore scarcely appear justifiable except for the fact, that in virtue of the developments in microchemical blood analysis which have taken place during the last decade, we felt ourselves able to approach the problem from a new standpoint. Observations regarding the passage into the milk of drugs admin- istered by mouth have been made for many years, and while such observations are mainly of a qualitative nature, and give as a rule no idea of the amount of absorption, they still indicate the possibility of producing experimentally changes in the mam- mary secretion by the production of changes in the composition of the blood. It is now generally conceded that it is possible by dietary measures to influence the concentration of certain of the inorganic and organic non-protein constituents of the blood, whereas in the case of certain other normally occurring constituents forced feeding either produces no demonstrable effect or at most a transitory rise. In the first class urea and phosphates may be mentioned, whereas in the latter striking examples are to be found in sodium chloride and in the salts of calcium. In view of these facts it would seem that it should in most cases be possible to predict the result of feeding experiments made with the object of producing changes in the composition of the milk (at least as regards a single constituent), provided data are on hand 315 316 Composition of Milk concerning the possibility of altering the concentration of this specific constituent in the blood. The experiments described below were undertaken with the purpose of obtaining experimental proof of the validity of the above hypothesis. In this paper are given the results obtained by the administration of urea and calcium chloride. Our experimental methods were as follows: An appropriate amount of the substance whose absorption was to be studied was dissolved in about 300 cc. of water and poured down the animal’s throat without the use of stomach tube. All samples of blood were taken by venepuncture from the external jugular vein. The analytical methods used were as follows: For Milk. Urea.—By the urease method according to the technique described by Denis and Minot (1). Calcitum.—By Lyman’s (2) method, slightly modified, as an extended experience has shown us that in experimental work better results are obtained if several standards of varying strength are provided instead of the single standard recommended by Lyman. For Blood. Urea.—By the method of Folin and Wu (3). Calcitum.—By the method of Lyman (2). Absorption of Urea. The ability of the body rapidly to absorb ingested urea is now too well known to require comment. It has been shown by Mar- shall and Davis (4) that when urea is injected intravenously it is stored in the muscles and in the various organs of the body in amounts approximately equal to the concentration found in the blood. Of recent years much experimental work has been carried out on the effects of high and low protein diets on the concentration of blood urea; in general the results of this work lead us to believe that it is possible, even in subjects with normal kidney function to increase the concentration of blood urea by relatively short W. Denis, W. R. Sisson, and M. Aldrich 317 periods of high protein feeding, and to decrease it by the use of low nitrogen diets. It was also shown some years ago by Denis and Minot (5) that by the administration of diets high in protein it was possible to obtain from cows, milk of high urea content, whereas the same animals when fed on low protein diets produced milk containing relatively small amounts of urea. These studies were unfortunately not accompanied by observations of blood urea. In Experiments 1, 2, and 3, the results of which are tabu- lated in Tables I, II, and III we have studied the relative effects of large doses of urea on the blood and milk. TABLE I. Absorption of Urea. Experiment 1.—Goat1, weight 31.8 kilos. An old animal, whose average yield of milk for 24 hours was about 150 ce. ; r Urea N Increase of Time. A del V ee of per ae of pase 1920 gm. ce. mg. per cent Beiveets S15 8.0... .. og ees vi 65 17.8 Chee Caner eee 5 WOROOne Pet eica aero 5 il hee 0 (0) mais eon aad 5 15 19 6.9 Liat) UM tee eat ae re =! 5 POO MD AMS yo rks 7 26 46.0 et ce Aerts area fe 3 95 433 .7 Naveed OU BI. nc cine <,0.0i0 37 29 62.9 As will be seen such treatment causes a rapid rise in the urea nitrogen fraction of both blood and milk. If we disregard the results of Experiment 1 in which no blood examinations were made, it will be noted that the urea content of the milk had increased to more than 30 per cent of its former value in 1 hour in Experiment 2, whereas in Experiment 3 a similar result was obtained in these specimens of milk taken 3 hours after the administration of urea. A similar relation exists between the two experiments as regards the final concentration of urea attained. In Experiment 2 the maximum figures were obtained 6 hours and 45 minutes after the administration of the initial dose, the rise being represented by an increase of 163 per cent in the milk and 206 per cent in the 318 Composition of Milk plasma, whereas in Experiment 3 after 11 hours the rise in the urea nitrogen of the milk was represented by an increase of 97 per cent. TABLE IL. Absorption of Urea. Experiment 2.—Goat 2, weight 22.5 kilos. A young animal in the first month of lactation. The average yield of milk for 24 hours was about 1 liter. ata i esc Volume of Urea N per 100 cc. fuer Sa N a fared milky») 22" ot eee Milk. Plasma. Milk. Plasma. 1920 gm. cc. mg. mg. per cent | per cent June 9, 9.00 a.m.... 600 14.4 15.0 ORT aa 5 IOS 5 ae 5 96 19 31.9 IPAS FOI sae 5 J SORES eee 40 37 15.69 | 126.0 ela” Sa tee 40 38 163.8 | 206.6 ORO 5.5 170 25.5 77.0 June 10, 8.30 a.m.... 420 12.4 TABLE III. Absorption of Urea. Experiment 3—Goat 3, weight 36 kilos. An old animal in the first month of lactation. The average 24 hour yield of milk was about 700 cc. ‘ : Increase of urea N Be adnan Volume of Urea N per 100 ce. eer ararniil Milk. Plasma. Milk. Plasma. 1920 gm. cc. mg. mg. per cent | per cent June 22, 6.00 a.m.... 350 13.4 20 6:10) 5 5 TO0ex"* 5 Sali. 5 Ole Strat 5 65 17.6 31.3 12° 00*mi es: 30 19 30 44.7 50.0 3.00 p.m.... 44 19 31 41.7 55.0 BOO eZ yd, 50 26.4 97.0 June 23, 9.00. a.m.... 320 29 116.4 5.00 p.m... 200 Re 14.1 June 24, 9.00 a.m.... 310 17 27.5 AE a W. Denis, W. R. Sisson, and M. Aldrich 319 Absorption of Calcium Chloride. Feeding experiments made with the object of modifying the calcium content of milk have usually been unsuccessful (6). Similar negative results have been the reward of investigators who have attempted to increase the calcium content of the blood by the administration of calcium salts (7). Itis not surprising, therefore, that our attempts to change the calcium concentration of the milk by feeding large doses of calcium chloride (as described in the tabulation of the results obtained in Experiments 7, 8, and 9, Tables IV, V, and VI) should also have yielded negative results as regards both milk and blood. TABLE IV. Absorption of Calcium Chloride. Experiment 7.—Goat 1. Fe fas. Volume of Per 100 ce. of milk. Increase. Face: Ca Cl Ca Cl 1920 gm. ce. mg. mg. per cent | per cent May 29, 9.00 a.m.... 85 15.9 156 SF30) * 2 10230) “* 2 158 156 0 0 ESO 2 12230) p.m... 2 37 158 180 0 15.3 2331 eee 13 158 189 0 15.3 In Experiments 7 and 8 unmistakable increase in the chloride concentration in the plasma and milk was observed. In Experi- ment 9, however, this increase was confined to the plasma, no change being noted in the milk. As it occurred to us that our negative results in the case of these experiments in which calcium chloride was administered by mouth might be due to slow absorption of this salt from the stomach and small intestine we performed a final experiment in which 1.87 gm. of calcium chloride contained in a volume of 75 cc. were injected intravenously. The duration of the injection was 3 min- utes. The plasma from a sample of blood taken 17 minutes after the end of the injection showed an increase in the calcium content of 150 per cent, 3 hours later the plasma calcium was still 48 per cent above the initial figure. Even with this unmistakable 320 Composition of Milk increase in the calcium concentration of the blood, however, no increase was noted in the calcium concentration of several samples of milk removed at intervals during the day. TABLE V. Absorption of Calcium Chloride. Experiment 8.—Goat 2. Per 100 ce. , Increase. CaCh |“ Vol=.|———-— > 3 | > Time. admin-jumeof| ilk. Blood. Milk. Blood. istered.| milk. Ca | Cl | Ca | Cl | Ca | Cl | Ca | Cl per | per | per | per 1920 gm. cc. mg. | mg. | mg.| m9. | ont! cent | cent | cent June 29, 8.50 a.m...... 320 | 160) 140) 8.6) 354 OR OO Se > var 4 TOZO0 ree 4 00 oa ae 4 12200 meee 4 44 161) 164) 8.5} 360} O |17.1) O]| 1.6 ' So O0MD Mise sete 43 160} 169} 8.6} 3880} O |20.7| O |10.1 June 30, 9.00 a.m...... 160 361; 0 1.6 TABLE VI. Absorption of Calcium Chloride. Experiment 9.—Goat 3. Per 100 ce. Increase. CaCl: | Vol-_ |————_—_| _—__—_——— Time. cae Bee ot Milk. Plasma. Milk. Plasma. Ca}; Cl. | .Ca Ch | Cai GlaiiGasimen 1920 gm. ct. mg.| mg.| mg.| mg. past ped bsok pie: June 29, 8.40 a.m...... : 420 163} 169) 8.7} 360 OOO. eee eeias 4 TOO0 Eas 4 1100.) Saree 4 2 OO sieey hander. 4 72 169) 8.6} 402 0 0 |10.1 310 p.m.).o. 60 164) 180} 8.7| 482) 0O 0 0 |20.0 June 30, 9.00 a.m...... 380 164| 170) 8.6) 360) O 0 0; 0 The results of the experiments described above would appear to lend support to the hypothesis outlined in the opening paragraphs of this paper. If we believe that the mammary tissue may act W. Denis, W. R. Sisson, and M. Aldrich 321 as a temporary storage place for certain substances that are not rapidly excreted our work would’suggest that both the chlorine ion and urea may be found in increased amounts in the milk when their concentration in the plasma, and probably also in the mammary tissue, rises to a high level. The calcium ion on the other hand apparently acts in an entirely different manner. The cause of this difference in behavior may be ascribed to the relative toxicity of calcium with the resultant lowering of the dosage, or to the fact that calcium cannot be retained in the mammary tissue but is TABLE VII. Absorption of Calcium Chloride. Experiment 10.—Goat 38. | Per 100 cc. Increase. Vol |__| Time. CaCleinjected.*| ume of) Milk. | Plasma. | Milk. | Plasma. muk, Ca [Cl | Ca. I) Clo Cay) Cl [5Car cel per | per | per | per 1920 HLS Ss mg. | mg. | mg.) Mg.) cent| cent | cent | cent July7, 9.30a.m. 360 | 200) 164/ 128) 350 10.385 “ |Injected 1.87 gm. CaCl. P 10.55.“ 32) 400 150) 14 cs I 55 | 200) 175 O66 1.50 p.m. 35 | 196] 180} 19) 350) O | 9.7} 48 O 210) 30 | 200) 173) 18] 350) O | 6.4) 1.4) 0 July8, 9.10a.m. 200 | 201) 180 0 | 9.7 * At 10.30 injected into the external jugular vein 75 ce. of a 25 per cent solution of calcium chloride. excreted almost immediately by the intestine and kidney. In view of the unmistakable increase in the chlorine concentration of the milk of the animals receiving calcium chloride we feel in- clined to accept this latter view. BIBLIOGRAPHY. 1. Denis, W., and Minot, A.S., J. Biol. Chem., 1919, xxxvii, 353. 2. Lyman, H., J. Biol. Chem., 1917, xxix, 169. 3. Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. 322 Composition of Milk 4, Marshall, E. K., Jr., and Davis, D. M., J. Biol. Chem., 1914, xviii, 53. 5. Denis, W., and Minot, A.S., J. Biol. Chem., 1919, xxxviii, 453. 6. Newman, J., Milch Ztg., 1893, xxii, 701. von Wendt, G., Skand. Arch. Physiol., 1909, xxi, 89. Fingerling, G., Landswirtsch. Versuch. Sta., OMS iocvieuts 7. von Fenyvessy, B., and Freund, J., Z. Immunititsforsch., 1913, xviii, 666. Heubner, W., and Rona, P., Biochem. Z., 1919, xciii, 187. Meigs, E. B., Blatherwick, N. R., and Cary, C. A., J. Biol. Chem., 1919, xxxvii. 45. Denis, W., and Minot, A.S., J. Biol. Chem., 1920, xli, 357. Clark, G. W., J. Biol. Chem., 1920, xii, 89. THE THERMOSTABLE ACTIVE AGENT OF PIG’S PANCREAS. By WALTER JONES. (From the Laboratory of Physiological Chemistry, Johns Hopkins Medical School, Baltimore.) (Received for publication, November 25, 1921.) More than a year ago the writer stated that yeast nucleic acid is decomposed into its nucleotides by a boiled aqueous extract of pig’s pancreas without giving rise to any increase in the acidity of the solution; and the crucial bearing of this circumstance upon the chemical constitution of yeast nucleic acid was discussed in detail.! The context of the article and of other writings on the subject indicated clearly that the word ‘“‘acidity’’ was intended to mean “titratable acidity” (not necessarily hydrogen ion concentration)? and the implication was definite that titrations had been made. In the experimental part (which was not otherwise complete) no specific description was given of a titration, the principal description being of an experiment dealing only with hydrogen ion concentration. This experiment Levene’ now criticizes in the following words: “The second experimental proof of Jones’ theory is the following: By a pancreas enzyme, yeast nucleic acid was cleaved to its nucleotides. At the starting point of the experiment the hydrogen ion concentration of the reacting mixture was brought to pH = 6.4, and at the end of the experi- ment there was no apparent change of the color of the indicator added to 1 Jones, W., Am. J. Physiol., 1920, lii, 203. 2 See Jones, W., Monograph on biochemistry, London, 1920, 47, ‘‘the increased acidity caused by the decomposition of two grams of nucleic acid should require about 8 cc. of tenth normal sodium hydroxide for neutralization toward phenolphthalein. There is no doubt about this. The nucleotides imitate the conduct of phosphoric acid toward alkalis and indicators.’’ 3 Levene, P. A., J. Biol. Chem., 1921, xlviii, 119. 323 324 Pancreatic Ferment the original solution. Hence the author concludes that no acid radicles could be liberated as the result of the hydrolysis. The reasoning is not correct. According to either theory, nucleic acid is a polyphosphoric acid and when brought to a pH = 6.4, it possesses considerable buffer effect. Furthermore, each nucleotide is a comparatively weak acid and when liberated does not affect the hydrogen ion concentration of the buffer very markedly. Since the dissociation constant of the nucleotides has not been measured, it is not possible to express the reaction in quantita- tive terms. Experimentally, however, we convinced ourselves that when a solution of guanosinphosphoric acid is brought to a pH = 6.4, it stands the addition of an equal volume of a solution of free guanosinphosphorie acid of the same concentration before any change of color of the indicator can be noticed. Taking further into consideration the fact that a solution of nucleic acid is not perfectly colorless, that an extract of the pancreas always contains a considerable quantity of phosphates and also is not colorless, one easily realizes that the argument of Jones carries but little weight.” This criticism would be more plausible if the following were ignored: 1. The pale yellow color of a boiled pancreas extract matches well the alkaline color of the indicator used (brom-cresol). The change of this indicator from its purple color to yellow can be detected just as well in a pale yellow solution as in distilled water when a proper arrangement for the observation is made. 2. A 2 per cent solution of Merck’s yeast nucleic acid is scarcely colored. But aside from this, the specimen of yeast nucleic acid that we have been using in this laboratory is snow-white and forms a colorless 2 per cent solution. A method will be described later for its preparation from yeast and from the various commercial sources of yeast nucleic acid. 3. All four of the nucleotides are strong enough acids to turn methyl orange. 4, Levene’s experiment with guanosinphosphoric acid has been repeated with both adenine nucleotide and guanine nucleotide. The results obtained are strikingly different from the result given by Levene. 15 cc. of a7 per cent solution of adenine nucleo- tide in hot water were cooled, colored with brom-cresol, and brought to a hydrogen ion concentration of pH = 6.4. The purple solution was divided into two equal parts, one of which 4Levene does not state what indicator he used. This is of vital importance. Walter Jones 325 was used for comparison while the other was carefully treated with a 7 per cent solution of adenine nucleotide. The first drop pro- duced a color change that could doubtfully be detected; the second drop produced an unmistakable change in the color of the indicator; the third drop so changed the indicator that its pre- dominant color was yellow. Essentially the same results were obtained with guanine nucleotide. 5. No titratable acidity is produced when yeast nucleic acid is decomposed into its nucleotides by the action of boiled extract of pancreas, as the following experiments will show. A boiled aqueous extract of pig’s pancreas was prepared accord- ing to the directions previously given! and 3 gm. of yeast nucleic acid were dissolved in 150 cc. of the warmed extract. This solution, in which the presence of the nucleic acid could be shown even after excessive dilution, was quickly cooled to the room temperature, placed in a burette, and compared with a tenth normal solution of sodium hydroxide using phenolphthalein as an indicator with the usual precautions. The results of ten titrations given in Table I show that for 5 cc. of the nucleic acid solution close to 5.60 cc. of the alkali were required for neutralization toward phenolphthalein. The remainder of this nucleic acid solution was preserved with a few drops of chloroform and allowed to digest in the thermostat at 38° for 24 hours. At the end of this time the nucleic acid had entirely disappeared, having been converted into its nucleotides. After cooling to the room temperature the solution was placed in a burette and again compared with the tenth normal solution of sodium hydroxide, using phenolphthalein as an indicator with the usual precautions. The results of eight titrations given in Table I show that for 5 cc. of the digested solution close to 5.75 cc. of the alkali were required for neutralization toward phenolphthalein. Weighed portions of adenine nucleotide were then added to 5 cc. portions of the digested fluid and the product was titrated as before. The titratable acidity was found increased by the theo- retical demand.*® 5 The nucleic acid dissolves readily in the warmed extract without the addition of alkali. 6 The experiment here reported was done after Levene’s article ap- peared, but the results do not materially differ from those of numerous older and recent experiments. 326 Pancreatic Ferment The two formulas which have been proposed in turn by Levene for yeast nucleic acid and the one proposed by Thannhauser and Sachs? demand the production of titratable acidity in each of the ten experiments corresponding to 4.60, 2.31, and 5.25 cc. of tenth normal sodium hydroxide, respectively. But no titratable acidity at all was found, although adenine nucleotide exhibits its theo- retical titratable acidity in a boiled aqueous extract of pig’s pancreas. It is true that no experimental evidence exists to show the points at which the individual nucleotide groups of yeast nucleic TABLE I. Before digestion (nucleic acid). After digestion (nucleotides formed). E | ; Alkali ela l Bere f gee Adenine calor nucleic |Alkali required.| °’C4 ae of nucleo- nucleotide |Alkali required.| ted for 5 aci Rocit ate tides added. ee. of used one formed, extract. eat PRIETO ce. ce. mg ce ce 3.42 3.80 5.50 eal 3.55 5.60 3.81 4.27 5.60 3.81 4.19 5.50 3.60 4.07 5.65 3.30 3.63 5.50 3.12 3.43 5.50 4.00 4.56 5.70 4.18 4.68 5.60 4.12 4.70 5.70 3.97 4.49 5.65 3.30 3.70 5.60 3.51 3.90 5.55 3.21 3.50 5.50 3.70 4.07 5.50 3.60 3.96 5.50 4.07 4.60 5.65 3.19 50 6.25 4.13 4.71 5.10 3.34 50 6.42 4.07 Used for comparison. 3.18 | Used for comparison. Miean v.12 5). Bt ae: 5.60 Mieamiceiee cg cians Uke eee 5/0 acid are joined to one another; but there exists abundant experi- mental evidence to show the points where they are not joined. The nucleotide linkages do not involve any one of the phosphoric acid groups and no formula for yeast nucleic acid can be accepted in which this kind of nucleotide linkage is assumed. The carbo- hydrate formula that I have used is arrived at only by exclusion and is intended specially to indicate the points where the nucleo- tide linkages do not exist. 7 Thannhauser, S. J., and Sachs, P., Z. Physiol. Chem., 1921, exii, 189. Walter Jones Be | In the same article Levene criticizes another kind of experi- mental evidence that I have adduced, and in the following words: “The curve expressing the rate of hydrolysis of yeast nucleic acid is identical with that of a mixture of the four nucleotides. Accepting the experiment as correct, what does it demonstrate? It proves that the union between individual nucleotides is more labile than that between the phos- phorie acid and the carbohydrate in each nucleotide. It is then self- evident that the first step in the hydrolysis of the nucleic acid molecule is the formation of four nucleotides. The further progress of hydrolysis of the nucleic acid is the same as of four nucleotides.” I raised this question myself long ago and dealt with it in the following words.® “Tt is of course possible to draw other conclusions that can be adjusted to the facts, but they all involve the assumption of curious coincidences and compensations which would cause phosphoric acid to be liberated with equal ease from different kinds of linkage, or that the liberation of phos- phoric acid from one kind of linkage is excessively slower than from another. After careful consideration of such matters, we believe we have drawn the correct conclusion.” The contents of the present article show that this statement did not exhibit poor critical judgment. In Levene’s article? is contained the following unfortunate sentence. “It is peculiar that Jones, in the latest edition of his monograph, in discussing the theories of the constitution of yeast nucleic acid, does not at all refer to the theory of the present writer.”’ It is not peculiar. I knew at the time what is contained in the present article. Had I known what is contained in Levene’s recent communication,’ I should not have referred to his theory of the constitution of thymus nucleic acid. CONCLUSION. The existence in the pancreas extract of an easily detected thermostable agent which decomposes yeast nucleic acid only as far as its nucleotides, is more interesting in its physiological significance than for any light that it throws upon the chemical 8 Jones, W., and Read, B. E., J. Biol. Chem., 1917, xxix, 126. 328 Pancreatic Ferment constitution of nucleic acid. The first alteration that yeast nucleic acid undergoes in its decomposition by tissue extracts is assumed to be the production of nucleotides. How then does it happen that extract of pig’s spleen and of other tissues, that do not contain the thermostable agent in question, can nevertheless bring about the progressive decomposition of nucleic acid with the formation of free phosphoric acid and free purine bases? It would appear either that there are two ferments which can decompose nucleic acid into its nucleotides, one of which is de- stroyed by heat, or that the decomposition of nucleic acid by tissue extracts does not proceed along conventional lines. Exam- ination of the matter is now proceeding. A RAPID COLORIMETRIC METHOD FOR THE QUANTITA- TIVE DETERMINATION OF THE INORGANIC PHOS- PHORUS IN SMALL AMOUNTS OF SERUM. By FREDERICK F. TISDALL. (From The Nutritional Research Laboratories of the Hospital for Sick Chil- dren and the Department of Pediatrics, University of Toronto, Toronto, Canada.) (Received for publication, November 25, 1921.) Principle of the Method. The principle of the method consists in the precipitation of the phosphorus, in a trichloroacetic acid extract of serum, as strych- nine phosphomolybdate, the isolation of the precipitate by the use of a centrifuge and small quantities of water, and the subse- quent development of a brilliant green color produced by the reduction of the molybdenum present in the precipitate. The reduction is accomplished by the use of potassium ferrocyanide and HCl. The Method. Precipitation of Protein.—1 cc. of serum is transferred to a 15 ce. centrifuge tube and to this are added 5 ce. of a 6 per cent solution of trichloroacetic acid. The mixture is thoroughly mixed with the aid of a glass rod and allowed to stand for 4 minutes. It is then centrifuged for 4 to 5 minutes at about 1,500 revolutions per minute and the supernatant fluid poured off. Precipitation of Phosphorus with the Strychnine Molybdate Reagent.—5 cc. of the supernatant fluid are measured into an ordi- nary 15 cc. graduated centrifuge tube, the outside diameter of which is 6 to 7 mm. at the 0.1 cc. mark. Water is added to bring the volume to 6 cc., followed by 2 cc. of the strychnine molybdate reagent which should be added drop by drop, and the tube shaken three or four times during the procedure. The contents of the tube are then thoroughly mixed by holding the tube at the upper 329 330 Inorganic Phosphorus in Serum end and tapping the lower end with the finger giving it a circular motion. The contents are allowed to stand for 10 minutes during which time they are thoroughly mixed twice as outlined above. Washing of Precipitate——After the 10 minutes has elapsed the tube is centrifuged at 1,500 revolutions per minute for 3 minutes, the supernatant fluid is poured off and the mouth of the tube wiped with a dry cloth. 3 cc. of water are allowed to run down the sides of the tube which removes any adherent supernatant fluid. The residual supernatant fluid (about 0.1 cc.) is thoroughly mixed with the added water by tapping the lower end of the tube with the finger giving it a circular motion, while the precipitate is disturbed as little as possible. The mixture is centrifuged for 1 minute at 1,500 revolutions per minute, the supernatant fluid is poured off and the above procedure repeated, making two wash- ings in all. Development of Color—After the final supernatant fluid has been removed, 2 cc. of a 1 per cent solution of NaOH are added and the contents mixed with the aid of a glass rod. This causes all the precipitate to go into solution. Water is added to 10 ce. and the contents are transferred to a 100 cc. glass-stoppered volumetric flask. Traces of the solution remaining in the centri- fuge tube are washed into the flask by means of two lots of 10 ce. of water, so that the total volume of fluid in the flask is 30 ce. 20 cc. of a 20 per cent solution of potassium ferrocyanide are then added, followed by 10 ec. of concentrated HCl. The flask is inverted two or three times and allowed to stand 10 minutes. Water is added to 100 ce., the contents are thoroughly mixed, and the color is read in the colorimeter against the standard. Preparation of the Standard.—1 cc. of a solution of KH:PO, con- taining 5 mg. of P per 100 cc. (219.3 mg. of KH;PO, (Merck) in 1,000 cc.) is measured into a graduated centrifuge tube, which contains 5 ec. of water, and the contents are thoroughly mixed. 2 ce. of the strychnine molybdate reagent are then added drop by drop. This step, and the washing of the precipitate and the development of the color, are carried out at the same time and in the same manner with both the standard and the unknown. The Volume of the Precipitate—The amount of precipitate obtained in the standard solution after it is centrifuged is almost exactly 0.1 cc. of volume. If the amount of precipitate obtained i F. F. Tisdall 331 in the unknown is 0.2 ce. or more, its solution (in 1 per cent NaOH) should be made up to a definite volume in the centrifuge tube and an aliquot taken which would contain approximately 0.1 ec. of the precipitate. If the amount of precipitate obtained in the unknown is about one-half the amount in the standard, its solution should be made up to 5 ee. and transferred to a 50 cc. volumetric flask with the use of two lots of 5 cc. of water. In all the subsequent steps the volumes used should be halved. Calculation—When the unknown is made up to 100 ec. and the standard solution is set at 20 in the colorimeter the calculation is as follows: 20 Unknown X 6 = mg. of P per 100 ce. of serum. When the unknown is made up to 50 ee. the result is divided by 2. Preparation of the Strychnine Molybdate Reagent. Solution A is prepared by dissolving 50 gm. of ammonium molybdate in 150 cc. of warm water. If not clear this solution should be filtered. Solution B consists of 2 volumes of concentrated HNO; and 1 volume of water. Solution C is prepared by pouring 1 volume of Solution A into 3 volumes of Solution B. Solution D consists of strychnine nitrate 7.5 gm., water to 500 ce. The water may be warmed to facilitate solution. 1 volume of Solution D is poured into 3 volumes of Solution C. This constitutes the strychnine molybdate reagent. The reagent should stand 24 hours before it is used. It will keep for at least 1 month. After the reagent has stood for 1 or 2 days a slight precipitate forms and when this occurs it should be filtered. - 2 ce. of the reagent will precipitate 0.2 mg. of P. Protocols. The development of the color on the addition of potassium ferrocyanide and HCl is in direct proportion to the amount of molybdenum present, as long as that amount does not exceed double the quantity, or is not less than half the quantity present 332 Inorganic Phosphorus in Serum in the standard. A solution of strychnine phosphomolybdate was prepared so that 2 cc. contained approximately the amount of precipitate obtained from 0.05 mg. of P (1 cc. of the standard solution). 1, 2, and 4 cc. of this solution were placed in 100 ce. volumetric flasks, the volumes made up to 30 cc. with water, and potassium ferrocyanide and HCl added. After 10 minutes the flasks were made up to volume and the color was read in the colorimeter with the flask containing 2 cc. of the solution taken as the standard at 20. Reading obtained with 1 ee. of solution TABLE I. Estimation of Known Amounts of Phosphorus in a Solution of KH2PO,. : ag to) 3 a 35 | & = S Volume of nd he; Es S “s > precipitate after = Reaches) ~_; | Reading.| Theory. | Error, =. =I centrifuging. s | a8 aS 5§ 8 © 4 6851-53 =p a § | ges] aa < oe nD < < cc mg ce cc ce. cc per cent 0.25 |0.0125| Less than 0.05. 50 | 10 5 40.8 40.0 —2 0.5 |0.025| About 0.05. | 50] 10 5) | 20'4" 2 ooo" iaaaes 1.0 |0.05 0.1 100 | 20 10 | Standard = 20 2.0 |0.10 0.2 200 | 40 20 19.8 20.0 +1 3.0* |0.15 0.3 100 |} 20 10 19.6 20.0 +2 4.0* |0.20 0.4 100 20 10 19.8 20.0 +1 * The solutions of the precipitate obtained from the 3 and 4 cc. samples were made up to 6 and 8 cc., respectively, and 2 cc. aliquots transferred to the 100 ce. flasks. = 39. Theory 40. Error + 2.5 per cent. Reading obtained with 4 cc. of solution = 10.3. Theory 10.0. Error —3 per cent. The amount of precipitate obtained and the development of the color are in direct proportion to the amount of P present in the sample in quantities varying from 0.0125 to 0.2 mg. of P. This is shown in Table I. Trichloroacetic acid in the concentrations used does not inter- fere with the precipitation and determination of the phosphorus present. The presence of Na, K, Ca, and Mg in concentrations compar- able to those found in serum and also the following organic com- a ov F. F. Tisdall 333 pounds do not interfere with the precipitation and determination of the phosphorus: Dextrose 300 mg. per 100 cc., urea 300 mg. TABLE II. Recovery of Phosphorus Added to Serum. Boma. | Teme? | Padded. | mfamd | epost. | =e mg. mg. mg. mg. per cent 10 0.078 0.0125 0.088 0.090 —2.2 10 0.078 0.025 0.102 0.103 —1.0 ' 10 0.078 0.050 0.127 0.128 —0.8 21 0.050 0.0125 0.062 0.062 +0.0 21 0.050 0.025 0.073 0.075 —2.6 21 0.050 0.050 0.105 0.100 +5.0 TABLE III. Inorganic Phosphorus Content of Normal Adult Serum (Twenty-Two Consecutive Determinations). Serum. Inorganic P per 100 ce. Serum. Inorganic P per 100 ce. mg. mg. 30 3.7 41 3.2 31 4.0 42 3.9 32 3.7 43 3.9 33 3.5 44 Bids 34 3.7 45 3.8 35 3.7 46 Sore 36 3.6 47 4.3 37 3.5 48 4.0 38 4.2 49 3.6 39 3.7 50 3.7 40 4.0 51 3.8 52 3.7 per 100 cc., uric acid 10 mg. per 100 cc., creatinine 20 mg. per 100 ce. (present as creatinine zinc chloride), creatine 5 mg. per 100 ce., and acetoacetic acid 100 mg. per 100 ce. The results given in Table II indicate that known amounts of phosphorus added to serum, the inorganic phosphorus content of which has been previously determined, may be quantitatively recovered. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 334 Inorganic Phosphorus in Serum Table III gives the results of the determination of the inorganic phosphorus in the sera of normal adults. The ages ranged from 20 to 35 years. The blood was removed between 9 and 11 a.m. andtheserum separated within13hours. The determinations were performed within 36 hours of the time the blood was obtained. Table IV gives the results of the determination of the inorganic phosphorus in sera of infants who showed no clinical evidences of rickets. TABLE IV. Inorganic Phosphorus Content of Normal Infant Serum. Serum. Age. Inorganic P per 100 ce. mos, mg. 10 1174 ae 19 13 4.6 20 14 5.8 21 16 5.0 22 10 5.3 53 f ed 55 36 6.4 URONIC ae RE ASORD OS TODD GOONC OURO EE OOEO OOD © 5.4 DISCUSSION. The use of a solution of strychnine molybdate for the precipita- tion of phosphorus was first used by Pouget and Chouchak (1). The reagent as devised by these authors had to be used immedi- ately after its preparation. Subsequently, Kober and Egerer (2) modified and improved the reagent so that it ““was stable and gave quantitative and constant results.” Other modifications have been made by Bloor (3), Medinger (4), Kleinmann (5), and Embden (6). The reduction of molybdie acid with the production of a blue color has been known for many years. Some time ago, Taylor and Miller (7) devised a method for the determination of small amounts of phosphorus which depended on the precipitation of ammonium phosphomolybdate with the subsequent determina- tion of the amount of molybdenum present by its reduction with phenylhydrazine. Quite recently a method has been devised by Bell and Doisy (8) which depends on the selective reduction of F. F. Tisdall 339 the molybdenum present as phosphomolybdic acid, in an excess of molybdic acid, with hydroquinone as the reducing agent. The present method depends on the precipitation of the phosphorus as strychnine phosphomolybdate, with the strychnine molybdate reagent as devised by Embden, and the reduction of the molyb- denum present in the precipitate by means of potassium ferro- cyanide and HCl. The precipitation of the phosphorus takes place very rapidly. 5 minutes after the addition of the strychnine molybdate reagent over 95 per cent of the inorganic phosphorus present is precipi- tated, and in 10 minutes the precipitation is complete, as no increase is obtained if the mixture is allowed to stand for } hour. The precipitate is slightly soluble in water, and if too much water is used for washing, erroneous results will be obtained. When the quantity of water indicated in this method is used it will be found that the amount of precipitate in solution in the second washing is so slight that it produces no appreciable error. Also the amount of the reagent left after the second washing is so small that if it is placed in a 100 ce. flask and potassium ferrocyanide and HCl added no perceptible green color can be observed. It is not necessary for the precipitate to be mixed with the added water to produce thorough washing. The green color produced by the reduction of the molybdate i is very stable and no special precautions are needed when it is being read in the colorimeter. If the solutions are allowed to stand a long time it will be found that the color tends to become more intense and of a bluish tint, this change taking place more rapidly in the weak solutions than in the strong ones. If a weak solution is read at 40 in the colorimeter with the standard at 20, in 3 hours time it will change to about 39 producing an error of +2.5 per cent, while a strong solution will change over night from 10.5 to about 11.0 producing an error of —5 per cent. No error will be obtained from this source if the reading is made during the first hour after the addition of the potassium ferrocyanide and HCl. The amount of potassium ferrocyanide added may be varied slightly without effecting the result. The amount added, how- ever, should be between 19 and 21 ec. The HCl should be added from a burette as a variation in the amount added effects the rapidity with which the color develops. The amount added 336 _ Jnorganie Phosphorus in Serum should be between 9.8 and 10.2 cc. It is also necessary that the volume of fluid in the known and unknown during the develop- ment of the color should be approximately equal. It is of interest to observe the concentration of the inorganic phosphorus in normal human serum, as determined by various other methods. Greenwald (9) found the “acid-soluble” phos- phorus to vary from 2 to 6 mg. of P per 100 cc. Marriott and Haessler (10) found from 1 to 3.5 mg. of inorganic phosphorus per 100 ce. and Bloor (11) 1.9 to 3.8 mg. of inorganic phosphorus per 100 ec. of normal adult serum. Feigl (12) found the “acid- soluble”? phosphorus to be under 4 mg. per 100 cc. Two determi- nations of serum from normal adults are recorded by Bell and Doisy (8) who found 3.5 and 3.9 mg. of inorganic phosphorus per 100 cc. Howland and Kramer (18) report an average of 2.1 mg. of inorganic phosphorus per 100 cc. in the serum of normal adults and an average of 5.4 mg. in the serum of twelve non-rachitic infants. The presence of a high inorganic phosphorus content in ‘the serum of non-rachitic infants reported in the present paper is in agreement with the high content, first recorded by these authors. The inorganic phosphorus content of normal adult serum as determined by the present method is slightly higher than recorded by most observers. Since it is possible that some of the ‘‘unknown phosphoric acid combinations” as well as the inorganic phosphorus might be extracted with the trichloroacetic acid, duplicate deter- minations were performed using trichloroacetic acid and acid ammonium sulfate as the extracting agents. Identical results were obtained. The possible extraction of some of the ‘‘unknown phosphoric acid combinations” by the strongly acid strychnine molybdate reagent was also considered. The supernatant fluid after its separation from the precipitate was allowed to stand 24 to48 hours. ie rs ‘ “ = : av AP Al. Js q ars ‘oSinGn bi aN Jy ‘ oly et Ci ae : aly ‘ t Ag Wy Ca we . (ito Har utilis: 1) ‘ > é : : ks Penh ape tee cotdiiiint [io bonita {. . mow Pha at aon arte ) en é 7 uf . i 7 te eed) pee as 1 fe- Eat WiO | edie ts nr ft fletponaes SL oe ond Terr, i ve (7 phe yt a.) ty a ¢ . : : 7 s ; P , r ios ay 7 VITAMINE STUDIES. IX. THE INFLUENCE OF THE DIET OF THE COW UPON THE QUANTITY OF VITAMINES A AND B IN THE MILK.* By CORNELIA KENNEDY anv R. A. DUTCHER. (From the Division of Agricultural Biochemistry, University of Minnesota, ; St. Paul.) WITH THE CooPERATION oF C. H. EcKLEs. (From the Division of Dairy Husbandry, Minnesota Agricultural Experiment Station, St. Paul.) (Received for publication, November 15, 1921.) The suggestion that Funk! offered in 1913 that there was prob- ably a definite relationship between the amount of vitamine secreted in the milk and that ingested in the food has become a well established fact. Since then many investigators have referred to this relationship, but it is only recently that any have carried out experiments with laboratory animals to prove it definitely. Without attempting to review completely the work done in this field attention may be called to a few important investigations. In 1916, McCollum, Simmonds, and Pitz? stated that the two factors, vitamines A and B, pass into the milk only as they are present in the diet of the mother. These investigators observed the ability of the female rat to rear its young when confined to rations known to be adequate for growth when sufficient amounts of vitamines A and B are added, and inadequate for growth when either of these factors is omitted. They had no means of deter- mining whether the omission of either vitamine A or B affected the amount of milk secreted by the rat and, therefore, whether the young received enough milk to permit normal growth. In * Published with the approval of the Director as Paper No. 285, Journal Series, Minnesota Agricultural Experiment Station. 1Funk, C., Biochem. J., 1913, vii, 211. 2McCollum, E. V., Simmonds, N., and Pitz, W., J. Biol. Chem., 1916, Svil oo. 339 340 Vitamines A and B in Milk 1918, Steenbock, Boutwell, and Kent* noting the variations in the amount of vitamine present in butter fat surmised that per- haps the rations on which the butter fat had been produced might be at fault. However, they further state that the ration is not the only factor to be considered as the butter fat produced by a cow fed exclusively on alfalfa hay was found to contain no demon- strable amounts of vitamine. They do not comment upon the fact that the butter had been kept in an unsalted condition in a poorly iced refrigerator for about 3 weeks and that mold had developed upon the surface; nor do they make any statement in regard to the quality of the alfalfa used. In 1920, Hess and Unger* suggested that an insufficiency of vitamine A in milk might be due to the winter fodder of the cow. In 1919, Barnes and Hume’ and Dutcher, Pierson, and Biester® announced simultaneously that they had noticed seasonal varia- tions in the antiscorbutic properties of cow’s milk. Experimental proof of this variation was later published by Hart, Steenbock, and Ellis,? Dutcher and coworkers,® and Hess, Unger, and Supplee.® Inasmuch as each of these investigators used a somewhat different method for studying this variation, the conclusion seems justified that the vitamine C content of cow’s milk can be influenced materially by the food of the cow. However, Hughes, Fitch, and Cave!’ have recently reported that the milks of ‘cows fed a vita- mine-rich ration and a vitamine-poor ration showed no difference as to vitamine C, but that both vitamines A and B weresdecidedly greater in the milk from a vitamine-rich ration. The same was true of the butter made from the milk produced on these two rations. 3 Steenbock, H., Boutwell, P. W., and Kent, H. E., J. Biol. Chem., 1918, xxxv, 517. ‘Hess, A. F., and Unger, L. J., J. Am. Med. Assn., 1920, Ixxiv, 217. 5 Barnes, R. E., and Hume, E. M., Biochem. J., 1919, xiii, 306. 6 Dutcher, R. A., Pierson, E. M., and Biester, A., Science, 1919, 1, 184. 7 Hart, E. B., Steenbock, H., and Ellis, N. R., J. Biol. Chem., 1920, xlii, 383. 8 Dutcher, R. A., Eckles, C. H., Dahle, C. D., Mead, S. W., and Schaefer, O.G., J. Biol. Chem., 1920-21, xlv, 119. * Hess, A. F., Unger, L. J., and Supplee, G. C., J. Biol. Chem., 1920-21, xlv, 229. 1° Hughes, J. 8., Fitch, J. B., and Cave, H. W., J. Biol. Chem., 1921, xlvi; p.1. C. Kennedy and R. A. Dutcher 341 Attention has already been called to the work carried on in this laboratory in regard to the variation in the vitamine C con- tent of milk due to the food of the cow. The experiments which were being carried out at that time offered an excellent opportunity for the study of vitamines A and B of milk produced under favor- able and unfavorable conditions. Up to this time only vitamine B, as it existed in the milk, had been studied, all investigations of vitamine A being made on the butter which had been separated _ from the milk. The work of Hopkins and Osborne and Mendel, the most notable experiments in which milk is used as a source of vitamine B, will be discussed later on with the experiments which embody this paper. EXPERIMENTAL. The experiments described in this paper are the first, in so far as we are aware, in which both vitamine A and vitamine B of cow’s milk have been quantitatively studied with respect to the changes which may occur in the milk due to a change from pasture to dry feed. In order to study such changes the amount of milk fed must necessarily be the minimum which would produce satisfactory growth. To find this minimum, experiments were begun in December, 1919, with the mixed milk from a Holstein and a Jersey cow which were on a vitamine-poor ration and whose milk was being used in another experiment.? Although at that time the actual details in carrying out the experiment differed somewhat from those which were finally adopted the results of these earlier experiments check very satisfactorily with those reported in this paper. The plan first followed was that used by many investigators; 1.¢€., to start with a small amount of milk to furnish the vitamine which had been omitted from the ration and gradually to increase the milk until an amount was being given which just gave satis- factory results. 5 cc. were arbitrarily chosen as the minimum amount. This method of procedure was not suited for our work as the rats lost in weight very rapidly on the low vitamine ration and their condition became so precarious that they did not respond to small additions of the milk which was probably becoming poorer in vitamines each day. From these results it was decided that —————— 342 Vitamines A and B in Milk 10 cc. daily were the minimum amount of milk which would furnish adequate amounts of vitamines A and B, but in order to be certain of the outcome duplicate groups of rats were started which received 15 cc. of milk daily. The milk used in these experiments was chosen to represent two types of milk: that produced by cows fed a ration presumably adequate in all respects, and that produced by cows fed a ration known to be deficient in vitamines but adequate in protein, mineral matter, and energy value. The ration fed the former group, which was the Station herd and includes Holstein, Jersey, Guern- sey, and Ayrshire cows, was made up as follows: Ofc: pe eee re res RANT Ren qian Rye ME Ot cms. 2 - 200 BRATS Scihel ok RSS ECan aie Gubieies Wiens 200 OBES ae sig eacs Sealed es cpaledeaea ag OE Oe Be siki s 0s esc 200 Oilimet lions: cone eC OES Gee Lee CeCe 140 Roughage was fed in the form of alfalfa hay and corn silage. During the summer months when the cows were on pasture this ration was continued but fed in smaller amounts than in the winter months. The ration which was fed to the latter group, which included two Holstein cows, consisted of equal parts of ground oats, ground barley, wheat middlings, and gluten feed with timothy hay and oat straw for roughage. This roughage was used for the first 2 months of the experiment, when the timothy hay was taken out as it was thought that it might be furnishing considerable amounts of vitamines to the ration. This ration is © undoubtedly poor in vitamine A and furnishes inadequate amounts of vitamine B, and is a poorer ration than is fed during the winter months on the dairy farms producing the milk for large cities. During the period this ration was fed a fair flow of milk was main- tained. The butter fat of the milk averaged 3.4 per cent which increased to 3.8 per cent after the cows had been on pasture for 3 weeks. These cows were placed on the vitamine-poor ration October 1, 1920, and continued on it until April 28, 1921, when the ration was changed to that of the Station herd. On May 16, 1921, the cows were turned out to pasture, the dry feed, however, being continued. The cows were milked morning and evening and samples of the mixed milk were taken for the feeding experiments. C. Kennedy and R. A. Dutcher 343 The time for starting the rats on their experimental rations was so chosen that the milk used would be representative of a milk produced in midsummer when all conditions of feeding and pasture were good; in fall when the pasture was beginning to fail; in mid- winter when the cows had to depend entirely on grain mixtures, silage, and hays; and in late winter and early spring when the effect of the change from winter feed to summer feed could be noted. The rats used in this experiment were healthy, normal rats selected from our breeding colony. As far as possible five or six rats weighing from 65 to 75 gm. each were selected for each group. These were kept in separate cages so that a record could be kept of their food intake. The usual laboratory precautions were taken in regard to the sanitary conditions of the cages, drinking bottles, and feeding dishes. The rats themselves were kept free of animal parasites by the application of pine oil.!! The milk in 10 or 15 ee. portions, as the conditions of the experi- ment called for, was given to each rat the first thing in the morning and after the first few weeks of the experiment had passed the milk was consumed within a few hours, at least before it became sour. A little difficulty was experienced at first in getting the young rats to drink all of their milk but this difficulty did not extend over a sufficiently long period to affect the results of the experiment. The basal ration used in our experiments was as free from vitamines as it is possible to make such a ration under the present conditions of our knowledge as to the nature of vitamines. The casein, which would be the most objectionable constituent in a vitamine-free ration because of adsorbed impurities, was made according to a method which is used in this laboratory and which we believe gives a very pure product.2 The preparation of vitamine B might carry small amounts of vitamine A. However, inasmuch as the wheat embryo, used as a source of vitamine B, was first thoroughly extracted with ether, the amount of this vitamine in the subsequent alcoholic extract would doubtless be very small as it has been reported" that ether removes vitamine 11 Kennedy, C., Science, 1921, liii, 364. 12 Palmer, L.S., and Kennedy, C., J. Biol. Chem., 1921, xlvi, 571. 183 Osborne, T. B., and Mendel, L. B., J. Biol. Chem., 1920, xli, 549. Steenbock, H., and Boutwell, P. W., J. Biol. Chem., 1920, xlii, 131. 344 Vitamines A and B in Milk A from plant tissue. Butter, melted at a low temperature and filtered, was used as the source of vitamine A. The salt mixture used in these rations was that given by McCollum." The dextrin was prepared by autoclaving commercial starch, acidified with 0.2 per cent citric acid, at a pressure of 15 pounds for 6 hours. This was then dried and ground. The basal ration used in this work was composed of purified casein 18 parts, salt mixture 3.7, agar 2, and dextrin to make up the 100 parts. When the milk used was to furnish the vitamine A the dextrin of the basal ration carried the alcoholic extract of 15 gm. of wheat embryo per 100 gm. of ration, and when the milk was to furnish the vitamine B, the basal ration carried 5 per cent of filtered butter fat. When the milk furnished both vitamines the basal ration with no additions was used. Group I, Charts I and II.—This group comprised the rats fed summer milk; that is, the milk produced by cows which were feeding in the pasture. These cows were not in the Station herd but were the two, a Jersey and a Holstein, which had been used during the previous winter in the preliminary experimental work having received the vitamine-poor ration given above. The cows had been turned into pasture and, in addition, given the herd ration on May 27, 1920, and their milk was not used for experi- mental feeding until July 1, 1920. The condition of the pasture was excellent at the time and continued to be so until the early part of August when the weather became very hot and dry for a period of approximately 3 or 4 weeks. Rains began again in September and restored the pastures which remained in good condition through October. The first hard frost came November 2,1920. It was thought, therefore, that the milk was representa- tive of good quality summer milk. The drying of the pasture during August influenced the vitamine content of the milk, especially the vitamine B, but this effect was not apparent until 3 or 4 weeks later (see Chart I, Curves 180 to 183). This would indicate that the cow can secrete vitamines A and B into her milk for some time after the vitamine supply in her feed has been di- minished, the vitamine content of her milk being kept up at the expense of her tissues and that when vitamines are again supplied 16 McCollum, E. V., Simmonds, N., and Pitz, W., J. Biol. Chem., 1916, xxv, 123. C. Kennedy and R. A. Dutcher 345 by her food they go first to her tissue and then to her milk. This postulate is further supported by the fact that the experimental animals do not respond immediately to the milk when the cows are changed abruptly from a vitamine-poor ration to one which is relatively rich. No butter fat or aaa blic extract embryo. 15icc. sfmmer milk. - Z| 28 A days 4, gv tter fat + 15 cc.|summoer milk Alcoholic pxtracit of 15 gramp wheat emb 15 cb. sum er milk. ae 2840402 ae A040 a0 Aa Coe sogeyensseavonae Ee a CuartI. Rats 140 to 151 inclusive, and Rats 180 to 187 inclusive were placed on the experimental ration July 1, 1920. The milk for these rats was supplied by the two cows which had been on a vitamine-poor ration from Jan. 17 to May 27, 1920, when they were turned out to pasture and, in addition, given the herd ration. Inasmuch as 4 weeks intervened between their going into pasture and the time when their milk was used for the experiment it was considered that the milk was representative of good quality summer milk. The effect of the drying of the pasture during August is apparent in the growth curves a few weeks later and is especially noticeable in the curves of the rats which were receiving their supply of vitamine B from the milk. This does not, however, demonstrate that vitamine B is diminished to a greater extent in the milk than vitamine A as the quantitative need of the rat for both vitamines has not yet been determined. Rats 148 to 151 inclusive which received both vitamines A and B from 15 ce. of this milk made normal growth which shows that the amount of each vitamine contained in 15 cc. of this milk is sufficient for normal growth. 346 Vitamines A and B in Milk The rats which were fed the milk produced in the fall months are included in this group due to the fact that for the first 2 months of the feeding period the cows were still on pasture. The fall milk was from the Station herd. There is no outstanding difference in the fall milk curves, Chart IJ, and the summer milk curves. Chart II, Fall }ilk. epee rmys | | | tT | TT tT | CS A A ee A ya ee | PPT ALP yt 4) oc PS a a PEC EEC CREPRERL CTL eet reset PA | | pA eH EFA Via SEG ame 2 Very Veo hu bee | | DAA Wadd ed | ian SRaGE. =P) | BES Pore Sholcl lok ale Se Poms aver a cobae AY a Cia Coe Zoe Cuart II. Rats 192 to 201 inclusive, Rats 254 to 262 inclusive, and Rats 211 to 214 inclusive were placed on the experimental ration Sept. 1, 1920. The milk used for these rats was supplied by the Experiment Station herd. During September and October the pastures were green and the grass abundant. The first hard frost came Nov. 2 when the cows were removed from the pasture. The curves for Rats 198 to 201 inclusive show that the amount of vitamine B in 10 ce. of this milk is inadequate to support normal growth after 6 to8 weeks. Comparison of these curves with those of Rats 304 to 307 inclusive (Chart VI) shows that the failure to make normal growth is not due to the fact that 10 cc. of milk will not furnish enough vitamine B but that 10 ce. of milk from cows feeding on poor pasture is not adequate. 190 150 ee C. Kennedy and R. A. Dutcher 347 The summer milk curves fall away from the normal from the effects of drying of pasture, and the fall milk curves start to fall off when the pasture season ends. Group II, Charts III, IV, and V.—This group includes the rats fed the winter milk produced by the two Holstein cows. These cows had been placed on the vitamine-poor ration in October, 1920, and the feeding of this milk to the rats began in December. This afforded 2 months in which the cows could get rid of the vitamines in their tissues. However, it could not be determined whether the milk was becoming poorer each day in vitamines and would finally become so depleted that it would no longer cause Chart III, Winter Eq ie ct of bryo 10 co. erin reise} eee | A A “ae geEECEEaEceeeaee pasa eect Cuart III. Rats 275 to 284 inclusive were placed on the experimental ration Nov. 30, 1920. The milk used for these rats was supplied by the two Holstein cows which had been on vitamine-poor rations since the first of October. 10 ce. of this milk do not furnish adequate amounts of either vitamine A or vitamine B. Rats 275 to 279 inclusive became very thin and nervous during the course of the experiment and two rats, Nos. 275 and 278, died of lung infection near the end of the experiment. Although Rats 280 to 284 inclusive suffered severely in their growth from a lack of vitamine A, none showed signs of eye trouble indicative of a deficiency of this vita- mine. May 1, 1921, the ration of the two experimental cows was changed to the herd ration and on May 16 the cows were turned out to pasture. This point is indicated by an asterisk on each curve. 348 Vitamines A and B in Milk Chart IV. butten fat gr alcoholio extrac uryo. |15 cc.| experimental milk 7 7 7 ae : ad i a a i™ \ | S@E GH 150 6 aay aa ie SUMP Agree ACr ie ea : an a 230 ae -- 1 : Ca] 110 : 70 | ; | 230 : 190 | a, PT A eS a CuartIV. Rats 285 to 299 inclusive were placed on the experimental ration at the same time as the rats in Chart III; the only difference in their ration being that they received 15 cc. of milk instead of 10 ec. The increased amount of milk had a decidedly beneficial effect on the rats’ growth curves. Although the increase in weight was normal or better than normal for a considerably longer period there was quite a long period when the rats were only able to maintain their weights. A comparison of this chart with Chart VI shows the great superiority of the quality of vitamine A contained in 10 ee. of winter milk from cows fed on an adequate ration over that contained in 15 cc. of milk from cows fed a ration known to be deficient in vitamines. The curves for Rats 285 to 289 inclusive show a greater divergence from the normal than do the other curves, but this is to be expected as there are two deficient factors in the ration of these rats while in the ration of the other rats there is but one deficient factor. Rats 291 and 294 died from an unknown cause. The cows were turned out to pasture at the point marked by the asterisk. C. Kennedy and R. A. Dutcher 349 growth, or whether it was as deficient at this time as it ¢ver would be, but still contained enough vitamine to cause some growth. 10 ce. of this milk do not furnish adequate amounts of either vitamines A or B after the first few weeks of feeding. The weights of the rats which received vitamine A and those which received vitamine B from 10 cc. of this milk began to fall away from the normal at about the same time although there is a slight extension of time in the case of the rats which received vitamine A from the Chart V. Late Winter and Karly —— Experimental Milk. exp Fimenthl mil pholid t of fl a Cc. @ mtal = A wy She 5 7 ave Sie ara EEE? SY ACN a es H i s é 5 rea i perc bath ns s e Cara | 2B 2ZeZERl. 602044 Do A a Ce ; CuHart V. Rats 331 to 346 inclusive were placed on the experimental ration Feb. 1, 1921. This group was started at this time to ascertain if the milk of the cows became progressively poorer in vitamines A and B as the length of time of feeding a vitamine-poor ration increased. Comparing this chart with Chart III, it is seen, in each group, that the rats stopped gaining weight at approximately 6 to 8 weeks after the beginning of the experiment. The curves in this chart (Chart V) appear better than those of Chart III because the rats were not on the experimental ration for as long a period, and, therefore, do not become as depleted. They also show a greater improvement after the cow’s ration improves. These curves indi- cate, though they do not demonstrate, that if the vitamines are stored in the body of the cow it is only for a short time. The curves of the group receiving 15 cc. are only slightly better than those receiving 10 cc. Rat 344 died from an unknown cause. co Ne a SJ 7 ie 350 Vitamines A and B in Milk milk. It is interesting to note that the male rats withstood the vitamine deficiency much better than the female rats. During the periods of the experiment although the rats which received vitamine A from 10 ce. of milk became thin, they had no xerophthalmia. Further proof of the paucity of vitamine A in this milk is shown in Chart [X. The rats whose curves appear in this chart obtained their supply of vitamine A from butter that had been made from cream separated from this milk, during the last 4 weeks of the experiment, when the milk was undoubtedly very poor in vita- mines. The butter was placed in small containers under an atmos- phere of CO, and stored in an ordinary refrigerator, the tempera- ture of which was approximately 10°C. Two groups of rats were fed purified rations carrying in one case 5 per cent of this butter / rd rai amlbaas Cuart IX. Rats 400 to 408 inclusive were placed on the experimental ration in the winter of 1921. The basal ration carried the alcoholic extract of 15 gm. of wheat embryo and in the case of Rats 400 to 404 inclusive, 5 per cent of butter fat made from the vitamine-poor milk, and in the case of Rats 405 to 408 inclusive, 9 per cent of the same butter fat. The asterisk indicates the point at which the fat in the ration of Rats 400 to 404 inclusive was increased to 15 per cent, and in the ration of Rats 405 to 408 inclusive to 20 per cent. That the rats did not improve very materially in this last group seemed to be due to the fact that the rats did not eat well of the ration because of its excessively greasy nature. Improvement in the former group took place on increase of the fat from 5 to 15 per cent but the curves never coincided with the normal. In all of our experimental rations 5 per cent of good butter fat has carried sufficient amounts of vitamine A for normal] growth. st Var rere ee Eye C. Kennedy and R. A. Dutcher 351 fat and in the other case 9 per cent. When the rats had ceased to grow the butter fat in each ration was increased to 15 and 20 per cent respectively. This increase gave only slightly better results. We believe that this deficiency is due to no other cause than to a lack of vitamine A in the ration of the cows as extreme care was exercised both in the separation and churning of the cream and in the subsequent storage of the butter. The rats receiving their vitamine B from 10 ce. of milk became very nervous and timid and two, Rats 275 and 278, died of lung infections. In other groups where there was a deficiency of this vitamine the same lung infection occurred. None of the rats in this group showed a quick or marked improvement in their weights after the cows had gone to pasture. Whether these rats were in too depleted a condition to improve after the cows had gone to . pasture, or whether the cows had become so depleted in vitamines A and B that 6 weeks was not a long enough period to show an improvement in the vitamine content of their milk, could not be definitely determined, although the rats of Chart V which had received the experimental milk for a shorter period and were in better physical condition showed a more decided upward trend in their curves. Perhaps the most striking inference which can be drawn from a comparison of Charts III and IV with Chart V is in connection with the observation that the rats on comparable rations grew to about the same weight before the vitamine defi- ciency became evident. As pointed out previously, there is no evidence to indicate that the milk fed to the rats of Chart V was any poorer in vitamine at the beginning of the experiment than that received by the rats in Charts III and IV, although the experiments labeled ‘‘ Winter Experimental Milk” began 2 months before those designated ‘‘ Late Winter and Early Spring Experi- mental Milk.’ As a matter of fact, the indications are rather that the quantitative deficiencies of the milk with respect to the vitamines were no greater in the late winter than several months earlier. The inference which we can draw from these data is either that the actual weight of vitamines required for growing rats of 70 to 150 gm. weight, a period of very active growth, is less than in larger rats, or that a certain period has been reached in the life cycle of the rat when the demand for vitamine is greatest. That this period is that of adolescence is indicated by the fact that a2 Vitamines A and B in Milk the rats in the experiment began to show signs of vitamine defi- ciency at the age when this species of animal attains sexual maturity. The rats which received 15 ec. of this same winter milk main- tained a normal growth curve for a considerably longer period than those which received 10 cc. but there was a long period when the rats were only able to maintain their weights. Group III, Charts VI and VII.—The milk used for this group was the winter milk from the Station herd. The rats receiving 10 cc. of milk daily made remarkably good growth; in fact, it equalled that of those receiving 15 cc. This is of interest in view of the fact that Osborne and Mendel" have reported that 16 ce. of whole summer milk were necessary for satisfactory growth with a ration that was supplying the vitamine A. A comparison of Chart VI with Chart IV shows the great superiority of the quality . of vitamine A contained in 10 cc. of the Station herd winter milk over that contained in 15 cc. of the winter milk from the two experimental cows. 15 cc. of the latter milk contained 0.526 gm. of butter fat and were much less efficient for growth than 10 ce. of the former milk which contained only 0.392 gm. of butter fat. From the excellent growth of these rats we can conclude that a winter ration for cows may be made perfectly satisfactory as far as vitamines A and B of the milk are concerned by the proper combination of grain and leafy foods. It will be noted that throughout the experiment there were only a few of the rats that reproduced. While it would have been desirable to have had reproduction, it was not the aim of the experi- ment to find a quantity of milk upon which perfect growth, as measured by the ability of the rat to reproduce at normal intervals would be attained, but rather to ascertain the effect of the cows fed on the quality and quantity of the vitamine produced in her milk. In order to obtain quantitative results in feeding it was, under the laboratory conditions at the time of the experiment, almost impossible to allow the rats to be together for more than a very short time each day and it is very difficult to obtain norma] reproduction under such conditions. There were five females, however, which gave birth to young and reared them successfully, 16 Osborne, T. B., and Mendel, L. B., J. Biol. Chem., 1918, xxxiv, 537. C. Kennedy and R. A. Dutcher 353 Attention may be called to the work of Hopkins" on the stimu- lating effect on growth of 2 cc. of cow’s milk as an addendum to an artificial ration consisting of casein, fats, starch, sugar, and inorganic salts; and to his later work in which he substantiates this result. Because of these results it seemed necessary, after some of the preliminary experiments of this investigation had been com- pleted and it was indicated that young rats could not grow on Chart VI. Winter Station Herd Milk. goholi¢ ot of a ryo + ce. Sta berd & Cuart VI. Rats 304 to 315 inclusive were started on the experimental ration Jan. 12, 1921. The milk fed to this group was furnished by the same herd that furnished the milk for the rats in Chart II. The rats receiving 10 cc. made remarkably good growth, in fact they equalled the growth of those receiving 15 cc. This fact demonstrates the value of an adequate combination of grain and leafy foods during the winter season. Rat 309 died of pneumonia. The time that the cows were turned out to pasture is indicated on the chart by an asterisk. 5 ec. of milk produced on a vitamine-poor ration, to further sub- stantiate this result by repeating the experiment but to use a summer milk known to produce satisfactory growth when fed at 16 Hopkins, F. G., J. Physiol., 1912, xliv, 425. 354 Vitamines A and B in Milk a higher level. Therefore, two groups of rats were used: one group received, in addition to a ration which was complete in all respects except for its vitamine A, 5 cc. of milk to supply this vitamine; and a second group received, in addition to a ration which was complete in all respects except for its vitamine B, 5 ce. of the milk to supply this vitamine. The growth curves for these rats found in Chart VIII demonstrate very conclusively that 5 ce. . Winter Station H a0 but te or Pr alcpholic| extract of as ets Whpat emp 15 oof Statjim herd mil ari Nee war avars) Fa ae aE I es ie ee | eee Ape f| TL ge fo, a ay laa) 150 Fd wha Cuart VII. Rats 316 to 330 inclusive were started on the experimental ration Jan. 12, 1921, and received, daily, 15 cc. of milk from the Experiment Station herd. These rats Byidenthy received sufficient vitamines from the start so that when the cows went into pasture any increase in the amount in their milk had little or no influence on the growth of the rats. That 15 ec. is ample to supply both vitamines A and B is shown by the curves for Rats 316 to 320 inclusive. Three rats, Nos. 323, 325, and 327, died from an unknown cause. The cows were turned out to pasture at the point marked by an asterisk. C. Kennedy and R. A. Dutcher 355 of good quality summer milk will not promote normal growth in young rats. It is possible that the ration of the cows from which Hopkins obtained the milk he used was much richer in vitamines than that of the cows furnishing the milk for our experiment. That it is entirely a question of the ration of the cows seems to be clearly shown by the various results published by the different investigators. Hopkins has described carefully planned experi- ments in which he obtained satisfactory growth on a remarkably small amount of milk added to a purified ration. Osborne and Mendel! have described equally carefully planned experiments in which they state that ‘‘not until at least 16 cc. of fresh milk per Summer Station Herd Milk. Cuarr VIII. Rats 201 to 210 inclusive were placed on the experimental ration in the summer of 1920. The milk used for these rats was the same as that used for the rats whose growth curves appear in Chart I. The failure to grow after the first 4 weeks is the result of a deficiency in vitamines A and B, owing to the small amount of milk (5 cc.) used as a source of these vitamines. day were supplied along with the food mixture, was anything approaching a normal rate of growth secured. Even this amount sometimes failed.” Later these investigators repeated this work using summer milk because they thought that the inferior quality of the milk, as a source of vitamine B, might be due to the winter ration of the cows, their earlier work having been carried out dur- ing the winter season when the cows were deprived of green pasture and were stall fed. In addition to the pasture grass, the cows were fed night and morning a ration consisting of corn gluten and 356 Vitamines A and B in Milk wheat bran together with hay and corn-stalks. Failure to grow became evident in approximately 25 days when the rats received 10 cc. of this milk, unpasteurized, from the beginning of the experiment. 15 cc. of the same milk barely sufficed as a source of vitamine B. These results of Osborne and Mendel indicate that milk is a poor source of vitamine B. In our experience with milk we find that 10 and even 15 cc. of milk are inadequate to furnish either vitamines A or B when the milk is produced by cows which are feeding on a ration in which these vitamines are deficient, but that 10 ee. of milk are amply sufficient to furnish either vitamine pro- vided the milk is produced by cows feeding on a ration which is adequate as to its vitamine content. It would, therefore, seem that milk becomes a good source of vitamine B when, in addition to feeding in pasture, the cow is given a grain and hay mixture rich in vitamines. Access to open pasture will not assure a ration rich in vitamines unless the pasture is always fresh and green. The feed in the pasture varies with the climatic conditions and in order to secure a milk uniform as to its vitamine content, it is necessary to give a good dry feed throughout the year. The unsatisfactory results obtained by the use of 15 cc. of summer milk, which Osborne and Mendel report, may be due to the fact that the dry feed ration of their cows was vitamine-poor and that practically the only source of vitamine B was pasture grass, a variable source. That milk, as a source of vitamine B, may compare favorably with vegetable sources of this vitamine is shown in results which we have obtained in other work in this laboratory. We have found that while the alcoholic extract of 10 gm. of wheat embryo for each 100 gm. of ration will produce normal growth, this amount is not always dependable; therefore, it has been our custom, for some time past, to use the extract of 15 gm. of embryo. This extract, dried on 100 gm. of ration complete in all other respects, gives no better growth curves than 10 cc. of our Station herd winter milk when used as a source of vitamine B. This amount of milk adds to the ration of a rat 1.26 gm. of milk solids, which would not greatly enhance a complete ration unless the presence of vitamine C in the milk exerts a beneficial effect or that the milk contains some other food accessory which we have not as yet rec- me more aye | |e Pe solar pene |< | | ae PAC oo BOM CARA AWA ea Hi of ia de cleat ome eC CCP CCPC er Le en oe Porth iE : Ld dt TA 7 Wi Po yIVEE ole the [oP [ial ee a es 451 ec 483 ae Cuart X. The curves of this chart represent the average food intake per week per rat. Record 'the dietary intake of each rat was kept throughout the experimental period and the average of 3 records comprising each group was used in computing these graphs. A comparison of the curves dws that the food intake is approximately the same, with slight variations, whether the animal is }ywing or merely maintaining its weight, which fact does not agree with the generally accepted theory jit the food intake corresponds very closely with increase in growth. The food intake of the rats ose growth curves are found in Charts III, IV, and V was approximately the same during the % 6 to 8 weeks of the experiment, when growth was normal, as during the remaining period of the deriment when only maintenance of weight was accomplished. The actual food intake of the rats ose growth was excellent (Charts I, II, and VII) was no greater than that of the other groups harts III, IV, and V) which grew normally during only the first 6 to 8 weeks of the experiment d then merely maintained their weights. Moreover, the food intake for the rats of the former Joups (Charts I, II, and VII) was less per gm. of body weight during the latter part of the experi- }ntal period than that of the rats of the latter groups (Charts III, IV, and V) during the same jriod of theexperiment. Itwouldseem, therefore, that the effect of the vitamine is not necessarily je of appetite stimulation but rather a stimulation of metabolic processes which promote growth. The food intake for Rats 304 to 315 inclusive, Chart VI, does not parallel that of any of the other joups. 357 358 Vitamines A and B in Milk ognized. And again, baker’s yeast which has recently come into great prominence as a rich source of vitamine B, has proved under the conditions of our experimentation, to be a less valuable source of vitamine B than we formerly supposed. We have found that when the yeast is mixed in the ration in the proportion of 10 gm. of yeast in each 100 gm. of ration that growth was not as good as when 10 ce. of our Station herd milk were used to supply the same vitamine. When the yeast was fed separately from the food mix- ture 0.6 gm. per day gave the same results as when the yeast was mixed in the ration in the proportion of 10 gm. to 100 gm. of ration. Osborne and Mendel!’ report that 15 cc. of summer milk are inferior to 0.2 gm. of brewer’s yeast. We have found that the brewer’s yeast, which we have been able to obtain, is much inferior in growth-promoting properties to baker’s yeast. SUMMARY. Two types of milk, one produced on a ration typical of that used on some farms during the winter season and known to be deficient in its vitamine content, and a second representing that produced on a ration carrying ample amounts of vitamines A and B, have been used in this investigation. Each milk was fed so as to show in as nearly a quantitative manner as possible its content of vitamines A and B. Growth curves are given which show the possibility of growth on low and high levels of each milk. The importance of feeding the cow a ration adequate as to its vitamine content is demonstrated. CONCLUSIONS. 1. The presence of vitamines A and B in cow’s milk is entirely dependent upon their oécurrence in the ration. 2. Stall fed cows will produce a milk rich in vitamines pro- vided their ration consists of a proper combination of grains and leafy foods. 3. A vitamine-rich milk is not necessarily correlated with access to pasturage. 4. 10 cc. per day of either winter or summer milk is adequate to furnish either vitamine A or B to a rat provided the ration of the cow carries each in amounts adequate to meet her requirements. '7 Osborne, T. B., and Mendel, L. B., J. Biol. Chem.; 1920, xli, 515. rive C. Kennedy and R. A. Dutcher 359 5. 5 ec. of the same milk that produced normal growth when used on a higher level does not furnish enough of either vitamines A or B to meet the requirements of growing, rats. 6. The effect of the vitamine is not necessarily one of appetite stimulation but rather a stimulation of metabolic processes which promote growth. In conclusion we wish to acknowledge the assistance of Mr. John W. Wilbur, formerly with the Division of Dairy Husbandry, in keeping careful oversight of the experimental cows whose milk was used in this investigation and in making weekly butter fat determinations on the milk. Sia ish Sata lew f | STUDIES ON THE ACETONURIA PRODUCED BY DIETS CONTAINING LARGE AMOUNTS OF FAT.* By ROGER 8S. HUBBARD anp FLOYD R. WRIGHT. (From the Laboratories of The Clifton Springs Sanitarium, Clifton Springs, New York.) (Received for publication, December 3, 1921.) The excretion of the acetone bodies—acetone, acetoacetic acid, and 6-hydroxybutyrie acid—in conditions in which the organism is not utilizing carbohydrate either through a de- ficiency of foodstuff of this kind in the diet or through the inability of the organism to metabolize the food when supplied, as in dia- betes mellitus, has attracted attention for many years, and a large amount of literature has collected on the subject. In two papers recently published, Shaffer (1921, a,b) has summarized this literature, and has suggested certain methods of studying the problem which are somewhat different from those which have been used before. In this paper are reported some experiments which were carried out along the lines suggested, and which appear to support the theses advanced in these two articles. In his first paper Shaffer (1921, a) reported experiments on the oxidation of mixtures of acetoacetic acid and glucose by alkaline hydrogen peroxide which showed that if there were present in the mixture one molecule or more of glucose for each molecule of acetoacetic acid, the acid was oxidized under suitable conditions of temperature, alkalinity, etc., but that if the relative concen- tration of glucose was less than this, the oxidation of the keto- acid was not as complete. In the second paper (Shaffer, 1921, b) * A preliminary report of the clinical side of the work discussed was read before the meeting of the New York State Medical Association in Brooklyn, May, 1921, by Floyd R. Wright; a portion of the work formed part of a thesis presented for partial fulfilment of the requirement for the degree of Doctor of Philosophy at Washington University, St. Louis, in June, 1921, by Roger S. Hubbard. 361 362 Studies on Acetonuria he studied the problem from the point of view of the metabolism of human subjects, and concluded that a reaction of a similar nature takes place in the body. Since the appearance of these two articles Woodyatt (1921) has published a paper in which the subject is discussed from the standpoint of the practical treat- ment of diabetes, and in which data are presented which support the conclusions pied above. The theory which has been developed in these papers, and on which the following paper is based, is that acetoacetic acid itself is not easily burned in the body, but that it forms with glucose, or with degradation products of glucose and related substances, a compound which is easily burned. To compounds which give rise, in the progress of metabolism, to acetoacetic acid the name ‘“‘ketogenic”’ is given, while the name ‘‘antiketogenic’”’ has been applied to compounds which furnish glucose or other related compounds with which the acetone body combines. The keto- genic compounds contained in the diets are the fatty acids con- tained in the fats and the a-amino-acids, leucine, tyrosine, phenyl- alanine, and possibly histidine which forms a part of the proteins. There is probably a molecule of the acetone bodies derived from each molecule of these compounds contained in the diet. The amounts and source of the antiketogenic compounds con- tained in the diet are more uncertain. Glucose and related sugars, as levulose, form one source of these substances, whether taken as the sugars themselves or as the more complex carbohydrates. Protein yields glucose when fed to the total diabetic in amounts which vary with the different kinds of the foodstuff, and some percentage of the protein should therefore be included with the carbohydrate in figuring the total antiketogenic intake. There is, too, considerable data which indicate that glycerol yields glu- cose under some conditions, and so fat, from which glycerol is produced by hydrolysis in the organism must also be considered as a possible source of antiketogenic compounds. The question is even more complicated than this, because it is not certain what derivatives of the glucose-forming a-amino-acids and glycerol act as antiketogenic compounds. Shaffer (1921, b) has pointed out this difficulty clearly. He states?! 1 Shaffer (1921, b), p. 458. R. 8S. Hubbard and F. R. Wright 363 “ . 6... «the two carbon residues from glycocoll and the three carbon residues from the other sugar-forming amino-acids may have direct and immediate antiketogenic (ketolytic) action without condensation to glu- cose, and the same may be true of glycerol.” To determine the border-line diet which should just produce an excretion of the acetone bodies Shaffer (1921, b) calculated the molecular equivalents of the ketogenic compounds from fat and protein, and of the antiketogenic compounds from carbohydrate, protein, and the glycerol residue of the fat to the extent of the glucose which could be derived from them. From the analysis of the data so obtained he concluded that a diet containing 10 per cent of the calories in the form of protein, 10 per cent as car- bohydrate, and 80 per cent as fat represented approximately the border-line diet. He studied this diet in the light of data con- tained in the literature and obtained experimentally, and showed that the theory was confirmed by such results as were available. In the experiments reported in this paper an attempt was made to study this diet described by Shaffer, and diets in which the relative amounts of carbohydrate and fat were somewhat varied. To obtain a method of graphic representation of the various diets in terms of their ‘‘ketogenic balance” the following plan was adopted. The excess antiketogenic material derivable from pro- tein, that is, the amount of glucose which protein would yield greater than that needed to bring about oxidation of the keto- genic material from the same protein, was calculated from the data presented by Shaffer (1921, b). He showed from the analyses of the a-amino-acid content of ox muscle given by Lusk (1917)? and from the glucogenetic power of protein, that there are twice as many gram molecules of antiketogenic substance (glucose) as of ketogenic compounds which can be derived from a given weight of this protein. Glucose is derived from ox muscle at the rate of 58 gm. for each 100 gm. of the protein ingested (Woodyatt, 1921), and there will be 29 gm. of excess glucose for each 100 gm. of this protein fed. The amounts of glucose and of acetoacetic acid which can be derived from different proteins vary, and 25 gm. have been chosen as a convenient average figure to express the amount of glucose available for additional antiketogenic 2 Lusk (1917), p. 77. 364 Studies on Acetonuria action from 100 gm. of protein. This calculation is similar to that suggested by Woodyatt (1921). To the excess glucose from protein was added the glucose taken in the diet, and the sum was multiplied by 1.5 (molecular weight of glucose = 180; molecular weight of stearic acid = 284; of palmitic acid = 256; of oleic acid = 282; average = 270; = = 1.5) to convert the result into terms of its fatty acid molecular equivalent. This product was divided by the fatty acid content of the diet (95 per cent of the fat) and the ratio was multiplied by 100 to give the resulting expression in the form of per cent. This formula for expressing the “‘ketogenic balance” of any diet is expressed as follows: 1.5 (weight glucose + 25 per cent weight protein) ses 95 per cent weight fat In preparing the charts in this paper the total carbohydrate content of the diet has been used instead of the glucose content. Such a substitution introduces an error, as the intake of starch in grams should be multiplied by 1.1 to give the correct amount of glucose to which it is equivalent, but the difference between the values is almost certainly within the limit of error, and the total carbohydrate content is more easily calculated from published tables. Before proceeding to a study of the experiments, attention must be called to some of the limitations and advantages of the formula given above. In the first place it is based on an assumption which did not hold exactly for any of the diets studied. The formula assumes that all of the fat contained in the diets was fed in the form of glycerides of the higher fatty acids—palmitic, stearic, and oleic—and such a diet could not be fed for any con- siderable period. In one of the experiments an attempt was made to approximate such a composition for a few days, as will be de- scribed below, but for the most part butter and cream formed a large percentage of the fat ingested. These fats contain rela- tively large amounts (up to about 8 or 9 per cent) of their fatty acids in the form of compounds of comparatively low molecular weight, and, therefore, yield more ketogenic material per gram than do fats not derived from milk. In the formula a figure R. 8. Hubbard and F. R. Wright 365 lower than 1.5 should be used to convert antiketogenic compounds expressed as glucose into molecular equivalents of fatty acids when these fats are included in the diet. Butyric acid, for ex- ample, has a molecular weight of 88, and if tributyrin were the only fat fed the sum of the antiketogenic compounds expressed as glucose should be multiplied by 0.49 (44°55) to express the fraction in terms of relative molecular concentrations. However, this error will not change the numerical value of the expression by more than 5 per cent; the error introduced by figuring from the carbohydrate content of the diet was of the same order of magnitude, and the two should practically compensate for each other. The second objection to the formula has been indicated already. It is impossible to be sure that the figure used to express the ex- cess antiketogenic material from protein is correct. The value will vary for different proteins as their content of leucine, tyro- sine, and phenylalanine varies, and will also vary because the glucogenic power of different proteins is different. The effect of this uncertainty upon the numerical expression is illustrated by the following figures. Ina diet in which 10 per cent of the calories is fed as protein, 10 per cent as carbohydrate, and 80 per cent as fat the numeric¢al value of the expression given above is 55 per cent. The values of similar expressions in which different figures express the excess glucose from protein would vary from 31 per cent if its antiketogenic power is neglected, to 71 per cent if its ketogenic power is neglected, and the glucose which can be de- rived from protein is figured at 60 per cent of the total weight. If protein contains both ketogenic and antiketogenic materials— and this is almost certainly the case—the different figures lie well within the limits of error with which such formulas can be applied to the study of actual diets. In case the antiketogenic effect of protein depends, not on glucose, but on the two and three carbon atom residues derived from the sugar-forming a-amino-acids the values of the expres- sion would be much higher than 71 per cent. In that case an a-amino-acid would figure three times as efficient as glucose if a two carbon atom residue takes part in the reaction as an anti- ketogenic compound, or twice as efficient if a three carbon atom residue takes such a part. It seems almost certain that it would THE JOURNAL CF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 ee Dr 366 Studies on Acetonuria be possible to detect such a marked effect as this experimentally, and an attempt has been made to interpret the data presented below in such a way as to solve this problem. In the formula given no account has been taken of the possible antiketogenic effect of the glycerol radicle present in the fats. This radicle is probably the most uncertain of the different possible sources of antiketogenic compounds contained in the diet, and the part which it plays in the reaction can be better discussed after a study of the data obtained experi- mentally. If different diets are fed, and the degree of acetonuria is noted and compared with some such numerical expression of their ketogenic balance as that suggested, the value of the ratio at which the excretion of the acetone bodies becomes normal will represent the condition of ketogenic antiketogenic equilibrium. From the numerical value of this ratio it should be possible to determine whether the glycerol residue figures as a source of anti- ketogenic compounds. If this residue yields glucose in the organ- ism, and this glucose acts as an antiketogenic compound, enough of such material would be furnished in each gram of fat to combine with one-sixth of the ketogenic material, so that only five-sixths of the total fatty acids will be free to combine with the antiketo- genic compounds from carbohydrate and protein; in this case acetonuria should develop and clear up when the diet has a value of 83 per cent. If glycerol in fat does not produce antiketogenic compounds in the organism, all of the fatty acid will take part in the reaction with antiketogenic material from other foods, and the border-line diet will have a numerical value of 100 per cent. If glycerol figures as an antiketogenic compound in the form of a three carbon atom residue, one-third of the total fatty acid will combine with it, two-thirds of it must be burned by the help of other foodstuffs, and the border-line diet should have a value of 67 per cent. In the experiments reported here an attempt has been made to make such comparisons, and to determine whether the glycerol residue of fat does possess an antiketogenic action. The study of diets which produce border-line acetonuria and at the same time maintain the body weight of the subjects is rather difficult. The diets are markedly different from those generally eaten, and many patients return a portion of the fat untouched. There is also a temptation to break the dietary R. 8. Hubbard and F. R. Wright 367 restrictions not unlike the temptation to which diabetics are sub- ject, and it would not be necessary for a patient to ingest much more carbohydrate than is furnished to spoil an experiment. In the series presented here only such cases are included as could be studied under rather close supervision in a department devoted exclusively to the study of nutritional diseases. We wish here to express our thanks to Dr. 8. T. Nicholson, Jr., the director of this department, and the dietitians and nurses attached to it for their cooperation in our experimental work. The series included two experiments on a normal subject, one of which has been previously reported in another connection, and studies on four cases of arthritis who were undergoing the dietary treatment recommended by Pemberton (1917) in which the car- bohydrate intake is reduced. These cases can probably be con- sidered as normals for the purpose of such a study, although Pemberton and Foster (1920) have stated that such patients show a slightly increased concentration of sugar in the blood and an abnormal rise in blood sugar after the administration of large doses of glucose. In all of the experiments the attempt was made to furnish enough food to each patient to maintain the body weight un- changed. To accomplish this the basal metabolism was deter- mined with the Benedict portable respiration calorimeter (Bene- dict, 1918) and enough calories were fed to allow for the main- tenance of basal equilibrium and for the probable activity of the patient. Usually the food provided for a bed patient was so calculated as to furnish 20 to 25 per cent more calories than his basal requirement called for, and this was found to be satisfac- tory for most of the subjects. It is desirable that the patients should be in nitrogen as well as in metabolic equilibrium, and at the same time that the protein content of the diet should be kept low to diminish the uncertain factor of its part in the ketogenic expression. In the experiments reported 10 per cent of the total calories were fed as protein in most of the diets studied. The relationship between the nitrogen intake and the output of nitrogen in the urine was determined, and it was found that there was little difference between them. If the excess antiketogenic compound had been figured from the urinary nitrogen instead of from the protein intake, the value of the ratio described would 368 Studies on Acetonuria not have varied beyond the limits of experimental error. The intake of carbohydrate and of fat formed, respectively, the sources of 10 and of 80 per cent of the calories in the basal diet, and of varying percentages—5 and 85 per cent, 15 and 75, 20 and 70 per cent—in the other diets studied. An attempt was made to feed each of these diets for a period long enough to determine the level of acetone excretion which corresponded to it, but it was usually necessary to change the more severe diets before such an equilibrium was established. The diets used were figured from the tables given in Joslin’s Diabetic Manual (Joslin, 1919); they were prepared under the direction of a competent dietitian, and food not eaten was weighed, and the proper allowance made in the record; a complete sample diet is given for one of the cases. While a majority of the patients ate the diets as furnished, two did not, and the results of the studies carried out on them are accordingly not wholly satisfac- tory. It has seemed best to include these cases in this report, - however, as they serve as a check upon the results obtained upon other subjects. The urines were sent to the laboratory daily. It was impossi- ble to control the completeness of the collection through creatinine determinations because the presence of the acetone bodies in urine interferes with the method of analysis (Morris, 1915), and all of the cases except one showed a large excretion of acetone on all of the more severe diets fed. This lack of suitable control of the accuracy of collection made it seem best to record and plot the concentration of the acetone bodies as well as their total excretion. In some specimens marked variations in volume, total nitrogen content, and ammonia nitrogen content almost certainly show failure to collect accurate 24 hour specimens. These daily urines were analyzed for acetone bodies by a method recently described (Hubbard, 1921) by which the acetone plus acetoacetic acid were determined together as acetone, and the B-hydroxybutyric acid was determined separately, also as acetone. Total nitrogen was determined by the direct Nesslerization method of Folin and Denis (1916), slightly modified to permit the use of the oxidizing and Nessler’s reagents, described by Folin and Wu (1919). Ammonia determinations were made by the permutit method of Folin and Bell (1917). In some instances R. S. Hubbard and F. R. Wright 369 other factors were studied which were connected more indirectly with the main problem. Total acidity was determined in the urine of two of the patients by the method described in Folin’s Manual (Folin, 1916),? and its hydrogen ion concentration in one of the experiments by a colorimetric method using the stand- ard universal buffer solution described by Acree, Mellon, Avery, and Slagel (1921). This solution was standardized before use against phosphate solutions of known hydrogen ion concentra- tion. Besides these determinations on the urine the stripped weight of the patients was recorded, and in most instances the tension of carbon dioxide in the alveolar air was estimated. This determination was carried out in Cases II and III] by the method of Marriott (1916) and in the other cases by the Fridericia method (Fridericia, 1914; Poulton, 1915). The results obtained are given in Tables I to VI, and plotted in Charts 1 to 6. In these charts the diet is indicated at the top in terms of both total food and of percentage of the calories fur- nished by the three main classes of foodstuffs. When the diet varied much from day to day the average intake was made the basis of this plot. The numerical value of the expression 100 x 1.5 (weight carbohydrate) + 25 per cent (weight protein) 95 per cent weight fat © = N per cent was plotted to correspond with the intake of food for each day, and the daily excretion of all the acetone bodies reckoned as the sum of the acetone which could be formed from them was plotted below it. Since slight increases of acetone above normal may be of considerable importance in this study, the plan was adopted, in two of the cases studied, of plotting the excretion of acetone bodies in terms of their concentration in the urine upon paper with logarithmic characteristics. This method shows differences which are actually slight but which may figure as large percentage increases because the normal amounts are small. In these plots the two fractions of the acetone bodies—acetone plus acetoacetic acid and §-hydroxybutyric acid—were plotted separately to show the relationship between the two fractions, and both were plotted as acetone to make the curves comparable. The concentrations 3 Folin (1916), p. 103. 370 Studies on Acetonuria instead of the total excretions were used because of occasional failure in the collection of accurate 24 hour specimens. Cases I and IV are reports of two experiments carried out on the same normal subject (one of the authors, R.S.H.). The data reported under the heading ‘‘Case I’’ have been previously presented (Hubbard, 1921) and are repeated here because a comparison with those obtained on the same subject in a later experiment show some things which are not as well brought out in other studies. The subject was a man 5 ft. 103 in. tall, who weighed 165 lbs. and who at the time of the first experiment was 28 years old. During both of the periods he did light laboratory work while the experiments were going on. The first series of results was obtained before the appearance of Shaffer’s papers in 1921 and the diet was differently planned from those used in the other experiments. The fat and carbohy- drate were fed in different relative amounts—multiples of 50 gm., as a study of the table shows—and an attempt was made to feed sufficient protein to keep the caloric intake constant. The subject had been living on a normal mixed diet up to the first day of the experiment. There was a slight nega- tive nitrogen balance during the first part of it, and a slight positive one after the diet had become more nearly normal, but the difference was not great enough in either case to affect the calculation of the probable keto- genic balance seriously. The development and clearing up of acetonuria is clearly shown in Table I and Chart 1. There was certainly an increased acetonuria on a diet which contained 250 gm. of fat, 50 gm. of carbohydrate, and 68 gm. of protein; this diet has a ketogenic balance of 42 per cent in terms of the formula suggested for expressing that balance. The excretion of the acetone bodies was slightly increased during the first 3 days of the experiment, but the increases were so slight that they cannot be attributed with certainty to the diet. The acetonuria completely cleared up when a diet having a ketogenic balance of 152 per cent was fed, and it seems certain that the border-line diet, that is, the diet representing ketogenie equilib- rium, must lie between the two extreme diets fed, and probably does not lie far from that fed at the start of the experiment which has a ketogenie balance of 97 per cent. Attention should be called to the gradual increase and decrease of acetonuria as it developed and cleared up; a study of the table and chart makes it seem improbable that the second diet was fed long enough to cause a maximum excretion of acetone to correspond with its composition. The experiment recorded under the heading ‘‘Case IV’’ was carried out on the same subject as was that recorded under ‘‘Case I.’’ The height and weight were approximately the same as those given in the preceding para- graph, and the age was 33 years. The basal metabolism measured 1,750 calories per day. The diets were calculated and fed as here described, and the subject ate the entire amount of every diet provided. The collec- tion of urine samples was accurate. There was some loss of weight during the first part of the experiment, but when diets were fed which caused an ov *9U0J99B JO SUIIO} UT possoidxo o1¥ SOIPOg 9U0499B OY} JO SUOT}RUTULIOJAp OY} JO S}[NSoxyT » “Eh ra. «6s F000) 0 F #200 aa! 06F'0 | #66 | OT9 Str ‘Z| 9 FZ! OST | 2°89 SZL || 2 Tt GL GG» = 9200 | FE €c0'0 8G 199°0 | 9 OT | OT8 Shr‘S | OFZ) OST | 2°89 G2T ) 21 CL Io» . 6000] FE 9¢0°0 €'§ 029°0 | $2°6 | 098 Shr'Z | 9 FZ} OST | 2°E9 GLbe | £01 GL 0B» lan 1900 | 8g LOT 0 £01 | 6780 | SIT | OFT | 8hF‘Z| 9'FZ] OST | 2°E9 CLT || 2 11 GL 61» . cse0)} 68 PLS 0 0°82 | 7960) F&I | 086 Shr ‘S| 9'FZ] OST | 2°89 C2 oe UE GL 1 Oe cy cP I GIT | c08'0 0°89 | F160] S°2r | OST'T | 112%] GZ 0S bos TSG |) OF 89 LT os mo «3c990)| 069 LP 0 08S | ST8'0;} 9 FL | ST8 LPL @ \) Soe Og F G8 Tee i teOL 89 Ol » & 9060) 88t | EO) SE | L820) SLT) 960'T | TIL} GL | OF | Fe8 | 19% | TOT | 89 ole» c60'0 | Gé ¢90°0 GL 9242°0 | 9 9T | 906 €2¢‘% | ¢°ST | OOT | $69 102 ST 96 tL» ST wo0| > 8100) FF | 8hL0] O'LT | OOT'T | G2g‘%| GST} OOL | 969 | 102 ST | 96 el» en ‘ a 020'0 | 32 900°0 L'0 8hL0} SCT | S&6 €492| 3 St} OOT | $69 10Z ST 96 GI “ABIN -Q “mB 29 QOT/' bu "UB 99 NOT /* bu “mb “UB 09 quao ad “Ub quad Jad 0) qua wad “UB LIGI 3 a eS eS eS EEE ee jan ein heroxp iu-g ‘orjoowtp + ouoqaoy | NPN fap ‘euUINJOA |‘sats0[BO] ‘ayeipAyoqieD “48 ‘ula}O1g * 978q io) ‘aulIy) “Jaq ~ *[ asp) ‘TI ATAVL 372 Studies on Acetonuria excretion of only very small amounts of acetone the weight was quite con- stant. There was a slightly negative nitrogen balance as based on the determination of urine nitrogen during the first part of the period of study, and a slight retention later when the diet fed was more nearly normal. The excretion of acetone increased from 40 mg. on the day before the experi- ment to about 1.25 gm. after 4 days on a diet which contained 10 per cent CAL 2570 CAL 2700 CAL 2460 R/S %968™ oo JPR 10.1 %68.5 9" rc PR 17% 7249" DIET A MMB $ Cy iss26100 DIET B cH 7.5%50.4 OVET CES ey aa ceziso FAT 69.5 26201 FAT 32.4% 151 FAT 63.7 4675 MARCH 1937 12 13 14 15 16 17 18 19 20 41 22 ——————— bas See CuHarrt 1. of the calories as protein, 10 per cent as carbohydrate, and 80 per cent as fat; this increase makes it seem probable that there were more ketogenic than antiketogenic compounds in the diet. When the relative amount of carbohydrate in the diet was increased the excretion of the acetone bodies diminished, but did not return to the normal values when the diet contained 20 per cent of the calories as carbohydrate and 70 per cent as fat; oleomar- garine and olive oil were substituted for the larger part of the butter fat in R. §. Hubbard and F. R. Wright 373 this diet for 3 days, the 7th, 8th, and 9th of August, but the excretion of the acetone bodies was not measurably’decreased further. However, the amounts of acetone excreted when this diet, which has a ketogenic balance of 108 per cent, was fed were not markedly different from those found on a less severe diet which had a ketogenic balance of 79 per cent, and were not very large in either case. The border-line of increased acetonuria appears to lie, for this case, between diets giving values of 78 and 108 per cent, although the value may be higher if the excretion of very small amounts of acetone is regarded as important in determining when ketogenic antiketo- genic equilibrium has been established. It is noticeable, from comparing these two experiments on the same individual, that the excretion of acetone depends on the ratio between fat and carbohydrate rather than on the fat content: of the diet. Detailed Diet. Case lV. July 15. Breakfast: Eggs, 2; Bacon, 20 gm.; Cream, 45 ce.; Butter, 15 gm.; Bread, 15 gm.; 10 per cent fruit, 70 gm. Dinner: Meat, 60 gm.; 5 per cent vegetable, 135 gm.; Potato, 30 gm.; Cream, 40 ce.; Cheese, 15 gm.; Bread, 15 gm.; Butter, 25 gm. Supper: Meat, 30 gm.; 5 per cent vegetable, 135 gm.; Bacon, 10 gm.; Cream, 35 cc.; Cheese, 15 gm.; 10 per cent fruit, 50 gm.; Bread, 15 gm.; Butter, 27 gm. Extras (one-third to each meal): Olive oil, 90 cc.; Lemon, 50 ce. Fore- noon—Olive oil, 30 cc.; Lemon, 15 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 16. Breakfast: Eggs, 2; Bacon, 20 gm.; Bread, 15 gm.; Cream, 60 cc.; Butter, 20 gm.; 10 per cent fruit, 80 gm. Dinner: Meat, 60 gm.; Bread, 10 gm.; 5 per cent vegetable, 120 gm.; Potato, 30 gm.; Cheese, 15 gm.; Cream, 60 cc.; Butter, 30 gm.; Water- melon, 45 gm. Supper: Meat, 30 gm.; 5 per cent vegetable, 120 gm.; Cheese, 15 gm.; Butter, 25 gm.; Bacon, 10 gm.; Cream, 60 cc.; Bread, 20 gm.; Watermelon, 45 gm. Extras (one-third to each meal): Olive oil, 90 ce.; Lemon, 20 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 17. Breakfast: Eggs, 2; Bacon, 20 gm.; Bread, 15 gm.; Cream, 60 cc.; Butter, 20 gm.; 10 per cent fruit, 80 gm. Dinner: Meat, 60 gm.; Bread, 10 gm.; 5 per cent vegetable, 120 gm.; Potato, 30 gm.; Cheese, 15 gm.; Cream, 60 ce.; Butter, 30 gm.; Water- melon, 45 gm. — EEE ee 374 Studies on Acetonuria Supper: Meat, 30 gm.; 5 per cent vegetable, 120 gm.; Cheese, 15 gm.; Butter, 25 gm.; Bacon, 10 gm.; Cream, 60 cc.; Bread, 20 gm.; Watermelon, 45 gm. Extras (one-third to each meal): Olive oil, 90 cc.; Lemon, 20 ec. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 18. Breakfast: Eggs, 2; Bacon, 20 gm.; Bread, 15 gm.; Cream, 60 cc.; Butter, 20 gm.; 10 per cent fruit, 80 gm. Dinner: Meat, 60 gm.; Bread, 10 gm.; 5 per cent vegetable, 120 gm.; Potato, 30 gm.; Cheese, 15 gm.; Cream, 60 cc.; Butter, 30 gm.; Water- melon, 45 gm. Supper: Meat, 30 gm.; 5 per cent vegetable, 120 gm.; Cheese, 15 gm.; Butter, 25 gm.; Bacon, 10 gm.; Cream, 60 cc.; Bread, 20 gm.; Watermelon, 45 gm. Extras (one-third to each meal): Olive oil, 90 ec.; Lemon, 20 ee. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 19. Breakfast: Eggs, 2; Bacon, 30 gm.; Bread, 10 gm.; Butter, 20 gm.; ‘ Cream, 30 ec.; Watermelon, 60 gm. Dinner: Meat, 60 gm.; 5 per cent vegetable, 185 gm.; Cheese, 15 gm.; Butter, 34 gm.; Cream, 30 cc.; Watermelon, 60 gm: Supper: Meat, 30 gm.; Bacon, 30 gm.; 5 per cent vegetable, 135 gm.; Cream, 30 cc.; Cheese, 15 gm.; Butter, 30 gm.; Watermelon, 60 gm. Extras (one-third to each meal): Olive oil, 80 ec., Lemon, 15 ec. Summary: Carbohydrate, 31.5 gm.; Protein, 63 gm.; Fat, 236 gm. July 20. Breakfast: Eggs, 2; Bacon, 30 gm.; Butter, 30 gm.; Cream, 90 cc.; Watermelon, 100 gm. Dinner: Meat, 45 gm.; Cheese, 15 gm.; 5 per cent vegetable, 180 gm.; Butter, 30 gm.; Cream, 60 cc.; Watermelon, 65 gm. Supper: Meat, 30 gm.; Bacon, 30 gm.; 5 per cent vegetable, 180 gm.; Cheese, 15 gm.; Butter, 27 gm.; Cream, 60 cc. Extras (one-third to each meal): Olive oil, 60 ec.; Lemon, 15 ce. Summary: Carbohydrate, 31.5 gm.; Protein, 63 gm.; Fat, 236 gm. July 21. Breakfast: Eggs, 2; Bacon, 30 gm.; Butter, 25 gm.; Cream, 60 cc.; Watermelon, 100 gm. Dinner: Meat, 45 gm.; 5 per cent vegetable, 180 gm.; Cheese, 15 gm.; Butter, 30 gm.; Cream, 90 cc.; Watermelon, 65 gm. Supper: Meat, 30 gm.; Bacon, 30 gm.; 5 per cent vegetable, 180 gm.; Cheese, 15 gm.; Butter, 27 gm.; Cream, 60 ce. Extras (one-third to each meal): Olive oil, 60 ce.; Lemon, 15 ec. Summary: Carbohydrate, 31.5 gm.; Protein, 63 gm.; Fat, 236 gm. : : | R. S. Hubbard and F. R. Wright 315 July 22. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 cc. Dinner: Meat, 45 gm.; 5 per cent vegetable, 90 gm.; Bread, 15 gm.; 10 per cent fruit, 40 gm.; Butter, 30 gm.; Cheese, 15 gm.; Cream, 60 cc. Supper: Meat, 30 gm.; Cheese, 15 gm.; Bacon, 25 gm.; 5 per cent vege- table, 90 gm.; 10 per cent fruit, 55 gm.; Bread, 20 gm.; Butter, 25 gm.; Cream, 60 cc. Extras (one-third to each meal): Olive oil, 60 cc.; Lemon, 15 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 23. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 ce. Dinner: Meat, 45 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 40 gm.; Bread, 15 gm.; Cheese, 15 gm.; Butter, 30 gm.; Cream, 60 ce. Supper: Meat, 30 gm.; Bacon, 25 gm.; Cream, 60 cc.; 5 per cent vege- table, 90 gm.; Butter, 25 gm.; 10 per cent fruit, 55 gm.; Bread, 20 gm.; Cheese, 15 gm. Extras (one-third to each meal): Olive oil, 80 ec.; Lemon, 10 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 242 gm. July 24. Breakfast: Eggs, 2; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 cc.; Bacon, 30 gm.; 10 per saat fruit, 100 gm. Dinner: Meat, 45 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 40 gm.; Bread, 15 gm.; Cheese, 15 gm.; Butter, 30 gm.; Cream, 60 cc. Supper: Meat, 30 gm.; Bacon, 25 gm.; Cream, 60 cc.; Butter, 25 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 55 gm.; Bread, 20 gm.; Cheese, 15 gm. Extras (one-third to each meal): Olive oil, 40 ec.; Lemon, 10 ec. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 202 gm. July 25. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 cc. Dinner: Meat, 45 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 40 gm.; Cheese, 15 gm.; Bread, 15 gm.; Butter, 30 gm.; Cream, 60 cc. Supper: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 55 gm.; Cheese, 15 gm.; Bread, 20 gm.; Butter, 25 gm.; Cream, 60 ce. Extras (one-third to each meal): Olive oil, 60 ec.; Lemon, 15 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. July 26. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 ce. 376 Studies on Acetonuria Dinner: Meat, 60 gm.; Bacon, 20 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 25 gm.; Cream, 45 ce. Supper: Bacon, 40 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 35 gm.; Cream, 45 cc. Extras (one-third to each meal): Olive oil, 30 ec.; Lemon, 10 cc. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. July 27. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 cc. Dinner: Meat, 60 gm.; 5 per cent vegetable, 90 gm.; Cream, 45 cc.; Bacon, 20 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 30 gm. Supper: Bacon, 40 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Cream, 45 cc.; Bread, 35 gm.; Butter, 30 gm. Extras (one-third to each meal): Olive oil, 30 cc.; Lemon, 10 ce. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. July 28. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 cc. Dinner: Meat, 60 gm.; Bacon, 20 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 30 gm.; Cream, 45 cc. Supper: Bacon, 40 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 30 gm.; Cream, 45 ce. Extras (one-third to each meal): Olive oil, 30 cc.; Lemon, 10 ec. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. July 29. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 ce. Dinner: Meat, 60 gm.; Bacon, 20 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Cream, 45 ec.; Butter, 30 ce. Supper: Bacon, 40 gm.; 5 per cent vegetable, 90 gm.; Potato, 30 gm.; Bread, 35 gm.; Butter, 30 gm.; Cream, 45 cc. Extras (one-third to each meal): Olive oil, 30 ce.; Lemon, 10 ce. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. July 30. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 ce. : Dinner: Meat, 45 gm.; Cheese, 15 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 40 gm.; Bread, 15 gm.; Butter, 30 gm.; Cream, 60 ce. Supper: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 90 gm.; Cream, 60 ec.; Cheese, 15 gm.; 10 per cent fruit, 55 gm.; Bread, 20 gm.; Butter, 25 gm. Extras (one-third to each meal): Olive oil, 60 ec.; Lemon, 15 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. ‘ ¥ b 7 . R. 8. Hubbard and F. R. Wright OtL July 31. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 cc. Dinner: Meat, 45 gm.; 5 per cent vegetable, 90 gm.; Cheese, 15 gm.; 10 per cent fruit, 40 gm.; Bread, 15 gm.; Cream, 60 ec.; Butter, 30 gm. Supper: Duck, 40 gm.; Bacon, 25 gm.; 5 per cent vegetable, 90 gm.; Cheese, 15 gm.; Bread, 20 gm.; Butter, 22 gm.; 10 per cent fruit, 55 gm.; Cream, 60 ce. Extras (one-third to each meal): Olive oil, 60 ce.; Lemon, 15 ce. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. Aug. 1. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 15 gm.; Butter, 20 gm.; Cream, 60 cc. * Dinner: Meat, 45 gm.; Cheese, 15 gm.; 5 per cent vegetable, 90 gm.; 10 per cent fruit, 40 gm.; Bread, 15 gm.; Butter, 30 gm.; Cream, 60 cc. Supper: Meat, 30 gm.; Cheese, 15 gm.; Bacon, 25 gm.; 5 per cent vege- table, 90 gm.; 10 per cent fruit, 55 gm.; Bread, 20 gm.; Butter, 25 gm.; Cream, 60 cc. Extras (one-third to each meal): Olive oil, 60 ec.; Lemon, 15 cc. Summary: Carbohydrate, 63 gm.; Protein, 63 gm.; Fat, 222 gm. Aug. 2. Breakfast: Egg, 1; Bacon, 30 gm:; Oats, 10 gm.; 10 per cent fruit, 110 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 60 cc.; Cream, 90 cc. Dinner: Meat, 30 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; 5 per cent vegetable, 60 gm.; Bread, 30 gm.; Bacon, 25 gm.; Butter, 30 gm.; Cream, 90 cc.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 10 per cent fruit, 100 gm.; 5 per cent vegetable, 60 gm.; Bread, 30 gm.; Potato, 30 gm.; Butter, 30 gm.; Cream, 90 ec.; Milk, 30 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 3. Breakfast: Egg, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 110 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 60 cc.; Cream, 90 cc. Dinner: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 cc.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 cc.; Milk, 30 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. 0 SE OE ONS a a eee ae a 378 Studies on Acetonuria Aug. 4. Breakfast: Egg, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 110 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 60 ce.; Cream, 90 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 cc.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; Potato, 30 gm.; 5 per cent vegetable, 60 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 ce.; Milk, 30 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 5. Breakfast: Egg, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 60 ce.; Cream, 90 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 ec.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 30 ec.; Cream, 90 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 6. Breakfast: Egg, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 110 gm.; Bread, 30 gm.; Butter, 30 gm.; Milk, 60 cc.; Cream, 90 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 ec.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 5 per cent vegetable, 60 gm.; Potato, 30 gm.; 10 per cent fruit, 100 gm.; Bread, 30 gm.; Butter, 30 gm.; Cream, 90 ec.; Milk, 30 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 7. Breakfast: Egg, 1; Egg white, 1; Bacon, 30 gm.; Oleo, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 200 gm.; Bread, 30 gm.; Milk, 60 ec.; Cream, 30 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; 5 per cent vegetable, 150 gm.; Potato, 30 gm.; 10 per cent fruit, 50 gm.; Bread, 30 gm.; Oleo, 30 gm.; Cream, 20 cc.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 5 per cent vegetable, 150 gm.; Potato, 30 gm.; 10 per cent fruit, 50 gm.; Bread, 30 gm.; Oleo, 30 gm.; Milk, 30 ee.; Cream, 30 cc. Extras (one-third to each meal): Olive oil, 30 ec.; Lemon, 10 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Ce iii, +S —— R. S. Hubbard and F. R. Wright 379 Aug. 8. Breakfast: Egg, 1; Egg white, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 200 gm.; Oleo, 30 gm.; Bread, 30 gm.; Cream, 30 ce.; Milk, 60 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; Potatoes, 30 gm.; 5 per cent vegetable, 150 gm.; 10 per cent fruit, 50 gm.; Bread, 30 gm.; Oleo, 30 gm.; Cream, 30 cc.; Milk, 60 ce. Supper: Egg, 1; Bacon, 30 gm.; 10 per cent fruit, 50 gm.; 5 per cent vegetable, 150 gm.; Oleo, 30 gm.; Potato, 30 gm.; Bread, 30 gm.; Milk, 30 cc.; Cream, 30 ce. Extras (one-third to each meal): Olive oil, 30 ec.; Lemon, 10 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 9. Breakfast: Egg, 1; Egg white, 1; Bacon, 30 gm.; Oats, 10 gm.; 10 per cent fruit, 200 gm.; Bread, 30 gm.; Oleo, 30 gm.; Cream, 30 cc.; Milk, 60 ce. Dinner: Meat, 30 gm.; Bacon, 25 gm.; Potato, 30 gm.; 5 per cent vege- table, 150 gm.; 10 per cent fruit, 50 gm.; Bread, 30 gm.; Oleo, 30 gm.; Milk, 60 ecce.; Cream, 30 ce. Supper: Egg, 1; Bacon, 30 gm.; 10 per cent fruit, 50 gm.; 5 per cent vegetable, 150 gm.; Oleo, 30 gm.; Bread, 30 gm.; Potato, 30 gm.; Milk, 30 ec.; Cream, 30 cc. Extras (one-third to each meal): Olive oil, 30 ce.; Lemon, 10 ce. Summary: Carbohydrate, 125 gm.; Protein, 63 gm.; Fat, 194 gm. Aug. 10. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 cc. Dinner: Meat, 60 gm.; Potato, 30 gm.; Bacon, 20 gm.; 5 per cent vege- table, 90 gm.; Bread, 35 gm.; Butter, 25 gm.; Cream, 45 ce. Supper: Bacon, 40 gm.; Potato, 30 gm.; 5 per cent vegetable, 90 gm.; Bread, 35 gm.; Butter, 35 gm.; Cream, 45 cc. Extras (one-third to each meal): Olive oil, 30 cc.; Lemon, 10 ce. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. Aug. 11. Breakfast: Eggs, 2; Bacon, 30 gm.; 10 per cent fruit, 150 gm.; Bread, 20 gm.; Butter, 30 gm.; Cream, 90 ce. Dinner: Meat, 60 gm.; Potato, 30 gm.; Bacon, 20 gm.; 5 per cent vege- table, 150 gm.; Bread, 35 gm.; Butter, 25 gm.; Cream, 45 ce. Supper: Bacon, 40 gm.; Potato, 30 gm.; 5 per cent vegetable, 90 gm.; Bread, 35 gm.; Butter, 35 gm.; Cream, 45 cc. Extras (one-third to each meal): Olive oil, 30 ec.; Lemon, 10 cc. Summary: Carbohydrate, 94 gm.; Protein, 63 gm.; Fat, 208 gm. 20 per cent cream was used. All food was weighed after cooking. Nn wT SS OOOO SS SSS 1280 | Z S82 |S6r°0 | O'9F |€29'0 |SF'6 | 090‘T 19¢°0 | 809 |F6¢°0 | T'9 |FIS'O | TOT | $26 : F160 OIT |809'0 | OSL |209'0 | F'OT | 0&8 crs 0} § 02 |L16'0 LIT |009°0 |22°6 | O82 . cs8&0} 6 &r |2se0| 6 EF (S090 026 | SSL 9620] FSS |26€0| Scr |S89'0 | € OL | 0&6 8SF'0| 9 FS 19680] GLP F920 066 | OF8 = 9620 | 26 (2990 | OTL |006'0 |F2'6 | O82 — L0G'0| FH9 (2090) GPL | ZIT | 2°01 | 028 ae 66 T PLT |202°0 | 396 | GIT | 6°OL | OFZ = Ole |- 790 | €0T|> Sar | F0'T 0z8 5 90 § P9E | SPT CLT | OC T | 9 TT | OF8 a £0'F 99 | 02 T PST | OST | 2 SI | COLT 16'S G8E | ZIG ZGL \n29: Dl SiGl- "OSE 1 5 LG F 99€ | Fr I OZI | OZ T | 9°SE | 002 ‘T cé T SZT {268 0 GIL |$26'0 | SCT | O82 8 826 0 CSI |oZP'0 | F 29 |002°0 | T&T |} 002 oS csg 0 | FG2 {4890 |) L218 |129'0 | FPL | 022 = 9F1 0; OIG |FE20!| SPE |46F0 | 6 CT | SZ9 WM 6910] S92 |0S20| T'6E \OSP'0 | SCT | 0F9 LOROR cee 900.05) “Sar | S25 Onl Geb |s00S “mwB \9999T/'bu\) “wb | 09 gp ,/ bu) “wb “ub 09 Laan 4 gnojesy © |NYEIN'| tetas, | fon ‘oulIQ —) ie) ine) ae ee — lle “UU OO Ie] -O0AV 92 |86r‘2 | 02] Ser S6r‘% | 02) Ser 0°92 |z0¢ ‘% or] ¢&9 Z0S ‘% O-| 8&9 €°92 |z0S ‘% Oo. | 8&9 00g ‘Z GI | +6 00g ‘Z GI | #6 8° FZ 1009‘ cI | 6 00S ‘% GI | +6 02 |z0¢‘4 | OL] &9 028°S e201. | 2-89. 789 ‘2 | 96 £9 F SL |Z0¢ ‘% or | $9 0°92 |Z0¢ % G ras Zz 92 |z0S ‘% G ce Z0S % G Ze 90¢ ‘% OI | £9 8 92 |90¢ ‘% or | 89 8°92 190 ‘% Or | 89 8 92 |90¢ ‘Z OL | §9 BO é é “BY quaowad | “ud Hee ‘apyeupAyoqiey "AI 980) ‘II GHIAV.L OL | POI OL | POL O8 | GGZ O8 | GGG O8 | GGG GL | 806 GL | 806 CL | 806 CL | 806 O8 | GG% ¢'8L | 60G I8 | GPG O8 | GEG G8 | 9&C- GS | 9&6 G8 | 98% O8 | GZS OS | GGG 08 | Z2z O8 | GGG é é quao dad “mb “4R ‘pid Or; & Or | £9 Or | &9 OL | €&9 Or | €9 Or | £9 Or | &9 oT | &9 OL |. &9 OL 69 L101 | & 6 69 Or | &9 OL | &9 Or | &9 OL} &9 OT | &9 OT | &9 Or | &9 OL | £9 é é quao sad ub “u19}01g yypsro yy | LT OT oI ”? Ajne 381 ‘T quourl10dxnf 10J posn oUo oY) SB OWES OY} SVM pouLLOj.od sv yUOUT -10dxo siqy uoym uodn qoofqns oy, ‘oU0JoOR JO SsuIIO} UT possordxo O18 SOIPOG OUOJOOV OY} JO SUOT}LUTUIOZOpP JO S}[NSoY R. S. Hubbard and F. R. Wright 8820 | 9°88 [6220] 9°02 |0F9'0 |E 2I | 098 8°S2 008 ‘% cl | +6 GL | 802 or | 9 el as LOT'0 | ESI IPel'O| 2°91 |PZ9°0 | 2°0T | 008 00¢‘Z| SI| 6 GL | 80% or | £9 Il ow Z1T'0 | 91% |09T'0 | 0°02 |88F'0 | O'OL | 008 | ZF 86r‘Z| 02] Sat OL |. P6T or | 9 OT ay gce'0 | 2&P |P6l'0| O€2 \zgr'0 [776 | O€8 | OF | 8'SZ |86r‘S | 02] Gar OL | 61 or} 9 6h Ph z60'0 | SOL |z91'0 | 0°02 |2TS'0 |St'6 | OS8 | OF 86h ‘Z| O2| Ser OL | F6I or} 9 Bo ay 760'0 | S8'IT |080°0 | O'OL (|8o'0 [02'S | 008 | OF | 2°92 |86F‘%| 02] S2r OL | P6T or} 9 eg 1280 | § Sh |e8c'0| GFE 0z8 Ip | 8°92 |86F‘% | 02] Ser OL | ¥61 Or | 9 Oat! 962'0 | 628 |P28'0} T'98 [08G'0 189°6 | 006 1°92 |86P ‘Z 02 | Gal OL | P61 Or | ¢9 g ‘any Studies on Acetonuria 382 _.. bhi Zol is StI%Zot HI} Aa Lia bb ZSi HESS DLFIid Sle “S$ HOM g igid £9 %0!) HO ‘wb £9 ZO! Ud | SESSHSZPZZAZ717ZACZZIZZACLCA ZA d WS SSSSASATSI OT TT im = bes si (el if | || a 5 || fe a || i at = = | fa ia | is ci uo’ glo Sb € TI IE OF bE BZ Lt OF SC be SUC (IT OF bI Bt Li 9! St 1Isnony ATHS por Ys re 960 S58 INS tzo %08 LY4 “wb £9 SO Yd ‘wb €9 %0) Yd ‘wb $9 %Ol Yd ANOLIJOV | Vv isla 9 10 Wt 12 ~ c 0 lo vt ~P N 7) ° Led c N oo x o Nn lo N x Led N La! N N ° N s —) = = b ‘ooo tt tt my ACETONE R. §. Hubbard and F. R. Wright HY DROXY- UTYRICO ACETIC ic) p 383 Cuart 2A, a . — a Pe ee a *9110}99" JO SUIIO} UT passoidxa o1¥v SOIpOg 9U0490v OY} UO SUOTYLUIUIIOJOp JO S};NSEY SO 0 OT. |\FIOo | OT 1820 922 | SPS Ig8e‘T | 2°99 |9T8‘T ST | 89 | I°SZ | 8ST OL| Sh | OL » 0g0'0 | €2 |s8c00}] ae |09z°0 09% | S'S 008 ‘T | 2:99 |9TS‘T GI | 89 | L°SZ | ecT Ol| ch 1/6 » 8200} O04 (|F20'0! LT 6920 08% €¢ |c6e‘T | 2:99 |9I8‘T Gt | 89 | I'S2 | 8ST Or; ch {8 » z99 |9TS ‘T SI | 89 | T°SZ | SST OF | Sha oe 0°99 1221 ‘T 6S 06 Th thre 918 ‘T ST | 89 | T°S2 | 8ST O1| SF |e » 7800] &2 |0900] IF (|ezz'0 OFS GG |09F ‘T | 2:99 |9I8 ‘T GI | 89 | T'SL | &ST OL} Sk ly 3 ss0'0 | 24% |9900| S'¢. \o0z'0 F8 bo 022 | 2:99 |och‘T | FOL | 2E | $62] 92T OL. . 86 "e" S 200] 62% [8900] ze \z6t'0 | 0¢°9 | F2z LG [008 ‘T | €°99 loze‘T | FOL | LE | $62] 92T OL | “S92 es. 3 5 r40'0| LY (|F90'0| 'T'F IP8z0 | OTL | 68z €°¢ |goc ‘I | $99 j9ar‘T | OT | 2E | $62 | 92T or] 9€ | IT ‘ze—W 5 960.0} 9S [FOTO] 82 {E820 8SZ FS OFZ ‘T | $99 |9cF‘T | FOL | Ze | $62 | 92T 0) ee a” A 5 1°99 |9tF‘T | FOL | 2E | S62 | OCT OL -c98 l- Zeus © &810) SSI lest'0| G:2r- (2080 | 20°2 | $92 | Ges lose‘T | 8:99 \szF‘T | 96 98 18-1 SOL e6 om ae <| 6ze0| 9:2 loizo| get rego | 20°9 | t6z G'S [00r‘T | 1°29 |SZF‘T | $'6 9& 18 | 81 | 9°6 Ge. )..08. 0 = eter 0} 92 6010) 69 © e080 06z br (G29‘T | 0°29 |8ZF‘T | 96 9g 18 | S&— | 9°6 a a © gro] 12t loet'o| et loro 942 | G'S |088‘T | 0°29 [SLT | 96 9g 18 | &&t |S°6 og lee @ 700] 2b |rFIO} G9 lese0] OLS] OTs | S'S \06T'Z | 0°29 IszF‘T | 96 | 98 1S |-6er | 6 | Be Pee cITO!] 19 |IZT0} & 6 9rE0 108 | &'S (962 °‘T | 229 827‘ | 96 | 98. je ge) Guu als Mk) Sa dae i} Slr ‘I | $6 98 18 | €s1 | 9°6 oF 10g 4 om OIL 0} TS lero | 9°9- \908'0 198 €°¢ loft ‘% | 2°29 |SZh‘T | 96 9g 18 | 81 | 9°6 Ye et ee r0'0'| &<¢ |2000)| 2°¢ \9z%'0 922 | GL’°S [092 ‘T | 0°89 |SzF‘T | 96 9g 181|-S2b 36 TS tag eee 6107012; 0- - |810°0")-8:0 = 0Sz:0 66z &G l0&L‘T | $89 |82zF‘T | $6 9g 18 | £81 | 9°6 aa figs ee 6000 | GO |2000] 20 {|s6z'0 8ce 09 02g ‘T é é é é é d é 91 “qaq ‘mB \-99 Q0]/ bu) -wb |-99 gy, /'Ou| -wd w6 |Nj'9'99| Ad 29 “By quaadad| wb |quaatad| "mb |,ua240d| “wb 1@el Ly Sea as boe N’HN aan “ploy ae seit ono ‘ayerpAyoquey BH | ula}OIg “qysta yy | —— ‘ayeq = euliy) “qoIqT i) = Ohh asD-) SORRCHEAAD a a Oe SRS Se —_ === R. S. Hubbard and F. R. Wright 385 Case II was a woman, Miss A. G., aged 47 years, whose height was 5 ft. 2 in. and who weighed 137 lbs. Her basal metabolism measured 1,335 calories per day. Like all of the pathological cases included in the series, she was a severe chronic arthritic of long standing. She made every effort to cooperate, but after a little more than a week she found it difficult to take the basal diet, and it was accordingly modified as shown in Table III and Chart 3. The diet fed during the first part of the experiment did not contain enough food to maintain the weight of the patient, and there was a progressive loss during the first 2 weeks. In four out of five determina- CAL 1478 CAL 1426 rg ae ed PR 9.5% 35-79 PRIO %o 35.79 PR 10% 45-49™ DIET A Mo gisv35.7 DET BOMD) cy iogn7.1 PET C2 cn 15% 68 FAT®} 132-7 FAT7942% 126 FAT 75 % 53 FEBRUARY MARCH Oe 16 §7 18 14 2041 2223 2492526272381 234567 8 4 '10 BRS a Bee ee er eee JENS Re SRERRERS ase Baas (a a a . 30 SSR @ || ee eet IB bel) Son Anes Shea TOTAL ACETONE CHART 3. tions of urinary nitrogen made during this period there was more found than was contained in the food taken, but on the day when the difference was largest, Mar. 1, the ratio used to express the ketogenic balance had a value of 56 per cent when calculated from the food fed and of 59 per cent when the protein burned was calculated from the urinary nitrogen; this diff- erence is almost certainly within the limits of experimental error. This case showed the smallest excretion of the acetone bodies of any in the series (the scale which has been used in plotting the acetone excre- tion is ten times as great as it is in the other cases), but still, when diets approximating the 10, 10, 80 per cent, diet were fed, the concentrations of ra Studies on Acetonuria 386 ———— 88's} soc | SIT S9T [926° 0/826 | LOE | 2°9 | S89 OTT} 062 |8SP'0 60Z |FLSO\cr SG | OZT | 24'S | O8E 029'0} 60 |F8Z'0| 6 88 |9ZF O)99'E | EST LG | 0GE 869°0] 9°96 |8EE0| FOF J9SF Olcr GS | 86T | L4°S | 0&2 F160 FIT \6SF'0| F 2G |2F9°0|99'9 | 28% | G's | 008 G26'0| +92 |6FE'0| 2°66 |9ZE OSES | SFI | 0-9 | OSE GI I GOT |S69'0} 288 |FLE 0/0G' 2 | 66T § 9 | S29 €TZ'0| 60% |80E°0} 206 |61G OSF'F | O&T €9 | OVE 69 T 91Z |069'0 SOT |882 0/6E°9 | LIT 9°9 | SZg CF6'0| OTT |8ZG'0| 8 FO |69F O/FG 2 | cae |S0'9 | STS 69'T ¢9% |§19 0 SOT |S0S O\ZF F | 8SZI |S8 ¢ | OF9 OF I 9ST |08¢'0| Z 6¢ |§62°0\228 | FFE | 9G | G68 69'T COP |SLZP'0|- SIT \00L O]ZF F | GIG [GGG | Och rey | SOI |8OF'0| & 6G |9T9 O/ET'9 | 9S | ES | 069 611 902 |Z1F'0| 22 {969° 0\0F'9 | 908 |SE°S | OSS GZL°0| GI \922'0| & HH \Ser O/GG'Z | ZcE& | FG | OL C&T'0| $ FS 12600} 291 |929°0/06'9 | 09G | Fg | OSS 910°0} 0% + |920°0} ZE jOT9 O/ST'S | O8@ | 24'S} STS “ul aniy al ‘wb |:09 go, /*bu| “mb | *wB |NI‘9'09| Wa “00 ee eee "prow oLAqnq *orjo08Ip is S, = 8 = -AxoipAy-g -+ du0jo0Vy Z Z = 8 : 5 : “ou Té | 0:0F L1G 0€ | o OF 0 OF && | P OP LE | 0 0F L&E | » OF LE | G OF PE FS | 0 OF O€ | 2 OF cé | 0 07 rE | 0 OF vé | 0 OF 6& | G OP FE vE T OF I OF “UU By *OO rey FY SIOM -O9ATV OD T8} “AI GTAVL 68°& 68°€ 68°€ 68°& F6.1 ¥6 1 F6 I *SOTIO[BD ¢ | 61 G | 61 CeeOL OT | 28 OT | LE OT | LE OT | Z§ OL R25 OLMPAs OT | Z& OL | 2& OT | 2ZE OL | LE OT} 28 OT | 2Z& OF 26. OWES é é Ma | wo e) ee pay a “7910 a i i ee ae GS | TfL | OT G3) TE |e Or cS | TfL | OT OS TiLrsh| > OL 08 | PEL} OT 08 | PET | OT O08 | FEL OT O8 ss eSle Or 08 | PEL Or 08 | P&I Or 08 | PEL OL O8 | FET OT 08 FET OL 08 | FET OL OS FEI OT 08 | FEI OT OS FEL OL é é é musa | mo | 12 z *ula}O1g €I ” 4 ey IT “18 lé6T ‘a18q 387 ~ =| a2 = ‘9103098 JO SUII9} UL possordxo 018 SOIPOG 9U0J99R OYY JO SUOTYRUIUIIOJOp Jo syNsoyy a MS a EE SE eS ES a a ARs EL Ser tee ns ee et ee SE |r YS a 6° OF TOGST | OS ek WO. 2h) On Leake - §80'0| #9 ({St0'0) 9°8 [09z Ole’ | Z6L | 2:9] OzS| 98 | TOFD | OS 92. O22 Opa 2a eee oy a 090'0| 2°24 |€90°0} 0°6 |98z°0\86'F | TIZ | T°9 | StIZ| 98 TOP 1) OS) 2 1 OF 1 21F | OL) Peers Gs TS 800| FF (|0S0°0} 2°9 |F8z'0160'S | Zoe |Gz°9| OFZ | 98 TOGET || SOe Gk 10k ar |) OL ICRR eer ae = GIO] Z°2ZT |SSl'O| 2 9T Iz6z'olee's | ose | 6S | Szs| ge TOG Ta 20C sO2. 02 RRL <\ OL 1 ZR act ry gy «892'0] 2°0E [06T'O} 2° |P8E'Ol0s's | F2E |go'G | 098 | gE | TOGST S|) OS.) -92 1 GL2 6) aT |) SO | ZO Ones ss a 6G'I] 992 |€Z¢°0] 2 €6 |0z9'010z'9 | g0E Ieg'¢ | OzO | GE COOT (9°6. |26) | TS REIS. | ZEN ee Se = G69'0| LIZ |2492'0| #08 joze'ol00'F | 6zZt | 2°¢ | oze!| ge goo |9°6 | 28 | TS: |} TPL 1G6. | Ze RR | 992°0| Z°9F |8ZT'0| g'Te |stg‘olze'¢ | 9F~ |gs'¢ | ozg | Ee | 209 T (9° | 28 11S PRT. 96 || JB tr |= SAT 0/ 0'0E jecr'0] 2°02 jose OjZ9'¢ | sez | 89] Go| EE Gor T.| OT | OS Wes TSSr Ol ee oe ow LE1'0| 9°92 |1ZI'0| 81% |ZFE'OlP8'S | HZ [PSS | Sco! 6z GOP Te) OT: OS ez eer |) OP erticg: aa, MM SIt‘o| ose |s90'0| o'1z sce olzo'r | zetr lege | cze! te Cop T | St 99 | 92 out | Ome |e | - 890°0/ 6'9 |FS0'0| O'OT {|2ZSF'0102'9 | 89L | 2:91] OF8| ZE G6m Ei) eSLalVOO.4)-G2. Gen | OT | 2E5 | a a Ps zcz'ol #92 |srz'0l 6'¢z lets olor'9 | Tez | 6°¢ | 096 CAP Te| OR. |Ome S20 (5220: |) OL | Ze le 8620} ST |8Z8'°0| 2 FL P29 0/189 | 26E | 6S | OTS 209 Teh MOT 2S 4)-08 FET | OTs ze | t “adv 999'0| G'28 |FZe°0| 2 6P 092 OFFS | TEz | 2:9] 092) TE 20 Te) SOT 2ees) 08> PEL |. OT 1-28 | Te L49°T| 281 |$Z9'0| SFL | OL'TI8h'2| Soe i¢o°S | OFS | OF BOS e200 te Za og | Pet. | (OTM Z8). 08. a, 8¢°Z} STS |1S8'0 COL | 28 1/9F' 2 | 962 | 9'S] O28] 8z | cog ‘T | OL] 26 | 08 | PEL] OT] Ze | 62 ‘aeW i ee es ee es ee =o - Te es Studies on Acetonuria 388 ‘p LUVHO 2 2 a a a By ea an PN ps RL: VA HNOLSOV AWLOL EINE ei a At : Z / leas 6 rrr TTT | | | IC Si bl ci TEV Of b 2 bo § & € TUE O€ BT Br lt ov Gz 47 Et Tt iW OF *bI Bi Li Of SI bi a ~ o ~ ~ nn -) ~ a] NX a Nn ”- ~ N N 5 —] nN = oo = ~~ = uw = = = foo ca, 1,000 ACE TONE PLUS ACETO- Acetic @ 500 PHY DROXY BUT YRIC O CONCEN TRATION es re EO 390 Studies on Acetonuria the acetone bodies varied from ten to one hundred times the values of those found when the subject was on a norma! diet. It seems possible that the continued loss of weight and negative nitrogen balance noted in this case may be accompanied by a liberation of glycogen which would serve as an additional (endogenous) source of antiketogenic compounds, but as other experiments, as Case IV, do not show evidences of such a process during loss of weight this explanation can only be offered tentatively. The figures actually obtained show that the border-line diet was very close to the basal one, and as this is the only case in the series in which this was found to be true, it seems reasonable that some such phenomena as those suggested above may have diminished the excretion. The results show also that the ratio between the fat and carbohydrate rather than the amount of fat ingested determines the acetone excretion, for less acetone was found when the diet contained 153 gm. of fat and 45 gm. of carbohydrate than when it contained 133 gm. of fat and 36 gm. of carbohydrate. The excretion of the acetone bodies was so low that the determination of creatinine was not interfered with, and the determinations of this compound showed a high ~ degree of success in collecting accurate 24 hour urines. Case III, Mrs. M. H., was a woman 28 years old, 5 ft. 3 in. tall, and weigh- ing 70 lbs. Her basal metabolism measured 1,240 calories per day. She had been suffering from severe chronic arthritis for 2 years, and was prac- tically helpless. A special nurse was assigned to the case, and both the nurse and the patient cooperated well in carrying out directions. Alto- gether the results of the study of this case were very satisfactory, but in some instances specimens of urine were unavoidably lost due to the condi- tion of the patient; the days on which these losses occurred are marked in the table with an interrogation point. The patient ate the entire amount of the diet provided at all times, and the diet furnished maintained the body weight throughout the experiment. It was possible to continue the base line diet long enough to determine the excretion of acetone caused by it with more accuracy in this case than in any other. Diets which had a lower ratio of carbohydrate to fat than did the basal diet caused a formation of larger amounts of acetone, and the change from one level of excretion to the other was gradual and not abrupt. A diet which had a ketogenic power of 108 per cent caused practically no increase in the excretion of acetone; one having a value of 78 per cent caused an excretion of distinctly increased amounts, although these amounts Were not great; while the basal diet—which has a value of 55 per ‘cent—caused an excretion of between 1 and 2 gm. From these figures it would seem that the value of the border-line diet must lie at about 78 per cent, unless considerable importance is attached to the formation of very slight traces of the acetone bodies. In this case small amounts of sodium bicarbonate were fed over a period of a few days after acetone excretion was thought to have reached an equilibrium which corresponded to the basal (10, 10, 80 per cent) diet. The sodium bicarbonate lowered the excretion of ammonia and of titra- —————— 9104998 JO SUIIO} UI possardxo O18 SOTPOd 9104908 OY UO SUOT}RUTUIIOJOp JO S}[NSOY, 3 Deitel eieesloWa st 446 1 Sere Tey “ 11r'0| ¥68 |I97'0| 2 6F O&¢ 669‘T | 9°22] 29 | 29 | SEL | ZO] OF 06 » £9¢°0 60T |9F2 0 | 009 |€62'0 OP If 00s ‘T | 1'9% | 89 | 6:89 | OST OL | S&P 6L €9¢°0 GOT |222'0 | GTS |€62'0 | ZL | OFE Z LG |008‘T | 1:92] 89 | 6'E9 | OST OL | SF 8ST» yw S410] FTE |A4PT 0} 8'G% |S9E'°0 | 90°F | OL E29 |2464‘T | 1°92 | 89 | 2°89 | GFT | ZO | OF LT » cl tTIl'0} #82 |860'0} 0° S@ |801'0 068 PE Zest | 9'9%-| 69 | 29 | 2ST | ZOOL | LP oT » a T61'0 | € 8 [FESO | L OP 00S 0°2¢ |008 ‘T 9% | 29 €9 | OST | S Or | SP BL) 9 l= OL T | 824) 49 | 619] GFT | SOT) SF bo» ve L6h'T | 818) 249 | 169] Tar | 96 | ge &I » aa 940°'0 | #62 |OOT'O | €'8E |OOT'O | GFT | 092 8'9¢ [008 ‘T 96; 29 v9 | OST OL | SP ol» gc00| 82 {S800 | OSS |88I'0 | 06°2 | 002 p LG [02'S | 8:06 | LIT | 212 | 84t | 08 SP TT 5 fy 440'0 | I' Ib |1800] 8 FF jOST'O | OY'T | OST 8& 092'G | 8°02 | ZIT | 212] S4t | 08 sh OL » m 962° O.| SIS |881°0 | OOF {FEO | 0O'E | OLS GZ0 €¢ | IL | 1 69)} &st | 68 GP Or pos qj GS98'"0 SST |Z9F'0 | 9°98 j099°0 | 00'S | OFS g°99 (020% | SET | C2 | 8°22.) S2r L8 cP 8» S OST OIF 0980 | 908 |SLZF'0 | 98'E | OFF 9€ 866 ‘T PL | 2 LL | OLT | 0°6 SP a | i 08% TLE |#S2'°0 Gol 029 #99 |7c0'% | 281} 24 | LL] 9ZT | 88 GP 9 » 3 63 °% SSP |St9'0 6cT j00S'0 | 00° | 00g 9€ 980'2 | 8S | 22 | 8242] PAE | 68 SP ria = 9F SG LOE |Fo6' 0 OZT |98Z°0 | 049 | OLL Z9G |6L8‘T | 9°6 cP | 8°08 | ZOT | 96 SP ee) =) Cr I SLT |S&Z'0 | 0:26 |88F'0 | 02'S | 008 1OL‘T | GOL | SF 62 | GFL | GOL | oF € » a Lo SST |889'0 | 0 €8 |8IS'0 | 022 | 0&8 0'L¢ 068 ‘T | $6 cP T8 | O24T | $6 SP Go» - veL0}| 268 |F6E0} 38h |989'0 | OTS | 028 686 ‘T | 1°6 or | 8 [le ter | ito OF T ‘dsny TZ 360°0| 921 |tz0°0| 26 {9080 | go's | O22 &P | &' Lg |910‘Z | 6'8 Gh | 228] PST | 68 cP LS) fat 0200; 62 |S000; 80 |FI€'0 069 é é é é é é é og Aine “mB \0 907 /0u) “mb |9099,/7 Ou) *wb “ud ne) “ULUL “By quaa wad | "mB \quaawad) "mb |quarwad| “mB Teel "pie Ady eel Rae dl N*HN a AC anu n the sale) ‘oy wupAYoqieg a | “M1001 IVT FY OM UBC ‘out tet qo "A aspg ‘A UITaVL Ee re oem 5 392 Studies on Acetonuria table acid, diminished the degree of acidity of the urine, and increased the tension of carbon dioxide in the alveolar air. The amount of the acetone bodies and their concentration in the urine were increased while the subject received the drug, and returned to the level previously established after it was discontinued. The periods before and after the alkali was given were short, but the changes were so marked that the experiment probably indicates a real increase in the excretion of these compounds. The results are similar to those recorded by Joslin’ and by Forssner (1911). As this was the only case in the series in which the effect of the administration of CALI870 CAL 2022 CAL 2190 are a PR 10% 459" PRY Yo 454m. - PRB %459™ PR 104%4 CIETA ae rosea DIET B MD awe tase DIET C EZ24 eu 21 24117 DIET D BS} cy 15.6768 | FAT 80% 162 FAT 17 % 169 PAT 71% 170 FAT7S % (34 JULY UST ox 12345678 9 10M 12 13 14 15 16 17 18 194 20 fees frees foam | be) = ET STIS. MOEA RSS | BRR . _) + | aed sea SE Bee occas a 7 TOTAL ACETONE SERBBrN/. Ae [ aed: bela a 4 Oa gn aft hoes PERCH Cuart 5. alkali was studied, it is impossible to do more than note what may be an accidental finding. Case V, Miss M. G., was a woman 22 years old, 5 ft. 23 in. in height, who weighed 127 lbs. Her basal metabolism measured 1,430 calories per day. This experiment was much less satisfactory than those described above; the patient failed to eat all of the food provided in any of the diets, and also, apparently, to collect 24 hour urines in a satisfactory manner. 4 Joslin (1917), pp. 394 and 395. ‘0U0499" Jo SUI} UT possardxo d.1v SoIpoq otoj99d" OY} Jo SUOTyRUIWIO4ap Jo S}[nsoy ae ae es ee 393 gs¢'0 | $92 | 26L'0 | TFT 098 ‘T 0c3‘T | #22 | OL | 849] 06 | 8'2t| OF OR: S GLO | LIT | SOTO} ESI | 9TS'0 OFS ‘T 9FZ‘T | 1°22 | 69 G9) 06 | 6 2I! OF Ga tas a 901'0 | 9ST | 2820] OIF 0s9 PSS TOSS | OL 19°99 | 06.) St TF S05 os OrF'0| 28S | 9280] 8 6E 0z8 0sz'T | 22] OL | 849] 06 | Sz | OF LG Ss et Fe 0) =2° 2h |928'0 | °Z' 2g OL €24¢ |0S2‘T | 2] OL | 849] 06 | S:zt| oF 9% %, = 8h 0] FE2 | 2090] G'9¢ 090 ‘T cre T | 9ST | 8¢ | e'¢9| 06 | ToL! Og OS Gs Mea 0.) -e 2S. (F220 | 28°02 008 ‘T 0S3‘T | 2 6T | 09 | 89! 06 91 | os Vor oy low 9F2'0 | 9'FZ | SST°O| 8ST 000 ‘T 0SZ‘T | 2°6L | 09 | 849] 06 9t | og SE. 25 : 1990 | SGP | 9880] 212 GPS ‘T cet ‘T | 2:02 | 09 | 9°99] 98 | ZBL! Zp Ber Sas Fy HOr'0 | F S92 | 2620] SI 066 ‘T 062‘T | 7 6T | 09 | 849! 06 9I | os Te <5 = S8G0"| 9°SZ | F120 | ~ S21 022 ‘T OLI‘T | 1:02 | 69 | I'e9] 28 | sor | GF OR: qi cor | 2&1 | 0610 | 8 St | Ose'0| SOL | oOz'T 7r0'T | 112) g¢ |¢'e9| og | pet | ge 6L 3; ss cz¢ |sor‘t | zer| 9¢ | 249! #8 | ToL! a Sit Bag zr Gcz'0 | ¢'Sz | 0920 | 0'¢z 000 ‘T BATE) 2:80 $9 2590") -o8= |.0° Ord Ge LY. as S 2020 | 2°61 GGO ‘T ooo 1 | 6'SI | 8¢ | #99 | 06 | 2° FL | GF Ot =, S Pe On| =) 61 (8010 | 2°e2 0L8 2261 | 681] 89 | 799] 06 | 2: FL | GF Gl Sy | tea 1 | 6ST | 89 | 799] 06 | 241 | GF Wi » an Meg) 2°02 | 261'0'|. 22s || TZ¢e°0 068 £L4G \0¢%‘T | 261] 09 | 8491 06 91 | o¢ Sh ts on LO) O It \-ccs'0} 0'2F OFZ ‘T 0Sz‘T | 261 | 09 | 849! 06 91 | OS ar Aine “mb = \9900,/6u\ “wb |a9Qg0T/'bu) ‘wo “ub “90 “By quaodad | "wb \quaavad| “wb |juaouad| -wo Ié6l ae oméynqtrospAy-g “OrgooRIp + ou0ja0V | N PFIN ante “QUINOA eo ‘opyerpAyoqieyg “qe “UlayOIg | 972104 ‘ayeqd ‘ou, “q9Iq ‘TA asp TA GTAVL 394 Studies on Acetonuria Although the weight varied somewhat from day to day it was fairly well sustained throughout the period of study. It seems probable in such a case that fat not taken in the diet is replaced in metabolism by tissue fat, although the possibility of an increased combustion of glycogen must also be considered. This case showed acetonuria when the diet had a keto- genic balance of about 80 per cent, and the acetonuria practically cleared up when a diet having a balance of 110 per cent was fed. CH 17 %57 CH 224%70 FAT 67.5% 88 FAT 64.8% 90 ¢——— SS SS SSS SS eS vOvmOODN’' 12 [3 14 15 16 17 1% 19 20 21 2% 23 24 25 26 27 28 29 30 3) CALI2Z00 : CALIZSO DIETA EZA AER 155% ATE pier B BSG {PR 12-8240 gm. suee SEES ORS eebon SCOP ET DERRY TTT ACETONE Fee can ee S| - e LeGnS@ SCcuuRaeem or Cuart 6. Case VI, Mrs. E. Y., was a woman 70 years old, 5 ft. 23 in. tall, who weighed 135 lbs. Her basal metabolism measured 1,185 calories per day. There was a special nurse assigned to the case, but there was a lack of coop- eration because the patient had a prejudice against protein food, and disliked large quantities of fat. This case was even less satisfactory than the preceding one. When the series of experiments was commenced the patient had been living on a diet low in carbohydrate for some time as a part of treatment for chronic arthritis. Acetone was found in her urine by the qualitative test used (Legal’s) when the diets did not seem to be severe enough to cause the presence of the compound, and it seemed desira- R. S. Hubbard and F. R. Wright 395 ble to determine whether there was an increased elimination of both the acetone bodies. There is no doubt that the patient showed such an increase, although no strictly normal values are available for comparison. The diet taken during the period of study contained, on the average 15 per cent of the calories as protein, 17 per cent as carbohydrate, and 68 per cent as fat, and had a ketogenic balance of 120 to 130 per cent. The results are different from those found on other subjects, and may perhaps be attributed to changes in the metabolism of the patient caused by her advanced age. There is one fact which is evident in all of the experiments; when the diet was changed the level of the acetone excretion changed to correspond, but changed gradually. Why these changes should have taken 3 or 4 days in some instances cannot be explained in an entirely satisfactory way. One factor which delayed the response was undoubtedly the time which it took in- gested fat to pass through the digestive, assimilative, and meta- bolic processes, but this did not seem adequate to account for the delay completely. It is possible that when there was a large excess of ketogenic compounds included in the diet, glycogen or other antiketogenic materials were furnished from the reserve supplies of the body in larger amounts than normal. If this was so not only would the changes be gradual, as was found to be the case, but also the amounts of acetone found during the period for which a given diet was fed would be lower than that expected from a calculation of the ketogenic balance of the diet. The data reported above are not sufficient to decide this question. What- ever may have been the cause of this gradual change in the ace- tone excretion, there are three facts which result from it: first, no decision concerning the acetone excretion which corresponds with a diet can be made until the diet has been fed for several days; second, analyses of the fat content of stools is not necessary because it would not be possible to decide to what acetone ex- cretion the figures would apply; third, it is useless to feed the diets—at least to feed the fat content of the diets—in small amounts taken frequently. In interpreting the meaning of the excretion of acetone when diets are fed, which are at or near the border-line of ketogenic antiketogenic equilibrium, there are certain possibilities which must be kept in mind. For instance, the body tissues may fur- nish part of the material burned, and this will be of ketogenic or 396 Studies on Acetonuria antiketogenic nature, as it may be derived from fat, protein, or elycogen. It is probable that a mixture of these is consumed in amounts which will appreciably affect the excretion of acetone when the subject is losing weight, and may so affect it at other times. The composition of the different materials burned may vary with the general nutrition of the patient, and with other factors not understood, factors which are perhaps similar to those which affect the storage of fat. Another fact which must be taken into consideration, especially when slight increases of acetone excretion are studied, is the probable variation in the mixtures of foodstuffs burned in the body at different times during the 24 hour periods. In the study of the effect of diets on the excretion of the acetone bodies dis- cussed here the interpretation has been based on the analysis of 24 hour specimens of urine; the results of these analyses have been discussed as if they represented not only the total, but also the average excretion for the periods. Such an assumption is not correct, because when the ketogenic and antiketogenic com- pounds are present in equivalent amounts, or when the antike- togenic material is present in excess, no matter how great that excess may be, no acetone bodies will be formed, while if the ketogenic material is in excess they will be formed. Although the nature of the results showed that it could not make much difference how the fat in any given diet was fed, it did seem possible, because carbohydrate is more rapidly assim- ilated, that feeding small amounts of this foodstuff frequently would reduce the error described above. In one of the cases in the series the carbohydrate was given at six times during the day, but practically the same amount of acetone was found as when it was given in three meals. It is possible that different combinations of foodstuffs may be simultaneously oxidized in different parts of the organism. If the cells were burning mixtures of foodstuffs which were at or near the border-line of ketogenic antiketogenic equilibrium, it is conceivable that slight disturbances of the blood and nutriment supplied to various parts of the body would lead to a local pro- duction of acetone. A similar explanation has been suggested to account for the rise in blood acetone found after the injection of small amounts of adrenalin chloride (Hubbard and Wright, R. S. Hubbard and F. R. Wright 397 1921). It does not seem probable that this effect can be of as much importance in producing an excretion of acetone as can differences in metabolism at different times during the day, but the possibility that there is some such source of acetone cannot be neglected. If the age of the patient studied in the sixth experiment has any effect upon the formation of the acetone bodies it was probably to exaggerate the effect either of the “local” or ‘“‘temporary”’ production of those compounds. The various uncertain factors which must influence the inter- pretation of the results of these experiments may be summarized as affecting three different parts of the study: one, as influencing the value used to express the ketogenic balance of the diets; two, as rendering uncertain the accuracy with which the excre- tion of the acetone bodies observed corresponds to that which should have been found; and, three the interpretation of the results in terms of ketogenic equilibrium. Uncertainties which affect the value of the ratio include: one, the use of a figure to convert the sum of the antiketogenic factors into terms of fatty acid which is based on the molecular weights of the higher fatty acids; two, the use of the total carbohydrate content of the diet instead of its glucose equivalent in calculating the antiketo- genic compounds; three, the use of the fat fed instead of the fat absorbed for measuring the amount of the ketogenic compounds; and, four, the uncertainty as to the correct percentage of pro- tein which figures as a source of antiketogenic material. Of these uncertainties the first is the only one which would certainly increase the apparent ketogenic value of the expression, the sec- ond and third sources of error would certainly increase its antike- togenic value, while the effect of the fourth is undeterminable. The errors which may have affected the experiments themselves are of various kinds: first, the subjects did not always eat all of the food which was provided; second, they may have eaten arti- cles of food which were not provided; third, enough protein was not eaten in all cases exactly to maintain the subjects in nitrogen equilibrium; fourth, enough food was not taken in all instances to maintain metabolic equilibrium; fifth, 24 hour specimens of urine were not always accurately collected. The first three sources of error mentioned were such as would lead to real or apparent in- creases in the amount of antiketogenic compounds metabolized, THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 398 Studies on Acetonuria as fat was the food which was left untouched, carbohydrate was the food which the subjects most craved, and a negative nitrogen balance was observed more frequently than was a positive one; the effect of the loss in weight and of the failure to collect urines accurately cannot be determined. A detailed examination of these sources of error has shown that either they could not be wholly avoided, or that their influence on the results was slight. The difficulties which affect the interpretation of the excretion of the acetone bodies influence the study of the results of the experiments more than do the other uncertainties met with in these experiments. These sources of uncertainty include: temporary production of the acetone bodies due to variations in the food- stuffs burned at different times during the day; local production of the acetone bodies caused by variations in different parts of the organism; and, possibly, the effect of glycogen drawn from the re- serve stores of the body. All of these except the last would lead to a production of acetone greater than the composition of the diets would indicate. In interpreting the value of the expression 1.5 (weight carbohydrate + 25 per cent weight protein) ous 95 per cent weight fat = N per cent which will express the condition of ketogenic antiketogenic equili- brium, the effect of these uncertainties, particularly of those af- - fecting the interpretation of small amounts of the acetone bodies, must be kept in mind. When the value of the expression was 100 per cent or more, acetone was not found in the urine except in very small amounts, and in two of the cases studied, the excre- tion decreased when the diets had this value to the normal level. In one other case such a diet failed to cause the appearance of a distinctly increased acetonuria, although the period of study (3 days) was perhaps not long enough to produce an equilibrium in the body. When diets were fed which gave numerical values between 55 and 60 per cent rather large amounts of acetone were excreted; there was a distinctly increased excretion also, except in Case II, when diets giving values of about 80 per cent were taken. It seems most reasonable to attribute the small amounts of acetone found on the diets which figured at 100 per cent to local and temporary production of the acetone bodies, and to conclude that values of R. 8. Hubbard and F. R. Wright 399 80 to 90 per cent approximately represent the diet in which the ketogenic and antiketogenic foods are present in equivalent amounts. It has been shown already that certain numerical values of the expression when ketogenic equilibrium is attained correspond to the different possible antiketogenic effects of the glycerol radi- cle: if glycerol does not figure as a source of antiketogenic com- pounds the value is 100 per cent; if glycerol is converted into glu- cose, and this glucose takes part in the reaction between ketogenic and antiketogenic compounds, the value is 83 per cent; if glycerol takes part in the reaction as a three carbon atom residue, the value is 67 per cent. A comparison of these values with the one which has been found experimentally to correspond with the con- dition of equilibrium makes it seem most probable that the gly- cerol residue of the fats does figure only to the extent to which it can yield glucose. These conclusions would be expressed mathe- matically as follows: 1.5 (weight carbohydrate + 25 per cent weight protein) 95 per cent weight fat = 83 per cent 100 X If this equation is transposed so as to express the amounts of pro- tein, fat, and carbohydrate which should be fed to. produce a condition of ketogenic equilibrium, the expression becomes: 1.9 (weight carbohydrate + 25 per cent weight protein) = fat. This expression is practically identical with that stated by Wood- yatt (1921):5 “2 X carbohydrate + protein = fat.” It is of course possible that too much stress has been laid upon “temporary” and ‘‘local”’ sources of traces of acetone, and that not enough emphasis has been placed upon glycogen as a source of antiketogenic compounds. If very small amounts of acetone result from an excess of antiketogenic material in the diet, 100 per cent probably represents the condition of equilibrium. In this case the expression would be: 1.5 (weight carbohydrate + 25 per cent weight protein) 95 per cent weight fat = 100 per cent 100 < and the expression for the relative amounts of food would become: 5 Woodyatt (1921), p. 133. ea 400 Studies on Acetonuria 1.42 (weight carbohydrate +25 per cent weight protein) = weight fat. This expression probably does not express the condition of keto- genic equilibrium correctly ; the one given above is almost certainly preferable. It seems reasonable to conclude from the experiments reported that the two and three carbon atom residues from the a-amino- acids do not figure directly in the antiketogenic reaction, but are condensed to glucose. If these residues did react with the keto- genic compounds the numerical value for each diet would be higher than they are reckoned here; acetonuria would develop and clear up at values of from 100 to 120 per cent, and traces of acetone would be found in some cases when the value was 150 per cent. The charts and tables recording these experiments have been examined for evidence of an adaptation of the organism to these diets high in fat with a consequent reduction of the amounts of the acetone bodies excreted. Folin and Denis (1915) have re- ported evidences of such an adaptation to starvation in three obese women studied by them, but there did not seem to be such a re- sponse to diets high in fat. When the basal diet—containing 10 per cent of the calories in the form of protein, 10 per cent in the form of carbohydrate, and the balance in the form of fat—was resumed after periods during which diets containing relatively more or less fat was fed, the excretion of acetone returned to the level first established if the base line diet was continued over a sufficient period. The method adopted of plotting the concentration of the acetone bodies upon paper ruled with logarithmic characteristics shows clearly the relationship between the two fractions of the acetone bodies discussed in an earlier paper (Hubbard, 1921). When large amounts of the acetone bodies were excreted the acetone from 6-hydroxybutyric acid was in excess of that from preformed acetone plus acetoacetic acid, but when the concentrations were only slightly increased the two fractions were as a rule nearly equal; in some cases the acetone from preformed acetone plus acetoace- tic acid was in excess. When acetonuria developed slowly it was this fraction which increased first, while the 8-hydroxybutyric acid increased later. The interpretation of these facts is com- plicated by differences in the kidney thresholds of the different acetone bodies. R. 8. Hubbard and F. R. Wright 401 Other results which have not yet been discussed include changes in alveolar carbon dioxide tension, the excretion of ammonia and of titratable acid, and changes in the reaction of the urine. The urinary ammonia roughly paralleled the acetone bodies except when sodium bicarbonate was added to the diet; during that per- iod the excretion of ammonia was markedly reduced, while that of the acetone bodies was somewhat increased. The alveolar carbon dioxide tension was somewhat lowered by the more extreme diets, and the values returned to normal when sodium bicarbon- ate was taken. The variations of the titratable acidity and hy- drogen ion concentration were little greater than those which are normally found; these were, of course, markedly affected by the administration of the alkali. CONCLUSION. A method has been suggested for expressing the ketogenic bal- lance of any diet mathematically. A series of six experiments has been described in which the effect of diets high in fat on the excretion of the acetone bodies by normal subjects was studied, and the results compared with this mathematical expression. From the results obtained the following conclusions have been drawn: (1) that the mechanism which controls the formation of increased amounts of the acetone bodies can be regarded as a molecular reaction or balance between ketogenic substances such as the fatty acids and antiketogenic substances such as glucose; (2) that protein figures as an antiketogenic compound only to the extent of the glucose which it can yield in the organism; (3) that glycerol, when fed as a part of the fat molecule figures as an antiketogenic compound only to the extent to which it forms glucose in the organism; and (4) probably that glycerol so fed does figure as an antiketogenic compound to the extent to which glycerol itself can yield glucose. Our thanks are due to Dr. Philip A. Shaffer for suggestions offered and for encouragement extended during the progress of the work described. 402 Studies on Acetonuria BIBLIOGRAPHY. Acree, 8S. F., Mellon, R. R., Avery, P. M., and Slagel, E. A., J. Infect. Dis., 1921 x1 Benedict, F. G., Boston Med. and Surg. J., 1918, elxxviil, 667. Folin, O., Laboratory manual of biological chemistry, New York and London, 1916. Folin, O., and Bell, R. D., J. Biol. Chem., 1917, xxix, 329. Folin, O., and Denis, W., J. Biol. Chem., 1915, xxi, 183. Folin, O., and Denis, W., J. Biol. Chem., 1916, xxvi, 473. Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. Forssner, G., Skand. Arch. Physiol., 1909, xxii, 349. Forssner, G., Skand. Arch. Physiol., 1911, xxv, 338. Fridericia, L. S., Hospitalstid., 1914, lvii, 585. Hubbard, R.S., J. Biol. Chem., 1921, xlix, 357. Hubbard, R. S., and Wright, F. R., J. Biol. Chem., 1921, xlix, 385. Joslin, E. P., The treatment of diabetes mellitus, Philadelphia and New York, 2nd edition, 1917. Joslin, E. P., Diabetic manual, Philadelphia and New York, 2nd edition, 1919. Lusk, G., The elements of the science of nutrition, Philadelphia and London, 3rd edition, 1917. Marriott, W. McK., J. Am. Med. Assn., 1916, Ixvi, 1594. Morris, J. L., J. Biol. Chem., 1915, xxi, 201. Pemberton, R., Am. J. Med. Sc., 1917, cliii, 678. Pemberton, R., and Foster, G. L., Arch. Int. Med., 1920, xxv, 243. Pautton, E. P., Brit. Med. J., 1915, ii, 392. Shaffer, P. A., J. Biol. Chem., 1921, a, xlvii, 433. Shaffer, P. A., J. Biol. Chem., 1921, b, xlvii, 449. Woodyatt, R. T., Arch. Int. Med., 1921, xxviii, 125. THE RESOLUTION OF HYDROXYASPARTIC ACIDS INTO OPTICALLY ACTIVE FORMS. By H. D. DAKIN. (From Scarborough-on-Hudson.) (Received for publication, December 7, 1921.) In a recent communication the synthesis and separation of two inactive forms of hydroxyaspartic acid have been described (1). The inactive form, more soluble in water, gave mesotartaric acid on treatment with nitrous acid and was designated as the anti compound, while the less soluble form gave racemic acid under similar conditions and was named the para compound. Each of these inactive acids contains two dissimilar asymmetric carbon atoms and should be resolvable into active components, giving a total of four active and two inactive forms. The resolution of the anti acid was readily effected by means of alkaloids as de- scribed in the present communication, but the para acid could not be resolved by this method although its finely crystalline alka- loid salts were subjected to exhaustive fractional crystallization. It appears that the alkaloid salts of the para acid are partially racemic compounds of the type described by Ladenburg (2). On _ turning to alternative biological methods for the resolution of the para acid it was found that no resolution could be effected by growing Penicillium glaucum in solutions of the sodium salt while some rather inconclusive evidence was secured of a slight resolu- tion by fermenting yeast used according to Ehrlich’s method (3). The small amount of dextro-rotatory acid thus obtained gave dextro-tartaric acid on treatment with nitrous acid. Since the Walden inversion rarely occurs with nitrous acid it is probable that d-hydroxyaspartic acid and d-tartaric acids are similarly constituted, and the same would be true of the levo forms. Both active forms of anti-hydroxyaspartic acid give inactive mesotar- taric acid on treatment with nitrous acid so that their relative configuration remains undecided. On heating either of the active anti acids with water at 125° partial conversion into the 403 404 Resolution of Hydroxyaspartic Acids para acid was effected, but the latter was invariably optically inactive. It is perhaps somewhat surprising that the active anti-hydroxy- aspartic acids should have as high a specific rotation as 12°. In the light of various theories of optical superposition it might be anticipated that a substance derived from internally compensated mesotartaric acid by the replacement of one hydroxyl group with a relative mass of 17 by an almost equally heavy amino group with a mass of 16, would have a vanishingly small rotation but this is evidently not the case. Of course the possibility still exists of a Walden inversion taking place in the action of nitrous acid on hydroxyaspartic acid, but at present the evidence is against such an assumption. Resolution of Inactive Anti-Hydroxyaspartic Acid. The inactive anti acid, prepared as previously described, gives _well crystallized salts with quinine, brucine, and strychnine and resolution may be effected by fractional crystallization of any of them. On the whole the most satisfactory plan is to separate the dextro acid first as strychnine salt and to use quinine for the separation of the levo acid. The morphine, cinchonine, and quinidine salts were not found helpful for purposes of resolution. Strychnine d-Anti-Hydroxyaspartate—10 gm. of anti-hydroxy- aspartic acid were heated on a water bath with 75 cc. of water and slightly less than the theoretical amount of strychnine was ~ added by degrees. The clear solution was then set aside to crys- tallize in an ice box. The salt crystallized readily in laminated plates which were filtered off and washed with 50 per cent aque- ous acetone. The yield of crude strychnine salt is approximately the theoretical amount calculated for the dextro component and was obtained apparently optically pure after two further crys- tallizations from water (50 ec.). The air-dried salt contains close to 4 molecules of water which were removed on drying at 120° over phosphorus pentoxide under greatly reduced pressure. The salt is insoluble in acetone and moderately soluble in methyl and ethyl alcohol. It crystallizes best from water in which it is very soluble when hot, but sparingly soluble at low temperature.! 1 The optical rotations recorded in this paper were observed in a highly sensitive Schmidt and Haensch polarimeter, for the use of which instru- ment I am indebted to The Rockefeller Institute for Medical Research. H. D. Dakin 405 0.1430 gm. air-dried salt lost 0.0183 gm. H,O at 120° = 12.9 per cent. Calculated for C2;H22N202 . G@,H,0;N S 4H.O = 15:0 a es Rotation. ec = 1.0 air-dried salt in water; 1 = 2.2; a = — 0.42° [a}*” = — 19.1° d-Anti-Hydroxyaspartic Acid.—The strychnine salt obtained as just described was dissolved in hot water and the bulk of the strych- nine precipitated with a slight excess of ammonia. The fil- trate was then extracted repeatedly with a mixture of amy] al- cohol and ether until the remaining trace of strychnine was re- moved. The solution was then concentrated under diminished pressure to a small bulk (15 ec.) and rendered just acid to Congo red by the cautious addition of dilute nitric acid. On allowing the solution to stand in a cool place the dextro acid separates out readily in the form of transparent thick wedge-shaped prisms which only become opaque after long standing. For analysis the acid was dried over phosphorus pentoxide at 60°. 0.1656 gm. substance : 0.1971 gm. COz, 0.0724 gm. H,O. C.H,;O;N. Caleulated. C 32.2, H 4.7. Found. aie ae 7 to Rotation. c = 2.0 in water; 1 = 2.2; @ = +0.53° [oly = +12.1° The dextro acid is slightly less soluble in water than the inac- tive acid, dissolving in about 45 parts of water at room tempera- ture compared with about 30 parts for the latter. The chemical properties of the active acid as expected closely resemble those of the inactive acid. An interesting fact was noted that on re- crystallizing a mixture of the dextro and inactive acids from water, the mother liquor which was at first dextro-rotatory, on standing in contact with the separated crystals gradually became entirely inactive while the separated dextro acid increased in amount. It is inferred that the inactive acid is a dl-mixture at any rate in solution at room temperature and not truly racemic, since under these circumstances the shifting of the equilibrium is readily com- prehensible. The yield of pure dextro acid from 10 gm. of the inactive compound was 3.2 gm. A slightly larger yield may be obtained by separating the acid as lead salt rather than by direct crystallization, but the optical purity of the product is apt to be impaired. The optical rotation of the dextro acid is increased about 30 per cent on addition of hydrochloric acid. Most of the soluble salts are also dextro-rotatory. 406 Resolution of Hydroxyaspartic Acids Quinine l-Anti-Hydroxyaspartate—The mother liquor from the strychnine salt of the dextro acid may be utilized conveniently for the preparation of this salt although it may also be obtained direct from the inactive acid. The strychnine mother liquors are precipitated with ammonia, and a crude levo acid obtained by neutralizing the concentrated filtrate with nitric acid using Con- go red as indicator. The acid crystallizes readily on keeping in a cool place. The crude levo acid (6 gm.) was suspended in 75 cc. of water and quinine base (15 gm.) added by degrees while heating on the water bath. The clear solution on cooling quickly deposited long needles of the salt which were filtered off on the following day. The yield of air-dried salt was 12.7 gm. and its specific rotation was — 96.4. On recrystallization from water (20 cc.) the rotation was prac- tically constant at —95.5. The salt contains close to 4 mole- cules of water of crystallization. 0.1830 gm. air-dried salt lost 0.0238 gm. H,O at 110° = 12.9 per cent. Calculated for CooH,4N202*C4H7O;N°4H2.0 =138.2 “ “ Rotation. c = 1.0 air-dried salt in water; 1 = 2.2; a = —2.10° [aly = —95.5° l-Anti-Hydroxyaspartic Acid—The quinine salt above de- scribed was dissolved in hot water and most of the quinine pre- cipitated by a slight excess of ammonia. The filtrate was re- peatedly extracted with chloroform to remove the remaining alkaloid and then concentrated under diminished pressure to about 15 ec. Dilute nitric acid was then added until the reaction was just acid to Congo red and the free acid crystallized readily. A single recrystallization from hot water gave an optically pure pro- duct. The acid crystallizes in wedge-shaped prisms, soluble in about 45 parts of water at room temperature and save for its sign of rotation it has properties identical with those of the dex- tro acid. For analysis it was dried at 60° over phosphorus pentoxide. 0.1747 gm. substance : 0.2051 gm. COs, 0.0781 gm. H.0. CyH;0;N. Calculated. C 32.2, H 4.70. Found. $632.0, '* 4g, Rotation. c = 2.0 in water; 1 = 2.2; a = —0.52° [a] = —11.9° H.. D:. Dakin 407 Action of Nitrous Acid on d- and I- Anti-Hydroxyaspartic Acids.— In each case 0.5 gm. of the active acid was dissolved in 20 ce. of water together with 0.5 cc. of concentrated hydrochloric acid. Silver nitrite (0.7 gm.) was added by degrees in the course of 24 hours. The mixture was allowed to stand for a further day after which the reaction was complete. Silver chloride was then fil- tered off and the filtrate concentrated to 15 cc. On examina- tion in a 2.2 dm. tube the solutions were found to be absolutely inactive and on addition of ammonia and calcium acetate a large yield of the characteristic calcium mesotartrate was at once ob- tained. It was evident that no trace of the active tartaric acids had been produced. Experiments on the Resolution of Inactive Para-Hydroxyaspartic Acid. Strychnine Para-Hydroxyaspartate—3 gm. of the para acid suspended in 25 to 30 cc. of water were heated on the water bath with the gradual addition of 6.6 gm. of strychnine. Thesaltcrys- tallized very readily in regular prisms and was filtered off and recrystallized twice more from 20 cc. of water. The air-dried salt contained 3 molecules of water of crystallization and its specific rotation was constant at —23.2°. The yield of thrice crystallized salt was close to 50 per cent. 0.1404 gm. lost 0.0140 gm. H,O at 110° = 10.0 per cent. Calculated for Co1Ha2N202*C4H70;N -3H2O0 = il0)si oe Rotation. ¢ = 1.0 air-dried salt in water; 1 = 2.2;a@ = —0.51° la] = —23.2° On decomposing the strychnine salt as already described for the anti compound, a good yield of the para acid was recovered. It was absolutely inactive in a saturated solution, as was also the mother liquor from which it had crystallized. On treating the mother liquors from the strychnine salt in the same way, the in- active acid was again recovered while the mother liquor which still contained a minute trace of strychnine had an angular rotation of only —0.04° in a 2.2 dm. tube. From these results it is clear that no resolution had been accomplished by the fractional cry- stallization of the strychnine salt. Ee CR eT am ae i 408 Resolution of Hydroxyaspartic Acids Cinchonine Para-Hydroxyaspartate-——The para acid (2.5 gm.) and cinchonine (5 gm.) were dissolved in 30 cc. of water. A separation of exceptionally fine clear hard hexagonal prisms readily occurred on cooling. The first crop of crystals (6.0 gm.) was re- crystallized twice more from water (20 cc.). The air-dried salt, representing about 45 per cent of the theoretical amount, con- tains rather more than 2 molecules of water of crystallization. 0.1152 gm. lost 0.0101 gm. H,O = 8.77 per cent. Calculated for C19H22N20 *CsH;0O;N -2H.O = 7.66 “ “ Rotation. ec = 1.0 air-dried salt in water; 1 = 2.0;a = +2.45° [a] = +122.5° On decomposing the cinchonine salt as already described, an absolutely inactive para-hydroxyaspartic acid was recovered. — The mother liquors from the cinchonine salt on similar treatment gave exclusively the inactive acid showing that no resolution had been accomplished. Brucine Para-Hydroxyaspartate—This salt which is extremely soluble in water was crystallized from alcohol. The acid (2.5 gm.) was neutralized with brucine (7.5 gm.) in hot aqueous solu- tion which was then evaporated to a syrup and stirred with about 4 volumes of absolute alcohol. The salt separates out readily in fine thin plates containing 4 molecules of water of crystallization. The anhydrous salt is extremely hygroscopic and was dried over phosphorus pentoxide at 110° under reduced pressure. > 0.3000 gm. air-dried salt lost 0.0355 gm. H.O 11.8 per cent. Calculated for C2;3H2sN204* C4H,O;N -4H,0 = 11.7 “ “ Rotation. ec = 2.0 air-dried salt in water; 1 = 2.2; a = —1.03° lel} = —23.4° The recrystallized brucine salt gave on decomposition only the inactive para acid and the same was recovered exclusively from the mother liquors, showing that no resolution. had been effected. Quinine Para-Hydroxyaspartate-—This salt readily crystallizes in masses of fine felted needles on warming 2.3 gm. of the acid with quinine (5.8 gm.) in 30 to 40 cc. of water. The salt was crys- tallized from water four times, but the acid recovered either from the salt or mother liquor was optically inactive. The air- dried salt retains 2 molecules of water of crystallization. The H. D. Dakin 409 anhydrous salt obtained by drying under reduced pressure at 120° is very hygroscopic. 0.1537 gm. air-dried salt lost 0.0108 gm. H.O = 7.02 per cent. Calculated for CapHasN Oz be C,H;0O;N . 2H.0 = 7.09 ce ds Rotation. c= 1.0 air-dried salt in water; 1 = 2.2;a@ = —2.55° [a] = —116° Action of Penicilliwm on Para-Hydroxyaspartic Acid.—A gram of the acid was converted into the monosodium salt by neutraliz- ing with sodium hydroxide, using litmus as indicator. The solu- tion was then diluted to 200 cc. with an inorganic nutrient solu- tion and heavily sown with Penicillium glaucum. The solution contained in a beaker was loosely covered with a clock-glass and no special precautions were taken to exclude infection with other organisms, since it appears that in many cases a satisfactory reso- lution is more often accomplished with mixed cultures than with a single pure organism. An excellent growth of the mold was obtained and at the end of 3 weeks the solution was just acidi- fied with acetic acid, boiled with a little charcoal, and filtered. The filtrate was entirely inactive and on recovering some of the unchanged acid by means of the lead salt, it also was found to be inactive. No resolution had been affected therefore. Action of Fermenting Yeast on Para-Hydroxyaspartic Acid.— These experiments were made substantially in accord with Felix Ehrlich’s (3) excellent method for the resolution of inactive amino- acids. 2 gm. of the para acid were converted into the monoso- dium salt and then mixed with a solution of cane-sugar (50 gm.) in water (500 cc.). 50 gm. of carefully washed yeast were then added and fermentation was allowed to proceed to completion which took 4 to 5 days at 28°. The filtered solution which contained no sugar was then concentrated under reduced pressure and the unchanged acid recovered by precipitation with lead acetate in neutral solu- tion. The lead precipitate was filtered off, well washed, sus- pended in hot water, and decomposed with hydrogen sulfide. The bulk of the acid recovered, which was about 0.8 gm., was opti- cally inactive and readily crystallized. The mother liquor, how- ever, had a definite rotation of + 0.19° in a 2.2 dm. tube and apparently represented a weak solution (less than 1 per cent) of the dextro and inactive acids On treatment with hydrochloric 410 Resolution of Hydroxyaspartic Acids acid and silver nitrite the rotation of the solution was not abolished as was the case with the active anti acids but was slightly in- creased (+0.24°). On concentrating the solution and adding potassium acetate, a separation of acid potassium tartrate was easily obtained. On dissolving the washed crystalline tartrate in dilute hydrochloric acid, a distinct dextro-rotation (+0.11°) was ob- served in a 2 dm. tube and this rotation was increased on addition of Walden’s uranium reagent. The results of this experiment, which were confirmed by repetition, made it appear probable that a small amount of resolution of the inactive para-hydroxyaspartic acid had been affected and that the levo component was -prefer- entially utilized by the yeast. Furthermore, it appears that the d-amino-acid gives d-tartaric acid on treatment with nitrous acid. The amount of resolution effected by yeast was insufficient to hold out much hope of isolating the pure active acid by its action. Addendum. The Probable Absence of the Hydroxyaspartic Acids in Casein.— The isolation of the two inactive and some of the active forms of hydroxyaspartic acid referred to in the preceding paper made it possible to search more intelligently for this amino-acid among the products of protein hydrolysis. A careful examination of the products from casein has failed to reveal its presence and fur- nishes additional evidence for the rejection of Skraup’s state- ments to the contrary? Owing to the relative instability of hydroxyaspartic acid to the prolonged action of acids, the casein was hydrolyzed by tryptic digestion over 5 months. The neutral monoamino-acids were removed by extraction with butyl alcohol and the residue was then precipitated with lead acetate in neutral solution. The lead precipitate was decomposed with sulfuric acid and the fil- trate warmed with excess calcium carbonate in order to remove most of the phosphates. The acid calcium salts of hydroxyas- partic acid are fairly soluble and should be formed in the filtrate if present. The filtrate was concentrated to about 200 ec. and the excess of barium hydroxide added with a view to obtaining the insoluble neutral barium salts of hydroxyaspartic acids. The * Compare Dakin, H. D., J. Biol. Chem., 1921, xlviii, 273. H. D. Dakin 411 precipitate on decomposition with sulfuric acid was found to con- tain a little organic phosphorus but very little (0.11 gm.) amino nitrogen. No hydroxyaspartic acid could be induced to crystal- lize even after ‘“‘seeding,’’ nor could its phenylisocyanate deriva- tive be obtained. It is concluded that no significant amount of hydroxyaspartic acid is formed by the tryptic digestion of casein. BIBLIOGRAPHY. 1. Dakin, H. D., J. Biol. Chem., 1921, xlviii, 273. 2. Ladenburg, A., Ber. chem. Ges., 1898, xxxi, 524. 3. Ehrlich, F., Biochem. Z., 1906, i, 8. OT a ae mp re — 0 _ fe imetctntse* THE HYDROGEN ION CONCENTRATION AND BICAR- BONATE LEVEL OF THE BLOOD IN PNEUMONIA.* By A. L. BARACH, J. H. MEANS, anp M. N. WOODWELL. (From the Medical Service of the Massachusetts General Hospital, Boston.) (Received for publication, December 2, 1921.) Certain studies of the acid-base equilibrium in disease were reported from this clinic a year ago (1). The method employed was in brief the determination of the carbon dioxide dissociation (or absorption) curve, and the carbon dioxide content of the patient’s arterial and venous blood, and from these data the construction of the carbon dioxide diagram of Haggard and Henderson (2). The method was described in the first paper (1). The further studies reported here were carried out in an identical manner, except that the Van Slyke (3) blood gas apparatus was used instead of the Henderson (4). The combined method of Van Slyke and Stadie (5) was used. In the present research all arterial or A-points, and all venous or V-points, were placed on the dissociation curve at its intersection with the abscissa represent- ing the carbon dioxide content of the blood as found by analysis. All the curves obtained are shown in Figs. 1 to 7. They are all for blood equilibrated with air and various tensions of carbon dioxide at the patient’s body temperature. Since in venous blood and in some pathologic arterial bloods there is an oxygen unsaturation, the curve obtained for fully aerated blood does not truly represent that of the blood as it exists in artery or vein. The effect of oxygen unsaturation according to * This paper is No. 26 of a series of papers on the physiology and pa- thology of the blood from the Harvard Medical School and allied hospitals, a part of the expense of which has been defrayed by a grant from the Proctor Fund for the study of chronic disease. A parallel and more exten- sive study of the acid-base balance in pneumonia has been carried out at the Boston City Hospital by Drs. Buckman, Adams, and Edwards. These results will be published shortly as a part of this series of papers. 413 414 pH and Bicarbonate Level Haldane and his coworkers (6) and to Peters, Barr, and Rule (7) is to shift the curve upwards. According to Haggard and Henderson (8) this phenomenon does not take place in oxalated blood. The blood used in this research was oxalated, but since Henderson and Haggard’s findings have not been confirmed by others we have here plotted the A- and V-points on the curves for fully oxygenated blood but have also calculated the possible effect of oxygen unsaturation by the formula of Peters, Barr, and Rule (7). The position of the points when so corrected is shown in the several figures by circles containing crosses, the uncorrected points by plain circles. As has been shown by Haggard and Henderson, the position of A- or V-points with respect to a series of radii drawn through the | zero point gives the hydrogen ion concentration of the blood. In the figures in this paper the diagonals for various hydrogen ion concentrations (pH) have been constructed as in the paper of Peters, Barr, and Rule (7), and in Table I is given the pH for the various points, both corrected for oxygen unsaturation, and un- corrected, as read off from these diagonals. The diagonals so constructed, however, are for a temperature of 38°C. Since the solubility of carbon dioxide in blood is one of the factors used in the construction of the pH diagonal, and since the solubility varies with the temperature, the position of the diagonal will also vary with the temperature. The solubility coefficient of earbon dioxide in whole blood according to Bohr (9) is 0.937 at 15°C. and 760 mm., and is 0.511 at 38°C. and 760 mm. If the temperature solubility coefficient curve is a straight line (which may legitimately be assumed for the present purpose), then the solubility coefficient at 40°C. and 760 mm. would be 0.474. Using this coefficient for the construction of a given pH line, we find that what it amounts to is that at 40° a given diagonal represents a pH that is lower by 0.04 than that at 38°. This correction has been made in the table by adding 0.01 to the pH as read from the diagonals for each half degree of temperature elevation above 38°. In the figures it is shown by moving the A- and V-points in relation to 38° diagonals, this being more satisfactory than drawing complete sets of diagonals for each patient’s tempera- ture. The points so corrected are shown by black dots. These black dots then probably represent the true pH of the blood as Barach, Means, and Woodwell 415 nearly as we can measure it by this method, since they are cor- rected both for oxygen unsaturation and for body temperature. The temperature correction is an approximation; it is probably maximal and sufficiently near the truth for the present purpose. The studies presented in this paper deal entirely with the blood of pneumonia patients. In the previous paper (1) the diagrams for the bloods of three pneumonia patients were shown. We felt at that time that there was some evidence, from the position of the A-points, of an increase in the hydrogen ion concentration of the blood of these patients. Peters has criticized this conclusion on the basis that the effect of oxygen unsaturation was not taken into account and that, if it were, the hydrogen ion concentration of the three bloods under discussion would be found quite normal. Whether this is a valid criticism or not will depend on who proves correct in the controversy over the effect of oxygen unsaturation on oxalated blood. We will not enter this controversy now. The object of the present research was by further study to try to dis- cover whether any change in hydrogen ion concentration or level of blood bicarbonate occurs in pneumonia. In our figures and table the pH, uncorrected, corrected for oxygen unsaturation, and corrected for both oxygen unsaturation and body temperature may all be found. In this communication we will present for discal seventeen carbon dioxide diagrams of the bloods of ten pneumonia patients. These patients were also made the subject of a study of the effect of oxygen therapy. Their histories have been reported in full in that connection by Barach and Woodwell (10), so that it will be unnecessary to repeat them here; the case numbers used here are the same as those used by Barach and Woodwell. All data except the actual curves will be found in Table I, the curves themselves in Figs. 1 to 7. Ten of the seventeen curves were obtained during the height of the disease and before the patients had received any alkali or oxygen therapy. The remain- ing seven were taken either after crisis or after treatment. The Blood pH in Pneumonia. The matter of the state of the acid-base balance in pneumonia may for convenience be separated into the related matters of blood reaction as indicated by hydrogen ion concentration (pH) it pte Fe — — et gegen gee einen il 22 23 17 13 20 10 21 19 * Case Nos. are the same as in Barach and Woodwell (10). Date. 1920 Nov. 22 SBD Dec. 2 saved 1s 1921 Jan. 4 vee 25 Sore ey Fan 1s Sd eas Feb. 15 Jan. 31 Feb. 2 i 3 Mar. 12 pat.” 1: st AG Diagnosis. Ae Lobar pneumonia. | 3rd “cc ““c * 7th 9th Bronchopneumonia. Septicemia. 7th Lobar pneumonia. 4th Bronchopneumonia. Septicemia. 21st 22nd Lobar pneumonia. 5th Lobar pneumonia. 7th Lobar pneumonia, Pulmonary tuber- culosis. 6th 8th 9th Lobar pneumonia. | 14th Lobar pneumonia. Pulmonary tuber- culosis. llth 7 A = Arterial; V = Venous. t Approximate only. 416 TABLE I—Blooc Oxygen Carbon Carbon g | satura- dioxide dioxide = tion. content. tension. =z eo 2 | & ATU Wah teat Vv A Vv vol. | vol. °c, | Per | Per | ner | per | mm. | mm. cent | cent cent | cent 38 .5|98 .8/66 .0/53 .5/56 3/37 .0 |44.0 39 .5|74. 4/54 8/47 .0/46.4/51.0 |49.5 38 .5|91 6/36 .0/47 .2/53 .7/36.0 |71.5 40.2/91.9 37.8 40.5 402/78 6/65 .951.6/52 .8)51.5 | 39 .5|93 9/77. 1/42 4/43 .8/39.0 39 .0/97 .7|85 .2/44 5/46 .5/24.5 40 3/92 .4 42.0 41.5 36 4/96 .5 50.6 45.0 40 .0/92 5/64. 1/36 6/41 .4/45.5 | 37 .0|96.2 50.0 44.0 391/77 .2|50.2/49 5/53 .1/46.0 [53.0 |7.27/7 39.112 3153.5(60.4)o0 2100. of 0. oto 39 .1/82.2|78.4/66 .7/70.0/60.5 (67.5 7.287 7.23) 7.27 38 .9/81.1 45 .6 46. 37 .9/88.5 48.1 44. oro 39 .0|87 .5|78 .6/40. 4/41 .2/30.0 |31.5 |7.38) | Rete Remarks. tempera- ture. No cyanosis. Recovered by crisis on 7th day. 7 .28|7 .27|7.31) Very sick, pneumonia (both lower lobes). Marked dyspnea. Marked | cyanosis. Very poor quality pulse. 7.32|7.40\7.33| Still very sick. A little less dyspnea and cyanosis. Later developed empyema and died. TPH Nearly moribund. Deep cyanosis (stagnant type of anoxemia). Rapid shallow breathing. Rapid feeble pulse. Died 2 hours after obser- vation. 7 .28|7.30|7.32| Very sick. Delirious. Moderate cyanosis and dyspnea. Died 6 hours after observation. 7 .29|7.30|7.32} Very sick. Stuporous. Extreme hyperpnea. No cyanosis. Consoli- dation of left base, rest of chest dry. .50|7 .53/7.52) Since yesterday given 30 gm. sodium bicarbonate with some relief to respiratory distress. 5 days later developed erysipelas and died. 7.30 Fairly comfortable. Slight cyanosis. Moderate dyspnea. Consolida- tion of right lower lobe. RaA&les at left base. 7.26 Had crisis on 9th day of disease. Now convalescent. 18 days since onset. .13/7.20/7.17| Moderately ill. No great respiratory distress. Slight cyanosis. Con- solidation of right upper lobe, rest of chest clear. 7.33 Recovered by crisis on 8th day of disease. Now well. Today is 25th day since onset. * 30|7 .33/7.32} Very sick. Moderate dyspnea. Marked cyanosis. RaAles in whole left chest. Consolidation of left base and below right clavicle. ” .15|7.16|7.17; Seems in extremis. Outspoken generalized pulmonary edema. ’ 28/7 .32|7.30| Better today. Improvement followed oxygen and alkali therapy. Later recovered from his pneumonia, but at home died of phthisis. 7.26 Severely ill. Moderate dyspnea. Slight cyanosis. 7.28 Crisis yesterday (15th day of disease). Very comfortable today. Slight dyspnea still. Later developed empyema. Operated. Recovered. both lower lobes. Rest of chest clear. Later developed phthisis. 418 pH and Bicarbonate Level and the level of blood alkali. The former is shown by the posi- tion of the A- and V-points. The pH of normal arterial blood as well as one can judge by the literature is not far from 7.35. Peters, Barr, and Rule (7) got this average for fully oxygenated blood of normal persons; their maximum was 7.42 and minimum 7.29. Roughly then, a pH between 7.30 and 7.40 may be considered normal. Now the average arterial pH of our ten sick pneumonia patients was as follows: pH Wncorrected 5.0.3 cas sis le.d8 gays tetniete | ona ey SE eee ere 7.26 Corrected for oxygen ‘unsaturstion: 7.2). 5ie07. 2 ae eee 7.28 Corrected for oxygen unsaturation and for body temperature. 7.31 The range of variation is somewhat greater than in normal arterial blood, the maximum being (in the case of the doubly corrected values) 7.42 and the minimum 7.20. It would seem then on the basis of the corrected values for pH that there is no constant tendency toward an abnormal hydrogen ion concentration of the arterial blood in pneumonia. The average figure of 7.31 cannot be regarded as outside the normal range. Certain individual cases, however, seem to present figures somewhat lower than do any normals; for example, Nos. 14, 21, 22, and 20, with pH of 7.27, 7.26, 7.27, and 7.20 respectively. These results seem to indicate that in certain cases of pneumonia the hydrogen ion con- centration actually is shifted in the direction of decreased alkalinity. Bicarbonate Level. The matter of the level of blood alkali is shown by the level of the dissociation curves. The extreme variations in the levels of the curves of normal persons are not known. In the previous paper a zone was shown within which all normal curves found in the literature (except those of Straub and Meier, 11) fell. This zone is shown again in the various figures of the present paper. Peters, Barr, and Rule (7) show a very similar zone. Their lower border is essentially the same as ours, the upper border a little higher in order to contain the rather high level curves of the subject J. P. The levels of curves of pathologic blood in relation to the normal zone can be seen in the various figures. To Barach, Means, and Woodwell 419 ) 10 20 30 40 50 60 PCO, Fic. 1. Carbon dioxide diagrams in Cases 11, 19, and 23 (lobar pneu- monia), and in No. 22 (bronchopneumonia). In this and in all following figures the arterial points, A, and the venous points, V, are shown by circles. The points corrected for oxygen unsatura- tion ure shown by circles containing crosses, and those corrected for body temperature as well by black dots. The shaded zone indicates the area within which we believe the curves of normal blood should fall. PCO. = pressure of carbon dioxide in mm. of mercury. V CO, = earbon dioxide content of the blood in volumes per cent. This figure shows three cases of lobar pneumonia with essentially normal diagrams, and one of bronchopneumonia with a slight depression in curve level and slight reduction in arterial pH. For further data on these cases and also on those in following figures see Table I of this paper and also the paper of Barach and Woodwell (10). 420 pH and Bicarbonate Level measure actually the height of the curves we can read off the earbon dioxide content either with respect to isohydric points along, let us say, the pH 7.35 diagonal, or perhaps more conven- iently for points of equal carbon dioxide tension, as for example along the 40 mm. ordinate. Peters, Barr, and Rule (7) found that at 40 mm. tension the average carbon dioxide content of all nor- mal subjects of their own and from the literature (except those of Straub and Meier which they excluded for the same reasons as did Means, Bock, and Woodwell) was 49.3 volumes per cent, the maximum 55.9 and the minimum 43.3. The average carbon dioxide content of our ten-sick pneumonia patients at 40 mm. carbon dioxide tension was 43.2 volumes per cent with a maximum of 54.5 and a minimum of 35.0. The pneumonia curves, there- fore, show a lower average level than the normals, but also a greater range of variation. Half of them are essentially within normal limits, half somewhat lower than normal. From this we should conclude that in pneumonia there might either be a normal level of blood alkali or a slightly reduced one. A low level of blood alkali is according to Henderson’s views (2) capable of one of two explanations, first as being due to non-volatile acidosis, second to acapnia. The latter is a condition of blood reaction more alkaline than normal, never of less alkaline; but our pneumonia diagrams show either a normal pH or one shifted in the direction of less alkalinity, hence those which show a low level curve must denote non-volatile acidosis rather than acapnia. The non-volatile acidosis when it occurs is slight; it is nothing like the marked lowering of blood alkali seen in diabetic or renal acidosis (see curves in first paper, 1). To summarize our conclusions to this point we should say that the series of pneumonia bloods as a whole suggests that there may be at the height of this disease either a normal pH and level of blood alkali or that on the other hand in certain cases there may be an acidosis. This acidosis may be an actual non-volatile acidosis (but of slight degree only) as shown by a dissociation curve below the normal level, or it may be simply a carbon dioxide acidosis, that is to say no lowering of the dissociation curve but a change in blood reaction, as shown by the position of the A-point in the acid direction. This latter suggests that sometimes there may be insufficient pulmonary ventilation with a resulting re- Barach, Means, and Woodwell 421 tention of carbon dioxide in the blood to an abnormal concen- tration. The two forms of acidosis may, theoretically at least, coexist. The attempt was made to see whether the acid-base data fur- nished any clue to prognosis. None could be found except that all the fatal cases showed a pH of not over 7.30; but against this must be put the recovery of Case 20 who had the most marked acidosis of all, pH 7.20. Between carbon dioxide capacity at 40 mm. and death or recovery, there was no relation at all. We also looked for possible relationships between pH and ar- terial oxygen saturation, and between carbon dioxide capacity at 40 mm. and arterial oxygen saturation, and again found none. A further study of the individual cases brings out one or two points which may perhaps be of interest. In the first place the situation in pneumonia, if our interpretation of the findings is correct, is that when abnormalities exist they are acidoses, either non-volatile or carbon dioxide, or both. An appreciation of this situation introduces one or two clear-cut indications for treatment. We have recently discussed these indications in their broader aspects elsewhere (12). It will suffice here to point out that an individual suffering from respiratory embarrassment may theo- retically be helped by having the level of his dissociation curve raised. This is true whether the curve is at a low level to start with or at a normal level but with a lowered pH. Raising a curve with a lowered pH may render the pH normal without change in carbon dioxide tension. This will mean that an insuffi- cient pulmonary ventilation becomes efficient without increasing in volume. In other words, as brought out by Henderson and Haggard (13), it requires less ventilation to maintain normal hy- drogen ion concentration at a given rate of carbon dioxide excretion with a high level of the dissociation curve than withalow one. To our minds, with two possible objections which we will discuss pres- ently, it would seem desirable for the pneumonia patient with respiratory embarrassment to have the level of his dissociation curve raised. Effect of Alkali Administration. That the dissociation curve can be raised by the administration of sodium bicarbonate has been proved—the case of the nephritic 422 pH and Bicarbonate Level for example in the previous paper (1) or of Case 10 in the present (Fig. 7), and experimentally by Haggard and Henderson (2). The two objections to alkali therapy which have been raised are these: In the first place, if, as the curve rose in level as alkali was given, the A-point passed to the left of the pH 7.35 diagonal, which it would do if a compensatory fall in pulmonary ventila- tion did not occur, we should be producing an alkalosis which in itself might be harmful. The second objection is one raised by Peters when we first reported this work, and that is that raising the curve and diminishing the pulmonary ventilation in pneu- monia might be harmful because it would produce or increase an existing anoxemia. Whether the phenomenon mentioned in the first objection oc- curs with any regularity we do not know. It did happen to a certain extent in the nephritic already referred to and in Case 17 of the present paper (Fig. 6), in neither instance, however, with any apparent ill etfects that could be attributed to alkalosis. In Case 10, however, a marked rise in the curve occurred with no alkalosis developing. It seems tous that the likelihood of pro- ducing a dangerous alkalosis is slight, particularly if the reaction of the urine is carefully followed and alkali administration stopped at once upon its becoming alkaline. Peters’ objection is a perfectly valid one. Anoxemia does often exist in pneumonia. Decreasing pulmonary ventilation might aggravate it. Giving alkali to diminish pulmonary ven- tilation, therefore, might be the diametrically wrong thing to do in pneumonia. The answer to the objection is, of course, that we should not only raise the dissociation curve by alkali administra- tion but at the same time abolish anoxemia by oxygen admin- istration. That anoxemia in pneumonia can be relieved or abolished by oxygen therapy has been proved in this clinic (10) and by others (14). Arterial and Venous Blood in Pneumonia. Another matter on which the present studies may throw some light is that of the relation between the reaction of arterial and venous blood in pneumonia. Peters, Barr, and Rule have already discussed this relationship in the bloods of normal persons. They state their findings in three normal subjects as follows: Barach, Means, and Woodwell 423 “The CO. tension of venous blood was found to vary between 42 and 72 mm. uncorrected for oxygen unsaturation, 39.5 to 58.5 mm. after cor- rection, with an average of 50.2 mm. The corresponding values for pH were 7.37 to 7.12 uncorrected, 7.40 to 7.22 corrected, with an average of 7.31.” 2 0 ia a6 4 90 40° SC” SY PCO, ; Fic. 2. Carbon dioxide diagrams in Case 13 (lobar pneumonia), I, on the 5th day of the disease and II, 9 days after the crisis. which was on the 9th day of the disease. In this case a curve just b after the crisis. elow the normal limit assumes 4 normal level The effect of oxygen saturation on carbon dioxide-combining the direction of keeping blood reaction nearly power is exerted in hree normals the average corrected arterial constant, of their t 424 pH and Bicarbonate Level and venous pH both being 7.31. Taking out of our ten sick and untreated pneumonia cases the seven in which we have observa- tions on both arterial and venous blood, we find an average pH corrected for oxygen unsaturation of 7.29 for both arterial and venous blood, and corrected for body temperature 7.32. Even in the four most acidotic of these bloods we find an average arterial pH, corrected for oxygen unsaturation and body tempera- ture, of 7.27, and for the venous 7.28. In pneumonia then (even in the presence of acidosis) there is, as there is in normal persons, little if any difference in pH between arterial and venous blood. Effect of the Crisis. In considering those cases of our series in which we have more than one curve, certain interesting features appear. In three cases for example (Nos. 13, 20, and 21) we have observations before and after the crisis. In each of these the second observa- tion showed a higher level of the curve than the first, thus: CO: content at 40 mm. Case No. ee | Riseanlevelloinees Before crisis. After crisis. vol. per cent vol. per cent vol. per cent 13 41.3 48.3 0 20 35.0 53.2 18.2 21 41.3 45.3 4.0 AV OTARG haf. Ce 5.. ONs ee ee ee? See ae TET be The pH of the arterial blood (corrected for oxygen unsaturation and for body temperature) of these same patients before and after crisis was as follows: Arterial pH. Case No. Before crisis. After crisis. 13 7.30 7.26 20 7.20 1.33 21 7.26 7.28 POWOPAG Gs isco a 7.28 7.29 Barach, Means, and Woodwell 425 In these three cases therefore, there was seen following a crisis a return of the blood acid-base balance toward normal in two directions. The available alkali as shown by the curve levels lower than normal before crisis in all three (slightly in Nos. 13 x 2 TO oy? ~ 4 Q “AS: vv rts As As Av y (6) o AAA A Ve q AAR A 4! eb e yo a 4 0 10 20 30 40 50 60 PCO, Fic. 3. Carbon dioxide diagrams in Case 20 (lobar pneumonia), I, on the 7th day of the disease and IJ, 17 days after the crisis, which was on the 8th day of the disease. This case showed a marked rise in curve level from a low position to a high normal position after the crisis and a rise in pH from 7.20 to 7.33. and 21 and definitely in No. 20) in each instance rose to an en- tirely normal level after the crisis. This is most marked in Case 20, which is not surprising as this patient had the most acidosis to start with and also in this case the second observation 426 pH and Bicarbonate Level was 17 days after the crisis, while in No. 13 the second observa- ions was only 9 days after crisis and in No. 21 only 1 day. The hydrogen ion concentration in two of the three cases showed a tendency to move from a less alkaline reaction than normal 9 O2°6 eee VP &% One a a 0 10 20 30 40 56 GO Fea Fia. 4. Carbon dioxide diagrams in Case 21 (lobar pneumonia), I, day before crisis, II, day after crisis. The crisis was on the 15th day of the disease. This case shows a slight rise in curve level the day after the crisis. toward a normal one after crisis. This phenomenon like the other is more noticeable in Case 20, the one most acidotic to start with. The changes undergone by the blood in these three cases before and after the crisis are shown diagramatically in Figs. 2, 3, and 4. aa Barach, Means, and Woodwell » 427 The effect on the acid-base balance of the two therapeutic measures which as suggested above and elsewhere (12), seem to us on theoretical grounds often indicated, either alone or coin- ou 0 g i | 60 t : q.! Wye x yy VWUALL / Ja Pr 0 10? 20% 36" 46° 50°60 PEO) Fic. 5. Carbon dioxide diagrams in Case 14 (lobar pneumonia), I, on 7th day and II, on 9th day of disease. Between Curves I and II he was treated intensively with oxygen. Here a slight rise in curve level and pH followed oxygen therapy but with no crisis. cidently, namely alkali and oxygen therapy, also receive some light from the present research. The outstanding effect of oxygen therapy is, of course, relief of anoxemia. This has been discussed in detail in other papers i — poe ene ee ee ee 428 pH and Bicarbonate Level (10, 12), but it is theoretically at least conceivable that relief of anoxemia may have in itself some beneficial effect upon acidosis. The recovery of normal blood alkali level just described, which Y\ || 0 10 20. 30. 40 505 6G) Fame Fic. 6. Carbon dioxide diagrams in Case 17 (bronchopneumonia), I, on the 21st and II, on the 22d day of the disease. Between the two he received 30 gm. of sodium bicarbonate. In this case a rise in curve level followed alkali administration and the pH was shifted to the alkaline side of normal, an alkalosis was produced but no harmful signs or symptoms of that condition appeared. occurs after crisis, could in the three cases under discussion hardly have been due to oxygen therapy for no one of these received oxy- gen except for a short period. Case 14, however, in which we have two sets of data 2 days apart, showed a rise in curve level Barach, Means, and Woodwell 429 and return of blood reaction from an acidotic to a normal one although he had had no crisis and on the second observation was still very ill though less dyspneic. He eventually died of em- pyema. The results in this case are shown in Fig. 5. From the $:9...9.40) .@ 0 4° sc Te?” Se aa 21 ge 4 ae } | 20 up SO: 40 oy BQiny GOnd GOs Fig. 7. Carbon dioxide diagrams in Case 10 (lobar pneumonia), I, on the 6th, II, on the 8th, and III, on the 9th day of the disease. Between Curves II and III he was treated with oxygen and was given 15 gm. of sodium bicarbonate. A marked rise in curve level without change in pH resulted. time of the first curve on the 7th day of the disease to that of the second curve on the 9th day, the arterial oxygen saturation rose from 74.4 to 91.6 per cent, due, it was believed, to the vigorous oxygen therapy which he received. As to whether the rise in THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 z 430 pH and Bicarbonate Level curve level from 41.8 volumes per cent at 40 mm. to 48.5, and that of corrected arterial pH from 7.27 to 7.40, were in any way the result of oxygen therapy one can only speculate. The problem, however, in that respect is worthy of further study. The effect of alkali administration was studied in two cases. In Case 17 (Fig. 6), between the first and second curves 30 gm. of sodium bicarbonate were given by mouth. Probably as a result of this, for there was no crisis, the level of the curve rose from 42.5 volumes per cent at 40 mm. to 51.3, and the corrected arterial pH from 7.30 to 7.53. In this case an actual alkalosis was produced, but no harmful effects due to that were noted and the dyspnea which had been present at the start was somewhat relieved. In Case 10, shown diagrammatically in Fig. 7, three curves were obtained. The first of these on the 6th day of the disease showed a curve at a normal level and perhaps a very slight carbon dioxide acidosis. On the 8th day of the disease the curve was essentially . the same as to level. No satisfactory A-point was obtained. Between the time Curve II was obtained and the next day when Curve III was obtained he was treated intensively with oxygen and was also given 15 gm. of sodium bicarbonate, the dissociation curve showed a rise of 8 volumes.per cent at 40 mm. and at the same time some relief in respiratory distress. SUMMARY. 1. Carbon dioxide diagrams of the bloods of ten new cases of pneumonia are presented. In three cases observations were secured before and after the crisis, in one case before and after oxygen therapy, and in two eases before and after the administra- tion of sodium bicarbonate. 2. The alkali of the blood in pneumonia as shown by the level of the carbon dioxide dissociation curve, that is to say by the car- bon dioxide capacity at a fixed carbon dioxide tension (40 mm.), was found to be sometimes within normal limits, sometimes some- what below normal limits. The average in the pneumonia group was 43.2 volumes per cent, while in normal persons Peters, Barr, and Rule found the average 49.3 volumes per cent. The lowest observed in pneumonia was 35.0 volumes per cent. ee ase Barach, Means, and Woodwell 431 3. The arterial pH in pneumonia as calculated from the car- bon dioxide diagram and corrected for oxygen unsaturation and body temperature showed an average of 7.31. Four of the ten bloods showed a pH below 7.30, which probably can be considered the lower border of normal variation. The lowest observed was 7.20. 4. Norelation between pH or dissociation curve level and degree of anoxemia or prognosis could be found. 5. In pneumonia patients, as in normal persons, there seems to be little or no difference in pH between arterial and venous blood. 6. In three patients studied before and after crisis there was an increase in curve level after crisis in each instance, and in one showing a marked shift in pH before crisis there was a normal pH after crisis. A rise in curve level and a return to a normal pH in the cases with lowered pH would seem to be among the phenomena that take place at or after the crisis. 7. The same phenomena occurred in one case without crisis but after vigorous treatment with oxygen. 8. In two cases the level of the curve was raised apparently by the administration of sodium bicarbonate, in one instance with the production of a slight alkalosis. 9. It is suggested that in pneumonia patients showing acidosis either in the sense of a low level of available blood alkali or of decrease in pH or combination of the two, the administration of sodium bicarbonate may be helpful by diminishing the work of the respiratory bellows. By such a procedure a pH less alka- line than normal may be brought to normal with no increase in ventilation because of a raising in the level of the dissociation curve. Or in a case with low curve but normal pH to start with, the raising of the curve may diminish the amount of ventilation necessary. The use of sodium bicarbonate should be carefully controlled, however, to avoid the production of alkalosis, and when anoxemia is present should be combined with oxygen therapy. BIBLIOGRAPHY. 1. Means, J. H., Bock, A. V., and Woodwell, M. N., J. Exp. Med., 1921, xxxili, 201. 2. Haggard, H. W., and Henderson, Y., J. Biol. Chem., 1919, xxxix, 163. 3. Van Slyke, D. D., J. Biol. Chem., 1917, xxx, 347. 432 pH and Bicarbonate Level oo or . Henderson, Y., and Smith, A. H., J. Biol. Chem., 1918, xxxiii, 39. . Van Slyke, D. D., and Stadie, W. C., J. Biol. Chem., 1921, xlix, 1. . Christiansen, J., Douglas, C. G., and Haldane, J. S., J. Physiol., 1914, xlviil, 244. . Peters, J. P., Jr., Barr, D. P., and Rule, F. D., J. Biol. Chem., 1920-21, xly, 489. . Haggard, H. W., and Henderson, Y., J. Biol. Chem., 1920-21, xlv, 215. . Bohr, C., in Nagel, W., Handbuch der Physiologie des Menschen, Brunswick, 1909, i, 63. . Barach, A. L., and Woodwell, M. N., Arch. Int. Med., 1921, xxviii, 394. . Straub, H., and Meier, K., Deutsch. Arch. klin. Med., 1919, exxix, 54. . Means, J. H., and Barach, A. L., J. Am. Med. Assn., 1921, Ixxvii, 1217. . Henderson, Y., and Haggard, H. W., J. Biol. Chem., 1918, xxxiii, 333. . Meakins, J., J. Path. and Bact., 1921, xxiv, 79. ANALYSIS AND COMPOSITION OF CORN POLLEN. PRELIMINARY REPORT. By R. J. ANDERSON anp W. L. KULP. (From the Biochemical Laboratory, New York Agricultural Experiment Station, Geneva.) (Received for publication, December 2, 1921.) INTRODUCTION. The literature dealing with corn and corn products is very extensive but we have failed to find in it any reference to the composition of corn pollen. In the process of fertilization and reproduction pollen plays a very important part. It would seem, therefore, that some knowledge of the kind and nature of the chemical compounds occurring in pollen would be of interest to plant physiologists. Different varieties of corn apparently pro- duce pollen which varies greatly in composition. This fact might be of importance in cross-breeding. The present investigation was undertaken in order to contrib- ute some information on the following points: (1) The approxi- mate composition of corn pollen, (2) the principal inorganic constituents of the ash, and (3) the principal organic compounds contained in the pollen grains. Unfortunately, stress of other work has prevented us from com- pleting the investigation, but we wish to publish this preliminary report giving the results which have been obtained up to the pres- ent time because our joint work will be interrupted during the coming year. EXPERIMENTAL. The pollen was gathered in the following manner: Corn tassels were cut off as the pollen sacs were opening. The tassels were spread out in thin layers on clean paper on the floor and allowed to dry. The pollen was then shaken out on clean paper and care- 433 434 Composition of Corn Pollen fully sifted. Many of the pollen sacs did not open and in order to obtain the enclosed pollen it was necessary to break them by rubbing or by passing the tassels through a coarse mill and then carefully sifting out the pollen. Practically all foreign matter was finally removed by sifting through very fine bolting cloth. It is important that fresh pollen be spread out in thin layers while drying as otherwise it will undergo very rapid decomposi- tion. One day about 100 gm. of ripe pollen obtained by shaking fresh corn tassels over paper were left over night in a dish which was loosely covered by a watch-glass. The pollen was a heavy yellow powder which appeared to be dry. The next morning it was found to consist of a semifluid gummy mass in which it was impossible to distinguish any individual pollen grains. We were unable at that time to make any investigation of the changes which had occurred and this observation is merely recorded to indicate the rapidity with which pollen may undergo spontaneous decomposition when conditions of temperature and moisture are favorable. In 1919 we obtained about 800 gm. of pollen from yellow dent Improved Leaming corn, and in 1920 some 2,000 gm. of pollen were gathered from White Flint Luces Favorite and a smaller quantity from pop-corn. The pollen obtained as outlined above formed a golden yellow, dense powder and had a strong but agreeable aromatic odor. 100 gm. of pollen occupied about 150 cc. of space. Determination of Moisture. On drying at 103°C. the pollen continued to lose in weight slowly and after 48 hours the loss amounted to about 6.5 per cent. All of this loss in weight was not due to loss of water because the dried pollen was dark brown in color and had lost practically all of the characteristic odor. After drying at 100°C. the color darkened and most of the odor was lost. It is evident, therefore, that in drying at these temperatures certain volatile constituents are lost and that some oxidation occurs. In order to obviate such losses of volatile principles and to pre- vent oxidation the pollen was dried for analysis at room tempera- ture in vacuum over sulfuric acid. The loss in weight on drying in this manner was 4.68 per cent and there was no noticeable change either in color or odor. R. J. Anderson and W. L. Kulp 435 Extraction of the Pollen with Various Solvents. A series of extractions was made to determine the amounts of material removed from the pollen by ether and alcohol during ° varying lengths of time. The results obtained are given in con- densed form in Table I. The percentages are all calculated to the original air-dry pollen. TABLE I. Ether and Alcohol Extraction of Corn Pollen. I'ther extraction; Soxhlet method.}| Alcohol extraction following ether extraction. Amount of leohol - Per- : 7 Per- 5 x! Weight Weight extract lg palen of se a Extraction. Rolle of centage soluble in extract. exixaht! extract. external ether, hrs. gm. gm. |per cent hrs. gm. gm. |per cent 8 5 |0.0538| 1.07 4* 5 |1.6166) 32.33} 14.93 24 10 |0.1320) 1.32 24 (Soxhlet) 10 |1.1405) 11.40 48 ‘| 20 |0.3170| 1.58 168 (Soxhlet) 20 (2.9272) 14.63 48 2 |0.0280) 1.40 4 weeks. tf 20 (6.8267) 34.13 48 2 |0.0285) 1.42 *5 gm. of pollen were suspended in 75 ec. of absolute alcohol and boiled under reflux condenser for 1 hour on the water bath. The alcoholic solu- tion was decanted and replaced by fresh alcohol which was in turn boiled for 1 hour. These operations were repeated four times. The alcoholic extracts were united, filtered, and evaporated to dryness and then dried in vacuum over sulfuric acid. + After extracting 20 gm. of pollen with absolute alcohol in a Soxhlet apparatus for 7 days, the pollen residue was suspended in about 100 ce. of absolute alcohol and boiled on the water bath under reflux condenser. The alcohol was decanted and renewed daily until all of the coloring matter was removed and the alcohol remained practically colorless. The time required was 4 weeks. The alcoholic extracts were united, the alcohol was distilled off, and the extract was dried in vacuum over sulfuric acid. It is interesting to note that extracting pollen with four por- tions of boiling absolute alcohol yields almost as high a percent- age of extract as was obtained after completely exhausting the pollen with absolute alcohol during a period of 4 weeks. A more complete extraction was obtained when the pollen was suspended in the alcohol in a small flask and digested under a reflux condenser on the water bath than when the pollen was con- tained in a thimble as in the usual Soxhlet method. / 436 Composition of Corn Pollen The dried ether extract was of a dirty green color and of a rather soft, wax-like consistency. This extract contained only a trace of phosphorus and consequently it could only contain a very small amount of phosphatide. The nature of this fat or wax- like material has not been determined. Extraction of Corn Pollen with Absolute Alcohol and Chloroform. It has been stated by Glikin (1) that the method of Rosen- feld (2) gave high yields of fat and lecithin, particularly in the analysis of animal tissues. This method consists in extracting the material for 3 hour in boiling alcohol and then extracting the residue for 6 hours with chloroform in a Soxhlet apparatus. We employed this method on corn pollen as follows: (a) 5 gm. of pollen after drying in vacuum over sulfuric acid, were placed in an extraction thimble and extracted for 2 hours by immersing in about 50 cc. of boiling absolute alcohol contained in a large test- tube. The thimble and contents were then rinsed with absolute alcohol. The solution was filtered, the alcohol evaporated, and the extract dried to constant weight in vacuum over sulfuric acid. The dry extract weighed 0.3815 gm. or 7.63 per cent of the air- dried pollen. (b) The pollen residue was extracted for 6 hours with chloroform in a Soxhlet apparatus. After evaporating the chloroform and drying as above the extract weighed 0.4215 gm. or 8.43 per cent. (c) The pollen residue was extracted a second time with chloroform for 6 hours. After evaporating and drying as above the extract weighed 0.0538 gm. or 1.08 per cent. (d) The pollen residue was extracted a third time with chloroform for 48 hours. On evaporating and drying as before there was ob- tained a semicrystalline material which weighed 0.0340 gm. or 0.68 per cent. The total yield of extract in the above operations was, there- fore, 0.8908 gm. or 17.82 per cent. These extracts were united and exhausted with absolute ether. After filtering, evaporating the ether, and drying, the ether-soluble extract weighed 0.6940 gm. or 13.88 per cent. The ether-insoluble material which remained was a semicrys- talline solid which was readily soluble in water. Evidently, there- fore, it was neither fat nor lecithin. R. J. Anderson and W. L. Kulp 437 Attention is called to the fact that a higher yield of extract was obtained by digesting the pollen in four successive portions of alcohol as shown in Table I than by the above alcohol-chloroform extraction. The ether-soluble part of the alcoholic extract was 14.93 per cent as against 13.88 per cent of alcohol-chloroform extract. The lecithin content of the ether-soluble portion of the alcoholic extract was also higher than in the alcohol-chloroform extract as will be shown below. But only a small amount of the ether-soluble material was phosphatides as shown by the low phosphorus content. The nature of the non-phosphatide part of the ether-soluble portion of the alcoholic extract has not been determined. It is interesting to note that while the maximum amount of ether extract obtained by direct extraction of the pollen with absolute ether was only 1.58 per cent yet the amount of alcohol- chloroform and the absolute alcohol extracts soluble in absolute ether was from 14 to 15 per cent of the weight of the pollen. It is probable that the membrane surrounding the pollen grains is nearly impermeable to ether while alcohol and chloroform per- meate the membrane and dissolve out fats and phosphatides to- gether with other substances. This assumption would account for the low percentages of ether-soluble material obtained in the direct extraction of the pollen with ether and for the much larger yields of ether-soluble substances in the alcoholic or chloroform extracts. The difficulty of completely extracting the soluble constituents is greater in the case of pollen than in other plant material be- cause it is practically impossible to rupture the pollen membranes by ordinary trituration. Prolonged grinding in a mortar, even after the pollen has been extracted with ether and alcohol, pro- duces only a small percentage of broken cells. Phosphatide Content of Corn Pollen. The absolute ether extract obtained from pollen contained, as previously stated, only a trace of phosphorus. A larger amount of ether-soluble or phosphatide phosphorus was contained in the alcoholic and the alcohol-chloroform extracts. The phosphorus in the alcoholic extract amounted to 0.19 per cent of the pollen. Nearly all of this phosphorus, or 0.139 per oor 2 ~ ee ee ee ee ee a ——— ae I 438 Composition of Corn Pollen cent was soluble in ether. Multiplying this number by the usual factor for lecithin we obtain 3.62 per cent of lecithin in pollen. In the alcohol-chloroform extract the ether-soluble phosphorus amounted to 0.104 per cent which corresponds to 2.72 per cent of lecithin. These figures indicate that the phosphatides are more completely removed from pollen by absolute alcohol than by the alcohol-chloroform treatment. Nitrogen in Corn Pollen and the Nitrogen Distribution in Pollen Extracts. The total nitrogen in the air-dried pollen was 4.30 per cent. 10 gm. of pollen were extracted with ether in a Soxhlet apparatus for 24 hours. It was then extracted with absolute alcohol for 24 hours. The alcohol was evaporated and the extract was taken up in ether as much as possible, filtered, and the ether evaporated. The ether extracts were united and were found to contain 0.14 per cent of nitrogen. The ether-insoluble portion of the alcoholic extract contained 0.18 per cent of nitrogen. The pollen residue after extracting with ether and alcohol was digested in water, filtered, and washed with water. The water- soluble nitrogen amounted to 0.50 per cent. The pollen residue from the above extractions, after drying contained 3.49 per cent of nitrogen. The above results are calculated to the original air-dried pollen. Approximate Composition of Corn Pollen. The figures given in Table II were obtained on analyzing the pollen obtained from three varieties of corn. The results are calculated to the water-free pollen. We do not feel that we can give any adequate reason for the striking differences found for starch and sucrose. Several deter- minations were made in duplicate and triplicate with concordant results. It is not impossible that the difference in composition depends upon varying degrees of ripeness but we tried, as far as possible, to gather all of this pollen when it was just ripe. It is not improbable that different varieties of corn may pro- duce pollens of different composition. However, until more work R. J. Anderson and W. L. Kulp 439 has been done on this subject we would only offer this explana- tion with some reserve. TABLE II. Analysis of Corn Pollen. Yellow dent | White flint corn. corn. Constituent. Improved Luces Polen eae Pollen gath-| Pollen eath-| eed 1920. ered 1919. ered 1920. per cent per cent per cent See ee esto tie. he rae «oes one 11.07 19.04 18.03 EEL Soe 30d divas oh Gale bots.ate dat ae 4.53 4.43 3.85 Reducing sugar as dextrose.............. 3.50 5.38 4.95 REE She's Ses os o% woe cane SA 9.09 2.97 14.18 Reemnneniiee fis oot tis. Gad Cetoacatane 10.60 ee eee 5.8) Crude fat or ether extract (average)..... 1.48 Ne so amin accua a ne CRE 3.46 3.83 Bates (2 2 LES RES Se St SE sat 0.63 SCARE PE hac. oles EEE RG Red, 0.34 Dateeinos.rs. 2 .).2-).2!. ween c astenet. 0.19 PEM MESFSALIV 251, «2551516! 15d, siete paw yb oh sae Sia 1.24 Analysis of the Pollen Ash. For analysis the pollen was ashed at a low temperature in an electric muffle and a pure white ash was obtained. The result of the analysis is given in Table ITT. TABLE III. Composition of the Pollen Ash from Yellow Dent Improved Leaming Corn. Constituent. Per cent. ih Na Ae kD sae ee 18 .92 RpEFERET fo. 55's so SSDs EE ha we bas oleae oe enw ee er 0.69 PPOTEATION (3 oooh 25S cc Spee a Mer din rac ae dita Sirs eps ae beers 0.80 ND go 5x x 05 ech RN a's ald uapaintenehne Ri diag e's 3.76 IRIN TEEN oso osc, 2h Cw sic , Liateare w sith hie. otal See = 1.02 SINE Ss o's nh ce eR gals 2 an hn aw A alee wR 4.60 PPEIECSETITER 325. 2s Ss cl SeRe Rt Rete ie nd se eee cia ole < wR ee ote Se 35.58 ueepba steer 9) 2 0-92 0c/5. ose kok » one Lak Pelee ae cae bis sai 0.69 Uli 27s. eh: i ne, Ane ree me Sees 0.25 EPEUERBTIUIYIN << 5 0-0'<.21 cc IE che bik we jo'w'b uo maya rmleye aie ate ote eters 0.22 440 Composition of Corn Pollen Separation of Certain Soluble Constituents of Corn Pollen. The pollen used in this investigation was obtained from the Improved Leaming variety of yellow dent corn which had been gathered in 1919. It was dried in vacuum over sulfuric acid. Extraction with Ether. The pollen, 590 gm. of dry material, was extracted with abso- lute ether during two 24 hour periods. After evaporating the ether and drying the extract in vacuum over sulfuric acid it weighed 9.0 gm. This is equal to 1.52 per cent of ether extract. The extract was of a dirty green color and of a soft, wax-like con- sistency. It was not further investigated. Extraction with Alcohol. The pollen residue was placed in a 2 liter flask, 1 liter of abso- lute alcohol was added and the mixture was heated to 60°C. under a reflux condenser for 1 hour. It was allowed to stand at room temperature for 15 hours and then heated to 60° for about 5 hours. It was filtered while hot on a Buchner funnel and washed with absolute alcohol. These operations were repeated three times with fresh portions of alcohol. The pollen residue was reserved for further examination. Examination of the Alcoholic Solution. The alcoholic extract was of a greenish yellow color and it measured about 4 liters. On cooling and standing over night in the ice chest there sepa- rated out a small amount of colorless crystals in the bottom of the flask. This material was filtered off and will be referred to later as ‘“‘Substance A.” The alcoholic solution was concentrated in vacuum at a tem- perature not exceeding 40° to about 300 cc. During the evap- oration of the alcohol a considerable amount of a crystalline sub- stance was deposited in the flask and the quantity increased on cooling and standing over night. The crystals were filtered off and washed in alcohol. This material will be referred to later as ‘‘Substance B.” R. J. Anderson and W. L. Kulp 441 The alcoholic solution was now taken to dryness in vacuum. There remained a thick oily substance which was mixed with some crystalline product. This oily residue was shaken with several portions of absolute ether in which the greater amount of the material dissolved. The ether-insoluble, semicrystalline substance was added to ‘‘Substance B” mentioned above. Preparation of the Amorphous Phosphatide. The ethereal solution was evaporated to a syrupy consistency and to it were added with constant shaking 600 cc. of acetone. A heavy sticky substance was precipitated which settled to the bottom of the flask. The dark-colored acetone solution was de- canted and the residue was washed thoroughly with acetone. The acetone solution and washings were concentrated to a thin syrup and again poured into 600 cc. of acetone when a fur- ther quantity of the sticky substance, similar to the first, sepa- rated. After decanting the mother liquor and washing with ace- tone this precipitate was added to the first amorphous phosphatide. Preparation of the Crystalline Phosphatide. The acetone mother liquor was allowed to stand in the ice chest for 2 days. A considerable quantity of nearly colorless, large, thin, plate-shaped crystals separated gradually. The crys- tals were removed, washed in acetone, and dried in vacuum over sulfuric acid. The dry substance weighed 4 gm. This crystalline material was found to be a phosphatide. It will be described later. The acetone solution was evaporated to dryness under reduced pressure. There remained a thick oily brown residue which, after drying in vacuum over sulfuric acid, weighed 26 gm. This ma- terial still contained a considerable amount of phosphatide be- cause it contained 0.55 per cent of phosphorus and 0.4 per cent of nitrogen, but it was not further examined. Purification of the Amorphous Phosphatide. The substance which was precipitated by acetone from the ethereal solution was dissolved in absolute ether. The ether solution was shaken with water and afterwards with a dilute solu- 442 Composition of Corn Pollen tion of sodium chloride. The emulsions which formed were broken up with much difficulty by adding sodium sulfate. The solution was finally dried with sodium sulfate, filtered, and the ether evaporated until a thin syrup remained. This was poured with constant stirring into 600 cc. of acetone. The phosphatide separated as a thick, pasty mass. The acetone was decanted and the phosphatide washed several times by thoroughly stirring with acetone. After drying in vacuum over sulfuric acid it weighed 11.5 gm. The two phosphatide preparations had a combined weight of 15.5 gm. This corresponds to a yield of 2.6 per cent. The amorphous phosphatide after drying formed a light yel- lowish brown, hard, brittle mass which could be powdered. It was not very hygroscopic. For analysis it was dried in vacuum over phosphorus pentoxide at the temperature of boiling chloro- form. Further drying at 78°C. did not cause any loss in weight. There was no perceptible change in color on drying at the above temperature and the loss in weight was only 1.09 per cent. Found. P = 3.86, N = 1.53 per cent. Ratio UNE ae The percentage of phosphorus and nitrogen and the N:P ratio corresponded very nearly to the values required for disteary] lecithin. Hydrolysis of the Amorphous Phosphatide. Without subjecting the substance to any further purification an attempt was made to determine quantitatively the amounts of choline, glycerophosphoric acid, and fatty acids after hydrolysis. In this experiment we followed the method outlined by Osborne and Wakeman (3) in their study of the hydrolysis of the phos- phatide from milk. We used 5.2390 gm. of the dry phosphatide and obtained 0.8519 gm. of choline platinum chloride, 1.2035 gm. of barium glycerophosphate, and 2.2646 gm. of fatty acids. The figures presented in Table IV are calculated from the above values. The choline platinum chloride after recrystallizing from water contained 32.21 per cent of Pt. (CcHi, ON Cl), Pt Cly. Calculated. Pt 31.64 per cent. R. J. Anderson and W. L. Kulp 443 The barium glycerophosphate was purified by precipitating it from aqueous solution with alcohokuntil a pure white amorphous preparation was obtained. The air-dried substance lost 8.30 per cent of water on drying at 105°C. in vacuum over phosphorus pentoxide and the weight remained constant on further drying at 130°C. On analysis the dried preparation gave: Ba = 39.90, P = 9.17 per cent. C;H;OcP Ba + 2H.O. Calculated. Ba = 40.01, P = 9.02 per cent. The analytical results agree with the theoretical composition of barium glycerophosphate plus 2 H.O. But the fact that this water could not be driven off at 130°C. in vacuum makes the purity of the preparation somewhat doubtful. Winterstein and Hiestand (4) obtained a barium glycerophosphate of similar com- TABLE IV. Cleavage Products of Amorphous Phosphatide. Calculated Constituent. Amount found. for disteary]l lecithin. gm. per cent per cent 0 AS re re oes 0.3335 6.36 14.99 Glycerophosphoric acid............ 0.6733 12.85 21.31 WERT CIO Sn. «5 55 5 Sessa eiassivie n« «i 2.2646 43 .22 70.26 position from the phosphatide which they had isolated from wheat flour. MacLean (5) calls attention to the difficulty of purify- ing the glycerophosphoric acid prepared from plant phos- phatides. Some evidence was found of the presence of another base be- sides choline in the phosphatide. After the choline platinum chloride had been filtered off the alcoholic solution was evaporated and the residue was taken up in water. The platinum was pre- cipitated by hydrogen sulfide and the filtrate was evaporated to dryness under reduced pressure. The residue was extracted at room temperature with absolute alcohol which left a small quantity of an insoluble white crystalline substance. This was recrystal- lized from hot 95 per cent alcohol and was obtained in colorless needle-shaped crystals which weighed 0.13 gm. From this sub- stance a gold double salt was prepared which crystallized from } 444 Composition of Corn Pollen water in large yellow needles. It contained 49.84 per cent of gold and melted at 132°C. (uncorrected), but we were unable to identify this substance. On hydrolysis of the phosphatide a mixture of saturated and unsaturated fatty acids was obtained. The percentage of iodine absorbed by the crude fatty acids was 49.01 determined by the Hanus method (6). The fatty acids were saponified and lead soaps were prepared and extracted with ether. The ether- insoluble lead salt was decomposed with hydrochloric acid and ex- tracted with ether. After evaporating the ether, the residue was recrystallized several times from absolute alcohol. The snow- white crystals melted at 63°C. (uncorrected). The melting point of palmitic acid is 62.6°C. and it is probable, therefore, that the saturated acid was nearly pure palmitic acid. Owing to the small quantity of unsaturated acid its nature could not be determined. The values found for the cleavage products of the amorphous phosphatide as indicated in Table IV are very much lower than is required for the formula of distearyl lecithin. This might be due to admixed impurities or possibly to the presence of carbo- hydrates. An attempt was made to determine the amount of carbohydrate present in the phosphatide. After hydrolyzing by boiling with 5 per cent sulfuric acid, as described by Winterstein and Hiestand (7), cooling, and neutralizing with sodium hydrox- ide only a very slight reduction was obtained on boiling with Fehling’s solution. The phosphatide contained, therefore, only a trace of carbohydrate. A complete analysis of the substance was made when it was found that, in addition to carbon, hydrogen, phosphorus, and nitrogen, it also contained sulfur. The results obtained on analysis are given below. Found. C 57.78, H 8.53, P 3.86, N 1.53, S 0.68 per cent. The carbon and hydrogen are much lower than is required for distearyl lecithin and the presence of sulfur would indicate that the substance is a mixture of phosphatide and sulfatide. R. J. Anderson and W. L. Kulp 445 Analysis of the Crystalline Phosphatide. The crystalline phosphatide which separated from the acetone mother liquor was of a slightly yellowish white color. After dry- ing in vacuum over sulfuric acid it was decidedly hygroscopic and on exposure to the air it became sticky. Qualitative analysis showed that it contained phosphorus and nitrogen but no sulfur. On drying at the temperature of boiling chloroform in vacuum over phosphorus pentoxide it lost only 0.93 per cent in weight and the weight remained constant on further drying at 78°C. On analyzing it for phosphorus and nitrogen the following re- sults were obtained. Found. P 1.74, N 1.53 per cent. Ratio N:P = 1.95:1. The N:P ratio is nearly as 2:1 and the substance is, therefore, probably a diaminomonophosphatide. Lack of time and material has prevented a complete examina- tion of the phosphatides of corn pollen. The results obtained so far indicate that at least two phosphatides are present. We hope to prepare more material and will report on a more complete investigation of the corn pollen phosphatides in a later publication. Examination of ‘“‘ Substance A.” It was mentioned previously that a small quantity of a color- less crystalline substance separated when the absolute alcoholic extract of pollen was allowed to stand over night. The crystals were filtered off, washed in a little absolute alcohol, and dried in vacuum over sulfuric acid. It weighed 0.2 gm. The substance was insoluble in water and very slightly soluble in cold alcohol, but readily soluble in ether, chloroform, and in hot alcohol. The substance was twice recrystallized from boiling absolute alcohol from which it separated on cooling in small transparent plates. The dry crystals were snow-white in color and they ex- hibited a fatty feeling to the touch. It gave in somewhat modi- . fied form the Salkowski and Liebermann reaction of cholesterol or phytosterol. The melting point, however, was sharp at 88— 89°C. (uncorrected). The small quantity of this substance pre- vented its identification but the crystal form and melting point correspond to those of myricy] alcohol. ee teen a a = . die ———_ ~ ais bs, 446 Composition of Corn’ Pollen Examination of ‘‘ Substance B.” This substance separated in crystalline form on concentrating the absolute alcoholic extract of pollen. The crystals were filtered off and washed in alcohol. The substance was readily soluble in water and it crystallized again on adding alcohol to the aqueous solution. Through an accident the larger part of the material was lost, but from the small quantity which was saved we obtained after recrystallizing four times 0.75 gm. of beauti- ful colorless needle-shaped crystals. It gave the reaction of Scherer and melted at 221°C. (uncorrected), thus showing that the substance was pure inosite. Quantitative Determination of Inosite in Corn Pollen. 60 gm. of the pollen were digested in 200 cc. of water with frequent shaking for 3 hours. It was then filtered through a layer of paper pulp and washed with water until 450 cc. of filtrate were obtained. ‘The filtrate was yellow in color and it possessed a strong odor of pollen. It was evaporated to 50 cc. on the water bath, filtered, and the inosite was isolated by the method of Mayer (8). The pure colorless characteristic inosite crystals finally obtained weighed 0.5 gm. which corresponds to a yield of 0.83. per cent, but considering the inevitable losses during the isolation and purification it is very probable that corn pollen contains not less than 1 per cent of free inosite. The crystals gave the Scherer reaction and melted at 221°C. (uncorrected). We have no doubt whatever that the substance was pure inosite and the analysis was therefore omitted. Extraction of the Pollen Residue with 70 Per Cent Alcohol. The pollen residue which remained after extracting with ether and alcohol, as already described, was digested in 1 liter of 70 per cent alcohol at room temperature, with occasional stirring, for several days. It was then filtered on a Buchner funnel and washed with 70 per cent alcohol until about 1,300 ee. of filtrate were obtained. This solution was concentrated under reduced pressure to about one-half of its volume. It was then heated nearly to boiling and alcohol was added until the solution turned cloudy. After standing in the ice chest several days a considerable amount of a crystalline substance had separated. R. J. Anderson and W. L. Kulp 447 The crystals were filtered off and washed in alcohol. They were dissolved in a little water, deeolorized with animal charcoal, and again brought to crystallization by adding alcohol. After recrystallizing four times 1.4 gm. of colorless needle-shaped crys- tals were obtained. The crystal form was characteristic of inosite. The substance gave the reaction of Scherer and melted at 221°C. (uncorrected). Since the reactions and properties of this substance indicated that it was pure inosite the analysis was omitted. After the above crystals of inosite had separated the mother liquor was concentrated to a thin syrup under reduced pressure at a temperature not exceeding 40°C. The syrup was taken up in a little water and precipitated with a solution of lead acetate. After settling, the precipitate was filtered on a Buchner funnel and washed with water. The filtrate was freed from lead by hydrogen sulfide and the excess of hydrogen sulfide was removed by a current of air. The solution was then decolorized with animal charcoal and concen- trated under reduced pressure. Sulfuric acid was added until the solution contained about 5 per cent of this acid. A concentrated solution of phosphotungstic acid was then added until no further precipitation occurred. After standing for several hours the precipitate was filtered and washed with 5 per cent sulfuric acid. Unfortunately, the filtrate was lost through an accident which prevented an examination of it for amino-acids and soluble car- bohydrates. The phosphotungstie precipitate was rubbed up in a mortar with an excess of barium hydroxide, filtered, and the precipitate thoroughly washed with water. The filtrate was acidified slightly with sulfuric acid, filtered from barium sulfate, and concentrated under reduced pressure to about 400 cc. The solution was then made up to 500 cc. with water. Nitrogen was determined in this solution by the Kjeldahl method and it was found to contain 0.8470 gm. of nitrogen. Through fractional precipitation with phosphotungstic acid we were able to separate the nitrogenous constituents into two prin- cipal fractions which were identified as choline and /-proline. The nitrogen recovered amounted to about 80 per cent, the balance being lost in the processes of separation and purification. a. nly: Ps — é. om ~~ 4 ~_ a a Se ee rs — tee 448 Composition of Corn Pollen Separation of Choline. An attempt was made to separate histone bases by the method of Kossel and Kutscher (9) as described by Steudel (10) but only traces of nitrogen were precipitated. After the solution was freed from barium and silver about 4 per cent of sulfuric acid and a very slight excess of phosphotungstic acid were added. The precipi- tate was filtered after standing over night and washed with 4 per cent sulfuric acid. The filtrate and washings were reserved for the preparation of [-proline. The phosphotungstic precipitate was decomposed with barium hydroxide, the excess of barium removed with carbon dioxide, and the filtrate made up to 500 ec. This solution was found to contain 0.2327 gm. of nitrogen determined by the Kjeldahl method. This would correspond to 2.01 gm. or 0.34 per cent of choline in corn pollen. The solution was concentrated under reduced pres- sure, filtered from a small quantity of barium carbonate, and evaporated in a vacuum desiccator over sulfuric acid. The thick syrupy residue which remained was taken up in a little alcohol and to it was added an alcoholic solution of picric acid until the solu- tion turned cloudy. On cooling the picrate separated in large needle-shaped crystals. These were filtered off and washed in a little alcohol. The filtrate and washings were concentrated and on cooling a further quantity of crystals were obtained which were added to the first lot. The picrate was recrystallized from a little hot water, filtered, washed with absolute alcohol, and dried in vacuum over sulfuric acid. The dry picrate weighed 2.67 gm. and it melted at 239-240°C. (uncorrected). The picrate was suspended in about 50 cc. of water and acidi- fied with hydrochloric acid and the picric acid was shaken out with ether. The aqueous solution was concentrated under re- duced pressure and finally dried in vacuum over sulfuric acid until a mass of colorless crystals remained. The crystals were very hygroscopic and rapidly liquefied on exposure to the atmos- phere. The xanthine and the Weidel reactions and the Kos- sel reaction for adenine were all negative. The substance was taken up in about 20 ec. of absolute alcohol, and ether was added gradually until crystallization began. ‘Tt was then placed in a freezing mixture for 1 hour. The crystals were R. J. Anderson and W. L. Kulp 449 filtered, rapidly washed with absolute ether, and dried in vacuum over sulfuric acid. In this manner the substance was obtained in long fine colorless needles. It was very hygroscopic and on ex- posure to the air the crystals rapidly liquefied. The crystals did not melt or show any change when heated to 260°C. The alcoholic solution of the substance gave an orange-colored precipitate with platinic chloride. The platinum double salt was prepared and after recrystallizing from hot water to which about 15 per cent of alcohol was added it was obtained in orange-colored octahedral crystals. The air-dried double salt melted with de- composition at 232°C. (uncorrected). On analysis, after drying at 105°C. in vacuum over phosphorus pentoxide, values were obtained which agree with the theoretical composition of choline platinum chloride. 0.1742 gm. substance: 0.0548 gm. Pt. (CsHisNO Cl). Pt Cly. Calculated. Pt 31.64 per cent. Found. “\ alee Preparation of I-Proline. After the choline fraction had been precipitated by phospho- tungstic acid, the filtrate was freed from sulfuric and phospho- tungstic acids by barium hydroxide. The filtered solution was evaporated under reduced pressure and the residue was taken up in a little hot water. A small quantity of barium carbonate was filtered off and the filtrate made up to 100 cc. with water. Ni- trogen was determined by the Kjeldahl method and the solution was found to contain 0.4396 gm. of nitrogen. The solution was acidified with hydrochloric acid and con- centrated under reduced pressure and finally dried in vacuum over sulfuric acid. A thick syrup remained which on scratching with a glass rod immediately crystallized. The crystals were digested in a little absolute alcohol, filtered on a Buchner funnel, and washed with a little cold absolute alcohol and ether. After drying in the air the substance weighed 3 gm. and it was a nearly pure white crystalline powder. It was very soluble in water and in hot 95 per cent alcohol. It also dissolved readily in hot abso- lute alcohol and on cooling it separated in colorless prisms. The alcoholic solution gave no precipitate with platinic chloride or with picric acid. The substance showed an acid reaction on litmus and on ignition it left no residue. 450 Composition of Corn Pollen The substance was twice recrystallized from hot absolute al- cohol and was obtained in beautiful colorless prisms. When heated in a capillary tube it melted with gas formation at 206— 207° (uncorrected). It contained nitrogen, but sulfur, phos- phorus, and halogens were absent. Boiled with copper oxide it gave off a peculiar odor and a deep blue-colored solution resulted. This solution was filtered and evaporated when a deep blue- colored amorphous copper salt was obtained which was com- pletely soluble in alcohol. The substance was analyzed after drying at 105°C. in vacuum over phosphorus pentoxide. There was no loss in weight on drying. 0.1470 gm. substance: 0.1051 gm. H,O and 0.2791 gm. COs. 0.1560. * i 16.9 ee. of nitrogen at 16° and 731 mm. Found. . GC 51.78, H 8.00, N 12.30 per cent. For C;H»NO2 = in iby Calculated. C 52.17, H 7.82, N 12.17 per cent. In aqueous solution the substance had a specific rotation of — 69.69°. The properties and composition of this substance agree with those of [-proline although the rotation is lower than given by Fischer (11). The identity with proline was established by con- verting it into dl-proline as described by Fischer (11). The cop- per salt was obtained in the form of the characteristic deep blue- colored crystals. The following results were obtained on analysis after drying at 105° in vacuum over phosphorus pentoxide. For (CsHsNO2)2 Cu + 2 H.O = 327.5. Calculated. H,O 10.99 per cent. Found. “« 10314 For (CsHsNO:2)2 Cu = 291.5. Calculated. Cu 21.78 per cent. Found. er 27S ea ee Some of the l-proline was treated with phenylisocyanate in alkaline solution and the resulting compound was converted into the anhydridé or hydantoin by heating with 4 per cent hydro- chloric acid. After recrystallizing four times from aqueous al- cohol, the reaction product was obtained in the form of fine colorless needles which melted at 143°C. (uncorrected). Pro- line hydantoin melts at 143°C. R. J. Anderson and W. L. Kulp 451 Corn pollen contains a relatively large amount of proline. The solution from which the proline was prepared contained 0.4396 gm. of nitrogen which corresponds to 3.6 gm. or 0.6 per cent of free /-proline in the pollen. Phosphorus in the Extracted Corn Pollen. The pollen residue after extracting with ether, absolute alco- hol, and 70 per cent alcohol was analyzed for phosphorus. Total phosphorus was determined after destroying the organic matter by the Neumann method. The total soluble and the inorganic phosphorus were determined in the extracts obtained on digest- ing the pollen residue during 5 hours in 1 per cent hydrochloric acid. The value for the organic phosphorus soluble in 1 per cent hydrochloric acid was obtained by subtracting the inorganic from the total soluble phosphorus. The results are given in Table V, TABLE V. Forms of Phosphorus in Corn Pollen after Extracting It with Ether, Absolute Alcohol, and 70 Per Cent Alcohol. Total soluble Total phosphorus. Inorganic phosphorus. | Organic phosphorus. phosphorus. per cent per cent per cent per cent 0.43 0.29 0.22 0.07 The organic phosphorus in Table V corresponds to the phytin phosphorus as determined in other plant material. Judg- ing by the very small difference between the total soluble and the inorganic phosphorus in pollen it appears rather doubtful if phytin or inosite hexaphosphoric acid is present in this material. Nitrogen in the Extracted Corn Pollen. The pollen residue remaining after extracting with ether, ab- solute alcohol, and 70 per cent alcohol was analyzed for nitro- gen. The material was extracted with the solvents mentioned in the table and the results obtained are given briefly in Table VI. In earlier investigations the presence of malic acid has been reported in the pollen of Phenix dactylifera by Fourcroy (12) and in the pollen of Typha latifolia by Braconnot (13) and Kresling (14) found tartaric and malic acid in the pollen of Pinus sylvestris. 452 Composition of Corn Pollen In the case of corn pollen we were unable to obtain any evi- dence of the presence of any of these acids. The only acid which we could find in aqueous extracts of corn pollen was phosphoric acid in the form of calcium phosphate. The soluble constituents which were isolated from corn pollen and identified are given in Table VII. The percentages are cal- culated to the dry pollen. TABLE VI. Nitrogen in the Extracted Corn Pollen. Nitrogen soluble | Nitrogen soluble Nitrogen:soluble in 1 per cent in 1 per cent Nitrogen soluble nit NaOH. hydrochloric acid.) *” Soh in distilled water. Total nitrogen. Digested on Digested on Dicested at 25°C, | Digested at 25°C. water bath for 24 | water bath for 24 ra iat arena for 24 hours. hours. hours. ; per cent per cent per cent per cent per cent 5.40 4.93 3.20 0.40 0.20 TABLE VII. Soluble Identified Constituents Occurring in Corn Pollen. Substance. Per cent. Amorphous phosphatide.. 20g ate 4.75 62 cee bee eee ae 1.94 Crystalline phosphatide.. 022i cv ec: ere ee eee eee eee 0.67 NAV ORSIEES AS isis cota ew Sata thane ea eae ane eee ee eerie 0.83 (Clive) bis (hs aes 4 Raber Arte te ted gata Sree i rn, th MG ES as 0.34 PeBroline 0s Rei. EE = ER. AR, Se noe te 0.60 Mayricyhaledhol.: .5i:i2.5 03 +.Gunriceh Seteee aes Trace Some attempts were made towards the isolation of proteins, nucleic acid, and certain carbohydrates from pollen but these experiments are still incomplete. We hope to present a fuller report on the above constituents in a later publication. SUMMARY. The approximate composition of the pollen from three varie- ties of corn has been determined and the results indicate a dif- ference in the composition of the pollen from different varieties of corn. A complete analysis of the ash of the pollen from one variety of corn is given. 24 save R. J. Anderson and W. L. Kulp 453 Evidence is presented which indicates the presence of at least two phosphatides in corn pollen.. One was an amorphous sub- stance which also contained sulfur but the other was a crystalline phosphatide. Relatively large quantities of free inosite, /-proline, and choline occur in corn pollen. BIBLIOGRAPHY. 1. Glikin, W., Arch. ges. Physiol., 1903, xev, 107; see also Biochem. Z.., 1907, iv, 235. 2. Rosenfeld, G., Centr. inn. Med., 1900, xxi, 833. 3. Osborne, T. B., and Wakeman, A. J., J. Biol. Chem., 1915, xxi, 539. 4, Winterstein, E., and Hiestand, O., Z. physiol. Chem., 1907-08, liv, 288. 5. MacLean, H., Lecithin and allied substances. The lipins, New York, 1918, 165. 6. Methods of analysis, U.S. Dept. Agric. Off. and Prov., Bull. 107, revised, 1911. 7. Winterstein, E., and Hiestand, O., Z. physiol. Chem., 1906, xlvii, 496. 8. Mayer, P., Biochem. Z., 1907, ii, 393. 9. Kossel, A., and Kutscher, F., Z. physiol. Chem., 1900-01, xxxi, 165. 0. Steudel, H., in Abderhalden, E., Handbuch der biochemischen Arbeits- methoden, Berlin, 1910, ii, 498. 11. Fischer, E., Z. physiol. Chem., 1901, xxxiii, 151. 12. Fourcroy, A. F., Ann. Mus. Nat. Hist. Nat., 1802, i, 417. 13. Braconnot, H., Ann. chim. et phys., 1829, xlii, 91. 14. Kresling, K., Arch. Pharm., 1891, cexxix, 389. + Hep Au . Th +7 y fines ist : bint a bis a Re fret yat ti +4) he ae eS . ee . ae obiterta ; “oy STS Se alin 7h x Ht feic Puedes > a . : or serum after coagulation) or tissue extracts Kage — THE ROLE OF CEPHALIN IN BLOOD COAGULATION. By ANDRE GRATIA anp P. A. LEVENE. (From the Laboratories of The Rockefeller Institute for Medical Research.) (Received for publication, November 30, 1921.) The process of blood clotting is very complex. As it occurs normally, it is the result of many factors not well understood. The progress made in recent years was due to the fact that ways were found to separate the many factors into groups, some of which control definite phases of the complex process of blood clot- ting. The phase best understood is that of the conversion of fibrinogen into fibrin. Four substances take part in this reac- tion. One is the substrate, and the other three combine to bring about the transformation of the substrate from a soluble into an insoluble state. Alexander Schmidt understood the process cor- rectly. More recent workers brought out many of the details of the process. The most recent contributions were made by Bor- det and his collaborators, and by Howell and his coworkers. The terms applied to each of the factors differed with the in- dividual author. Since the blood clotting experiments here re- ported were carried out by a worker of Bordet’s school, the nomen- clature employed in this publication is of that school. In terms of that school the process of fibrin formation is expressed by the following diagram. Plasma at the ‘ moment of coagulation| serozyme in presence of Ca ions — thrombin Platelets or certain lipoids) Fibrinogen = fibrin 456 Cephalin in Blood Coagulation Howell accepts the interplay of all four of the substances in the process of blood clotting but holds a different view on the réle of eytozyme, which, according to Howell, plays no part in the actual transformation of fibrinogen into fibrin. Howell also disa- grees with other workers in his view on the chemical nature of eytozyme. Alexander Schmidt, Wooldridge, Bordet, Delange and others regarded the substance as lecithin. This opinion was based on the thermostability of the substance and on its solubility in alcohol. Howell on the contrary came to the con- clusion that the active substance was another phosphatide; namely, cephalin. Howell and his coworkers have also made an attempt to associate the activity of the phosphatide with a definite pecu- liarity of its chemical structure. At the time of the work of Howell it was generally accepted that the unsaturated acid en- tering into the structure of cephalin differed from that of lecithin. From cephalin, linolic acid was isolated and from lecithin, oleic acid. According to Howell and McLean the higher unsaturation of the fatty acid is the factor which lends to cephalin its property of being an agent in the formation of thrombin. The present communication is a mere note dealing not with the entire problem of fibrin formation but only with the chemical nature of cytozyme. Is cytozyme lecithin or cephalin? Since the work of Howell and his coworkers, the knowledge of the chemical structure of phosphatides has made considerable prog- ress. In the light of this progress the conclusions regarding the chemical nature of cytozyme required reinvestigation. The recent work on lecithin has brought out the fact that there exist forms of this substance which contain in their molecule a fatty acid of still higher unsaturation than the one previously isolated from cephalin. In the light of the theory of Howell and McLean, one might have expected the new form of lecithin to play the same part as cephalin in fibrin formation. Furthermore, recent work on cephalin has brought out the fact that the material handled by the older writers under the name of lecithin was in reality a complex mixture and not a uniform substance. The components of this mixture were found to be identical in character with those of another complex mixture described by previous writers under the name of cuorin, or heparphosphatide. Cuorin and the cephalin of the older writers -_ “fs A. Gratia and P. A. Levene 457 differed one from another in the proportions of some of their com- ponents. Both substances were found to consist of true cephalin, true lecithin, and of the same substances in a state of partial decomposition. The character of the decomposition products varied from sample to sample. Yet, Howell and his coworkers observed that cephalin and cuorin acted in the process of blood coagulation antagonistically to one another. On the other hand, a substance was recently prepared which was free from the decomposition product of lecithin and of ceph- alin and which contained 75 per cent of undecomposed cephalin and 25 per cent of undecomposed lecithin. Whether or not the substance contained impurities undetectable by the present meth- ods of analysis, cannot be stated. It is self-evident that it became important to compare the cy- tozymic function of the three substances; namely, of ordinary lecithin, of lecithin which contained the fatty acid of a high degree of unsaturation, and of the new cephalin material. Ina way also the present materials were mixtures. Ordinary lecithin con- tains a small proportion of the new form. The new form still contained a very small proportion of the older type. The cepha- lin contained a small proportion of lecithin. Yet even such material was sufficient to bring out the fact that lecithin, regard- less of its form and of its origin, possesses no cytozymic action. On the other hand, material containing 75 per cent of undecom- posed cephalin and 25 per cent of lecithin possesses unusually high cytozymic action. It is still active in a concentration of At"). The coagulation experiments were carried out by Dr. Gratia who followed the routine customary in Bordet’s school. The plan and the details of the experiments follow. EXPERIMENTAL PART. Oxalated plasma from which most of the platelets have been removed by centrifugation contains only a small amount of cyto- zyme and consequently clots very slowly when recalcified, but clots quickly if some cytozyme is given back in form either of platelet suspension, tissue juice, or lipoidic tissue extract. This offers means of testing the cytozymic properties of a given lipoid by measuring the accelerating influence of the lipoid on the coagula- tion of a plasma almost free from platelets. 458 Cephalin in Blood Coagulation When an oxalated plasma has been strongly centrifugalized and then recalcified, the few remaining platelets contain just enough cytozyme to react with but a small part of the serozyme and thus yield only a small amount of thrombin. The plasma clots slowly and a great excess of unutilized serozyme is found in the serum after coagulation. Such a serum is rich in serozyme and is an excellent reagent to test the cytozymic properties of a given lipoid. If cytozyme even in very small amount is added to this serum, an active production of thrombin immediately results and this mixture is able to clot an equal volume of fibrino- gen or oxalated plasma in a few minutes. This is the serozyme- cytozyme reaction of Bordet and Delange. In our researches we have submitted our different lipoids to both tests. The materials used were prepared as follows: Preparation of the Reagents. 1. Lipoidic Emulsions.—1 per cent emulsions of our three lipoids were made in saline solution. As a controla similar 1 per cent emulsion was made with lipoidic extract of tissue which was known to possess strong cytozymic properties. When necessary, higher dilutions of these suspensions were made in the course of the experiments. 2. Oxalated Plasma Free from Platelets —A rabbit was carefully bled from the carotids with a paraffined cannula. Precautions were taken to avoid the contact of the blood with tissue juice and 9 parts of blood were received in 1 part of a 1 per cent solution of sodium oxalate in saline solution, and thoroughly mixed. This 1 per cent oxalated blood was centrifugalized at high speed during 1 hour and the clear supernatant plasma removed from the cells with a pipette. For use in the experiments 1 part of this oxa- lated plasma (O. P.) was recalcified with 4 parts of a 0.35 per cent solution of calcium chloride in saline solution (Ca). 3. Serum Rich in Serozyme.—A few cc. of oxalated plasma were recalcified as above described. When coagulation began, the re- calcified plasma was defribinated with a glass rod. The serum obtained was kept at room temperature until the next day. As thrombin is very labile, the small amount of thrombin left after this very slow coagulation disappears quickly and the next day the serum containing nothing but a large excess of serozyme is ready for use. —————<—<— et A. Gratia and P. A. Levene 459 4. Fibrinogen.—Instead of the so called pure solution of fibrino- gen, ‘‘dioxalated plasma” (F) was used as a test for thrombin. This very convenient reagent was prepared according to the technique of Bordet and Delange; 7.e., 1 part of 1 per cent oxa- lated plasma was diluted with 4 parts of a 2 per cent solution of sodium oxalate in saline solution. A. Egg Lecithin. Experiment I. 0.25 ec. O.P. + 1 drop saline solution + 7 ec. Ca = 110’ aoe EL OS ope lecithineg ee 0.25“ “ +1 “ eytozyme ag eh / ee 20 Egg lecithin exerts only a slight accelerating influence on the coagulation of recalcified oxalated plasma. Experiment IT. 0.25 ce. serozyme + 1 drop saline solution ..5’... + 0.25ec. F = « O25: Ss <3 +1 “ egg lecithin . =. Otte 6S ae 0.25 “ . +i seytozyme 1/100 ....5'... 025.45: 0.25 “ = tines tectthin 1/100 ..5'... - O25 % "oe 25.“ . =--E* = Weytozyme-t/1, 000; .5'.¢s74- 0.25 “0 © = 25! 0:25 “ ° -- E> aecithin 1/1,000..5'... 0.25" “= a Egg lecithin is thus inactive at higher dilutions. In the fol- lowing series the tests were allowed to react at longer intervals. 460 Cephalin in Blood Coagulation Experiment IV. 0.25 cc. serozyme + 1 drop lecithin ... 5’... + 0.25 ec. F = «© (0), -4ay ce. °C. | calories| calories 119 | 0.298 m NaOH....... 50 4.282] 56 8.3 120) || (A ee 50 | 4.209) 56 8.3 MPOS29S AG eo ac sees 50 4.166) 56 8.3 22 | OS 50 4.183) 56 8.3 HeoeOezosia eee 50 4.211) 56 8.3 OPO R29 Me "sche 50 4.214) 56 8.3 Pe eOt4G mw Se see. es 100 2.216} 111 16.8 Wow Ont4On Skee 100 2.225) 111 16.8 138 | 0.181 um KOH........ 100 2.706} 111 16.8 PEO RI Sh Wee oo 100 2.665} 111 16.8 142 | 0.250 m Na.CO3;..... 100 | 1.513) 109 16.2 143 | 0.250 mM Se ee 100 | 1.541} 109 16.2 144 | 0.250m ae Ree 100 | 1.481) 109 | 16.2 243 | 0.067 m Na;HPO,....| 100 | 0.323) 111 | 17.0 257 | 0.067 Mm 7 eee 00. | 0.298) tit Fee 261 | 0.067 m eee 100. ,| 0.328] TES | Teo 258 | 0.067 m KH2PQ,..... 100 | 0.147} 111 | 17.0 Formula for calculation: (7) = {[4) x ()] — @)} x 1,000 (2) X (3) +11,100 calories at 18°C., for the reaction of solutions of carbonic acid with solutions of sodium hydroxide. The heat of neutralization of strong acids and strong bases at 22°C. is +13,600 calories. Using the above experimental re- ae herr” sults, the heat of ionization of carbonic acid to bicarbonate is 472 Heat of Reaction of Oxygen —2,910 calories. Thomsen (6) found it to be —2,800 calories at 18°C., and Kendall (10) calculated by the isochore that it would be —2,830 calories at 22°C. Heat of Oxidation of Pyrogallol Solutions.—The chemistry of the reaction of oxygen with pyrogallol is not known, and varies with dilution and other factors. Berthelot (11), however, has shown that each molecule of sodium pyrogallate absorbs 3 atoms of oxygen under most conditions. No inactive gas was used in the fore period, but oxygen was bubbled through a solution of 0.5 m sodium hydroxide, and after 5 minutes a weighed amount of crystalline pyrogallol was added (Table III). TABLE III. Heat of Oxidation of Pyrogallol. (1) (2) (3) (4) (5) (6) @ Heat No. Pyrogallol. | Solution. Bae a Sere pestis Avene Le : molecule pytogallol. gm cc °C ole calories calories degree 114 0.050 100 0.288 112 79, 500 +65, 000 il4a 0.100 100 0.825 112 113, 700 115 0.050 50 0.712 57 99, 800 116 0.050 50 0.810 57 113, 500 +95, 900 116a 0.050 50 0.783 57 109, 700 117 0.050 50 0.824 57 115, 300 (4) X (5) X 126 (2) Formula for calculation: (6) = The heat of solution of pyrogallol, and its heat of complete neutralization, were measured by Berthelot (12) and by de Forcrand (13). Averages of their values (—3,590 and +13,440) were subtracted and a further correction was made for the heat of solution of 3 gram atoms of oxygen, to give the final result of -++95,900 calories. - The three varieties of gas reactions studied above indicate the reliability of the method, and demonstrate that it can attain an accuracy of about 1 per cent. Such an accuracy was deemed more than sufficient for the study of the oxygenation of hemoglobin in solution. See E. F. Adolph and L. J. Henderson 473 Experiments with Hemoglobin Solutions and Blood. Heat of Reaction of Oxygen and Carbon Monoxide with Hemo- globin Solutions—The quantitative results depend upon the fact that 1 molecule of oxygen or of carbon monoxide combines chemi- cally with that amount of hemoglobin which contains 1 atom of iron. The molecular weight of hemoglobin has been assumed to be 16,700. Corrections have been made for the heats of solution of the gases, both in measuring heats of reaction and in comparing the values of the equilibrium constant for oxyhemoglobin (K) and for carbon monoxide hemoglobin (k) at various temperatures. Several investigators have measured the heats of reaction of oxygen and of carbon monoxide with hemoglobin in blood. With- out attempting a critique of their methods, a summary of their results is presented in Table IV. In addition to the experiments of four authors included in the table, the work of Camis (18) and of Meyerhof (19) should be mentioned. Camis obtained a nega- tive heat of formation of oxyhemoglobin, though his results have usually been interpreted otherwise. Meyerhof found roughly that when fresh blood was oxygenated, the heat produced just balanced that absorbed due to carbon dioxide carried out of solution. To avoid as many complications as possible, it was decided to work chiefly with purified hemoglobin solutions. Washed cor- puscles from defibrinated beef blood were dialyzed by a method which has been already described (20). Before use the solu- tion of dialyzed hemoglobin was boiled at 40°C. in vacuo to re- move all gases. It was cooled to 22°C., measured into the cal- orimeter without exposure to air, and subjected to a stream of hydrogen. After a few minutes the flow of gas was stopped while a sample of solution for analysis was pipetted from the calor- imeter. At the end of the passage of the reacting gas a second sample was taken for analysis. All oxygen and carbon monoxide analyses were performed by the method of Van Slyke (21). In every case a correction was made for dissolved gases. There is some evidence that the quanti- ties thus measured may be too large (22). Carbon dioxide was measured with the same apparatus, using Van Slyke’s technique (23), except that lactic acid replaced sulfuric acid as the reagent (1) Tem- Heat eat |Heat of Average - | Tem- Gas | devel- | developed 4 Hemoglobin 0 ei rs d solution] molal hee preparation. oe ais # eh per gm. molecule as Bas bs Oxygen + hemoglobin. Of. “G: 5 fase calories| calories |calories| calories Berthelot | Defibrin-| 9 | 0.115) 20.2 | 0.895) 14,960 (14). ated 0.108| 18.3 | 0.918 re 3,000 sheep blood. Torup (15). | Crystal- | 17 | 0.050) 4.8 | 0.754) 12,600 line 0.042) 3.7 | 0.658) 11,000|| _ 11.300 horse 0.041) 3.7 | 0.623) 10, 400; Hb: 0.046; 4.6 | 0.678} 11, 300} Barcroft Crystal- | 16 | 0.1388] 11.9 | 1.82 | 30,400 and Hill line 1.98 | 33, 070;| 3,300) 27, 600 (16). dialyzed 1.75 | 29, 230 dog Hb. Du Bois- Defibrin-| 20 | 0.17 | 19 1.06 | 17,700 Reymond ated 0.16 | 21 1.09 | 18, 200 (aly). horse 0.382 | 34 1.21 | 20,200 blood. 0.214) AL 1.33 | 22, 200?| 3,200] 19,650 0.24 | 23 1.39 | 23, 200 0.34 | 26 1.73 | 28,900 0.30 | 23 1.77 | 29,600 43 | 0.08 | 10 0.97 | 16,200 | 2,400) 13, 800 .|Crystal- | 20 | 0.04} 6.7 | 0.80 | 18,400 line 0.12 | 16.2 | 0.99 , 165,00 horse 0.02 | 2.3 | 1.14 | 19,000 3,200), tans Hb. 0.02 | 2.3 | 1.14 | 19,000 Defibrin-| 20 | 0.04] 5 1.05 | 17,500 ated 0.08 | 10 1.05 | 17,500?) 3,200) 15,100. horse 0.09 | 10 1.20 | 20, 000 blood. ; Only par- tially re- duced. Previous Measurements of Heat of Reaction. (3) TABLE IV. (4) 6) (6) 474 (7) (8) (9) E. F. Adolph and L. J. Henderson A75 TABLE IV—Concluded. (1) (2) (3) (4) (5) (6) (7) (8) (9) Heat Heat Tem- 4 Heat of| Average hathar Hemoglobin poe pera- Pic dare pe solution] _molal 7 preparation.| P ture Pp Pp ealcula-| heat of ture. -— ed. |per gm.} molecule rise. Hb. Hb. ted. reaction. —q~“«“ Carbon monoxide + hemoglobin. °C; wa By a calories| calories |calories| calories Berthelot | Defibrin-} 9 | 0.998 1.08 | 18,030 (14). ated 0.125 1.16 | 19, a sisal Jk sheep blood. Du Bois- | Defibrin-| 38 | 0.17/19 | 1.2 | 20, fee one Reymond ated 0.23 | 33 0.9 | 15,000)| ’ (17). horse 17, 800 blood. 20: | 0.19 | 22 1.2 | 20,000 0.20 | 17 1.6 | 26, 700-| 2, 900 1.2 | 20,000 to avoid the precipitation of hemoglobin. Frequent analyses of evacuated samples show that the average amount of carbon di- oxide present was less than 2 volumes per cent. Any fall in temperature due to the expulsion of carbon dioxide was automatically corrected for in the graphical method of cal- culating the cooling (Fig. 4), since the fall is continued after the oxygenation is complete. Any heat production due to internal metabolism of blood, such as Meyerhof (19) measured, was also compensated by the correction. A third error similarly cared for was the evaporation of octyl alcohol. Evaporation of water was prevented by the complete saturation of the gases going into the calorimeter; even failing this it was likewise offset by the method of experimentation. Reaction of Oxygen with Hemoglobin Solutions—All the com- plete measurements with oxygen are given in Table V. The average result for 35 measurements at 22°C. turned out to be +10,050 calories per gram molecule of oxygen, standard de- viation +2,320 calories (22 per cent). When grouped statisti- cally, these data give a slightly skewed probability curve with a mode at +11,200 calories. The heat of solution of this quantity of oxygen was calculated to be +3,100 calories, leaving +6,950 calories as the average heat of reaction at 22°C. Temperature Seat oe 476 Heat of Reaction of Oxygen A smaller number of measurements were made at 38°C. The average result for these eight determinations was +8,050 calories per gram molecule of oxygen, +1,700 calories (21 per cent). The heat of solution at this temperature was +2,700 calories, leaving + 5,350 calories as the average heat of reaction at 38°C. Time -minutes Dialyzed hemoglobin solutions Fia. 4. le recorded during thermochemical experi- ments with hemoglobin solutions. For both temperatures together, the average was +6,650 calories. The lower value at body temperature as compared to room temperature, while within the limit of error, is in agreement with the principles of thermodynamics. A smaller value of Q at body temperature was also obtained by du Bois-Reymond (17). It should be noted that the amount of heat produced when hemoglobin unites with oxygen is not definitely correlated with (1) No. E. F. Adolph and L. J. Henderson TABLE V. Heats of Reaction of Oxygen with Hemoglobin. (2) Solution. Dialyzed whole blood..... “ “ce “ “ec whole blood..... Corpuscles + 0.9 per cent NaCl. Dialyzed corpuscles. ..... “ “ Corpuscles + 0.9 per cent NaCl. Dialyzed corpuscles. ..... | “cc “ NaCl. Waeate plod... cco... | Dialyzed corpuscles....... “ “ ! 02002 m KPO... <.. Dialyzed corpuscles + 0.003m KH2PO,......... Dialyzed corpuscles. ..... “ “cc 0.074 n NaHCO3........ Dialyzed corpuscles + 0.068 n NaHCO........ “ce wn KOH Dialyzed corpuscles “ce “ | 0.074) 0.045 | 0.063) 0.070 0.086 0.075 0.079 0.100 (4) Solu- tion. 100 100 94 88 477 (5) (6) (7) (8) H Heat | Initial | Final eared capac | HbO2. | HbO>. |per gram ity. molecule HbO2. calories 7 aiken per cent|per cent! calories 106 13.3 |10, 700 105 17.5 | 5,400 106 13.0 | 8,900 102 15.3 | 8,700 101 1.2 | 13.5 | 9,400 99 | 3.7 | 13.4 |10,700 96 | 11.2 | 29.9 | 5,300 99 | 6.4 | 22.1 |10, 400 92 0.9 | 18.5 {11,600 93 0.6 | 37.3 |11, 100 105 0.8 | 11.38 |13, 400 103 | 1.4 | 17.1 |11,600 99 | 1.1 | 23.3 |11, 100 103 | 1.6 | 17.0 |11,500 102 | 1.3 | 19.2 |14,000 100 | 3.1 | 23.0 |12, 400 99 | 2.9| 19.1 | 8,600 103 | 0.9 | 17.1 |11,500 103 1.0 | 16.3 | 6,700 104 | 1.1 12.1 |12,800 110 | 1.0 | 16.4 |13, 000 74 | 2.3 | 14.5 |11,600 96 7 | 12:70 |-7; 900 101 8.5 | 18.4 {11,200 103 | 1.4] 15.7 | 8,700 103 | 1.0 | 16.0 |10,000 102 | 2.5 | 15.4 |10,200 99 | 2.0 | 15.1 |10,600 89 | 5.7 | 22.3 |10, 200 478 Heat of Reaction of Oxygen TABLE V—Concluded. (1) (2) (3) (4) (5) (6) (7) (8) pont | ‘gol. | 2ee8'| Taseiat | inal’ | erate = . pera- olu- _| Initia ina evolve No. Solution. ture tion. gees HbOsz. | HbOsz. |per gram rise. y- molecule HbO.. 22°C .—Concluded. 2G: jon per cent|per cent| calories Dialyzed corpuscles + 0.046 N lactic acid 0.109} 102 | 104 1.6 | 20.7 |10, 100 Dialyzed corpuscles 0.132 90 1.4 | 21.6 |11, 200 (79 “ce 0.076 106 4.2 | 14.6 |13, 100 0.048 103 5.0 | 18.7 | 9,900 0.026 104 4.6 | 14.8 | 4,600 0.080 100 5.3 | 15.6 |13, 800 38°C. _ 293 | Dialyzed corpuscles. ..... 0.094) 98 | 102 0.3 | 20.8 | 7,800 293 a as Elen Drege, 0.049) 95 99 | 18.5 | 20.8 {11,700 294 yt ee AN ioe 0.064; 67 “fl 1.9 | 19.3 | 6,500 2944 s Se A See 0.074) 85 89 5.5 | 19.0 | 9,600 296 “s Gah bee 0.081} 97 | 103 1.0 | 18.0 | 8,400 296a as SE AE eee 0.066} 97 | 1038 1.1 | 18.8 | 6,600 296b Se AR od eae 0.086) 92 94 1.2 | 23.0 | 6,700 297 | Corpuscles + 0.9 per cent ING ee Scat ebe temas 0.151; 96 93 1.9 | 36.4 | 7,100 (3) X () X 16,700 x 100 (4) X [(-©)] Formula for calculation: (8) = the presence of salts, acids, alkalies, or other proteins. More- over, the age, dilution and previous treatment of the solutions, and substitutions of air for pure oxygen have no controlling influence. The latter portion of the oxygenation evolves as much energy as the total oxidation; the strict proportionality between combined oxygen and temperature production is shown in Fig. 5. The heat of solution of the oxygen merely dissolved in blood is negligible, since the solubility of oxygen is so small that no heat of solution in water could be demonstrated directly. Moreover, it is nearly compensated by the displacement of hydrogen from solution. 479 E. F. Adolph and L. J. Henderson uUIqO[SoWOY YIIM JUoWTIOdxXO UB SulaNp Uolrywinyes UdSAXO UT puv oINYvIOdUIO} UI SoduLYD °G “YY o p Cc A] y vO} jt v YU : 5 02 iz ¢ y ow a ie) a ’ E12 HZ Saynuiw ~ aur “Oe - aunqvied Wo] Heat of Reaction of Oxygen 480 ose ‘II] I°St | ST ZOL 002 ‘TZ| 6° ZT | 8°9 OOT 002 ‘8I| ¢°9T OOT ‘TI] OFT | O'F oL 002 ‘FI 000 ‘8z] 0's | GT OOT 002 ‘TZ} 6 91 | 8g 16 002 ‘FI! 0'9T | 6°0 68 008 ZI] 2°91 | 9'T 06 aalbop §9140]09 | sailojpa |7Uad wad) 2Wa9 “ad 1 4 $9140]D0 “poe olny -[us -+- ofuloy UlOIZ OD “plow olny -[ns + ofuI0y WOIy OD ‘plow orang -[ns ++ oFUdoOy WLS OD ‘plow olny -[ns -+ oluLoy WoL, OD ‘pre DMN J[NS + O1[BXO WOIJ OD ” ” .! ‘SBS SUTYBULWIN IT] ‘saposndio0o pezAyeiqy 60 ar ” ”» ” ” ‘TOBN Jus rod 69 + saposndi0g “Poorqd efouM ‘saposndi09 pozA]viqy _POola)e[OUm. ‘saposndioo pazA[viq, ‘TOBN Jueo sed G79 + saposndsog ee ee Ss SS ea a ee ee Gime oat,| we ‘oodH | “oan | wat ey dod | peur | pemwuy sqedag . ate yCoyT IBA: pedo V | -joaep yoo (01) (6) (8) (1) (9) ee ee SE ee ee Ee Se ea “UOLy So10S “OSLL any -viod “Ula, L, (¥) “pozPeNoBvanyp *pasn oinjxtul QO (s) “uorzn 19s (2) “ungo/bowar yyun QO fo uoyovay fo qa py ‘IA GHTAVL ‘ON (1) 481 E. F. Adolph and L. J. Henderson 002 ‘8 | P'9E 098 ‘Z | 9° $1 000 ‘g 00¢ ‘€ | I°ST 080 ‘Z | # ST 66 00r 06 16 16 v6 [(2) — (8)] X (g) X 00L‘9T X (9) X (¥) = (6) :uOTZe[No[wo IO}; BpnUUAIO,T RE ER hr ii he Ae TE eA ee, hae eee eS &80°0 -[ns -+- b60°0 -[ns + 0€0°0 -[ns + 6600 “Ins + “plow o1imny OIULOF WOIE OO "plow orang OIULIOJ WIE OL) “plow olny OIULLOF WOLF OO “plow ory OIWIOJ WOIE OD ‘IOBN Juoo sod 69 + saposndsog ”? ”? *soposndioo pezA|viqy a ee le i *pozeuedAxGO RSME aDh ar vic cs aan ee Ee EL aa Ee ss Rel ee —_, 482 Heat of Reaction of Oxygen The standard deviation of 22 per cent in these measurements is perhaps as significant as the magnitudes themselves. The preliminary experiments with inorganic reactions have indicated that an accuracy of about 1 per cent can be attained. That such an agreement was not obtained with whole blood, and that pre- vious workers on this problem have not agreed, indicates that important influences still uncontrolled are at work. The result obtained for the heat of reaction of oxygen with hemoglobin, approximately +7,000 calories, is lower than that obtained by Berthelot (14), Torup (15), Barcroft and Hill (16), or du Bois-Reymond (17). Reaction of Carbon Monoxide with Hemoglobin Solutions.—Only a few measurements of the heat developed in this reaction were made, using gas generated from formic acid and sulfuric acid. Similar results were obtained with illuminating gas, which is com- ‘posed chiefly of hydrogen, carbon monoxide, and hydrocarbons (Table VI). In all cases hydrogen was used to give a fore period _ (Fig. 4, Curve III). Control measurements with water showed no appreciable heat of solution. The average value for eight measurements was +17,500 cal- ories per gram molecule, 5,600 calories (82 per cent). The heat of solution of a gram molecule of carbon monoxide is +2,800 calories, leaving +14,700 calories. This is very close to the result of Berthelot (14). A series of four experiments in which carbon monoxide re- placed oxygen gave a smaller development of heat, averaging +3,000 calories. The heats of solution of oxygen and carbon monoxide in this case compensate each other. If the compara- tive values for oxygen and carbon monoxide, +7,000 and +14,700 calories are correct, it is to be expected that +7,700 calories will develop when carbon monoxide replaces oxygen. That only +3,000 calories were evolved, indicates the importance of other factors which are either experimental or chemical. Reaction of Carbon Dioxide with Defibrinated Blood—The amount of energy liberated when carbon dioxide combined with solutions of dialyzed hemoglobin varied tremendously, depend- ing upon the ionic equilibria of the solution. But when carbon dioxide combined with whole blood the results were remarkably uniform, even with the blood of different individuals. The course > ‘ ae a a 7 E. F. Adolph and L. J. Henderson 483 Fia. 6. Temperature changes during a thermochemical measurement using carbon dioxide and beef blood. 1 — eS “syUotMOINsvoU 4ySIO-AyUIM} JO OFvVIOAV UB ST poOoTq [euIIOU IOJ yurod OL “OoSS FB OPIXOIp UOTE YQIM poorq Jooq JO Jo}]T T JO UoLyBInyes oy} Aq PoATOAD YvoHT *L “OTT Pir 21990] PEPPY poeld ayrucqae> wnipos paPpPyy N oVo N S90 NsSoo Heat of Reaction of Oxygen - PeAjcAs v3} —] = Salacjvo 484 E. F. Adolph and L. J. Henderson 485 of a typical experiment is given in Fig. 6. The blood was usually laked, octyl alcohol being present. A series of twenty-eight measurements in which carbon dioxide was passed into fully oxygenated blood gave an average result of +513 calories per liter (standard deviation +12 per cent). The defibrinated beef blood used included samples from eight different animals. The reverse reaction in ten measurements averaged —476 calories. When the blood was evacuated and saturated with hydrogen, the passage of carbon dioxide produced +486 calories per liter in six determinations. Interesting results were obtained by adding known amounts of lactic acid or sodium carbonate to the blood; then passing car- bon dioxide into it. The data are plotted in Fig. 7. It will be seen that 0.065 Nn lactic acid was sufficient to exhaust the acid- combining power of the blood. Added carbonate increased the heat production in strict proportion to the available alkali. DISCUSSION. The first worker to measure the heats of reaction of hemoglobin compounds, Berthelot (14), was interested in locating the heat production of the animal body. When he found that oxygen liberated considerable energy in combining with hemoglobin, he supposed that much heat was produced in the lungs, and calcu- lated this to be about one-seventh of the human body’s energy output. He suggested, however, that most of this portion of energy was used in the lungs in vaporizing water. Berthelot supposed that the liberation of carbon dioxide in the lungs in- volved little or no energy exchange. From the results of the present experiments the heat exchange in respiration can be calculated. In saturating 1 liter of blood, carbon dioxide liberates 513 calories; oxygen liberates 84 calories. Complete saturation with oxygen requires an amount of gas which is only one-seventh of the carbon dioxide required for saturation, so that when the respiratory quotient is 1 these two reactions compensate almost exactly in chemical energy change. In the - living body this is approximately true both in lungs and in tissues. The dissociation equilibrium of oxyhemoglobin was first rep- resented mathematically by Hiifner (24). Various discrepancies 486 Heat of Reaction of Oxygen between theory and fact have presented themselves, which have [Eb] [Os [HbOs] Nevertheless, the data of Barcroft and Roberts (25) showed that solutions of hemoglobin can be obtained for which K is constant at a given temperature for all oxygen tensions. The data of Bert (26) and of Hiifner (27) show that the dis- sociation of oxyhemoglobin in whole blood increases with tempera- ture. Barcroft and King (28) have plotted the actual curves at several temperatures, not only for blood but also for dialyzed hemoglobin solutions, and it is readily observed in their data that K, as defined above, is many times greater at 38°C. than at 14°C. The measurements of Barcroft and Hill “ show the same change in dissociation with temperature. Henri (29) was the first to suggest that by means of the iso- chore a relationship might be found for oxyhemoglobin between the temperature coefficient of K and the heat of reaction, Q@. He showed that the heat of reaction ought to be very large, since K increases sixfold for a temperature change of 20°. Barcroft and Hill (16) again calculated this relationship, and found that @ must have a value of approximately 28,000 calories. These authors considered both K and Q as factors in a heterogeneous equilibrium; v72., between gaseous oxygen and dissolved hemo- globin. In this paper their data are recalculated in terms of dissolved oxygen, corresponding to the form of our own data. Calculations based upon the present series of experiments yield an average value for Q of roughly 7,000 calories. This series includes 43 determinations, as contrasted with 24 determinations by four other authors. Individual determinations in this series vary from 1,500 to 10,900 calories, while the results of others vary from 10,200 to 29,870 calories. In Fig. 8 the variation of K with temperature is plotted, assum- ing that Q is 7,000 calories. In the same graph is plotted the experimental variation of K as found by Barcroft and Hill (16). In the first case K increases 1.4 times for 10°, in the second case 3.1 times. K is plotted here in arbitrary units, which represent approximately its true value when oxygen concentration is cal- culated in gram molecules. led to modifications of the Hiifner equation: K = E. F. Adolph and L. J. Henderson 487 It is evident that the experimental values of K and Q are not related by the simple isochore of van’t Hoff. Careful criticism has revealed errors in the technique neither of our measurements of Q nor of Barcroft’s (16, 28) measurements of the temperature co- efficient of K which are sufficiently great to explain away the discrepancy. It is, therefore, a real one. It seems worth while to recognize a number of unmeasured factors in hemoglobin systems. The measured value of Q on the Temperature =iGe Fic. 8. Variation with temperature of the equilibrium constant for the dissociation of oxyhemoglobin, as calculated by van’t Hoff’s isochore. one hand, undoubtedly includes energy derived from the aggre- gation and ionic dissociation of hemoglobin, at least one and per- haps both of which occur during oxygenation. The measured value of the temperature variation of K, on the other hand, prob- ably includes the following factors: 1. Change in the solubility of oxygen with temperature. This is the only one of the factors which can be corrected for, and when calculations are applied to Barcroft’s (16, 28) data, they are altered relatively slightly. 488 Heat of Reaction of Oxygen 2. Change in the aggregation equilibria of hemoglobin and of oxyhemoglobin. While Hill (80) and Haldane (31) have been able to calculate the amount of aggregation of molecules which may take place when hemoglobin gives up oxygen, it is not known how this ratio of equilibrium constants varies with temperature. 3. Change in ionic dissociation of hemoglobin and of oxyhemo- globin. The deductions of Henderson (32) and the experiments of Adolph and Ferry (20) show that probably the ionization of hemoglobin changes its equilibrium with variation in combined oxygen. It is impossible to distinguish this change from the changes due to other shifts in ionic equilibria, all of which are probably influenced greatly by temperature. 4. The redistribution of cations associated with hemoglobin and with oxyhemoglobin. There is little doubt that each salt of either substance differs in its behavior toward oxygen. 5. The redistribution of all ions and molecules not chemically related to hemoglobin or oxyhemoglobin. This is at a minimum in dialyzed preparations. Barcroft (33) has drawn the conclu- sion that there are at least two ways in which the marked influence of salts is exerted; first, by changing the aggregation of hemoglo- bin, and secondly, by changing the ionization of hemoglobin. There may in addition be a direct influence of salts. In the light of these considerations it is perhaps not surpris- ing that simple theory and complex fact are apparently at vari- ance. The discrepancies may be summarized for the present in the statement that the active mass of hemoglobin differs from its measurable properties in its behavior toward oxygen. It is of interest to apply the isochore to the equilibrium be- tween carbon monoxide and hemoglobin. The equation k = Lib] _1CO} 6) best down jhy) Haldahe) (30) jacana [HbCOl as been shown by Haldane g (34) to be subject to the same conditions as that for oxyhemoglobin. By measuring the equilibrium when both oxygen and carbon mon- oxide are present in solution at several different temperatures, they showed that the change with temperature in the k for car- bon monoxide is from 5 to 10 per cent higher than in the K for oxygen. Applying the isochore to their data one would expect to find g for carbon monoxide hemoglobin approximately 8 per cent greater E. F. Adolph and L. J. Henderson 489 than Q for oxyhemoglobin, or about 7,500 calories, if our average for Q is correct. Our measurements of g, in common with those of Berthelot (14) and of du Bois-Reymond (17), show that the actual g is very much larger than the heat of reaction for oxygen, about 15,000 calories. Reversing the use of the isochore, it is found that k increases rapidly with temperature in such a case. Moreover, the experiments in which carbon monoxide replaced oxygen which was already combined with hemoglobin show an evolution of heat, not of 500 calories nor of 8,000 calories, but of 3,000 calories. The influences of acids, alkalies, salts, and dilution upon the dissociation of carbon monoxide hemoglobin are the same as those upon oxyhemoglobin (31, 34). The temperature effect and the energy of reaction are the only chemical properties studied so far that are known to differ for the two compounds; and these two quantities are not correlated in the expected manner in the case of either compound. SUMMARY. 1. The technique and apparatus used in measuring the energy exchange in gas-liquid reactions are described. 2. The thermochemical method is shown to be useful in study- ing (1) the location of animal heat production, (2) the velocity of reactions, (3) the amount of oxygenation and reduction of hemoglobin, (4) the neutralizing power of solutions, and (5) the heat of reaction as applied in the use of van’t Hoff’s isochore and in the measurement of chemical affinity. 3. It is shown that neither the isochore nor the mass law can be applied directly to the oxyhemoglobin system, under the limita- tion of present analytical methods. Several unmeasured factors that occur in every hemoglobin system are suggested as partially responsible for this. This investigation has been aided by a grant from the Elizabeth Thompson Science Fund. BIBLIOGRAPHY. 1. Pfaundler, L., Ann. Phys. Chem., 1866, exxix, 102. 2. Landolt, H., Bornstein, R., and Roth, W. A. Physikalisch-chemische Tabellen, Berlin, 4th edition, 1912. 490 Heat of Reaction of Oxygen * ey 3. Hillersohn, S., and Stein-Bernstein, Arch. Physiol., 1896, 249. 4, Bordier, H., Compt. rend. Acad., 1900, exxx, 799. 5. Bohr, C., and Bock, J., Ann. Phys. Chem., 1891, xliv, 318. 6. Thomsen, J., Thermochemische Untersuchungen, Leipsic, 1882, iv. 7. Berthelot, M., Thermochemie, Paris, 1897. 8. van’t Hoff, J. H., Etudes de dynamique chimique, Amsterdam, 1884. 9. Winkler, L. W., Ber. chem. Ges., 1891, xxiv, 89. 10. Kendall, J., J. Am. Chem. Soc., 1916, xxxviii, 1480. 11. Berthelot, M., Compt. rend. Acad., 1898, exxvi, 1066, 1459. 12. Berthelot, M., Ann. chim. et phys., 1873, series 4, xxix, 433. 13. deForcrand, R., Ann. chim. et phys., 1893, series 6, xxx, 56. 14. Berthelot, M., Compt. rend. Acad., 1889, cix, 776. 15. Torup, S., Festschrift fiir Olof Hammarsten gewidmet, 1906. 16. Barcroft, J., and Hill, A. V., J. Physiol., 1909-10, xxxix, 411. 17. du Bois-Reymond, R., Arch. Physiol., 1914, 237. 18. Camis, M., Arch. ital. biol., 1907, xlviii, 261. 19. Meyerhof, O., Arch. ges. Physiol., 1912, exlvi, 159. 20. Adolph, E. F., and Ferry, R. M., J. Biol. Chem., 1921, xlvii, 547. 21. Van Slyke, D.D., J. Biol. Chem., 1918, xxxiil, 127. 22. Smith, A. H., Dawson, J. A., and Cohen, B., Proc. Soc. Exp. Biol.and Med., 1919-20, xvii, 211. 23. Van Slyke, D. D., J. Biol. Chem., 1917, xxx, 375. ' 24, Hiifner, G., Z. physiol. Chem., 1882, vi, 94. 25. Barcroft, J., and Roberts, F., J. Physiol., 1909-10, xxxix, 143. 26. Bert, P., La pression barometrique, Paris, 1878. 27. Hiifner, G., Arch. Physiol., 1890, 1. 28. Barcroft, J., and King, W.O.R., J. Physiol., 1909, xxxix, 374. 29. Henri, V., Compt. rend. Soc. biol., 1904, lvi, 339. 30. Hill, A. V., Biochem. J., 1918, vii, 471. 31. Douglas, C. G., Haldane, J. S., and Haldane, J.B.S., J. Physiol., 1912, xliv, 275. 32. Henderson, L. J., J. Biol. Chem., 1920, xli, 401. 33. Barcroft, J., The respiratory function of the blood, Cambridge, 1914. 34. Hartridge, H., J. Physiol., 1912, xliv, 22. —————— THE PHYSIOLOGY OF THE PHENOLS. I. A QUANTITATIVE METHOD FOR THE DETERMINATION OF PHENOLS IN THE BLOOD. By K. F. PELKAN. (From the George Williams Hooper Foundation for Medical Research, University of California, San Francisco.) (Received for publication, December 12, 1921.) Liver function and related problems have interested several workers in this laboratory during the past few years. As a part of this research program it seemed advisable to examine critically the conjugation of phenolic substances in the body. The mechan- ism of absorption, conjugation, distribution, and excretion can be studied when a suitable method is at hand. The method described and controlled below is sufficiently accurate to make possible a careful study of these phenolic substances in the blood and body tissues. Benedict and Theis (1) in 1918 made an attempt to apply the urinary phenol method of Folin and Denis (3) to the blood. The principal modification consisted in a separate determination of uric acid by Benedict’s method and an addition of sodium bisulfite to stabilize the color. Their conclusion from an examination of a number of pathological cases was that blood contains an average of 4.70 mg. of phenol per 100 cc., that it contains no conjugated phenols, and that the polyphenols appear to represent about one- third of the total phenols. From all the experimental evidence presented by other workers we must regard as an established fact that the phenols are formed in the intestinal tract by bacterial decomposition of tyrosine, and that they are excreted largely in the urine. Only a small percent- age is excreted with the feces (4). Also, it has been conclusively shown that a large percentage of the urinary phenols exists in the conjugated form (9). Since we have no reason to believe that conjugation occurs either in the kidneys or in the bladder we may 491 492 Physiology of the Phenols. I safely assume that at some time or other the conjugated phenols pass through the blood prior to excretion. It should, therefore, be possible to demonstrate and measure these substances in the blood stream. An examination of Benedict and Theis’ method shows that there are several factors which may well prevent the demonstration of minute quantities of conjugated phenols. The separate determi- nation of uric acid not only complicates the method considerably, but involves an unavoidable experimental error. The boiling of the blood filtrate down to less than half the volume is the process primarily responsible for their inability to demonstrate conjugated phenols in the blood. Benedict and Theis state that only 85 per cent of resorcinol added to the blood is recovered by their method and that phenol itself added to the blood disappears completely during the boiling of the filtrates. There are sufficient reasons to believe (9) that the bulk of the conjugated phenols is made up of two very volatile phenols—p-cresol and phenol—and it is, therefore, logical to assume that the boiling of the filtrates - in Benedict and Theis’ method is responsible for the disappearance of the conjugated phenols. However, it is possible to modify Folin and Denis’ original method so that it may be applied to blood. It seemed important to eliminate heat as a factor in the precipitation of the blood pro- teins and in the concentration of the filtrate. The precipitation of proteins with tungstic acid, as described by Folin and Wu (5) is an excellent substitute for the boiling acetic acid; the precipitation is fully as complete, and the reaction takes place at room tempera- ture. Inordertoremove thelast traces of protein, aluminum cream is added. It was further found that with a standard of 5 mg. of phenol per 100 cc. set at 20 in the Duboseq colorimeter, the color developed in the blood filtrate is easily and accurately readable. This eliminates the concentration of the filtrate by boiling as carried out by Benedict and Theis. Another objection to Bene- dict’s method is the separate determination of uric acid by an entirely different and time-consuming method. In order to sim- plify this step a number of uric acid precipitants, such as Morris’ (8) zine salt and Curtman and Lehrman’s (2) nickel salt were tried, but, while the precipitation of uric acid is complete, the excess zinc or nickel is difficult to remove from the solution, and K. F. Pelkan 493 if not removed interferes with the final color reaction. Folin uses silver lactate in lactic acid as a precipitant for uric acid, but the difficulty with lactic acid lies in the fact that it gives a blue color with Folin’s phenol reagent; a fact which was overlooked by Folin and may in part be responsible for his high figures. Never- theless, this method of precipitating the uric acid was retained, but the error caused thereby is now corrected by addition of a corresponding amount of lactic acid to the standard. Benedict’s innovation of stabilizing the color by addition of sodium sulfite TABLE I, Recovery of Phenols Added to Blood.* After addition of 5.81 mg. Before addition of phenols. p-cresolt to 100 cc. of blood. mg. per 1,000 cc. mg. per 1,000 cc. 24.6 74.0 29.1 78.9 25.6 76.0 Addition of 5 mg. phenol. 25.2 75.0 * The blood proteins must be precipitated immediately after the addi- tion, as prolonged standing destroys a part of the phenols. 7 5.81 mg. of p-cresol are equal in color production to 5 mg. of phenol. The colorimeter reading is expressed as phenol. TABLE II. Accuracy of Separate Simultaneous Determinations of Phenol. Determination. Dog 1. Dog 2. Dog 3. Ls a ee 28.4 36.25 31.0 ECOG ae... . ss 28.75 ay ou 30.6 Whirdes soe. ok: 28.0 37 .50 is omitted. Due to the more complete precipitation of the pro- teins and the higher dilution in which the readings are made, the dirty green tinge which sometimes occurs with Benedict’s method does not appear. Finally another change has been made in the method: due to the higher dilution no addition of water to make up to 50 or-100 ce. as in Folin’s or Benedict’s methods is necessary, and there is substituted a manipulation of the final filtrate in graduated test-tubes which greatly increases the accuracy of the method and makes possible the detection of conjugated phenols in very minute quantities. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 494 Physiology of the Phenols. I Method. The entire method is very simple and rapid. 10 cc. of blood are added to 50 cc. of distilled water in a 100 ec. Erlenmeyer flask. Then 10 cc. of 10 per cent sodium tungstate and 10 ee. of $ N sulfuric acid are added, the flask is closed with the thumb or rubber stopper, and vigorously shaken for a few seconds. To precipitate the proteins completely 10 ec. of aluminum cream are added and the flask is again shaken. The contents are trans- ferred to a 100 cc. centrifuge tube and centrifugalized for 45 minutes. The supernatant fluid is filtered to the 45 cc. mark in a narrow 50 cc. graduate, 5 ec. of a 5 per cent solution of silver lactate in 5 per cent lactic acid are added, and the graduate is well shaken for 1 minute. After centrifugalization and filtration, the filtrate is ready to be examined for phenols. This last step is carried out as follows: only two narrow test-tubes are required —one graduated at 15 cc. and the other at 10 ec.—in which both the total and the free phenols are determined. Thus, any error due to the graduation of two sets of test-tubes is avoided. The procedure for the determination of free phenols is this: the 15 ce. tube is filled to the mark with the filtrate, 1 cc. of the phenol reagent! is added, and the tube is shaken. The excess silver pre- cipitates out, the solution is filtered to the mark into the 10 ce. tube, and 5 ee. of 20 per cent sodium carbonate are added. This solution is now transferred to another test-tube in which the color develops to its maximum in about 20 minutes. The two gradu- ated test-tubes are meanwhile used for the determination of total phenols. The 15 cc. tube is again filled to the mark with the same filtrate, 5 drops of concentrated HCl are added, and the tube is placed in a water bath at 100°C. for 10 minutes. Boiling of the contents of the tube is avoided and no loss of volatile phenols occurs, as was shown by repeating the determination with known amounts of phenol. If the tube has a diameter of 14 to 15 mm., 1 Bell’s modification of the reagent (Folin, O., and Denis, W., J. Biol. Chem., 1916, xxvi, 508) is used, since the HCl which it contains is needed for the precipitation of the excess silver. It contains: 100 gm. of sodium tungstate, 20 gm. of phosphomolybdiec acid, 50 ec. of phosphorie acid 85 per cent, 100 ec. of concentrated HCl. This is gently refluxed for 2 hours with 750 ce. of water, and at the end of the period of heating made up to 1,000 cc. K. F. Pelkan 495 TABLE III. Removal of Uric Acid by Precipitation. After addition of 5 mg. of Before addition of uric acid. uric acid to 100 ec. of blood. Total phenols per 1,000 cc. Total phenols per 1,000 cc. mg. mg. 23.5 23.8 25.5 25.5 29.0 29.5 TABLE IV. Influence of Amino-Acids Normally Present in the Blood. After addition of casein digest Before addition of casein digest. 20:ng. of anne Mito) Total phenols per 1,000 cc. Total phenols per 1,000 cc. mg. mg. 34.2 36.8 31.4 35.0 24.6 27.5 TABLE V. Demonstration of Conjugated Phenols in Human* and Dog’s Blood by the Author’s Method. Blood. Total. Free. Conjugated, 1,000cc. | 1,000 ce. | 1,000 cc. | Ber cent Human. Decompensated heart..... 31.2 28.5 240 8.6 MEAGcUEGEOL ATMS: - os. as BY SB” 4.0 10.7 iam By. . Spas by. so eraeutac 42.8 38.0 4.8 18.2 Dog. MSGIEITIA lee cca 5 «ais eg eee Oe 28.2 1.2 4.1 a2 6 30.1 2.4 7.4 38.1 Soe 2.9 “a6 34.0 Seno 125 4.4 27.4 24.4 3.0 10.9 * This table is not intended to give average values for human blood (for which purpose a much greater number of determinations must be made) but merely to demonstrate the presence of conjugated phenols. 496 Physiology of the Phenols. I the volume of the contents on cooling at the end of exactly 10 minutes is back to the graduation mark, so that no adjustment of volume is necessary. Then 1 cc. of the phenol reagent is added and the solution is treated in the same way as in the determination of free phenols. The difference between the total phenols and the free phenols represents conjugated phenols. The standard is prepared as follows: 5 cc. of the stock solution of resorcinol (Benedict and Theis, 1), containing 5.81 mg. are placed in a 100 cc. volumetric flask, 0.5 ec. of concentrated HCl and 10 ce. of the silver lactate-lactic acid solution added, centri- fugalized or filtered, and the filtrate is manipulated in the gradu- ated test-tubes in the same manner as the blood filtrate in the determination of free phenols. ei ed Recently there have been published several criticisms of Folin’s phenol reagent (6, 7, 9), from which it appears that this reagent gives a blue color with a great many substances, and that it is by no means specific for phenols and the closely allied hydroxy-acids. It must be said, however, that most of these substances are nor- mally not present in the blood, and that the normal constituents of the blood which interfere with the reaction can either be readily removed (uric acid, Table III), or are in such small concentration (amino-acids, Table IV), that their presence accounts, at best, for only a negligible fraction of the color developed. Further, in experiments such as reported in the later papers of this series, where we are dealing with the measurement of injected or ingested phenols, the color-producing effect of these interfering substances is entirely eliminated by a preliminary determination. SUMMARY. 1. A method is described for the determination of phenolic substances in blood, which is based on Folin’s method for the determination of phenolic substances in urine. 2. Contrary to the opinion of Benedict and Theis (1) conjugated phenols are present in human as well as in dogs’ blood and can be demonstrated with the above method. I am indebted to Professor W. R. Bloor and Dr. C. L. A. Schmidt of the Department of Biochemistry for their helpful advice and criticism in the development of this method. OND OV OO ND ps K. F. Pelkan 497 BIBLIOGRAPHY. . Benedict, S. R., and Theis, R. C., J. Biol. Chem., 1918, xxxvi, 95. Curtman, L. J., and Lehrman, A., J. Biol. Chem., 1918, xxxvi, 157. Folin, O., and Denis, W., J. Biol. Chem., 1916, xxvii, 305. Folin, O., and Denis, W., J. Biol. Chem., 1916, xxvi, 507. Folin, O., and Wu, H., J. Biol. Chem., 1919, xxxviii, 81. Gortner, R. A., and Holm, G. E., J. Am. Chem. Soc., 1920, xlii, 1678. . Levine, V. E., Science, 1920, lii, 612. . Morris, J. L., J. Biol. Chem.; 1916, xxv, 205. . Tisdall, F. F., J. Biol. Chem., 1921, xlvi, 409. THE PHYSIOLOGY OF THE PHENOLS. ¢ II. ABSORPTION, CONJUGATION, AND EXCRETION. By K. F. PELKAN anp G. H. WHIPPLE. (From the George Williams Hooper Foundation for Medical Research, University of California, San Francisco.) (Received for publication, December 12, 1921.) With a suitable method at hand for the quantitative estimation of phenolic substances in the blood we took up a study and review of the factors concerned in the normal and abnormal metabolism of phenols in the body, directing particular attention to the liver, Many investigators have studied this question and have mar- shalled experiments to prove or disprove that the conjugation of phenols takes place in the liver, or again in the kidney, or else- where in the body. The best evidence to date is in favor of conjugation taking place in the liver, but in the face of contradic- tory experiments it must be admitted that much of this evidence is indirect and the question is at least open to debate. We may now review some of the recent work which concerns our thesis. Baumann (1) noted a temporary accumulation of conjugated phenols in the liver of a dog poisoned with phenol, and showed that he could obtain nineteen times more phenol-sulfuric acid from that organ than from other organs or the blood. He suggested that the liver is the organ primarily concerned with the synthesis. Christiani and Baumann (4) tied off the ureters of a dog and showed that no accumulation of ethereal suifates occurs in the blood. From this they concluded that, at least when the ureters are tied, the kidneys take no part in the conjugation of phenols. In other and more convincing experiments they ligatured all renal vessels and poisoned the dog with phenol. Synthesis occurred to the same extent as With normal dogs. From these experiments they concluded that if the kidneys are af all concerned with this reaction they are concerned with it to a negligible extent only. In the pursuit of the question of the partici- pation of the kidneys in this synthesis Baumann and Herter (2) had already tried to perfuse the kidneys with blood containing phenol] and sodium sulfate and had been unable to show any synthesis. Kochs (14) ground up liver, kidney, pancreas, and muscle, and added phenol plus sodium 499 500 Physiology of the Phenols. II sulfate. He states that a moderate amount of conjugation was demon- strable in each case, and that negative results were obtained with thymus gland. Landi (15) repeated Kochs’ experiments but could find no con- jugation. From perfusion experiments with some of the organs he con- cludes that the intestine is the seat of the synthetic process. Lang (16) found small amounts of conjugated phenols in the urine of geese whose livers he had extirpated, and thinks that the liver is not the exclusive, although the most important, organ of conjugation. Herter and Wake- man (11) added equal amounts of phenol to blood and ground brain, muscle, kidney, and liver, and showed an increasing disappearance of phenols in the order given. They do not believe, however, that this disappearance is due to a conjugation with sulfuric acid; rather they favor a chemical destruction or a ‘“‘loose combination of the phenol molecules with the molecules of the substance.’’ Salta (19) thought that he could determine the place of conjugation by an analysis of various organs for phenol- sulfuric acid. The largest amount was found in the liver. Then follow in decreasing amounts in the order given: muscles, lungs, intestine, stom- ach, nerves. No conjugated phenols were found in the brain. From this he reasons that all these organs play a part in the synthesis. Embden and Glaessner (6), in a number of carefully controlled perfusion experiments, show that the conjugation takes place almost exclusively in the liver, although very small amounts of ethereal sulfates were found in the lungs and kidneys. The work of Herter and Wakeman (11) is sometimes cited (5) as proof that other organs than the liver, 7.e., the intestinal epithelium, kidney, muscle, brain, blood, etc., have the ability, although to a lesser extent, to conjugate phenols with sulfuric and glucuronic acids. This statement, which is not in accordance with observable facts, is based on a misapprehension of the work of the first named authors. Although their experiments showed a disappearance of phenols when known amounts were added to and left in contact for some time with ground liver, muscle, kidney, etc., there is no evidence that this disappearance is due to a syn- thesis of phenol-sulfuric or glucuronic acids. In fact, Herter and Wakeman themselves state! that “the synthesis of indoxyl potas- sium sulphate cannot be accomplished by extirpated cells. It also seems improbable that the dead cells convert phenol into phenol-sulphuric acid.” They then attempt to show? that this destruction of phenols is not in the nature of an oxidative process since they were unable to recover such oxidation products of phenol 1 Herter and Wakeman (11), p. 317. ? Herter and Wakeman (11), pp. 317-318. K. F. Pelkan and G. H. Whipple 501 as hydroquinone or pyrocatechol. By subjecting liver pulp to various ferment-destroying agents (alcohol, boiling water, hot air, bichloride of mercury) and finding no decrease in the behavior towards phenol, these authors consider it highly improbable that ferments play any important rdle in this disposal of phenols, and they finally speculate? as to the probability of some “loose com- bination of the phenol molecules with the molecules of the cell substance.” It should be strongly emphasized that there is little experi- mental evidence for the statement that other organs than the liver have this ability to conjugate phenols. If the transforma- tion of phenols by dead tissues is due to conjugation, it should be possible to hydrolyze these conjugated phenols by the addition of TABLE VI. Blood and Liver Incubated with Phenols. Blood + 5 mg. phenol Same hydrolyzed Blood blank. per 100 ce. and Same -+ 5 me. with 5 drops 2 hrs. incubation. K2S20s per 100 ce. , HCland heat. mg. per 100 cc. mg. per 100 cc. mg. rer 100 ce. mg. per 100 cc. 2.8 7.6 f 7.4 7.45 o.2 7.9 8.2 ak: Liver as above. 6.6 9.8 7.25 10.5 10.0 9.4 10.6 a few drops of HCl and the application of heat, and so recover the total amount of added phenol. That this is not the case is shown in Table VI in which are recorded the results of an experiment similar to those of Herter and Wakeman. Instead of using Millon’s reagent the method described in the first paper of this series was used. Known amounts of ingested phenols cannot be completely recovered in the urine or feces, and this loss is probably due to a similar reaction by which organ pulp disposes of phenol, 2.e., chemical oxidation, ferment action, or Herter and Wakeman’s “loose combination.” 3 Herter and Wakeman (11), pp. 319-320. 502 Physiology of the Phenols. II Tauber (22) fed phenol to a dog and found that as the dose was decreased, the amount of phenol so ‘‘oxidized”’ was increased. On feeding 240 mg. of phenol in water per os, he found that 53 per cent of the ingested phenol was oxidized in the body through oxalic acid to carbon dioxide. Dubin (5) found that 68.7 per cent of ingested phenol and 50.6 per cent of p-cresol could be recovered from the urine. When tyrosine was given, only 17.7 per cent of the amount of phenol corresponding to the amount of tyrosine administered was recovered. Friedlinder (9) recovered only 25 per cent of ingested cresol from the urine of his dogs. Numerous other workers have noted the fact that ingested phenols cannot be quantitatively re- covered from the excretions. Siegfried and Zimmermann (20) fed p-cresol to a dog and recovered both p-cresol and phenol, the sum of both amounting to 32 to 48 per cent of the p-cresol fed. Jonescu (13) administered cresols to dogs and noted that they were oxidized in the order of ascending toxicity. The percentage oxidation was as follows: m-cresol 50 to 53 per cent, o-cresol 65 to 69 per cent, and p-cresol 73 to 76.5 per cent. As an example of the powerful oxidation to which phenols may be subjected in the organism, Jaffé (12) has shown that the ingestion of benzene increases the urinary output of muconic acid, and suggests that this is due to the cleavage of the benzene ring into a straight chain compound. The whole subject of oxidases and phenolases is exhaustively treated in Oppenheimer’s text- book (18) and need here not be further considered. (‘‘Die Wirkung dieser Phenolasen ist neben der Oxidation, die niemals tiefgreifend ist, haufig eine Kondensation mehrerer Molekiile. Das hingt mit ihrem physiolo- gischen Zwecke der Schaffung unléslicher Stoffe zusammen.’’) This oxidizing action on the phenols which has been noted by Tauber to increase as the amount of volatile phenol present de- creased, would explain why normally there are such small amounts of conjugated phenols in the blood: the volatile phenols produced in the intestinal tract are probably to a large extent rapidly oxi- dized by the intestinal mucosa and the liver—and those which escape this process are completely conjugated by the liver. When the oxidizing mechanism is overwhelmed by large amounts of phenol, as occurs when phenol is ingested, more of that substance reaches the hepatic tissues unchanged and the relative amount of conjugated phenols is greater. Ordinarily this oxidizing mechan- ism is sufficient and much larger amounts of volatile phenols than are normally produced in the intestine are oxidized within a few minutes. This serves to strengthen the argument that free volatile phenols as such cannot remain in the blood stream for any considerable period. It now appears probable that in its effort to dispose of phenol which reaches the circulation from the intestinal tract, the body K. F. Pelkan and G. H. Whipple 503 makes use of two methods: one a process of oxidation which takes place to a considerable extent in the epithelial lining of the intes- tinal tract and in the liver and to a lesser extent in the other organs; the other a process of conjugation with sulfuric and glucuronic acids which takes place, as we shall later show, exclusively in the liver. As regards the nature of the phenolic substances in the blood, there is no reason to suppose that they are essentially different from those in the urine. The most important literature dealing with that subject has been summarized by Folin and Denis (8), Dubin (5), and Tisdall (23). EXPERIMENTAL OBSERVATIONS. It seemed necessary to control diet factors in these experiments because of the observations of Underhill and Simpson (24) to the effect that urinary phenols vary directly with the protein intake, and with an increase in output of total phenols the percentage of free phenols remains constant. Some similar observations are recorded in Table VII which gives controls for the various diets under laboratory conditions. Dogs weighing 20 to 35 pounds were used in these experiments. Mixed diet indicates a liberal amount of mixture of table scraps including cooked meat, macaroni, potatoes, bread, and bones. The most important volatile phenols present in the body are p-cresol and phenol (7) and we, therefore, limited out experiments to these two substances focussing our attention upon p-cresol, as we found that complete conjugation occurs with it more rapidly than with any other phenolic substance investigated. At first we injected known amounts of phenol in a watery solution into the jugular vein of the animal and analyzed for total and free phenols the blood drawn just prior to and at stated intervals after the injection. Charts A and B give typical results of some of these early experiments. These experiments show that 5 minutes. after the injection of 500 mg. of p-cresol or phenol the “theoretical concentration” of about 500 mg. per 1,000 cc. of blood has fallen to 80 mg. per 1,000 cc. We say “theoretical” advisedly, for the disappearance of phenols from the blood stream is so rapid that, due to the time lost during the slow injection this concentration is never approximated. The dog weighed 24 pounds and we may 504 Physiology of the Phenols. II assume an approximate blood volume of 1,000 ce. This almost instantaneous disappearance of injected phenols is, we believe, due to two factors; a rapid and relatively uniform distribution through- out all tissues of the body, and oxidative processes also occurring throughout the body although not with equal rapidity in all tissues. ‘The amount of conjugated phenols in the blood 5 minutes after the injection is only slightly greater than before injection, but it increases rapidly to a maximum about 1 hour after injection. At the end of the hour the free phenols have reached their pre- TABLE VII. Diet Factors and Blood Phenols. Total phenols innate Total phenols Gaua a Doc. | pep k piece phase, | Per uumee |) aan Exclusive carbohydrate diet. High protein diet. mg. per cent mg. per cent 21-56 26.5 Ae 39.4 8.0 21-95 23 .4 55 42.2 7.0 21-95 24.0 6.4 45.9 8.4 21-98 SL67 7.4 38.8 6.6 21-80 25.0 8.0 43.0 7.8 5 to 8 days fasting. _ Mixed diet. 21-98 28.1 9.0 36.8 4.0 21-105 2OnD 6.3 37.9 6.1 21-80 29.0 “a0 28.2 11.0 21-80 32.4 7.4 35.0 5.8 injection level and the conjugated phenols fall slowly (6 to 12 hours) back to their normal level. As these curves are typical of a large number of determinations made on different dogs, it can be safely concluded from their examination that p-cresol is the more quickly oxidized or conju- gated substance, for the fall of the free phenols back to the original level occurs sooner in every experiment. The injection of such large amounts of highly poisonous sub- stances invariably produces severe systemic reactions. Immedi- ately after injection the pupils are widely dilated, the animal has convulsions and is unable to stand. In most cases the animal becomes clinically normal within 5 minutes. In some cases K. F. Pelkan and G. H. Whipple 505 animals which have previously reacted moderately to repeated injections will not survive the unit dose. These intravenous injections are always dangerous and may. destroy a valuable standardized animal. Because of this difficulty we examined the OD R S) (O)¢ Tey) os wo SS. CIP iSO Oe (Oe dn. (Crt Phenol in t000 cc. blood Ee ur Co cy ol (NOM ON Sy LS) CuartT A. reaction to phenol solutions administered by stomach tube, and found that doses lethal for intravenous injection often produced no symptoms when given by stomach. An examination of Chart C shows that absorption from the intestine is sufficiently rapid and regular to be substituted in our experiments for intravenous in- 506 Physiology of the Phenols. II jection. In fact, the ingestion of phenols is superior to the in- jection as it will be seen that complete conjugation of ingested phenols is more conspicuous. This, no doubt, is due to the fact that all the ingested phenols must pass through the liver before v i Nh) fae) qn 6 3G = ‘ige | fi ye NIK a a | eee es ee 2 tel = 0 ees |r | po = O o . = a : 70+ Se eel : ae 4 Q 20 a. | fI-~CLESOL woo ig & S| ie || ail 9) Pee Hrs. 0 — ayn rae) mr A “4 Cuart B. reaching the general circulation, while injected phenols are first distributed to all the tissues of the body and reach the liver much more slowly. The slight irregularities which sometimes occur in the ascending limb of the curve may be due to differences in stom- ach and intestinal absorption. In this connection it is interesting EE K. F. Pelkan and G. H. Whipple 507 to recall the statement of Sollmann and Hanzlik (21) that the absorption of phenols is not a steady process, because of a slowing of the local intestinal circulation by these substances. A standard dose of 50 mg. in 10 ec. of water per pound body weight given by stomach tube has been used in all experiments reported in this paper except where otherwise stated. A closer examination of the curves obtained by plotting the results of consecutive blood analyses reveals the following: immediately after ingestion both the total and free phenols begin to rise—the free somewhat slower than the total. The total phe- nols reach their greatest concentration between 3 and 1 hour after Phenol in 1,000 cc. blood Cuart C. ingestion and after that time decline very slowly (6 to 12 hours) to their former level. The free phenols, after a slight rise within the first 20 or 30 minutes, quickly return to their old level (4 to 1 hour). In many cases the free phenols fall somewhat below that level. The resulting differences between the total and free phenols represent conjugated phenols which are plotted at the bottom of the charts. It will be seen that the conjugated phenols reach their greatest concentration 1 hour after ingestion and at this time the conjugation of ingested phenols is complete; an obser- vation which corroborates Baumann (3), who showed that 1 hour after the intravenous injection of phenols the free phenols dis- 508 Physiology of the Phenols. II appeared from the system. After reaching the greatest concen- tration within 1 hour, the conjugated phenols are slowly excreted, corresponding to the slow fall of the total phenols. From the fact that in a healthy animal the conjugation of very large amounts of ingested phenols is complete within 1 or 2 hours, it would appear unlikely that free volatile phenols are normally present in the blood. The conjugation of the small amount of volatile phenols produced in the intestine must be very rapid within the liver, and it is improbable that any free volatile phenols reach the general circulation under normal physiological conditions. The bulk of the blue color (90 per cent and more) which is pro- duced by Folin’s reagent with the blood filtrate is due to substances other than true phenols. It may be due, among others, to hy- droxy-acids and unidentified protein decomposition products as well as to carbohydrates and related substances (10, 17); the remainder of 10 per cent or less is produced by the conjugated phenols. These three tables (Tables VIII, IX, and X) supplement each other and bring out several interesting points. The various organs compared with blood (Table VIII) contain substances which react with the phenol reagent and give larger figures for their “phenol” content. We have no reason to suppose that these reacting substances are true phenols. The injection of phenol (Table IX) gives an immediate increase in phenol content of the blood and tissues. There is probably a pretty uniform distribution of the phenols in the blood and tissues but we record in analyses about 40 mg. increase in blood and 25 to 30 mg. increase in the parenchymatous organs. After an hour’s interval following a phenol injection (Table X) we note a uniform distribution of conjugated phenols in the blood and organs. From these facts we wish to assume that free phenols, when injected into the blood stream, are distributed throughout the living tissues. After conjugation by the liver they again diffuse out to the tissues before being excreted (Table X). When phenols are given by stomach no such distribution of free phenols occurs because they are conjugated by the liver before reaching the gen- eral circulation. This last fact explains why doses of phenol, K. F. Pelkan and G. H. Whipple 509 lethal when injected into the blood stream, show no effects when given by stomach tube, but are lethal once again when adminis- tered by stomach tube to a dog whose liver has been seriously injured. TABLE VIII. Distemper Dog, Killed with Chloroform. Organ. Total phenols | Free phenols Conjugated per 1,000 gm. | per 1,000 gm. phenols. mg. mg. per cent ot Ras 22.3 20.8 1.5 ME ac pana saw acta 65.8 65.8 0.0 en cetera 71.3 70.1 1.2 0 ee 59.4 57.9 1.5 TABLE IX. Dog Killed by Injection of Phenol into Jugular Vein, Death 5 Minutes after Injection. Total phenols | Fr henol j Organ. per Lp gen. pee LUE |. raikereae ; mg. mg. per cent RISTO. , | Sic «01a bid be SEES 60.0 56.6 3.4 oS Ee Pee er ee 92.4 91.2 1.2 ME, © o6aio ds ov Magee ee 96.8 90.1 6.7 . -. See 93 .2 92.2. 1.0 EME es. cs ss oa cc ain Soren Leia 88 .4 88 .4 0.0 TABLE X. Dog Killed by Chloroform 1 Hour after Injection of Phenol. Total phenols | Free phenols Conjugated Organ. per 1,000 gm. | per 1,000 gm. phenols. mg. mg. per cent EL D00| ss! i a ee tee ls ae 50.0 3a.0 16.7 RRR Cc hia. ds. Learns a" 117.6 100.0 17.6 USVI a ok SR gs 2 ee Re 117.6 100.0 17.6 SERER es it 59.4 54.1 8.9 Apr. 21. Eck fistula operation. Apr. 27, p-cresol ingestion. Before ingestion............... 21.1 20.7 1.4 After fe Le TS re 25.0 24.4 2.4 oon Lo OOS Gee ee ye 28 .4 26.70 6.0 17.8 30 AS 5 io ae ape ire For ee 30.0 27.0 10.0 29.2 1 hour.. 30.7 26.7 in 37.4 = DR ee ae ere 2 30.0 24.7 ATT 55.0 fe Oa eee ee eee 7a | 22.4 10.8 57.5 May 13. p-Cresol ingestion, conjugation much decreased. Before ingestion............... 35.0 34.0 2.8 After ee POLIMINULES. 5.55240. i eee 37.8 36.2 4.2 21.4 DUS toe aa 39.4 37.0 6.1 31.8 A lS 2. se, Se 39.4 Sane 9.4 61.3 1 hott ec sae 42.0 40.0 5.0 14.3 MENSITS S255. . ooo ee eet 41.0 39.0 4.9 16.6 516 Studies of Liver Function. III TABLE. XI—Concluded. Conjugated F : Total Free . Conjugation Time. phenols per | phenols per phescie of added 1,000 ce. 1,000 ce. ae rete phenols. May 20. Ligation of hepatie artery. May 24, p-cresol ingestion. mg. mg. per cent per cent Before imeestion......- «cece 27.0 25.0 7.4 After " 1S acau balls bee ARREARS eo cic Bic one 315), 7/ 3Bd.7/ 5.6 0.0 SOM ses sss Ee 41.5 39.0 6.0 3.4 112 hYo\b a See MR ec cio So <5 45.4 42.6 Gao, 4.3 DAN OUTS S58 dais ce.cpnacle See 47 .6 45.2 on 2.0 AN. cc erotic aioe Sees 40.8 36.7 10.0 15.2 May 25. Death. Hepatic insufficiency Experimental Protocol of Dog 21-100 (See Table XI). Mar. 25, 33 lbs., healthy. Standardized with 1,650 mg. of p-cresol after 4 days carbohydrate diet. No reaction. Apr. 16, 75 minutes chloro- - form’after 4 days fasting. Apr. 18, 35 lbs., 1,750 mg. of p-cresol, no reac- tion. Conjugation about one-half normal. Apr. 21, Eck fistula operation. External surface of liver shows that the chloroform injury has not been completely repaired—the centers of the lobules are hyperemic and stand out distinctly from the opaque periphery. Apr. 27, 30 lbs., 1,500 mg. of p-cresol. 4 hour after ingestion the dog is severely intoxicated. Conju- gation about one-third normal. Operative wound is partly open and infected. May 1, wound sewed up. May 8, wound again open and sewed up. May 18, slight muscle tremors—prominent distension of abdominal cutaneous veins. 32 lbs., 1,600 mg. of p-cresol. No reaction, but very weak. Conjugation about one-third normal. May 20, no tremors, no ascites, collateral circulation less prominent than on May 13. 30 Ibs. Arch of hepatic artery tied in two places at 11 a.m. 1,500 -mg. of p-cresol by stomach tube immediately after operation. Very severe reaction which may, in part, have been due to the ether. Temperature at 1 p.m. 36.2°; at 2 p.m. 37.2°; at 3 p.m. 37.9°. Conjugation one-fourth normal. May 24, very weak and slight muscle tremors. Hematocrit red cells 38 plus. 25 Ibs. 1,250 mg. of p-cresol, severe reaction. Conjugation less than one-fourth normal. May 25, died 5 days after ligation of hepatic artery. Autopsy.—May 25. Thorax, heart, and lungs normal. Spleen fibrous and rather pale. Serous surfaces clean except plastic adhesions and yellow- ish fibrinous exudate about site of hepatic artery operation. Stomach and intestines not abnormal. Kidney and pancreas negative. Hepatic artery ligated in two places and completely occluded. Eck fistula about 6 mm. long and clean. There must have been consid- erable flow through this opening or it would have closed. The ligature K. F. Pelkan and G. H. Whipple 517 above the Eck fistula on the portal vein did not completely occlude the lumen. Lumen probably about 1 mm. in diameter. The knot must have slipped before the second tie was made. Liver is atrophic but not exactly like the usual Eck fistula specimen. It is not as translucent. The lobules are small and brown at the margins— yellow in the centers, probably necrotic. Some areas are swollen, yellow, and opaque. The lobules here are larger and yellow—necrosis probable. These areas are not numerous, but pretty sizable, 2 X 2 X 5 or 6 em. for the largest—the volume is estimated as about one-tenth or less of the liver parenchyma. Microscopic Examination.—Spleen, much pigment and phagocytes. Liver, some sections show extensive hyaline necrosis, involving the centers of the lobules up to 90 per cent in extent—few liver cells escape at the mar- gins of the lobules. Other sections show central atrophy alone (usual Eck picture) with phagocytes full of lipochrome pigment. Others show some evidence of repair, enlarged liver cells, mitoses, etc. Other organs of no importance. Most instructive are experiments performed on two Eck fistula dogs. Standardized on Mar. 25, Dog 21-100 showed 95 per cent conjugation within 30 minutes (Table XI). On Apr. 16 a liver injury of about 60 or 70 per cent was produced by chloroform. Table XI, experiment on Apr. 18, shows slightly less than 50 per cent reduction of function. On Apr. 21 the Eck fistula operation was performed. On Apr. 27 the test with p-cresol revealed a one-third normal conjugation. A repetition of this test on May 13 shows a slight improvement in function. The arch of the hepatic artery was ligated in two places on May 20 and on testing with p-cresol we found that there remained less than 3 per cent of the original capacity of the liver to conjugate phenols. Death of the dog was due to liver insufficiency. The story of the other Eck fistula dog is similar. Table XII shows that Dog 21-105 exhibited perfectly normal conjugation when standardized on a carbohydrate diet. On Apr. 5 we performed an Eck fistula operation and a conjugation test on Apr. 12. It appears that only 27.8 per cent of the added phenols had been conjugated 2 hours after ingestion, whereas before the operation 100 per cent conjugation occurred in 20 minutes—the function of the liver on Apr. 12 we may estimate as about 10 per cent of normal. On Apr. 20 another test was made and some improvement noted. The conjugation at this time amounted to about 25 per cent of normal. On May 16 we ligated the hepatic artery in one place, leaving some of the collaterals to the liver patent. It will be seen that the liver function after this operation amounted to somewhat more than 5 per cent of normal. TABLE XIl. Eck Fistula Experiments. ae Dog 21-105. Black, female. Conjugated A 5 Total F . (Gi Time. heme per phenol per Lares of adage F 1,000 ce. 1,060 ce. of total. phenols. Mar. 31. Standardization on carbohydrate diet. mg. | mg. per cent per cent Before ingestion............... 24.3° (ODS 3.7 After ee 10aminutes!:...). ose Ses 26.0 200 9.6 94.0 SD ad 6 ooo sss cee 28 .2 2250 22.0 100.0 SOniaen) | c.. | ee et 29.0 2285 22.4 100.0 [Ghoury) tik OLA eee S580 21.8 Sat 100.0 Dhours! 9) °.3 3 SE Nees 33.0 23.0 30.0 100.0 Apr. 5. Eck fistula operation. Apr. 12, p-cresol ingestion. Before Mgestion..«:. 2% tae - 25.5 24.7 3.1 After Ke LOE MInUMbCS en eet ee ashi 31.2 LOA 14.4 D200. ier. EA 2 Bet eae rene ee ye 34.2 3202 5.8 13.8 SO ee 8 os een er eee 36.8 34.5 6.2 ise Whours 348 19. 9 ae ee aoe She Gi 34.5 8.5 19.6 DiWOurse. oh, a Pe 36.3 8225 10.4 27.8 Apr. 20. 15 days after Eck fistula operation. Before iIngestionceas-eeeeeee ae 27.0 26.7 ibeal After ss VORMIMUWtESs../.).cae eee cee 38.0 30-2 4.7 13.6 DOGS mph Sade chee 38.0 Buyeil AG 23.6 11) AR ears eet Bo Gets ae St 42.5 38.5 9.4 23.9 TL hOUrRe eke eee 42.0 Soni 15.0 40.0 DPN OUTS aes cote cen cian ene By 0) 30.0 18.9 67.0 Seite Srey 2b ae Se Se) P2590) 25:3 100.0 May. 16. Immediately after partial ligation of hepatic artery. Conju- gation greatly decreased. Before ingestion............... 26.7 26.0 226 After . AS aminutesss. 2a. ho. ee ee 43.5 Alay, 4.1 6.5 SOM) billy: at. iesl ae 47.0 44.7 4.8 7.8 : MEN GUE F882 iso ee ee 5215 48.1 8.3 14.3 LARNOUTASE. 52. 24d! the eS 52.0 46.9 9.8 17.3 ; OA) ae Shade Sn On ek ae ate 8 40.9 35.0 14.4 36.6 : 5 VDT se idiot atye 37.0 30.2 18.3 59 2 May 18. Death. Hepatic insufficiency and peritonitis. 518 gee K. F. Pelkan and G. H. Whipple 519 Experimental Protocol of Dog 21-105 (See Table XII). Mar. 27, carbohydrate diet begun. Mar. 31, 29 lbs., slight distemper. Standardized with 1,450 mg. of p-cresol. Noreaction. Apr. 6é, Eck fistula operation. No food. Apr. 6, water and bread. Apr. 7, regulation diet: carbohydrates, vegetables, kaolin, and bones. No meat or milk. Apr. 12, 26 lbs., slight distemper. Moderate ascites. 1,300 mg. of p-cresol by stomach tube. Moderately severe reaction. 4 hour after ingestion dog falls down occasionally when trying to walk. 1 hour after dog does not fall down but is still very shaky. Phenol conjugation is about one-tenth normal. Apr. 20, 30 lbs., no distemper. Severe ascites. Collateral circu- lation on abdominal wall is prominent. 1,500 mg. of p-cresol by stomach tube. Moderate reaction. A half hour after ingestion the dog sways from side to side when trying to walk but does not fall. Phenol conjuga- tion about one-fifth normal (a little better than on Apr. 12). May 16, 26 lbs., moderate ascites. Ligature on hepatic artery in only one place, 1,300 mg. of p-cresol. Moderate reaction. Dog has apparently recovered late in afternoon. Phenol conjugation less than one-tenth normal. Hema- tocrit red cell 42 per cent. May 18, dog died 32 to 38 hours after last operation. Autopsy.—May 18. Ascitic fluid less than at operation—now about 100 cc. The fluid is turbid and the serous surfaces are injected—there is a perforation in the first third of the duodenum and shortly before death there evidently had been an escape of intestinal contents with recent peri- tonitis. Thorax, heart, and lungs negative. Blood clots normal. Spleen small and fibrous; not much blood. Adhesions about the site of the first operation are numerous and bled easily at the second operation. Liver small, decidedly yellow, due to fat. Lobules show much injury (fat) and perhaps necrosis—there is no edema and the veins are clear. Fistula is about 5 mm. in length and is less than one-half its original length. The edges are smooth and there is no thrombosis. The passage of blood through it was obviously difficult and caused the portal stasis, development of collaterals, and ascites. Intestinal tract negative in general. Duodenum shows a sharp punched out area about 3 X 1 em. due probably to ligature of the hepatic artery. Kidneys negative. Microscopic Examination——Much fat in liver cells (large and small droplets), central necrosis is abundant—about one-third of cell lobules. Kidneys negative. Death explained by liver injury plus terminal duodenal perforation. Chloroform poisoning gives considerable information as to the conjugation of phenols in the liver. It also points out the fact that this method is inadequate to measure the high limits of liver capacity. For example, given a chloroform injury of approx- imately one-third of the liver parenchyma we may expect a normal or almost normal conjugation of phenols after administration of 520 Studies of Liver Function. III the unit dose with the usual routine technique. This means that our test does not reveal the maximum capacity of the normal liver in conjugation of phenols. When we have a chloroform liver injury of one-half to two-thirds of the liver lobule we can then demonstrate a considerable impairment of liver conjuga- tion (compare Table XI, Experiment of April 18). With more advanced liver injury we note a great drop in liver conjugation (compare Table XIV, Experiment of April 22). Finally with a lethal chloroform injury involving the greater part (90 per cent or more) of each liver lobule (Table XIII, Dog 21-113) we observe zero liver conjugation. Phosphorus injury gives a similar picture of impaired liver function (Table XIII). Itis more difficult to estimate the amount of liver injury in phosphorus poisoning as cell necrosis is not a conspicuous feature of the injury. We note two experiments in Table XIII to show liver function impairment. The injury was fatal in one experiment and very severe in the second animal, yet there was a certain amount of phenol conjugation in both experiments. We observed no impairment of liver function as shown by phenol conjugation in bile fistula dogs. It is interesting to recall that one liver function test (phenoltetrachlorphthalein) shows a dis- tinct fall in output in bile fistula dogs indicating that the chronic cholangitis which is usually present in such animals interferes with the elimination of phenoltetrachlorphthalein but not with the conjugation of phenols. Another important control is given by dogs sick with distemper, as it might be argued that intoxication of any sort might disturb this conjugation of phenols. One animal (Table XVI) was pro- foundly prostrated with acute distemper and died shortly after the completion of the experiment yet a normal conjugation fol- lowed the administration of the unit dose of p-cresol. Autopsy showed the familiar lesions of acute distemper and nothing else. Under these conditions therefore the capacity of the liver to con- jugate phenols is in nowise impaired. From an examination of Table XI, Experiment of April 18, it will be seen that following chloroform injury the initial concen- tration of total phenolic substances in the blood is more than twice, and, in some cases (fatal chloroform injury, Table XIII), K. F. Pelkan and G. H. Whipple 521 > TABLE XIII. Chloroform and Phosphorus Injury of Liver. Dog 21-113. Mongrel, female. Conjugated . : Total Free : Conjugation Time. phenols per | phenols per Leos of added 1,000 cc. 1,000 ce. a total: phenols. May 2. Chloroform anesthesia, 1} hours. Liver necrosis almost total. Conjugation zero. May 3. p-Cresol by ingestion. mg mg. per cent per cent Before ingestion............... 103.0 99.2 3.0 After ee 1.) 0G 111.0 107.0 3.6 25 Lo es eee iD 117.0 37 3.8 a, Eg eS ee eee 124.0 121.0 2.4 0.0 _ i. 2 ee ee 135.0 131.0 3.0 0.6 0 ee ee 133.0 130.0 Ze 0.0 Dog died 2 days after injury. Microscopic examination shows only 1 to 2 rows of surviving liver cells about portal veins. Dog 21-112. Black, male. Phosphorus injury, fatal. Conjugation impaired. Before ingestion............... 42.5 42.0 ee, After se Li) Tc ee 49.5 47.0 S25 28.5 ERM TS... ca cen ene Gy hel, 47.6 7.6 BY fr SE ee ee 52.6 46.0 12S 60.4 L LDOVhA. a3 ele 60.5 50.0 17.3 Seo 2S IL 5 eee Rares 58.5 38.5 34.2 100.0 2 mg. of phosphorus per pound body weight, in oil subcutaneously. Test made 1 day after injury. Dog died 36 hours after injection. Autopsy shows fatty degeneration of liver. Dog 21-76. Mongrel, male. Phosphorus injury, sublethal but severe. Conjugation impaired. Before ingestion............... 26.7 26.0 220 After re MermInutes... 2... 5. eens 32.6 27.0 EGE1 83.0 meee <> SL rredirseeweress 35.6 29.2 10% 61.3 3) GiMRISSeee Pe 2 ee 40.8 31.5 22-4 61.0 _ LS Tee 40.0 28 .2 29.5 83.5 LS A ee 34.2 25 .2 26.3 100.0 2 mg. of phosphorus per pound body weight, in oil subcutaneously. Test made 2 days after injury. Dog recovered very slowly. 522 Studies of Liver Function. III four times as high as in the blood of fasting normal dogs or dogs ona carbohydrate diet. We have no definite knowledge regarding all the factors responsible for this pronounced increase, but it is TABLE XIV. Chloroform Liver Injury and Controls. Dog 21-80. Brown, female. Conjugated Time. Penn Ag enone per phenols in yeh viet 1,000 ce. 1,000 cc, | Percentage | phenols. Apr. 15. Standardization after 4 days fasting. mg. mg. per cent per cent Belorengestion.s: ..(eateree eee 30.8 29.4 4.5 After “¢ OMMIMUbeSs. cet ee ene Sa 30.2 14.2 98.0 P10) te LMR D OME IEG SA ane pape 0 37.8 30.2 20.1 89.0 oie ccs | be) Miedt ie oem penataae 38.5 28.0 27.8 100.0 NOUR Acie ak acotaae cana 40.0 28.5 28.7 100.0 PON OULSs ea cnco65 vena stk Rete 36.4 28.2 22.5 100.0 Apr. 20. Chloroform anesthesia 1 hour, 20 minutes. Apr. 22, p-cresol by ingestion. Beforeingestion.s... se: aces 31.3 30.0 4.1 After sf 1Omimiuties:: sos 05. he eee 43.5 41.0 5.7 9.8 Zia aegs nil oor ghetieths sk awa 48,5 43.6 10.1 20.9 Se aa Rees cme ee 47.8 41.8 12.5 28.4 thOUR, oh. = «ogee... bet ae hee 46.5 40.0 14.0 34.2 DZ HOUMA, ee. ake siete eee 40.0 33.0 17.5 65.2 Effect of repeated bleedings. Apr. 28. Water given instead of phenol solution. Before ingestion.......... Pte 27.2 26.8 1.4 After os UOseowanhiS ey ey eee Sone ea co” 27.0 26.9 0.4 DO) os oeerae reenter 26.8 26.8 0.0 UR ee ns Re theotpettece Ate 2ao PH fea? 1.0 UaROUP. tetas roe eee toes 26.9 PAVETS 0.7 PAMOUESstntye cece he Da 26.5 26.4 0.4 interesting to speculate about some of them. It has been shown by Dubin (3) that in a fasting dog the concentration of urinary phenols, after an initial drop rises considerably. In our cases K. F. Pelkan and G. H. Whipple BUR TABLE XV. Bile Fistula Control. Bile fistula dog (old white bull, female). Conjugated : . Total Free : Conjugation ime. henol henol phenols in f added re Phenols per | phenale per | Percentage | ot edded Before ingestion............... 22.0 21.5 onc After & MORNIN GOS et orn cisieicics cin fs oe ol .2 28.3 9.3 20 3! eA SP PIETDNG » ite oe 34.5 22F 34.2 89 1S 010i: cae ele eee a Oe 31.1 2227 27.0 86 2) ENCES Eee 30.8 20.8 32.4 100 Bile fistula of over 2 years duration—dog normal. TABLH XVI. Acute Distemper Intoxication—Control. Dog 21-98. Black-brown, long haired female. y Conjugated : : Total ¥ : Time. tiecmie per Snanols per Lair aes ete ce a 1,000 ce. 1,000 ec. Shtotall phenols. Mar. 23. Standardization. mg. mg. per cent per cent Before ingestion.:............. 28.3 27.8 lee After fe TOFmmMutest jis evs aes 37.0 3a.8 10.0 36.5 2\0) | GUS Bas MS Peon Pr es 43.1 37.4 13.2 34.7 210) "a ees ee ee 47.1 40.0 15.0 35.1 LL LVS IES ee a a ae OST AES 41.7 28.9 30.7 91.7 BITES TINS Siecle. ¢-.1s.«.c, cymes Sia 25.4 32.6 100.0 Apr. 13. Terminal distemper. Detore ingestion............... 29.0 28 4 20) After - HObminutes.<)). 972 eee eek 38.0 33.0 listed | 48.9 OA) - ESSERE SES stn en a 42.4 37.0 12h 36.0 Sil). SME eS onc pertctcneer 45.2 38.3 15.2 38.9 REL OUIT © 5 os, bic! ohare 40.0 29.8 20.0 | 87.2 Very severe distemper, toxic condition, slight reaction to cresol after 10 minutes, from which dog completely recovered in 30 minutes. Died of distemper after 1 hour. Postmortem: No pneumonia. 524 Studies of Liver Function. III of chloroform injury the period of fasting (8 to 4 days) was certainly not long enough to account for the tremendous increase; this is further controlled by the normal fasting dogs (Table XIV and others) in which no such increase occurs. It has been pointed out that the increase of urinary phenols in prolonged fasting is due to a great destruction of proteins. This cell destruction may be one of the factors producing the rise of blood phenols in liver injury. As mentioned before, proteins and their decomposition products such as tyrosine, tryptophane, and other easily oxidiz- able substances give the blue color with the phenol reagent, and since the increase of conjugated, volatile phenols in these cases is not proportional to the increase in total phenols (the conjugated phenols being practically the same as in normal dogs) the bulk of the additional color-producing bodies may be proteins and their decomposition products other than phenols, or sugars and related substances (Gortner and Holm, 4; Levine, 5). It was shown in the preceding paper that a large part of the absorbed phenols is destroyed by a process of oxidation entirely different from the synthetic process by which conjugated phenols are produced. It is possible that this oxidative destruction which goes on in most tissues, but to greatest extent in the liver and in the intestinal epithelium, is inhibited or at least greatly reduced by the presence of injured liver tissue, thus adding to the increase in phenolic substances. This seems to be substan- tiated by the fact that in dogs injured by chloroform the abso- lute rise in blood phenols after ingestion of p-cresol is always greater than in normal dogs. Table XI, Experiment of April 18, and Table XIII, Dog 21-113 show further that with chloroform injury the excretion of ingested phenols may be less rapid than in normal dogs—another less important factor in the increased concentration of phenolic substances in the blood. Delprat and Whipple (2) showed that there is no impairment of renal function following a chloroform anesthesia as measured by the elimination of phenolsulfonephthalein. In accounting for this increased concentration of “phenolic reacting substances” in blood in eases of chloroform injury, we have, in addition to whatever unknown factors may play a part, these possibilities: (1) an increased destruction of body proteins, particularly liver proteins, by chloroform; (2) an inhibition or K. F. Pelkan and G. H. Whipple 525 lessening of the oxidative destruction of absorbed phenolic sub- stances; and (3) a slowing of the excretion of phenolic substances. It is interesting to note that this rise in blood ‘‘phenols’” seldom occurs in dogs whose liver has been injured with phosphorus. Even fatal phosphorus injury shows only a slight increase over the normal level (Table XIII). Nor does this increase occur with bile fistula or Eck fistula dogs. This indicates again the pres- ence of certain “‘ phenol-reacting”’ substances in the blood of dogs poisoned with chloroform but not necessarily true phenols alone. DISCUSSION. The cause of death in Eck fistula dogs has long been a puzzle to physiologists. It has been claimed by many workers that this peculiar intoxication was due to the absorption of the toxic amino- acids as it is well known that heavy meat feeding will precipitate the characteristic intoxication. Unpublished experiments of Van Slyke and Whipple show that there is no abnormal heaping up of amino nitrogen in the blood during periods of meat feeding in Eck fistula dogs. No amino-acids, such as have been noted in severe cases of chloroform poisoning and fatal liver injury, appear in the urine of these Eck fistula dogs. The experiments given in this paper indicate clearly that the Eck fistula liver is incapable of normal conjugation of one toxic radicle, (p-cresol). This disability is noted during periods of normal health as indicated by clinically normal reactions. It is at least possible that this im- pairment of the conjugating powers of the liver is responsible for the toxic developments in the Eck fistula dog. We note that there is no heaping up in the blood of the Eck fistula of any phe- nol-reacting substances. It will be of considerable interest to study this reaction in the Eck fistula dogs on a high meat diet and during periods of the characteristic Eck fistula intoxication. It is to be recalled that these Eck fistula dogs were maintained on a diet of rice, bread, milk, and bones. It is significant to note in the tables that severe poisoning with phosphorus will not cause a great rise in the total blood phenols but equally severe poisoning with chloroform will give very high figures for total blood phenols. One suspects that cell necrosis which is so conspicuous in chloroform poisoning is responsible for this difference. This reasoning suggests that a considerable THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. L, NO. 2 526 Studies of Liver Function. III part of the total phenol of the blood in chloroform poisoning may be due to ‘‘ phenol-reacting substances’’ quite apart from the phenolic substances. The method we have used is of interest to investigators but as yet of little practical value to internists. p-Cresol is too toxic to be used clinically but further study may enable us to suggest some non-toxic radicle which will test these synthetic activities of the liver parenchyma. It is of some significance that this method is a specific test of liver function, as we have evidence that other body cells are not concerned in the conjugation of phenols. SUMMARY. The conjugation of phenols in the body is in nowise disturbed by bleeding periods, by the presence of a bile fistula nor by a lethal intoxication (distemper). Under the conditions of these experiments the reaction to ingestion of a unit dose of p-cresol is uniform and associated with a constant amount of phenol con- jugation which can be measured quantitatively. The presence of an Eck fistula modifies this reaction and reduces the amount and speed of phenol conjugation. At times the Eck fistula liver function of conjugation may fall to one-third or even to one-tenth of the normal. When the liver circulation is fur- ther impaired by partial ligation of the hepatic artery in an Eck fistula dog, we may observe a fall in phenol conjugation to 3 or 5 per cent of normal. Liver exclusion, therefore, will eliminate phenol conjugation. The presence of a slight liver injury due to chloroform or phos- phorus may not modify the phenol conjugation. Extensive liver injury due to these poisons will always lessen phenol conjugation. Extreme and fatal liver injury (chloroform) will reduce phenol conjugation to zero. These observations lead us to conclude that phenol conjugation is a function of liver parenchyma cells and of no other body cells. BIBLIOGRAPHY. oo . Davis, N. C., and Whipple, G. H., Arch. Int. Med., 1919, xxiii, 612. . Delprat, GC. D., and Whipple, G. H., J. Biol. Chem.; 1921, xlix, 229. . Dubin, H., J. Biol. Chem., 1916, xxvi, 70. . Gortner, R. A., and Holm, G. E., J. Am. Chem. Soc., 1920, xlii, 1678. . Levine, V. E., Science, 1920, lii, 612. or em CO bo THE EFFECT OF HYDROGEN ION CONCENTRATION UPON THE DETERMINATION OF CALCIUM. By ALFRED T. SHOHL. (From the Department of Chemical Hygiene, School of Hygiene and Public Health, the Johns Hopkins University, Baltimore.) (Received for publication, December 12, 1921.) MecCrudden’s (1, 2) method has come to be recognized as the best for the accurate determination of calcium. The method is to precipitate the calcium as oxalate in the presence of suffi- cient ammonium chloride to hold the magnesium oxalate in solu- tion, and of sufficient acid to hold the calcium oxalate partly in solution. Sodium acetate is then added to decrease the acidity and to precipitate the rest of the calcium oxalate on the crystals already formed. Thus, large crystals are made which are easy to filter and are not contaminated with occluded magnesium or calcium phosphate. The amount of acetate is selected so as to give a solution not acid enough to dissolve the calcium oxalate, nor alkaline enough, if cold, to allow calcium phosphate to pre- cipitate. These directions are empiric and must be followed accurately, as the author cautions.2 Just what determines the amount of acetate is not definite, for he states that in the ash of feces 15 instead of 8 cc. are required.’ The sodium acetate added regulates the acidity of the solution. Unfortunately, at the time McCrudden wrote his article, Séren- sen’s (3), Michaelis’ (4), and Clark’s (5) monographs on hydrogen ion concentration were not published. In 1913, Hildebrand (6) of the Bureau of Standards showed the necessity of a systematic investigation of analytical methods with regard to the hydrogen ion concentration. Such information is not as yet available for 1 MecCrudden (1), p. 99. 2 McCrudden (1), p. 100. 3 McCrudden (2), p. 198. 528 pH and Determination of Calcium many types of analysis. Except that of Kramer and Tisdall (7) there is none for calcium. We haye, therefore, critically examined McCrudden’s method in relation to hydrogen ion concentration. Theory Underlying the Precipitation of Calcium as Oxalate. The determination of calcium in the presence of magnesium and phosphate depends primarily upon the solubility product of the various precipitates involved; secondarily, upon the hydro- gen ion concentration. As a development of the last 30 years in the chemistry of solutions, the theory of ionization has been applied to the problem of precipitation of nearly insoluble sub- stances with great success. Solubility Product.— Stieglitz (8) and Noyes (9) have presented very carefully in their manuals of qualitative analysis the rela- tion between ionization and precipitation. The basic principle is the mass law, which can be stated.® (At) x (B-) LER ce In non-mathematical terms, when a substance is present in water an equilibrium is established between the concentration of the undissociated salt in solution and the concentration of the ions. The value, K, is a constant depending upon the nature of the par- ticular salt. Or, when there are two ions present which form an insoluble compound, precipitation occurs until the product of the concentration of the ions divided by the concentration of un- dissociated salt is a constant which is the K for that salt. A simpler relation which holds with sufficient accuracy for prac- tical purposes is that the product of the ions is a constant. This constant is known as the “‘solubility product.” (At) x (B) = K Obviously, we decrease the amount of either ion present by in- creasing the other ion. If we double A* we halve B-. Hence, adding more A+ causes more and more complete precipitation * The question of the problem of occlusion and the conditions for obtain- ing large crystals have been discussed by McCrudden (1), p. 99. 5 Parentheses about a symbol mean concentration. Thus, H* means hydrogen ion; (H+) means hydrogen ion concentration. Alfred T. Shohl 529 of B-. Thus, one adds an excess of oxalate ions, so that the cal- cium is more completely precipitated. The value of the solu- bility constant is very important for it determines which salt will precipitate. The salt having the smaller solubility product will precipitate and the other will remain in solution. [If all the factors are known a quantitative expression can be calculated rigorously from the mass law. The solubility product determines whether calcium will be precipitated as a phosphate or carbonate or oxalate in solutions containing these acids. The solubility products of the various salts to be considered are given in Table I. Effect of Hydrogen Ion Concentrations.—Acids on dissociation always give hydrogenions, H+. The extent of the acidity depends upon the concentration of the hydrogen ions, (H+) (10). The effect of hydrogen ions in precipitation can best be discussed under three heads: (a) the effect on basic salts, (b) the change in the ioniza- tion constant of the acid radicals, and (c) the suppression of ioni- zation of weak acids. The effect of acid on basic salts is to depress the ionization of the hydroxyl] ions; for substances in solution are related to the ioni- zation of water. Water itself gives H* ions and also OH™~ or basic ions. The relation of the ions is expressed by the mass law: (H+) x (OH-) =r1h=14 (H,0) ay When one adds H* ions to water one depresses the ionization of the hydroxy] ions since the product of these ions is a constant. Thus, in the case of magnesium hydroxide, by adding acid one decreases (OH) to such an extent that in acid solution its solubility prod- uct is never reached and hence no magnesium hydroxide can precipitate. By adding acid one changes the ionization constant of the acid. Thus, in the case of the tri-basic phosphate the addition of H+ ions gives the following reaction: PO, + H+ = HPO, + H+ = HPO, The tri-basic salt can only exist in alkaline solution and as one in- creases the hydrogen ion concentration the di-basic and mono- basic salt must be formed. Thus, in an acid system there is no calcium phosphate but only the more soluble di-basic and mono- 530 pH and Determination of Calcium basic calcium phosphates. The solution, if quite acid, about pH 3.0, will have the same effect on calcium oxalate. a H+ + (G0) — (HC.0) By adding acid to a solution containing ions of a weak acid one converts the highly ionized salt into a slightly ionized acid, ac- cording to the equation: os 2H+ + (C,0,) = HC.0, Thus one removes oxalate ions from the solution by adding acid. If the ionization is repressed below the solubility product no pre- cipitate is formed. Each of these factors: the effect on basic salts, the change in the ionization constant of the acid, and the suppression of ioniza- tion, influence the precipitation. Therefore, we shall consider each of the salts that may be formed in the course of the analysis of calcium in the presence of magnesium and phosphates, in an acid solution of pH 4.0 to 6.2. Calecum Oxalate and Magnesium Oxalate-—These salts are pre- cipitated in the presence of an excess of oxalate ions. The ques- tion is whether there is any danger of precipitating magnesium with the calcium. This has been discussed in part by Kramer and Tisdall (7). The solubility product of these salts is suffi- ciently different so that Gooch (11) recommends that even in the presence of ten times the amount of magnesium it is not necessary to carry out double precipitation. In acid solution this differ- ence in solubility product is even more marked. Acid added to a‘solution of calcium and magnesium oxalates favors the precipitation of the calcium in two ways. It forms the acid salts which are more soluble. The acid salt of calcium oxalate, at this acidity (pH 4.0 to 6.2) is not formed in sufficient amount to have any appreciable effect on the solubility. Nostudy of the magnesium salts has been made, but from general evidence it is concluded that this would be more readily affected by acids and hence more soluble. Second, the addition of acid suppresses ionization, in this case from about 40 to about 30 per cent. The diminished ioniza- Alfred T. Shohl 531 tion affects the calcium oxalate much less than the magnesium oxalate as the former has a much smaller solubility product. Any acidity not great enough to form acid calcium oxalate will be a factor in preventing the formation of magnesium oxalate and in favoring the precipitation of calcium oxalate. Magnesium Hydroxide and Magnesium Ammonium Phosphate.— Both salts are extremely insoluble but since they do not occur in acid solutions they do not precipitate. Hildebrand (6) says that the former precipitates at pH 7.4 to 8.0 and that the latter first appears at pH 6.6. Mono-, Di-, and Tri-Basic Calcium Phosphate-—The salts of cal- cium form a very difficult problem since as found by Cameron (12) and his collaborators and many others, these salts vary in their composition according to the source, are decomposed by water, and take months to come to equilibrium. For orientation we attempted to find the pH of saturated solu- tions of the mono-, di-, and tri-basic salts of calcium phosphate. These salts (Baker’s analyzed) were washed with distilled water. The acidity of the supernatant liquids, determined by the colori- metric method (5), became roughly constant at room tempera- ture in 24 hours and gave the following values: CaHy, (POs). = pH 5.0 CaH PO, =pH7.4 Ca;(PO.)2 = pH68 _ We found that no calcium solution more acid than pH 4.0 which contained phosphates yielded a precipitate after boiling, but Patten: and Mains (13) report a precipitate at pH 2.3 at 26°. But the more acid the point at which these salts precipi- tate, the greater is the proportion of phosphate to calcium. And, as one can see by glancing at Table I, the more acid the salt the greater is the solubility product. Therefore, at pH 4.0 to 6.4 there is no danger of any calcium being present as phosphates. At this acidity the calcium phosphates have a solubility product more than a million times as great as calcium oxalate. Brea- zeale (14) remarks it is quite safe to precipitate calcium phos- phate quantitatively as calcium oxalate by adding oxalic acid. Therefore, there is no danger of the precipitation of calcium acid Doe pH and Determination of Calcium phosphate if the precipitation of calcium oxalate is carried out at the proper acidity. TABLE I. Data on Solubilities.* Salt. Per liter. Mols per liter. Solubility product. gm. aN Ls O21 OF 0 kein MRrere Sd. - 0.0055 0.000044 1:9 xX 10" Mo(OH) 5... ::~.eeee 0.009 0.00015 3.5 X 107” MeNHiPO). ..<.o eee 0.05 0.0068 3.0 X 1077 MipCiOght i.e 0.302 0.0027 4.8 X 10-5 Gas(PO) ai...) slacererete 0.01 0.00033 2.8. 100 Os BO mete tI 0.2 0.00147 2.0 X 107° Call (LO){)2.) Secor 18.0 0.77 4.0 X 107 *The figures are compiled from Landolt-Bérnstein, Physikalisch- chemische Tabellen; A. Seidell, The solubilities of inorganic substances; The Chemische Kalender, 1914—Dictionary of Solubilities, etc. EXPERIMENTAL APPLICATION OF THE THEORY. It is important to verify, first, the most acid limit of acidity; second, the least acid limit; and third, the best method of obtain- ing the desired acidity. Most Acid Limit—The most acid limit is the point at which calcium oxalate begins to be converted into the more soluble acid calcium oxalate. McCrudden has determined the amount of sodium acetate, which, under the conditions of his procedure will prevent the solution of calcium oxalate.6 We determined colorimetrically the hydrogen ion concentration of solutions pre- cipitated according to his directions. In all solutions more acid than pH 4.4, which contain less than 6 ec. of sodium acetate, the results are low. Further experiments in Table II show that correct determinations are made at pH 4.0, which is, therefore, the most acid limit for the determination of calcium oxalate. The Least Acid Limit.—The least acid limit is the point at which magnesium ammonium phosphate and magnesium hydroxide precipitate. According to Hildebrand this is pH 6.6 to 7.6. To fix this point experimentally we carried out the precipitation 6 McCrudden (1), p. 90. Alfred T. Shohl 533 TABLE II. The Amount of Sodium Acetate and the Resulting Acidity.* ify 20 per cent aoditer sieiate, | P= calculated. | pH determined. Calcium. ce. mg. 1 0 1.3 30.1 2 a 2.8 2.8 32.0 3 5 4.0 4.0 33.1 4 6 4.4 4.4 33.1 5 10 4.8 4.8 33.3 6 20 5.3 5.2 33.1 7 50 5.7 5.6 33.2 Pheory.... 33.2 * In solutions more acid than pH 4.0 the results are low; 50 ec. of sodium acetate are not an excess and give correct results. of calcium at varying acidities and determined the amount by both the gravimetric and volumetric methods. By this procedure it is possible to determine faulty results and also the cause of error. Titration with permanganate determines the oxalates. If phosphates are present and contaminating the calcium oxalate they will not affect the result. By the gravimetric method one determines the calcium as oxide and the oxalates are destroyed. Phosphates will cause the results to be too high. Reasoning thus, one can deduce the following: Method. Precipitate consisting of: Gravimetric.| Volumetric. Breer Comte AONE. ooo aca» obirinin. a s.0ee ne ap nae he Correct. | Correct. x 4 and calcium phosphate............ High. Low. eg “magnesium ammonium phos- ASU pe SA se 61 os 5 rr # Correct. Calcium oxalate and magnesium oxalate............ * High. cs are published, the aim being to provide concise but comprehensive reviews of — the recent literature and present status of various subjects in Physiology, using | this term in a broad sense to include Bio-chemistry, Bio-physics, Experimental Pharmacology and Experimental Pathology. Each volume will consist of approximately twenty articles of twenty-five pages each, a total of five hundred pages per volume. The numbers will be issued quarterly in January, April, July and October of each year. Subscriptions will be received in advance only. The subscription priee is: $6.00 per volume in the United States, net postpaid $6.25 per volume in Canada, net postpaid $6.50 per volume elsewhere, net postpaid $2.50 per single number, net postpaid Subscriptions should be sent to D. R. Hooxrr, Managing Editor ‘ 1222 Sv. Paun STREET, BALTIMORE, MD. JOURNAL OF IMMUNOLOGY OFFICIAL PUBLICATION OF THE SOCIETY FOR SEROLOGY AND HEMATOLOGY AND THE AMERICAN ASSOCIATION OF IMMUNOLOGISTS Edited by ArrHuur F. Coca Vol. VI, No. 3 CONTENTS MAY, 1924 aiNeGre se Nee ee a re Bee eee eee ee ee Moyer §. FLEISCHER AND NaTALIE ARNSTEIN. Specificity of Anti-organ Sera...............seseeeeceees 223 F.M. Hunroon ann §.H.Craia. Polyvalent Antibody Response to Multiple Antigens................. 235 Current Volume: Vol. VI, 192i—Price, $5.00 Domestic; $5.5@ Foreign Back volumes oan be furnished: Vols. I-V incl. Price per set: $24.00, United States, Mexico, Cuba; $25.00, Canada; $26.00, othercountries. Prices are net postpaid Subscription Order for the Journal of Immunology WILLIAMS & WILKINS COMPANY, Mount Rorat anp Guitrorp AVENUES, BauTimorE, Mp., U.S.A. We | enclose { LO.) \ for subscription to the current volume of JouRNaL or ImMmuUNOLOGT. Please send Back Volumes (Please state whether Dr., Prof., eto.) AUTOR nies ccs cuiSanibles ++ ccccccncwaesinin<.« ws «1 5 aninnenle= © p< sete nee eee een eee eccedens Hii HLTA HU AN OUTFIT FOR Basal Metabolism Estimations AS DESIGNED BY Dr. Cameron Vernon Bailey of the Laboratory of Pathological Chemistry \\ New York Post Graduate Medical \ School and Hospital For detailed description see September issue Journal of Laboratory and Clinical Medicine rE /) 4 ¥ \ Sg <5 a AMERICAN MADE PRO 81 77 REPRESENT QUALITY OF HIGHEST STANDARD MANUFACTURED BY C. M. SORENSEN CO., Inc. 177 East 87th Street DEPT. K New York ‘* Eternal Vigilance is the Price of Accuracy ’’ FOR LY / YEARS The Traditions and Standards developed in our shops since 1881 are reflected in the accuracy and quality of each piece of apparatus made by us. The scientific worker cannot afford to overlook the painstaking care and pride of workmanship summed up in the trade mark We are specialists in the manufacture of apparatus made to order. We are the original manufacturers of the Van Slyke CO: apparatus and amino nitrogen apparatus, the Van Slyke-Cullen Urea apparatus, Folin’s tubes and pipettes, The Fredericia apparatus and many other types now widely used. If you are not already one of our many customers WHY NOT? FULL LINE OF LABORATORY SUPPLIES THE EMIL GREINER COMPANY ESTABLISHED 1881 Folin Blood 55 Fulton Street New York, N. Y. vipette A Bloodless Field is promptly produced by the appli- cation or hypodermatic injection of Suprarenalin Solution, 1:1000 —the stable and non-irritating preparation of the Suprarenal active principle. The e. e.n. and t. men find it the premier product of the kind. Ischemia follows promptly the use of Elixir of Enzymes is a potent and pala- 1:10000 Swprarenalin Solution -slightly warmed (make |:10000 solution by add- ing | part of Suprarenalin Solution to 9 parts of sterile normal salt solution). In obstetrical and surgical work Pituitary Liquid (Armour), physiologically stand- ardized, gives good results—l4 c. c. am- poules obstetrical—l c.c. ampoules sur- gical. Either may be used in emergency. = aa = ——) LABORATORY PRODUCTS table preparation of the ferments active in acid environment—an aid to digestion, corrective of minor alimentary disorders and a fine vehicle for iodides, bromides, salicylates, etc. As headquarters for the organothera- peutic agents, we offer a full linz of Endo- crine Products in powder and, tablets (no combinations or shotgun cure-alls), Armour’s Sterile Catgut Ligatures are made from raw material selected in our abattoirs, plain and chromic, regular and emergency length, iodized: regular lengths, sizes 000—4- Literature on Request ARMOUR “> COMPANY CHICAGO The Journal of General Physiology Edited by JACQUES LOEB, New York, N. Y. Me. W. J. V. OSTERHOUT, Cambridge, Mass. The Journal of General Physiology is devoted to the explanation _of life phenomena on the basis of the physical and chemical consti- - tution of living matter. The Journal of General Physiology is issued bimonthly, one volume of about 600 pages appearing in a year. Contributions should be sent to the editors of The Journal of General Physiology, The Rockefeller Institute for Medical Research, Avenue A and 66th Street, New York, N. Y., or 60 Buckingham Street, Cam- bridge, Mass. The papers should be limited preferably to twenty printed pages, not counting the space occupied by illustrations. Authors receive 100 reprints of their papers free of charge; addi- tional copies may be obtained at cost. The subscription price per year (one volume) is $5.00, payable in advance. Single copies cost $1.00. Remittances should be made by draft or check on New York, or by postal money order, payable to The Journal of General Physi- ology, Mount Royal and Guilford Avenues, Baltimore, Md., or Avenue A and 66th Street, New York, N. Y. CONTENTS OF VOL. IV, No. 3, JANUARY 20, 1922 Norturop, JouN H. The inactivation of trypsin. I. Norturop, Jonn H. The inactivation of trypsin. II. The equilibrium between trypsin and the inhibiting substance formed by its action on proteins. Norrurop, Joun H. The inactivation of trypsin. III. Spontaneous inacti- vation. OsteRHouT, W. J. V. Direct and indirect determinations of permeability. Harvey, E. Newron. Studies on bioluminescence. XIV. The specificity of luciferin and luciferase. REDFIELD, ALFRED C., and Bricut, EvizanerH M. The effects of radium rays on metabolism and growth in seeds. Crozier, W.J. Correspondence of skin pigments in related species of nudibranchs. Powers, Epwin B. The physiology of the respiration of fishes in relation to the hydrogen ion concentration of the medium. UsuLennutTu, Epuarp. The effect of iodine and iodothyrin on the larve of salamanders. IV. The rdéle of iodine in the inhibition of the metamorphosis of thymus-fed salamanders. Untennuti, Epuarp. The influence of feeding the anterior lobe of the hypophy- sis on the size of Ambystoma tigrinum. Fenn, WaLuAcE O. The temperature coefficient of phagocytosis. Brooxs, MatitpA MoupENHAUER. The penetration of cations into living cells. Lors, Jacqurs. The origin of the electrical charges of colloidal particles and of living tissues. Published by The Rockefeller Institute for Medical Research Avenue A and 66th Street, New York, N. Y. 11 THE AMERICAN JOURNAL TROPICAL MEDICINE OFFICIAL ORGAN OF THE AMERICAN SOCIETY OF TROPICAL MEDICINE ahi Editor-in-Chief H. J. NICHOLS Medical Corps, U. S. Arm Army Medical School, Washirigron? D.C. Advisory Editorial Board B. K. ASHFORD G. W. McCoy Medical Corps, U. S. Army, San Juan, Porto Rico Director, Hygienic Laboratory, U. S. P. H.S., C. C, Bass Washington, D. C. Tulane University, New Orleans, La. K. F. MEYER M. F. Bovp University of California, San Francisco, Calif. University of Texas, Galveston, Texas B. H. Ransom C. F. Craic Department of Agriculture, Washington, D. C. Medical Corps, U. S. Army, Army Medica! School, R. P. STRONG Washington, D. C. Harvard University, Cambridge, Mass. GEORGE Dock Washington University, St. Louis, Mo. A SIMON FLEXNER Rockefeller Institute, New York City ee R, Suet F acen Washi a Wiiuram Krauss urgeon Genera . 5. Navy, Washington, D.C. Memphis, Tenn. Wa. S. Teaver) © 5 W. D. McCaw Johns Hopkins University, Baltimore, Md. Asst. Surgeon General, U. S. Army, Army Medical E. J. Woop School, Washington, D.C. Wilmington N.C. . J. SMITH University of Pennsylvania, Philadelphia, Pa. Ex Officio Advisory Editorial Board The American Society of Tropical Medicine J. M. Swan, President GEorRGE Dock, Councillor K, F. Meyer, Ist Vice-President C. L. Fursusn, Councillor V. G. HEISER, 2nd Vice-President J. F. SILER, Councillor S. K. Sruon, Sec’y and Treasurer J. H. Waite, Councillor A. J. Smitu, Asst. Sec’y and Treasurer C. S. BuTLer, Councillor Issued Bimonthly; six issues a year, 85 pages, approximately, an issue; one volume a year is planned at present. Price, net postpaid: $5.00, United States and dependencies, Mexico, Cuba; $5.25, Canada; $5.50, other countries. Order from WILLIAMS & WILKINS COMPANY Publishers of Scientific Journals and Books BALTIMORE Us. Sate wee THE FSENTATION of essential data condng vitamines to succeeding groups of studemas become increasingly difficult with the slopment of research in this field. The litere itself has assumed a bulk that precludesiing the student to original sources except ipse instances when they are them- selves tome investigators. The demand on the parthe layman for concise information about tlew food factors is increasing and worthy ¢ention. For all of these reasons it has seenvorth while to collate the existing data anc it in a form which would be avail- able for student and layman. Such is the purpose he Vitamine Manual. A Timely - New Book It HAen called a manual since the arrange- meims to provide the student with workingerial and suggestions for investiga- tion as as information. The bibliography, the dat:he epee on vitamine testing, the tables the subdivision of subject matter have aen arranged to aid the laboratory workerl it is the hope that this plan may make anual of especial value to the stu- dent tigator. The details necessary to laboratvestigation are separated from the more ¢ historical aspects of the subject, an arrangt that will be appreciated by the lay rezs well as the student. N°? LOGIES are made for data which cblication shall be found obsolete. The wubject is in too active a state of in- vestig:to permit of more than a record of events: their apparent bearing. (Froipreface of THE VITAMINE MANUAL) about the new food factors BY WALTER H. EDDY ASSOCIATE PROFESSOR PHYSIOLOGICAL CHEMISTRY Teachers College, Columbia University The Vitamine Manual will prove useful to Medical Men, Public Health Officers, Dietitians, Nutrition Experts, Chemists, and to all individuals interested in promot- ing correct habits of living and correct habits of diet. The Vitamine Manual will serve as a valuable text for classroom purposes WILLIAMS & WILKINS COMPANY Publishers of Scientific Journals and Books Wriiams & WiLkrins Company, Publishers of Scientific Journals and Books, Baltimore, Maryland, U. S. A. I (or) We enclose $...... for......copy (copies) of THE VITAMINE MANUAL. Price, $2.50 a copy, net postpaid 5 Copies - - - - 10% discount 25 Copies - - - - 15% discount 10 Copies - - - - 12% discount 50 Copies - - - - 20% discount (Signed) ..... Ba totetacts « 1a Oe ieveks «= s.s(eis didiew-ee BF SO. OR GOODS CD OUOUL BGGress es... he Gtecaheeins oi , Oem eeeer ee ee ee eer ee seer er ee eesee that answers the demand for concise information THE VITAMINE MANUAL “ PHYSIOLOGICAI ABSTRACTS a a Monthly, price 5s. net. Annual Subscription, post free,}. This Journal is issued by the PHYSIOLOGIC SOCIETY, acting in cooperation with num ous physiological organisations in Great Brita America, and other countries. | a staff of competent abstractors both at home a The Editor, Professor Hatirsurron, is assisted. abroad. | The Journal aims at issuing promptly abstracts the papers published throughout the world, physiological and allied sciences (including pl physiology). A number is published at the beginning of ea month. Volume VI begins with the April (1921) issue. Most of the numbers of preceding volumes can still be | tained from the Publishers at the price of 4s. per No. few numbers which are now scarce are 8s. per No. Subscriptions are only taken for April to March of next year, and are payable in advance. PUBLISHED BY H. K. LEWIS & CO., Ltd. 28 Gower Place LO lity ~~ 4 (; F. a Wee ie fe Oe Ui irre eP i RORENLsi.. . (PIAL 2S 80. QP The Journal of biological pa chemistry rl (17.27) Biological & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY rate Kia tues Fah der ree recat L pans ee re + Soe Sars esa Hin aa Fs Sea Site G ie ay # aise ws m Ca Fo ath bate mat Sees ae wee ieee “by Saree ye evi lirl et oe Vor wneeat payer reat pst ae * of iglblas heat Vaiss eet , sie ats en eaeniat as Be eats lots styleg tye eves Sohn aired eel esd Crt AR aa evi? Wet - pee yise se tity 4 iad vee apotti rere! ei iatet it ia Sieh bets nse’ ‘a ibe pepe erie ateereiny os Fe bebe ha $7 ree ts, Se. re -3 ; taeiitie tret $233 AL bibs stitlige Heebaayeededy aebegijogeg ésdh600°