1224 HARDEN AND YOUNG: GLYCOGEN FROM YEAST. CXXV.-GZycogen from Yeast. By ARTHUR HARDEN and WILLIAM JOHN YOUNU. THE occurrence in the yeast cell of a substance which yields a reddish- brown coloration with iodine appears to have been first observed by Errera in 1882 (L'EpipZasnze dcs Ascomyc8te.s et Ze Glycogdne des vgqgttaux, Bruxelles, 1885, and Compt. Tend., 1885, 101, 253), although the existence of a carbohydrate reserve substance had been surmised by Pasteur in 1859. On account of this reaction, it was considered to be glycogen, and Errera attempted to extract it from the cell, but did not succeed in obtaining it sufficiently pure t o be certain of its identity with ordinary glycogen. It was next examined by Cremer (Munchener Med. Wochenschr., 1894, 41, 525), who stated that he prepared it from yeast and purified it by Briicke's method (precipitation of the proteids by potassium mercuri- iodide) and by fractional precipitation, 13 grams being obtained from 250 grams of dried yeast.He describes it as a neutral white powder, of specific rotation [ a ID + 198.9". It gave a red coloration with iodine, formed an opalescent solution in water, was converted by boiling withHARDEN AND YOUNG: GLYCOGEN FROM YEAST. 1225 dilute hydrochloric acid into dextrose, and was inverted by saliva, pancreatic juice, and diastase, A much more detailed investigation on this subject was made by Clautriau (Etude chirnique du Glycogdne. Mem. Coummgs, Acud. Roy. Belg., 1895, 53), who prepared glycogen from yeast, from two species of fungi, and from the rabbit, and carefully compared their properties.Clautriau made use of yeast which had been enriched in glycogen by treatment with wort containing 1 2 per cent. of cane-sugar, and after repeatedly boiling it with '1 per cent. aqueous potash and washing with water disintegrated the cells by making them into a solid mass with silica, chalk, and potassium silicate, and then grinding the whole to a fine powder. The glycogen was then extracted by boiling water, a,nd was freed from mucilaginous matters by the repeated production in its solution of a precipitate of calcium phosphate, the last traces of gum being removed by saturating the solution with salt and then adding ammonium sulphate in excess. The glycogen was precipitated from the diluted solution by an excess of iodine, the precipitate treated with sulphurous acid, and the glycogen precipitated by the addition of two volumes of alcohol, and finally washed with absolute alcohol and ether and dried in the air.He thus obtained it free from nitrogen, but containing 1-3-15 per cent. of ash, and observed that in its general properties it agreed with the substance described by Cremer. He found, moreover, that all the samples prepared by him, both from animal and vegetable sourcee, had the composition 6C,H,,O,,H,O, and were members of a group of closely allied substances, but were not identical in every respect. Extraction, and Pzcrij%ation of Yeast Glycogen.-Well washed and pressed yeast was mixed with an equal weight of fine, white sand and the cells broken by grinding for two hours in the disintegrator described by Rowland (J.Zhysiol., 1901, 27, 53), the mass being cooled during the operation by liquid carbon dioxide. The ground-up mass was then poured into two or three times its bulk of boiling water and boiled for about two hours. The liquid was separated in a centrifuge and the residue again boiled with water aud separated, the two extracts being then mixed. To the united solutions, 1 per cent. of sodium phosphate was added, together with an equivalent amount of calcium chloride, according t o Clautriau's method, and the whole mass, after having been neutralised with ammonia, mas heated on the water-bath, the calcium phosphate filtered off, the filtrate evaporated dowu, and the glycogen precipitated by the addition of an equal volume of alcohol. This operation was repeated as long as the glycogen appeared sticky and viscous on precipitation with alcohol, the calcium phosphate bringing down with it most of the mucilaginous material, The glycogen was then redissolved i n water, and the liquid1226 HARDEN AND YOUNG : GLYCOGEN FROM YEAST, first saturated with salt, then with ammonium sulphate, and allowed t o stand in a cool place for three days, By this means, the last traces of gum mere precipitated, the glycogen remaining in solution.After filtration, the liquid was freed from the greater part of the salt and ammonium sulphate by dialysis, and the glycogen precipitated by alcohol. To obtain the glycogen free from proteids, it was further found essential to treat it by the method described by Pfluger for the purifi- cation of animal glycogen (P’ilgsr’s Archiv, 1900,81, l), which consists in precipitating the glycogen from a solution containing 3 grams of potash and 10 grams of potassium iodide per 100 C.C.by the addition of half a volume of alcohol, filtering off, and washing first with a mixture of 300 C.C. of alcohol, with 400 C.C. of aqueous 3 per cent. potash containing 40 grams of potassium iodide, and then with 50 per cent. alcohol. After this treatment, the glycogen was freed from ash by being redissolved in water, and the solution precipitated with its own volume of alcohol in the presence of a trace of acetic acid, this opera- tion being repeated several times. Finally, it was again redissolved in water, precipitated with 1 volume of alcohol, filtered through silk, washed well with 50 per cent.alcohol, then with absolute alcohol, and finally with ether which had been carefully freed from acid, and was then allowed to dry in the air. The processes of solution in water and precipitation by alcohol must be repeated until the dried material is found to be free from ash. The purification was always carried on until 0*2-0°4 gram of the dried glycogen gave no ammonia by Kjeldahl’s process, and left no weighable amount of ash on ignition. The juice obtained by pressing out ground yeast with kieselguhr, according to Buchner’s method of obtaining zymase, may also serve as a source of glycogen, an equal volume of alcohol being added and the aqueous extract of the precipitate then treated as described above.Since the amount of glycogen contained in yeast varies very con- siderably with the condition and past history of the yeast, the yield of pure dry glycogen obtained also varies, but it may be taken to be on the average about 2 per cent. of the weight of the pressed yeast. For the purposes of comparison, samples of glycogen were prepared from rabbit’s liver and from oysters, Pfliiger’s method of extraction and purification being employed, and the purification carried to the same point as with the yeast glycogen; 570 grams of rabbit’s liver yielded 8.7 grams, and 1617 grams of oyster (1 gross) yielded 12 grams of pure dry glycogen. Cornpodtion.-Much diversity of opinion exists as to the com- position of glycogen both of animal and vegetable origin, AlthoughHARDEN AND YOUNG: GLYCOGEN FROM YEAST. 1237 Kekulh, in 1858 (Chern.Plmrm. CsntraZbZ., 3, 301), determined the empirical formula to be C,H1,O,, a number of investigators since that date, among whom may be mentioned Euppert (Zeit. physiol. Clmn., 1894, 18, 138), Chittenden (Annakn, 1875, 178, 266), Kiilz and Borntrager (PJuger’s Archiv, 1881, 24, 19, where a bibliography of the literature on this point is to be found), Vintschgau and Diet1 (Pjuger’s Archiv, 1878, 17, 163) and Clautriau (Zoc. cit.) have obtained figures agreeing with those required for the formula 6C,H,o0,,H,0, whilst Boehm and Hoffmann (Arch. exp. Path. Pharm , 1879, 10, 12) found numbers agreeing with those required for the formula 1 1 C6H,,05,H,0. On the other hand, Klincksieok, in 1861, in a single analysis of glycogen from the human liver (AnmaZen, 1861, 118, 229) and Bizio (Compt.rend., 1867, 65, 176), who employed glycogen from mollusca, obtained the formula C,H,,O,. Still more recently, Norking (PJEiiger’s Archiv, 1901, 85, 320) has obtained a similar result, using glycogen from horse flesh carefully freed from nitrogen and ash, and dried to constant weight a t looo. All these authors appear to have dried their glycogen at temperatures from 100-115° in an air-bath or a water-oven, with the exception of Bizio, who dried some samples at the ordinary temperature in a dry vacuum. Among these authors, Clautriau has paid special attention to the composition of glycogen extracted from yeast and fungi, and quotes a series of ten analyses referring to glycogen from four distinct sources (rabbit, yeast, Boletus, and Amanita).The glycogen in every case was dried at 105-110° in an oven until constant in weight, and the results, after allowing for the ash, agreed without exception with the formula 6C,Hlo0,,H,0, which requires C = 43.63 ; H = 6-26 per cent. I n two other analyses of glycogen from the rabbit, Clautriau ob- tained 44.1 and 44.0 per cent. of carbon, but in both these eases he found that the glycogen had become altered during drying, probably owing to the presence of small traces of acid, and yielded a solution in water which reduced Fehling’s solution freely. He attributes the high numbers obtained by some of the investigators already quoted to a similar cause. Our first analysis of oyster glycogen was made with material dried in the air at looo until constant in weight, and gave numbers even lower than those of Clautriau : 0.2092 gave 0.3308 CO, and 0.1247 H,O.It was found, however, that all the water could not be removed by C = 43.1 ; H = 6.6. 6C,H,,O,,H,O requires C = 43-63 ; H = 6-26 per cent.1228 HARDEN AND YOUNG: GLYCOGEN FROM YEAST. simply heating at looo in the air, and that a further quantity was lost when the glycogen was heated a t 100' in a vacuum over phosphoric oxide. The samples used for the following analyses were all heated in this way until constant in weight, the boat containing the glycogen being kept and weighed in a stoppered weighing tube and transferred t o the combustion tube as rapidly as possible. I n every case, a separate portion was dried in a similar manner, dissolved in water and tested for reducing'substances, but on no occasion could any reduction be observed. The two samples of yeast glycogen were separate preparations : I.Yeast glycogen : (a) 09249 gave 0.3639 CO, and 0.1283 H,O. C= 44-13 ; H = 6.34. (6) 0.1311 ,, 0.2132 CO, ,, 0.0743 H,O. C=44*35; H=6*30. 11. Oyster glycogen : ( a ) 0.1974 gave 0,3189 GO, and 0.1120 H,O. C544.06; H=6*30. ( b ) 0,1643 ,, 0,2669 CO, ,, 0.0924 H,O. C=44*30 ; H= 6-25. 111. Rabbit glycogen : (a) 0.2559 gave 0.4167 CO, and 0,1462 H,O. C = 44.41 ; H = 6.34. (6) 0.2611 ,, 0.4232 GO, ,, 0.1458 H,O. C=44*20 ; H=6*20. C6H,,0, requires C = 44-44 ; H = 6.17 per cent. 6C,HI,0,,H ,O requires C = 43-63 ; H = 6.26 per cent. These results all agree, within the limits of permissible experimental error, with those required for the formula C,H,,06, thus confirming the formula originally obtained by Kekul6 and a t the same time showing the identity in composition of the glycogen derived from yeast with that derived from animal soiirces.As none of the investigators who obtained low numbers for glycogen dried their material in a vacuum, it seems probable that their results were due to the presence of small quantities of water, the amounts of which would naturally vary with the humidity of the atmosphere in which the glycogen was heated. Speci$c Rotation.-The opalescence of aqueous solutions of glycogen renders the determination OF its specific rotation extremely difficult, and hence the results obtained by different observers differ somewhat widely, varying From + 213O to + 1 8 4 O .The results hitherto obtained f o r the glycogen of yeast are those of Cremer, [a],= 198.9", and of Clautriau, [a]= = + 184*5O, or, taking glycogen as C6HIOO5, [aID = 187.9". Using a Landolt-Lippich polarimeter, reading to O - O l O , with sodium light i t was only found possible t o employ solutions containing 0.1-0-2 gram per 100 C.C. Since glycogen which has been thoroughly driedHARDEN AND YOUNG: GLYCOGEN FROM YEAST. 1239 dissolves very slowly in water, it was found advisable to employ air- dried glycogen, the weight of the dried glycogen being found by drying a separate portion of the same sample in a vacuum over phosphoric oxide at 100' and allowing for the same percentage of water.Each of the following results was obtained as the mean of two independent sets of 15 readings made by two observers : I. Oyster glycogen : ~=0*1083; 2=1; u E = + 0.207O; [a]:"= +191*2O. 11. Rabbit glycogen : 111. Yeast glycogen : ~ = 0 . 1 2 0 4 ; Z=1; ar= + 0.23 ; [u]g"= 191.1 . ( a ) i. c=0*1730; Z=2; UY= 0.682; [u]Y= +197*1O. ii. c=0*1937; Z=2; UY= 0.790; [u]F= 203.5 . (b) i. c=0*1558; Z=2; uT= 0.616; [a]""= 197.7 . ii. c=0*1739; Z=2; ~1,7"= 0,678; [u]:"= 194.9. Two separate estimations of each of two different preparations of yeast glycogen were made, the mean of the four being [ u ] ~ ~ " + 198.3'. This result is slightly higher than that of Clautriau and agrees well with that of Cremer. The difference between the value for yeast glycogen and that obtained for animal glycogen, in view of the low concentration of the solutions employed, cannot be considered as indicating any essential difference between the properties of the two substances. OpaZesce~ce.-Clautriau states that the glycogen which he obtained from yeast produced a much less opalescent solution in water than the specimens which he prepared from fungi and the rabbit, and estimated the opalescence of the yeast glycogen solution at about one-fourth of that of the others.The opalescence diminished and gradually disappeared when the solution was preserved under sterile conditions for a few days. It must be remembered that Clautriau's glycogen contained 1-3.15 per cent. of ash, We have also found that a solution of yeast glycogen is less opalescent than solutions of animal glycogen, a 1 per cent.solution of glycogen from the oyster possessing roughly 2.5 times the opalescence of a similar solution of yeast glycogen. Rabbit glycogen was found to yield a slightly more opalescent solution than the oyster glycogen. A comparison of two eeparate preparations of yeast glycogen showed that one of them yielded a solution which was somewhat more opalescent than that of the other, but the difference did not exceed about 5 per cent. The opalescence showed no signs of diminution when the solutions were preserved under sterile conditions for a fortnight,1230 HARDEN AND YOUNG: GLYCOGEN FROM YEAST Reaction with Iodine,--The coloration produced with iodine has also been observed to vary for specimens of glycogen derived from different sources.Thus, Clautriau found that yeast glycogen gave a darker coloration than that from fungi and from the rabbit, and that the coloration produced by the yeast glycogen disappeared at a somewhat higher temperature than that of the others, whilst Weinland (Zeit. Biol., 1901, 41, 69) observed that the glycogen derived from certain parasitic worms gave a much weaker reaction with iodine than mammalian glycogen. The three varieties of glycogen examined by us differed in the intensity of their colorations with iodine. The t i n t produced by rabbit glycogen was slightly deeper than that given by the yeast glycogen, whilst both of these were much stronger than that produced by the oyster glycogen. On heating, the coloration of the oyster glycogen disappeared before that of the others.Comparing two samples of yeast glycogen from separate preparations, i t was found that the coloration produced with iodine was in one case 25 per cent. deeper than in the other, The reaction was in all cases carried out by adding 2 C.C. of a 1 per cent. solution of iodine in 1.5 par cent. potassium iodide solution to 10 C.C. of a 0.2 per cent. solution of the glycogen, as recommended by Clautriau. Acid Hydrolysis.-Many observers have noticed that glycogen does not yield the calculated amount of dextrose when i t is boiled with dilute acids, and in this respect it resembles starch (Sachsse, Chem. Centr., 1877, 8, 732). Kiilz and Borntrager (P’idgeq*’s Archiv, 1881, 24, 28), in a long series of experiments with glycogens of different origins, obtained very varying results, the weight oE sugar obtained varying from 95.34 to 117.6 per cent.of that of the glycogen employed, the calculated amount for their glycogen, 6C,H,00,,H,0, being 109-09 per cent. They also determined the rotations of the solutions obtained and found that these corresponded with the dextrose calculated from the copper reductions within about 1-4 per cent. Still more recently, Nerking (PJzXger’s Arohiu, 1901, 85, 320) has shown that glycogen derived from horse-flesh only yields 97 per cent, of the calculated amount of dextrose when 0.2-0*4 gram of the glycogen is boiled for 3-5 hours with 100 C.C. of 2.2 per cent. hydro- chloric acid. He also found, as the result of a number of separate experiments, that a longer or shorter period of action or a weaker acid all produced a lower result. Since no observations appear to have been made hitherto on the course of the hydrolysis of a single sample of glycogen, the behaviour of the three varieties of glycogen referred to above towards diluteHARDEN AND YOUNG: GLYCOGEN FROM YEAST.1231. acid was examined by heating them in 1 per cent. solution in semi- normal hydrochloric acid (1 *82 per cent. HCl) in a boiling water-bath, the course of the reaction being followed by removing samples at intervals, cooling rapidly to 1 5 O , observing the rotation in a 2 cm. tube, and then making up 10 C.C. of the liquid to 25 C.C. with 5 C.C. of normal potash and water, and estimating the reducing power of 20 C.C. of this solution with Fehling's solution by Brown, Morris, and Millar's method (Trans., 1897, 275).The experimental numbers were in every case plotted on squared paper against the time, a curve drawn through them, and the values taken from the curve. The results obtained show that in every case the rotation of the solution falls, a t first rapidly and then very slowly, whilst the cupric reducing power increases at first rapidly, and then slowly, until a maximum is reached -+ Time in hours. Sugar from cupric reduction. --... Botation in degrees. after which it slowly decreases, This is shown in the accompanying curves, in which the actual rotation and the dextrose calculated from the cupric reducing power are plotted against the time for a sample of yeast glycogen. In the following table, the numbers for each specimen of glycogen are given under two headings : (1) dextrose calculated from the cupric reducing power, and (2) dextrose calculated from the rotation, the latter being calculated on the assumption that the rotation of a 1 per cent.solution of dextrose at 15" in P 2 cm. tube is 1-05'. The average deviation from the smoothed curve of the numbers expressing the dextrose calculated from the cupric reducing power is only 0.7 per cent., and in only four observations does the deviation exceed 2 per cent. :1232 HARDEN AND YOUNG: GLYCOGEN FROM YEAST. Yeast glycogen. -. Rabbit glycogen. a. - Time in Hour: 1 2 3 4 5 6 7 8 9 10 11 12 13 1 4 15 16 - Oyster glycogen. 1'433 1.160 1.093 1'082 1.075 1.072 1.070 Dextrose calculated from - 0-700 0.940 1.035 1.058 1.065 1'068 Reduc- tion..I__ - 0.900 1.010 1'055 1.075 1 *08 4 1.088 1.086 1 -084 1 '083 1.081 1.078 1'069 - - - - - - Rota- tion. - 1 - i 1.055 I 1.052 I I - - 1.218 1.177 1 -1 43 1'122 1.098 1 -088 1.077 1.065 1-06 1.05 1'047 - -- - Reduc- tion. 0.916 1.013 1'051 1.070 1.078 1-082 1.085 1.083 1.081 1 '080 1-072 - - - - - Rota- 1, Reduc- tion. tion. Rota- tion. 3'24 2.20 1 :470 1-206 1.123 1'088 1'070 1.060 1 '055 1 -052 1'050 1.050 1 '048 1'047 - - Reduc- tion. 0.952 1 '032 1'065 1,080 1'089 1.090 1'092 1.090 1'088 1.087 1.082 - - - - - Rota- tion. 1.370 1.181 1.128 1'112 1.105 1 '1 02 1'101 1'100 1'100 1 '100 1 '099 - - - - - I n thecase of the rabbit glycogen, the maximum was obtained after 8 hours, and amounted to 96.3 per cent. of the calculated amount (the temperature in this case was slightly below 100' for the first hour).Of two separate samples of yeast glycogen, one yielded the maximum of 97.9 per cent. of the calculated in 7 hours, and the other a maximum of 97.6 per cent. also in 7 hours, Finally, the oyster glycogen produced a maximum of 98.3 per cent. of the calculated amount in 7 hours. It will be seen that when the maximum reducing power is attained, the rotation of the solution correspond# with an amount of dextrose almost exactly equal to that calculated from the reducing power. The differences observed are not greater than might be expected under the conditions of the experiment, a polarimeter reading of only O*0lo corresponding with a difference OF 1 per cent. in the dextrose. No special examination of the early stages of hydrolysis has as yet been made. From a consideration of the various properties of glycogen from yeast and from animal sources, it appears that no well marked difference exists betwesn these substances. All the three varieties examined by us have the same compositionROBERTSON: ATOMIC AND MOLECULAR HEATS OF FUSION. 1233 and the same optical activity. The differences observed between yeast glycogen and animal glycogen as regards opalescence, reaction with iodine, and behaviour towards dilute acids are not greater than those which exist between the two specimens obtained from the rabbit and the oyster. How far the degree of opalescence in solution and of coloration with iodine are characteristic of glycogen from a definite source is not a t present known, but the difference in this respect between the two samples of yeast glycogen examined by us and the divergence of our results in these particulars from those obtained by Clautriau seem t o indicate that these properties cannot be considered as essential. We are a t present engaged on the study of the early stages of acid hydrolysis of yeast glycogen and of the action on it of diastatic enzymes, and in particular of that of yeast itself. JENNER INSTITUTP OF PREVENTIVE MEDICINE.