CCLXV. REACTIONS OF PYRUVIC ACID WITH
`THIOLACETIC ACID AND CYSTEINE.
`BY ERNST FRIEDMANN AND JOSEPH GIR9AVIJ1US.
`From the Biochemical Laboratory, Cambridge.'
`
`(Received 27 July 1936.)
`
`BAUMANN [1885,2] observed that the action of ethyl-, phenyl- and p-bromophenyl-
`mercaptans on pyruvic acid led to the exothermic formation of substances which
`could be recrystallized from benzene and possesMed well-defined melting-points.
`Though stable in the solid state and in benzene solution, these compounds were
`rapidly broken down to their original constituents when dissolved in water, and
`were regarded by Baumann as having the general structure RS . C(OH)(CH3).
`COOH. When treated with dry HCI the thiophenylpyruvic acid passes into the
`mercaptol (RS)2C(CH3) . COGH. The mercaptals and mercaptols [Baumann,
`1885, 1,2] have been extensively studied both by Baumann and his school and by
`other investigators, but the unstable "hemimercaptals" and "hemimercaptols"
`were little investigated until the recent work of Schonberg & Schuitz [1927],
`Levi [1932] and Schubert [1935; 1936] appeared. Their work is of interest both
`from the point of view of the chemistry of the formation of mercaptals [Fromm,
`1889; Levi, 1932] on which the present paper and recent work [Girsavicius &
`Heyfetz, 1935; 1936, 1] throw some light, and also from a biological aspect.
`Lohmann's [1932] well-known discovery that the transformation of methyl-
`glyoxal into lactic acid by glyoxalase requires the presence of reduced glu-
`tathione as a necessary and specific co-enzyme, together with the observations of
`Kuhnau [1931], of Lohmann [1932] and of Jowett & Quastel [1933], that gluta-
`thione reacts in aqueous solution with methylglyoxal to form a fairly unstable
`compound, have led to the hypothesis [Jowett & Quastel, 1933] that a hemi-
`mercaptal-like compound of methylglyoxal with glutathione forms a necessary
`intermediate stage in the enzymic reaction. Further evidence for this view has
`been advanced by Platt & Schroeder [1934] and by Girsavicius & Heyfetz [1936,
`2]. Possiblythe biological significance ofreactions ofthis type may extend beyond
`their participation in glyoxalase action [see Kuhnau, 1931; Bersin, 1935].
`In the following experiments the compounds formed between pyruvic acid
`on the one hand and thiolacetic acid [Baumann, 1885, 2; Bongartz, 1886] or
`cysteine on the other have been investigated more closely, the process of their
`formation and its reversal being studied.
`
`Reversible combination of pyruvic with thiolacetic acid.
`Thiolacetic acid (2-12 g.) and pyruvic acid (2.03 g.) in substance were mixed;
`the mixture became hot, then set to a mass of white crystals mixed with a good
`deal of viscous liquid. After cooling, the product was stirred up with ether and
`filtered through a sintered glass funnel. Yield: 2-63 g. (64 %) of crystalline
`substance. The ether washings (20 ml.) gave on evaporation a small additional
`crop. Ice cooling during the reaction did not appreciably alter the yield.
`1 The work described in this paper was carried out in 1933, while both authors were working
`at the Cambridge Laboratory.
`
`( 1886
`
`Eton Ex. 1047
`1 of 6
`
`

`

`1887
`
`PYRUVIC ACID AND SU COMPOUNDS
`The following investigation of the crystalline compound shows that it is made
`up of the two reactants in equimolecular amounts and that it is exceedingly labile
`in aqueous solution.
`Titration of -SH group. 61 mg. of the compound were dissolved in 5 ml. of
`ice-cold water and immediately titrated with 0-0996 N iodine to the first stable
`yellow tint. Iodine uptake was as rapid as speed of titration permitted. 3-50 ml.
`were required. Calculated for complete oxidation of the thiol group: 3-40 ml.
`[cf. Lucas & King, 1932].
`Titration of -CO group. 59 mg. titrated according to Cook & Clift [1932].
`Found: 6-58 ml. iodine. Calculated: 6-58 ml.
`Isolation of components. (1) 5-647 g. of the addition product were dissolved in
`30 ml. of water and 5 g. of NaHS03 added. The solution was repeatedly ex-
`tracted with ether and the combined ethereal extracts dried over Na2SO4;
`evaporation ofthe ether left 2-95g. of oily substance (102 % of original thiolacetic
`acid). Vacuum distillation (102-112°/10-16 mm.) gave 2-04 g. (70 %) of thiol-
`acetic acid, identified iodimetrically and acidimetrically.
`(2) 0-4975 g. of the
`compound in 10 ml. of water was treated with 50 ml. of 2N HCI, containing
`0-8 g. of 2: 4-dinitrophenylhydrazine. The precipitate was twice purified by
`dissolving in M Na2CO3 and reprecipitating by acidification. Yield: 0-752 g.,
`corresponding to 94-1 % of the theoretical amount of pyruvic acid dinitro-
`phenylhydrazone. In another experiment 91-6 % was obtained.
`Considering the losses involved in the isolation methods, these results, to-
`gether with the above described titrations, demonstrate the easy dissociation in
`aqueous solution of the addition compound. No conclusion can however be
`drawn from these experiments as to whether the compound is actually incapable
`of existing in aqueous solution, or whether it attains an equilibrium, dissociating
`more or less according to the dilution and the extent to which one or other com-
`ponent is removed.
`
`Reaction of pyruvic acid with cysteine.
`Titration methods were incapable of telling us anything about the state of
`the pyruvic-thiolacetic compound in aqueous solution, or even whether any
`reaction takes place, when the components are mixed in solution. The lower
`reactivity of the products obtained by the interaction of pyruvic acid with
`cysteine permits a more profitable application of iodine titration. We attempted
`also to obtain a clearer picture by a parallel study of the changes in rotatory
`power undergone by natural l-cysteine in presence of pyruvic acid.
`The readings were taken with Hg green (A= 5461 .) in a 2 dm. tube. The
`times were measured with a stopwatch and alternate readings were taken ap-
`proaching from the right and from the left.
`Fig. 1 shows the course of change of rotation of 0-2 M cysteine in aqueous
`solution in presence of 0-2, 0-4 and 0-6 M pyruvic acid. The solutions were always
`kept long enough for the rotation to reach a final value (several days). These end-
`values were: for 0-2 M pyruvic acid - 5.320, for 0-4 M -7-2°, and for 0-6 M
`- 7-44°. The experiments were carried out at 26-27'.
`Fig. 2 shows a number of curves obtained with cysteine and pyruvic acid in
`alcoholic solution (in view of the possible dissociating effect of water), at 370
`+ 0.10 in a jacketed polarimeter tube. The cysteine concentration was again
`0-2 M and those of pyruvic acid 0-2 and 0-4 M. The end-values, after several
`days in a thermostat at 370, were: 0-2 M pyruvic acid - 7-52°, 0-4 M - 9.560.
`An inspection of both sets of curves shows that they have their origin at some
`point corresponding to a weak negative rotation, which, as far as can be judged, is
`
`Eton Ex. 1047
`2 of 6
`
`

`

`1888
`
`E. FRIEDMANN AND J. G1RbAVICIUS
`the more marked, the higher the pyruvic acid concentration. Since cysteine itself
`is weakly dextr6rotatory (arrow in Fig. 1 shows the rotation of 02 M cysteine
`alone), we are clearly dealing with two successive effects: a practically in-
`stantaneous shift of rotation (of about 10 with our concentrations) with reversal
`of sign, succeeded by the gradual development of a strong laevorotation.
`
`00
`
`0
`
`0~~~~~~~~~~~~~~~0
`
`0
`
`~~~~~~~~~~~~~~~~~0
`
`-8°D
`
`50
`
`100
`
`150
`
`-8°1
`
`1
`
`acid
`
`Min.
`Min.
`Fig. 2.
`Fig. 1.
`In view of the complexity of these reactions, of the initial shift of rotation
`which seems to imply a rapid first stage' and the variable end-point which seems
`to indicate the attainment of an equilibrium, it is not surprising that the reaction
`courses observed fail to fit the ordinary equations. With equimolecular concen-
`trations of pyruvic acid and cysteine the bimolecular reaction constants increase
`with time; the unimolecular constant is reasonably uniform in the experiment
`in water, but rapidly falls in the alcoholic solution. In all the other experiments
`the bimolecular constant falls with time, the unimolecular constant of course
`even more so.
`
`Some colour reactions.
`An attempt to obtain some of the typical -SH reactions in presence of
`pyruvic acid provided additional evidence of a combination involving the
`sulphydryl group of the cysteine.
`FeC13. Added to an alkaline solution of cysteine, FeCl3 gives a purple colour,
`which fades on standing, but reappears on admitting oxygen [Harris, 1922;
`Michaelis & Barron, 1929]. To an alkaline (ammonia) mixture of 1 ml. M/4
`cysteine + 2 ml. M/4 Na pyruvate was added 0.1 ml. M/100 FeCl3. The purple
`colour faded more rapidly than in absence of pyruvic acid and reappeared less
`strongly on shaking; after repeating several times the cycle of reduction and re-
`oxidation the colour fails to reappear again. In absence of pyruvic acid the pro-
`cess can be repeated almost indefinitely. If the cysteine and pyruvic acid are
`allowed to stand a short while before adding FeCl3 the colour reaction is faint and
`rapidly vanishes irreversibly.
`1 Unpublished observations by one of us with P. A. Heyfetz have demonstrated a similar
`phenomenon in the reaction of GSH with methylglyoxal (iodimetric titration).
`
`Eton Ex. 1047
`3 of 6
`
`

`

`1889
`
`PYRUVIC ACID AND SH COMPOUNDS
`Nitroprusside. In ammoniacal solution cysteine gives immediately the well-
`known purple colour; this slowly fades to a brownish yellow. Pyruvic acid, under
`the same conditions, gives a slowly developing blue colour. On adding nitro-
`prusside to an ammoniacal solution of cysteine and pyruvic acid (in excess) a
`transient purple colour is observed, followed by the development of the blue
`pyruvic acid colour. Between the two a colourless interval is sometimes noticed,
`showing that the colour due to cysteine has actually vanished and is not merely
`covered by the deeper colour due to excess pyruvic acid.
`Methylene blue reduction. The reduction of methylene blue by cysteine in
`alkaline solution is inhibited by pyruvic acid, though not very strongly (Thunberg
`tube).
`
`Titration experiments.
`Absolute alcoholic solutions of cysteine hydrochloride (0-19 M) and of
`pyruvic acid (21 x 0-19 = 0 475 M), or mixtures of the two solutions were added
`by means of 0-5 or 1 ml. Ostwald pipettes to ice-cold 1-2 N HCl containing 1 ml.
`of 0-1014 N iodine and one drop of starch solution. The excess iodine was
`titrated with 0-1001 N Na2S203 from a microburette allowing an accuracy of
`0-001 ml.
`(1) 3 ml. of 1-3 N HCI+0-5 ml. cysteine solution+0-2 ml. pyruvic acid
`solution were mixed ice-cold. The mixture was kept in ice for various periods
`before adding iodine and (as rapidly as possible) titrating. As in all further ex-
`periments, the iodine consumption is given in terms of 0-1 N iodine. Cysteine
`alone (no pyruvic acid): 0-958, 0-955 ml.; with pyruvic acid: after 2 min.
`0-907 ml., after 10 min. 0-881 ml., after 15 min. 0-789 ml.
`(2) 15 ml. of the cysteine solution + 6 ml. of the pyruvic acid solution (that is,
`an absolute alcoholic solution, 0-136 M in each of these substances) were kept
`for 4 days. 0-5 ml. samples were added to a mixture of 1 ml. 2 N HCI +1 ml.
`
`4- 0-2 -y
`
`0
`
`0
`
`2
`
`6
`
`8
`
`4
`Min.
`Fig. 3.
`0-1 N iodine, cooled in a freezing mixture to -5° to -6°; on adding the alcoholic
`solution the temperature rose to about 0°. The titration vessel containing the
`mixture (a short wide tube with a tapered bottom) was transferred to a beaker
`containing ice in dilute brine, so that its temperature up to and during the
`titration was kept at -2° to 00. At the moment of adding the cysteine-pyruvic
`acid mixture to the iodine a stopwatch was started, which was stopped at the
`
`Eton Ex. 1047
`4 of 6
`
`

`

`1890
`
`E. FRIEDMANN AND J. GIRSAVICIUS
`end of the titration. The titration was started after the iodine had been acting
`for various periods. Fig. 3 shows the course of iodine uptake up to 8 min. The
`contrast with the experiments with thiolacetic acid, where the whole of the
`iodine corresponding to the -SH introduced is immediately consumed, is striking.
`It is not clear from these results whether the gradual uptake of iodine is
`entirely due to its action on the compound, or whether a certain amount of
`dissociation takes place spontaneously when the alcoholic mixture is added to
`water+ HCl. The following experiment was designed to answer this question.
`0 5 ml. of the alcoholic mixture of pyruvic acid and cysteine was added to 1 ml.
`HCl+ 1 ml. iodine, as above, or else only to HCI, the iodine being added some
`time later. The total times up to the end of the titration, were kept as nearly as
`possible alike.
`
`Iodine uptake
`ml.
`Iodine at once. Titration finished in 2 min. 12 sec.
`0-082
`Iodine at once. Titration finished in 8 min. 14 sec.
`0-213
`Iodine added after 64 min. Total time 8 min. 15 sec.
`0 097
`In the following experiment the alcoholic solution was added to 4 ml.
`2 N HCI + 3 ml. of water; iodine present or added later.
`
`Iodine uptake
`ml.
`
`Iodine at once. Titration finished after 7 min.
`0-476
`Iodine after 64 min. Total time 7 min. 4 sec.
`0-173
`Little, if any, sulphydryl is free in the aqueous HCI solution in absence of
`iodine. This accords well with the polarimetric observations, which showed that
`the reaction follows much the same course in water and in alcohol. (It must be
`noted, that on adding the alcohol mixture to the dilute HCI not only does a
`change in the nature of the medium take place, but also a considerable change in
`volume). Judging from the second series of figures higher dilution increases the
`effect of the iodine and perhaps has itself a certain dissociating effect.
`
`CONCLUSIONS.
`Girsavicius & Heyfetz [1935] have recently studied in some detail the
`reaction between glutathione and methylglyoxal in aqueous solution. Two
`main conclusions were reached: (a) a true equilibrium is established between the
`free components and the reaction product; (b) the reaction rates in both directions
`depend on the pH, being slow in strongly acid solution and extremely rapid as
`the solution approaches neutrality (compare Girsavicius & Heyfetz [1936, 1]
`where this observation is elaborated). The results described in the present paper
`fit in with the view that here too reactions of the same kind as that between
`glutathione and methylglyoxal may take place, and that such differences as are
`found, for instance between cysteine and thiolacetic acid, are expressions of dif-
`ferent reaction rates and equilibrium constants. Baumann [1885, 1,2], in investi-
`gating the products of spontaneous reaction between pyruvic acid and various
`mercaptans, already mentioned the differences in their properties, including their
`stability. Great differences in the reactivity (reaction rates) ofthe -SH group, in
`accordance with the structure of the molecule of which it forms a part, are also
`mentioned by Michaelis & Schubert [1934]. It must be admitted, however, that
`the great difference in behaviour shown by the compounds of pyruvic acid with
`thiolacetic acid on the one hand and with cysteine on the other suggests the
`possibility of a different type of bond. Schubert [1935] has shown that with
`
`Eton Ex. 1047
`5 of 6
`
`

`

`PYRUVIC ACID AND SH COMPOUNDS
`
`1891
`
`eysteine and with thiocarbamide1) methylglyoxal forms not addition, but cyclic
`condensation compounds, apparently involving the nitrogen as well as the
`sulphur. Recently [1936] he has extended this observation to the compounds of
`cysteine with certain aldehydes. It is true that Schubert claims [1936] that the
`compound of cysteine with pyruvic acid is an addition, not a condensation com-
`pound; but the reaction he describes proceeded only for one day at about O and
`may, perhaps, not have been complete. At any rate we consider that the nature
`of the compound finally formed between pyruvic acid and cysteine, a comnpound
`fairly stable to iodine and endowed with strong laevorotation, still remains to be
`settled.
`
`SUMMARY.
`Reactions taking place in the systems pyruvic acid + thiolacetic acid and
`pyruvic acid+ cysteine were investigated. Pyruvic acid in substance combines
`exothermically with pure thiolacetic acid to form a white crystalline compound.
`Dissolved in water this behaves like a mixture of free pyruvic acid and free
`thiolacetic acid, both in its reactions with iodine or bisulphite and in the
`possibility of isolating the original components from the aqueous solution. Cys-
`teine in aqueous or alcoholic solution reacts slowly with pyruvic acid, as shown
`polarimetrically and by iodimetric titration. The compound is reasonably stable,
`as judged from its behaviour towards iodine and some colour reagents. A
`hypothesis is outlined which connects these reactions with the recently investi-
`gated reaction of glutathione with methylglyoxal, and some alternative explana-
`tions of the reaction of pyruvic acid with cysteine are discussed.
`1 The condensation compound of methylglyoxal with thiocarbamide seems to have been first
`obtained by Sjollema & Kam [1916].
`
`REFERENCES.
`
`Baumann (1885, 1). Ber. dtsch. chem. Ges. 18, 258.
`(1885, 2). Ber. dtsch. chem. Ges. 18, 883.
`Bersin (1935). Ergebn. Enzymforsch. 4, 68.
`Bongartz (1886). Ber. dtsch. chem. Ges. 19, 1933.
`Cook & Clift (1932). Biochem. J. 26, 1788.
`Fromm (1899). Liebigs Awenn. 253, 135.
`Girsavicius & Heyfetz (1935). Biochem. Z. 276, 190.
`(1936, 1). Biochemnia, Mlloscowv (in the Press).
`(1936, 2). Biochemia, Mloscowi (in the Press).
`Harris (1922). Biochem. J. 16, 739.
`Jowett & Quastel (1933). Biochem. J. 27, 486.
`Kiihnau (1931). Biochem. Z. 243, 14.
`Levi (1932). Gazz. chim. ital. 62, 775.
`Lohmann (1932). Biochem. Z. 254, 332.
`Lucas & King (1932). Biochem. J. 26, 2076.
`Michaelis & Barron (1929). J. biol. Chem. 83, 191.
`& Schubert (1934). J. biol. Chem. 106, 331.
`Platt & Schroeder (1934). J. biol. Chein. 104, 281.
`Schonberg & Schiitz (1927). Ber. dtsch. cheein. Ges. 60, 2344.
`Schubert (1935). J. biol. Chern. 111, 671.
`(1936). J. biol. Chem. 114, 341.
`Sjollema & Kam (1916). Rec. Trav. chim. Pays-Bas, 36, 180.
`
`Eton Ex. 1047
`6 of 6
`
`