An observation of remarkable interest, which promises to throw light on several important features of the biochemistry of yeast, was made in 1911, and has since then formed the subject of detailed investigation by Neuberg and a number of co-workers.

It was found that yeast had the power of rapidly decomposing a large number of hydroxy-and keto-acids [Neuberg and Hildesheimer, 1911; Neuberg and Tir, 1911; see also Karczag, 1912, 1, 2]. The most important among these are pyruvic acid, CH₃·CO·COOH, and a considerable number of other aliphatic a-keto-acids which are decomposed with evolution of carbon dioxide and formation of the corresponding aldehyde:—

The reaction is produced by all races of brewer's yeast which have been tried, as well as by active yeast preparations and extracts and by wine yeasts [Neuberg and Karczag, 1911, 4; Neuberg and Kerb, 1912, 2]. The phenomenon can readily be exhibited as a lecture experiment by shaking up 2 g. of pressed yeast with 12 c.c. of 1 per cent. pyruvic acid, placing the mixture in a SchrÖtter's fermentation tube, closing the open limb by means of a rubber stopper carrying a long glass tube and plunging the whole in water of 38–40°. Comparison tubes of yeast and water and yeast and 1 per cent. glucose may be started at the same time, and it is then seen that glucose and pyruvic acid are fermented at approximately the same rate [Neuberg and Karczag, 1911, 1]. If English top yeast be used it is well to take 0·5 per cent. pyruvic acid solution and to saturate the liquids with carbon dioxide before commencing the experiment. The production of acetaldehyde can be readily demonstrated by distilling the mixture at the close of fermentation and testing for the aldehyde either by Rimini's reaction (a blue coloration with diethylamine and sodium nitroprusside) or by means of p-nitrophenylhydrazine which precipitates the hydrazone, melting at 128·5° [Neuberg and Karczag, 1911, 2, 3]. [p082]

As the result of quantitative experiments it has been shown that 80 per cent. of the theoretical amount of acetaldehyde can be recovered. The salts of the acids are also attacked, the carbonate of the metal, which may be strongly alkaline, being formed. Thus taking the case of pyruvic acid, the salts are decomposed according to the following equation:—

Under these conditions a considerable portion of the aldehyde undergoes condensation to aldol [Neuberg, 1912]:—

This change appears to be due entirely to the alkali and not to an enzyme since the aldol obtained yields inactive -hydroxybutyric acid on oxidation [Neuberg and Karczag, 1911, 3; Neuberg, 1912]. The various preparations derived from yeast which are capable of producing alcoholic fermentation also effect the decomposition of pyruvic acid in the same manner as living yeast. They are, however, more sensitive to the acidity of the pyruvic acid, and it is therefore advisable to employ a salt of the acid in presence of excess of a weak acid, such as boric or arsenious acid, which decomposes the carbonate formed but has no inhibiting action on the enzyme [Harden, 1913; Neuberg and Rosenthal, 1913].

As already mentioned the action is exerted on a-ketonic acids as a class and proceeds with great readiness with oxalacetic acid, COOH·CH₂·CO·COOH, all the three forms of which are decomposed, with a-ketoglutaric acid, and with a-ketobutyric acid. Hydroxypyruvic acid CH₂(OH)·CO·COOH is slowly decomposed yielding glycolaldehyde, CH₂(OH)·CHO, and this condenses to a sugar [Neuberg and Kerb, 1912, 3; 1913, 1]. Positive results have also been obtained with diketobutyric, phenylpyruvic, p-hydroxyphenylpyruvic, phenylglyoxylic and acetonedicarboxylic acids [Neuberg and Karczag, 1911, 5].

Relation of Carboxylase to Alcoholic Fermentation.

With regard to the relation of carboxylase to the process of alcoholic fermentation, nothing definite is yet known. As Neuberg points out [see Neuberg and Kerb, 1913, 1] the universal presence of the enzyme in yeasts capable of producing alcoholic fermentation, and the extreme readiness with which the fermentation of pyruvic acid takes place create a [p083] strong presumption that the decomposition of pyruvic acid actually forms a stage in the process of the alcoholic fermentation of the sugars. On the other hand Ehrlich's alcoholic fermentation of the amino-acids (p. 87) provides another function for carboxylase—that of decomposing the a-ketonic acids produced by the deaminisation of the amino-acids. It must be remembered in this connection that carboxylase is not specific in its action, but catalyses the decomposition not only of pyruvic acid but also of a large number of other a-ketonic acids, including many of those which correspond to the amino-acids of proteins and are doubtless formed in the characteristic decomposition of these amino-acids by yeast. Carboxylase undoubtedly effects one stage in the production of alcohols from amino-acids, whether it is also the agent by which one stage in the alcoholic fermentation of sugar is brought about still remains to be proved.

A comparison of the conditions of action of carboxylase and zymase has revealed several interesting points of difference. Neuberg and Rosenthal [1913] have observed that the fermentation of pyruvic acid by maceration extract commences much more rapidly than that of glucose and interpret this to mean that in the fermentation of glucose a long preliminary process occurs before sufficient pyruvic acid has been produced to yield a perceptible amount of carbon dioxide. The long delay (3 hours) which they sometimes observed in the action of maceration juice on glucose is however by no means invariable (see p. 46), but in any case indicates that the sugar fermentation can be affected by conditions which are without influence on the pyruvic fermentation. A similar conclusion is to be drawn from the fact that the pyruvic acid fermentation is less affected by antiseptics than the glucose fermentation [Neuberg and Karczag, 1911, 4; Neuberg and Rosenthal, 1913], chloroform sufficient to stop the glucose fermentation brought about by yeast or dried yeast being usually without effect on the fermentation of the pyruvates either alone or in presence of boric or arsenious acid. A more important difference is that carboxylase decomposes pyruvic acid in the absence of the co-enzyme which is necessary for the fermentation of glucose [Harden, 1913; Neuberg and Rosenthal, 1913]. This can be demonstrated experimentally by washing dried yeast or zymin with water (see p. 63) until it is no longer capable of decomposing glucose (Harden), or by allowing maceration extract to autolyse or dialyse until it is free from co-enzyme (Neuberg and Rosenthal). The zymase of maceration extract is moreover inactivated in 10 minutes at 50–51°, whereas after this treatment the carboxylase is still active. [p084]

The only conclusion that can be legitimately drawn from these highly interesting facts is that if the decomposition of pyruvic acid actually be a stage in the alcoholic fermentation of glucose the soluble co-enzyme is required for some change precedent to this, so that in its absence the production of pyruvic acid cannot be effected.