A comparison of alanine and glucose as precursors of serine and glycine.

The importance of glycine in animal and plant metabolism as a precursor of many nitrogenous compounds, such as creatine, porphyrins, chlorophyll, uric acid and nucleic acid purines is now well established (for a recent review, see Arnstein, 1954). Relatively little is known, however, about the metabolic origin of this amino acid from non-nitrogenous precursors. That glycine can be synthesized by mammals has been recognized for almost half a century (Magnus-Levy, 1907; Wiechowski, 1906), but conclusive proof that it is a nutritionally dispensable amino acid has been obtained only recently in feeding experiments with synthetic diets (Rose, Burr & Sallach, 1952). The metabolism of glycine is closely linked to that of serine, since both amino acids are readily converted into one another both in vivo (Shemin, 1946; Sakami, 1948) and in vitro (Blakley, 1954; Kisliuk & Sakami, 1954). The biosynthesis of either serine or glycine from three-carbon or twocarbon compounds in sufficient amounts would therefore be adequate for the animal's requirements, and both mechanisms have in fact been proposed for the biosynthesis of glycine. Thus, the conversion of [2-14C] and [1-18C]pyruvate into glycine in vivo has been reported (Anker, 1948), and it was suggested that serine was formed as an intermediate. More recently it has been shown that glycollic acid (Weinhouse & Friedmann, 1951; Chao, Delwiche & Greenberg, 1953) and glyoxylic acid (Weinhouse & Friedmann, 1951; Weissbach & Sprinson, 1953) can be used for the biosynthesis of glycine, presumably by reductive amination of glyoxylic acid. It is not clear, however, whether these C2 acids are formed in sufficient amounts to account quantitatively for glycine biosynthesis; the available evidence concerning the relative importance of C3 and C2 precursors has so far been inconclusive, although it has been shown that, on a diet free of both serine and glycine, the biosynthesis of serine by the rat significantly exceeds that of glycine (Arnstein & Neuberger, 1953b). The recent observation that the biosynthesis of glycine from non-nitrogenous precursors is markedly reduced in pteroylglutamic acid de* Visiting worker; permanent address: Chemical Institute, Faculty of Science, Zagreb, Yugoslavia. ficiency (Arnstein & Stankovic, 1956) suggested to us that it should be possible to differentiate between these alternative mechanisms by comparing the effect of pteroylglutamic acid deficiency on the biosynthesis of glycine and serine from a non-nitrogenous precursor, such as glucose, since this vitamin deficiency reduces the metabolic interconversion of the two amino acids (Elwyn & Sprinson, 1950b; Totter, Kelley, Day & Edwards, 1950). The results of such experiments, which are now reported, show that the biosynthesis of serine from glucose or alanine is independent of the glycineerine reaction, whilst an important part of glycine must be synthesized from serine. Neither alanine nor pyruvate, which arises from it by transamination, were efficient serine precursors when compared with glucose, and it is suggested that serine is formed from a C3 precursor which is closely related to an intermediate of glycolysis, such as a triose or glyceric acid. Some of these results have already been reported in a preliminary communication (Arnstein, 1955).