THE SYNTHESIS OF D-ERYTHRO- AND D-THREO-a-AMINO-

In 1931 Klenk and Diebold (1) reported that sphingosine, upon oxidative degradation, gave rise to myristic acid and an optically active dihydroxyaminobutyric acid, [o(]A3 = 33.4”, which these authors characterized as an cY-amino-p ,r-dihydroxy-n-butyric acid. 5 years later Fischer and Feldmann (2), starting from D-glyceric aldehyde, synthesized an optically active oc-amino-p ,r-dihydroxy-n-butyric acid, [a]:’ = 13.7”, and, in order to explain the difference in the specific rotation of their amino acid and that of Klenk and Diebold, suggested that the amino acid with [LY]~’ = 13.7“ was a mixture of the two expected disastereoisomers, one of which was identical with, or an antipode of, the amino acid obtained from sphingosine. We repeated the synthesis of Fischer and Feldmann and obtained, as they did, the amino acid with [cY]:~ = -13.7” in good yield. However, in addition another cr-amino-p , r-dihydroxy-n-butyric acid, with [cr]i4 = l&O’, was isolated from the reaction mixture. It is the purpose of this communication to show that the a&no acid with [LY]~~ = -13.7” is not a mixture of disastereoisomers, but is D-threo-a-amino-p , y-dihydroxy-nbutyric acid, and that the amino acid with [cr]E4 = 16.0” is D-erythroar-amino-p , r-dihydroxy-n-butyric acid. An a-benzamido-p, y-dihydroxy-n-butyrolactone, m.p. 210-211°, was obtained when the cY-amino-p , y-dihydroxy-n-butyric acid with [o(]E4 = 13.7” was benzoylated. This la&one when treated with phefiylhydrazine gave an a-benzamido-fi , y-dihydroxy-n-butyr-(N-phenyl)hydrazide, [a]:4 = -15.9”. The partial hydrolysis of this phenylhydrazide led to the regeneration of the a-benzamido-p , y-dihydroxy-n-butyrolactone, m.p. 210-211°, and when this lactone was hydrolyzed the original cr-amino-