SOLUBLE, NITROGENOUS CONSTITUENTS OF PLANTS 39 
was sought by co-chromatography of natural and synthetic products on ion-exchange 
resins by the method of SPACKMAN, STEIN AND Moore®*, the evidence in Fig. 14 was 
obtained. This shows that the synthetic product was 98°% the unnatural diastereomer, 
namely allo-y-hydroxyvaline, and it contained only 2% of the natural diastereomer. 
These two compounds had superimposed upon paper chromatograms. Final proof 
requires a synthesis of sufficient of the natural diastereomer to permit optical resolu- 
tion and direct comparison of the optically active forms with the natural product. 
Structural Relationships of Certain New Amino Acids 
CHO COOH COOH COOH 
HOCH & NHoCH HCNH 2 NHoCH 
CH5 OH CHEOnhin =,.4) 4 CHaOH si aan! HCOH 
! 
CH3 
l- glyceraldehyde l-serine D-serine Le-threonine 
(natural) 
CHO CHO COOH COOH COOH 
HOCH HOCH NHdCH NHoCH & NHoCH 
ert Pieonea OCR chs. GH eas Gcria: 
CHo5OH CH50H CH OH CHOAc & CH OH 
Threose Erythrose Homoserine O-Acetylhomo:  ls~-threo- 
ees serine shydroxyvaline 
On oxidation of natural y-hydroxyvaline to the methylaspartic acid, it was possible 
to show that the natural compound gave /-methylaspartate. BARKER? had been able 
to show that a specific enzyme converts one of the four possible 6-methylaspartic 
acids to mesaconic acid and ammonia and, on request, he tested the methylaspartic 
acid from natural y-hydroxyvaline and found it to be a substrate for this enzyme. 
These observations, therefore, establish the configuration of the hydroxyvaline from 
Kalanchoe, and its relations to serine and to threose are shown in the chart above, 
together with the structure of homoserine and its o-acetyl compound (Fig. 15). 
It is now possible to state the structure of another hydroxy acid which occurs in 
plants. From apples, an amino acid was isolated by UrRBAcH and by Hurm_. These 
two isolates were shown to be identical by their infrared-absorption spectra (HULME 
AND STEWARD!®), but there was no conclusive distinction between the two alternative 
structures A and B (see p. 40). 
By the application of mass spectroscopy to these compounds, BrEMANN has now shown 
that the masses corresponding to formula B as shown at C establish formula B as 
References p. 42 
