Io M. WINITZ 
to distinguish it from the neighboring purple glutamine spot”. Serine and threonine 
may be identified either by the red-brown color developed upon treatment of the paper 
first with periodate and subsequently with the Nessler reagent *°-*’, or by the violet 
color produced by reaction first with alkaline hypochlorite and then with 1, 2-dinitro- 
benzene?’. Sulfur-containing amino acids may be detected by their bleaching action 
when the paper is sprayed with a solution of platinic iodide”: 76 29, and glycine by 
the green color it gives with o-phthalaldehyde*; with this latter reagent, histidine and 
tryptophan exhibit a blue-green color and a gray color, respectively. An interesting 
color reagent is I, 2-naphthoquinone-4-sulfonate which in sodium carbonate solutions 
produces reddish spots with proline and hydroxyproline, a green spot with glycine, 
and blue-gray to violet colors with all other common amino acids*?: , 
These represent only a few of the available, more or less specific color reactions 
exhibited by individual amino acids. They have been treated here at length to re- 
emphasize the fact that all too many amino acids are already known which display 
identical chromatographic behavior in a variety of solvent systems, and may even 
reveal identical elemental analyses. Hence, in the search for new amino acids, addi- 
tional criteria may sometimes be necessary in order to expose those elusive amino 
acids that may otherwise evade detection by remaining concealed behind known 
amino acids of closely related structure. 
STRUCTURAL DETERMINATION BY MEANS OF CHEMICAL DEGRADATION 
If we now return to the condensation product of pyruvic acid and glycine considered 
earlier, it will be recalled that the unique color reactions exhibited by this material, 
by eliminating from consideration y-hydroxyglutamic acid, suggested that it might be 
/-hydroxy-f-methylaspartic acid. In order to establish the structure of this material 
more firmly, degradative studies were undertaken. Now it is well known that most 
a-amino acids liberate one mole of carbon dioxide per mole of compound upon being 
subjected to manometric ninhydrin—CO, analysis*’. Certain compounds such as 
aspartic acid and the diaminodicarboxylic acids, like a, e-diaminopimelic acid and 
cystine, are exceptional in that two moles of carbon dioxide per mole of amino acid 
are liberated. Some years ago, we had occasion to synthesize and subsequently analyze 
various substituted aspartic acids, such as -methylaspartic acid*, 6-hydroxyaspartic 
acid and a-aminotricarballylic acid®®, and noted that these compounds also liberated 
two moles of carbon dioxide per mole of amino acid (Fig. 3). As the unknown synthetic 
condensation product behaved in a comparable manner upon manometric ninhydrin— 
CO, analysis, a substituted aspartic acid structure was suggested and this constituted 
further evidence in favor of a 6-hydroxy-f-methylaspartic acid structure. y-Hydroxy- 
glutamic acid, on the other hand, behaved like the usual a-amino acid in that only 
the expected one mole of carbon dioxide was released. Further evidence favoring the 
B-hydroxy-(-methylaspartic acid structure was revealed by the fact that the material 
was converted to /-methylaspartic acid upon reduction with hydriodic acid and that 
it showed a positive periodate reaction, typical of an a-amino-f-hydroxy acid. The 
structural elucidation of this compound illustrates only a few of the degradative 
processes employed in the characterization of a new amino acid. As these processes 
are numerous in number and vary markedly with the nature and structure of the amino 
acid under consideration, their further elaboration is here precluded. 
References p. 22/24 
