THE Jon~ivar. OF BIOLOGICAL CHEYISTRY

In pursuing quantitative histochemical investigations aimed at demonstrating the biochemical architecture of human frontal cortex (l-4), we found it necessary to determine ganglioside sialic acid in submicrogram amounts. Even when adapted to the microscale, spectrophotometric methods lack the sensitivity required for studying nervous tissue samples containing less than 0.3 pg of sialic acid. Since fluorometric methods usually possess a sensitivity 100to lOOO-fold that of spectrophotometric methods, the possibility of devising a fluorometric technique for sialic acid measurement was investigated. N-Acetylneuraminic acid of brain gangliosides is a 9 carbon or-keto acid. Such compounds readily undergo decarboxylation on treatment with hot mineral acids. Formation under such conditions of the deacylated, decarboxylated derivative of NAN,’ 2-deoxy-4-amino-octose, has been suggested by Gottschalk (5) as the basis for the reaction of NAN in (a) the diphenylamine test; (b) the resorcinol-HCl test; and (c) humin formation at concentrations of hot mineral acid as low as 0.1 N. The 2-deoxy sugars and in general aldehydes of the type R-CHZ-CHO will react wit,h 3,5-diaminobenzoic acid when heated in mineral acids to yield highly fluorescent quinaldines. The product of the Doebner and Von Miller quinoline synthesis involving acetaldehyde and 3,5-diaminobenzoic acid has been characterized by Velluz, Amiard, and Pesez (6), who also reported (7, 8) that deoxyribose and DNA undergo the reaction. The principle was applied subsequently to the ultramicro assay of DNA by Kissane and Robins (9), of succinic semialdehyde by Salvador and Albers (lo), and of 2-deoxy-n-glucose by Blecher (II). When NAN was heated at 100” under conditions similar to those used in the DNA method (9), a green fluorescent product was obtained, as described in a preliminary report (12). Hexoses, however, were serious interfering compounds in the method as it was originally carried out, presumably because an intermediate in the conversion of hexose to levulinic acid by hot mineral acids is w-OH-levulinic aldehyde, an a-methylene aldehyde (13). In a search for reaction conditions which would favor the synthesis of the sialic acid product over that of hexoses, systematic studies were made of the effects on the fluorescence yield of variations in 3,5-diaminobenzoic acid and HCl concentrations as well as in temperature and duration of the heating period. Conditions were found enabling reduction in the interference by hexoses t,o 1.6% on a molar basis. The sensitivity of t.he resulting method may be indicated by the fact that the