54 A. GOTTSCHALK 



product released enzymatically from the inhibitory mucoproteins. Subsequent 

 work was, therefore, focused at the identification of the split product, the 

 elucidation of its molecular structure, and its location and linkage within the 

 inhibitory mucoids. 



The split product shared its unusual properties with an acid first isolated in 

 crystalline form by Blix (1936) from bovine salivary mucin. This acid was 

 shown to have the composition C 13 H 21 O 10 N and to contain an iV-acetyl group, 

 an O-acetyl group (which was very easily lost), a reducing group, a primary 

 alcohol group, and a carboxyl group (Blix et al., 1955, 1956). Digesting urine 

 mucoprotein with the virus enzyme, Klenk et al. (1955) confirmed Gottschalk's 

 results, crystallized the split product, and identified it as iV-acetylneuraminic 

 acid. Neuraminic acid was first isolated as the methoxy derivative from brain 

 gangliosides (Klenk, 1941). The properties of iV-acetylneuraminic acid 

 (NANA) suggested that it differed from Blix's acid only by the lack of the 

 O-acetyl group. Therefore, from Blix's data it would have the formula 

 C u H 19 9 N; but Klenk et al. (1955) assigned to NANA the formula C 12 H 2i O 10 N. 

 This discrepancy and the many possible interpretations of such formulas made 

 it impossible at this stage to conceive a molecular structure of neuraminic 

 acid. Obviously the preparation from acetylated neuraminic acid of a structur- 

 ally well-defined degradation product was required. This information was 

 provided by the isolation of pyrrole-2-carboxylic acid (PYCA) from alkali- 

 treated bovine submaxillary gland mucoprotein (BSM) and urinary mucopro- 

 tein (UM) and by the degradation of the enzymatically released split product 

 to PYCA by weak alkali (Gottschalk, 1953, 1954b, 1955a). On the basis of 

 these data NANA was visualized as an aldol condensation product of N- 

 acetylhexosamine with pyruvic acid, and neuraminic acid as the correspond- 

 ing condensate of hexosamine with pyruvic acid; the degradation of NANA to 

 PYCA, involving loss of acetyl and of water and reverse aldolization, was 

 depicted, as shown in Fig. 1 (Gottschalk, 1955b, 1956a). The correctness of 

 the proposed structure was strongly supported by the synthesis in 20 % 

 yield of PYCA from D-glucosamine and pyruvic acid (Gottschalk, 1955b, 

 1957a) and by the fragmentation of NANA to iV-acetylglucosamine, C0 2 , and 

 a 2-C compound (Kuhn and Brossmer, 1956b). It was conclusively proved by 

 the chemical synthesis of NANA from iV-acetyl-D-glucosamine and carbo- 

 xylated pyruvic acid (Cornforth, et al., 1957). The crystalline synthetic 

 product was identical with crystalline NANA from biological sources with 

 regard to melting point, elementary analyses (C, H, N), infrared spectrum, 

 X-ray powder diagram, specific optical rotation, and chromatographic 

 behavior (Cornforth, et al., 1958). By the synthesis, the general structure 

 of NANA and its stereochemistry at C , C 7 and C 8 were defined. Recently 

 Comb and Roseman (1958) have shown that the N-acetylhexosamine frag- 

 ment of NANA is N-acetylmannosamine. Apparently under the alkaline 



