402 VITAMIN Bi2 



tography of the aerated solution allowed separation of two fractions which 

 still possessed absorption maxima at 280 m/x. 



C. CONSTITUTION 



Soon after the initial announcements of the isolation of vitamin B12, the 

 investigation of the general nature or class of the compound began. Acid 

 hydrolysis failed to indicate the presence of amino acids'^ but did reveal the 

 presence of a single ninhydrin-reacting constituent.^ The presence of phos- 

 phate ion and ammonia was recognized in the hydrolyzates.-''' -^ Analogy 

 with other naturally occurring compounds possibly related metabolically to 

 vitamin B12 suggested the possibility of a pterin, nucleotide, or polypyrrole 

 structure. Evidence from physical and chemical experiments, particularly 

 ultraviolet absorption data,'^ eliminated a pterin-like structure. Hydrolytic 

 fragments characteristic of the purines have likewise been lacking.* How- 

 ever, alkaline fusion has produced pyrroles^^ as judged by their characteris- 

 tic color reaction with p-dimethylaminobenzaldehyde (Ehrlich reagent). 

 More extensive information on the exact nature of the pyrrole-like moiety 

 will surely be forthcoming when appropriate methods to degrade the stable 

 cobalt complex* have been devised. The biological relationship of vitamin 

 B12 to the general hematopoetic metabolism seems to indicate that vitamin 

 B12 is a polypyrrole related in some way to hemin or the bile pigments. 



1. 5,6-DlMETHYLBENZIMIDAZOLE 



Despite the fact that acid hydrolysis of vitamin B12 fails to completely 

 rupture the cobalt complex, it has been found to liberate certain portions 

 of the molecule and, in doing so, to destroy its biological activity and bring 

 about changes in the ultraviolet absorption spectrum. Hydrolysis of vi- 

 tamin B12 in 6 TV hydrochloric acid at 150° for 24 hours produced 5,6- 

 dimethylbenzimidazole (I), which was isolated by continuous extraction of 

 the alkaline hydrolyzate with chloroform.-^- ^^ The general class identifica- 

 tion of this degradation product was determined by appraisal of its molecu- 



CHi 



\/\N 



CH; 



CeHeCOCl 



NaOH 



CH- 



\/ 



NHCOC6H5 



NHCOCfiHs 



I II 



lar composition and by comparison of its ultraviolet absorption spectrum 

 with that of known bcnzimidazoles. The characteristic ultraviolet absorp- 



2' B. Ellis, V. Petrow, and G. F. Snook, J. Phann. Pharmacol. 1, 2cS7 (19-49). 



28 B. Ellis, V. Petrow, and G. F. Snook, J. Phann. Pharmacol. 1, 950 (1949). 



29 N. G. Brink and K. Folkcrs, J. Am. Chem. Soc. 71, 2951 (1949). 

 3« N. G. Brink and K. Folkers, J. A?/i. Chem. Soc. 72, 4442 (1950). 



