406 VITAMIN Bi2 



nent appears therefore, to be phosphorylated ribazole, if no fragments re- 

 main undetected in the hydrolyzate. A ribazole phosphate was isolated as 

 the barium salt. Since the ribazole phosphate fails to react with periodate, 

 it is evident that it contains no a-glycol grouping, and the phosphoryl 

 group then must be located at C2 or C3 of the ribose moiety. ^^ The exact 

 location of the phosphate grouping awaits further study. 



3. Dg-l-AMINO-2-PROPANOL 



At an early stage in the progress of the structural elucidation of the vi- 

 tamin B12 molecule, it was noted that acid hydrolyzates, when subjected 

 to paper strip chromatography, showed the presence of a single ninhydrin- 

 reacting substance,^ later found to be Dg-l-amino-2-propanol.^^ 



An extended investigation of this substance by highly skilled paper strip 

 chromatographic techniques followed. ^^ Very small samples of vitamin B12 

 were subjected to hydrolysis in 2% hydrochloric acid at 100° for 6 hours. 

 Chromatography on paper strips using a n-butanol-acetic acid mixture 

 separated the ninhydrin-reacting fragment from the fluorescent benzimida- 

 zole derivatives. The ninhydrin-reacting substance was eluted and found to 

 be transparent to ultraviolet light ; it was concluded that the compound was 

 an aliphatic base. By analogy with known naturally occurring substances, 

 it was found that 2-aminopropanol showed identical behavior on paper 

 strip chromatography.-^' " Further investigation, however, proved that the 

 two substances were not identical. ^^' ^^ Oxidation of 2-aminopropanol with 

 permanganate produced alanine which could be identified by the ninhydrin 

 color reaction on a paper strip chromatogram. The ninhydrin-reacting sub- 

 stance from vitamin B12, on the other hand, gave an oxidation product 

 which was not alanine but which gave a yellow color with ninhydrin. 



The identification of the ninhydrin-reacting substance as Dg-l-amino-2- 

 propanol was accomplished by classical chemical means. ^^ The acid hydrolj'- 

 zate of vitamin B12 was extracted with butanol to remove the colored cobalt 

 complex. The amino fragment was benzoylated and isolated as the dibenzo- 

 ate. Since this compound was optically active, the structural possi!)ilities 

 were limited to compounds with an asymmetric center. The dibenzoate was 

 hydrolyzed, and the solution containing the amino fragment was oxidized 

 with sodium metaperiodate. From the oxidation mixture, acel aldehyde and 

 formaldehyde were identified as dimedone derivatives. This information in 

 conjunction with microanalytical data limited the possible structures to 

 an aminopropanol with adjacent finictional groups. Synthesis of D^-l-amino- 



36 D. E. Wolf, W. H. Jones, J. Valiant, and K. Folkors, J. Am. Chcm. Soc. 72, 2S20 



■ (1950). 

 " B. Ellis, V. Petrow, and G. F. Snook, /. Phnrm. Pharmacol. 1, 735 (1949). 



38 G. Cooley, B. Ellis, V. Petrow, J. Pharm. Pharmacol. 2, 128 (1950). 



39 G. Cooley, B. Ellis, V Petrow, J. Pharm. Pharmacol. 2, 535 (1950). 



