II. CHEMISTRY AND INDUSTRIAL PREPARATION 45 



Recently Harris ct al}^^ have determined the biopotency of neovitamin 

 A alcohol and neovitamin A acetate. They found that the alcohol has a 

 potency of 87.1 ± 3.25% of the potency of all-/ra7js vitamin A and the 

 acetate has a potency of 83.2 ± 4.28% of the potency of all-^mns vitamin 

 A acetate. Furthermore, the authors came to the conclusion that in a mix- 

 ture of the two stereoisomers each isomer exerts its characteristic physi- 

 ological eflfect independently of the presence of the other isomer. The same 

 authors also studied the extent of interconversion of the two stereoisomers 

 in vivo. They found that rats have the ability to convert neovitamin A to 

 the all-/rans form, and vice versa, and they seem to store in their livers a 

 mixture of the two isomers containing approximately 12% of neovitamin 

 A and 88 % of the all-/rans vitamin A, regardless of which pure isomer is 

 fed. This ratio seems to be difTerent with different species; in fish livers it 

 is about 35:65. 



Chemical interconversion of neovitamin A and vitamin Ai has also been 

 accomplished through their anthraquinone /3-carboxylate esters by treat- 

 ment with traces of iodine in benzene solution. The neovitamin A ester was 

 converted to the correspondhig vitamin Ai ester to the extent of about 70 % 

 in 2 hours at 25°, whereas the reverse was accomplished under similar con- 

 ditions to the extent of only 30%.^^^ 



From the chemical and biological behavior of neovitamin A, it was con- 

 cluded that one of the double bonds must have the cis configuration, and, 

 since neovitamin A dehydrates at a much slower rate than vitamin Ai, 

 the cis double bond must be the one closer to the hydroxyl group. This 

 reasoning is not entirely clear, since the dehydration can be almost com- 

 pletely prevented by interposing a triple bond between the hydroxyl group 

 and the double bond in the ring. For example, 1,2-dehydrovitamin A, 

 which forms both the red and the yellow anthraquinone )S-carboxylate 

 esters, is not easily dehydrated with methanolic hydrogen chloride. ^^^ Fur- 

 thermore, the red color of the anthraquinone jS-carboxylate and the position 

 of the ultraviolet absorption maximum have not been adequately explained 

 on the basis of the cis configuration. It is well known that the cis isomers 

 in the carotenoid group (Section II. A) absorb at lower wavelengths than 

 the corresponding all-/mns isomers. Moreover, the Z-cis vitamin A alde- 

 hj'dc seems to follow this rule.'" 



c. Evidence for the Possible Existence of Other Stereoisomers 



Mention has already been made of the 3-cis vitamin A aldehyde which 

 can easily be converted to the 3-as vitamin A. Hubbard and Wald have 

 reported recently '^^ that natural vitamin A concentrates can be converted 

 in vitro to rhodopsin in the presence of opsin and two enzymes, liver alcohol 



'66 p. L. Harris, S. R. Ames, and J. H. Brinkman, J. Am. Chem. Soc. 73, 1252 (1951). 

 '6' VV. Graham, D. A. van Dorp, and J. F. Arens, Rec. trav. chim. 68, 609 (1949). 



