METABOLISM OF CAROTENOIDS AND VITAMINS A 547 



that vitamin A2 administered to rats appeared unchanged in the livers, 

 blood, and retinae of the animals, although the vitamin Ai which had been 

 present in the blood was still tenaciously held. Vitamin A2 was found to 

 have the capacity to assume all the functions in the rat usually mediated 

 by vitamin Ai. Thus, rats receiving only vitamin A2 grew normally, lived 

 longer, and their xerophthalmia Avas cured. Their coats appeared glossy, 

 and the rats became lively, ate well, and were in excellent condition. 

 Females receiving their vitamin A only in the form of vitamin Ao Avere 

 mated, and one-half became pregnant and produced litters. However, 

 only one female was able to raise her pup. In general, male rats seemed to 

 respond to vitamin A2 better than did female rats, as far as growth was 

 concerned. 



Morcos and Salah^^- observed that when vitamin Ai was given to the 

 Nile fishes, the air-breathing edible African catfish {Clarias lazera), and the 

 bolti, a Nile food-fish (Tilapia nilotica), in which vitamin A2 is the only 

 form ordinarily stored, administered vitamin Ai was present in the liver in 

 unchanged foim. It is apparent that the two forms of vitamin A are not 

 readily in tercon verted in the animal. 



The precursors of vitamin A2 are still unknown. Collins and associates'^*^ 

 conclude that vitamin A2 may result from preformed vitamin A2 in the 

 diet, from the conversion of 18-carotene or of other provitamins A to vita- 

 min Ao, or from a specific provitamin A2. This might be an animal carote- 

 noid such as astaxanthin. However, there is not direct evidence sup- 

 porting the last suggestion. Presumably vitamin A2 may exist in stereo- 

 isomeric forms comparable to those of vitamin Ai. These have not as yet 

 been prepared. 



b. Cis-Trans Forms of Vitamin A and Their Biopotencies. Vitamin Ai 

 can exist in four isomeric forms. Of the five double bonds in the vitamin A 

 molecule, only the four present in the side-chain can contribute to stereo- 

 isomerism. According to the theoretical considerations of Pauling,"*^ and 

 of Zechmeister,^-^ '^^- steric hindrance should prevent double bonds 2 and 

 4 from assuming a cis position; this would leave only bonds 3 and 5 free 

 to change from the trans to the cis configuration. On this basis, the folloA^dng 

 four vitamin A stereoisomers w^ould be possible :^^ (1) all-irans- Vitamin A, 

 (2) 3-cis- vitamin A, (3) 5-c^s- vitamin A, (4) 3, 5-di-czs- vitamin A. For the 

 formulas of these compounds the reader is referred to The Lipids, Vol. I, 

 p. 696. 



In addition to all-/ra?is- vitamin A, neovitamin A {3,3,4-tTi-trans-5-cis 



'32 S. R. Morcos and M. K. Salah, Nature, 167, 117-118 (1951). 

 "3 L. Pauling, Fortschr. Chem. org. Naturstoffe, 3, 203-235 (1939). 



