54 VITAMINS A AND CAROTENES 



in accordance with the methods outlined in Fig. 6 failed to give satisfactory 

 yields of the desired intermediate products, owing perhaps to the inherent 

 limitations of these methods as discussed elsewhere. The most successful 

 method utilized the C2o-vitamin Ai acid methyl ester (LVI, m.p. 55 to 56°) 

 as the starting material. This was allowed to react at 0° in chloroform with 

 N-bromosuccinimide, thereby replacing one of the hydrogen atoms of car- 

 bon 3' with bromine. Dehydrobromination of the unstable bromo com- 

 pound was effected by heating with 4-phenylmorpholine. After separation 

 and chromatography, a 25% over-all yield of the crystalline C2o-vitamin 

 A 2 acid methyl ester (m.p. 45 to 47°) was obtained. This was readily con- 

 verted to vitamin A2 with lithium aluminum hydride. Owing to the in- 

 stability of vitamin A2, it was necessary to add small amounts of cc-to- 

 copherol to the final crude product before it was isolated in the pure form. 



A prehminary announcement was made recently^ ^^ of an alternative syn- 

 thesis of vitamin A2 based on the condensation of 2,6,6-trimethyl-l-ethy- 

 nylcyclohexen-2-ol-l (CII) with the ketone LXXVII via the Grignard re- 

 action and following the sequence of reactions shown in Fig. 8. 



Vitamin A2 undergoes the same reactions as vitamin Ai, so that the 

 derivatives shown in Table XIV have been prepared by following analogous 

 procedures. Like vitamin Ai, vitamin A2 is expected to exist in four stereo- 

 isomeric forms, but no work has been published along these lines. Vitamin 

 A2 has not yet been crystallized and may therefore be a mixture of stereo- 

 isomeric forms. 



8. ReHYDRO VITAMIN A^^^ 



When vitamin A-deficient rats were fed with crystalline anhydro^dtamin 

 A, a growth-promoting vitamin was isolated from their livers which was 

 neither vitamin A nor anhydrovitamin A. The ultraviolet spectrum of this 

 vitamin had three maxima at 330, 351, and 369 mpL, respectively, and the 

 antimony trichloride color had a maximum at 612 mju. Its biopotency was 

 about one-fifteenth the potency of vitamin Ai. Shantz assigned the tenta- 

 tive structure CIII to this vitamin. Since no degradative studies were 

 done on this vitamin, it is highly doubtful whether it has structure CIII 



H3C CH3 ~„ „„ 



I ' I ' 



|=CH— CH=C— CH=CH— CH=C— CH2— CH2OH 



— CH3 



CIII Rehydrovitamin A(?) 



1'^ N. A. Milas, P. Davis, and D. D. Grassetti, Paper presented before the 2nd Intern. 

 Congr. Biochem., Paris. July, 1952. 



