Dehydration o f Bile Acids and Their Derivatives . III . Dehydration o f Hyodesoxycholic Acid . * A New Pathway o f the Bile Acid to Progesterone

Chenodesoxycholic and ursodesoxycholic acids are a-3, a-7and a-3, /-7-dihydroxycholanic acids, respectively' and so they are C7epimers. In the previous papers`' we have reported that both epimeric acids underwent dehydration by means of concentrated hydrochloric acid as well as of metal chlorides, and that chenodesoxycholic acid, because of trans elimination, was dehydrated much more easily than its epimeric acid, yielding a mixture of 3-hydroxycholenic acids 24. 4~. Based upon these experimental results, we (1944) reached a conclusion that the 9-hydrogen atom must be trans to the methyl group at C103~; and indeed later we found in Fieser's book 5, that the same conclusion was drawn by Shoppee (1946) under the quite analogous considerations, which support the X-ray work of Bernal (1940). Hyodesoxycholic acid was also subjected to such a dehydration procedure and found to remain almost unchanged 3, even though it has a hydroxyl group (C6) adjacent to a tertiary hydrogen atom (C5), just as chenodesoxycholic acid has. We accordingly were inclined to assume that the 6-hydroxyl group of the bile acid may be cis in reference to the 5-hydrogen atom. Recently Moffett and Hoehn6~, however, have proved that the reverse is true. On the way of the sequence of dehydration experiments of bile acids, we have found phosphorus oxychloride*** to be the only tool for dehydration of hyodesoxycholic acid among the reagents so far examined. The methylester of hyodesoxycholic acid (I) was heated with this reagent in pyridine and afforded an unsaturated compound (III) in a satisfactorily good yield, being simultaneously halogenated. By dehalogenation, followed by hydrolysis, it furnished quantitatively ,Q-3-hydroxycholenic acid (IV), melting at 235-236°, and a mixed