16-Substituted steroids. XV. A new synthetic method for 16-ketosteroids.

In 1941 (1) the senior author entered the field of 16-substituted steroids with the view of preparing all of the three possible 16,17-epimers of natural estriol. Since that time two of these three epimers, 16-epiestriol (isoestriol-A) (2) and 17-epiestriol (3), have become available through our efforts and those of the Swiss investigators. In 1952 the present authors made an attempt to prepare the missing 1,3,5(10)-estratrien-3,16p, 17~~ trio1 by epimerization of 3-methoxy-1,3,5(10)-estratrien-16P, 17/I-diol-17 p-toluenesulfonate. However, much to our surprise, alkaline saponification of the latter compound resulted in actual cleavage of p-toluenesulfonic acid from the molecule, yielding the 16-ketosteroid in very satisfying yield. This new synthetic approach to 16-ketosteroids is far superior to our original method (4) which involved the Clemmensen reduction of the 16-keto17/3-hydroxysteroid. The Clemmensen reduction always, rather strangely, yields a mixture of 16-ketosteroid and 17-ketosteroid, and often this mixture of 16and 17-ketones is intractable. Furthermore, the Clemmensen method is too drastic for some unsaturated compounds in the steroid series. The present synthesis has wide applicability. We describe here its use in the estrogen series, giving our original experiments along with later much improved procedures. For instance, it is now possible to start with estrone and operate through the seven synthetic steps of benzoylation, nitrosation, saponification, zinc-acetic acid reduction (5), ditosylation, sodium borohydride reduction, and cleavage and to obtain pure estrone-16 in 54 per cent over-all yield. In keeping with our other structural assignments at Cl6 (6), we have designated the carbinol formed on the sodium borohydride reduction of 16ketoestradiol-3,17-di-p-toluenesulfonate as possessing the p configuration.