November 19Gf
HIGHLY P U R I F I E D ~a-AhDIIOS'I'ASE DERIVATIVES
hydroxyandrost-4en-3-one followed by Wolff-Kishner reduction, as described by Brutcher and Bauer.7 Attempts to effect complete purification via the lip-acetate, nip 70" (lit.8 mp iO-i2"), aiid the lip-tol~ieiie-p-sulfoiiate,mp 138-140" (lit.0 nip 139-140') were only partly successful. Treatment of t,he lip-tosylate with LiAlH4 in refluxing ether or dioxane for 12 hr regenerated the alcohol in 60% yield only. 5a-Androstan-17-one.-Jones' oxidation of 5a-aiidrostan-lipol with chromic acid' afforded 5a-androstan-li-one in S5c;b yield. Rigorous purification by column chromatography on alrirnina and recrystallization from methanol gave the pure ketone as feathery plates, mp 124.8-125' (lit.l0 mp 119-121"), 1738 cni.-' Purified 5a-Androstan-17p-ol.-A solution of pure sa-aiidrostan-1T-one in methanol was reduced with S a B H 4 t,o give the 1i'p-alcohol (82 yield). Repeated layer chromatographic pririficatioii n i t h benzene-diethyl ether ( 5 : 1) as the developing solvent yielded pure 5a-androstaa-lip-o1, mp 166.8-168.0' (lit." mp 165.5-1663'). Glpc on SE-30 and QF-1 columns showed the steroid impurity level to be 0.0005c,,. Purified 5~-Androstane.-Reductiori of pure 5a-aiidrostan17-oiie with 855; hydrazine hydrate and KOH in ethylene glycol6 gave a 95 .& yield of 5a-aiidrostatie. Column chromatography 011 alrtmina followed by several recrystallizations from methanol afforded the pure hydrocarb(iii as wafer thin plates, mp 48.248.9' (lit.6 mp 5G-51'). Glpc on SE-30 and QF-1 columris showed the sterciid impurity level to be less than 0 . 0 0 0 0 2 ~ . Biological.-The highly purified samples of sa-androstan-lip01 and sa-androstane were assayed by inunction to the l-day-old \\'bite Leghorn chick's comb. The details of the method have been published.l2 Briefly, the method consists in the application of the standard, testcisterorie, and the test compounds in 0.05 ml of abs ,lute alcohol once daily for i days. Twenty-four hours after the last applioatioii the chicks were sacrificed, and comb weights and body weights were deterniiiied. The results are expressed as a comb ratio defined as the milligrams of comb per gram of body weight. Table I lists the results of three experiments indicating the chick comb response to 5a-androstane, 5a-androstan-lip-ol, and testmterone applied directly to the comb. Statistically significant increases i n comb ratio were found for total doses of 90, 270, 600, 1500, 4500, and 7000 pg of &-androstane. The two lowest doses of 10 and 30 pg were inactive. -411 total doses of 5~androstan-lip-olfrom 10 to 7000 pg were highly effective in increasing the comb ratios.
!):
