18%
July, 1947
COMMUNICATIONS T O T H E EDITOR THE STERIC STRUCTURE OF BSTRIOL ARD RELATED STEROIDS
described, cis-oid isoestrio1-Al0 (acetonide, m. p. 183.5-184.5', uncor.) and trans-oidg estriol are Sir: obtained. Isoestriol-A and estriol are therefore, Stodola, et ad.,'#* have shown that the zinc- respectively, A1.**C-estratriene-3, 16(a),17(a)-triol acetic acid reduction of 16-oximino-17-keto- and A1#a*6-estratriene-3, 16( 17(a)-triol. steroids followed by acetylation results in the Ruzicka and co-workersl1*l2have recently preformation of 16-keto-17-acetoxysteroids.The pared in the estrogen and androstane series 16,17structure of the intermediate a-keto1 was not de- glycols which they designated as A1-aJ-estrat.rienetermined. 3,16(a),17(a)-trioll and androstane-3 (P), 16(a), We have found that 16-keto-17-acetoxysteroids 17(a)-tri01,~~ respectively. These two triols must made by the Stodola method possess the 17-group be cis-oid a t Cl6-Cl7, since they were formed in the a-position, since As-androstene-3(j3),17- through the osmium tetroxide addition to the diol-16-0ne &acetate* so prepared readil forms a proper A%teroid. However, as neither of the 16-diethyl thioacetala (m. p. 136.5-138 , uncor.) two triols is identical with isoestriol-A or with the which, upon hydrogenolysis,' yields the known saturated androstene-triol (m. p. 251-2S0, unAK-androstene-3(@),17(a)-diol diacetate'J identi- cor.) of Butenandt, we suggest that the two comfied by comparison with an authentic sample. pounds are in reality A1J6-estratriene-3,IS(@, Furthermore, we feel certain that no rearrange- 17(8)-triol and androstane-3 (g), IS(@,17(8)-triol. ment occurs in the acetylation of a-ketols pre(10) M. H. Huffman and H. H. Darby, TEISJOURNAL, 66, 150 pared by the Stodola method, and that these (1944). ketols are also 16-keto-l7(a)-hydroxysteroids, (11) V. Relog, L. R u d c h and P. Widand, H c b . Chim. A d o , S8, (1945). because: (1) the acetylated a-keto1 may be 255(12) L. Ruricka, V. Prelog and P. Wielnnd, W . .SO, 1608 (1945). saponified to regenerate the original a-keto1 from DEPARTKBNT OF BIOCHEMISTRY which it was prepared; (2) the same acetylated SOUTHWESTERN MEDICALCOLLEGE M u N. HUFFMAS a-keto1 results on esterification by long standing DALLAS,TEXAS MARY H A RLon ~ in pyridineacetic anhydride or by almost inRECBIVEDJars 11,1947 stantaneous acetylation according to Ciusa and S o l l a ~ z o the , ~ yields being virtually identical in both instances. Hydrogenation of Ab-androstene-3(8),17(a)diol-16-one diacetate with Raney nickel in ethanol Sir: (Stodola, et aL1)or in ethyl acetate-ethanol (our In a previous communication,1 it was pointed confirmation) yields Butenandt's tri0l.l Since out that the form of the A-dc curve of nthis triol is known to be cis-oid at C ~ O - C its I ~ , ~octadecylpyridonium chloride in water is similar structure must be AE-androstene3(8),16(a),17- to that of the paraflinchah electrolytes in general (a)-triol. If A6-androstene-3(B),17(a)diol-ls- (i. e., A decreases abruptly a t the critical concenone (Stodola prepared, m. p. 200-202', uncor.) tration), but that in water-methanol mixtures is reduced with sodium amalgam in excess dilute containing 10 to 35% of methanol (by wt.), A ethanolic acetic acid (40'), both Butenandt's passes through a maximum a t concentrations in triol (cis-oid) and Hirschmann's triol8 (trans- the neighborhood of the critical concentration in oid) may be obtained. Hirschmann's triol (trans- water. Maxima in the A-dc curves in suitoid at c16-c179) is therefore A5-androstene-3(@), able water-methanol mixtures, but not in water, 16( 17(a)-triol. have also been found for the following n-octaIf, in the estrogen series, A1*a~6-estratriene-3,17decyltrimethylammonium salts: chloride, bro(a)-diol-I&one (Stodola prepared, m. p. 234- mate and formate. On the other hand, the corre237', uncor.) is reduced with sodium amalgam as sponding nitrate, bromide and oxalate exhibit ( 1 ) F. H . Stodola, E. C. Kenddl and B. F. McKenrie. J . Orx. the usual "break point" phenomenon in mixtures Chrm., 8 , 841 (1941). containing 0 to 38% of methanol. (2) F. H. Stodola and E. C. Kenddl, ibid.. T, 338 (1942). We now find that the existence of a maximum in (3) S. Bernstein and L. Dorfmnn, T u n JOURNAL, a,1152 (1946). (4) L. Ruricka and A. Wettstdn, Hrlo. Chim. Acfo, 18, 1264 A is not limited to water-methanol mixtures as is (1935). clearly shown in Fig. 1, where A has been plotted (5) A. Butroandt and G. Hanisch. 2. physroi. Chrm.. SST, 89 as a function of 4 for n-hexadesylpyridonium io(1936). date (curve 1) and n-octadecylpyridonium iodate (8) R . Ciusa nod 0. .%llasro. AIR. cham. opplicolo. SS, 72 (1943). (curve 2) in water at 25'. In the first mentioned (7) A. Buteoandt. J . Schmidt-ThomC and T . Weisr. Bn.. TS, 417 case, the peak is approximately 1 A unit high, and (1939). in the second, approximately 10 A Units. ( 8 ) H.Hinrhnuoo. J . Biol. C h . ,160,863 (1948).
a),
B
a),
(9) M . N. Huffmon and M . H. Lott. ibid.. lU,786 (1946).
(1) B v m . Criegrr and h u m , Tmm J o m n l ~m, . 1187 (1040).
