The Stereochemical Course of The Catalytic Hydrogenation of Ring a

Mechanism of estrogenbiosynthesis. III. Stereochemistry of aromatization of C19 and C18 steroids. John D. Townsley and Harry J. Brodie. Biochemistry 1...
0 downloads 0 Views 259KB Size
3766

COMMUNICATIONS TO THE EDITOR

Vol. 84

The more likely pathway for C-1 hydrogen elimiThe proof for the distribution of tritium in the nation would involve either displacing it by an two reduced products is outlined in the accompanyactivated group “Y” (for example hydroxyl) ing flow sheet. Before determining the distribution which can later be eliminateds (pathway 11) or of tritium a t C-1, the tritium a t C-2 and other eliminating i t as a hydride ion concerted with the labile positions was removed by refluxing with KOH removal of the C-19 oxygenated group (pathway in aqueous methanol. Experiments have shown 111). This latter pathway could take place either that the percentage of tritium lost from these posithrough the mechanism shownl0 or by a four- tions ranges from 40 to 600j0.* For clarity, tritium center type (cyclic) elimination involving an is shown a t the positions of highest concentration in activated C-19 oxonium ion intermediate (not the reduced compounds 111, IV, and V. All reshown.) From present knowledge, these types of ductions were carried out with tritium gas over 5yo reactions could reasonably be expected t o involve palladium-charcoal in dioxane solvent. The strucT P N H and oxygen and the ID-hydrogen. Experi- tures of all compounds were proven from mixture ments are now in progress to define the more likely melting points and by comparing their infrared pathway.ll spectra with authentic samples. The radiochenlical Acknowledgment.-We wish to thank Drs. H. J. purity was established by purifying to constant Ringold and M. Gut for their active interest in specific activity using paper chromatography and recrystallization. various aspects of this problem. Androst-l-ene-3,17-dione(I) was reduced and (9) The group “Y” also could be eliminated in concert with the C-10 then refluxed with potassium hydroxide in aqueous group to form a C - l ( l 0 ) double bond. methanol to give compound 111. Compound I11 (10) This is similar to a mechanism previously proposed (M. then was treated with dichlorodicyanoquinone Hayano, H. J. Ringold, V. Stefanovic, &Gut I. and R. I. Dorfman, (DDQ,6 and incubated with Bacillus sphericus Biochem. Bioghys. Reseavch Commun., 4, 454 (1961)) except t h a t i t involves the 16-hydrogen. (ATCC 7055)6 to give compound IIIA and IIIP, (11) This work was supported in part by U. S. Public Health with losses of 89 and %yoof the tritium, respecService grants A-2672, FF-258, and training grant CRTY-5001. This showed that the reduction and subtively. (12) Trainee in the Training Program for Steroid Biochemistry. sequent dehydrogenation were quite stereospecific (13) Post-doctorate Trainee in the Training Program for Steroid Biochemistry. for addition and removal of tritium a t C1. That the TOMAS MORATO removal of tritium was from the alpha position was WORCESTER FOUNDATION FOR KARLRAAB’* indicated from the literature on these reactions606 EXPERIMENTAL BIOLOGY HARRY J. B R O D I E ~and ~ was proven conclusively in a recent study by SHREWSBURY, MASS. MIKAHAYANO RALPHI. DORFMAN Ringold, Gut, Hayano and Turner,’ who showed the l a that with 5a-androstane-3,17-dione-la-H2 1, 1962 RECEIVED AUGUST hydrogen (deuterium) is lost exclusively. They also showed that the reduction of the androst-l-eneTHE STEREOCHEMICAL COURSE OF THE 3,17-dione with deuterium proceeds almost exCATALYTIC HYDROGENATION OF RING A clusively alpha a t C-1. UNSATURATED STEROIDS To determine whether the enzymatic and DDQ Sir: reactions were as specific for the elimination of the In the reactions of double bonds in ring A of la-hydrogen with a C-4 double bond present steroids, attack usually occurs predominantly on I11 was oxidized to compound IV using the standthe alpha face since this is often less hindered due to ard bromination-dehydrobromination technique.s the presence of the 109-methyl group.’ Indeed, Compound IV then was oxidized to compound some workers have made “alpha face attack” a IVA with DDQ. Here 76y0 of the tritium was general rule for, in a recent publication,2 it was as- lost, indicating that some of the 10-hydrogen was sumed, based on the type of work cited in reference eliminated with the C-4 double bond present. 1, that in the reduction of steroids with tritium Dehydrogenation with B. sphericus gave the same on palladium catalyst, the label goes mostly into the result as with the saturated compound, i.e., 10s.; of 93% of the tritium (IVB), showing that in this 1a and 2a positions. For the study of the inechariisrn of estrogen bio- reaction the la-hydrogen is lost preferentially bynthesis, we wished to prepare a C-1 tritiated with the A4-compound to the same extent as with androst--l-ene-3,1’i-dione of known configuration. 5a-androstane-3,17-dione. ’To do this 17P-hydroxy-androsta-1,4-dien-3-one Finally, 17p - hydroxy - androsta - 1,4 - dien - 3was reduced with tritium gas over palladium.a We one (11) was reduced catalytically to give a mixture can report now that under these conditions the of product^.^ The A4-compound was isolated attack a t C-1 is predominantly befa in a ratio of (4) M. Gut and M. Hayano in “Advances in Tracer Methodology,” about 3: 1. In contrast, however, when androst-l- Vol. I, S. Rothchild, ed., Plenum Press, New York 11, N. Y., in press. ene-3,17-dione is reduced in the same manner the Osinski and Vanderhaeghe have r e p r t e d (ref. 3) t h a t after the reducintroduction a t C-1 is over 90% alpha as expected. tion of the AI bond of the bismethylene dioxide of prednisone no loss (1) For examples of alpha attack during reduction epoxidation and glycol formation, see L. P. Fieser and Mary Fieser, “Steroids,” Reinhold Publishing Corp., New York, N. Y.. 1969, pp 254, 266, 271-274. The less common beta attack has been shown in hydrogenations of A‘ and A’ compounds where a ring junction is involved (pp. 272, 274). (2) L. R. Axelrod and J. W. Goodzieher, J . Clrn. Endocrinol. and Melabolism. 22, 537 (1962). (3) Compare P. Osinski aud H. Vanderhaeghe, Rec. haw. chrm., 79, 216 (1960).

of tritium was noted on treatment with potassium hydroxide. AS can be seen, this is contrary to results obtained with C-l,a-tritiated testosterone and androstenedione. ( 6 ) H. J. Ringold and A. Turner, Chcm. and Ind., 211 (1962). (6) M. Hayano, H. J. Ringold, V. Stefanovic, M. Gut and R. I. Dorfman, Biochem. Biophys. Res. C o m m . , 4, 454 (1961). (7) H. J. Ringold, M. Gut, M. Hayano and A. Turner, Tcfrahedron Lelfers, in press. ( 8 ) G. Rosenkranz, 0.Mancera, V. Gntica and C. Djerassi, J. A m . Chcm. SOC..72, 4077 (1050).

3767

COMMUNICATIONS TO THE EDITOR

Oct. 5 , 1962 0

8

0

DDQ(II1a) B. sphaericus(I1Ib)

H," I'd

O

'

H

t

1113 - 0 11 1, -0 07

111 - - L O O *

2, t i n 1 in acetone 3 , Zn i i i IIAc

& & & B. sphaericus\IVb) DDQ(1Va)

0

B. sphaericus(Vb) DDQWa) ~

0 IVa = O 24 v =1.00 b =007 Va = 0 6 8 b =075 Numbers refer to the relative specific activities of reactants and products. 0

IV = l o o

*

from the mixture by paper chromatography, oxidized to the diketone using the Jones method9 and equilibrated with base to give V. Compound V was oxidized both with DDQ to give compound VA (68y0 retention of tritium) and with B. sphaericus to give compound VB (75% retention of tritium). The remaining tritium was shown t o be a t C-1 by oxidizing another sample of compound V to androsta-1,4,6-triene-3,17-dioneusing chloranil1° and then DDQ.6 When this compound was treated with acid it rearranged to 1-methylestrone acetate which was inactive (specific activity