Steroids. Stereochemistry of D-Homoannulation

droxy-17a/3-methyl-D-homoandrostan-17-one (Xa), whichcould be equilibrated with alkali to a mixture containing ca. 70% of the Cna^-CHs (equatorial CHg...
1 downloads 0 Views 841KB Size
644

FAVSTO RAMIRE7 [CONTRIBUTION FROM THE

Val. 'i$

.\NU STANI.EY ST.\FIE J

DEPARTMENT OF CHEMISTRY OF

COLUMBIA U N I D R S I T Y ]

The Nitrous Acid Deamination of 17P-Hydroxy-20-amino-C,, Steroids. Stereochemistry of D-Homoannulation. I1 B Y FAUSTO RAhlIREZ

AND STANLEY STAFIEJ'

RECEIVED JUNE 22, 1955 The action of nitrous acid on 3p,17p-dihydroxy-20~-amino-17-isoallopregnane (IXa) yielded, stereospecifically, 3b-hydroxy-17ap-methyl-D-homoandrostan-17-one (Xa), which could be equilibrated with alkali t o a mixture containing ca. 70% of the Cl?,p-CH3 (equatorial CH3) and 30% of the C17,a-CHa (axial CH3) isomers. 3~-Hydroxy-17aa-methyl-D-honioandrostan-17-one (XIa), previously obtained stereospecifically from the nitrous acid deamination of 3p,17a-dihydroxy-20aaminoallopregnane (XIIIa), was equilibrated t o the same mixture of C17&epimeric D-homosteroids. A discussion of the stereochemistry of ring-D enlargement in the steroids based on these observations is presented.

We have recently reported on the course of the nitrous acid deamination of a steroidal 20-amino alcohol having a 17a-hydroxy configuration2 and an asymmetric center a t Cz0,such as IA (Chart 11). The present study is concerned with the deamination of a 20-amino alcohol having the 17,B-hydroxy configuration, such as IIB (Chart 11). Enough information is now available to permit a more detailed interpretation of the stereochemistry of the ring-D enlargement in this type of steroids. 3/3,17/3-Dihydroxy-20-oximido17-isoallopregnane 3-monoacetate (VIIIb, Chart I) was prepared in nearly quantitative yield from 3/3,17fl-dihydroxy17-isoallopregnan-20-one 3-monoacetate (VIIb),3 a result which does not bear out the assumed general unreactivity of the Czo-carbonyl in Cl,,-oriented side chains in steroid^.^ 3@,17P-Dihydroxy-17isoallopregnan-20-one diacetute (VIIc),3,6*6 on the other hand, failed to yield an oxime even under rather drastic condition^.^ Catalytic hydrogenation of the oxime 3-monoacetate VIIIb in acetic acid solution in the presence of platinum oxide, followed by mild hydrolysis with potassium carbonate, afforded in 54% yield a stereochemically pure amine diol regarded as 3/3,17/3-dihydroxy-20/3amino-17-isoallopregnane (IXa) . The formation of a C.o-epimer of IXa in the reduction could not be demonstrated, although the results of the nitrosation experiments described below suggest that if (1) From part of t h e Ph.D. Thesis of S. Stafiej. D u P o n t Postgraduate Teaching Fellow, 1954-1955. (2) F . Ramirez a n d S. Stafiej, TRISJOURNAL, 7 7 , 134 (1955). I n this reference, t h e 17-keto-17a-methyl-D-homosteroidobtained in t h e rearrangement of 3~,17a-dihydroxy-20a-aminoallopregnane( X I I I a , Chart I) was tentatively assigned a 17ab-methyl configuration (Clr.-CHs equatorial). It will now be shown t h a t this inference was incorrect and t h a t compounds X V I I a , X V I I b and X V I I I of ref. 2 have a 17aa-methyl configuration (Clr.-CH, axial), a s in X I a , XIb and X I V of Chart I. (3) (a) R. B. Turner, THISJ O U R N A L , 75, 3484 (1953); (b) C. W. Shoppee and D. A . Prins, Helv. C h i m . Acta, 26, 185 (1943). (4) (a) L. F. Fieser and RI. Fieser, Exprrientia, 4 , 285 (1948); (b) L. F. Fieser, ibid., 6, 312 (1950). T h e oxime of As-17-isoallopregnenS 3B,17b-diol-ZO-one has been reported (H. E. Stavely, T R ~JOURNAL 62, 489 (1940)). ( 5 ) (a) R. B. Turner, ibid., 75, 3489 (1953); (b) L. Ruzicka, K. Gatzi a n d T. Reichstein, Helv. C h i m . A c t a , 22, 626 (1930). (6) (a) I. Salamon a n d T. Reichstein, i b i d . , 3 0 , 1616 (1947); (bf A . H . Soloway. W. J . Considine, D. K. Fukushima a n d T. F. Gallagher, THISJOURNAL, 76, 2941 (1954). (7) T h e 176-hydroxy-20-keto steroids are available by t h e hydration of 17~-hydroxy-17a-ethynyl steroids utilizing mercuric salts (ref. 3 and H. E. Stavely, THIS JOURNAL, 63, 3127 (1941)). T h e 178acetoxy-20-keto steroids a r e available from 178-acetoxy-17a-ethynyl steroids b y more convenient procedures (ref. 6). T h e acetoxy-ketones lead t o D-homosteroids under hydrolytic conditions and t h u s could not be vtillzed in t h e present invrstigation.

the Czoa-amine (cf. IIA, Chart 11) is produced, the amount must be quite small. The configuration8 assigned to the amine diol IXa a t Czo is based on the results of the nitrosation experiments, interpreted in the light of a stereospecific mode of rearrangement of IXa, as discussed in the next section of this paper. The action of nitrous acid on 3/3,17P-dihydroxy20P-amino-17-isoa11opregnane (IXa) gave in 68% yield a C21H8402 keto alcohol to which structure and configuration Xa, ;.e., 3P-hydroxy-17aB-methyl-Dhomoandrostan-17-one, has been assigned. No other carbonyl compound was detected in the product of the nitrosation, which was explored also via a 2,4-dinitrophenylhydrazonederivative. That the keto alcohol obtained had a reactive methylene position was demonstrated by formation of 36hydroxy - 17a@- methyl - 16-hydroxymethylene - Dhomoandrostan-17-one (XV). That the methyl group a t had the more stable conjguration, that is the equatorial Cl,,p-methyl conformation, was established by equilibration experiments in alkaline medium. The keto alcohol X a gave on alkaline treatment a mixture of CITa-epimers, consisting of approximately 70% of 17,p-methyl isomer (Xa) and 30% of 17aa-methyl isomer (XIa). These figures are based on the specific rotations of the pure isomers and of the equilibrium mixture and were confirmed by separation of the isomers by fractional crystallization. 3B-Hydroxy- 17aa-methyl-D -homoandrostan-17-one (XIa), available2 also from the nitrosation of 3/3,17a-dihydroxy-20~~-aminoallopregnane (XIIIa), could likewise be converted to the same mixture of CI,,-epimers (Xa and XIa) on alkaline treatment. Comparisons among the steroisomeric series were also carried out with the 2,4-dinitrophenylhydrazones,the acetates X b and XIb and the diones XI1 and XIV. TH;

xv

yJ0 /I H

%HOH

Table I summarizes the results of an examination of the infrared spectra of the D-homosteroids kindly performed by Miss F. Herling of the Sloan-Ketter(8) An independent assignment of configuration a t Clo among 17ahydroxy-20-amino steroids (IA, IB, C h a r t 11) based on t h e known configurations of 17a-hydroxy-20-hydroxy steroids, is possible (ref. 2). For reasons discussed below, such a n assignment is not possible in t h e l7g-hydroxy series (IIA, IIB,Chart 11).

STEREOCHEMISTRY OF D-HOMOANNULATION

Feb. 5, 1956

CHARTI D N P = 2,4-dinitrophenylhydrazone;constants reported: ~ in: ethanol, chloroform, m.p. (Kofler); [ a I z 5taken e acetone, pyridine. (I

CHB

CHI

I c=o

I

C=NOH

Pi?--VIIa, R = R ' = H VIIb, R = Ac, R' = PI, 185' VIIC, R = R' = 4 c , 226'

VIIIa, R = H VIIIb, R = Ac, 253O, -15Od

CHI

I

R'NH-C-H

RO/ Xa, R = H, 210", -EAob S b , R = Ac, 162", -4joC, -52" Xa, DXP, 237"

I

XIa, R = H, 222O, -227'6 XIb, R = Ac, 1,27", -49"b XIa, DNP, 212

I

IXa, R = R' = 221", I X b , R = R' = 260°,

H, -12OG Ac, -290b

drogenation (followed by CrOBoxidation) of 30-hydroxy-l7a?-methyl-A5-D - homoandrosten - 17 - one acetate ("neopregnenolone acetate"), Ruzicka and Meldahlg have obtained a 3~-hydroxy-l7a!-methylD-homoandrostan-17-one acetate of m.p. 174-175' (no rotation given). This substance is probably identical with Xb (Chart I) and thus had a C17apCH3 configuration. Since "neopregnenolone acetate" was in turn made from what is now known to be 30-hydroxy-l7afl-methyl-17acy-hydr0~y-A~D-homoandrosten-17-one 3-monoacetatelo by a sequence of reactions involving replacement of the CIWOH by bromine and removal of the bromine with zinc dust in acetic acid, i t seems that the over-all change involved retention of the Cua-COnfiguration. The alkaline hydrolysis of the 174175' acetate was reported9 to yield a 30-hydroxy17a?-methyl-D-homoandrostan-l7-one, m.p. 222224' (no rotation given), which has m.p. closer to that of the less stable isomer X I a than to the more stable isomer Xa. However, in the absence of rotation data a final conclusion is unwarranted. The 17a?-methyl-D-homoandrostan-3,17-dione obtained by oxidation of the 222-224' hydroxy ketone was saidg to melt a t 195-198O (no rotation given), which does not distinguish between XI1 and XIV. A hydroxymethylene derivative presumably identical with XV was also d e ~ c r i b e d . ~ I n the course of the present work, the recently reported conversion of 17c~-aminomethyl-A~-androstene-30,17fl-diolto A6-D-homoandrosten-30-01-17aone (ii -+ iv) l1 was repeated, with results which are similar to those obtained by the Swiss group.

Discussion

XII, 196', -350b

I

------7 CHI

I

11-C-XIIR'

XIIIa, R = R' = H, 197"; -20Db XIIIb, R = R ' = Ac, 291 , -30°b

645

XIV, 19S0, -6"b

ing Institute for Cancer Research. These data are fully consistent with the structures shown. Several correlations with previously reported coinpounds can be made. From the catalytic hy-

The course of the nitrous acid deamination of amino alcohols seems quite susceptible to steric strain in the transition state, which is assumed to be of a bridged-type.12 The reaction thus appears to be particularly suitable for an exploration of the steric factors operating in the environment of the (217- and Cno-positions of the steroid m o l e c ~ l e . ~ The following stereochemical analysis of the possible course of the rearrangement of 17-OH, 20-NH2 C21-steroids epimeric at both C17-and Cz0-positions accommodates the observations made during the present study. Reference is also made to the situation encountered among 17-OH, 20-NH2 C ~ O steroids1'r13having no asymmetry a t C20. The results of the analysis are summarized in Table I1 with reference to Chart I1 and to Figs. la, 2a, 1 and 2.14 For example, it can be seen from Table I1 that of the two possible modes of rear(9) L. Ruzicka and E. F. Meldahl, Hclv. Chivz. A c t a , 23,364 (1940). (10) (a) L. Ruzicka and H. F. Meldahl, i b i d . . 21, 1760 (1938); (b) 22, 421 (1939); (c) R . J. Cremlyn, D. L. Garmaise and C. W. Shoppee, J . Chem. Soc., 1847 (1953). (11) H. Heusser, P. T h . Herzig, A . Furst and P1. Plattner, Hclr. Chim. Acto, 83, 1093 (1950). (12) (a) D. Y . Curtin and M . C Crew, THIS J O U R N A L . 77, 354 (1955); (b) D. Y . Curtin and S. Schmuckler, ibid., 7 7 , 1105 (1955). and previous references. (13) (a) M. W. Goldberg, J . Sice, H. Robert and PI. P L t t n e r , H e l v . Chim. A c t a , 30, 1441 (1947); (b) M . W. Goldberg and E, Wydler, i b i d . , 26, 1142 (1943); (c) M . W. Goldberg and R . Monnier, i b r d . . 23, 376 (1940); (d) N. L Wendler, D. T a u b and H. L. Slates, TEIS JOURNAL, 77, 3559 (1955). (14) Only carbonyl products are considered i n t h e stereochemical analysis of t h e deaminations. For conventions and nomenclature see ref. 4.

TABLE 1 ISFRARED SPECTR-4 O F D-HOMUSTEROIDS Bands due t o unperturbed cyclic methylene groups and bands due to C-H bending of acetate-methyl are not included c=o fa?) CHz (CClr) Frequency, cm

li-keto

-1

C H I (CCW

ClS

1713 ( 5 . 8 4 /A) 1420 (7.04 p ) 1388 (7.21 p ) 1713 1424 1388 1 4 2 0 ~(Ix-ouI) 1388 1717-171 t' 1715 1427 1389 3p-Hydroxy-17a~-methyl-D-honiodndrostan-l7-one (Xd)" 1426 13ss 1714 38 Acetoxp-17~B-1nethpl-D-hoxn~~rid~ost~~i-17-0ne (Xb)e 1717 1.422-141 S d 1389 l~a~-~Iethyl-D-hornoa1i~lro~tdrl-3,17-d~01ie (XII) 17 14- 17i 3 142i-1422 1387-1386 17-Keto-17aa-methyl-17ap-hydroxy-D-hutno' 1722 1'122-1421 1390 17-Keto-17a~-rnethyl-l7~a-hydroxy-D-h~1no' 1G9i" ;3.S(3 /A) , ., ,, 7 379- 1378 17a-Keto-17~-methy1-170c-hydroxy-D-honiof 0-H stretching a t 3625 cm-1 (CS2). Acetate: C-=O stretching a t 1735 cm.-I ( C S 1, C-0 stretching a t 1243 C I I I . - ~ (CS2). 3-Keto and 17-keto. ii CI6,Czand CCmethylenes. e Acetate: C-0 stretciii ig a t 1727-1733 an.-' (bro:id), C-0 stretching a t 1243 cm.-1 (CS2). f Reference compounds. 0 1ia-Keto. 3~-Hydroxy-17aa-methyl-D-homoandrostan-lione (Si2 I' 3p-Acetoxy-17acu-methy1-D-honioa1idrostdn-17-one (XIli)h 17aa-i'vIethyl-D-homoa1idrostan-3,17-dione (XIY)

,

TABLE 11 STEREOCHEMISTRY O F D-HOMOANSUI,ATIOS OF h I S O ~ l . C O I l O L S Atnirio alcohol Groups in Chart I1 Bond Transition state b opposition

1) I l i ~ m < ~ t e r i i i d

Chart I1

Ollill.

I-I~ IA:L Fig. la, 11 = CH3 OII-CH3; Cit;-Ii 17-Kcto, ~ ~ : L c u - C(11:)' Fig. 2a, R = c& C13-CHa; OH-€€ li:t-Keto, 17,3-CH, (VI)' OH-H; C16-CHn 17-I