in absoliik EtOH using IT4 Rariey nickel" ratalyst. The catalyst was removed by filtration and the EtOH was evaporated at reduced pressure. The residue was recrystallized from pe- troleum ether (60-70') affording 0.5 g (427;) o f desired amino alcohol 3: mp 149-151": ir (CEICl,), 2.78, 2.96, 3.33, 8.41, 3.30, 6.25, 6.3.>, 6.70, 6.92, 7.40, '3.85, 10.66 p ; nmr (CDCl,), 6 7.70 (multiplet, aromatic ortho protons), 7.34 iriniltiplet, aromatic ireeta arid para protons), 3.00 (WI;~ = I9 cps. axial methine proton

2(a)-Phenyl-trans-decalin-2(e),3( e)-diol 3-Tosylate (12).- - T o 9 (1.05 g, 0.0043 mole), dissolved in 20 ml of a n h y d r o i ~ pyri- dine, was added p-tol~~eiiesiilfoiiyl chloride (2.0 g, 0.0 1 mole) and the solution was allou-ed to jtarid at room temperature for 48 hr. I t 0 w:w added and the reaiilting oil was scratrhed with a glaw rod to pr(~iriole crystallization. The solid was removed by filtra- t ion aiid recrystallized from petroleiim et,her (60-70°) affording 1.0 g (.i8F;) of t,osylate 12: mp '39-100°; ir (C€IC13), 2.78, 3.42, : 3 . X , 6.27, 6.70, 6.92, 7.40, 8.,X, 9.13, 0.74, 10.33, 10.82, 11.25, 1 1.W, 11.98 p ; nnir (CCl,), 8 7.2-8.0 (multiplet, aromatic pro- toris), 4.80 jqiiartet, ./$):, = 11 cps, ./:$,, = 6 i'p, axial rnethinr pr(i tc~ii at (X), 2.43 (singlet, ArCI13).

3(a)-Amino-2(a)-phenyl-trans-2(e)-decaIoi (2).--Conipoiind 12 (1.0 g, 0.0023 mole) was placed iti a steel bomb arid the bomb was c~)olctl in 1)ry Ice-.\le2C0, TI) the bomb was added cn.

a t c-3). r h d . ( C I B H ? ~ ~ O ) c, H, x.

.171ai. (('L3112SS04) C, I i .

(17) 1 1 . R. Billica a n d €1. Adkins in "Organic Syntheses," ( b l l . Vol. 111, E. C. Horning, Ed . , John Wiley and Sone, Inc. , i i ew 1-ork, S. Y., 1955, p 180.

I00 nil of liqiiid XI&. for 24 hr. d v e d in CNCI,. was evaporated. ether (60-- iOo) affording 0.80 g (30 116--1 I f " : ir (C1ICYj'), 2 . W j 2.9'7, 6.71, 6.92, 7.40, 9.8>, 10.01, 10.28, I multiplet, arcrrnatic proi.ons), :1.8:2 i 11.1 ' ? = f i cps, cqii:it(~~,ial inethineproton a t ('-3 I.

3( e)-Amino-2(e)-phenyl-trans-2(a)-decalol (4 I.--- ('~JInp~Jlill(i 15 ((1.0 g 0.004 mole) w a s dissolved it1 100 i d o f xbsoliite EfOfl satiirateti with S I I . , arid material war hydrogenated ritider T i l kg/cm2 of HB using It.4 Raney Xi c*atalyst.17 Ttic c~ntalyst w:i* removed by filtraiiciri and the solvent FWS evaporated at i.cdrlcwl preasure. The iwidiir wah chromatographed 011 si1ic:i gcl i.\Ierck 0.03--0.20 nirn) eluting with cyc.lohexaiie-l':tO.~(. [ I : 1 i affording O..? g i3O';j of h i r e d amino alrohcll 4: nip 146- 148'; it , iCHClZ), 2.78, ?.<I:, 3,i+i, 3.52, 3.31, (5.25, 6.:34, (i.70, (;.!I?, 7.6:;, 8.60, !l.t71, 9.H6, 1lI.l~0, 10.26 p : i1rnr (('lK'l:4, ('F4C'OLil I , 6 7,4li ihro:i(I singlet, :ironrii~ic, ~ ~ I I I I ) I I S ~ , 3.5: i I i ' I :- 10 (p, r n r - ~ t i i i lc .

~ 1 l ' " t e ~ r l :i1 C-31. ~ I n t r r . ~ ( ~ I ~ l I ~ ~ X O ~ ( $ , 1 1 , s. Acknowledgment. - 'l'hc autht lrs grutefullJ- :~ckrlo\vl-

edge support ( i f this project bj. the S a t i o n d Institutcs ( i f Health Grant He-08555. The authors \\.ish tci VS-

press their appreciaticiri t ( J Drs. -1. 13. LaPidus imd 1'. S . l'atil and JIr . S . 1-Ictt.y f(ir performiiig the vas defereiis :1sS:tyh.

The bomb was miled atid 1ie;ried at 120 The presslire was released and the residue was dis-

The CHCl3 solution w a ~ filtered and t,he solvent The residue was rec~ystallized from petroleiim

of amino alr~ohol 2: in 4, 3.42, S.,j1, 6.2.7, 6." 7 p ; iinir (C'I)Cl:!), 6 7

. i nn / . ((>l611&()) (1, I I , S .

-. ... . .

Synthesis and Myotrophic-Androgenic Activity of Substituted 2a.3a-Methan0-5a-androstane Derivatives'

J~ASFRED E. KOLFF, SHEUE-YANS CHESG, A N D Wxhrwox Ho

Uepuitment of l'harrrznceutical Chemistry, School o j Pharmacy, I*ncierscty of Culifornrcl, Ban Pr(z7~ccsc0, ('alcJornia 05122

Rpceiaed February 17, 1966

The p i epar ation and aiidrogeiiic-niyotropliic testing of analogs of 17 having substituents on the q clopropj-1 ring were undertaken in an effort to obtain information regarding the steric and electronic requirements in the A ring of anabolic-androgenic androstanes. Treatment of 176-hydroxyandrost-2-ene acetate with ethyl di- azoacetate in the presence of anhydrous CuS04 gave 2~~,3a-(~-carbethoxymethano)-5a-androstan-l7p-ol acetate which u a s converted to a variety of substituted cyclopropane derivatives. The most potent is the aldehyde 15 uhich i i more active than testosterone propionate in the myotrophic test and is much less androgenic.

Studies in this laboratory have resulted in the proposal2 that anabolic-androgenic androstanes are bound to their receptor by a @-face a-bond to an sp2 sjsteni in the> A ring. The pronounced anabolic androgenic activity of 2a,3Lu-meth,zrio-ja-androstati- 17p-01 (17)2 was taken : ~ b evidence for this hypothesis. Recent on steroidal cpisulfides, bioisosteric with thcsc mctliauo steroids, has shown that the L)a,3cr isomer:, indeed have high pareiiteral activit? . whereas the 2p,:$? isomers :ire es~entistlll mictive. This is ii i

harmony with our proposal.3"' 011 the other hand, L ) L ~ , ; ~ C Y - and 2@,3&steroidal difluorocyclopropanes have s i m i l a r activitj .i

To gain furthrr informatioil in this area, the prep- :u'ntion o f analogs of 17 having altered electron

(1) (a) This investigation was supported in par t by a Public Health Service

density at the $pp" centers was undertakcri. If a a bond is important, the strength of the binding would hi different in such analogs. Since @-face binding i i :tssunied to be involved, the preparation of substituted c\-clopropane aiialogs of 17 should be feasible since thc substituent groups should not interfere sterically with drug receptor binding.

The SJ nthetic plan involved the prepaixtion of 2a, 3 a-carbet hox~niethaiio-~cr-aridrostari-l7p-01 are- t x t e (7) :iq a cuninion intermediate for thc other derivatives. This material wa6 prepared 1)) ilic reactioii of 17/3-hydroxyandrost-2-eiie acetate5 (2) with ethyl tlazoacetate. Llltliough carberie inter- mediates have been proposed in the reaction of diazo compounds with olefins,6 the reaction failed when thv reagents were heated at 120-180°, or were irradiated i i i

toluene or hexaiie solution with a medium-vressurc . . research grant (AM-05016) f rom-the National Insti tute of Arthritis and h l e t a h l i c Diseases, U. 9. Puhlic Health Service. (h) Portions of this work are taken from the P h m . thesis or s.-T. Clienp, University California, San Francisco, 1966.

mercury arc. On the other hand, a 45% of 7 realized \dm the reagents B%re heated in the

These results point) to presence of anhydrous CuS04. (2) 31. E. n'olff, I\-. I I o , and R. Iiwok, J . .Med. C h e m . , 7, 577 (1964). ( 3 ) (a) P. 1). Klimstra, E. F. Kuttine, and R . E. Counsell. i b i d . , 9, tixi

(1966). (b) T h e Searle group did not interpret their da ta in this way. ( 4 ) ( a i I,. 11. Knox, E. Velarde. S. Berger, I). ('uadriello. P. R-. Laudis,

and .\. I ) . ('russ. .I. d r r r . Clirm. Soc., 85, 1881 (1963); (b) E'. 3 . iiin<,l and it . K Lhr fman , S l e r o c d ~ , 3 , 109 ( Ig64) .

l j l .\. Ilowers, A. D. Cross. ,J. 13. Edwards, €1. L'arpio, 11. Calaada, and

(6) J. Hinq "1)ivalent Carbon," The Ronald Press ( 'o. , Sen. l u r k , N. 1 ., 1:. IlPnot, J . M e d . (Them., 6, 156 (1963).

lil64, j l l2?.

July 1968 h l Y O ' l ~ R O P H I C - i ~ N ~ R o G E N I C ~ C ~ , ~ ~ : - ~ I E T H A ~ ' O - - Da-ANDROSTASES 865

the intervention of ;in organocopper intermediate rather than a carbene, but further research is needed to clarify this point.

1, It = I* 2, R = CII,

3, isomer -4 4, isomer B

5 , I t = 011; I t ' = 011 11, It = T I ; 11' = 011 6, It = 011; I t ' = 0 12, It = H; It' = 0 7, It = OCJla; It' = 0 A c 8, I t = OCOgC2Hj; R' = 0 9, It = C1HaN; R' = 0

10, IL = C ~ H ~ K T ; R' = 011

13, R = (2113; R' = 011 14, R = CH,; R' 0

18, R = OAc 19. R = 0

The configuration of the cyclopropane ring and substituent in 7 was established in the following manner. The presence of a single C-19 methyl resonance in the nmr spectrum indicated that only one isomer was at hand. Moreover, the expected position of this isomer could be calculated by use of the Zurcher values7 for substituents in &-androstane and by the Tori equation8 for calculating the anisotropic shielding effect of the cyclopropane ring.

3 cos20, - 1 A6 (ppni) = -6.67 X cm3

i = 1 Rt3 where R, is the distance in -1 between the midpoint of a C-C bond of the cyclopropane ring and a C-19 proton and 0, is the acute angle which the line R, makes with the C-C bond. An expected shielding of +0.273 ppni was calculated by the Tori equation for 2/3,3/3- iiiethatio-jcu-androstarie, whereas the corresponding effect in the ~ C Y , ~ C Y isomer was calculated to be only +0.010 p p m Experimentally, in compounds 6-8, 10, :itid 16, the effect of the cyclopropane ring mas found to differ by only -0.049 to +0.011 ppm from the Zurcher value^.^" These results are compatible only with a 2oi,:3a: fusion,

Chemical studies were in harmony with this assign- ment. Thus a modified Hunsdieclter reactiong*'O on 6 gave a mixture of the epimeric bromides 19 as shown by the nmr spectrum. The formation of two bromides is not unexpected in view of the probable free-radical mechanism of this reaction." Reduction of 19 with

( 7 ) N. S. Bhacca and D. H . Williams, ".lpplications of Nmr Spectroscopy in OrEanic Chemistry," Holden-Day, Inc., San Francisco, Calif., 1964, p 19.

( 8 ) li. T o r i arid K. Kitahonoki, J . A m . Chem. Sac., 87, 386 ( l !>G5) . ( M I J. Criatol and \V. C. Firth. J . OTY. Cliem., 26, 280 (1961). (10) J. .\. Ilavis, S. Carroll, J. Bunds, and I). Johnson, ibid., 30, 415

(11) R. H. Johnson and R. K. Ingliam, Chem. Reu., 56, 219 (1956). (196,5).

lithium tri-t-butoxyaluminohydride followed by re- ductive dehalogenation of the mixture of halides in the presence of Raney nickel gave 17 identical in all respects with an authentic sample,2 together with a single isomer of starting material. Thus i t is clear that the least hindered cyclopropane is formed in the addition reaction, and therefore we assign the p con- figuration to the carboxyl group as we11.12 The CY

position is hindered by interaction with protons a t the la: and 5a positions. It is also evident that these same factors are involved in the difference in stability of the two bromides toward catalytic reductive de- halogenation, but i t is not possible to specify which isomer reacts preferentially.

Hydrolysis of 7 gave the acid 5 which was oxidized to the ketone 6 with Jones reagent. This was sub- jected to a modified Curtius13 reaction in order to obtain the corresponding amine. The mixed anhydride 8 could be isolated and upon treatment with sodium azide smoothly gave the amine 12. Reduction at C-17 with NaBH4 gave the 17~-alcoholll .

One of the goals of this work was the formation of a cyclopropene ring fused to C-2 and C-3, in order to study the consequences of increasing the p character of the 2 and/or 3 carbon. Possible routes to the desired cyclopropenes included dehydrohalogenation of hal- ocy~lopropanes'~ and Hofmann elimination of cyclo- propylamines. 15,16 The required quaternary amine 14 mas prepared by quaternization of 12 with methyl iodide and alkali. However, no cyclopropenes were obtained from either method of preparation, in spite of very careful work-up procedures.

The reaction of the mixed anhydride 8 with ethyl- enimine gave the amide 9. Reduction with LiAlH4 gave the aldehyde 15 and the alcohol 16. The last compound was also obtained by reduction of 15 with KaBH,. The preparation of both isomers of 2,3- methano-Sa-estran-17P-ol 3 and 4 by means of the Simmons-Smith reaction is included in the Experi- mental Section.

Discussion The data from the pharmacological t e ~ t i n g ' ~ ~ ' ~ are

displayed in Table I. The most active compound is the cyclopropyl aldehyde 15, which is more active than testosterone propionate or the corresponding un- substituted cyclopropane derivative 17 in the my- otrophic test. At the same time, the compound has low androgenic action.

The next ranking compound in activity is the hy- droxymethylene derivative 16 which has more than 10% the activity of testosterone propionate in the my- otrophic test and less than 10% in the androgenic assay. Compound 16 has roughly one-tenth the activity of 15 in both tests.

(12) The usual u,P dotted-solid-line convention is employed to descrilie the cyclopropane substituent. T h e ,4 substituent is attached to the upper- most bond in the planar structural formula.

(13) J. Weinstock, J . Org. Chem., 26, 3511 (1961). (14) C. L. Osborn, T. C. Shields, B. 4 . Shoulders, J. F. Krause, H. V.

(15) N. Y. Demyanov and N. N. Doyarenko, Eer. , 66, 2200 (1923). (16) N. J. Schlatter, J . A m . Chem. Soc., 63, 1733 (1941). (17) Pharmacological teats mere performed at T h e Endocrine Labora-

(18) L. G. Hershberger, E. G. Shipley, and R. K. AIeyer, Proc. SOC. Ezp t l .

Cortezand, and P. D. Gardner. J . A m . Chem. Soc., 87, 3158 (1965).

tories, Xadison, Wis.

Bid. ]Wed., 83, 175 (1953).

S66

Cast rate coiitrol Testosterone propioiiate ( 0 . 3 ) P .i (6 .0) 1'

P 7 ( 6 . 0 )

C'astrate coiitrol Testosterone piopioiiate ( 0 . 3 ) P 10 ( 3 . 0 ) P 13 ( 3 . 0 ) I' 1.5 ( 0 . 3 ) P I(i (:;.I)) I' 1 i',

I'

CastrLLLe co11lrol Testosterone propioilate (0.3) i"' I T (0 . :<) P

1 entral prostate

14 .0 i 0 .70 :13,4 Z!Z 2 .00 <0.001 18.5 I 0 4G (I. 02 2 0 . 9 i 2 . i 0 XSr

14. 1 z 0.03 44 .7 i. 5 . 3 1 <0. 001 17.3 i- 2 64 s s 19.2 + 0.98 < O . 0 2

16.1 = 1.O.i 41 .i i 4.62 <o. 01 l2, . . i Z!Z 0.05 0 . 0 2 deciesse 1 *-, 0 Z!Z 0 89 1- ;;

1h.O i. 0.67 i34.8 i 4.06 <o. 01 16.0 i 0.70 NS 14.2 i 0.49 < O . 01 decrease 22 .2 r= 1.23 <o 0'2 24 S L , j . i O

S:: 16 0 i 1.77 SP

1 4 . 2 = 0 42 40 6 * :i 11 <0.001 21 2 f 3 34 iYs

Seniinal veairie

Serie, .I I l . . ? .b 0 . 7 1 24.4 .t I . S l <0. 01 1 G . ti :T: 1 . 03 0.0, l . i . . i ;i L.4h xs Series €3 11.8 * (1 .31 :31.e i. 1.80 <0. (I01 1 2 . 6 i 0.74 SS 19.6 i 0.39 <o. 001

deriea ( '

l l . . i i 0 . 3 7 27.4 3 . 7 9 ( 0 . 0 1 9 , r; i 0 , 14

0.01 tlec.re:irc 12.2 i 1.09 x;; Series 1) 1 2 . 1 i 1 . 0 1 1 G . X i 1.17 < o . 02 14.4 I 1 . 7 0 NS 12.1 .& 0 . 5 4 SS 21.8 2 0.27 <o. 001 19.:; * ] . ( I2 ( '0 . 0.001 12 .6 2 0.41 XS

Series 1.; 11.4 .1- U.62 :;1.4 = 1.112 <0 ,001 14.1 ;t 0.70 <O. 0.;

I ) 11

11 i o . 0.i

Alean i standard error. Free alcohol. SS = not sigiiificant. 1's. tealosteroiie propionute.

Clornpouiids of still lower activity are 3, 7! and 11, than 3-10yo the potency of the stan- 4, 5, 10, 13, and 18 are inactive. Thus,

in t'lie substituted c\-clopropane, t8he order of activity is CHO > H >> CH&H > C,O?Et, H,N >>> C02H, C'OSC'H,C'H,. . \ l e s s+ , Br.

Tivo ni:i,jor conclusions can he drawn from these (lata. Ic'irst, act'ivity c n ~ i he ret'airied or enhanced in hubntituted compounds. This would not be expectetl on steric grounds if a-face adsorption were involved, t)(it is in harmony with adsorption on the p face of the st croitl.

Srt:otitily, x t i v i t y of :I givwi :ui:thg is ( I t + w i r i t i ( v I 1 1 ) ucturc of thc substitueiit' 011 t,he cycloprop:m(,

villa. Thorefore, it, is now est'ablished tha t in principle such ;t series of arialogs could shed light on the elcc-

tronic reyuirenieiits in the A ring, :is outliiicd pix'- viously. Howver, the sniall number of act i1.c. coiii-

pounds niakes the present series inntlequ:itc: for 1 t i is purpose, and a further discussion must aw\.nits thr coiii-

plet.iori of another series of compounds, ciirretit1)- being carried out in our laboratory.

Experimental Sectioni9 The i i i e ~ 1 1 ~ L j l b eriiploj ed iri itie androgeiiic-riiy0l rol~l i ic a.

has been discusbed previously.? 2[,3E-Methan0-5a-estran-l7,9-01 (Isomer A ) (3j.- .A s t i r n d mi l -

ture of 8.5 g (0.10 mol?) of Zn--Cu couple, 30 g of CH~12, :rritl I iO iirg of I ? irr 200 n i l OF :tilliyrlroirs Et20 W:IS heat14 i i r i t l i b i , reflu\ for . I lir. '1'1i(~ii : 3 , 0 g 10,00!1 i i i i i I(~) i i L 1 2 ciissolvcd iii : i i i l~ j .dr (~ i i~ 1,:12() \\:t$ a t l t l c ~ l . Tlic, iiiistiirc: \vas heated under reffils for !!O Iir :IIIII them filtered tliroiigli :iliiiiiiti:i, ivashed with dili itc IICI sciliitioii

:iiid I120, :iiid clricd (S:i.>S04 1. Aftrr rvaporation of the. wIv(-tit

July 1968 MYOTROPHIC-ANDROGENIC 2aj3a-J1ETHANO-5a-ANDROSTANES S67

the residue qas dissolved in 100 ml of 570 methanolic KOH solution. It was refluxed for 0.5 hr and the product was ob- tained by evaporation of the MeOH and addition of HzO. It was recrystallized from RIeOH to afford 1.2 g of material, mp 105-110". Further recrystallization gave the analytical sample, mp 114-115', [ a I 2 * ~ +77" ( c 1, CHCl3). Anal. ( C I & ~ O O ) c, H.

From the mother liquid there was obtained 0.4 g of isomer B (4), mp 96-102". Recrystallization from aqueous MeOH gave the analytical sample, mp 104-105", [a]28D +62" (c 1, CHC13). Snal. (C19H300) C, H. 2a,3a-(p-Carboxymethan0)-5a-androstan-17p-o1 (5).-A solu-

tion of 1.0 g (0.0025 mole) of 7 in 100 ml of 5% methanolic KOH was refluxed for 30 min, concentrated in vacuo, and poured into 300 ml of ice-water. Upon acidification to pH 1 with 2070 HC1 crystalline product precipitated. It was filtered and dried to afford 0.8 g (96%) of 5, mp 258-260'. Several recrystalliza- tions from NeOH furnished the analytical sample, mp 265-266", [ a l Z o ~ +14O ( c 0.4, dioxane).

2a,3a- (pCarboxymethano)-5a-androstan-17-one (6).-A solu- tion of 1.20 g of 5 in 200 ml of acetone wa8 allowed to react wit'h 2 ml (excess) of 8 S CrOs solution a t 27". After 20 min, the excess reagent \vas destroyed with i-PrOH and the mixture was filtered. After the addition of a small quantity of Hz0, the mix- ture was evaporated under reduced pressure to give 0.98 g (82Yc) of product, mp 280-281'. Several recrystallizations from N e C N furnished the analytical sample, mp 282-283", [a] 2 0 ~

f76" (c 1, CHClr), nmr 0.851 ppm (19-H). Anal. (CSIHIOOH) C, H.

2a,3a- (p-Carbethoxymethano)-5a-androtan-17p-o1 Acetate (7). -A mixture of 4.0 g (0.0126 mole) of 25 and 0.4 g of anhydrous CuSO4 was heated to 120" and 7.0 g (0.06 mole) of ethyl diazo- acetate was added dropwise. The mixture was kept a t 120" for 15 min under stirring and 10 ml of 10% H0.4~ was added to decompose the excess reagent. The resulting mixture was cooled and extracted with Et20 and the extracts were combined, lyashed with 5 5 SaHC08 solution, and dried (XanSO4). The EtrO solution was evaporated in vacuo and the oily residue was treated with 2 ml of cold MeOH to afford, after filtration, 2.3 g (46%) of the crude product, mp 120-125". Several recrystal- lizat,ions from MeOH gave the analytical sample, mp 148-EO", [aIzo~ -15' ( c 1, CHC13), nmr 0.778 ppm (19-H). Anal. (Cz~Has04) C, H.

Ethylcarbonic 2a,3a-(p-Carboxymethano)-5a-androstan-l7- one'Anhydride @).-A solution of 0.40 g (0.0012 mole) of 6 in 60 ml of Me2C0 and 6 ml of HzO was cooled to 0" and 0.15 g (0.0014 mole) of Eta?; in 2 ml of Me2CO was added. While maintaining the temperature a t O", a solution of 0.144 g (0.0013 mole) of ethyl chloroformate in 2 ml of Me&O was added slowly. The mixture was stirred at 0" for 1 hr and the solvent was evaporakd giving 0.33 g (69%) of crude product, mp 143-144". Several recrystallizations from MeCN furnished the analyt,ical sample, mp 14,5-147', [ C Y ] ~ O D +173' (c 1, CHCl,), nmr 0.820 ppm (19-H).

201,301- [ p-Carbo(l-aziridyl)methano]-5~-androstan-17-one (9). -To a stirred mixture of 0.10 g (0.0024 mole) of ethylenimine and 0.125 g (0.0012 mole) of MeSHz in 5 ml of CeHB was added 0.46 g (0.0012 mole) of 8 in 3 ml of CeHe during 1 hr a t 0" and the reaction was stirred for an additional 1 hr a t 0". The solvent was evaporated in vacuo and the gummy residue was treated with a small amount of cold hexane. The precipitated product (0.23 g) (52%) was recrystallized from hexane to furnish the analytical sample, mp 142-144', [a]'OD +91" (c 1, CHC13). Anal. (C23H33-

2a,3a- [ p-Carbo( 1-aziridyl)methano] -5a-androstan-17p-o

Anal. (CzlH3203) c, H.

Anal. (C24H340j) H ; C: calcd, 71.61; found; 71.20.

NO,) C, H, X.

(19) Melting points were determined with a Thomas-Hoover apparatus equipped with a corrected thermometer. I r spectra were obtained with a Beckman IR-8 or Perkin-Elmer 337 instrument. Microanalyses were per- formed hy the Microanalytical Department, University of California, Berkeley, Calif. Nmr spectra were obtained a t a field strength of 60 MHz on samples in CDCk solutions on a Varian A-80h instrument, using TMS as in- ternal standard. When only small amounts of sample were available, a Varian C-1024 computer was used for t ime averaging. Optical rotations were ob- tained in a 0.5-dm tube with a Rudolph photoelectric polarimeter. Glpc was carried out using a Barber-Coleman Model 5000 system employing 1.83-m U-tuhe columns of 2%; SE-30 on Gas Chrom Q or Z, He carrier, Nn flame detection, column temperatures of 220-240°, "on column" injection a t 290°, and deteotor temperatures of 250'. Where analyses are indicated only by symbol of the elements or functions, analytical results obtained for those elements or functions were within 3~0.4% of the theoretical values.

(IO).-A solution of 0.100 g of 9 and 0.2 g of LiAlH (t-BuO)a in 10 ml of anhydrous T H F was kept a t 0" for 1 hr. The product was isolated by Et20 extraction and recrystallized from MeCK to give the analytical sample, mp 175-176", [aIzon -16' ( c 1, CHC13), nmr 0.788 ppm (19-H). Anal. (C23H3jN02) C, H, N. 2~,3a-(@-Aminomethano)-5~-androstan-l7@-ol Hydrochloride

(ll).-The crystalline free amine derived from 1.50 g (0.0044 mole) of 12 was dissolved in RleOH and reduced with 0.6 g of NaBH4. The excess NaBH4 was decomposed by addition of 5% HCl and the mixture was concentrated in vacuo and poured into ice water. The aqueous solution was made alkaline and extracted with EtnO. The product was rather insoluble inEtzO and partially precipitated. The 0.85 g (637,) of crude amine obtained was dissolved in anhydrous Et20 and acidified with saturated ethe- real HC1 solution. The precipitated salt was filtered and recrys- tallized from EtOH-Et20 to give the analytical sample, mp 279- 280°, [ a l Z o ~ +22" ( c 0.5, CH30H). Anal. (CzoH34C1NO) C, H, x.

2a,3a- [ p-Aminomethano] -5a-androstan-17-one Hydrochloride (l2).-Compound 6 (2.50 g, 0.0076 mole) was converted to 8 as described. To the resulting reaction mixture was added a solution of 0.815 g (0.012 mole) of Nay2 in 6 ml of H20. The mixture was stirred for 1 hr and poured into 500 ml of ice water. The white precipitate was collected by filtration and dried to afford 2.42 g ( 7 3 7 , ) of the crude azide, mp 122" (evolution of N2).

This azide (2.42 g) was dissolved in 12 ml of toluene and heated on a steam bath until no more nitrogen was evolved (1 hr). Removal of toluene in vacuo afforded a yellowish oil which was shown to be almost pure isocyanate by its infrared spectrum ( v g f i 2230 cm-1). The isocyanate was suspended in 16 ml of 20% aqueous HC1 and the mixture was heated under reflux for 1 hr, during which time the amine HC1 precipitated. Recrystalliza- tion from lYc aqueous HC1 gave 2.00 g (78Yc) of colorless needles. Further recrystallization from CH30H-Et20 gave the analytical sample [aylZo~ +92" (c 1, CH,OH). ilnal. (CZoHa2C1NO) C, H, N. 2a,3a-(p-Dimethylaminomethano)-5~-androstan-l7@-0~ meth-

ochloride (13) was obtained from 11 with Me1 and KzCOB in MeOH. A solution of 0.36 g of the methiodide in 5 ml of MeOH was passed through 10 g of IRA-400 resin which was washed with MeOH previously. The eluates were collected until neutral and evaporated to dryness. Several recrystallizations from EtOH-Et20 gave the analytical sample, mp 280-283' dec,

2a,3a-( @-Dimethylaminomethano)-5a-androstan-17-one meth- ochloride (14) was obtained from 12 with Me1 and KOH in MeOH solution. The crude product was recrystallized from H2O to give 0.43 g (60%) of the quaternary iodide, mp 272-276". One additional recrystallization from EtOH-Et20 raised the melting point t o 280-283'.

A solution of the iodide in hIeOH was passed through IRA-400 resin previously washed with RleOH. Recrystallization from EtOH-Et20 gave the analytical sample, mp 274-275' dec, [ a l 2 0 ~ +So (cO.5, CHIOH).

2a,3a-(~-Aldehydomethano)-5a-androstan-17~-01 (15).-Re- duction of 0.68 g (0.0019 mole) of 10 in 180 ml of Et20 with 0.07 g (0.0019 mole) of LiA1H4 a t 0" for 1 hr, decomposition with 10 ml of cold 5 il; HzSO~, and work-up using preparative tlc on silica gel gave 16 (27y0) and 15 (55%). The product 15 had mp 158- 160°, [ a l Z o D +30° ( c 1, CHC1,). 2a,3a-(~-Hydroxymethylmethano)-5a-androstan-17p-ol (16).

-A solution of 0.100 g of 9 in 15 ml of EtOH was treated with 0.04 g of NaBH4. The crude product (98%) was crystallized from MeCY-RIeOH ( 7 : 3 ) to give the analytical sample, mp 217- 218O, [aIzon + 24' ( c 0.5, dioxane), nmr 0.762 ppm (19-H). Anal. (CZ1Hr402) C, H. 2a,3a-(~-Bromomethan0)-5a-androstan-l7@-01 Acetate (18).-

Compound 19 was reduced with L ~ A ~ H ( ~ - B U O ) ~ in T H F solu- tion in the usual way. Acetylation of the resulting alcohol with Ac2O in pyridine solution gave 18. I<ecrystallization from EtOII gave a sample, nip 159-160°, which was a mixture of bromo epimers a5 shown by a multiplet a t 146-157 Hz in the nmr spectrum. 2a,3a-Methano-5a-androstan-17p-ol.-To a mixture of ap-

proximately 2.0 g of Raney Ni and 0.2 g of the product of the reduction of 19 with L ~ A ~ H ( ~ - B U O ) ~ in 150 ml of MeOH was added 1.0 g of KOII ill 5 nil of H20. The resulting mixture wa5 bhakeii under HZ a t 2.1 kg/cm2 for 6 hr. The catalyst was removed, the filtrate was concentrated in Vacuo and HzO was added giving 0.2 g of crude product. Upon separation by preparative tlc (35T0

[ a l Z o D -31" (C 1, CHaOH). A d . (CaoHaoClNO.H,O) C, H, N.

Anal. ( C L ~ H ~ ~ C ~ N O ) C, H, N.

Anal. (C21H3202) C, H.

Anal. (C~2H~~Br02) C, H, Br.

petrolcniii ether (bp :K- t iOo) in Et2() was iised in the deveIo1)- riieiit), there was obtained 0.076 g of 2cu,::cu-metliniiol-.ia-an- tli,ustaii-17p-ol, mp 129-130" after re allization from LleOII. I r aiid rinir spectra arid mixture melti jint were intfistirigiiisli- :ibk from those of t,he aiithentic sample.2

The cornpoiind with lower I t* valiie was isolated i i i ii yield oi O.O!l:$ g aiid was shown t,o be a single isomer of 20,30-(bromo-

I*:tOlI gave colorl

I I r ip le t , l'icu-H) ppiii.

Sq-n thesis of 6,7-Difluorome thylene Corticoids'

The .syiit h i s of Gn,ia- aiid (i~,7~-difl11oronieth~leue corticoids bj. itdditioii of "diflicorijc.:ii,~Jeiie" lo selwred l1 G-:3-keto steroids is described. Tlie ob-erved potentiation uf corticoid d i v i t y bj- boob a- a i d $-face difluoro- riiethylene adduct- is incon4steiit with an antiiriflamni:itc,1.y receptor site which requires biriding to rings A arid 13 of the 3teroid molecvile.

'1'11~~ clriorriious effort c~xperitlrtl 111 the 5 ) nthehis aiid modification of cortisone has led to the accumulstioii of :t corisiderablc bod3 of empirical kiiowledge relating .triicture to antiinflammatorj activitj . Within thr I a Y t two decade5 ever) position of the cortisone moleculr 11:~s been subject to scrutiiil m d clieniical iiiodificatioii Tliese efforts have resulted in t h e discoverj of a number ot activity-enhancing aiid Ltctivitj -modifj irig group< which alorie or in combination have led to the &- vclopment of several clinically useful corticoid\.

.llthough the prirnaq locu' of corticoid action i y

iiiiI\riowri, the h j potliesis is generallj, accepted that biological action i i the rehult of a i l interaction n i th n complement arj i cceptor sit(,. Coriiitlerddc specultr- 1 i o i i a i to the iiature :uid geometi v of the receptor \ i t c has l ~ d to the suggestion that corticoids interact w i t h :L qurface complementar? to :I portion of the $ f:iw of the steroid molecule 8 -6

h:is further defintd the propertie;. of t l i i i

iwc.ptor k ) j \ugge\ting iigid geomctrj with p i o v i ~ i o ~ i ~ for bpecific hiriding to the 118-h~ d r o q 1 and to tlie .3- : i i t ( l 20-keto group5 of 11) clrocortiv)itc.7 Al(lditioii:il i i i t csixctioiii arc' proviclctl hj the wriiiiistion o f 1,oridoii

Sarctt

t I ) l'iil>lic~atiun 340 f rom tlie SJ iiiex Inatitiite Sirmiid C'llciiiiatly. l o r itiil,lication iW4 see 1'. I T . Nelson, J. \V. 1111 .i. Edwardr, and J . f l . 1;ried. .I. A m . Chen. S O C . , in prrss. This I on is also part Y I of the wr iw, l l r t l i j ienation of Unsaturated Ketones. Par t V: G. Tarzia. S. H. l ) y w n , I. '1.liarrison. .J. .\, Edwards, an(1.J. 11. E'ried, Steroida, 9 , 387 (1967). -1 irurtion of this material was presented ai tile symposium on .\ntiinAamma- tor>- .\genrs sponsored by tlie lledicinal Chemistry Section a t tlie 134tli

eetinv of t h e American Chemical Swiety, ('liicago, Ill., Sept 196;. lite of IIornione Bioloiiy, Ssntex Research. Sarr t t , A ? < I L . .V.Y. d c o d . Sca.. 82, 802 (19.59).

, 1) 1,. I f . Sarr t t , '1. . i . Patcliett, and 9. 1.. Steelinan, I ' royr . D r a y R e a . . 5 ,

( a ) J. Fried, "Alc~clianisin o i .Xction of Steroid Iformones," Tile Ala?-

(0) I. E. Rush, I ' t i u r m a c o l . Rei,., 14, 447 11968). ( 7 ) The %ketone can be replaced by a 2,3-fused heterocyclic ring v i t h a

cmisiderat~le enhancement, of biological activitl-; rf. R. Hirschmann, N. G. Stt*intmx, t'. Buciiechachrr, .T. i f . Vrieci, , ~ I < . I . ~ I I I ; L I I , .I. .{VI. ( " h r , , n . Sw., 85, 120 (l!il :iii<I II. l l r u z i k , S t e r o i d , > . 2. :{!)!I ( l ! i t i 3 ) :

1 . i (l!)ti3).

Alillan %., New York, N. Y., 1961, p 232 .

. k'riwl, . J . .lfeil. C l i u n i . . 7, S X L (l!jU4); E'. I>el<)igcieri, ( ' .

, 11. ( 'hiaiaiuuntv and ( ' . I.'t.rruri. Fmrniuco, 20, 280 I l9 l i , j i ,

forceb over tho total of tlie 6 face o f the steroid niolecul(~. AUteriiatively, BuJhb envisages little if aiiy binding to

rings -1 :tiid R with the major interactions being provided by the l l~ -hy t l roxy l , rings C and D, arid tlic. - d e chain. With this hypothesis, the requirements of the receptor \rould riot be inconsistent with axial @-face B-ring s u h h t i t u ~ ~ t h . The illactivity of GP-halo :tnd 68-methyI corticoids. an important consideratiori i i i tlie previouh p l ~ p o s a l , ~ ~ ~ \\ rationalized by the t-ugg~~ct ion t h a t a gerieral distortion of the steroid molecule due to iiitrarnolecular interaction with tliv axial (i-suhstituent irit erfewd wit!) biridirig.6t8

(:learly a cr1tic:iI evaluation of these hypo thew iiiu't bahed o11 biologicall\, active compounds '' 111 particular, :ti1 activc cvrticoid substituted n-it ti I d k y p-f ace €3-ring substituents would provide evideiicct t i l bupport of tlie p r o p o d that rhgs A and B are not iiivolvrtl in biiidiiig to thc complementary surface of I lit1 i'cceptol'.

\I7(> 1iar.c rtwiitly reportcd ail c+ficient method foi*

tiit. prt.p;tr:itiori of fi,i-difluoromethylrrie steroids.11) " The :tpplication ( i f thcw findings to the corticoitl v r i w providrd itti opportunity to further evaluate t h v iwjuiremeiits for bio1ogic:il activitj,.

.Itltlitioii of "( l i f l i io roc : t r t )e i te" to thc dicrioiici\ la , b g : ~ i (1 :t mixture of products from which tlie (io!,-

;cr-difluoron~ethyleiic adducts 2a, b were isolated aftcir

( X ) Tile contrasting antiinflainmatory activities ,if UP- aiiri 6e-iiiwnj ticoids can also attril,iitPd t,ri a inore rapid metaltolic reduction Oi t l i t .

3 ketone in the t i @ - H u i m zeries; cf. 11. J. Rinpoid, 5. Rainachandran, ami

(9 ) Since the IJioiotriral >;, Forcliielli, Biocliim. Bio

(,liemica1 and pilynicachem tion, distribution, and rnetabolisni, absence of biological activity cannot Ire safely ascribed t a tlir lack of binding and, therfore, inferences ns to t l i v geometry of the receptor 3it.e based on inact i re compounds a r e open to que-- tiun.

n , and .J. I [ . Fried, Tetrahedron L e t t e r s , 3281 o n , 1,. Kirkham, and .J. 33. Fried, ibid. . :198i

, 9 , :3hT ( J ! i t i i ) ; ( ' . 1 3 ~ ~ 1 , H . I < e r k o ~ , N. IT. I)yaun, 1. '1. I / $ , (;. 1 , ~ 5 5 1 . I ) . (:iaiinini, .J. \, l ,; ,irar,ih, ami