1738
B JARTE
LOKEN,
S.KAUFMANN, G. ROSENKRANZ AND F. SONDHEIMER
aqueous methanol was refluxed for 18 hours. The cloudy solution was filtered while hot, then concentrated until turbid, cooled, 50 ml. of water added, and the solids collected by filtration; yield 1.95 g., XEi:E 19,680 a t 240 mfi. The infrared spectrum indicated the loss of the 21-acetate and substantial replacement of hydrazone groups by oxime group. A n d . Calcd. for CzlHs~OaS~: r’J, 7.18. Found: N, 8.61. (B).-A mixture of 9.18 g. of cortisone acetate 3-monosemicarbazone, 2.78 g. of hydroxylamine hydrochloride,
[CONTRIBUTION FROM
THE
Vol. 78
3.28 g. of sodium acetate, 45 ml. of acetic acid and 275 ml. of methanol was refluxed for 15 hours. The methanol wa% removed by distillation, the residue poured into water and the precipitated solid removed by filtration; yield 6.8 g. (7201,), m.p. 165-175” dec., 20,900 at 240 mM. I t s infrared spectrum was identical with that of authentic cortisone acetate 3,20-bis-oxime ( I I I c ) , Anal. Calcd. for C Z ~ H Z ~ O N, ~ N 6.49. ~: Found: K, 6.76. BLOOMFIELD, SEW JERSEY
RESEARCH LABORATORIES OF SYNTEX S.A,]
Steroids. LXXV1I.l Synthesis and Reactions of 16P-Oxygenated Pregnan-20-one Derivatives BY B JARTE LO KEN,^^ S. KAUFMANN, G. ROSENKRANZ AND F. SONDHEIMER~~ RECEIVED OCTOBER28, 1955 Oxidation of AS~16-pregnadien-3fi-ol-20-one acetate ( I ) with alkaline hydrogen peroxide has been shown t o give 16P,17Poxido-17-iso-Aa-pregnen-3~-ol-20-one acetate ( I I I a ) in low yield in addition t o 160r,17a-oxido-As-pregnen-3~-ol-20-one acetate (11). The &oxide I I I a adds hydrogen bromide with the formation of 17a-bromo-A~-pregnene-3~.16P-diol-20-one 3monoacetate (V), iaentical with the product obtained from A5,16-pregnadien-3P-ol-20-one acetate ( I ) by successive addition of bromine to the A5-doublebond, addition of hypobromous acid t o the A”-double bond and treatment with sodium iodide (I VI ‘-f VI1 + V). The bromohydrin V on treatment with potassium hydroxide and reacetylation gives back the @-oxide 3-mOnOIIIa, whereas catalytic hydrogenation over a palladium-charcoal catalyst produces A6-pregnene-3P,16@-diol-20-one acetate (VIIIa), which can be dehydrated to As,18-pregnadien-3p-ol-20-one (I). Lithium aluminum hydride reduction and 3-monoacetate (VIIIa) gives A6-pregnene-3@,16@,20fl-triol trisubsequent acetylation of A6-pregnene-3~,16~-diol-20-one
-
acetate (IXb), identical with the acetylated lithium aluminum hydride reduction product of the “diosone” X I derived from diosgenin X. Catalytic hydrogenation of the triacetate I X b yields the known allopregnane-3P,16P,208-trioltriacetate ( X I I ) and the stereochemical configuration of the various 166-oxygenated pregnan-20-one derivatives is therefore fixed.
160-Hydroxylated pregnan-20-one derivatives such as A5-pregnene-3&16P-diol-20-one3-monoacetate (VIIIa) are of interest since their 16-7methyl-6-acetoxyvalerates are the so-called “diosones” (e.g., XI), which are important intermediates in the industrial conversion of steroidal sapogenins (e.g., diosgenin, X) to the adrenal and the sex hormones. I n contrast to the 16a-hydroxypregnan20-ones which may be obtained by several chemical methods3 as well as by microbiological means,4 only one synthetic route to 1@3-hydroxypregnan-20-ones (unsubstituted a t C-17) has been de~cribed.~ This route proceeds from the corresponding A16-pregnen20-one by addition of hypobromous acid to yield the 16~-hydroxy-17c~-bromopregnan-20-one, followed by debromination with zinc.6.6 The stereochemistry a t C-16 and C-17 of the resulting 160hydroxypregnan-20-one was not, however, rigidly established. ( I ) Paper LXXVI, E. Batres, R . G6mez, G. Rosenkranz and F. Sondheimer. J. Otg. Chem., 21, 240 (1956). (2) (a) Instituto Centro American0 de Investigaciones y Tecnologfa Industrial, Guatemala, C.A.; (b) Department of Chemistry, The Weizmann Institute of Science, Rehovoth, Israel. (3) H. Hirschmann, F. B . Hirschmann and J. W. Corcoran, Federal i o n Proc., 12, 218 (1953); J. Org. Chem.. 20, 572 (1955); H. Hirschmanu, F. B . Hirschmann and G. L. Farrel, THISJOURNAL, 75, 4862 (1953); W. Cole and P. L. Julian, J . Org. Chem., 19, 131 (1954); S. Bernstein, M. Heller and S. M. Stolar, THIS JOURNAL, 76, 5674 (1954). (4) D. Perlman, E. Titus and J. Fried, ibid., 7 4 , 2126 (1952); E. Vischer, J. Schmidlin and A. Wettstein, H e h . Chim. Acfa, 37, 321 ( 1954). (5) G. Gansau, Doctorate Thesis, Technische Universitat, BerlinCharlottenburg, 1952. ADDEDIN PROOF.--S. Bernstein, M . Heller and S. M. Stolar (THISJOURNAL, 77, 5327 (1955)) have now described another route to such compounds. (6) In addition, syntheses of 168,17a-dihydroxypregnan-20-ones have been described recently (a) by H. H. Inhoffen. F. Blomeyer and K. Bruckner [Bcr., 87, 593 (1954)l aad (b) by K . Heusler and A. Wettstein [ i b i d . , 8 7 , 1301 (195411.
We first became interested in this subject when investigating the oxidation of A5.18-pregnadien-3P01-20-one acetate (I) with alkaline hydrogen peroxide. This reaction is known’ t o produce a n oxide in about 95yo yield, shown to be the 16a,l7ar-oxide I1 by its subsequent reactions. The formation of this isomer is in accord with the known predominant a-attack of steroidal 16(17)- and 17(20)-double bonds.8 We have now isolated a by-product in 1-27” yield, for which the lGP,17P-oxide structure IIIa ( 16j3,17~-oxido-17-iso-A5-pregnen-3~-ol-20-one acetate) was first proposed on the basis of the elemental analysis, the infrared spectrum and the comparatively high negative rotation which was expected for a 17-isopregnane derivative.8a This formulation for the substance was subsequently proved to be correct through its further transformations (vide i n f r a ) . Although by-products derived by @-attack of 17-ketoandrostanes (containing a 17(“20”)-double bond) have been isolatedjgthe above appears to be the first recorded instance of the formation of a product derived by @-attack of a steroid with a 16(17)-doublebond. The reaction of 16a,17a-oxido-A6-pregnen-3P01-20-one acetate (11) with hydrogen bromide in acetic acid is knownI0 t o yield 16P-bromo-A6-pregnene-30,17a-diol-2O-one 3-monoacetate (IV) by opening of the oxide ring in such wise as to produce the new substituents in the quasi-axial configura(7) P. L. Julian, E. W. Meyer, W. J. Karpel and 1. R. Waller, THIS JOURNAL, 7 2 , 5145 (1950). (8) L. F. Fieser, Erpericnlia, 6, 312 (1950). (8a) Cf. A. Butenandt and L. Mamoli, Bcr.. 68, 1847 (1935); A. Butenandt and G.Fleischer, ibid., TO, 96 (1937). (9) L. Ruzicka and H. Kggi, Relo. Chim. A c t a , 19, 842 (1936): K. Miescher and W. Klarer, ibid.. PP, 962 (1939). (10) Cf.H.J . Ringold, B. LGken, G.Rosenkranz and F. Sondheimer T H I SJOURNAL, 78, 818 (1956).
April 20, 1956
16P-OXYGENATEDPREGNAN-20-ONE DERIVATIVES
1739
H Ac
tion. The 16@,17@-oxideIIIa has now been which proved to be identical in every respect with shown also to produce a bromohydrin with hydro- the bromohydrin V derived from the 16/3,17p-oxide gen bromide in acetic acid, to which we similarly IIIa. assign the structure 17a-bromo-As-pregnene-3@,- In order to obtain As-pregnene-3P,16P-diol-2016P-diol-20-one 3-monoacetate (V) with the sub- one 3-monoacetate (VIIIa) i t was necessary t o stituents a t C-16 and C-17 in the quasi-axial con- remove the bromo grouping from the bromohydrin figuration. Whereas we have found the 16P-bromo- lr. This debromination was best effected through 17a-hydroxy compound IV to be reconverted read- hydrogenation in methanol solution over a 5% palily to the a-oxide I1 on being heated with pyridine ladium-charcoal catalyst in the presence of ammoa t 90°, the 16P-hydroxy-17a-bromo compound V nium acetate (to prevent reduction of the A5was recovered unchanged under these conditions. double bond), as had been the corresponding 16pBy heating with potassium hydroxide in aqueous bromo-l7a-hydroxy isomer IV.l0 By this means 3-monoacetate methanol, however, the bromohydrin V smoothly A5-pregnene-3P,16P-diol-20-one lost the elements of hydrogen bromide and pro- (VIIIa), further characterized as the diacetate duced 16PI17P- oxido - As- isopregnen - 3P - 01 - 20 - one VIIIb, was obtained in 92% yield. That no gross (IIIb) ; on acetylation the corresponding acetate rearrangement had occurred in the reduction step IIIa was obtained from which the bromohydrin V was shown through chromatography of the diacetate VIIIb on alkaline alumina, when the expected had been derived. Gansaub has shown that the reaction of A4816- @-eliminationtook place" and A6,l6-pregnadien-3ppregnadiene-3,20-dione with N-bromoacetamide 01-20-one acetate (I) was produced. The correin aqueous acetone results in the addition of hypo- sponding conversion of the 3-monoacetate VIIIb to bromous acid to the A16-doublebond with formation I did not occur with alumina, but could be brought of a bromohydrin to which the structure 17a- about through heating with acetic acid and hydrobromo-A4-pregnen-16~-ol-3,20-dione was assigned. chloric acid. The corresponding reaction with A5~l6-pregnadien- The debromination of 17a-bromo-A4-pregnen30-01-20-one acetate (I) required the protection of 16P-ol-3,20-dione to A4-pregnen-16~-ol-3,20-dione the As-double bond as the 5,g-dibromide VIPGa (corresponding to our conversion of V to VIIIa) which with N-bromoacetamide in aqueous acetone had been carried out by Gansau6 in about 20% smoothly produced 5,6,17-tribromopregnane-3@,- yield by means of zinc dust.12 The 16P-hydroxy16P-diol-20-one 3-monoacetate (VII) . The A5- pregnane rather than the 16@-hydroxyisopregnane double bond was regenerated by treatment of the (11) Cf.G. P. Mueller, R. E. Stobaugh and R. S. Winniford, last-mentioned substance with sodium iodide in THIS JOURNAL, 7 6 , 4888 (1953). acetone. This procedure yielded 17a-bromo-A6(12) In accord with this, w e have found t h e reduction of V to VIIIa pregnene-3p,16P-diol-20-one 3-monoacetate (V) , with zinc dust to proceed only in low yield.
1740
BJARTELOKEN,S. KAUFMANN, G. ROSENKRANZ AND F. SONDHEIMER
Vol. 78
structure was favored by the German author on the and allowed t o stand for several weeks at room temperature. I n this way a precipitate with m.p. 130-134", [a]D -45', basis of molecular rotation differences and the as- separated, which m-as chromatographed on acid-Tvashed sumption was made that no inversion occurred in alumina. Hexane eluted first some of the a-oxide I1 and the reduction step. Since the "vicinal action" of then material with m.p. ca. 140'. The latter was rechrothe 1GP-hydroxy group is unknown and the assump- matographed on alumina and the fractions with m.p. above 165", eluted with hexane, were crystallized from methylene tion regarding the retention of configuration is not chloride-methanol. This procedure produced i i i 1-270 necessarily justified, we felt it desirable to provide yield the 8-oxide I I I a with m.p. 176-178", [ a ]-108", ~ no additional evidence for the stereochemistry as- high-intensity absorptiou in the ultraviolet. signed to the debromination product VIIIa through dnul. Calcd. for ClaH8204:C, 74.16; H , 8.66. Found: its identification with a known This C, 74.00; 13, 8.66. (b) From 17a-Bromo-Aj-pregnene-3p, 16p-diol-20-one 3was achieved through reduction of T'IIIa with litli(V).--A solution of 3.5 g. of potassium liyium aluminum hydride. The A5-pregnene-3,16,20- Monoacetate droxide in 5 cc. of water and 20 cc. of methanol was added to triol thus obtained gave a triacetate which was a suspension of 10 g. of the broinohydrin Y (described bcidentical with a sample prepared by the lithium alu- low) in 100 cc. of methanol and the mixture \ Y : L ~ boiled ut:der minum hydride reduction and subsequent acetyla- reflux for 1 hour. IVater was added and the product mt, isolated with methl-lene chloride in tlie usual w a y . Cry,tion of A5-pregnene-3B,lC,3-diol-20-one 3-acetate tallization from methylene chloride-methanol furnislietl 6.1 16-y-methyl-6-acetoxyvalerate ("diosone") (XI).l 4 g. of the &oxide I I I b with m.1,. 143-145', [ a ] D - 110'. Moreover, catalytic hydrogenation of the S-double Acetylatioii of this materi:il ~vithpyridiii bond of the triacetate proclucctl a compound, iden- dride (1 hour :it 90")and iub-eyuent c chloride-methanol produced tified with a substance shown by Hirschmann, ei methylene all) of the 0-oxide acetate 1113 n-ith m.p. 175-177", la111 uZ.,I5 t o he allopregiiane-3,3,1G3,20$-triol triacetate -109". The m.1). w ~ i bundepressed O i l admixture lvith 3 (XII). The triol obtained by the lithium alumi- sample prepared by method a . 16~,17LJ-Oxido-17-iso-A~-pregnen-3Q-ol-2Q-one ( IIIbi .nuin hydride reduction of VIIIa and XI is there2 g . of the 8-oxide acetate 111:) (obfore A5-pregnene-3,J,16,3,20B-triol (IXa) and the ) . 3 g. ~i potassium hydroxide, 1 cc. of A5- pregnene - 38,163 -diol - 20 - one 3 - monoacetate etlimol was boiled under rcflus fur 1 structure for VIIIa is rigidly established. This in liour. .4cidification xrith acetic acid, evapor,itioii to -mall turn provides final evidence for the 16b-hydroxy- volume and addition of water gave a precipitate n.hicli OII froiri niethj-lerie cliloride--metli:~:iol yielded 17cr-brorno structure for the hroiiiohydrin T: and the crystallization 1.62 g. fY17,) of the &oxide I I I b with m . p . 117-148', 16,6,1i$-oxidostructure for the oxide IIIa. l u ] O -11Za. The m , p . w:i> undcpre>sed on admixture with Acknowledgrnents.-II*e would like to express the sample of I I I b de,cribed above, obtained by base treatour thanks to Dr. H . Hirschrnann of the Depart- ment of the hromclhydri:~1.. .4nal. Calcd. for Cr,H,jcO,,: C, 75.32; H , 9.15. I'(JuI!~: merit of Xedicine, n'estern Reserve Unive:sity, c , 76.37; €1, 9.13. for interesting corr'espondence and for carrying out 5,6,17-Tribromopregnane-3~,16p-diol-20-one 3-Monothe identification of our allopregnarie-3B,lBP,~O~-acetate (VII) .-~5~'G-Pregnadien-3~-01-20-orie acetate triol triacetate (XII) and to Dr. G. Gansau for was converted t o the 5,G-dibromide VI (m.p. 132-124 , ng us with a copy of tlic thesis mentioned in [a10 -52',), as described by Iiihoffeii, et ~ 1 . 6 " X-Bromoacetctmide (170 g., freshly crystallized from chloroform) was footnote 5. added t o a stirred solutioii of 170 g. of the dibromide V I ExperimentalLo in 7.2 1. of acetone and 720 cc. of water a t room temperature.
(a)
16p,17p-Oxido-17-iso-A6-pregnen-3p-ol-20-one Acetate (IIIaj , (a) As By-product in the Hydrogen Peroxide OxidaAcetate (I).-The oxition of A5,'E-Pregnadien-3,3-ol-20-one acetate (I) with dation of ~6,i6-pregnadien-~,~-ol-20-one alkaline hydrogen peroxide was carried out as described by Julian, et al.' The total product was acetylated (acetic anhydride-pyridine, 1 hour a t 90") and the isolated product was triturated with hexane a t room temperature. The insoluble material, obtaiued in 'JO-Rj7, yield, was almost pure 16a,l7a-oxido-A~-pregneii-3&ol-20-one acetate (11) withm.p. 156-15X3, [ a ]-10". ~ T h e hexane washiiigs werc evaporated to small volume (13) An indicrtion t h a t t h e debromination product VIIIa had the 168-hydroxypregnane structure a s has the "diosone" X I (d ref. 15) was provided b y t h e fact t h a t 'I'IIIa a n d its diacetate X'IIIb undsrwent dehydration n i t h hot aqueoiis acetic acid a t almost identical r a t r s as did t h e "diosone" XI (der Experimental Sectionj (11) T h e triacetate is probabl, also identical with t h e 15-pregnene.3,16,'2O-trioI trixcetate obtained h y K . E. X a r k e r , D. I.. Turner, K . B. TVagner, P. K . Ulshafcr, H . 11. Crooks and E. L . Wittle, THISJ O U R X A L , 63,774 110-&1)I b y ?ileerxr.ein-Ponndorf reduction and subsequent acet>:alion oi the "rliosoiie" X I , (1)) H. Hirschmann, 1'. R . Iiirschmann a n d h l . A. Dau5, J . H i o i . C h e h z . , 178, 751 (1919). T h e substance had been obtained !iy these norkers b o t h b y a repetition of LInrker's degradation of tigogenin, involving preparstion of t h e "diosonr," catalytic hydrogenation over a platinum catalyst, saponification a n d acetylation (Marker, et ai., r e f . 14) a n d also b y a series of redctinns involving t h e urinary steroid allopregnane-313,lRu,SO8-triul. (10) lleltinl: p u i t i t s i i i o iinw>ri-vciecl, l'nlcss rioted o t h r r a i . c , rulatiuns n e r e < l < , t c r i i i i i i u l .it ?(I" i i i clilc,rtlfrption in the ultrziviolet. U'ater (13 1.) \vas therefore added and the mixture was extracted several time, with methylciie chloride. The organic extract was washed with water, dried and evaporated almost to dryness under reduced pressure. Addition of methanol and cooliilg produced X4 g. (42%i of the tribromo compound VI1 \r.it!i m.p. 151. -133'. A further purified sample shon.ed n1.p. 136-137.3', iuln -42', 110 high-inteniity ahsorptioii in the ultrnviolct .
17n-Bronio-Ah-~regnene-3~~,16p-diol-20-one 3-Monoacetate (V). (a; From 5,6,17-Tribromopregnane-3,16p-diolZQ-one 3-hlonoacetate (VII).--Sodium iodide (56 g.1 w i s iidded t o :I tirreti ,ul colorless solution wah poured into 2.6 1. of water, whereby :i.n nil separated w!iieh boo11 solidified. Tlie 1irccipit:tte \la:: collected, dried and crystallized from aqueous rncthanol. 111 this \\-a>26.8 g. (87Y0i of t!ic brornohydriri \. with u1.p. If56-1S3" \r.:i\ obtaitird. A furthrr pllrificd specililell . l l t l \ Y < ~ L l 1 1 1 . 1 1 . 1.78 159". j C Y I l 1 - i-4". .I!/CI/.L':11\,
