17a- [3-Furyl]estradiol
Journal of Medicinal Chemistry, 1975, Vol. 18, No. 6 581
Derivatives
Oxidation of Furans. 3.' Estrogenic Properties of Lactones and Anhydrides Derived from the Oxidation of 17a-[3-Furyl]estradiol Derivatives Yvon Lefebvre* and Clara Revesz Ayerst Research Laboratories, Montreal, Quebec, Canada. Received December 23,1974 The oxidation of 17a-[3-furyl]estradiol derivatives and the estrogenic properties of the resulting isomeric 17,21- and 17,23-dihydroxy-19,24-dinor-l7~-chola-l,3,5(10),20(22)-tetraenoic acid(21+23) y-lactones as well as those of the related 17-hydroxy-19-nor-17a-pregna-1,3,5(lO)-triene-2O,2l-dicarboxylic acid anhydrides and y-lactones are described. Of these, only lactones 3c,e,h and 5b retained the same degree and profile of estrogenic activity as the starting 17n-[3-furyl]estradiols. %Substituted furans may be converted t o isomeric butenolides. For example, treatment of 3-isopropylfuran with N-bromosuccinimide afforded 4-hydroxy-2-isopropyl-2butenoic acid y-lactone, while reaction with peracetic acid acid y-lactone.2 gave 4,4-dihydroxy-3-isopropyl-2-butenoic 17(3-[3-Furyl]androstanes,derived from digitoxigenin, afforded, when similarly treated, modified cardenolides,2y3 particularly actodigin, a new isocardenolide with a promising cardiotonic profile of activity different from t h a t of t h e corresponding digitoxigenin d e r i ~ a t i v e . ~ , ~ Therefore, in view of the interesting estrogenic properties exhibited by 17-[3-furyl]estradio16 and 7a,8a-epoxy17- [3-furyl]estradiol derivative^,^ it appeared interesting to convert t h e furan group into butenolides and to evaluate t h e effect of this modification on activity. T h e present paper deals with t h e preparation and biological properties of t h e resulting 17,21- and 17,23-dihydroxy-19,24-dinor17a-chola-1,3,5(10),20(22)-tetraenoicacid(21-23) y-lactones (2 and 3), as well as those of t h e related maleic and succinic anhydrides 6 and 7 and acid lactones 8. Chemistry. Treatment of 17~~-[3-furyl]-3-methoxyestra-1,3,5(10)-trien-17-o16 (la, R = CH3) with an excess of N - bromosuccinimide in aqueous dioxane followed by reduction of t h e reaction mixture (lactones 2a and 2b) with zinc and acetic acid afforded 2b. On the other hand, reaction of l a with rn-chloroperbenzoic acid in chloroform in t h e presence of acetic acid and sodium acetate gave a mix-
4
5a. R' =OH b, R 1 = H
6
7
1
R' 8a, R2 = H
b. R' = CH,
2a, HI = Br b, R1 = H c,
la, R = C H , b, R=c-CjH1o c, R = A c
R1=OH
d, R'= OAc
0
3a, R b, R c. R d, R e, R f, R g. R h. R
CH,; R' = OH CH,; R' = OAc CH,; R' = H c-C,H,,; R' = OH c-c,H,,; R' = H = Ac; R' = OH = Ac; R' = OAc = Ac; R' = H
= = = = =
ture of isomeric hydroxy lactones 3a ( R = CH3; R1 = OH) a n d 2c (R' = OH), the latter being formed only in minor quantities. T h e two lactones were separated as the acetates 3b ( R = CH3; R1= OAc) and 2d (R1 = OAc). Reduction of 3b with sodium borohydride, followed by acidification, afforded lactone 3c ( R = CH3; R1 = H ) , t h e positional isomer of 2b. In view of the high degree of estrogenic activity exhibited by 3c (see below), the analogs 3e, 3h, and 5b were similarly prepared from t h e appropriate fury1 derivatives lb," 1c,6 and 4.7 Finally oxidation of the hydroxy lactone 3a with chromic acid yielded t h e maleic anhydride 6, which upon reduction with zinc and acetic acid gave the succinic anhydride 7. T h e latter compound readily dissolved i n aqueous sodium hydroxide and acidification of the alkaline solution afforded t h e acid lactone 8a (R2 = H ) , readily converted t o its methyl ester 8b (R2 = CH3) in a conventional manner. Pharmacology. T h e oral estrogenicity of the compounds was determined in two standard assays.
(a) Allen-Doisy Test? with a Slight Cornification of t h e vaginal epithelial cells in ovariectomized rats was the end poirft of the experiment. T h e results
582 Journal o/ Medicinal Chemistry, 1975, Vol. 18, No. 6 a r e expressed as t h e dose necessary to induce cornification in 50% of the animals (EDSO).T h e EDSOwas calculated by using an average of 40 (20-100) animals per compound. Dose-response curves were used to determine the ED50 graphically.9 There were at least four dose levels for the determination of t h e EDSOfor each compound.
(b) Uterotrophic Assay in Immature Intact Mice.lo This test was done a t a minimum of five doses for each compound. Five to ten animals were used at each dose level. T h e results are expressed as the minimum effective dose at which a threefold increase in uterine weight over t h a t of t h e controls was obtained. T h e results a r e given in Table I along with data on mestranol which was used as standard. In t h e Allen-Doisy assay, compounds 3c, 3e, 3h, and 5b were highly active, being about 4-20 times as potent as mestranol and exhibiting essentially t h e same degree of activity as the corresponding 1 7 ~[3-furyl]estradiol derivatives.6~~ With the exception of 3b, which was weakly active, all the other lactones a n d the anhydrides were devoid of activity. I n t h e uterotrophic assay the described lactones and anhydrides were much less active t h a n mestranol. In summary, the conversion of t h e fury1 group i n 17a-[3furyllestradioi derivatives into lactones or anhydrides was generally detrimental to t h e estrogenic activity as determined i n t h e Allen-Doisy assay and only compounds 3c, 3e, 3h, a n d j b retained t h e same degree and profile of estrogenic activity as the starting 17~-[3-furyl]estradiols.
Experimental Section Tht. iwnpiiunds were analyzed for C and H and 2a also for Br. Where the elements are indicated the results were within 0.4% of the c~~lculated values. The NMR, ir, and uv spectra were in agreement with the proposed structures. The rotations were taken in 1% chlori~tc~rni o r methanol solutions at 24-25'. The melting points are imcorrected, 17,23-Dihydroxy-3-methoxy-19,24-dinor- l7a-chola- 1,3,5(10),20(22)-tetraen-2l-oic Acid(21-23) y-Lactone (2b). N Bromosuccinimide (j.05 g) was added by portions to a solution of l 7 ~ ~ - / ~ - f u r y l ] 3 - m e t h o x y e s t r a ~ l , 3 , 5 ( l O ) ~ t r i e n(la, ~ l 75.0 - o l g) 6 in dioxane t 2550 ml) and Hz0 (20 ml). The reaction mixture was stirred at room temperature for 30 min, diluted with HzO, and extracted with E t & The solution was washed to neutrality (NaHC03-H20), dried (MgSOd). and evaporated. The crude residue (6.70 g) was dissolvrd in AcOH (335 ml) and the solution was stirred for 1 hr a t room temperature with Zn dust (33.5 9). The metal was filtered and the filtrate diluted with H20. The resulting solid was washed with H 2 0 arid dried. This solid was chromatographed on silica gel m d the fractions eluted with CsHs-EtzO (9:l) were combined and crystallized from CH2CIz-MeOH to yield 2b (2.3 g, 44%): mp 171174': [(>]L) f 7 2 . 1 " (CHC13). Anal. (C23H2804) C, H. In another run the intermediate bromo lactone 2a was isolated i n low yield. :V-Bromosuccinimide (11.1 g) was added by portions I ( ; t~ iolution of l a (11.1g) in dioxane (55 ml) and HzO (45 ml). The reaction was allowed to proceed as above. After work-up the resillized from CH&lZ-MeOH and CHzClz-hexane to afford 2a (2.6 g, 18%):mp 194-196O; [a]D -27.9' (CHC13). Anal. (C2jH2:0J3r) Rr: C: calcd, 61.74: found, 62.64. H: calcd, 6.08; found 6.39. Peracid Oxidations. Preparation of 17,21,21-Trihydroxy19,24-dinor-17a-chola-1,3,5(10),20(22)-tetraen-Z3-oic Acid(21--23) y-Lactones (3 and 5, R = OH). The general procedure for the preparation of the hydroxy lactones 3 and 5 (R = OH) may he illustrated as follows. m- Chloroperbenzoic acid (8596, 23.65 g) was added to a mixture of la (17 g), NaOAc (17 g), AcOH (17 g), and CHCX1 (850 ml). After stirring for 1 hr a t room temperature, Et& was added and the solution was washed to neutrality (NaHC03-H20), dried (MgSOd), and evaporated. The residue was
