J. Am. Chem. SOC.1984, 106, 7283-7285
7283
Table 11. Control Exoeriments with "0, ~~~~
~
composition, % origin of formic acid 17 46 37 ( I ) 2HC180180Na" 20 41 39 (2) same as I , incubated 60 minb 43 38 19 (3) same as 2 65 20 min 35 (4) from incubation of 4, H = *H, in 62 38 (5) same as 4, incubated 60 min "Svnthetic samole used also in Exoeriments 2 and 3. GC-MS analvses were carried out as described.l' *For conditions of incubation, see text (4 and 5 are averages of duplicate experiments). 139 for 2HC160'80CH2C6H5 (Table 11, entries 4 and 5). These that at least 50% of the molecules (6a) were labeled with lSO at results confirm the Akhtar et aL9 observations that the "third" the 2P-hydroxy moiety. The silyl ether was cleaved with aqueous mole of oxygen is incorporated into the extruded formic acid. acetic acid and, after TLC (hexane-EtOAc 2:1), homogeneous To evaluate the operation of pathway C (Scheme 11), 19-DT [2P-I8O,19-3H]-2P-hydroxy- 1O~-formylandrost-4-ene-3,17-dione aldehyde 4 was incubated with placental aromatase in 180H2(30% (6b) was obtained. Two sets of experiments were then carried out. In the first excess of lSO) in the air. The rationale of the experiment was experiment, 6b [ 100 pg (3 X IO4 dpm of 3H) in each of five flasks] based on the premise that, if Fishman's hypothesis is correct, the was incubated in Tris-buffer (pH 7.4) with placental aromatase 1602 should be utilized for C-2 hydroxylation to give 6b,2P-160H, for 1 h at 37 "C, under nitrogen, as previously d e ~ c r i b e d . ~At J~ which will then aromatize with the incorporation of l80from the the termination of the reaction, the contents of the flasks were water into the formic acid. The recovered formic acid (60% yield) rapidly combined, the mixture was acidified and frozen in liquid contained only l 6 0 . These results when taken together with the nitrogen, and the formic acid was recovered by l y o p h i l i ~ a t i o n . ~ ~ ~results ~ on the aromatization of [180]-6bexclude the operation of The derived sodium formate was then converted to benzyl forpathway C, Scheme 11. ~ n a t e . ' ~The . ~ second set of experiments was carried out exactly Our results show that the oxygen atom of the [180]-2p-hydroxyl as above4J5 but without placental aromatase, and the recovered of 6b was not incorporated into the extruded formic acid derived from C-19. It may therefore be concluded that the mechanism formic acid was also converted to benzyl formate. Each of the proposed by Fishman et al.'&13 via enzymatic formation of 2ptwo samples of benzyl formate contained tritium (1.35 X IO5 dpm), indicating that ca. 90% of the substrate 6b was aromatized. The hydroxy- 10~-formylandrost-4-ene-3,17-dione (6b) and its nonGC-MS" of the two samples were recorded, and the results are enzymatic aromatization is not an obligatory pathway of estrogen summarized in Table I. The benzyl formates showed peaks at biosynthesis by placental aromatase. m l z 136 and 137, but none was present at m l z 138 indicating the absence of I 8 0 . Usually, variable amounts of endogenous [ 160]formicacid were recovered from the placental aromatase preparations. It is Reaction of Ketene Alkylsilyl Acetals with therefore of importance that the benzyl formate derived from Bromoform-Diethylzinc. An Unprecedented aromatization of [I80]-6bin the absence of placental aromatase Cyclopropanation Reaction gave formic acid (analyzed as benzyl formate) which contained only l60.This benzyl formate could not be contaminated with Gtrard Rousseau* and Nasser Slougui a formate of endogenous origin and must have originated solely Laboratoire des Carbocycles, UA(CNRS)478 from C-19 of the [2P-180H]-6bsubstrate. Universitd de Paris-Sud, Bdtiment 420 To determine whether l 8 0 was exchanged (lost) under the F-91405 Orsay Cedex, France experimental conditions, ZHC180180Na (80% I8O enrichment) Received July 24, 1984 was ~ r e p a r e dand , ~ an aliquot of the salt was converted to benzyl f ~ r m a t e . IA ~ -second ~ aliquot was incubated with human placental In conjunction with our program dealing with the reactivity aromatase (1 h, at 37 O C , in the air), and the formic acid was of ketene alkylsilyl acetals and carbenes, we have reported that recovered by lyophilization of the acidified m i x t ~ r e . ~ The J~ the reaction of those species with chlorocarbenes provides a GC-MS analyses of the two [1802]ben~yl formates indicated that convenient route to a-substituted a$-ethylenic esters via a-bond no detectable loss of l 8 0 occurred (Table 11, entries 1-3). cleavage of an intermediate cyclopropane (eq 1I). We have also incubated [ 19-2H]-1OP-formylandrost-4-ene3,17-dione 4, H = 2H] in an atmosphere of Two sets of incubations (in duplicate) were carried out for 20 and 60 min with human placental aromatase4J5 in an atmosphere of lSOz(98% excess). The recovered formic acids were analyzed (as benzyl formates) by GC-MS," and all four samples showed ions at m l z Recently it has been shown that the reaction of cyclopropanone ethyltrimethylsilyl acetal with titanium(1V) chloride results in (14) The synthesis was carried out by treating [19-'H]-6P-bromo-19the formation of an ester homoenolate.2 This result suggested acetoxyandrost-4-ene-3,17-dione with '802-labeledpotassium acetate in glacial that chlorocyclopropanoneacetals could lead to a carbenoid species I80-labeled acetic acid to give [2@-1802,19-3H]-2&19-diacetoxyandrost-4ene-3,17-dione. Saponification followed by selective silylation of the resulting by cyclopropane b-bond ring cleavage (eq 2) if the reaction was 28,19-diol furnished 19-hydroxy-2P-(fert-butyldimethyIsilyl) ether which on carried out in the presence of a Lewis acid. mild oxidation gave the required 6a. The proton NMR of 6a showed signals at 6 9.78 (1 H, s, 19-CHO),5.85 (1 H, s, C4-H), 3.98 (1 H, br s, 2a-H), and 0.82 (3 H, s, C18-H) in accordance with published data.1° The MS of [160]-2,9-silylether 6s showed ions at m / r 402 (M' - 28) and 373 (M+ C4H9),while the MS of [180]-2,Y-silylether 6a showed ions at m/z 402 (49%), 404 (51%). 373 (45%), and 375 (55%). (15) Caspi, E.; Arunachalam, T.; Nelson, P. A., manuscript in preparation. (16) Corina, D.L. Anal. Biochem. 1977, 80, 639. (17) A Varian Model 3700 GC instrument equipped with a glass capillary column (25 m) coated with OV-101 was used. Injection port temperature 270 "C;column temperature 7C-280 OC;temperature gradient 3 "C/min. The GC was linked to a Varian-MAT Model 312 mass spectrometer via a direct inlet. Spectra were recorded at 70 eV.
0002-7863/84/1506-7283$01.50/0 , I
,
c:
CI
R I A O R 3 R2 OSiMe,
M-X (Louis a c i d 1
-
R2 Me3Si0
products ( 2 )
OR3
( I ) Slougui, N.; Rousseau, G.; Conia, J. M. Synthesis 1982, 58. Slougui, N.; Rousseau, G. Synth. Commun. 1982, 12, 401. (2) Nakamura, E.; Kuwajima, I. J . Am. Chem. SOC.1983, 105, 651.
0 1984 American Chemical Society
1284 J . Am. Chem. SOC.,Vol. 106, No. 23, 1984
Communications to the Editor
Table I. Cyclopropane Ester Synthesis from Ketene Acetals entry
ketene acetal
cyclopropane esteP yield, %
a
b
Scheme 1. Postulated Mechanism of the Bromoform-Diethylzinc-Ketene Acetal Reaction
80
w
82
COOEt
72
C
d
(9
COOMe'
55
e
63
f
58
g
4
80
Scheme IIP Preparation of (*)-Sabinene
8
COOMe'
h
18 OEt I
II i
10
j
40
6 COOEt
a All reactions were performed at 20 "C. 5 mmol of ketene acetal, 2.5 mL of pentane, and 3.75 m L of 1 M Et,Zn in pentane (1.5 equiv) were mixed under argon; 0.655 m L of CH Br, (1.5 equiv) was added dropwise (30 min); then, oxygen bubbling for 1 minS initiated the reaction. After 2 h at room temperature and a standard workup, the products were isolated via liquid chromatogThe reaction conducted with the E : Z = 8 9 : l l or the raphy. E:Z =
