Synthesis of 21,21-difluoro-3β-hydroxy-20-methylpregna-5,20-diene and 5,16,20-triene as potential inhibitors of steroid C17(20) lyase

Article abstract of DOI:10.1016/S0968-0896(02)00434-0

Novel 21,21-difluorovinyl steroids, designed as difluorinated C₂₀₍₂₁₎ enol mimics of pregnenolone, were targeted as potential mechanism-based inhibitors of C₁₇₍₂₀₎lyase, a crucial enzyme in the biosynthesis of testosterone. Addition of (difluoromethyl)diphenylphosphine oxide reagent to 17-acetyl steroids was the approach chosen for the construction of these compounds. Of particular interest were the abnormal Wittig products which formed during attempted preparation of the triene 9. The target difluoroolefin 3 was found to be a moderately potent, time-dependent inhibitor of the enzyme.

Full text of DOI:10.1016/S0968-0896(02)00434-0

Bioorganic & Medicinal Chemistry 11 (2003) 427–431
Synthesis of 21,21-Difluoro-3ꢀ-hydroxy-20-methylpregna-5,20-diene
and 5,16,20-Triene as Potential Inhibitors of Steroid C17(20) Lyase
Philip M. Weintraub,a,* Amy K. Holland,b Cynthia A. Gates,c William R. Moore,d
Robert J. Resvick,,,a Philippe Beye and Norton P. Peetf,*
aAventis Pharmaceuticals; Rte 202-206, PO Box 6800, Bridgewater, NJ08807-0800, USA
bVankel, 13000 Weston Parkway, Cary, NC 27513, USA
cVertex Pharmaceuticals Inc., 130 Waverly St., Cambridge, MA 02139-4242, USA
dLocus Discoveries, Inc., 512 Township Line Rd, Blue Bell, PA 19422, USA
eArQule, 19 Presidential Way, Woburn, MA 01801-5140, USA
fAurigene Discovery Technologies, 8-A Preston Court, Bedford, MA 01730, USA
Received 22 May 2002; accepted 29 August 2002
Abstract—Novel 21,21-difluorovinyl steroids, designed as difluorinated C20(21) enol mimics of pregnenolone, were targeted as
potential mechanism-based inhibitors of C17(20) lyase, a crucial enzyme in the biosynthesis of testosterone. Addition of (difluoro-
methyl)diphenylphosphine oxide reagent to 17-acetyl steroids was the approach chosen for the construction of these compounds. Of
particular interest were the abnormal Wittig products which formed during attempted preparation of the triene 9. The target
difluoroolefin 3 was found to be a moderately potent, time-dependent inhibitor of the enzyme.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
which displayed potent, time-dependent inhibition of
C17(20) lyase.9 This report is an extension of our workon
enol mimics as potential C17(20) lyase inhibitors.
Prostate cancer is a leading cause of cancer-related
deaths among men.1 Treatment is primarily aimed at
blocking the synthesis and action of androgens. Andro-
gen deprivation can be achieved surgically or through
the use of LHRH analogues in combination with anti-
androgens. A key enzyme in the biosynthetic pathway
for the synthesis of androgens is C17(20) lyase.2,3 This
enzyme converts the C21-progestins to C19-androgens by
cleavage of the two-carbon acetyl side chain.
The fluoro-olefins which we designed as C17(20) enol
mimics are compounds 1 and 29 as shown in Figure 1.
A variety of C17(20) lyase inhibitors of different
mechanistic types and structural classes has been
described.4 Our laboratory has been interested in devel-
oping mechanism-based inhibitors of this enzyme and
we have prepared a number of C17-cyclopropylamines,5
C17-cyclopropyl ethers,6 and C17-heterocyclic com-
pounds7 for this purpose as well as combined C17(20) lyase/
5a-reductase inhibitors.8 We recently reported fluoro-ole-
fins which were C17(20) enol mimics of pregnenolone,
*Corresponding authors. Tel.: +1-908-231-2572, e-mail: philip.
weintraub@aventis.com (P. M. Weintraub); Tel.: +1-978-804-8099,
e-mail: norton_p@aurigene.com (N. P. Peet).
,Deceased 28 April 1998.
Figure 1. C17(20) and C20(21) fluoroolefin enol mimics.
0968-0896/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00434-0

428
P. M. Weintraub et al. / Bioorg. Med. Chem. 11 (2003) 427–431
Table 1. Inhibition of cymologous monkey testicular C17(20) lyase with
fluoroolefin enol mimics of pregnenolone and related compounds16
These compounds are potent inhibitors of cynomolgous
monkey testicular C17(20) lyase (Table 1). From the pre-
incubation experiments, the Z-isomer 1 is clearly
demonstrating time-dependent behavior. Thus, we
wanted to explore the activity of novel difluoroolefin 3,
which we felt would provide a somewhat analogous
mimic of the C20(21) enol of pregnenolone.
Compd
Enol
mimic
Concentration
(mM)
Preincubation
(min)
Inhibition
(%)
1
C17(20)
C17(20)
C20(21)
C20(21)
10
10
1
0
40
0
78
94
49
72
1
40
Starting with pregnenolone (4), we prepared pregneno-
lone acetate (5) and treated this ketone with (difluoro-
methyl)diphenylphosphine oxide10 as shown in Scheme
1. A mixture of products was obtained which included
the desired difluorolefin 3 (20%). Also produced in 14%
yield was 6, the acetate of 3, along with pregnenolone
(30%) and pregnenolone acetate (7%). As a reference
compound, we also prepared the simple, nonfluorinated
vinyl compound 711 as also shown in Scheme 1.
2
3
6
10
10
1
0
40
0
88
94
54
60
1
40
1
1
0.1
0.1
0
40
0
10
40
11.5
8
40
1
1
0.1
0.1
0
40
0
7
12
5
40
5
Since compound 3 did display inhibition of the target
enzyme (Table 1), although modest, we decided to pre-
pare the Á16 analogues as we have previously found that
this additional double bond imparted enhanced inhibi-
tory activity to a series of C17-heterocyclic com-
pounds.7,12 As shown in Scheme 2, diene 8 was treated
with (difluoromethyl)diphenyphosphine oxide to pro-
vide a separable mixture of the desired 21,21-difluoro-
3b-hydroxy-20-methylpregna-5,16,20-triene (9) and
phosphonate 10. Compound 10 is probably produced
by addition of the C3-alcoholate of 8 to the difluoro-
methyl phosphinate with subsequent elimination of the
relatively stable difluoromethyl anion. Similar anamo-
lous reaction products have been observed in the
difluoromethenylation of ketones.13,14
7
C20(21)
C20(21)
C17(20)
1
1
0
40
70
60
9
1
1
0
40
64
62
14
1
1
0
40
100
85
In an attempt to circumvent the production of the
undesired phosphinate ester 10, a side reaction which
was consuming starting material 8, we chose to protect
the alcohol. Treatment of 8 with TBDMS-Cl gave silyl
Scheme 1. Reagents: (a) Ac2O; (b) Ph2P(O)CHF2; (c) Ph3P+CH3Brꢀ.
Scheme 2. Reagents: (a) Ph2P(O)CHF2; (b) TBDMS-Cl; (c) DMSO, 160–170 ꢁC, 4 h; (d) 3-bromotoluene, reflux, 3 h.

P. M. Weintraub et al. / Bioorg. Med. Chem. 11 (2003) 427–431
429
ether 11 which was subsequently treated with (difluoro-
methyl)-diphenylphosphine oxide. The presence of the
silyl ether changed the course of the reaction, such that
diphenylphonic ester 12 was isolated. Recognizing that
compound 12 could represent an intermediate on the
path to the desired difluoroolefin via a potential thermal
elimination of diphosphosphonic acid, we attempted to
effect this elimination at elevated temperatures. Thermal
gravimetric analysis (TGA) showed no weight loss to
210 ꢁC at which temperature a pronounced loss occurred.
And lastly, the conformationally rigid enone 14 dis-
played potent enzyme inhibition but no time-depen-
dency. Since compound 14 is a Michael acceptor, the
potential exists for it to interact with an active site
nucleophile. However, if this interaction is operative, it is
a fast process.
Conclusion
In summary, we have extended our studies on fluor-
oolefin pregnenolone enol mimetics from the C17(20) to
the C20(21) mimetics. Indeed, our target C20(21) pregne-
nolone enol mimetic, difluoroolefin 3, was synthesized
and shown to be a moderately potent, time-dependent
inhibitor of cymologous monkey testicular C17(20) lyase.
Thermolysis of 12 in dimethylsulfoxide at 160–170 ꢁC
for 4 h gave a new product which was isolated and
identified by spectral analysis as a,b-unsaturated ketone
14. It seems likely that 14 emanated from sulfonium ion
13 by oxidative 1,2-elimination of a proton, analogous
to the Swern oxidation. Several possible routes to 13
can be envisioned. Direct SN20 displacement of
diphenylphosphinic acid by DMSO can give 13. Alter-
natively, E1 or E2 elimination of Ph2POꢀ2 with sub-
sequent trapping of the resulting carbonium ion A by
dimethyl sulfoxide can give 13. One can also propose a
[3,3]-sigmatropic rearrangement of 12 to give 15 which
then could suffer E1 elimination of Ph2POꢀ2 in the polar
DMSO and subsequent trapping by solvent. The possi-
ble intermediancy of 15 is suggested by the thermolysis
of 12, in 3-bromotoluene for 3 hr at reflux temperature
from which compound 14 was isolated.15
Experimental
General
Melting points were determined with a Thomas-Hoover
capillary melting point apparatus and are uncorrected.
TLC analyses were performed with MerckDC-F254
silica gel plates, with visualization by UV light and
alkaline permanganate. Flash chromatography was
performed with Mercksilica gel 60 (0.040–0.063 mm).
NMR spectra were recorded on Unity 400 or Gemini-
300 spectrometers in CDCl3, unless otherwise stated. 1H
and 13C NMR signals are reported in ppm from tetra-
methylsilane (s, d, t, m and br for singlet, doublet, tri-
plet, multiplet and broad, respectively) and coupling
constants are reported in hertz (Hz). Infrared spectra
were recorded on a Perkin–Elmer Model 1800 or Matt-
son Galaxy 5020 FT-IR spectrophotometer. Mass spec-
tral data were collected at 70 eV on a Finnigan MAT
4600, Mat TSQ-700 or VG Analytical Ltd ZAB2-SE
mass spectrometer and computerized for HRMS. Com-
bustion analyses were performed using a Perkin–Elmer
Model 2400 elemental analyzer and fell within ꢂ0.4%
of calculated values. The organic extracts were dried
over magnesium sulfate or sodium sulfate prior to eva-
poration in vacuum on a rotary evaporator. Anhydrous
solvents and starting materials were purchased from
Aldrich Chemical Co. and used as obtained with the
following exception: dehydroepiandrosterone was pur-
chased from Proquina.
Results and Discussion
The novel compounds prepared in this study were eval-
uated for their ability to inhibit cynomolgous monkey
testicular C17(20) lyase at two concentrations, with and
without preincubation of enzyme with the potential
inhibitor. These results are presented in Table 1. Also
included in Table 1 are the two C17(20) fluoroolefin enol
mimics (Z-isomer 1 and E-isomer 2) which were pre-
viously prepared,9 for comparison.
The target molecule of this study, difluorolefin C20(21)
enol mimic 3, was a time-dependent inhibitor of the
monkey enzyme at the 1 mM concentration, although
not quite so potent as the C17(20) enol mimics 1 and 2.
Compound 3 was the only new compound in Table 1 to
display time-dependency, although three additional
compounds 7, 8 and 14 were potent inhibitors of the
enzyme at both test concentrations.
21,21-Difluoro-20-methylpregna-5,20-dien-3ꢀ-ol (3) and
21,21-difluoro-20-methylpregna-5,20-dien-3ꢀ-ol Acetate
(6). To a solution of lithium diisopropyl amide (4.00
mmol) in THF (50 mL) cooled to ꢀ78 ꢁC was added a
solution of (difluoromethyl)diphenylphosphine oxide
(1.01 g, 4.00 mmol) in THF (2 mL). After 50 min, a
solution of pregnenolone acetate (1.43 g, 4.00 mmol) in
THF (15 mL) was added dropwise. The reaction was
warmed to room temperature and heated at reflux for
90 min. The cooled reaction was quenched by pouring
into water. The products were extracted into methylene
chloride and purified by flash chromatography eluting
with hexane ꢀ20% ethyl acetate (33ꢃ50 mL) and hexane–
40% ethyl acetate (8ꢃ100 mL).
Compound 6, the O-acetate of inhibitor 3, was essen-
tially inactive, which underscores the necessity of the
free hydroxyl group at the 3-position for active site affi-
nity. Interestingly, compound 7, the dihydrogen analogue
of inhibitor 3, was a potent inhibitor but did not display
time-dependency. This suggests that the fluoro groups do
provide a special contribution to enzyme inhibition.
Compound 9, the 16,17-unsaturated version of inhibitor
3, was a potent inhibitor of the monkey enzyme but was
not time-dependent. Perhaps the conjugation imposes a
geometry which is suitable for enzyme affinity but does
not facilitate a time-dependent process.

430
P. M. Weintraub et al. / Bioorg. Med. Chem. 11 (2003) 427–431
The first material to elute from the column was con-
centrated to give 215 mg (13.7%) of 6, mp 124–130 (aq
MeOH): IR n 1734, 1250 cmꢀ1; MS (FAB) m/e 415
solid, 10 (241 mg, 2.9%); 1H NMR d (CDCl3) 0.81 (3H,
s), 1.05 (s), 2.26 (3H, s, C21–Me), 4.15–4.33 (1H, m, C3–
H), 5.28–5.36 (1H, m, C6–H), 6.71 (1H, t, C16-H), 7.41–
7.57 (6H, m), 7.78–7.90 (4H, m).
1
(85%, M+Na), 333 (100%, M+1-AcOH); H NMR d
32;(CDCl3) 0.65 (3H, s), 1.05 (s), 1.56–1.61 (m, C21–
Me), 4.53–4.68 (1H, m, C3–H), 5.34–5.42 (1H, m, C6–
H); F NMR d 32;(CDCl3)-92.65 (d, J=55 Hz), ꢀ93.35
(dq, J=3.3, 55 Hz). Anal. calcd for C24H34F2O2: C,
73.44; H, 8.73. Found: C, 73.86; H, 9.19.
3ꢀ-(Dimethyl tert-butylsilyloxy)-pregna-5,16-dien-20-one
(11). To a solution of alcohol 8 (5.0 g, 15.9 mmol) in
pyridine (30 mL) was added tert-butyldimethysilyl
chloride (2.7 g, 18.2 mmol) and dimethylaminopyridine
(0.213 g). After 2.5 h, additional TBDMS-Cl (1.35 g, 9.1
mmol) and DMAP (106 mg) were added and stirring
was continued for 3 days. The reaction mixture was
partitioned between CH2Cl2 and H2O. The organic
layer was washed with H2O, twice with 1 N HCl, with
brine, dried and concentrated. Purification by flash
chromatography (SiO2, hexane–15% EtOAc) gave 11
(6.03 g, 80.5%), mp 150–152 ꢁC.
The next material to elute from the column was con-
centrated to give 100 mg (7.0%) of recovered starting
material.
The third material to elute from the column was con-
centrated to give 281 mg (20.0%) of 3, mp 150–151 ꢁC
(EtOAc–hexane): IR n 3428, 1734 cmꢀ1; MS (CI) m/e 351
(8%, M+1), 350 (18%, M+), 349 (25%, M-1), 333
(100%, M+1-H2O), 313 (15%, 333-HF); (EI) m/e 350
(100%, M+); 332 (45%, M–H2O), 335 (12%, M–CH3),
317 (99%, 335-H2O), 213 (95%); 1H NMR d(CDCl3) 0.64
(3H, s), 1.02 (s), 1.54 (t, C21–Me), 3.46–3.58 (1H, m, C3–
H), 5.33–5.38 (1H, m, C6–H); F NMR d (CDCl3)ꢀ92.60
(d, J=2.7 Hz), ꢀ93.32 (dq, J=56 Hz). Anal. calcd for
C22H32F2O: C, 75.39; H, 9.20. Found: C, 74.70; H, 9.03.
21,21-Difluoro-3ꢀ-(dimethyl tert-butylsilyloxy)-3ꢀ-(di-
phenylphosphoryloxy)-20-methylpregna-5,16-diene (12).
To a stirred solution of 1.5 M LDA/THF (9.5 mL, 14.2
mmol) at ꢀ78 ꢁC was added a solution of Ph2P(O)CHF2
(3.55 g, 14.0 mmol) in THF (7 mL). After 1 h, the enone
11 (6.03 g, 14.1 mmol) in THF (45 mL) was added and
the reaction mixture was allowed to warm slowly to
room temperature. It was then heated at reflux for 2 h
and stirred overnight at room temperature. The reaction
was quenched with H2O and extracted with CH2Cl2.
Concentration of the organic layer and purification by
flash chromatography and crystallization (Et2O) gave
The last material to elute from the column was con-
centrated to give 382 mg (30.2%) of pregnenolone.
20-Methylpregna-5,20-dien-3ꢀ-ol (7). Mp 125–126 ꢁC
[CH2Cl2-hexane; reported mp 134–135 ꢁC (MeOH)11];
IR n 3400 cmꢀ1; MS (CI) m/e 315 (15%, M+1), 314
(27%, M+), 297 (100%, M+1–H2O); 1H NMR d
(CDCl3) 0.60 (3H, s), 1.02 (s), 1.77 (s, C21–Me), 3.28–
3.38 (1H, m, C3–H), 4.72+4.86 (2H, s+s, C21-CH2),
5.34–5.37 (1H, m, C6–H). Anal. calcd for C22H34O: C,
84.02; H, 10.90. Found: C, 83.71; H, 10.81.
12, (692 mg, 19.7%): IR n 1439, 1231, 1090, 982 cmꢀ1
;
MS (CI) m/e 681 (10%, M+1), 679 (21%, M-1), 549
(28%, M+1-TBDMS-OH), 219 (100%, Ph2PO2H+H);
1H NMR d (CDCl3) 0.07 (6H, s, 2ꢃSi-CH3), 0.89 (9H,
s, tert-Bu-CH3), 0.95 (s), 0.99 (s), 1.69 (d, C20–Me),
3.42–3.54 (1H, m, C3–H), 5.30–5.35 (1H, m, C6–H),
5.86–5.90 (1H, m, C16–H), 6.22 (1H, t, CHF2), 7.38–
7.55 (7H, m), 7.74–7.89 (4H, m);
F NMR d
21,21-Difluoro-20-methylpregna-5,16,20-trien-3ꢀ-ol (9)
and 3ꢀ-(diphenyl-phosphoryloxy)-pregn-5-en-20-one (10).
To a solution of Ph2P(O)CHF2 (2.05 g, 8.13 mmol) in
THF (8 mL) at ꢀ78 ꢁC was added a 1.5 M LDA/THF
solution (22.3 mL, 33.4 mmol). After 1 h, the steroid 8 (5.0
g, 15.9 mmol) in THF (50 mL) was added. The reaction
was stirred at ꢀ78ꢁC for 1.5 h, warmed to room tempera-
ture, then heated under reflux for 2 h. The reaction was
quenched with H2O, extracted into CH2Cl2 and the pro-
ducts separated by flash chromatography (SiO2, CH2Cl2–
2% MeOH). Product containing fractions were pooled,
concentrated and rechromatographed (hexane–35%
EtOAc) The least polar product fractions were combined,
concentrated and the residue crystallized (CH2Cl2–hexane)
to give 9 (54 mg, 0.95%); mp 106 ꢁC (EtOAc–hexane); UV
l 223 nm (lg "=3.769); MS (CI) m/e 349 (18%, M+1), 348
(30%, M+), 347 (28%, M-1), 331 (10%, M+1–H2O); 1H
NMR d (CDCl3) 0.91 (3H, s), 1.04 (3H, s), 1.79 (t,
C21ꢀMe), 3.47–3.60 (1H, m, C3–H), 5.34–5.41 (1H, m, C6–
H), 5.63–5.67 (1H, m, C16–H); 19F NMR d ꢀ88.58 (d,
J=45 Hz), ꢀ93.34 (dd, J=2.5, 46 Hz). Anal. calcd for
C22H30F2O: C, 75.83; H, 8.68. Found: C, 75.22; H, 9.09.
(CDCl3)ꢀ128.89 (dd, J=5, 129 Hz); P NMR Áꢀ80.01.
Anal. calcd for C40H55F2O3PS: C, 70.55; H, 8.14.
Found: C, 70.37; H, 8.18.
21,21-Difluoro-3ꢀ-(dimethyl-tbutylsilyloxy)-20-methyl-
pregna-5,17(20)-dien-16-one (14). Material recovered
from the crystalization filtrate of phosphoryl diene 12
was suspended in DMSO and heated at 160–170 ꢁC for 4
h. The solids dissolved and the solution gradually turned a
blood-red color. The cooled solution was poured into
H2O and extracted with Et2O. After drying and con-
centrating, the resulting residue was flash chromato-
graphed (SiO2, hexane–25% EtOAc). Product containing
fractions were pooled, concentrated and crystallized from
CH2Cl2–hexane to give 14 (102 mg) mp 139–140 ꢁC; IR n
3422, 3229, 1721, 1642 cmꢀ1; MS (CI) m/e 365 (67%,
M+1), 347 (55%, M+1-H2O), 345 (100%, M+1-HF),
1
327 (20%, M+1-HF-H2O); H NMR d (CDCl3) 1.04
(3H, s), 1.10+1.11 (3H, s+s, 2:1), 1.94+2.16 (s+s, C20–
Me, 2:1), 3.49–3.62 (1H, m, C3–H), 5.33–5.40 (1H, m, C6–
H), 6.63 (0.33H, t, CHF2), 7.54 (0.67H, d, CHF2); F NMR
d (CDCl3)ꢀ116.62 (dd, J=55, 322 Hz), ꢀ119.47 (dd,
J=55, 322 Hz), ꢀ120.26 (dd, J=56, 324 Hz), ꢀ122.08 (dd,
J=56, 324 Hz). Anal. calcd for C40H55F2O3PS: C, 70.55;
H, 8.14. Found: C, 70.37; H, 8.18.
Fractions of the most polar material from the column
were combined and concentrated to a white, crystalline

P. M. Weintraub et al. / Bioorg. Med. Chem. 11 (2003) 427–431
431
References and Notes
ran, T. P.; Johnston, J. O. US Patent 5,486,511 1996, Chem.
Abstr. 1995, 122, 214328.
1. De Coster, R.; Wouters, W.; Bruynseels, J. J. Steroid Bio-
chem. Molec. Biol. 1996, 56, 133.
2. Chung, B.-C.; Picado-Leonard, J.; Haniu, M.; Bienkowski,
M.; Hall, P. F.; Shively, J. F.; Miller, W. L. Proc. Nat. Acad.
Sci. U.S.A. 1987, 84, 407.
3. Bradshaw, K. D.; Waterman, M. R.; Couch, R. T.; Simp-
son, E. R.; Zuber, M. X. Molec. Endocr. 1987, 1, 348.
4. Njar, V. C. O.; Brodie, A. M. H. Current Pharm. Des. 1999,
5, 163.
9. Burkhart, J. P.; Weintraub, P. M.; Gates, C. A.; Resvick,
R. J.; Vaz, R. J.; Friedrich, D.; Angelastro, M. R.; Bey, P.;
Peet, N. P. Bioorg. Med. Chem. 2002, 10, 929.
10. Edwards, M. L.; Stemerick, D. M.; Jarvi, E. T.; Mat-
thews, D. P.; McCarthy, J. R. Tetrahedron Lett. 1990, 31,
5571.
11. Carlson, R. M. K.; Popov, S.; Massey, I.; Delseth, C.;
Ayanoglu, E.; Varkony, T. H.; Djerassi, C. Bioorg. Chem.
1978, 7, 453.
5. Angelastro, M. R.; Laughlin, M. E.; Schatzman, G. L.;
Bey, P.; Blohm, T. R. Biochem. Biophys. Res. Commun. 1989,
162, 1571.
6. Angelastro, M. R.; Marquart, A. L.; Weintraub, P. M.;
Gates, C. A.; Laughlin, M. E.; Blohm, T. R.; Peet, N. P.
Bioorg. Med. Chem. Lett. 1996, 6, 97.
12. Also finding similar results were Jarmon, M., Barrie, S. E.,
Llera, J.M. J. Med. Chem. 1998, 41, 5375.
13. Phillion, D. P.; Cleary, D. G. J. Org. Chem. 1992, 57,
2763.
14. Piettre, S. R.; Cabanas, L. Tetrahedron Lett. 1996, 37,
5881.
7. Burkhart, J. P.; Gates, C. A.; Laughlin, M. E.; Resvick,
R. J.; Peet, N. P. J. Bioorg. Med. Chem. 1996, 4, 1411.
8. Weintraub, P. M.; Gates, C. A.; Angelastro, M. R.; Cur-
15. Thermolysis of 15 in DMSO could not be performed due
to lackof material.
16. For experimental details, see ref 8.