Inhibitors of sterol synthesis. Synthesis and spectral properties of 3β-hydroxy-25,26,26,26,27,27,27-heptafluoro- 5α-cholestan-15-one

Article abstract of DOI:10.1016/S0009-3084(97)02656-X

3β-Hydroxy-25,26,26,26,27,27,27-heptafluor 5α-cholestan-15-one (4) has been prepared as part of a program to synthesize 15-ketosterols that are not readily metabolized to cholesterol or side-chain oxygenated species. Saponification of 3β -acetoxy-5α-chola-8(14),23-dien-15-one (5) followed by lithium-ammonia reduction with a bromobenzene quench gave 3β-hydroxy-5α-chol-23-en-15-one (6). Addition of (CF₃)₂CFI to 6 in the presence of triethylborane gave an iodide preparation, which was reduced to 4 with tributyltin hydride (71% overall yield of 4 from 5). The 23-iodide preparations consisted of 6:1 mixtures of (23R)-3β -hydroxy-23-iodo-25,26,26,26,27,27,27-heptafluoro 15-one (9a) and its C-23 epimer 9b with variable amounts of 4. Compound 4 was also prepared by lithium-ammonia reduction of the Δ⁸¹⁴ analogs of 4 and iodides 9a and 9b. The presence of small amounts of 6 in the latter product suggested a side reaction involving cleavage of the C24-C25 bond with loss of a (CF₃)₂ CF· radical. Also prepared were 25,26,26,26,27,27,27-heptafluoro-5α-cholestane-3β,15α-diol, its 15β epimer, the 7α-methyl analog of 4, 3β-hydroxy-7α-methyl-5α-cholestan-15-one (16), and (25R)-3β,26-dihydroxy-5α-cholestan-15-one. Full ¹H and ¹³C-NMR data of high precision with complete signal assignments are given for all new compounds. Definitive ¹H-NMR stereochemical assignments of the C-24 protons were established for most sterols with a C₈H₁₇ side chain based on analysis of the downfield H-24 resonance in a 750-MHz spectrum of 16. Detailed electron-impact mass spectral data are presented together with a summary of major fragmentation patterns for 15-hydroxy-and 15-ketosteroids with and without a Δ⁸¹⁴ bond.

Full text of DOI:10.1016/S0009-3084(97)02656-X

Chemistry and Physics of Lipids
86 (1997) 95–119
Inhibitors of sterol synthesis. Synthesis and spectral properties
of
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one
Abdul U. Siddiqui, Shankar Swaminathan, Xiangdong Su, William K. Wilson,
George J. Schroepfer Jr. *
Departments of Biochemistry and Cell Biology and of Chemistry, Rice Uni6ersity, 6100 Main St., Houston, TX 77005, USA
Received 16 August 1996; accepted 19 December 1996
Abstract
3i-Hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one (4) has been prepared as part of a program to
synthesize 15-ketosterols that are not readily metabolized to cholesterol or side-chain oxygenated species. Saponifica-
tion of 3i-acetoxy-5h-chola-8(14),23-dien-15-one (5) followed by lithium-ammonia reduction with a bromobenzene
quench gave 3i-hydroxy-5h-chol-23-en-15-one (6). Addition of (CF3)2CFI to 6 in the presence of triethylborane gave
an iodide preparation, which was reduced to 4 with tributyltin hydride (71% overall yield of 4 from 5). The 23-iodide
preparations consisted of 6:1 mixtures of (23R)-3i-hydroxy-23-iodo-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-
15-one (9a) and its C-23 epimer 9b with variable amounts of 4. Compound 4 was also prepared by lithium-ammonia
reduction of the D8(14) analogs of 4 and iodides 9a and 9b. The presence of small amounts of 6 in the latter product
suggested a side reaction involving cleavage of the C24–C25 bond with loss of a (CF3)2CF“ radical. Also prepared
were 25,26,26,26,27,27,27-heptafluoro-5h-cholestane-3i,15h-diol, its 15i epimer, the 7h-methyl analog of 4, 3i-hy-
1
droxy-7h-methyl-5h-cholestan-15-one (16), and (25R)-3i,26-dihydroxy-5h-cholestan-15-one. Full H and 13C-NMR
data of high precision with complete signal assignments are given for all new compounds. Definitive 1H-NMR
stereochemical assignments of the C-24 protons were established for most sterols with a C8H17 side chain based on
analysis of the downfield H-24 resonance in a 750-MHz spectrum of 16. Detailed electron-impact mass spectral data
are presented together with a summary of major fragmentation patterns for 15-hydroxy- and 15-ketosteroids with and
without a D8(14) bond. © 1997 Elsevier Science Ireland Ltd.
Keywords: 15-ketosterols; NMR; Mass spectrometry; Liquid-ammonia reduction
* Corresponding author.
0009-3084/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved.
PII S0009-3084(97)02656-X

Chemistry and Physics of Lipids
86 (1997) 95–119
Inhibitors of sterol synthesis. Synthesis and spectral properties
of
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one
Abdul U. Siddiqui, Shankar Swaminathan, Xiangdong Su, William K. Wilson,
George J. Schroepfer Jr. *
Departments of Biochemistry and Cell Biology and of Chemistry, Rice Uni6ersity, 6100 Main St., Houston, TX 77005, USA
Received 16 August 1996; accepted 19 December 1996
Abstract
3i-Hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one (4) has been prepared as part of a program to
synthesize 15-ketosterols that are not readily metabolized to cholesterol or side-chain oxygenated species. Saponifica-
tion of 3i-acetoxy-5h-chola-8(14),23-dien-15-one (5) followed by lithium-ammonia reduction with a bromobenzene
quench gave 3i-hydroxy-5h-chol-23-en-15-one (6). Addition of (CF3)2CFI to 6 in the presence of triethylborane gave
an iodide preparation, which was reduced to 4 with tributyltin hydride (71% overall yield of 4 from 5). The 23-iodide
preparations consisted of 6:1 mixtures of (23R)-3i-hydroxy-23-iodo-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-
15-one (9a) and its C-23 epimer 9b with variable amounts of 4. Compound 4 was also prepared by lithium-ammonia
reduction of the D8(14) analogs of 4 and iodides 9a and 9b. The presence of small amounts of 6 in the latter product
suggested a side reaction involving cleavage of the C24–C25 bond with loss of a (CF3)2CF“ radical. Also prepared
were 25,26,26,26,27,27,27-heptafluoro-5h-cholestane-3i,15h-diol, its 15i epimer, the 7h-methyl analog of 4, 3i-hy-
1
droxy-7h-methyl-5h-cholestan-15-one (16), and (25R)-3i,26-dihydroxy-5h-cholestan-15-one. Full H and 13C-NMR
data of high precision with complete signal assignments are given for all new compounds. Definitive 1H-NMR
stereochemical assignments of the C-24 protons were established for most sterols with a C8H17 side chain based on
analysis of the downfield H-24 resonance in a 750-MHz spectrum of 16. Detailed electron-impact mass spectral data
are presented together with a summary of major fragmentation patterns for 15-hydroxy- and 15-ketosteroids with and
without a D8(14) bond. © 1997 Elsevier Science Ireland Ltd.
Keywords: 15-ketosterols; NMR; Mass spectrometry; Liquid-ammonia reduction
* Corresponding author.
0009-3084/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved.
PII S0009-3084(97)02656-X

96
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
The goal of the present study was to prepare
1. Introduction
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-
cholestan-15-one (4), a 15-ketosterol in which
neither conversion to cholesterol nor side chain
oxidation at C-26 (or C-25) is anticipated. A
preliminary account of a portion of this work
has been reported previously [26]. Structures of
15-ketosterol 1, its saturated analog 2, and their
25,26,26,26,27,27,27-heptafluoro analogs 3 and 4
are shown in Fig. 1.
3i-Hydroxy-5h-cholest-8(14)-en-15-one (1) is a
potent inhibitor of sterol synthesis in cultured
mammalian cells and lowers the levels of 3-hy-
droxy-3-methylglutaryl coenzyme A reductase ac-
tivity [1–6]. 1 not only suppresses the synthesis
of cholesterol but it also is a potent inhibitor of
the in vivo intestinal absorption of cholesterol
[7,8]. 1 serves as an alternate substrate for acyl
coenzyme A-cholesterol acyltransferase (ACAT)
and inhibits the oleoyl-CoA-dependent esterifica-
tion of cholesterol by hepatic and jejunal micro-
somes [9]. In addition, dietary administration of
1 to rats is associated with a lowering of the
levels of ACAT activity in jejunum [10]. Oral
administration of 1 causes a reduction of serum
cholesterol levels in rodents [11] and non-human
primates [12,13]. The lowering of total serum
cholesterol levels in non-human primates is asso-
ciated not only with a reduction of the levels of
low-density lipoprotein plus very low-density
lipoprotein cholesterol but also an elevation of
the levels of high-density lipoprotein cholesterol
[12,13]. This unique combination of actions, gen-
erally considered favorable for potential use in
the treatment and/or prevention of atherosclero-
sis, forms the basis for our continued interest in
1 and its analogs.
2. Experimental procedures and results
2.1. Materials and methods
Triethylborane, tributyltin hydride, silica gel,
and bis(trimethylsilyl)trifluoroacetamide (BSTFA)
were purchased from Aldrich Chemical Company
(Milwaukee, WI). 2-Iodoheptafluoropropane was
obtained from Strem Chemicals, Inc. (Newbury-
port, MA) or PCR, Inc. (Gainesville, FL). 2,2%-
Azobisisobutyronitrile (AIBN) was obtained from
Janssen Chimica (San Diego, CA). Syntheses of
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-
cholest-8(14)-en-15-one (3) [24], its 7h-methyl
derivative 15 [27], 3i-hydroxy-7h-methyl-5h-
cholest-8(14)-en-15-one (14) [28], (25R)-3i,26-di-
hydroxy-5h-cholest-8(14)-en-15-one (18) [29,30],
and 3i-acetoxy-5h-chola-8(14),23-dien-15-one (5)
[24] have been described previously.
Detailed studies of the in vitro and in vivo
metabolism of 1 have demonstrated its conver-
sion to cholesterol and to polar metabolites, with
the latter process apparently initiated by oxida-
tion at C-26 [14–23]. In an attempt to obtain
more potent analogs of 1, we have prepared
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-
cholest-8(14)-en-15-one (2), a 15-ketosterol in
which side-chain oxidation was blocked by the
fluorine substitution [24]. Although the F7-D8(14)
-
15-ketosterol 2 showed favorable actions upon
administration to rats, significant metabolism
of 2 to 25,26,26,26,27,27,27-heptafluorocholes-
terol was observed [25]. We therefore sought
the preparation of analogs of 1 in which both
its side chain conversion to form polar meta-
bolites and its conversion to cholesterol was
blocked.
Fig. 1. Structures of 3i-hydroxy-5h-cholest-8(14)-en-15-one
(1), 3i-hydroxy-5h-cholestan-15-one (2), and their 25,26,26,
26,27,27,27-heptafluoro derivatives 3 and 4.

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
97
steroids), H2O, trimethylsilanol, HI, or combina-
tions thereof are marked by an asterisk. Ions
showing exact masses within 93.0 millimass
units of the indicated or implied assignments are
marked by †. Exact masses from the ZAB in-
strument are reported as the average of ꢀ5
scans; standard deviations were typically 1–1.5
millimass units. Ions A, B, C, D, E, F, G, H,
and I, M−Ring A, and M−C1...C3 (loss of
C-1, C-2, and C-3 and attached atoms) are
defined in Fig. 2, and the same designations are
used for the corresponding ions from 3i-
trimethylsilyl (TMS) ether and 7h-methyl deriva-
tives. TMS ether derivatives were prepared by
treating the steroids with a 1:1 mixture of
BSTFA and pyridine for 1 h under nitrogen at
room temperature, followed by evaporation to
dryness at 40°C under nitrogen. Samples for gas
chromatography (GC) were injected in hexane (1
vl). GC-MS analyses were carried out isother-
mally on a 60-m DB-5 column (60 m×0.25 mm
i.d.; 5% phenyl-95% methyl polysiloxane; 0.1
vm film thickness; J&W Scientific Inc., Folsom,
CA) with split injection (ZAB instrument) or on
a 15-m DB-5 column with falling needle injec-
tion (Extrel ELQ-400 instrument). GC analyses
were done on a Shimadzu 9A (splitless injection,
30-m DB-5 column) or a Perkin-Elmer Sigma
2000 (split injection, 60-m DB-5 column).
Melting points (m.p.) were measured with a
Thomas-Hoover apparatus in sealed, evacuated
capillary tubes. Infrared (IR) spectra were ob-
tained on a Mattson Galaxy 6020 Fourier-trans-
form spectrometer; samples were analyzed as
KBr pellets or thin films obtained by evapora-
tion of CH2Cl2 solutions onto NaCl plates.
Thin-layer chromatography (TLC) was carried
out on silica gel G plates (Analtech, Newark,
DE) or aluminum-backed silica gel 60 plates
(EM Separations, Gibbstown, NJ). Components
of the plates were visualized after spraying with
5% ammonium molybdate(VI) in 10% sulfuric
acid followed by heating. Solvent systems for
TLC were ethyl acetate-hexane 4:6 (SS-1), ether-
benzene 1:1 (SS-2), or ethyl acetate-hexane 7:3
(SS-3). High performance liquid chromatography
(HPLC) was carried out isocratically with a Wa-
ters 600 pump operating at 1 ml/min, UV detec-
tion at 210 nm, and a 5-vm Customsil ODS
reversed phase column (250 mm×4.6 mm i.d.;
Custom LC, Houston, TX). Unless otherwise
specified, column chromatography was per-
formed on silica gel (70–230 mesh), and
medium
pressure
liquid
chromatography
(MPLC) was done in a glass column dry-packed
with silica gel (230–400 mesh) using a 460
mm×10 mm i.d. column (20 g silica gel), 960
mm×10 mm i.d. column (40 g silica gel), a 460
mm×25 mm i.d. column (ꢀ120 g silica gel), or
a Lobar column (EM Separations, Gibbstown,
NJ). The preparation of silica gel impregnated
with AgNO3 and the adsorption of samples
onto silica gel have been described previously
[31]. Fraction volumes were 20 ml unless noted
otherwise.
Low-resolution and high-resolution mass spec-
tra (MS) were recorded on a VG ZAB-HF dou-
ble sector instrument with an electron energy of
70 eV and direct inlet sample introduction, ex-
cept for high-resolution MS of 4 and 6, which
were recorded on a Kratos MS-50DA spectrom-
eter at the Midwest Center for Mass Spectrome-
try (Lincoln, NE). Low-resolution mass spectral
data are given as m/z (relative intensity, sug-
gested assignment). Ions attributable to losses of
CH3, side chain (SC), SC+2H (for D23
Nuclear magnetic resonance (NMR) spectra
were measured as described previously [24,32] on
a Bruker AMX500 (500.1 MHz, 25°C for 1H,
22°C for 13C), an IBM AF300 (75.5 MHz for 13C,
ca. 22°C), or a Bruker AC250 (235.4 MHz for
19F, ca. 22°C) spectrometer in CDCl3 solution and
referenced to internal tetramethylsilane (1H),
CDCl3 at 77.0 ppm (13C), or internal CFCl3 (19F).
Coupling constants were derived from line spac-
ings of resolution-enhanced one-dimensional (1D)
spectra or, at lower precision, from f1 or f2 cross
1
sections of 2D spectra. Also, H-NMR spectra of
16 and 19 were obtained on a Varian Unity-Plus
spectrometer (750.2 MHz) at 26°C in 1 mM
CDCl3 solution. PC Model (Serena Software,
Bloomington, IN) was used for modeling of
steroids by molecular mechanics and to predict
1H-NMR vicinal coupling constants. Spin simula-

98
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
Fig. 2. Summary of suggested major mass spectral fragmentation patterns for 15-hydroxy- and 15-ketosteroids with and without a
D8(14) bond. Fragmentation of 3i-acetoxy, 3i-TMS ether, and 7h-methyl derivatives is usually analogous, and the same
designations (ions A–I) are used for the corresponding ions from these derivatives.
tion was done with gNMR 3.6 (Cherwell Scientific
Publishing, Oxford, UK) with compensation for
comparisons [24,33] in conjunction with 2D spec-
tral results by strategies described previously [32].
In the case of 12a, the 13C-NMR assignments for
C-8 and C-14 were confirmed by an HMBC (het-
1
minor chemical shift inaccuracies. H- and 13C-
NMR assignments were made from chemical shift

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
99
eronuclear multiple bond correlation) experiment
showing a cross peak connecting the H-18 signal
at l 1.03 to a 13C signal at l 140. 19F-NMR
spectra were interpreted as described previously
for other F7 sterols [24]. The purity of steroid
samples was determined by HPLC, TLC, GC, and
1H-NMR (500 MHz, after sufficient magnification
of the vertical scale in the methyl and l 2–6
spectral regions to detect a 1% impurity).
gave 5h-chol-23-ene-3i,15i-diol (7b), fractions
41–42 gave 5h-chol-23-ene-3i,15h-diol (7a), and
fractions 48–50 gave 5h-chol-8(14),23-ene-
3i,15i-diol (8). Characterization of diols (1H-
and 13C-NMR chemical shifts are given in the
order used in Table 1 and Table 3 with designa-
tions of precision described in footnotes therein):
1
7a (ꢀ97% purity by H-NMR), GC-MS (ZAB
instrument, 250°C) of the bis-TMS ether (single
component at tR 31.3 min), 504 (9, M+), 489*
(16), 435* (6), 433* (31), 414* (17), 399* (8), 324*
(5), 309* (12), 255* (8), 197 (100); MS, 360.3022†
(1, M+), 342*† (5), 319† (6, ion D), 301† (23, ion
D−H2O), 289*† (8), 283† (4, ion D−2H2O), 263†
(11, ion F2), 235† (13, ion G3), 217† (100, ion
G3−H2O), 107 (27); 1H-NMR, l 0.978, 1.707,
1.802, 1.396, 3.597, 1.570, 1.292, 1.121, 1.326†,
1.284†, 1.125, 1.898, 1.574, 0.679, 1.509, 1.275†,
1.201, 1.904, 1.043, 3.959, 1.743, 1.914, 1.410†,
0.697, 0.820, 1.441†, 0.908, 1.792, 2.140, 5.757,
4.990†, 4.992†; 1H-1H-NMR coupling constants,
Table 2; 13C-NMR, l 36.93, 31.28, 71.03, 37.98,
44.53, 28.49, 32.25, 34.96, 54.11, 35.37, 21.00,
39.95, 43.89, 63.43, 73.68, 40.39, 53.16, 13.32,
12.30, 35.20, 18.35, 40.42, 136.91, 115.90; 7b (ꢀ
2.2. 3i-Hydroxy-5h-chol-23-en-15-one (6) by
lithium-ammonia reduction of 5 (tert-butanol
quench)
3i-Acetoxy-5h-chola-8(14),23-dien-15-one (5;
504 mg) was dissolved in ether and cooled in a
dry ice-acetone bath. Dry liquid ammonia (ꢀ100
ml) was condensed in the flask followed by the
addition, in one portion, of lithium (260 mg).
After stirring for 15 min, tert-butanol (20 ml) was
added and the resulting mixture was cautiously
poured onto ice. After extraction with ethyl ac-
etate, the organic phase was evaporated to dry-
ness and subjected to chromatography on silica
gel (150 mm×10 mm i.d. column; elution with
ethyl acetate-hexane 1:9 (250 ml) and ethyl ac-
etate-hexane 15:85 (250 ml)). Evaporation of frac-
tions 14–22 gave 6 (180 mg; 40% yield): m.p.
153–154°C; single component on TLC in SS-1 (Rf
0.47) and SS-2 (Rf 0.46); IR, wmax 3350, 2926,
2857, 1734, 1640, 1449, 1383, 1132, 1076, 1045,
995, 910 cm−1; GC-MS (ZAB instrument, 250°C)
of the TMS ether (tR 37.7 min), 430 (44, M+),
415* (100), 340* (12), 333 (9, ion F1), 325* (5),
217 (7, ion G3−TMSOH), 216 (7, ion G2−
TMSOH); MS, 358.2862 (82, M+), 343*† (5),
340*† (2), 325*† (4), 317† (11, ion D), 299† (7, ion
D−H2O), 289*† (3), 261† (100, ion F1), 234† (32,
ion G2), 233† (13, ion G1), 219† (15, C15H23O),
217† (72, ion G3−H2O), 216† (41, ion G2−H2O),
215† (14, ion G1−H2O), 201† (18, C15H21), 190†
(20, C14H22), 165† (31, ion E), 108† (90, C8H12),
107† (89, C8H11); 1H-NMR, Table 1; 1H-1H-NMR
coupling constants, Table 2; 13C-NMR, Table 3.
Diol byproducts were eluted from the above
column with ethyl acetate-hexane 3:7 (500 ml),
ethyl acetate-hexane 4:6 (500 ml), and ethyl ac-
etate-hexane 1:1. Evaporation of fractions 31–32
1
90% purity by H-NMR; contains 5% 10), GC-
MS (ZAB instrument, 250°C) of the TMS
derivatives (tR 27.7 min for the bis-TMS ether, tR
37.6 min for 3i-TMS ether), 504 (14, M+), 489*
(31), 435* (8), 433* (37), 414* (52), 407 (20), 399*
(18), 372 (10), 346 (11), 324* (11), 309* (20), 283
(10, ion D−2TMSOH), 255 (12), 197 (100), 157
(57) and 432 (3, M+), 417* (35), 414* (23), 399*
(13), 373 (11, ion D−H2O), 346 (11), 283 (18, ion
D−H2O−TMSOH), 255* (12), 217 (100); MS,
360.3011† (1, M+), 342*† (4), 319† (14, ion D),
301† (32, ion D−H2O), 289*† (4), 283† (14, ion
D−2H2O), 263† (3, ion F2), 235† (9, ion G3), 217†
1
(100, ion G3−H2O), 107 (31); H-NMR, l 0.981,
1.719, 1.812, 1.415, 3.596, 1.571, 1.301, 1.133,
1.33†, 1.33†, 1.009, 1.905, 1.725, 0.695, 1.500,
1.303, 1.088, 1.920, 0.868, 4.193, 2.416, 1.348,
1.057, 0.951, 0.844, 1.623, 0.932, 1.818, 2.176,
1
5.767, 4.988†, 4.992†; H-1H-NMR coupling con-
stants, Table 2; 13C-NMR, l 36.96, 31.50, 71.30,
38.13, 44.94, 28.54, 31.48, 31.40, 54.74, 35.59,
21.09, 41.27, 42.24, 60.90, 70.43, 40.92, 55.91,
14.78, 12.27, 35.39, 18.65, 40.50, 137.03, 115.87; 8,

100
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
Table 1
1H-NMR chemical shifts for 3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one (4) and other 15-oxygenated steroids
described hereina–c
D8,23
D8(14),23 D23
D0
23-I
9a
D8(14)
23-I
12a
D0
D0
15h-OH
13a
D0
15i-OH
13b
D0
7h-Me
16
D0
7h-Me
17
D0
26-OH
19
11
10
6
4
H-1h
H-1i
H-2h
H-2i
H-3h
H-4h
H-4i
H-5h
H-6h
H-6i
H-7h
H-7i
H-8i
H-9h
H-11h
H-11i
H-12h
H-12i
H-14h
H-15
H-16h
H-16i
H-17h
H-18
1.142
1.807
1.862
1.465
3.630†
1.665
1.34
1.200
1.725
1.848
1.371
3.646
1.681
1.272
1.419
1.475
1.347
1.579
4.136
0.977
1.699
1.810
1.404
3.586
1.566
1.281
1.115
1.30†
1.30†
0.837
2.662
1.707†
0.635
1.555
1.271
1.353
2.100
1.670†
0.983
1.702
1.813
1.409
3.588
1.570
1.286
1.120†
1.30†
1.30†
0.837
2.656
1.721
0.640
1.572
1.283
1.365
2.129
1.666†
1.203
1.728
1.853
1.375
3.648
1.685
1.276
0.980
1.702
1.813
1.407
3.589
1.568
1.282
1.117
1.30†
1.30†
0.841
2.660
1.709†
0.642
1.566
1.277
1.368
2.113
1.678
0.980
1.708
1.804
1.397
3.597
1.572
1.293
1.121
1.29*
1.31*
1.129
1.894
1.575
0.684
1.518
1.282†
1.213†
1.911
1.051
3.957
1.698
1.877
1.409
0.693
0.822
1.384†
0.920
1.373
1.092
1.430
1.635
1.968†
2.032†
0.983
1.721
1.814
1.417
3.599
1.575
1.303
1.134
1.31*
1.33*
1.013
1.898
1.726
0.701
1.509
1.308
1.100
1.928
0.875
4.199
2.362
1.319
1.047
0.948
0.846
1.560
0.940
1.417
1.108
1.452
1.623
1.971†
2.051†
0.997
1.704
1.805
1.390
3.607
1.474
1.260
1.360
1.082
1.578
0.999
1.703
1.808
1.392
3.609
1.478
1.262
1.363
1.087
1.579
0.979
1.701
1.810
1.405
3.587
1.567
1.281
1.116†
1.29***
1.31***
0.836
2.662
1.707
0.635
1.556
1.272
1.356
2.113
1.664
1.34
1.476
1.365†
2.161
1.880
1.486
1.355
1.581
4.122
2.797
1.903
0.935
1.559
1.28
1.30
2.078
1.834
2.796
1.909
0.941
1.572
1.28
1.32
2.078
1.844
1.852
1.647
1.538
1.246
2.093
1.863
1.666
1.553
1.261
2.123
2.07
2.04
1.918
1.293
2.311
2.442
1.99
2.03
0.986
0.950
1.638
1.010
1.794
2.159
5.751
2.398
2.071
1.471
0.986
0.717
1.682
1.022
1.876
2.177
5.762
2.453
1.793
1.57
0.751
0.811
1.60
1.005
1.838
2.139
5.760
2.408
1.871
1.668
0.796
0.819
1.751
0.974
1.311
1.831
2.331
2.133
1.575
1.033
0.725
1.835
0.983
1.286
1.888
2.410
1.764
1.566
0.750
0.813
1.520
1.013
1.367
1.134
1.464
1.654
1.976†
2.051†
2.425
1.779
1.559
0.752
0.812
1.481
0.982
1.287
1.038
1.349
1.177
1.095†
1.140
2.411
1.772
1.572
0.762
0.814
1.528
1.013
1.371
1.136
1.466
1.657
1.979†
2.050†
2.421
1.769
1.553
0.741
0.811
1.481
0.986
H-19
H-20
H-21
H-22R
H-22S
H-23R
H-23S
H-24R
H-24S
1.302†
1.072†
1.28**
1.34**
1.094*†
1.327*†
4.326
2.822
3.028
4.337
2.827†
3.036
5.018†
5.007†
5.016†
5.016†
5.022†
5.009†
aData obtained at 500 MHz in 10–30 mM CDCl3 solution at 25°C and referenced to the Si(CH3)4 signal.
bChemical shifts given to two (three) decimal places are generally accurate to 90.01 (90.001) ppm except that values marked by
† are accurate to about 90.02 (90.003) ppm. R and S denote pro-R and pro-S hydrogens, or, in the case of the C-24 protons of
D23 steroids, the pro-E and pro-Z hydrogens. Stereochemical assignments marked with an asterisk may be interchanged.
cAdditional signals: 16, H-25, 1.514, H-26, 0.864, H-27, 0.868, 7h-CH3, 0.838; 17, 7h-CH3, 0.841; 19, H-25, 1.604, H-26, 3.428 and
3.493, H-27, 0.912.
1H-NMR (100 mM solution), l 1.127, 1.712,
1.833†, 1.370, 3.629, 1.655†, 1.28, 1.30, 1.44, 1.26,
1.793†, 2.729, 1.673†, 1.607, 1.504, 1.055†, 1.907,
4.632, 2.380, 1.440, 1.080, 1.037, 0.719, 1.643†,
NMR coupling constants, Table 2; 13C-NMR, l
36.45, 31.36, 71.08, 38.01, 44.26, 29.29, 28.98,
133.63, 49.25, 37.12, 19.77, 37.31, 42.13, 145.34,
69.91, 38.11, 54.28, 19.36, 12.73, 34.46, 18.95,
40.26, 136.84, 115.99.
1
0.943, 1.865†, 2.208, 5.763, 4.992†, 5.000†; H-1H-

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
101
Table 2
1H-1H NMR coupling constants for 3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholestan-15-one (4) and other 15-oxygenated
steroids described hereina
7a
7b
8
11
10
6
9a
12a
4
13a
13b
16
17
19
1h-1i
1h-2h
1h-2i
1i-2h
1i-2i
13.2
3.9
13.6
3.2†
4.0
13.3
13.3
13.2
3.9†
13.7
13.2†
3.7
13.2
4.0†
13.5
3.7†
3.7†
13.1†
4.0†
13.7
13.1
3.7
13.0†
4.0
13.1
3.9
13.3† 13.2
13.1
3.9†
13.2†
4.3
13.7
3.4†
4.0
3.9
3.9
13.7
3.4†
4.0
13.8
3.4†
3.8
13.8†
3.6†
3.6†
13.8† 13.6
3.5†
13.9†
3.6†
3.6†
13.5† 13.8
3.4
3.8
3.5†
4.1
3.4†
4.0
4.0
4.0
3.5†
4.0
2h-2i
2h-3h
2i-3h
12.9
4.8†
11.3
12.8
4.7
11.3
12.6
4.9†
11.3
12.8
13.0
5.1†
11.3
12.8
4.8†
11.3
12.9
4.6†
11.3
12.7†
12.9
4.7†
12.8
4.9†
11.3
12.9
4.9†
11.3
4.7
11.2
4.7†
11.1
4.9†
4.8†
11.2
11.0
11.1† 11.3
3h-4h
3h-4i
4.8†
11.0
5.0
11.0
4.9†
4.7
10.9
5.1†
10.9
4.8†
11.0
4.7†
11.1
4.6†
10.9
4.7†
11.0
4.9
11.0
5.0
10.8
4.9
10.8
4.9†
10.9
10.9†
4h-4i
4h-5h
4i-5h
12.5
12.2
2.8
12.2†
3.0†
12.5
12.4†
12.2† 12.5†
2.9†
12.6
12.5
3.0
12.1
2.5†
12.0
2.7†
12.5†
12.5
12.5†
"
"
"
"
12.2†
12.5
12.4†
12.1† 12.5†
12.6
12.5† 12.4
12.5†
12.5†
5h-6h
5h-6i
3.0†
2.9†
2.8†
13.0
2.7†
12.1†2
12.0†
12.8†
6h-6i
6h-7h
6h-7i
6i-7h
6i-7i
2
5.1
"
12.7†
7.0
2.2
13.0
4.2
"
"
12.7‡
6.0†
2.1
"
"
"
13.5
2.3
13.5
2.3
"
5.5‡
3.5‡
12.3‡
3.5‡
3.4†
3.4†
2.2†
3.3†
3.3†
3.4†
3.4†
3.4†
3.5†
3.5†
3.4‡
12.3†
13.4†
4.4†
4.0†
13.9
3.4†
13.1
5.0
5.0†
3.4‡
13.2†
7h-7i
7h-8i
7i-8i
12.3†
11.8†
12.8
11.4†
3.7
14.3
13.1
10.6†
3.8†
14.3
13.1
10.8†
3.4†
12.6†
11.7‡
4.0
12.9
11.4†
3.8
3.4†
4.8
4.7
3.4‡
8i-9h
8i-14
10.6
10.9
10.5
11.2
10.2†
10.3†
10.2†
10.4†
10.5
11.2
11.0† 11.1†
11.32 10.92 10.52
10.1†
10.6†2 10.7†2
10.4†2 10.9
9h-11h
9h-11i
4.1
12.1
4.1
12.5
7.1
10.5
4.0†
4.0†
7.1
4.0†
4.3†
4.2
12.5
4.4†
3.9†
10.4
12.2†
12.1†
10.7† 12.3†
12.2†
11.6†
12.1†
11h-11i
11h-12h
11h-12i
11i-12h
11i-12i
13.8
3.5†
3.5†
13.8†
"
3.2
11.0
5.0
14.0
3.5
3.5
14.0
3.7
14.0
3.5†
3.5†
13.3†
13.1‡
3.9†
13.8†
4.1†
3.2
4.1
3.5
13.9†
3.5
3.2†
2
3.0†
3.6†
3.2†
3.3†
3.4†
2
12.8†2
14.1† 13.3†
13.1‡2 13.5‡
"
"
3.5†
6.1
3.2†
3.5†
12.4
3.5
3.3†
3.9
3.4†
12h-12i
12.8†
9.0
12.4
5.6
12.6
13.6
12.7
12.2
12.2
12.3
12.3
9.0
12.5
5.5
12.5
14-15
15-16h
15-16i
3.4
9.4
7.8
2.2
8.3
6.0
3.3
9.3
7.7
2.1
16h-16i
16h-17h
16i-17h
14.2
8.6
14.5
8.7
10.3
13.4
7.2
12.5
18.4
7.8
12.4
18.8
8.5
9.4
18.7
8.8
9.5
18.3
7.9
12.4
18.7
8.5
9.4
14.3
8.4
14.4
8.8
18.9
8.7
9.7
18.7
8.4
9.5
18.8
8.6
9.6
9.9†
2
9.9†
10.3†
17h-20
2
10.2
9.7
8.8‡ 10.0
2
9.6†
9.7†2 10.0†" 10.3† 10.3†
2
10.1
20-21
20-22R
20-22S
6.2†
7.8†
3.0
6.6
8.4
3.4
6.7
6.7
8.8
3.1†
6.7
8.3†
3.4
6.2
7.9†
3.0
6.4
6.6
6.2
8.3
6.2†
2
6.7
6.5
2.5†
8.7
6.2†
6.4
2.5
8.6†
9.2†
3.2†
3.4
2.7†
2.2†
8.4
22-22
13.8
13.8
13.8
13.8
13.9
13.9
14.8†
14.5† 13.6
13.0
13.3

102
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
Table 2 (continued)
7a
7b
8
11
8.3
10
8.3†
6
9a
12a
4
13a
13b
16
17
5.6†
19
22R-23R
22R-23S
22S-23R
22S-23S
23-23
8.3
6.1
8.1
6.1
8.5
6.0
2
5.6†
4.4
6.1
6.0
6.0
12.6†
11.9† 10.8
4.7
10.8
4.7
10.6†
12.8
5.2
23R-24R 10.5
23R-24S 16.7
23S-24R
10.5
16.7
10.5
16.7
10.4
16.7
10‡
9.4
4.1
10.8†
5.0†
10.7†
5.2†
4.0†
4.6†
9.9
4.5‡
10.2
5.3
4.6†
23S-24S
24-24
11.2†
15.0‡2
11.3‡
2
11.0†
2.4†"
"
"
2.0†2
"
1.5†
15.9†
16.0
2
13.12 16.0‡2
aData obtained at 500 MHz in 10–150 mM CDCl3 solution at 25°C. Estimated accuracy is ca. 90.2 Hz except for couplings
marked by † (90.5 Hz) or ‡ (91 Hz). Pairs of protons strongly coupled at 500 MHz are indicated by " (Dw/J B1.5) or 2
(1.55Dw/J53). R and S denote pro-R and pro-S hydrogens, or, in the case of the C-24 protons of D23 steroids, the pro-E and pro-Z
hydrogens, respectively. Most compounds also showed J1h-19 of ꢀ0.6 Hz, J2h-4h of ꢀ2.1 Hz, and various allylic couplings (0.5–2
Hz). 14h-15-Ketosteroids showed J14h-16h of ꢀ1.1 Hz. Iodides 9a, and 12a showed JH24R-F25, ꢀ13.5 Hz, JH24S-F25, ꢀ29 Hz, and
other F7 sterols showed JH24R-F25, 19.5† Hz, JH24S-F25, 21.2† Hz. 26-Hydroxysterol 19 showed J25-26, 6.4 and 7.0 Hz, J26-26, 10.4 Hz,
J25-27, 6.7 Hz. 7h-Methyl sterols 16 and 17 showed J7i-CH3 of 7.2 Hz. Additional side chain couplings for 16: J24R-25, 6.4 Hz; J24S-25
7.0 Hz; J25-26, 6.6 Hz; J25-27, 6.7 Hz.
,
2.3. 3i-Hydroxy-5h-chola-8(14),23-dien-15-one
(10)
Evaporation of fractions 132–148 gave 3i-hy-
droxy-5h,14i-chola-8,23-dien-15-one (11; 178
mg): gum; IR (thin film) wmax 3400, 3075, 2928,
2861, 1730, 1640, 1038, 737 cm−1; high resolution
MS, calcd. for C24H36O2, 356.2715, found
356.2729; MS, 356† (100, M+), 341*† (73), 323*†
To a solution of acetate 5 (5 g, 12.5 mmol) in
degassedmethanol(50ml)anddegassedtetrahydro-
furan (25 ml) was added potassium carbonate (1.72
g, 12.5 mmol). The reaction mixture was stirred
underargon for3h at roomtemperatureandpoured
into water. After extraction with ethyl acetate
(3×150 ml), the combined organic layers were
washed with water (3×200 ml), dried over sodium
sulfate, and evaporated to a white solid (4.5 g) that
was subjected to MPLC (100 g of silica gel, 70–230
mesh; elution with ethyl acetate-hexane 15:85).
Evaporation of fractions 154–266 gave 10 (4.28 g;
96% yield): m.p. 135–136°C; single component on
TLC in SS-1 (Rf 0.37) and SS-2 (Rf 0.35); IR wmax
3400, 3077, 2922, 2855, 1697, 1688, 1632, 1613, 1229,
1127, 1090, 1049, 907, 665 cm−1; GC-MS (ZAB
instrument, 290°C) of the TMS ether (tR 12.8 min),
428 (100, M+), 426 (3), 413* (18), 410* (4), 395*
(6), 341* (5), 338* (14), 336 (9), 323* (46), 305* (15),
269* (14), 251* (20), 243 (10); high resolution MS,
calculated for C24H36O2, 356.2715, found 356.2710;
MS, 356† (100, M+), 341*† (16), 338*† (8), 323*†
(43),306(5),297† (6,ionD−H2O),287*† (13),269*†
(24), 259† (8, ion C), 251*† (10), 121 (32), 107 (37);
1H and 13C-NMR and 1H-1H-NMR coupling con-
stants, Table 1, Table 2 and Table 3.
1
(62), 297 (8, ion D−H2O), 269* (5), 105 (49); H
and 13C-NMR and 1H-1H-NMR coupling con-
stants, Table 1 Table 2 and Table 3. A similar
reaction of 5 (9.75 g) gave 10 (7.93 g; 91% yield)
and 11 (521 mg).
2.4. 3i-Hydroxy-5h-chol-23-en-15-one (6) by
lithium-ammonia reduction of 10 (bromobenzene
quench)
Lithium metal cut into pieces ꢀ3–5 mm on a
side (150 mg; 22 mmol) was added to undistilled
anhydrous ammonia (100 ml) cooled in a dry
ice-acetone bath. When all the lithium had dis-
solved (ꢀ3–5 min), a cold solution of D8(14),23-15-
ketosteroid 10 (1 g; 2.8 mmol) in anhydrous ether
(50 ml) was added rapidly to the blue solution,
and the mixture was stirred for 5 min. The reac-
tion was quenched by addition of bromobenzene
(1 ml) in one portion, poured onto ice, and ex-
tracted with ethyl acetate (3×100 ml). The com-
bined organic extracts were washed with water

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
103

104
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
(3×150 ml), dried, and evaporated to a white
mixture was stirred for 2 h at room temperature.
The mixture was passed through a silica gel
column (70 mm×5 mm i.d.) using hexane (50 ml)
and ethyl acetate-hexane 1:1 as the eluting sol-
vents. Fractions 50 ml in volume were collected.
Evaporation of the contents of fraction 4 gave a
crude preparation of iodide epimers 9a and 9b (70
solid (1.16 g). The crude product (96:4 mixture of
1
6:10 by H-NMR) was subjected to MPLC on a
Lobar column (310 mm×25 mm i.d.; Lichroprep
Si 60, 40–63 vm; elution with hexane (200 ml)
and ethyl acetate-hexane 16:84. Fractions 78–117
1
contained 10 (46 mg; ]95% purity by H-NMR).
1
Fractions 51–73 were evaporated to give 6 (947
mg; 94% yield) containing 1–2% of the 14i
epimer by 1H-NMR. An analytical sample was
obtained by recrystallization from methanol: m.p.
152–153°C; single component on TLC in SS-1 (Rf
mg; 28:5:1 mixture of 9a, 9b, and 4 by H-NMR):
single component on TLC in SS-1 (Rf 0.47) and
1
SS-2 (Rf 0.47); H- and 13C-NMR of 9a, Table 1,
1
Table 2 and Table 3; resolved H-NMR signals of
9b: l 1.663† (H-14h), 2.465 (ddd, 18.2, 8.6, 1.2
Hz, H-16i), 1.746† (H-16i), 1.640† (H-17h),
0.759 (s, H-18), 0.814 (d, 0.5 Hz, H-19), 1.54
(H-20), 1.076 (d, 6.6 Hz, H-21), 1.93‡ and 2.01‡
(H-22), 4.284† (H-23), 2.72† and 2.83† (H-24).
To a solution of the iodide preparation (ꢀ70
mg) and AIBN (8 mg) in dry tetrahydrofuran (10
ml) was added tributyltin hydride (0.46 ml) under
argon. After 6 h, the solvent was evaporated and
the resulting residue was dissolved in CH2Cl2 (1
ml) and subjected to silica gel column (190 mm×
10 mm i.d.) chromatography, with successive elu-
tion with hexane (300 ml), ethyl acetate-hexane
5:95 (500 ml), and ethyl acetate-hexane 1:9. The
contents of fractions 70–77 were pooled (48 mg)
and recrystallized twice from methanol to give the
saturated F7-15-ketosterol 4 (21 mg): m.p. 146–
147°C; single component on TLC in SS-1 (Rf
0.58) and SS-2 (Rf 0.39) and by HPLC (tR 6.9
min; methanol-water 95:5); GC-MS of the TMS
ether (ZAB instrument), major component (tR 8.2
min, 98%), MS, 600 (42, M+), 585* (100), 510*
(42), 335 (20, ion E); minor component (tR 7.1
min, 2%), MS, 407 (29), 335 (100, ion E); IR, wmax
3400, 2924, 2849, 1736, 1317, 1223, 1157, 1132,
1042, 939, 719 cm−1; GC-MS (ZAB instrument,
250°C) of the TMS ether (tR 39.5 min), 600 (31,
M+), 585* (100), 510* (15), 495* (4), 471 (7), 335
(10, ion E), 155 (22); high resolution MS,
528.2837, calcd. for C27H39O2F7, 528.2838; MS,
528† (100, M+), 513*† (12), 510*† (15), 495*† (9),
477*† (3), 456† (6, M−Ring A), 438† (3, M−
Ring A−H2O), 335† (55, ion E), 319† (7, SC+
C6H8), 295† (2, SC+C3H4O), 289*† (13), 276† (9,
M−SC−CH), 271*† (6), 266† (7, SC+C2H3),
261† (93, ion F1), 259† (4, C17H23O2), 253*† (4),
1
0.47) and SS-2 (Rf 0.42); \99% purity by H-
NMR. Four repetitions of this procedure on a 2-g
scale furnished a crude product (8.35 g), which
was purified by MPLC to give 6 (7.44 g; 93%
yield).
2.5. 3i-Hydroxy-5h-chol-23-en-15-one (6) by
lithium-ammonia reduction of 10 (NH4Cl quench)
Reduction of D8(14),23-15-ketosteroid 10 (1 g; 2.8
mmol) in ether (50 ml) with lithium (300 mg; 43
mmol) in ammonia (100 ml) was carried out as
above except that the reaction was quenched by
addition of NH4Cl (3 g) in one portion. After
work-up and MPLC as above, evaporation of
fractions 53–74 gave 6 (346 mg; 34% yield), which
1
was identified by H-NMR. Evaporation of frac-
tions 79–117 gave a residue (72 mg, chiefly 10 by
1H-NMR). Evaporation of fractions 123–162
gave unidentified 3i-hydroxy-D23-15-ketosteroid
material (81 mg) showing the following distinctive
1H-NMR signals: lH 2.81 (dd, 18.5, 9.6 Hz), 2.70
(dd, 18.8, 9.0 Hz), 2.56 (dq, 12, 3 Hz), 2.46 (dq,
12.5, 3.4 Hz), 0.951 (d, 6.6 Hz), 0.926 (d, 6.6 Hz),
0.833 (s), 0.829 (s), 0.806 (s), 0.799 (s). Elution of
the column with ethyl acetate gave a yellowish oil
1
(360 mg) showing the following H-NMR methyl
signals: lH 1.161 (s), 0.944 (d, 6.9 Hz), 0.876 (s).
2.6. 3i-Hydroxy-25,26,26,26,27,27,27-heptafluoro-
5h-cholestan-15-one (4)
To 6 (50 mg) in hexane (12 ml) was added
sufficient 2-iodoheptafluoropropane (0.3 ml) to
dissolve the steroid. Triethylborane (0.1 ml; 1 M
solution in hexane) was added, and the resulting

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
105
234† (56, ion G2), 219† (16, C15H23O), 217† (37,
be separated by argentation chromatography but
not by chromatography on silica gel. Completion
of reaction was judged by H-NMR (disappear-
ion G3−H2O), 216† (59, ion G2−H2O), 201† (19,
1
1
C15H21), 190† (26, C14H22), 108† (89, C8H12); H,
13C, and 19F-NMR and 1H-1H-NMR coupling
constants, Table 1, Table 2 and Table 3.
ance of olefinic signals at l 5.02 and 5.76 or
H-16h signals at l 2.45), MS (disappearance of
ion at m/z 358), or by TLC (2:8 ethyl acetate-hex-
ane, three developments), in which the product
showed a bluish spot at slightly lower Rf than the
greenish spot of the starting material. Relative
amounts of 9a, 9b, 4, and 6 were estimated by the
relative intensities of the H-19, H-21, and H-16h
signals in resolution-enhanced 500-MHz 1H-
NMR spectra. 19F-NMR spectra (Table 3) were
also useful for quantitating trace levels of 4 in
iodide preparations. The D8(14) iodide preparations
(predominantly 12a and 12b) were analyzed simi-
larly. The H-16h resonance was particularly useful
for quantitation of D8(14) steroid mixtures because
the absence of the 1-Hz J14h-16h coupling observed
in saturated 15-ketosteroids simplifies the cou-
The above conversion of 6 to 4 was modified
for work on a larger scale. Compound 6 (4 g, 11.1
mmol) was dissolved in hexane (400 ml) by warm-
ing, followed by addition of triethylborane (4.44
ml; 1 M solution in hexane) and 2-iodohep-
tafluoropropane (7.82 ml; 55 mmol). The solution
was stirred overnight under argon at room tem-
1
perature. After H-NMR analysis indicated com-
pletion of the reaction, the mixture was
evaporated to a yellow-brown oil (7.38 g) that was
purified on silica gel (80 g; 230 mm×13 mm i.d.
column; elution with hexane (250 ml), ethyl ac-
etate-hexane 5:95 (250 ml), ethyl acetate-hexane
1:9 (500 ml), ethyl acetate-hexane 15:85 (500 ml),
and ethyl acetate-hexane 2:8). Evaporation of
fractions 116–293 gave a crude iodide preparation
as a nearly white solid (5.26 g; chiefly 6:1 mixture
1
pling pattern to a doublet-doublet. H- and 19F-
NMR spectra in Fig. 3 illustrate these points.
Iodide preparations were also analyzed by re-
versed phase HPLC (water-methanol 1:7): tR 12.9,
14.7, and 15.4 min for 4, 9b, and 9a. A 5:1:1
1
of 9a and 9b with 4% of 4 by H-NMR): m.p.
131–132°C; single component on TLC in SS-1 (Rf
0.48) and SS-2 (Rf 0.43); IR, wmax 3400, 2934,
2857, 1738, 1308, 1292, 1225, 1159, 1146, 1045
cm−1. Fractions 85–115 gave a yellowish solid
that was combined with the main product from a
similar 2-g reaction to give, after purification, 4
(4.15 g; 86% combined yield). Reduction of a
similar iodide preparation (5.2 g; 8.0 mmol) with
tributyltin hydride (2.95 ml; 11.0 mmol) and
AIBN (100 mg), followed by MPLC purification,
gave a white solid (3.98 g). Three recrystalliza-
tions from methanol to remove small amounts of
the 14i epimer and organotin impurities fur-
nished 4 (3.49 g; 83% yield): m.p. 146–147°C;
1
mixture of 9a, 9b, and 4 (portions of H-NMR
spectrum shown in Fig. 3, panels D and E) was
subjected to HPLC (two injections, 100 vg total).
An early fraction showed 1H-NMR signals for
H-3h, H-7i, H-16h, H-18, H-19, and H-21 identi-
cal in chemical shift (90.002 ppm) and coupling
patterns to those of 4. An intermediate fraction
(tR 14.7 min) contained a 2:1:1 mixture of 9b, 9a,
and an unidentified component (possibly a 14i
1
sterol; H-NMR, l 0.764 (s), 0.860 (d, 6.6 Hz)).
Later fractions contained 9a (]99% purity by
1H-NMR).
1
single component by TLC; \99% purity by H-
The crude reaction product 4 typically con-
tained ꢀ10% of the 14i epimer and considerable
organotin material. The mother liquors from re-
crystallization of 4 were enriched in tributyltin
species (NMR signals at lH 1.64 (m), 1.36 (m),
1.30 (m), 0.93 (t, 7.3 Hz) and lC 27.8, 26.8, 17.5,
13.5) and the 14i epimer (NMR signals at lH
2.46 (m, H-7i), 2.37 (dd, 19.9 and 10.1 Hz,
H-16h), 1.162 (s, H-18), 0.898 (d, 6.9 Hz, H-21),
0.760 (s, H-19)). NMR assignments for the 14i
epimer correspond to those of the non-fluorinated
NMR.
2.7. Analysis of the iodide preparations and
saturated F7-15-ketosterol 4 for purity
Iodide preparations contained a ꢀ6:1 ratio of
9a and 9b and variable amounts of 4 and 6.
Incomplete consumption of the starting material 6
ultimately resulted in contamination of the final
15-ketosterol product 4 by 6, a mixture that could

106
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
Fig. 3. 1H- (500 MHz) and 19F-NMR (235 MHz) spectra of iodide preparations. Panel A shows a portion of the 19F-NMR spectrum
(region containing resonances for F-26 and F-27) of a 7:1:3 mixture of (23R)-3i-hydroxy-23-iodo-25,26,26,26,27,27,27-heptafluoro-
5h-cholest-8(14)-en-15-one (12a), the 23S epimer 12b, and 3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholest-8(14)-en-15-one
(3). Panel B shows the 1H-NMR spectral region containing the H-16h resonances of the 7:1:3 mixture, and panel C shows the
1H-NMR methyl region. Panels D and E show portions of the 1H-NMR spectrum (H-16h and methyl regions) of a 5:1:1 mixture
of the corresponding saturated sterols 9a, 9b, and 4. The absence of signals for H-16h and H-21 of 3i-hydroxy-5h-chola-8(14),23-
dien-15-one (10) and its saturated analog 6 in panels B–D demonstrate that these compounds are not present in the iodide
preparations. The 1H-NMR spectra were resolution enhanced; vertical and horizontal scales are not comparable between panels.

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
107
14i epimer [33]. Traces of the organotin impu-
rity were easily detected at 500 MHz by the
triplet at l 0.93, but this signal may be ob-
scured at lower fields by methine proton reso-
nances of 3 (H-1h and H-7h). Traces of the 14i
epimer could be detected in the 1H methyl re-
gion at 500 MHz; at 200–300 MHz, the best
criterion of purity may be a clean doublet-dou-
blet at l 2.41 showing no extraneous lines, even
under resolution enhancement. Examination of
hexane (100 ml), ethyl acetate-hexane 1:9 (100
ml), and ethyl acetate-hexane 15:85. Evaporation
of fractions 32–67 gave a crude iodide prepara-
tion (324 mg; 88% yield; 1H- and 19F-NMR
analysis (Fig. 3, panels A–C) showing a 7:1:3
mixture of 12a, 12b, and 3 containing about 2%
each of (23E)-3i-hydroxy-25,26,26,26,27,27,27-
heptafluoro-5h-cholesta-8(14),23-dien-15-one and
an unidentified sterol and B0.5% of the starting
D8(14),23 steroid 10: m.p. 160–162°C; single com-
ponent by TLC in SS-1 (Rf 0.48); IR, wmax 3400,
2934, 2861, 1701, 1622, 1308, 1292, 1225, 1150,
1044 cm−1; high resolution MS, calcd. for
C27H36O2F7I, 652.1648, found 652.1642; MS
652† (63, M+), 637* (10), 634* (11), 619*† (19),
524*† (53), 509* (17), 506* (8), 491* (15), 287*
(17), 285 (9), 269*† (63), 267 (9), 265 (9), 259
(13, ion C), 251* (29), 237 (28), 213 (31), 107
(100), 105 (91) as well as ions attributable to 3,
including 526 (74), 511 (10), 508 (14), 493 (28),
387 (10) and ions (e.g. 287, 269, 259, 251) con-
1
the l 2.2–2.4 region of the 500 MHz H-NMR
of thrice-recrystallized 4 indicated none of the
14i epimer or tributyltin material (detection
limits 0.5 and 0.2 mole percent, respectively).
2.8. Saturated F7-15-ketosterol 4 from
F7-Z8(14)-15-ketosterol 3
As in the procedure for preparing 6 from 10,
a cold solution of 3 (50 mg; 0.095 mmol) in
ether (10 ml) was added to lithium (50 mg; 7.2
mmol) in ammonia (25 ml) at −78°C. The mix-
ture was stirred for 8 min, followed by quench-
ing with bromobenzene (1 ml). Work-up as
above gave a pale yellow solid (51 mg) that was
subjected to MPLC (18 g silica gel, 70–230
mesh; elution with hexane (100 ml), ethyl ac-
etate-hexane 1:9 (500 ml); ethyl acetate-hexane
15:85). Evaporation of fractions 76–87 gave a
white solid (38 mg; 76% yield): m.p. 146–147°C;
single component on TLC in SS-1 (Rf 0.52) and
SS-2 (Rf 0.44); 1H, 13C, and 19F-NMR spectra
essentially identical with those of authentic 4.
1
taining contributions from 3; H, 13C, and 19F-
NMR and 1H-1H-NMR coupling constants,
1
Table 1, Table 2 and Table 3. Partial H- and
13C-NMR data for 12b: lH 2.372 (dd, 18.1, 7.7
Hz, H-16h), 2.045 (dd, 18.1, 12.5 Hz, H-16i),
1.544† (m, H-17h), 0.994 (s, H-18), 0.719 (s, H-
19), 1.090 (d, 6.5 Hz, H-21), 1.925† (dd, 14.8,
6.7 Hz, H-22), 2.047† (ddd, 14.8, 8.5, 4.0 Hz,
H-22), 4.31† (m, H-23R), 2.738† (ddd, 18.5†, ꢀ
16, ꢀ8 Hz, H-24), 2.848† (ddd, ꢀ20, ꢀ16,
ꢀ5 Hz, H-24) and lC 42.0 (C-16), 51.0 (C-17),
18.8 (C-18), 12.9 (C-19), 19.7 (C-21), 49.7 (C-
22).
2.9. Mixture of (23R)-3i-hydroxy-23-iodo-25,26,
26,26,27,27,27-heptafluoro-5h-cholest-8(14)-en-
15-one (12a) and its 23S epimer 12b
2.10. Lithium-ammonia reduction of
F7-Z8(14)-15-ketosterol iodide preparation (12a
and 12b)
To the D8(14),23 steroid 10 (200 mg, 0.56
mmol) in hexane (50 ml) was added successively
triethylborane (0.55 ml, 1 M solution in hexane)
A cold solution of a 7:1:3 mixture of 12a, 12b,
and 3 (250 mg; containing B0.5% 10) in anhy-
drous ether (25 ml) was added rapidly to a well-
stirred blue solution of Li (100 mg) in undistilled
ammonia (50 ml) and the mixture was stirred for
10 min at −78°C. The reaction was quenched by
rapid addition of bromobenzene (1 ml) and
poured onto ice. The resulting mixture was ex-
and 2-iodoheptafluoropropane (1.96 ml,
7
mmol), and the solution was stirred in the dark
1
overnight at room temperature. After H-NMR
analysis showed completion of the reaction, the
mixture was poured directly onto a silica gel
column (12 g), and the product was eluted with

108
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
tracted with ether (3×50 ml), washed with water
(3×50 ml), dried, and evaporated to a yellow oil
(0.26 g). H-NMR analysis showed 4 accompa-
and water (100 ml) were added, and the mixture
was extracted with ethyl acetate (2×50 ml). The
combined organic extracts were washed with wa-
ter, dried over Na2SO4, and evaporated to a
white solid (250 mg). The crude product was
subjected to MPLC on silica gel (20 g; elution
with hexane (100 ml) and ethyl acetate-hexane
16:84). Evaporation of fractions 48–76 gave
15i-hydroxy epimer 13b as a white solid (184
mg; 73% yield): m.p. 161–162°C; single compo-
nent on TLC in SS-1 (Rf 0.36) and SS-2 (Rf
0.33); GC-MS (ZAB instrument, 290°C) of the
3i-TMS ether (tR 12.0 min, 98%), MS, 602 (7,
M+), 587* (100), 584* (44), 569* (15), 512* (12),
497* (3), 494* (17), 479* (19), 455 (10), 345 (74,
ion F2 or M−SC−H2O), 335 (14, ion E), 332
(4), 306 (8, ion G2), 255 (82), 217 (56, ion G3−
TMSOH), 107 (42) and the 3i,15i-bis-TMS
ether (tR 9.5 min, 2%), MS, 659 (16, M−CH3),
584* (44), 494* (17), 479* (22), 367 (100, SC+
128), 345 (15, ion F2 or M−SC−TMSOH),
255* (18), 129 (28); IR, wmax 3400, 2934, 2851,
1312, 1221, 1159, 1134, 1032 cm−1; high resolu-
tion MS, calcd. for C27H41O2F7, 530.2995, found
530.2976; MS, 530† (1, M+), 512*† (31), 497*†
(6), 494* (1), 479*† (2), 331† (3, C15H18F7), 273*†
(6), 263† (40, ion F2), 255*† (4), 234† (32, ion
1
nied by a few minor components, including ꢀ
4% of 6. The crude product was subjected to
MPLC on silica gel (20 g, 70–230 mesh; elution
with hexane (100 ml) and ethyl acetate-hexane
15:85). Evaporation of fractions 21–32 gave a
white solid (120 mg) consisting of 4 with the
same minor components, and this composition
was unchanged by recrystallization from
methanol. A portion (75 mg) of this mixture was
subjected to MPLC on silica gel-AgNO3 (20 g;
elution with hexane (50 ml) and ethyl acetate-
hexane 15:85). Evaporation of fractions 30–37
1
gave 4 (62 mg; ꢀ99% purity by H-NMR): m.p.
146–147°C; single component on TLC in SS-1
1
(Rf 0.54) and SS-2 (Rf 0.45); H-NMR virtually
identical with data for 4 in Table 1. Evaporation
of fractions 44–50 gave 6 (3 mg; ꢀ99% purity
by 1H-NMR): m.p. 152–153°C; single compo-
nent on TLC in SS-1 (Rf 0.51) and SS-2 (Rf
1
0.42); H-NMR and MS virtually identical with
those of 6 described above.
Similar lithium-ammonia reduction of
a
60:10:3:1 mixture (1 g) of 12a, 12b, 10, and 3 for
8 min gave, after chromatography, a white solid
1
1
(548 mg; 5:1 mixture of 4 and 6 by H-NMR)
G2), 217† (100, ion G3−H2O), 107 (37); H, 13C,
and a later-eluting material (117 mg) consisting
of 1:1:5:5 mixture of 6, 12a, 3, and (23E)-3i-hy-
droxy-25,26,26,26,27,27,27-heptafluoro-5h-chol-
esta-8(14),23-dien-15-one. The latter component
was identified by NMR signals at lH 6.28 (ddd,
15.7, 8.7, 6.3 Hz), 5.55 (dd, 20.0, 15.8 Hz), 1.008
(d, ꢀ6.6 Hz), 0.999 (s) and lC 139.3, 115.9,
38.8, 34.5. The mixture of 4 and 6 (250-mg por-
tion) was separated on silica gel-AgNO3 into 4
(214 mg; m.p. 147–148°C) and 6 (29 mg; m.p.
152–153°C).
and 19F-NMR and 1H-1H-NMR coupling con-
stants, Table 1, Table 2 and Table 3.
Evaporation of fractions 88–146 gave 15h-hy-
droxy epimer 13a as a white solid (49 mg; 20%
yield): m.p. 185–186°C; single component on
TLC in SS-1 (Rf 0.27) and SS-2 (Rf 0.26); GC-
MS (ZAB instrument, 290°C) showed the
3i,15h-bis-TMS ether (tR 10.0 min, 99.7%), MS,
674 (2, M+), 659* (19), 584* (30), 568 (6), 494*
(15), 479* (20), 367 (100, SC+128), 345 (14, ion
F2 or M−SC−H2O), 255* (19), 129 (19); IR,
wmax 3400, 2932, 2855, 1312, 1219, 1159, 1132,
1040 cm−1; high resolution MS, calcd. for
C27H41O2F7, 530.2995, found 530.2980; MS, 530†
(4, M+), 512*† (81), 497*† (27), 494* (3), 479*†
(6), 399† (5, C20H26F7), 320 (4), 273*† (18), 263†
(33, ion F2), 255*† (8), 234† (22, ion G2), 217†
2.11. 25,26,26,26,27,27,27-heptafluoro-5h-choles-
tane-3i,15h-diol (13a) and its 15i epimer 13b
To a solution of saturated F7-15-ketosterol 4
(250 mg; 0.47 mmol) in anhydrous ether (50 ml)
was added LiAlH4 (36 mg; 0.95 mmol), and the
mixture was stirred at room temperature under
argon for 2 h. Saturated aqueous NH4Cl (2 ml)
1
(100, ion G3−H2O), 107 (56); H, 13C, and 19F-
NMR and 1H-1H-NMR coupling constants,
Table 1, Table 2 and Table 3.

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
109
2.12. 3i-Hydroxy-7h-methyl-5h-cholestan-15-one
(16)
Chromatographic purification as described for 14
gave, after evaporation of fractions 31–47, 17
(156 mg): m.p. 177–178°C; single component by
TLC in SS-1 (Rf 0.58) and SS-2 (Rf 0.47); GC-MS
(ZAB instrument, 290°C) of the TMS ether (single
component at tR 12.0 min), 614 (27, M+), 599*
(100), 524* (27), 509* (7), 485 (11), 467 (8), 335
(ion E), 231 (6, ion G3−TMSOH), 230 (5, ion
G2−TMSOH), 169 (73); high resolution MS,
calcd. for C28H41O2F7, 542.2995, found 542.3000;
MS, 542† (93, M+), 527*† (29), 524*† (15), 509*†
(10), 335† (100, ion E), 319† (7, SC+C6H8), 303*†
(12), 285*† (12), 275† (28, ion F1), 266 (8), 248†
(33, ion G2), 233† (21, C16H25O), 231† (32, ion
G3−H2O), 230† (14, ion G2−H2O), 206 (33), 205
As in the first procedure for preparing 6 from
10, a cold solution of the 7h-methyl-D8(14)-15-ke-
tosterol 14 (250 mg; 0.60 mmol) in ether (25 ml)
was added to lithium (100 mg; 14.4 mmol) in
ammonia (50 ml) at −78°C. The mixture was
stirred for 15 min, followed by quenching with
bromobenzene (1 ml). Work-up as for 10 gave a
1
white solid (258 mg), whose H-NMR spectrum
showed a 5:4 mixture of 16 and an unidentified
compound together with traces of 14. The crude
product was subjected to MPLC on silica gel (20
g; elution with hexane (100 ml) and ethyl acetate-
hexane 12:88). Evaporation of fractions 21–36
gave 16 (128 mg): m.p. 158.5–159.5°C; single
component on TLC in SS-1 (Rf 0.62) and SS-2 (Rf
0.47); GC-MS (ZAB instrument, 290°C) of the
TMS ether (tR 18.5 min), 488 (42, M+), 473*
(100), 398* (23), 383* (8), 359 (13), 347 (7, ion
F1), 231 (11, ion G3−TMSOH), 230 (13, ion
G2−TMSOH), 209 (24, ion E), 169 (66); high
resolution MS, calcd. for C27H48O2, 416.3654,
found 416.3654; MS, 416† (100, M+), 401*† (23),
398*† (23), 383*† (6), 303*† (17), 285*† (12), 275†
(37, ion F1), 248† (40, ion G2), 233† (31, C16H23O),
231† (60, ion G3−H2O), 209† (99, ion E), 192
1
(43), 192 (39), 123 (74), 122 (53); H, 13C, and
19F-NMR and 1H-1H-NMR coupling constants,
Table 1, Table 2 and Table 3. Evaporation of
fractions 53–68 gave 15 (30 mg).
2.14. (25R)-3i,26-dihydroxy-5h-cholestan-15-one
(19)
As in the first procedure for preparing 6 from
10, a cold solution of (25R)-3i,26-dihydroxy-5h-
cholest-8(14)-en-15-one (18; 100 mg; 0.24 mmol)
in tetrahydrofuran (25 ml) was added to lithium
(100 mg; 14.4 mmol) in ammonia (50 ml) at
−78°C. The mixture was stirred for 7 min, fol-
lowed by quenching with bromobenzene (1 ml).
Work-up as for 10 gave a white solid (0.11 g; 2:1
1
(48), 123 (76), 122 (65); H- and 13C-NMR and
1H-1H-NMR coupling constants, Table 1, Table 2
and Table 3. Evaporation of fractions 42–67 gave
a mixture (21 mg) containing 14. Elution of the
column with methanol gave an unidentified sterol
1
mixture of 19 and 18 by H-NMR), which was
subjected to MPLC (20 g silica gel; elution with
hexane (100 ml) and ethyl acetate-hexane 3:7).
Evaporation of fractions 44–57 gave 19 (56 mg):
m.p. 200–201°C; single component by TLC in
SS-2 (Rf 0.28) and SS-3 (Rf 0.56); GC-MS (ZAB
instrument, 290°C) of the bis-TMS ether (single
component at tR 32.2 min), 562 (14, M+), 547*
(100), 472* (38), 457* (25), 433 (22), 361* (36),
333 (16, ion F1), 297 (24, ion E), 271* (9), 216 (18,
ion G2−TMSOH); high resolution MS, calcd. for
C27H46O3, 418.3447, found 418.3456; MS, 418
(M+, 62), 403* (6), 385* (5), 360 (4), 346 (3), 317
(3), 289*† (35), 276 (20), 261† (100, ion F1), 234†
(49, ion G2), 225† (27, ion E), 219 (21), 217 (32,
ion G3−H2O), 216 (55, ion G2−H2O), 190 (31),
108 (91); 1H- and 13C-NMR and 1H-1H-NMR
1
(84 mg) showing H-NMR signals at l 3.64 (tt),
2.71 (td, 6.3, 2.6 Hz), 2.54 (m), 2.45 (dd, 12.2, 4.0
Hz), 1.03 (s, H-18), 0.93 (d, 6.8 Hz, H-21), 0.89
(d, 7.0 Hz), 0.86 (d, 6.6 Hz), 0.85 (d, 6.6 Hz), 0.83
(s, H-19).
2.13.
3i-Hydroxy-7h-methyl-25,26,26,26,27,27,27-hepta
fluoro-5h-cholestan-15-one (17)
Lithium-ammonia reduction of F7-7h-methyl-
D
8(14)-15-ketosterol 15 (250 mg) as described for
14 gave a crude product (261 mg), whose 1H-
NMR spectrum showed a 6:1 mixture of 17 and
15, with small amounts of unidentified sterols.

110
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
coupling constants, Table 1, Table 2 and Table
3. Evaporation of fractions 71–119 gave 18 (23
mg).
and the potential instability of the F7 moiety to
dissolving metal conditions [37]. These consider-
ations suggested early removal of the 3i-acetate
protecting group needed for the side chain cleav-
age and careful placement of the D8(14) reduction
in the synthetic scheme.
2.15. Nuclear O6erhauser enhancement (NOE)
difference experiments
Our first synthesis (Fig. 5) began with addition
of lithium to a solution of the known [24] D8(14),23
-
The following signals were observed in NOE
difference spectra of the saturated F7-15-ketos-
terol 4: irradiation of H-16h: H-16i, H-17h, H-
22R; irradiation of H-18: H-8i, H-11i, H-12i,
H-16i, H-20; irradiation of H-19: H-2i, H-4i,
H-7i, H-8i; irradiation of H-21: H-12i, H-20;
irradiation of H-22S: H-16h, H-16i, H-20, H-
22R, H-24R; irradiation of H-23S at l 1.64: H-
3h (evidently from concomitant irradiation of the
HDO signal), H-16h, H-21. In many cases, NOE
difference signals were obscured by multiplets
arising from spin population transfer. Irradiation
of H-18 of diol 13a led to strong enhancements
for signals of H-8i, H-11i, H-12i, H-15i, H-
16i, and H-20. Irradiation of H-18 of diol 13b
gave enhanced signals for H-8i, H-11i, H-12i,
and H-20.
15-ketosteroid acetate 5 in liquid ammonia, fol-
lowed by quenching with tert-butanol and then
ice. Although this reaction accomplished both
reduction of the D8(14) bond and removal of the
3i-acetate group in a single step, the yield of the
desired D23-15-ketosteroid 6 was only 40%, owing
largely to the formation of 15-hydroxy byprod-
ucts, i.e. D23-3i,15h-diol 7a, D23-3i,15i-diol 7b,
and D8(14),23-3i,15i-diol 8. Similar byproducts
have been reported in analogous reductions of
D
8(14)-15-ketosterols [35,38,39]. A markedly im-
proved yield was obtained by first hydrolyzing
acetate 5 to 10 (91–96% yield) and by modifying
the conditions of the lithium-ammonia reduction,
i.e. adding 10 in ether to the blue lithium-ammo-
3. Discussion
3.1. Chemical syntheses
One approach to the synthesis of the saturated
F7-15-ketosterol 4 follows a route similar to that
used to prepare F7-D8(14)-15-ketosterols 3 [24] and
15 [27]. As outlined in Fig. 4 these compounds
were prepared from the corresponding non-
fluorinated sterols, which were converted to the
24-hydroxy-C24 steroids by cleavage of the C8H17
side chain with trifluoroperacetic acid. Subse-
quent elimination of H2O via the selenide, addi-
tion of (CF3)2CFI to the resulting D23 steroid,
removal of the 23-iodide, and hydrolysis of the
3i-acetate gave the F7-D8(14)-15-ketosterols
[24,27]. Adaptation of this approach to the syn-
thesis of 4 was complicated by the likely instabil-
ity of saturated 15-ketosterols to the harsh
conditions of the side-chain cleavage [34], their
tendency to undergo epimerization at C-14 in a
mildly acidic or basic environment [33,35,36],
Fig. 4. Outline of the previously reported [24,27] synthetic
routes to F7-15-ketosterols from the corresponding unfluori-
nated 15-ketosterols.

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
111
Fig. 6. Fate of the C-23 free-radical postulated as an interme-
diate in the conversion of D23 steroids to 23-iodo-25,26,
26,26,27,27,27-heptafluorosterols. The radical intermediate
normally reacts with C3F7I in a chain-propagating step to give
the 23-iodo-F7-sterol, but the starting D23 steroid and sterols
with a C8H10F7 side chain may also be formed. The same C-23
free radical may be an intermediate in the reduction of 23-io-
dides by either lithium-ammonia or tributyltin hydride, but the
D23 steroid byproduct is not observed with the latter reagent.
Fig. 5. One chemical synthesis of 3i-hydroxy-25,26,26,26,
27,27,27-heptafluoro-5h-cholestan-15-one (4) from 3i-ace-
toxy-5h-chola-8(14),23-dien-15-one (5).
nia solution and quenching with bromobenzene
(ꢀ93% yield of 6). We have previously shown
that quenching with bromobenzene results in
much higher yields of saturated 15-ketosterols
than quenching with tert-butanol or ammonium
chloride [33].
made sufficiently soluble by using an excess of
(CF3)2CFI, a measure that circumvented masking
the 3i-hydroxyl as a tert-butyldimethylsilyl ether
or other hydrophobic derivative. Under these con-
ditions, the addition reaction proceeded at a rea-
sonable rate albeit more sluggishly than with
15-ketosteroid acetates used previously [24,27].
The resulting product consisted of a 6:1 mixture
of the C-23 iodide epimers1 9a and 9b together
with variable amounts of the saturated F7-15-ke-
tosterol 4, an interesting byproduct that evidently
arose when the postulated C-23 radical intermedi-
ate incorporated hydrogen instead of iodine (Fig.
6). The corresponding byproduct 3 was also ob-
served in the preparation of iodides 12a and 12b,
and reanalysis of other iodides reported previ-
The free-radical addition of (CF3)2CFI to a D23
olefin can tolerate a free hydroxyl group, but 6
was only sparingly soluble in hexane. In other
solvent mixtures, such as benzene-hexane, the re-
action was intolerably slow. However, 6 could be
1 The identity of the principal minor component consistently
present in iodide preparations has never been rigorously estab-
lished as the 23S epimer (e.g. 9b). However, the 1H-NMR
chemical shifts for H-18 and H-19 (l 0.759, 0.814) of this
minor component are incompatible with those anticipated for
a 14i-15-ketosterol [33], and the 19F-NMR coupling pattern
for F-25 (t of septet, 19 Hz and 7 Hz) is incompatible with a
–CH(CH3)CH2CH(CH2I)CF(CF3)2 side chain resulting from
reversed addition of (CF3)2CFI to the D23 bond.

112
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
saturated F7-15-ketosterol 4. The four-step synthe-
sis of 4 from 5 in 71% overall yield (Fig. 7) was
much more efficient than the original three-step
scheme shown in Fig. 6. It is noteworthy that all
the reactions described herein were performed
without any protection of the 3i-hydroxyl group.
An important factor in lithium-ammonia reduc-
tions is the temperature of reaction. Reduction of
D
8(14)-15-ketosteroids can be carried out at temper-
atures much lower than the boiling point of ammo-
nia (−33°C). Under conditions of low temperature
and quenching with bromobenzene, we have suc-
cessfully reduced the F7-D8(14)-15-ketosterol 3 to the
saturated F7-15-ketosterol 4 in respectable yield
(76%). In view of the reactivity of halides under
dissolving metal conditions [37], it is noteworthy
that the F7 moiety was unaffected by this transfor-
mation, which represents an efficient alternative
synthetic route to 4. This methodology was used to
prepare the 7h-methyl-F7-15-ketosterol 17 in a
single step from the corresponding D8(14)-15-ketos-
terol.
We also investigated the lithium-ammonia reduc-
tion of the F7-D8(14)-15-ketosterol iodide prepara-
tion (60:10:3:1 mixture of 12a, 12b, 10, and 3) in
an attempt to avoid the use of tributyltin hydride
(see below). However, the yield of 4 was modest
(49%) due to incomplete reduction and side reac-
tions, such as dehydroiodination. One byproduct
was the D23 steroid 6, which appeared to originate
from iodides 12a and 12b as well as from the D23
steroid 10. Lithium-ammonia reduction of another
F7-D8(14)-15-ketosterol iodide preparation (7:1:3
mixture of 12a, 12b, and 3, containing B0.5% 10)
also showed 6 as a byproduct. The formation of 6
from 12a or 12b suggested cleavage of the C24–C25
bond by loss of (CF3)2CF“ from the postulated
C-23 free-radical intermediate (Fig. 6). This expla-
nation is compatible with our observation that, in
Fig. 7. Synthetic routes to 3i-hydroxy-25,26,26,26,27,27,27-
heptafluoro-5h-cholestan-15-one (4) and related compounds.
ously [24,27] indicated that analogous byproducts
are routinely present in variable amounts2. Reduc-
tion of the iodide preparation (containing 9a, 9b,
and 4) with tributyltin hydride gave the desired
2 Amounts of 3 in D8(14)-15-ketosterol iodide preparations
(12a and 12b) varied from about 3% to 30%, and levels of 4 in
the corresponding saturated iodide preparations (9a and 9b)
were 2–15%. Reexamination of 1H-NMR data for the 7h-
methyl iodide acetate preparations [27] showed the presence of
ꢀ15% of the 3i-acetate derivative of 15, and 19F-NMR data
reported in [27] are compatible with this analysis. Reanalysis
also showed that iodide preparations in the synthesis of the
F7-D8(14)-15-ketosterol 2 contained from B5% to 15% of the
3i-acetate derivative of 2, and that iodide preparations in the
synthesis of 25,26,26,26,27,27,27-heptafluorocholesterol con-
tained about 3% of 3h,6h-diacetoxy-25,26,26,26,27,27,27-hep-
tafluoro-5i-cholestane. Proposed assignments for some mass
spectral fragmentations of the two iodides in [24] should be
reevaluated in light of the present findings. These byproducts
are desirable in syntheses of F7-15-ketosterols described herein
and previously [24,27], but in other reactions of the 23-iodide
(e.g. elimination), their presence may lead to contamination of
the product. Although the mechanism and reaction conditions
leading to the formation of byproducts with a C8H10F7 side
chain have not been established, observation of these byprod-
ucts is noteworthy in view of recent interest in alternative
hydrogen sources for radical reactions [40].

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
113
the preparation of iodides 9a and 9b from olefin 6,
occasionally NMR analysis of an aliquot of the
reaction mixture indicated completion of reaction,
but the isolated iodide product contained some
starting olefin. No such D23 products were noted in
any tributyltin hydride reductions, a result that
might be attributable to a shorter lifetime of the
C-23 radical intermediate. If 6 is not excluded fully
from iodide preparations, it must ultimately be
removed from the F7-15-ketosterol 4 by argenta-
tion chromatography.
When the saturated F7-15-ketosterol 4 was pre-
pared by tributyltin hydride reduction of the iodide
preparation (9a and 9b), the crude product was
contaminated with a few percent each of the 14i
epimer and organotin impurities. Tributyltin con-
taminants were not a problem in previous syntheses
of F7-15-ketosterols [24,27] because they were re-
moved during the saponification of the 3i-acetate.
In the present work, the 14i epimer of 4 and
organotin impurities were removed by repeated
recrystallization, after which the 500-MHz 1H-
NMR spectrum showed no trace of either material.
As anticipated from studies of the unfluorinated
15-ketosterols [33], chromatographic methods were
generally of limited value for analyzing mixtures of
C-14 epimers because of low detector response
(HPLC with UV detection), lack of separation
(TLC on silica gel), or possible epimerization
during analysis (GC)3.
Also reported herein are syntheses of saturated
15-hydroxysterols (Fig. 7) and additional saturated
15-ketosterols (Fig. 8). The saturated 15h-hydroxy
and 15i-hydroxyl F7-sterols 13a and 13b were
prepared by lithium aluminum hydride reduction
of 15-ketosterol 4. The saturated 7h-methyl-15-ke-
tosterol 16, its F7 analog 17, and the saturated
26-hydroxy-15-ketosterol 19 were obtained by
lithium-ammonia reduction of the corresponding
D
8(14)-15-ketosterols 14, 15, and 18.
In crude products from hydrolyses of D8(14)-15-
3 The extent of epimerization varied considerably from in-
strument to instrument. Negligible (B2%) epimerization was
observed for 2, its 14i epimer, 4, 6, 16, 17, and 19 with the
ZAB instrument, on which the GC injector is cleaned monthly.
Another GC instrument showed 2–4% epimerization, and two
others, on which the injector is cleaned infrequently, showed
extensive (up to 50%) epimerization. For example, despite
\99% purity of 4 by 1H-NMR, GC-MS of 4 on the Extrel
instrument showed two components: major component (4; tR
7.8 min, 90%), MS, 528 (43, M+), 335 (77, ion E), 261 (92, ion
F1), 234 (70, ion G2), 216 (66, ion G2−H2O), 108 (100); minor
component (14i epimer of 4; tR 6.9 min, 10%) m/z 335 (100,
ion E), 289* (19). Among the four GC instruments, no obvi-
ous correlations could be found between epimerization and
injection mode (split, splitless, falling needle), carrier gas (ni-
trogen, helium), column length (15, 30, or 60 m), column
temperature (250–290°C), or injector temperature (270–
290°C).
ketosteroid esters, we have sometimes observed a
minor component showing slightly greater mobility
on TLC than the major product. In the hydrolysis
of acetate 10, the minor (ꢀ5%) component was
isolated and identified as the 14i-D8-15-ketosteroid
isomer 11. NMR results and molecular modeling
indicated that 11 has an unusual conformation,
which will be described elsewhere (W.K. Wilson
and G.J. Schroepfer, Jr., unpublished results). The
formation of ꢀ5% of an analogous 14i-D8-15-ke-
tosterol has been observed when the parent D8(14)
-
15-ketosterol 1 was subjected to acidic conditions
[41].
4 Deviations from 90.001 ppm precision are attributable to
concentration effects, small errors in analyzing resonances of
strong coupled protons, and minor solvent impurities. 1H-
NMR data for steroids in Table 1 and for diols 7a and 7b were
measured at concentrations of 530 mM; data for 8 (measured
at 100 mM) are of lower precision. Chemical shifts measured
from concentrated samples (50–150 mM) generally differed by
50.002 ppm from those from dilute samples. Some solvent
impurities also affected chemical shifts. For example, 1 equiva-
lent of methanol in 7a affected shieldings by up to 0.010 ppm.
Chemical shifts could usually be estimated to 90.001 ppm in
COSYDEC spectra (digital resolution in f1 of 0.003 ppm after
zero-filling from 256 to 512 points) as judged by their similar-
ity to shieldings from 1D spectra.
Fig. 8. Chemical syntheses of additional saturated 15-ketos-
terols (16, 17, and 19).

114
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
3.2. NMR analyses
signments in the C8H17 sterol side chain has been
hampered by spectral overlap and second-order
effects that persist even at high magnetic field
strengths. Stereochemical assignments have been
determined for the side-chain protons of
25,26,26,26,27,27,27-heptafluorosterols [24], which
show nearly first order couplings at 500 MHz, and
these results together with NMR data for 25-fluoro-
cholesterol have been used to assign signals of the
C-22 and C-23 protons of cholesterol and other
sterols by chemical shift comparisons [44]. How-
ever, only tentative stereochemical assignments
could be made for the C-24 protons. It has been
observed [33] that saturated 14h-15-ketosterols
show increased spectral dispersion of the side-chain
1H-NMR resonances relative to other sterols and
that the consequent reduction in overlap and sec-
ond-order effects permits rough analyses of their
coupling patterns at 500 MHz. However, these
analyses did not furnish unequivocal stereochemical
assignments for the C-24 proton signals.
In the present work, we have obtained a 750-MHz
1H-NMR spectrum of the 7h-methyl-15-ketosterol
16 and investigated the coupling patterns of the
C-24 protons and the upfield C-23 proton, both by
first-order analysis and by spin simulation. The
coupling constants of the C-24 protons were too
similar to distinguish the pro-R from the pro-S
signal, but spin simulation was more fruitful. Sim-
ulations were done assuming the downfield H-24
signal to be pro-S (Fig. 9, panel A) or pro-R (Fig.
9, panel C) based on the assumption of a predom-
inantly extended side chain [33,44] with J23-24 values
of ꢀ5 Hz for gauche couplings and ꢀ10 Hz for
anti couplings. The former simulation showed good
agreement with the experimental spectrum (Fig. 9,
panel B). More specific and convincing evidence was
provided by examination of the roof effect, in which
two strongly coupled proton signals each consist of
a set of strong inner lines and a set of weak outer
lines. Each strong line has a weak-line counterpart
at the mutual coupling constant. The difference in
intensity between the strong and weak lines is large
when the chemical shift difference is small relative
to the mutual coupling constant. The roof effects
shown in Fig. 9 demonstrate that the coupling
constant between H-23S and the downfield H-24
proton is about 5 Hz and therefore that the
downfield H-24 proton has the pro-S configuration.
The saturated F7-15-ketosterol 4 was character-
ized by 1H, 13C, and 19F-NMR and by GC-MS, IR,
GC, HPLC, TLC, and melting point. Full 1H- and
13C-NMR assignments were established for 4, its
synthetic precursors, and analogs. The 13C-NMR
assignments of C-8 and C-14 of 12a were reconfi-
rmed by an HMBC experiment in view of recent
work [42], in which these assignments were reversed
1
for a D8(14)-15-ketosterol. The H-NMR chemical
shifts were generally determined to 90.001 ppm4
as described previously [32] from 1D and f1-decou-
pled 1H-1H correlation spectra. This high precision
was crucial in the analysis of the iodide preparations
and detection therein of a byproduct having a
C8H10F7 side chain. The use of 1H-NMR to analyze
the composition of iodide preparations is illustrated
in Fig. 3.
The NMR data for 4 were shown to be in full
accord with the assigned structure by comparison
with NMR data for 3i-hydroxy-5h-cholestan-15-
one [33] and the F7-D8(14)-15-ketosterol 3 [24]. Diols
1
7a and 8 were identified by H- and 13C-NMR
chemical shift comparisons with 5h-cholestane-
3i,15h-diol [33] and 5h-cholest-8(14)-ene-3i,15i-
1
diol [43]. Comparison of the H and 13C chemical
shifts and 1H-1H coupling constants of diols 7a and
7b showed that 7b is the 15i epimer of 7a. Diols
13a and 13b were identified by chemical shift and
coupling constant comparisons with diols 7a and 7b
and by NOE difference experiments. Coupling
patterns for iodides 9a and 12a and NOE difference
experiments for the saturated F7-15-ketosterol 4
reaffirmed conformational analyses and stereo-
chemical assignments established previously for
side-chain protons of F7 sterols [24]. Notably,
enhancement of the H-24R signal upon irradiation
of the H-22S resonance supported the previous [24]
stereochemical assignments of the C-24 protons.
Evidence for a significant population of the +
gauche rotamer of the C17-C20-C22-C23 dihedral
was provided by the enhancements of signals for
H-16h upon irradiation of H-23S and H-16h and
H-16i upon irradiation of H-22S. This minor
side-chain conformer has been noted previously for
cholesterol and D8(14)-15-ketosterols [44].
1
Establishing stereochemical H-NMR signal as-

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
115
Fig. 9. Comparison of experimental and simulated 750-MHz 1H-NMR spectra of 7h-methyl-15-ketosterol 16: panel A, spectrum
simulated with H-24R at l 1.095, H-24S at l 1.140; panel B, observed spectrum (resolution enhanced); panel C, spectrum simulated
with H-24R at l 1.140, H-24S at l 1.095. Other simulation parameters are given above (coupling constants) or in Table 1 (chemical
shifts). The simulation parameters for panel A were manually adjusted slightly to improve the match with the experimental
spectrum; no attempt was made to optimize the parameters for the simulation in panel C. The downfield H-24 resonance consists
mainly of 16 lines, 8 of which are more intense due to the roof effect resulting from the proximity of the upfield H-24 resonance.
Of these 8 lines, the roof effect due to the nearby H-23S resonance causes 4 lines to be more intense (marked by closed circles) and
4 to be weaker (open circles). Each of these 4 stronger lines is separated from its weaker counterpart by the coupling constant to
H-23S. The other set of 8 lines of the downfield H-24 resonance are similarly marked in panels A and B with closed diamonds
(stronger lines) and open diamonds (weaker lines). The line marked in panels A and B by an arrow is from the upfield H-24 signal.

116
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
Because the side-chain conformation of satu-
tert-butyldiphenylsilyl [51] ethers show mainly ions
corresponding to fragmentation of the silyl group.
Many D8(14),9(11)-15-ketosterols and D8(14)-15-ketos-
terols having a 9h-hydroxy or 9h-fluoro substituent
show fragments corresponding to loss of ketene
[46]. Mass spectra of saturated 15-ketosterols of
14i configuration are typically dominated by a
single distinctive fragment ion, ion E of Fig. 2
([33,39,47], and references therein). When unsatu-
ration is present in the side chain, M−SC−2 is
usually observed together with or in place of the
M−SC ion [52,53], and such losses were noted
herein for D23 steroids. In the present work, ions F1,
G1, G2, and G3 were observed in much higher
abundance in spectra of saturated 3i-hydroxy-15-
ketosterols than in spectra of their TMS ethers, and
the opposite trend was noted for M−CH3. Ion D
was significant in spectra of some 3i-hydroxy-D23
steroids but absent in spectra of the TMS ethers.
Several other points deserve mention. Fragmen-
tation of 15-oxygenated sterols often occurs in ring
D, with the positive charge remaining either on the
steroid nucleus or side chain fragment. In the latter
case, ions from sterols with a C8H17 side chain are
often obscured by the forest of hydrocarbon frag-
ment ions in the m/z 100–200 region. Sterols with
the 25,26,26,26,27,27,27-heptafluoro substitution
provided the opportunity to identify some of these
ions, of which ion E is the most notable. These F7
side-chain ions were recognizable, even at resolu-
tions as low as 1000, because of their unusually
small exact mass increment. As observed previously
[24,27], no losses of HF were noted for the F7
sterols despite the presence of these ions in the mass
rated 15-ketosterols is similar to that of other
sterols [33], the stereochemical assignments estab-
lished for the C-24 protons of 16 apply to choles-
terol and other sterols having a C8H17 side chain
(excluding sterols with a 17h or 20S configuration,
some sterols with a 14i configuration, and other
sterols with unusual side-chain conformations).
This result combined with previous findings [32]
provides the first truly complete signal assignments
for the 1H-NMR spectra of cholesterol and related
C27 sterols.
3.3. Mass spectral findings
For purposes of sterol structure determination,
GC-MS is about 1000 times as sensitive as NMR.
Consequently, it is worthwhile to investigate the
elaborate relationship between chemical structure
and fragmentation tendencies in electron-impact
spectra. Determining the exact structures of frag-
ment ions and their mechanisms of formation is a
daunting task requiring MS-MS and/or extensive
isotopic labeling, but important clues about struc-
ture can often be obtained from less precise knowl-
edge, such as whether an ion contains the side chain
or ring A. The present work includes mass spectral
data for 15-oxygenated sterols with a variety of
functionality in the side chain and steroid nucleus.
These findings combined with previous MS results
[22–24,27–30,33,34,39,43,45–50] can be summa-
rized in a comprehensive scheme of fragmentation
tendencies for 15-ketosterols and 15-hydroxys-
terols with and without a D8(14) bond. The scheme
shown in Fig. 2 contains exact structures suggested
by isotope labeling studies ([39] and references
therein, [47,49]) and less precisely defined frag-
ments indicated by high-resolution MS data and
comparisons of fragmentation tendencies among
sterols with various substituents in the side chain
and steroid nucleus.
spectrum of 25-fluorocholesterol [44]. Unlike D8(14)
-
15-ketosterols, the saturated 15-ketosteroids de-
scribed herein gave no dehydrogenation artifacts
[23], did not readily form TMS enol ethers under
standard derivatizing conditions [23], and gave
M−2 ions [24] in negligible abundance. The 7h-
methyl substituent did not markedly affect the
fragmentation pathways of saturated 15-ketos-
terols except for the appearance of a major ion of
m/z 169 in the TMS ether derivatives. The mass
spectral fragmentations observed herein are com-
patible with those reported previously for D8(14)-15-
ketosterols[22–24,27–30,34,39,45,46,48–50], satu-
rated 15-ketosterols [33,39,47], D8(14)-15-hydroxys-
terols [43,48], and saturated 15-hydroxysterols [33].
The fragmentation pathways shown in Fig. 2
generally account for the bulk of the major ions
(m/z \250) observed in mass spectra of 15-oxy-
genated sterols. However, fragmentation patterns
of 15-oxygenated sterols may be markedly altered
by derivatization or minor changes in functionality.
For example, tert-butyldimethylsilyl [29,33] and

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
117
droxy-3-methylglutaryl coenzyme A in CHO-K1 cells.
Biochem. Int. 1, 223–228.
4. Summary
[3] F.D. Pinkerton, A. Izumi, C.M. Anderson, L.R. Miller,
A. Kisic and G.J. Schroepfer, Jr. (1982) 14h-Ethyl-5h-
We have described the synthesis of saturated
15-ketosterol 4 from both the D8(14)-15-ketosterol
3 and its synthetic precursor 5. The nature and
probable origin of byproducts is discussed, and
analytical methodology needed to ensure a final
product of high purity is described. Detailed MS
data of the steroids described herein together with
results reported previously have led to a compre-
hensive scheme to account for the mass spectral
cholest-7-ene-3i,15h-diol,
a potent inhibitor of sterol
biosynthesis, has two sites of action in cultured mam-
malian cells. J. Biol. Chem. 257, 1929–1936.
[4] F.R. Taylor, S.E. Saucier, E.P. Shown, E. Parish and
A.A. Kandutsch (1984) Correlation between oxysterol
binding to a cytosolic binding protein and potency in the
repression of hydroxymethylglutaryl coenzyme A reduc-
tase. J. Biol. Chem. 259, 12382–12387.
[5] T.N. Pajewski, F.D. Pinkerton, L.R. Miller and G.J.
Schroepfer, Jr. (1988) Inhibitors of sterol synthesis. Stud-
ies of the metabolism of 5h-cholest-8(14)-en-3i-ol-15-one
in Chinese hamster ovary cells and its effects on activities
of early enzymes in cholesterol biosynthesis. Chem. Phys.
Lipids 48, 153–168.
[6] F.D. Pinkerton, C.H. Spilman, D.P. Via and G.J.
Schroepfer, Jr. (1993) Differing effects of three oxysterols
on low density lipoprotein metabolism in Hep G2 cells.
Biochem. Biophys. Res. Commun. 193, 1091–1097.
[7] G.J. Schroepfer, Jr., A. Christophe, D.H. Needleman, A.
Kisic and B.C. Sherrill (1987) Inhibitors of sterol synthe-
sis. Dietary administration of 5h-cholest-8(14)-en-3i-ol-
15-one inhibits the intestinal absorption of cholesterol in
lymph-cannulated rats. Biochem. Biophys. Res. Commun.
146, 1003–1008.
1
fragmentation of 15-oxygenated sterols. H-NMR
stereochemical assignments were determined for
the C-24 signals of 7h-methyl-15-ketosterol 16
and by analogy for those of cholesterol. This
result represents the culmination of a series of
1
efforts to establish the complete H-NMR signal
assignments for cholesterol and related C27
sterols.
Acknowledgements
This work was supported in part by the Ralph
and Dorothy Looney Endowment Fund, the
Texas Advanced Technology Program (3604021
and 3604026), and the Robert A. Welch Founda-
tion (C-583). Some high-resolution mass spectra
were measured at the Midwest Center for Mass
Spectrometry with partial support by the National
Science Foundation, Biology Division (Grant No.
DIR 9217262). The Rice 500 MHz NMR Facility
was established with the support of NIH grant
RR05759 and the W.M. Keck Foundation. We
thank the Sealy Center for Structural Biology and
Dr. Edward P. Nikonowicz for obtaining 750-
MHz NMR spectra.
[8] J.S. Brabson and G.J. Schroepfer, Jr. (1988) Inhibitors of
sterol synthesis. The effects of dietary 5h-cholest-8(14)-en-
3i-ol-15-one on the fate of [4-14C]cholesterol and [2,4-
3H]5h-cholest-8(14)-en-3i-ol-15-one after intragastric
administration to rats. Chem. Phys. Lipids 47, 1–20.
[9] L.R. Miller, D.H. Needleman, J.S. Brabson, K.-S. Wang
and G.J. Schroepfer, Jr. (1987) 5h-Cholest-8(14)-en-3i-
ol-15-one. A competitive substrate for acyl coenzyme
A:cholesterol acyl transferase. Biochem. Biophys. Res.
Commun. 148, 934–940.
[10] D.H. Needleman, K. Strong, K.A. Stemke, J.S. Brabson,
A. Kisic and G.J. Schroepfer, Jr. (1987) Inhibitors of
sterol synthesis. Effect of dietary 5h-cholest-8(14)-en-3i-
ol-15-one on ACAT activity of jejunal microsomes of the
rat. Biochem. Biophys. Res. Commun. 148, 920–925.
[11] G.J. Schroepfer, Jr., D. Monger, A.S. Taylor, J.S. Cham-
berlain, E.J. Parish, A. Kisic and A.A. Kandutsch (1977)
Inhibitors of sterol synthesis. Hypocholesterolemic action
of dietary 5h-cholest-8(14)-en-3i-ol-15-one in rats and
mice. Biochem. Biophys. Res. Commun. 78, 1227–1233.
[12] G.J. Schroepfer, Jr., E.J. Parish, A. Kisic, E.M. Jackson,
C.M. Farley and G.E. Mott (1982) 5h-Cholest-8(14)-en-
3i-ol-15-one, a potent inhibitor of sterol biosynthesis,
lowers serum cholesterol and alters the distribution of
cholesterol in lipoproteins in baboons. Proc. Natl. Acad.
Sci. USA 79, 3042–3046.
References
[1] G.J. Schroepfer, Jr., E.J. Parish, H.W. Chen and A.A.
Kandutsch (1977) Inhibition of sterol biosynthesis in L
cells and mouse liver cells by 15-oxygenated sterols. J.
Biol. Chem. 252, 8975–8980.
[2] L.R. Miller, F.D. Pinkerton and G.J. Schroepfer, Jr.
[13] G.J. Schroepfer, Jr., B.C. Sherrill, K.-S. Wang, W.K.
Wilson, A. Kisic and T.B. Clarkson (1984) 5h-Cholest-
8(14)-en-3i-ol-15-one lowers serum cholesterol and in-
(1980) 5h-Cholest-8(14)-en-3i-ol-15-one,
hibitor of sterol synthesis, reduces the levels of activity of
enzymes involved in the synthesis and reduction of 3-hy-
a potent in-

118
A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
duces profound changes in the levels of lipoprotein
[24] S. Swaminathan, W.K. Wilson, F.D. Pinkerton, N. Gerst,
M. Ramser and G.J. Schroepfer, Jr. (1993) Inhibitors of
sterol synthesis. Chemical synthesis and properties of
3i-hydroxy-25,26,26,26,27,27,27-heptafluoro-5h-cholest-
8(14)-en-15-one and 25,26,26,26,27,27,27-heptafluoro-
cholesterol and their effects on 3-hydroxy-3-methylglu-
taryl coenzyme A reductase activity in cultured mam-
malian cells. J. Lipid Res. 34, 1805–1823.
[25] A. Izumi, F.D. Pinkerton, S.O. Nelson, J.St. Pyrek,
P.J.G. Neill, J.H. Smith and G.J. Schroepfer, Jr. (1994)
Inhibitors of sterol synthesis. Submicromolar 14h-ethyl-
5h-cholest-7-ene-3i,15h-diol causes a major modification
of the sterol composition of CHO-K1 cells and a marked
change in cell morphology. J. Lipid Res. 35, 1251–1266.
[26] S. Swaminathan, A.U. Siddiqui, F.D. Pinkerton, N.
Gerst, W.K. Wilson and G.J. Schroepfer, Jr. (1994) In-
hibitors of sterol synthesis: 3i-hydroxy-25,26,26,26,
27,27,27-heptafluoro-5h-cholestan-15-one, an analog of a
potent hypocholesterolemic agent in which its major
metabolism is blocked. Biochem. Biophys. Res. Commun.
201, 168–173.
cholesterol and apoproteins in monkeys fed a diet of
moderate cholesterol content. Proc. Natl. Acad. Sci. USA
81, 6861–6865.
[14] D.J. Monger, E.J. Parish and G.J. Schroepfer, Jr. (1980)
15-Oxygenated sterols. Enzymatic conversion of [2,4-
3H]5h-cholest-8(14)-en-3i-ol-15-one to cholesterol in rat
liver homogenate preparations. J. Biol. Chem. 255,
11122–11129.
[15] D.J. Monger and G.J. Schroepfer, Jr. (1988) Inhibitors of
cholesterol biosynthesis. Further studies of the
metabolism of 5h-cholest-8(14)-en-3i-ol-15-one in rat
liver preparations. Chem. Phys. Lipids 47, 21–46.
[16] G.J. Schroepfer, Jr., A.J. Chu, D.H. Needleman, A.
Izumi, P.T. Nguyen, K.-S. Wang, J.M. Little, B.C. Sher-
rill and A. Kisic (1988) Inhibitors of sterol synthesis.
Metabolism of 5h-cholest-8(14)-en-3i-ol-15-one after in-
travenous administration to bile duct-cannulated rats. J.
Biol. Chem. 263, 4110–4123.
[17] J.S. Brabson and G.J. Schroepfer, Jr. (1988) Inhibitors of
sterol synthesis. Studies of the distribution and
metabolism of 5h-[2,4-3H]cholest-8(14)-en-3i-ol-15-one
after intragastric administration to rats. Steroids 52, 51–
68.
[27] S. Swaminathan, A.U. Siddiqui, N. Gerst, F.D. Pinker-
ton, A. Kisic, L.J. Kim, W.K. Wilson and G.J.
Schroepfer, Jr. (1995) Inhibitors of sterol synthesis.
Metabolism-based design and construction of
a new
[18] G.J. Schroepfer, Jr., A. Christophe, A.J. Chu, A. Izumi,
A. Kisic and B.C. Sherrill (1988) Inhibitors of sterol
analog of 3i-hydroxy-5h-cholest-8(14)-en-3i-ol-15-one
and its effects in cultured mammalian cells and in rats. J.
Lipid Res. 36, 767–786.
synthesis.
A
major role of chylomicrons in the
metabolism of 5h-cholest-8(14)-en-3i-ol-15-one in the
rat. Chem. Phys. Lipids 48, 29–58.
[28] A.U. Siddiqui, N. Gerst, L.J. Kim, F.D. Pinkerton, A.
Kisic, W.K. Wilson and G.J. Schroepfer, Jr. (1994) In-
hibitors of sterol synthesis: effects of a 7h-alkyl analog of
3i-hydroxy-5h-cholest-8(14)-en-15-one on 3-hydroxy-3-
methylglutaryl coenzyme A reductase activity in cultured
mammalian cells and on serum cholesterol levels and
other parameters in rats. Chem. Phys. Lipids 70, 163–
178.
[29] H.-S. Kim, W.K. Wilson, D.H. Needleman, F.D. Pinker-
ton, D.K. Wilson, F.A. Quiocho and G.J. Schroepfer, Jr.
(1989) Inhibitors of sterol synthesis. Chemical synthesis,
structure, and biological activities of (25R)-3i,26-dihy-
droxy-5h-cholest-8(14)-en-15-one, a metabolite of 3i-hy-
droxy-5h-cholest-8(14)-en-15-one. J. Lipid Res. 30,
247–261.
[19] G.J. Schroepfer, Jr., T.N. Pajewski, M. Hylarides and A.
Kisic (1987) 5h-Cholest-8(14)-en-3i-ol-15-one. In vivo
conversion to cholesterol upon oral administration to a
nonhuman primate. Biochem. Biophys. Res. Commun.
146, 1027–1032.
[20] G.J. Schroepfer, Jr., A. Kisic, A. Izumi, K.-S. Wang,
K.D. Carey and A.J. Chu (1988) Inhibitors of sterol
synthesis. Metabolism of [2,4-3H]5h-cholest-8(14)-en-3i-
ol-15-one after intravenous administration to a nonhu-
man primate. J. Biol. Chem. 263, 4098–4109.
[21] T.N. Pajewski, J.S. Brabson, A. Kisic, K.-S. Wang, M.D.
Hylarides, E.M. Jackson and G.J. Schroepfer, Jr. (1989)
Inhibitors of sterol synthesis. Metabolism of [2,4-3H]5h-
cholest-8(14)-en-3i-ol-15-one after oral administration to
a nonhuman primate. Chem. Phys. Lipids 49, 243–263.
[22] J.St. Pyrek, J.L. Vermilion, T.W. Stephens, W.K. Wilson
and G.J. Schroepfer, Jr. (1989) Inhibitors of sterol synthe-
sis. Characterization of side chain oxygenated derivatives
formed upon incubation of 3i-hydroxy-5h-cholest-8(14)-
en-15-one with rat liver mitochondria. J. Biol. Chem. 264,
4536–4543.
[23] J.St. Pyrek, W.K. Wilson, S. Numazawa and G.J.
Schroepfer, Jr. (1991) Inhibitors of sterol synthesis. Char-
acterization of trimethylsilyl dienol ethers of 3i-hydroxy-
5h-cholest-8(14)-en-15-one. Applications in the analysis
of mitochondrial metabolites of the 15-ketosterol by gas
chromatography-mass spectrometry. J. Lipid Res. 32,
1371–1380.
[30] A.U. Siddiqui, W.K. Wilson and G.J. Schroepfer, Jr.
(1994) Inhibitors of sterol synthesis. An improved chemi-
8(14)-15-ketosterols having
cal synthesis of 26-oxygenated D
the 25R configuration. Chem. Phys. Lipids 71, 205–218.
[31] A.U. Siddiqui, W.K. Wilson, S. Swaminathan and G.J.
Schroepfer, Jr. (1992) Efficient preparation of steroidal
5,7-dienes of high purity. Chem. Phys. Lipids 63, 115–
129.
[32] W.K. Wilson, R.M. Sumpter, J.J. Warren, P.S. Rogers,
B. Ruan and G.J. Schroepfer, Jr. (1996) Analysis of
unsaturated C27 sterols by nuclear magnetic resonance
spectroscopy. J. Lipid Res. 37, 1529–1555.
[33] A.U. Siddiqui, W.K. Wilson, F.D. Pinkerton and G.J.
Schroepfer, Jr. (1994) Inhibitors of sterol synthesis. Syn-

A.U. Siddiqui et al. / Chemistry and Physics of Lipids 86 (1997) 95–119
119
thesis and spectral properties of 3i-hydroxy-5h-
cholestan-15-one and its 14i-epimer and their effects on
3-hydroxy-3-methylglutaryl coenzyme A reductase activ-
ity. Chem. Phys. Lipids 74, 1–15.
and on 3-hydroxy-3-methylglutaryl coenzyme A reductase
activity in CHO-K1 cells. Steroids 59, 310–317.
[45] S. Swaminathan, F.D. Pinkerton, S. Numazawa, W.K.
Wilson and G.J. Schroepfer, Jr. (1992) Inhibitors of sterol
synthesis. Chemical syntheses and spectral properties of
3i-hydroxy-5h-cholesta-8(14),24-dien-15-one, 3i,25-dihy-
droxy-5h-cholest-8(14)-en-15-one, and 3i,24-dihydroxy-
5h-cholest-8(14)-en-15-one and their effects on 3-hydroxy-
3-methylglutaryl coenzyme A reductase activity in CHO-
K1 cells. J. Lipid Res. 33, 1503–1515.
[34] J.E. Herz, S. Swaminathan, F.D. Pinkerton, W.K. Wilson
and G.J. Schroepfer, Jr. (1992) Inhibitors of sterol synthe-
sis. A highly efficient and specific side chain oxidation of
3i-acetoxy-5h-cholest-8(14)-en-15-one for construction of
metabolites and analogs of the 15-ketosterol. J. Lipid Res.
33, 579–598.
[35] D.H.R. Barton and G.F. Laws (1954) Some oxidation
products of ergosta-7:14:22-triene-3i-yl acetate (ergos-
terol B3 acetate). J. Chem. Soc. 52–63.
[36] A.R. van Horn and C. Djerassi (1967) Optical rotatory
dispersion studies. CVII. Factors governing the relative
stability of hydrindanones. Syntheses of 17-alkyl-15-keto
steroids. J. Am. Chem. Soc. 89, 651–664.
[37] D. Caine (1976) Reduction and related reactions of h,i-
unsaturated carbonyl compounds with metals in liquid
ammonia. Org. Reactions 23, 1–258 (notably pp. 59–60).
[38] S. Ikegawa, J. Kinoshita, K. Takeshita, H. Sato and M.
Tohma (1989) Preparation and comparative properties of
antisera against glyco-3i,12h-dihydroxy-5-cholen-24-oic
acid linked to bovine serum albumin at the C-3, C-15h,
and C-24 positions. Chem. Pharm. Bull. 37, 2147–2152.
[39] I. Midgley and C. Djerassi (1973) Stereochemical course
of the reduction of 5h-cholest-8(14)-en-7-one by metal-
ammonia; a ready synthesis of 14i-steroids. J. Chem.
Soc. Perkin Trans. 1, 155–161.
[40] D.H.R. Barton, D.O. Jang and J.C. Jaszberenyi (1993)
The invention of radical reactions. 32. Radical deoxy-
genations, dehalogenations, and deaminations with di-
alkyl phosphites and hypophosphorous acid as hydrogen
sources. J. Org. Chem. 58, 6838–6842.
[46] A.U. Siddiqui, S. Swaminathan, F.D. Pinkerton, N.
Gerst, W.K. Wilson, H. Choi and G.J. Schroepfer, Jr.
(1994) Inhibitors of sterol synthesis: synthesis and spectral
properties of derivatives of 3i-hydroxy-25,26,26,26,
27,27,27-heptafluoro-5h-cholest-8(14)-en-15-one
fluori-
nated at carbon 7 or carbon 9 and their effects on
3-hydroxy-3-methylglutaryl coenzyme A reductase activ-
ity in cultured mammalian cells. Chem. Phys. Lipids 72,
59–75.
[47] A.R. van Horn and C. Djerassi (1967) Mass spectrometry
in structural and stereochemical problems. CXXVIII Fur-
ther studies on the mass spectrometry of 15-ketosteroids.
Steroids 9, 163–175.
[48] H.-S. Kim, W.K. Wilson, N.D. Kirkpatrick, F.D. Pinker-
ton, S. Swaminathan, M.D. Hylarides and G.J.
Schroepfer, Jr. (1992) Inhibitors of sterol synthesis.
Chemical synthesis of 7h-ethyl and 16h-ethyl derivatives
of D8(14)-15-oxygenated sterols and their effects on 3-hy-
droxy-3-methylglutaryl coenzyme A reductase activity in
CHO-K1 cells. Chem. Phys. Lipids 62, 55–67.
[49] J.St. Pyrek, W.K. Wilson and G.J. Schroepfer, Jr. (1987)
Inhibitors of sterol synthesis. Spectral characterization of
derivatives of 5h-cholest-8(14)-en-3i-ol-15-one. J. Lipid
Res. 28, 1296–1307.
[41] W.K. Wilson, M.E. Wheeler, F.D. Pinkerton, J.St. Pyrek
and G.J. Schroepfer, Jr. (1991) Inhibitors of sterol synthe-
sis. Characterization of i,k-unsaturated analogs of 3i-
hydroxy-5h-cholest-8(14)-en-15-one and their effects on
3-hydroxy-3-methylglutaryl coenzyme A reductase activ-
ity in CHO-K1 cells. J. Lipid Res. 32, 1215–1227.
[42] Y. Sugo, Y. Inouye and N. Nakayama (1995) Structures
of nine oxygenated 4-methylene sterols from Hachijo
marine sponge Theonella swinhoei. Steroids 60, 738–742.
[43] S. Swaminathan, F.D. Pinkerton, W.K. Wilson and G.J.
Schroepfer, Jr. (1992) Inhibitors of sterol synthesis.
Chemical synthesis and spectral properties of (25R)-5h-
cholest-8(14)-ene-3i,15i,26-triol, a potential metabolite
of 3i-hydroxy-5h-cholest-8(14)-en-15-one and its effects
on 3-hydroxy-3-methylglutaryl-coenzyme A reductase in
CHO-K1 cells. Chem. Phys. Lipids 61, 235–242.
[50] J.E. Herz, W.K. Wilson, F.D. Pinkerton and G.J.
Schroepfer, Jr. (1991) Inhibitors of sterol synthesis. Syn-
thesis and spectral properties of 3i-hydroxy-24-dimethyl-
amino-5h-chol-8(14)-en-15-one and its effects on HMG-
CoA reductase activity in CHO-K1 cells. Chem. Phys.
Lipids 60, 61–69.
[51] Y. Ni, A. Kisic, W.K. Wilson and G.J. Schroepfer, Jr.
(1994) Inhibitors of sterol synthesis. Tritium-labeled 26-
hydroxycholesterol of high specific activity from
a
byproduct of the Clemmensen reduction of diosgenin. J.
Lipid Res. 35, 546–559.
[52] I.J. Massey and C. Djerassi (1979) Structural and stereo-
chemical applications of mass spectrometry in the marine
sterol field. Synthesis and electron impact induced mass
spectral fragmentation of D24 and D24(28)-3i-hydroxy-D5
sterols. J. Org. Chem. 44, 2448–2456.
[44] W.K. Wilson, S. Swaminathan, F.D. Pinkerton, N. Gerst
and G.J. Schroepfer, Jr. (1994) Inhibitors of sterol synthe-
sis. Effects of fluorine substitution at carbon atom 25 of
cholesterol on its spectral and chromatographic properties
[53] S.G. Wyllie and C. Djerassi (1968) Mass spectrometry in
structural and stereochemical problems. CXLVI. Mass
spectrometric fragmentations typical of sterols with unsat-
urated side chains. J. Org. Chem. 33, 305–313.
.
.