Efficient chemoenzymatic synthesis, cytotoxic evaluation, and SAR of epoxysterols

Article abstract of DOI:10.1021/jm9003973

A library of diastereomerically pure epoxysterols, prepared by combining chemical and enzymatic methodologies, was evaluated for cytotoxicity toward human cancer and noncancer cell lines. Unsaturated steroids were oxidized by magnesium bis(monoperoxyphthalate) hexahydrate in acetonitrile, and the resulting epimeric epoxides were enzymatically separated using Novozym 435 or lipase AY. Some of the synthesized epoxysterols have potent cytotoxicity and higher activity on cancer cell lines HT29 and LAMA-84.

Full text of DOI:10.1021/jm9003973

J. Med. Chem. 2009, 52, 4007–4019
4007
Efficient Chemoenzymatic Synthesis, Cytotoxic Evaluation, and SAR of Epoxysterols
Joa˜o F. S. Carvalho,† M. Manuel Cruz Silva,† Joa˜o N. Moreira,‡,§ Se´rgio Simo˜es,‡,§ and M. Luisa Sa´ e Melo*,†
Centro de Estudos Farmaceˆuticos, Laborato´rio Qu´ıımica Farmaceˆutica, Faculdade de Farma´cia, UniVersidade de Coimbra,
3000-548 Coimbra, Portugal, Laborato´rio de Tecnologia Farmaceˆutica, Faculdade de Farma´cia, UniVersidade de Coimbra,
3000-548 Coimbra, Portugal, Center for Neurosciense and Cell Biology, UniVersity of Coimbra, Portugal
ReceiVed March 27, 2009
A library of diastereomerically pure epoxysterols, prepared by combining chemical and enzymatic
methodologies, was evaluated for cytotoxicity toward human cancer and noncancer cell lines. Unsaturated
steroids were oxidized by magnesium bis(monoperoxyphthalate) hexahydrate in acetonitrile, and the resulting
epimeric epoxides were enzymatically separated using Novozym 435 or lipase AY. Some of the synthesized
epoxysterols have potent cytotoxicity and higher activity on cancer cell lines HT29 and LAMA-84.
Introduction
molecular oxygen as oxidant, among other methods, give access
to ꢀ-epoxides. Steroid epoxidation methods were recently
reviewed by Salvador et al.26 However, the majority of these
methods fail to give diastereomerically pure R- and ꢀ-epoxys-
terols.
Biocatalysis has emerged as a valuable strategy to access
epimerically pure compounds by taking advantage of the
stereoselective properties of enzymes.27,28
Oxygenated derivatives of cholesterol, known as oxysterols,
represent a group of biomolecules gaining much attention in
recent years due to their relevant bioactivities.1-4 Oxysterols
have been detected in appreciable quantities in human tissues
and fluids5 and have been implicated in a number of important
biological processes such as regulation of cholesterol homeo-
stasis, atherosclerosis, and cytotoxicity.5-8
Naturally occurring 5R,6R-9-11 and 5ꢀ,6ꢀ-epoxysteroids12,13
have shown relevant cytotoxic effects in tumor cells. For
instance, the 3ꢀ-hydroxy-5ꢀ,6ꢀ-epoxy moiety is present in
natural withanolides like tubocapsenolide A, recently identified
as a potent inhibitor of MDA-MB-231 cell proliferation.13
Epoxides are three-membered cyclic ethers that have specific
reactivity patterns owing to the highly polarized oxygen-carbon
bonds in addition to a highly strained ring.14 Some epoxides
are responsible for electrophilic reactions with critical biological
targets such as DNA and proteins, leading to mutagenic, toxic,
and carcinogenic effects.15-17
Biocatalytic epoxidation can be achieved through oxidizing
enzymes such as monooxygenases (e.g., cytochromes P450s or
nonheme monooxigenases) or haloperoxidases (e.g., chloro-
peroxidases).29,30 P450s are direct asymmetric epoxidation
catalysts, whereas haloperoxidases produce epoxides indirectly
from alkenes by initial halohydrin formation and subsequent
ring closure by halohydrin epoxidase. Several limitations have
hampered their synthetic application, however, a considerable
effort in protein engineering has been done, and enzyme direct
evolution can be in the future the solution to overcome such
limitations.31,32
Other methodologies for the synthesis of epimerically pure
products can be performed by the conjugation of chemical and
enzymatic approaches. In this way, the use of hydrolytic
enzymes can be a useful tool to separate epimeric mixtures of
epoxides.
Specifically, epoxide hydrolases can be effectively used for
hydrolytic kinetic resolution.33 They seem to be a promising
tool for the discrimination of epoxide mixtures because they
act directly on the epoxide ring, being independent of any other
functionality. Furthermore, epoxide hydrolases are cofactor
independent, present good stability, and can be used in the
presence of organic solvents. However its application has been
limited due to the small number of enzymes available, limited
substrate acceptance, and low catalytic efficiency.
On the other hand, lipases are among the most frequently
used enzymes, because they are ease to handle, commercially
available, stable, even in organic solvents, and are cofactor-
independent. Lipases are able to perform enantioselective
hydrolytic reactions and catalyze the formation of a wide range
of ester and amide bonds. Moreover, these enzymes can be used
for the diastereomeric resolution of epoxide mixtures by
acylation/deacylation of a hydroxyl group closely located to the
stereocenter.
Although there are many applications for racemic epoxides,
the demand of enantiomerically pure epoxides has increased in
the recent years. Enantiopure epoxides are useful synthetic
intermediates for the chemical and pharmaceutical industries
because their synthetic versatility renders them valuable inter-
mediates for further enantiocontrolled transformations. They are
essential in the synthesis of important drugs such as (S)-timolol,
a ꢀ-adrenergic blocking agent, indinavir, a HIV protease
inhibitor, and diltiazem, a calcium channel blocker. In the steroid
field, several 4,5-epoxysteroids are intermediates in the synthesis
of biologically active compounds like aromatase inhibitors18,19
and anti-HIV agents20,21 among others.22 Furthermore, ring-
opening of 5R,6R-epoxysteroids gives access to contraceptives,
anti-inflammatory agents, and enzymatic inhibitors.23-25
The stereoselectivity of most steroid epoxidation methods is
dictated by the higher hindrance of the ꢀ-face on the steroid
scaffold. Peroxyacids afford predominantly 5R,6R-epoxides,
while potassium permanganate/inorganic salts, chiral ketones/
oxone, and transition metal complexes in the presence of air or
* To whom correspondence should be addressed. Phone: +351 239
859990. Fax: +351 239 827030. E-mail: samelo@ci.uc.pt.
†
Centro de Estudos Farmaceˆuticos, Laborato´rio Qu´ıımica Farmaceˆutica,
Faculdade de Farma´cia, Universidade de Coimbra.
Our group has studied the ability of lipases to separate
mixtures of 5,6-epoxysteroids through selective acylation of the
3ꢀ-hydroxyl group or deacylation of the acetoxy counterpart.34
‡
Laborato´rio de Tecnologia Farmaceˆutica, Faculdade de Farma´cia,
Universidade de Coimbra.
§
Center for Neurosciense and Cell Biology, University of Coimbra.
10.1021/jm9003973 CCC: $40.75  2009 American Chemical Society
Published on Web 05/27/2009

4008 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
Scheme 1. MMPP Epoxidation of Cholest-5-enes
CarValho et al.
Scheme 2. MMPP Epoxidation of Cholest-4-enes and
Cholest-5-enes
In this study, the stereoselectivity of lipases toward epoxys-
terols was further explored in order to obtain pure epoxides from
the diastereomeric mixture obtained by chemical oxidation.
Compounds with extra hydroxyl groups in the steroid nucleus
were studied under the same discriminative conditions. Fur-
thermore, the hydrolysis of 3ꢀ-acetoxy esters in the presence
of the sensitive oxirane functionality was explored.
Specifically, we report the synthesis of a series of cholestan-
4,5- and -5,6-epoxide mixtures by a fast and high-yielding
magnesium bis(monoperoxyphthalate) hexahydrate (MMPP)
oxidative protocol35 and the enzymatic discrimination of the
diastereomerically pure R- and ꢀ- epoxides using lipases AY
(from Candida cylindracea) and Novozym 435 (from Candida
antarctica) as complementary stereoselective catalysts for the
acylation/deacylation of the epoxysterols. Diastereomerically
pure epoxides were tested for cytotoxicity in two cancer cell
lines (HT29 and LAMA-84) and in a normal fibroblast cell line
(BJ). Some of the epoxides prepared showed high cellular
toxicity, and structure-activity correlations were drawn.
The use of an equimolar amount of MMPP under reflux in
acetonitrile revealed to be a very fast and efficient method to
obtain epoxysterols, as summarized in Table 1. Epoxidation of
compounds 1, 7, 8, 9, and 11 was already explored by our group
using the above-mentioned oxidative procedure. Herein we
move our attention to the synthesis of epoxysterols bearing an
allylic 7-hydroxyl group. 7-Oxygenated steroids are widespread
in mammals, birds, fish, and plants, and there is evidence about
the importance of the 7-hydroxylation pattern for biological
function.41
The preparation of ∆5-7-oxygenated cholesteryl derivatives
started with the allylic oxidation of cholesteryl acetate using
t-butyl-hydroperoxide (tBHP) and copper iodide.42 The de-
scribed procedure was modified by adding tBHP portionwise
(5 portions each, 45 min, during 3 h) to the reaction mixture.
7-oxo-Cholest-5-en-3ꢀ-yl-acetate was isolated by flash chro-
matography in 50% yield. The addition of tBHP portionwised
resulted in higher product yields, probably due to the avoidance
of the free-radical chain decomposition of tBHP.43
The 7R-hydroxy derivatives were obtained by reduction of
the corresponding 7-oxo-derivatives with L-Selectride.44 In our
procedure, the reduction of 7-oxo-cholest-5-en-3ꢀ-yl acetate with
this reagent and additional stirring (35 min.) with aqueous
solutions of NaOH and H2O2 afforded, after flash chromatog-
raphy, the pure 7R-hydroxy-choleste-5-en-3ꢀ-yl acetate (6) and
cholest-5-ene-3ꢀ,7R-diol (5) in 45% and 46% yield, respectively.
7ꢀ-Hydroxy derivatives were easily prepared in a fast and
highly stereoselective manner from 7-ketocholesteryl acetate
using sodium borohydride in the presence of cerium trichlo-
ride.44 The pure 7ꢀ-hydroxy-cholest-5-en-3ꢀ-yl acetate (4) was
obtained in 95% yield. Cholest-5-ene-3ꢀ,7ꢀ-diol (3) was isolated
after alkaline hydrolysis in 90% yield. The same reduction
protocol was used to convert cholestenone into cholest-4-en-
3ꢀ-ol (9), although being less selective.
Results and Discussion
The known bioactivity of natural and synthetic epoxysterols
prompted us to prepare a library of such cholestane derivatives
by combining chemical and enzymatic procedures to further
correlate their structure with cytotoxicity.
Some of the epoxides chemoenzymatically synthesized herein
are oxidative analogues of sterol compounds produced in vivo,
such as 4ꢀ-hydroxycholesterol, a quantitatively important ox-
ysterol in circulation,36,37 and 7R-hydroxycholesterol, which is
the first step product in the most important pathway for bile
acid biosynthesis.38 7ꢀ-Hydroxycholesterol, present in minor
levels in plasma, is related to the presence of oxidatively
modified LDL and oxidative stress.5,39 Moreover, 7ꢀ-hydroxy-
cholesterol and 5ꢀ,6ꢀ-epoxycholesterol derivatives are known
to exert cell death by apoptotic processes in a human monocytic
U937 cell line.40 Therefore, oxysterol epoxides, being direct
oxidative intermediates of endogenously oxysterols, can possibly
have interesting biological roles in vivo.
Chemical Synthesis of Epoxides. A protocol for MMPPa
-
The whole 7-hydroxy-cholest-5-ene derivatives (3, 4, 5, and
6) prepared were then oxidized using MMPP in acetonitrile
under reflux. It is well recognized the syn-directing effect of
allylic hydroxyls, both in acyclic45 and cyclic compounds as
steroids,46 in the outcome of epoxidation stereoselectivity.
Herein, we have observed the same effect, specifically in the
epoxidation of cholest-5-ene-3ꢀ,4ꢀ-diol (7), where the introduc-
tion of an allylic ꢀ-hydroxy group in ring A, in comparison to
cholesterol, results in the higher formation of the ꢀ-epoxide
(entries 1 and 8, Table 1). Thus it is expected that an allylic
mediated epoxidation in acetonitrile under reflux, recently
developed by our group,35 was applied to access the epoxysterols
shown in Schemes 1 and 2.
a Abbreviations: MMPP, magnesium bis(monoperoxyphthalate) hexahy-
drate; LDL, low density lipoprotein; tBHP, tert-butyl-hydroperoxide; MDR,
multidrug resistance; N435, Novozym 435; LAY, lipase AY; VAC, vinyl
acetate; OCT, n-octanol; ATCC, American Type Culture Collection; DSMZ,
Deutsche Sammlung von Mikroorganismen and Zellkulturen GmbH; PS,
penicillin and streptomycin; CM, culture medium; iFBS, inactivated fetal
bovine serum.

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4009
Table 1. MMPP Epoxidation of ∆4- and ∆5-Cholestenes
entry
substrate (mmol)
MMPP (mmol/mmol)a
solvent (mL/mmol)
temp
time
product
yield (%)b
epoxide ratio R:ꢀ (%)c
1
2
3
4
5
6
7
8
1/1.200
3/0.248
3/0.507
4/0.206
4/0.235
5/0.175
6/0.180
7/0.550
8/0.550
9/0.646
11/0.268
1.1
2.5
1.1
2.6d
1.1
1.1
1.1
1.1
6.1e
1.1
1.1
15/CH3CN
15/CH2Cl2/H2Of
30/CH3CN
7.5/CH2Cl2/H2Of
30/CH3CN
30/CH3CN
30/CH3CN
15/CH3CN
50/CH3CN
30/CH3CN
15/CH3CN
reflux
rt
reflux
rt
reflux
reflux
reflux
reflux
reflux
reflux
reflux
10 min
53 h
1a,1b
3a,3b
3a,3b
4a,4b
4a,4b
5a,5b
6a,6b
7a,7b
8a,8b
9a,9b
11a,11b
83
92
89
83
86
78
82
82
93
83
80
78:22
52:48
44:56
56:44
50:50
89:11
86:14
43:57
76:24
22:78
99:1
15 min
12 days
25 min
25 min
40 min
25 min
6 h
9
10
11
10 min
2 min
a mmol of MMPP per mmol of substrate. b Isolated yield by flash chromatography. c Calculated by 1H NMR integration in the crude product. d Addition
of MMPP portionwised until complete substrate consumption: initial addition of 1.1 mmol of MMPP, with subsequent additions of 0.5 mmol each 3 days.
e Addition of MMPP portionwised (1.017 mmol per h) until complete substrate consumption. f Addition of 100 µL of H2O per mmol of MMPP.
Table 2. Enzymatic Discrimination of Epoxycholestanesa
entry
substrate (mg)
epoxide ratio R:ꢀ (%)b
enzymec (mg)
reagentd(mL)
time (days)
producte
1a (77%) + 2b (20%)
1a,1b (18%, 13:87) + 2a,2b (80%, 94:06)
1b (16%) + 2a,2b (82%, 91:09)
3a (44%) + 4b (55%)
3b (29%) + 4a,4b (70%, 64:36)
3b (44%) + 4a,4b (53%, 79:21)
3a,3b (61%, 75:25) + 4a,4b (37%, 09:91)
5a (87%) + 6b (9%)
7a (42%) + 13b (31%) + 14b(24%)
7b (10%) + 13a,13b (88%, 87:13)
8a,8b (98%, 76:24)
1
2
3
4
5
6
7
8
1a,1b/75
1a,1b/140
1a,1b/130
3a,3b/81
3a,3b/35
3a,3b/60
4a,4b/75
5a,5b/50
7a,7b/176
7a,7b/200
8a,8b/35
9a,9b/110
9a,9b/90
78:22
78:22
78:22
44:56
44:56
44:56
50:50
89:11
43:57
76:24f
76:24
21:79
21:79
N435/42
VAC/0.5
VAC/1
VAC/2
1
1
3
1
3
1
4
1
7
1
7
5
14
LAY/218
LAY/880
N435/180
LAY/65
LAY/53
LAY/450
N435/55
N435/1505
LAY/305
LAY/400
N435/1434
LAY/1040g
VAC/1.0
VAC/0.35
VAC/0.6
OCT/0.2
VAC/0.35
VAC/2
VAC/0.35
OCT/0.08
VAC/1.3
VAC/1
9
10
11
12
13
9a (20%) + 10b (76%)
9a, 9b (74%, 19:81) + 10a,10b (24%, 32:68)
b
a Reactions performed in toluene and shaken at 200 rpm at 45 °C. Calculated by H NMR integration. c Abbreviation for Novozym 435: N435; and
lipase AY: LAY. d Abbreviation for vinyl acetate: VAC; and octanol: OCT. e Isolated yield by flash chromatography; diastereomeric mixtures calculated by
1H NMR integration. f Obtained through the mild chemical deacylation hydrolysis using an excess of sodium carbonate in methanol of the epoxide mixture
8a,8b. g Portionwise addition of lipase AY: 512 mg each week.
1
hydroxyl group in ring B of the steroid (compounds 3, 4, 5,
and 6) gives rise to the same effect. We have reported35 that
moist MMPP (1.1 mmol per mmol of substrate) in dichlo-
romethane at room temperature affords 5,6-epoxides with high
stereoselectivity although with long reaction times. Nevertheless,
7-hydroxy derivatives had not been studied yet under such
reaction conditions. 7ꢀ-Hydroxy-cholest-5-en-3-yl acetate (4)
was submitted to moist MMPP/CH2Cl2, at room temperature,
and the result was a small amount of epoxide after 3 days, as
assessed by TLC analysis. Therefore we made consecutive
additions of MMPP (0.5 mmol per mmol of substrate each 3
days) until complete substrate consumption. With this approach,
an amount of 2.6 mmol of MMPP for mmol of substrate was
needed to complete the reaction in 12 days. The reaction gave
an epoxide mixture of 56:44 R/ꢀ in 83% isolated yield (entry
4, Table 1). Applying the acetonitrile/reflux protocol, the
reaction rate was significantly increased, being complete in 25
min and only a 10% excess addition of MMPP was required.
The reaction afforded a 50:50 diastereomeric mixture of
epoxides in a high yield (entry 5, Table 1). Summarizing, not
only the reaction was much faster, as it has been high yielding
with less oxidant requirement and with no significant difference
in terms of stereoselectivity.
It is also possible to verify that the faster the epoxidation
reaction is, the bigger the amount of the hydroxyl syn-directed
effect product is formed.
When a 7R-hydroxyl substituent is present (compounds 5 and
6), the reaction outcome changes dramatically, with a predomi-
nance of the R-epoxide (higher than 85%), as expected (entries
6 and 7). The R-epoxide formation was greater than in the case
of cholesterol 1. The reaction rate was lower when compared
to the epoxidation of the 7ꢀ-derivatives. Therefore, although
some hydroxyl syn-directed effect can be observed, an additional
R-face blockage by the 7R-hydroxyl is imposed to the peracid,
hampering the reactivity.
Stereoselective Lipase Esterification of 3ꢀ-Hydroxy-
epoxysterols. Steroidal epoxide mixtures, due to their close
elution rates, are difficult to separate and isolate with good
yields, even when highly stereoselective epoxidation methods
are applied. A convenient approach to overcome such difficulty
is the use of combined chemical and enzymatic procedures.
To obtain isomerically pure epoxysterols, we explored a
biocatalytic approach developed by our group34 in which
epimeric epoxides were separated by irreversible lipase catalyzed
acylation in organic media. Specifically, Novozym 435 was able
to convert in a stereoselective manner 5ꢀ,6ꢀ-epoxycholesterol
(1b) into the corresponding 3ꢀ-acetate derivative, leaving
unreacted the major isomer, the R-epoxysteroid (1a). As a result,
5R,6R-epoxycholesterol (1a) and 5ꢀ,6ꢀ-epoxycholesteryl acetate
(2b) were obtained in excellent isolated yields (77% and 20%,
respectively) and 100% diastereomeric excess, as confirmed by
1H NMR analysis (entry 1, Table 2).
By exploring the reactivity of an homoallylic derivative of
compound 4, the cholest-5-en-3ꢀ,7ꢀ-diol (3), the same conclu-
sions can be drawn. Moist MMPP/CH2Cl2, added in a single
portion (2.5 mmol per mmol of substrate), at room temperature,
the reaction needed 2 days to reach completion, producing an
epoxide mixture of 3a,3b, 52:48 R/ꢀ ratio (entry 2, Table 1).
In reflux acetonitrile, the reaction was complete in 15 min,
rendering an epoxide mixture with a slight predominance of
the ꢀ-epoxide (44:56 R/ꢀ ratio, entry 3, Table 1).
Complementarily, when epimeric epoxysterols 1a and 1b
were used as substrates for lipase AY catalyzed acylation, the
R-epoxide was the preferred substrate by the enzyme but some

4010 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
CarValho et al.
Scheme 3. Enzymatic Transesterification of 3ꢀ-Hydroxy-5,6-epoxycholestanes
Scheme 4. Chemoenzymatic Discrimination of 4,5- and 5,6-Epoxycholestanes
ꢀ-isomer was also acylated. Lipase AY proved to be less
selective and reactive than Novozym 435. In fact, after 1 day,
no diastereoselection could be achieved (entry 2, Table 2) and
3 days were needed to obtain 100% diastereomeric excess of
the less reactive epimer 5ꢀ,6ꢀ-epoxycholesterol (1b) (entry 3,
Table 2).
In the presence of an additional 7ꢀ- (compounds 3a and 3b)
or 7R-hydroxyl group (compounds 5a and 5b), Novozym 435
showed total regioselectivity for the 3ꢀ position and the same
stereoselectivity for acylating the ꢀ-epoxysteroid was observed
(Scheme 3). 3ꢀ,7-Dihydroxyl-5R,6R-epoxy-cholestane deriva-
tives (3a and 5a) and the corresponding 3ꢀ-acetoxy-5ꢀ,6ꢀ-epoxy
derivatives (4b and 6b) were isolated in excellent yields and
100% diastereomeric excess (entries 4 and 8, Table 2).
On the other hand, the presence of a second hydroxyl group
at C-4 provides different results using Novozym 435. As shown
in Scheme 4, 3ꢀ,4ꢀ-dihydroxy-epoxysterols 7a and 7b were very
slowly converted by Novozym 435 into a mixture of compounds.
Clearly, Novozym was still selective for the acylation of the
3ꢀ-OH in the ꢀ-epoxide. However, acylation at 4ꢀ-OH also
occurred. After 7 days, the 5R,6R-epoxy-diol (7a), the 3ꢀ-
acetoxy-5ꢀ,6ꢀ-epoxy derivative (13b), and the 4ꢀ-acetoxy-
5ꢀ,6ꢀ-epoxy derivative (14b) were recovered isomerically pure
by flash chromatography (entry 9, Table 2).
We have previously reported the lipase-catalyzed acylation
of 2,3- and 3,4-vicinal diols of steroids.47 3ꢀ,4ꢀ-Dihydroxys-
teroids were used as substrates for a panel of enzymes and,
although Candida rugosa lipase and lipase AY were the best
enzymes, Novozym 435 also accepted the diol. Herein, the
presence of an additional 5,6-epoxy group hinders the access
to Candida rugosa lipase and lipase AY active sites. The results
obtained with Novozym 435 agree with our previous results:
preference for the 3ꢀ-OH instead of 4ꢀ-OH47 and preference
for the ꢀ-epoxide.34 However, the presence of other functional
groups close to the 3ꢀ-OH nucleophile renders the reaction
slower.
Using lipase AY as catalyst, we studied the acylation of an
epimeric mixture of 5,6-epoxycholestane-3ꢀ,7ꢀ-diol (3a,3b 44:

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4011
Table 3. Mild Enzymatic Alcoholysis and Transesterifications of 5,6-Epoxycholestanesa
entry
substrate (mg)
enzyme (mg)
reagent (mL)
time
productb
1
2
3
4
5
6
1a/30
4b/50
6a/40
11a/60
12a/45
13b/30
lipase AY/60
vinyl acetate/0.3
octanol/0.15
octanol/0.15
vinyl acetate/0.6
octanol/0.15
octanol/0.1
2 days
20 days
4 days
1 day
3 days
20 days
2a (94%)
3b (56%)
5a (94%)
12a (97%)
11a (96%)
7b (52%)
Novozym 435/30
lipase AY/400
lipase AY/120
lipase AY/450
Novozym 435/90
a Reactions performed in toluene and shaken at 200 rpm at 45 °C. b Isolated yield by flash chromatography.
56). After 3 days, complete conversion of the R-epoxy derivative
was observed. However, by TLC control, the reaction seemed
to be running faster than the one for 5,6-epoxycholesterols,
1a,1b. In fact, after 3 days, 5ꢀ,6ꢀ-epoxycholestane-3ꢀ,7ꢀ-diol
(3b) was successfully isolated isomerically pure, but in a low
yield (29%), due to the acylation of the ꢀ-epimer along with
the R (entry 5, Table 2). Consequently, we repeated the reaction
using a lesser amount of enzyme and left it running for only
one day (entry 6, Table 2). This procedure resulted in a great
increase in the amount of the ꢀ-epoxide derivative isolated
(44%).
Then we studied the lipase AY acylation of epoxide mixtures
with an additional ꢀ-allylic hydroxyl substituent in ring A. A
mixture of 5,6-epoxycholestane-3ꢀ,4ꢀ-diol (7a,7b 76:24) was
separated using the same amount ratio of lipase AY found to
obtain a good percentage recovery of the 5ꢀ,6ꢀ-epoxycholes-
tane-3ꢀ,7ꢀ-diol (3b) in one day. In effect, isolation of 10% pure
5ꢀ,6ꢀ-epoxycholestane-3ꢀ,4ꢀ-diol was achieved after 1 day of
reaction (entry 10, Table 2).
(6a) and 5R,6R-epoxy-7-norcholestan-3ꢀ-yl-acetate (12a), af-
fording quantitative hydrolysis in 4 and 3 days, respectively
(entries 3 and 6, Table 3).
Stereoselective deacylation of 3ꢀ-acetoxy-5,6-epoxysteroids
was already achieved with the use of lipase AY and octanol as
acyl acceptor, exploring its R-stereopreference in epoxide
mixtures where the R-epoxide was the minor epimer. Three days
were needed to isolate the corresponding 3ꢀ-acetoxy-5ꢀ,6ꢀ-
epoxy-derivatives in 100% diastereomeric excess.34 We have
explored lipase AY’s ability to discriminate acetoxy-epoxy
mixtures by selective alcoholysis of 3ꢀ-acetoxy-5,6-epoxys-
teroids containing an allylic 7-hydroxyl substituent. As men-
tioned above, we had noticed that alcoholysis of 7R-hydroxy-
5R,6R-epoxycholestan-3ꢀ-yl-acetate (6a) was quantitative in 4
days. To perform the stereoselective deacylation of 7ꢀ-hydroxy-
5,6-epoxy-cholesteryl-3ꢀ-acetate mixture (4a,4b, 50:50), we
used a 12-fold amount of the lipase and stopped the reaction in
4 days (entry 7, Table 2). Although the TLC analysis of the
reaction suggested a successful separation of the pure 4b
1
Because of the close proximity of the epoxide and the
nucleophile for enzyme acylation, ring A epoxysterols consti-
tute interesting substrates to perform such enzymatic reactions.
We intended to evaluate the ability of both enzymes to acylate
the 3ꢀ-hydroxyl group when a neighboring 4,5-epoxy function-
ality is present. Although the isomeric 4R,5R- and 4ꢀ,5ꢀ-epoxy-
3ꢀ-hydroxysterols 9a and 9b were distinguished by Novozym
435, which again showed preference for the acylation of
ꢀ-epoxide (entry 12, Table 2), lipase AY revealed to be
inefficient, with practically no stereoselectivity and low reactivity
(entry 12, Table 2). Comparing to the 5,6-epoxysterols 1a and
1b, the presence of the epoxide in the 4,5-position hinders the
access of the nucleophilic C3-OH to the acyl enzyme, resulting
in longer reaction times (entries 1 and 12; 3 and 13, Table 2).
Finally, we decided to investigate the acylation of 5R,6R-
epoxy-7-norcholestan-3ꢀ-ol (11a) by lipase AY. Comparing to
5R,6R-epoxycholesterol (1a), the 5-membered B-ring of 11a
has a positive effect on the reaction rate (entries 1 and 5, Table
3), and after one day of reaction, 12a was isolated in 97%.
Lipase Catalyzed Deacetylation of 3ꢀ-Acetoxy-epoxys-
terols. Lipase catalyzed deacetylation of 3ꢀ-acetoxy-epoxys-
terols was explored using n-octanol as nucleophile. The
deacylation of pure 4b, 8b, and 10b was pursued using
Novozym 435 catalyzed alcoholysis in toluene. The respective
epoxyalcohols 3b, 7b were obtained in low isolated yields (about
50%, entries 2 and 4, Table 3) after a long reaction time (20
days). 4ꢀ,5ꢀ-Epoxycholestan-3ꢀ-ol (9b) was difficult to separate
by flash chromatography from n-octanol. Thus we used another
acyl acceptor, n-pentanol, which can be eliminated by vacuum,
but the reaction rate was even lower. For this reason, we adopted
a mild chemical hydrolysis procedure using sodium carbonate
in methanol for the synthesis of 9b, which was successfully
isolated in 96% yield after 24 h of reaction.
compound by complete deacylation of the R-epimer 4a, H
NMR of the two isolated fractions proved that complete
diastereoselectivity was not achieved and longer time would
be needed. Moreover significant deacylation of the ꢀ-epimer is
observed, which makes lipase AY not suitable for high yielding
epoxide diastereoselection.
In addition, lipase AY did not deacylate a mixture of 5,6-
epoxycholestane-3ꢀ,4ꢀ-diyl-diacetates (10a,b) (entry 11, Table
2, Scheme 5) due to the steric hindrance imposed by the two
bulky acetate substituents.
Biological Evaluation. Taking advantage of the chemoen-
zymatic approach described above, several isomerically pure
epoxycholestanes were prepared. Compound purity was con-
trolled by gas chromatography and proved to be >95% for all
compounds, except for 5ꢀ,6ꢀ-epoxycholestane-3ꢀ,4ꢀ-diol (7b),
which undergoes thermal degradation in the conditions used for
analysis.
The evaluation of the cytotoxicity was performed on the sterol
starting materials, compounds 1, 3, 4, 5, and 7 and the
corresponding epoxysteroids 1a, 1b, 3a, 3b, 5a, 7a, 7b, 11, 10a,
10b, 11a, 13b, and 14b. Cytotoxicity was evaluated by the
Alamar Blue assay based on the cellular reduction of the blue
resazurin to the pink resorufin by the living cells.48,49
The in vitro cell toxicity effects mediated by the different
synthesized compounds were studied on HT29 (human colon
adenocarcinoma; adherent epithelial cell line) and LAMA-84
(human chronic myeloid leukemia in blast crisis; suspended/
slowly adherent cell line) cancer cell lines, as well as on
noncancer cells BJ (human foreskin fibroblast; adherent cell
line). Compounds were primarily screened in the HT29 cancer
cell line due to the straight relation of sterols and colon tissues,
where cholesterol is both absorbed and synthesized.50,51 The
compounds exhibiting higher cytotoxic effects were further
tested in the other cells.
On the other hand, lipase AY proved to be efficient in the
alcoholysis of 7R-hydroxy-5R,6R-epoxycholestan-3ꢀ-yl-acetate

4012 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
CarValho et al.
Scheme 5. Enzymatic Alcoholysis of 3ꢀ-Acetoxy-5,6-epoxycholestanes
a
Table 4. Cytotoxicity of the Compounds Synthesized (IC50 in µM), in
3a and 3b. Although the ꢀ-epoxide 3b is more active than 3a
on HT29 cells, less activity toward LAMA-84 and BJ was
observed.
Three Human Cell Lines after 48 h of Exposure
entry
compound
HT29
LAMA-84
BJ
The results indicate that the presence of a second hydroxyl
group at position C-7 has an impact on bioactivity. While
cholesterol 1 is not cytotoxic, its 7ꢀ-OH and 7R-OH derivatives
are cytotoxic (compounds 3 and 5, respectively). In the presence
of the epoxide group, the cytotoxicity also increases with the
addition of a hydroxyl at C-7. As shown in Table 4, the
R-epoxides 5a (with a 7R-OH group) and 3a (with a 7ꢀ-OH
group) are more cytotoxic than epoxycholesterol 1a. Moreover,
the ꢀ-epoxide 3b is 4.5-fold more cytotoxic than compound 1b.
Comparing the 3ꢀ,7-dihydroxy-5R,6R-epoxysterols 3a and
5a, the stereochemistry of the hydroxyl group at C-7 influences
the activity, the 7ꢀ-OH derivative 3a being almost 3-fold more
cytotoxic in the HT29 cell line.
On the other hand, the comparison of 4ꢀ-hydroxy-5ꢀ,6ꢀ-
epoxycholesterol 7b with its derivative 14b clearly indicate that
the presence of the 4ꢀ-acetate does not affect significantly the
activity toward these cell lines. However, the protection of the
3ꢀ-hydroxyl group with an acetate, 13b, leads to a significant
decrease in cytotoxicity. The same conclusion can be drawn
when comparing the oxysterols 3 and 4, where again the
introduction of the acetate substituent in position C-3 does not
favor cytotoxicity.
The presence of an additional 4ꢀ-hydroxy group in the
cholesterol structure does not confer cytotoxicity (compound
7). However, the R- and ꢀ-epoxy derivatives of 7 are cytotoxic
(compounds 7a and 7b). Nonetheless, the 4ꢀ-hydroxy group
has a lower effect in the cytotoxicity of the epoxysterols as
compared to the 7R- and 7ꢀ-hydroxy derivatives.
The position of the epoxide in the steroid nucleus also has
an impact on the cytotoxicity, 9b (a 4ꢀ,5ꢀ-epoxide) being more
active than 1b (a 5ꢀ,6ꢀ-epoxide) in the HT29 cell line. However,
compounds 9a and 1a, both R-epoxides at 4,5- and 5,6-positions,
respectively, have comparable low toxicities.
1
2
3
4
5
6
7
8
1
1a
1b
3
3a
3b
4
5
5a
7
7a
7b
9
9a
9b
11a
>50
>50
37.6
6.4
13.3
8.3
29.4
15.5
33.8
>50
38.2
17.7
>50
61.9
22.6
40.9
28.2
15.9
21.9
1.15
4.8
4.1
7.0
20.1
13.4
19.5
9
10
11
12
13
14
15
16
17
18
19
20
8.0
20.4
31.4
11.2
13b
14b
cisplatin
doxorubicin
8.1
9.0
0.76
14.4
2.7
a IC50, the concentration that inhibits 50% of cellular growth. Data are
presented as the mean of at least three separate experiments for each cell
line.
As shown in Table 4, the compounds exhibited different
degrees of cytotoxicity toward the cell lines tested. The
oxysterols synthesized induce higher cytotoxicity in the LAMA-
84 cell line as compared to HT29 cells, as displayed in Figure
1. The selectivity was studied by incubating the compounds with
both cancer (HT29 and LAMA-84) and noncancer cell lines
(BJ) and, generally, they are more cytotoxic toward cancer cell
lines than normal cells, except compound 3a, which exhibited
similar effects on both HT29 and BJ cells (Table 4, entry 5 and
Figure 1b).
The introduction of an epoxy moiety in the steroid nucleus
led to different results. In the case of nonsubstituted ring B ∆4-
and ∆5-cholestenes, the epoxidation of the molecules leads, in
general, to increased cytotoxicity. Indeed, epoxides 1b, 11a,
and 11b are cytotoxic, whereas the corresponding olefins 1 and
11 are not cytotoxic, although epoxide 1a is also not cytotoxic.
However, when a hydroxyl group is present in position C-7
(compounds 3 and 5), the introduction of the epoxy moiety
decreases activity (compounds 3a, 3b, and 5a).
Finally, comparing the influence of the ring B size, the
R-epoxide functionality seems to be more favorable in a five-
membered ring, as shown in Table 4. Compound 11a is more
cytotoxic than 1a and proximally cytotoxic to 1b.
A brief and schematic representation of these structure-activity
relations for cytotoxicity is presented in Figure 2.
Cisplatin and doxorubicin are two of the most effective
chemotherapeutic agents used nowadays for the treatment of a
broad range of cancers52 and were used as reference drugs for
the cytotoxic evaluation. Our results indicate that LAMA-84
cells are more sensitive to these drugs as compared to HT29 or
The stereoisomery of the epoxides is important for bioactivity.
ꢀ-Epoxides 1b, 7b, and 9b are more active than the correspond-
ing R-epoxides 1a, 7a, and 9a. The exceptions are the epoxides

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4013
Figure 1. Dose-dependent effect of oxysterols (3, 3a, 3b, 9b) on cell viability of different cell lines (cancer: HT29, LAMA-84, and noncancer:
BJ). Data shown are mean ( standard error of at least three independent experiments.
In fact, cholesterol is highly abundant in humans and shows no
cytotoxicity. However, a simple oxidation can transform cho-
lesterol into a cytotoxic molecule, 1b, and the same for the
conversion of 4ꢀ-hydroxycholesterol, 7, one of the most
abundant oxysterols in humans, into 7b.
The synthesis of oxysterols involves a panel of chemical
reactions, mainly oxidative. A recent methodology developed
by us to access epoxysteroids in a fast and straightforward
manner35 was further investigated herein.
The biocatalytic acylation/deacylation at the position C-3 of
5R,6R- and 5ꢀ,6ꢀ-epoxysterols was previously reported by our
group.34 This biocatalytic approach was extended and has found
a useful application to obtain pure bioactive sterols. The
stereoselective enzymatic acylation of 4R,5R- and 4ꢀ,5ꢀ-epoxy-
3ꢀ-hydroxysteroids has been studied for the first time. Further-
Figure 2. Graphical depiction of the SAR for cell cytotoxicity based
on the IC50 results.
BJ cells and as is the case for the oxysterol compounds
synthesized. In fact, cisplatin has a similar cytotoxic potency
more, the stereo- and regioselective acylation of dihydroxy-
epoxysterols was explored, disclosing the ability of the lipases
for both HT29 and BJ cells, besides colorectal carcinomas, one
of the leading causes of cancer-related deaths in U.S.,53 are
known to be resistant to chemotherapy, namely involving
cisplatin or even doxorubicin because they generally express
high levels of the MDR1 gene that encodes the P-glycoprotein
(Pgp) responsible for the extracellularly drug transportation.54
Thus oxysterols 3, 3b, 7b, and 9b constitute a group of
compounds with interesting cytotoxic properties due to their
higher activity on cancer cells. Moreover, compounds 3 and
3b showed a better cytotoxic profile than cisplatin for all cells
tested.
tested to act on the 3ꢀ-position of steroids with striking
regioselectivity by leaving unreacted the C-7 hydroxyl groups.
Novozym 435 was shown to be the ideal enzyme to perform
the stereoselective discrimination of several epoxysterols by
acylating the ꢀ-epoxy isomer.
The bioactivity of the epoxysterols obtained was studied by
evaluating their cytotoxic potential in two cancer cell lines and
in a noncancer cell line. The results obtained show that these
molecules are cytotoxic in a dose-dependent manner at micro-
molar range and have higher activity on the tumor cells (Table
4; Figure 1), being the bioactivity strongly affected by the
stereochemistry of the 4,5- and 5,6-epoxide and the surrounding
hydroxyl groups at 3-, 4- and 7-positions (Figure 2).
Conclusion
Oxysterols have become increasingly recognized as bioactive
compounds and much work has been done in order to identify
cellular receptors for oxysterols and to correlate them with novel
in vivo functions.4
Oxysterols can be generated endogenously by simple autoxi-
dation of cholesterol or by enzymatic reactions involving specific
hydroxylases, and it is postulated that the actions of these
compounds are mediated by oxysterol binding proteins.55
Therefore, could not oxysterols and their epoxy derivatives be
important local regulators of cell growth and/or death in vivo?
7ꢀ-Hydroxycholesterol (3) is a widely studied oxysterol,
mainly due to its possible involvement in the atherosclerosis
process. Indeed, 7ꢀ-hydroxycholesterol is one of the main
cytotoxic oxysterols present in oxidized LDL, foam cells, and
fatty streaks.56 Moreover and among the oxysterols studied, 7ꢀ-
hydroxycholesterol is known to be one of the most potent
inducers of cell death in cultured cell lines,57,58 which are
normally related to the apoptotic pathway.59,60 Our results
corroborate the reported cytotoxic activity of 7ꢀ-hydroxycho-

4014 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
CarValho et al.
lesterol.58 In addition, we show that this oxysterol is more active
toward tumor cell lines than to the normal cell line BJ.
Recent studies describe the cytotoxicity of epoxysteroids such
as 5R,6R-epoxycholesterol (1a)61 and 5ꢀ,6ꢀ-epoxycholesterol
(1b)62 and the potential of induction of apoptotic cell death.63
Nonetheless, there are no reports on the biological activities of
synthetic epoxy derivatives of the common oxysterols such as
4ꢀ-hydroxycholesterol and 7R- or 7ꢀ-hydroxycholesterol.
In summary, we describe herein an efficient chemoenzymatic
methodology for the preparation of isomerically pure epoxys-
terols. A recent protocol using MMPP allowed the efficient
synthesis of a library of epoxysterols. The biocatalytic discrimi-
nation of the resulting epimeric epoxysterols gave access to pure
isomers through mild and efficient lipase-catalyzed acylations
and deacylations. From the library synthesized 5ꢀ,6ꢀ-epoxyc-
holestane-3ꢀ,7ꢀ-diol, 3b revealed to be the most cytotoxic and
to display higher activity toward tumor cells. In fact, compounds
3b and 3 exhibited very similar cytotoxic and selectivity
behavior as it can be observed in the curve of the dose-dependent
effect on cell viability (Figure 1), suggesting that both com-
pounds share the same mechanism responsible for cell death in
the cell lines tested. To the best of our knowledge, the
cytotoxicity of compound 3b had not been studied before.
The cytotoxic evaluation of several epoxy derivatives revealed
structure-activity relationships for this type of sterols as
summarized briefly in Figure 2.
last addition, the reaction mixture was stirred at reflux temperature
for another 1 h. Then reaction solvent was concentrated under
vacuum and the residue dissolved in diethyl ether and filtered under
celite to eliminate the copper iodide. Organic phase was left stirring
with Na2SO3 (10% aq soln) for 2 h and then extracted with diethyl
ether. The resulting organic phase was washed with NaHCO3 (sat.
aq soln) and water, dried over anhydrous Na2SO4, and concentrated
under vacuum. 7-oxo-Cholest-5-en-3ꢀ-yl acetate (1027 mg, 51%)
was isolated using column chromatography (petroleum ether, ethyl
acetate 10:1) IR (film) 3024, 2953, 2867, 1734, 1669, 1471, 1373,
1
1268, 1239, 1041, 751 cm-1. H NMR (300 MHz, CDCl3) ppm
0.68 (3H, s, 18-CH3), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz,
26-CH3 and 27-CH3), 0.92 (3H, d, J ) 6.6 Hz, 21-CH3), 1.21 (3H,
s, 19-CH3), 2.05 (3H, s, 3ꢀ-CH3COO), 4.72 (1H, ddt, J ) 11.5,
11.5, 5.4, 4.4 Hz, 3R-H), 5.70 (1H, d, J ) 1.7 Hz, 6-H). 13C NMR
(75 MHz, CDCl3) ppm 11.9, 17.2, 18.8, 21.1 (CH2), 21.2, 22.5,
22.8, 23.7 (CH2), 26.2 (CH2), 27.3 (CH2), 27.9, 28.5 (CH2), 35.7,
35.9 (CH2), 36.1 (CH2), 37.7 (CH2), 38.2 (C-10), 38.6 (CH2), 39.4
(CH2), 43.0 (C-13), 45.3, 49.7, 49.9, 54.7, 72.1, 126.6 (C-6), 163.8
(C-5), 170.2 (CH3COO), 201.9 (C-7). Then, to a solution of 7-oxo-
cholest-5-en-3ꢀ-yl acetate (677 mg; 1.529 mmol) in THF (8 mL)
and MeOH (4 mL) containing CeCl3 ·7H2O (575.8 mg; 1.529 mmol)
was added NaBH4 (144.6 mg; 3.822 mmol). The mixture was stirred
at rt for 15 min and concentrated under vacuum. The residue was
dissolved in diethyl ether, washed sequentially with HCl (5% aq
soln), NaHCO3 (10% aq soln) and water, dried over anhydrous
MgSO4, filtered, and evaporated to yield the 7ꢀ-hydroxycholest-
5-en-3ꢀ-yl acetate, 4 product. Flash chromatography (petroleum
ether, ethyl acetate 2:1) afforded the pure compound (650 mg; 96%)
mp 110-111 °C (EtOH); lit.65 109-110 °C. IR (film) 3326, 2948,
1
2872, 1734, 1471, 1379, 1245, 1036, 757 cm-1. H NMR (300
Experimental Section
MHz, CDCl3) ppm 0.69 (3H, s, CH3-18), 0.86 and 0.87 (each 3H,
2d, J ) 6.6 Hz, 26-CH3 and 27-CH3), 0.92 (3H, d, J ) 6.6 Hz,
21-CH3), 1.06 (3H, s, 19-CH3), 2.04 (3H, s, 3ꢀ-CH3COO), 3.86
(1H, td, J ) 7.9, 2.0, 2.0 Hz, 7R-H), 4.62 (1H, ddt, J ) 10.7, 10.7,
6.4, 4.2 Hz, 3R-H), 5.31 (1H, dd, J ) 2.0, 2.2 Hz, 6-H). 13C NMR
(75 MHz, CDCl3) ppm 11.8, 18.7, 19.0, 21.0 (CH2), 21.4, 22.5,,
22.8, 23.8 (CH2), 26.3 (CH2), 27.7 (CH2), 28.0, 28.5 (CH2), 35.7,
36.2 (CH2), 36.5 (C-10), 36.6 (CH2), 37.6 (CH2), 39.4 (CH2), 39.5
(CH2), 40.7, 42.9 (C-13), 48.1, 55.4, 55.8, 73.2, 73.4, 126.2 (C-6),
142.3 (C-5), 170.5 (CH3COO). MS (ESI) m/z (%): 443.5 (100)
[M - H]+, 429.3 (48), 371.9 (58), 368.3 (77), 331.7 (53), 268.9
(68), 221.0 (61).
Chemistry. General Methods. Cholesteryl acetate, cholesterol,
and cholestenone were purchased from Sigma-Aldrich Co. The other
starting materials were prepared as reported in literature35 and as
described below in the following sections. All reagents and solvents
were purchased from Sigma-Aldrich Co. All commercial available
chemicals were used without purification. Lipase AY and Novozym
435 were obtained from Amano and Sigma, respectively. Enzymatic
reactions were performed in a New Brunswick Scientific C24
Incubator Shaker. TLC monitoring was done using Kieselgel
60HF254/Kieselgel 60G and revealed with Komarowsky’s reagent.64
Flash column chromatography was performed in a Bu¨chi automated
system using a Borosilikat 3.3 column and silica gel 60 (230-400
mesh ASTM). Melting points were determined on a Bu¨chi melting
point B-540 and are uncorrected. IR spectroscopy was performed
on a Jasco FT/IR 420 spectrophotometer. 1H, 13C, and DEPT NMR
spectra were recorded on Bruker Avance 300 MHz and 500 MHz.
Mass spectra were recorded on a Thermo Finnigan mass spectrom-
eter, model LCQ Advantage MAX. The degree of the epoxysteroids
purity was controlled by gas chromatography carried out on an
Agilent Technologies 6850 Network GC system using an Agilent
19091Z-41 3E HP-1 methyl siloxane capillary (30.0 m × 320 µm
× 0.25 µm) and with a flux mixture of H2 (40 mL/min), N2 (45
mL/min), and air (450 mL/min). Compounds were dissolved in
chloroform for injection. Elution run started in 100 °C for one min
and increased 35 °C/min until 300 °C. Retention times of the epoxy
compounds were between 10 and 15 min. Purity was confirmed to
be >95% for all compounds, except for 5ꢀ,6ꢀ-epoxycholestane-
3ꢀ,4ꢀ-diol (7b), which undergoes thermal degradation in the
conditions used for analysis. High resolution electrospray mass
spectra (HR ESI-MS) were acquired with a Fourier transfer ion
cyclotron resonance (FT-ICR) APEX II model (Bruker Daltonics)
equipped with a 4.7 T cryomagnet (Magnex). Samples were
dissolved in the specific solvent and injected into the instrument
equipped with its own ESI source. Spectra were recorded in the
HR mode with resolutions ranging from 20000 to 30000.
Cholest-5-ene-3ꢀ,7ꢀ-diol (3). To a solution of 7ꢀ-hydroxyc-
holest-5-en-3ꢀ-yl acetate (4, 407 mg; 0.915 mmol) in ethanol (12
mL) and CH2Cl2 (4 mL) was added NaOH 15% (1.8 mL; 6.862
mmol). The mixture was stirred at rt for 2 h, poured into HCl (5%
aq soln) until neutralization, and evaporated to dryness. The residue
was extracted with ethyl acetate and the resulting organic phase
washed with water, dried (anhydrous Na2SO4), and evaporated to
yield the crude product. Column chromatography of the crude
product (petroleum ether-ethyl acetate, 2:1 to 1:1) afforded the
pure cholest-5-ene-3ꢀ,7ꢀ-diol (3, 350 mg, 95%) mp 178-179 °C
(EtOH); lit.66 177-178 °C. IR (film) 3306, 2930, 2867, 1466, 1379,
1
1053, 949 cm-1. H NMR (300 MHz, CDCl3) ppm 0.69 (3H, s,
CH3-18), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.92 (3H, d, J ) 6.6 Hz, CH3-21), 1.05 (3H, s, CH3-19),
3.55 (1H, ddt, J ) 11.3, 11.3, 5.0, 4.3 Hz, 3R-H), 3.85 (1H, td, J
) 8.0, 2.2, 2.2 Hz, 7R-H), 5.29 (t, J ) 2.2, 2.2 Hz, 6-H). 13C NMR
(75 MHz, CDCl3) ppm 11.8, 18.7, 19.1, 21.0 (CH2), 22.5, 22.8,
23.8 (CH2), 26.4 (CH2), 28.0, 28.5 (CH2), 31.5 (CH2), 35.7, 36.2
(CH2), 36.4 (C-10), 36.9 (CH2), 39.5 (CH2), 39.5 (CH2), 40.8, 41.7
(CH2), 42.9 (C-13), 48.2, 55.4, 55.9, 71.4, 73.3, 125.4 (C-6), 143.4
(C-5). MS (ESI) m/z (%): 403.9 (100) [M + H]+, 367.1 (62), 365.0
(68), 301.1 (48), 286.3 (81), 278.7 (45), 104.8).
7r-Hydroxycholest-5-en-3ꢀ-yl acetate (6) and Cholest-5-ene-
3ꢀ,7r-diol (5). To a solution of 7-oxo-cholest-5-en-3ꢀ-yl acetate
(250 mg; 0.565 mmol) in THF (10 mL) was slowly added
L-Selectride 1 M (675 µL; 0.675 mmol) at -70 °C (methanol bath
cooled in Cryocool apparatus). The mixture was stirred at -65 °C
for 9 h. Reaction was hydrolyzed with water (0.8 mL) and warmed
Substrate Synthesis. 7ꢀ-Hydroxycholest-5-en-3ꢀ-yl acetate
(4). To a mixture of cholesteryl acetate (1950 mg, 4.549 mmol)
and copper iodide (520 mg, 2.729 mmol) in CH3CN (30 mL) at
reflux temperature was added portionwised t-butyl hydroperoxide
(70% wt soln) (5 × 1.171 mL each 45 min, 45.49 mmol). After

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4015
up to 0 °C. The organoborane was oxidized with NaOH (6 M aq
soln, 0.8 mL) and H2O2 (30% aq soln, 0.8 mL) for 35 min. Then
most of the THF was removed under reduce pressure, and the
residue diluted with water. The crude product was extracted with
CH2Cl2, dried with anhydrous Na2SO4, filtered, and concentrated
under vacuum to yield a white crude product (230 mg). Column
chromatography of the crude product (petroleum ether-ethyl acetate,
2:1 to 1:3) afforded in order of elution the pure 7R-hydroxycholest-
5-en-3ꢀ-yl acetate (6, 112 mg, 44.6%) mp 139-140 °C (EtOH);
lit.67 139-140 °C. IR (film) 3420, 3035, 2942, 2867, 1734, 1471,
cm-1. 1H NMR (300 MHz, CDCl3) ppm 0.67 (3H, s, 18-CH3), 0.86
and 0.86 (each 3H, 2d, J ) 6.5 Hz, 26-CH3 and 27-CH3), 0.90
(3H, d, J ) 6.5 Hz, 21-CH3), 1.05 (3H, s, 19-CH3), 4.15 (1H, m,
3R-H), 5.27 (1H, d, J ) 1.7 Hz, 4-H). 13C NMR (75 MHz, CDCl3)
ppm 11.9, 18.6, 18.9, 21.0 (CH2), 22.5, 22.8, 23.8 (CH2), 24.2
(CH2), 28.0, 28.2 (CH2), 29.5 (CH2), 32.2 (CH2), 33.1 (CH2), 35.3
(CH2), 35.8, 35.9, 36.1 (CH2), 37.3 (C-10), 39.5 (CH2), 39.8 (CH2),
42.4 (C-13), 54.4, 56.1, 56.1, 68.0 (C-3), 123.2 (C-4), 147.8 (C-5).
MS (ESI) m/z (%) 385.6 (83) [M - H]+, 329.1 (65), 311.5 (84),
294.8 (57), 279.3 (66), 249.1 (100), 111.0 (96).
Chemical Synthesis of Epoxysterols. General Procedure for
the Epoxidation Using MMPP. According to the data presented
in Table 1, to a solution of the steroid in the appropriate solvent,
MMPP was added as a single portion. The suspension was stirred
for the periods of time, and temperatures indicated in Table 1, until
total substrate consumption (TLC control). Then reaction was
stopped by cooling and filtered and concentrated under vacuum.
The white solid residue was dissolved in ethyl acetate (saturated
in water) and the resulting organic phase was washed with Na2SO3
(10% aq), NaHCO3 (sat. aq), and water, dried with anhydrous
Na2SO4, filtered, and evaporated to yield a white foam crude
product.
1
1373, 1245, 1030, 757 cm-1. H NMR (300 MHz, CDCl3) ppm
0.68 (CH3-18), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3
and 27-CH3), 0.92 (3H, d, J ) 6.6 Hz, CH3-21), 1.00 (3H, s, CH3-
19), 2.03 (3H, s, 3ꢀ-CH3COO), 3.84 (1H, m, 7ꢀ-H), 4.64 (ddt, J
) 10.4, 10.4, 6.2, 4.2 Hz, 3R-H), 5.63 (dd, J ) 5.2, 1.5 Hz, 6-H).
13C NMR (75 MHz, CDCl3) ppm 11.6, 18.2, 18.7, 20.6 (CH2), 21.4,
22.5, 22.8, 23.7 (CH2), 24.2 (CH2), 27.5 (CH2), 28.0, 28.2 (CH2),
35.8, 36.1 (CH2), 36.7 (CH2), 37.4, 37.4 (C-10), 37.9 (CH2), 39.1
(CH2), 39.45 (CH2), 42.1 (C-13), 42.1, 49.4, 55.8, 65.2, 73.4, 124.7
(C-6), 145.2 (C-5), 170.4 (CH3COO). MS (ESI) m/z (%): 445.7
(42) [M + H]+, 452.9 (82), 348.7 (53), 309.8 (68), 291.9 (100),
278.8 (46), 172.7 (82), 147.5 (73), and the pure cholest-5-ene-
3ꢀ,7R-diol (5, 105 mg; 46.2%) mp 188-189 °C (EtOH); lit.67
186.5-189 °C. IR (film) 3337, 3030, 2936, 2867, 1466, 1379, 1053,
949, 757 cm-1. 1H NMR (300 MHz, CDCl3) ppm 0.69 (3H, s, CH3-
18), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-CH3),
0.93 (3H, d, J ) 6.6 Hz, CH3-21), 1.00 (3H, s, CH3-19), 3.59 (1H,
ddt, J ) 11.0, 11.0, 5.4, 4.2 Hz, 3R-H), 3.86 (1H, ddd, J ) 5.2,
3.5, 1.5 Hz, 7ꢀ-H), 5.61 (dd, J ) 5.2, 1.5 Hz, 6-H). 13C NMR (75
MHz, CDCl3) ppm 11.60, 18.2, 18.7, 20.7 (CH2), 22.6, 22.8, 23.7
(CH2), 24.3 (CH2), 28.0, 28.2 (CH2), 31.3 (CH2), 35.7, 36.1 (CH2),
37.0 (CH2), 37.4 (C-10), 37.5, 39.1 (CH2), 39.5 (CH2), 42.0 (CH2),
42.1 (C-13), 42.2, 49.4, 55.8, 65.3, 71.2, 123.8 (C-6), 146.2 (C-5).
MS (ESI) m/z (%): 403.9 (50) [M + H]+, 384.2 (100), 367.3 (48),
349.8 (61), 246.6 (34), 216.8 (53).
7r-Hydroxy-5r,6r-epoxycholestan-3ꢀ-yl Acetate (6a). To a
solution of 7R-hydroxy-cholest-5-en-3ꢀ yl-acetate (6, 80 mg, 0.180
mmoL) in refluxing CH3CN (5.4 mL) was added MMPP (98 mg,
0.198 mmoL). The reaction mixture was stirred under reflux
temperature for 40 min and then cooled to room temperature,
filtered, and concentrated under vacuum. The residue was dissolved
in ethyl acetate and washed sequentially with Na2SO3 (10% aq),
NaHCO3 (sat. aq), and water, dried with anhydrous Na2SO4, filtered,
and evaporated to yield a white foam crude product (82 mg, 99%).
Flash chromatography (petroleum ether, ethyl acetate 6:1) of the
crude product afforded the pure 7R-hydroxy-5R,6R-epoxycholestan-
3ꢀ-yl-acetate (6a, 50.6 mg, 61%). IR (film) 3515, 2950, 2867, 1746,
1
1468, 1363, 1248, 1036, 870 cm-1. H NMR (300 MHz, CDCl3)
ppm 0.61 (3H, s, 18-CH3), 0.85 and 0.86 (each 3H, 2d, J ) 6.6
Hz, 26-CH3 and 27-CH3), 0.90 (3H, d, J ) 6.6 Hz, 21-CH3), 1.08
(3H, s, 19-CH3), 2.02 (3H, s, 3ꢀ-CH3COO), 3.23 (1H, d, J ) 4.7
Hz, 6ꢀ-H), 3.86 (1H, bs, 7ꢀ-H), 4.94 (1H, tt, J ) 11.3, 11.3, 5.0,
5.0 Hz, 3R-H). 13C NMR (75 MHz, CDCl3) ppm 11.6, 15.9, 18.7,
20.2 (CH2), 21.3, 22.5, 22.8, 23.6 (CH2), 24.4 (CH2), 27.0 (CH2),
28.0, 28.1 (CH2), 32.2 (CH2), 35.5 (C-10), 35.7, 35.8 (CH2), 36.1
(CH2), 36.8, 37.1, 38.7 (CH2), 39.4 (CH2), 41.8 (C-13), 48.7, 55.4,
62.6, 64.7, 68.8 (C-5), 71.1, 170.2 (CH3COO).
Cholest-5-ene-3ꢀ,4ꢀ-diol (7). To a solution of cholesterol 1
(1200 mg, 3.104 mmol) in toluene (16 mL) a hot solution of
selenium dioxide (960 mg, 8.652 mmol) in acetic acid 98% (12
mL)) was added. The reaction mixture was stirred for 90 min at
90 °C and then stopped by the addition of sodium acetate (2.4 g).
Reaction mixture was stirred for 10 min at room temperature and
then filtered and extracted with toluene. The toluene solution was
washed with water, dried under anhydrous Na2SO4, and concen-
trated under vacuum to give a brown crude product. Flash
chromatography (gradient of petroleum ether-ethyl acetate, starting
from 8:1) afforded the 7 pure compound (624.8 mg, 50%) mp
176-177 °C (EtOH); lit.66 176-177 °C. IR (film) 3408, 2940,
Enzymatic Acylation of Epoxysterols. 5r,6r-Epoxycholestan-
3ꢀ-ol (1a) and 5ꢀ,6ꢀ-Epoxycholestan-3ꢀ-yl Acetate (2b). To a
solution of 5,6-epoxy-cholestan-3ꢀ-ol (75 mg, R/ꢀ ratio 78:22) in
toluene (4.5 mL) and vinyl acetate (0.5 mL), Novozym 435 (42
mg) was added and the reaction shaken 200 rpm at 45 °C. After
24 h, the enzyme was filtered and the solvent evaporated. Flash
chromatography (petroleum ether, ethyl acetate 7:3) afforded in the
following elution run the 5ꢀ,6ꢀ-epoxycholestan-3ꢀ-yl-acetate (2b,
16.6 mg, 20% yield). IR (film) 2948, 2867, 1728, 1466, 1373, 1245,
1
2870, 1457, 1365, 1072, 966, 841, 762, 641 cm-1. H NMR (300
MHz, CDCl3) ppm 0.68 (3H, s, 18-CH3), 0.86 and 0.87 (each 3H,
2d, J ) 6.6 Hz, 26-CH3 and 27-CH3), 0.91 (3H, d, J ) 6.6 Hz,
21-CH3), 1.18 (3H, s, 19-CH3), 3.56 (1H, td, J ) 11.8, 4.0, 4.0 Hz,
3R-H), 4.13 (1H, d, J ) 4.0 Hz, 4R-H), 5.68 (1H, dd, J ) 4.9, 1.9
Hz, 6-H). 13C NMR (75 MHz, CDCl3) ppm 11.8, 18.7, 20.5 (CH2),
21.0, 22.5, 22.8, 23.8 (CH2), 24.2 (CH2), 25.3 (CH2), 28.0, 28.2
(CH2), 31.8, 32.0 (CH2), 35.8, 36.0 (C-10), 36.1 (CH2), 36.9 (CH2),
39.5 (CH2), 39.6 (CH2), 42.26 (C-13), 50.1, 56.0, 56.9, 72.4, 77.2,
128.8 (C-6), 142.7 (C-5). MS (ESI) m/z (%) 401.7 (68) [M - H]+,
387.7 (24), 293.7 (35), 270.4 (27), 265.2 (100), 201.0 (27).
Cholest-4-en-3ꢀ-ol (9). To a solution of cholestenone (400 mg,
1.040 mmol) and cerium trichloride heptahydrate (387.5 mg, 1.040
mmol) in a mixture of tetrahydrofuran (6 mL) and methanol (3
mL) at 0 °C was added slowly and portionwised sodium borohy-
dride (78.7 mg, 2.080 mmol). The reaction mixture was stirred at
0 °C for 15 min and after concentrated under vacuum. The residue
was dissolved in ethyl acetate and organic phase washed sequen-
tially with HCl (5% aq soln), NaHCO3 (sat. soln), and water, dried
over anhydrous MgSO4, filtrated, and evaporated to yield a white
crude product (398 mg, 99%). Pure cholest-4-en-3ꢀ-ol (9, 362 mg,
90%) was obtained by flash chromatography (petroleum ether/ethyl
acetate 16:1) mp 135-136 °C (EtOH); lit.68 133.5-134.5 °C. IR
(film) 3354, 2928, 2867, 1654, 1465, 1377, 1045, 878, 854, 753
1
1041, 757 cm-1. H NMR (300 MHz, CDCl3) ppm 0.64 (3H, s,
18-CH3), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.88 (3H, d, J ) 6.6 Hz, 21-CH3), 1.00 (3H, s, 19-CH3),
2.03 (3H, s, 3ꢀ-CH3COO), 3.08 (1H, d, J ) 2.4 Hz, 6R-H), 4.77
(1H, tdd, J ) 11.9, 10.0, 4.8, 4.8 Hz, 3R-H). 13C NMR (75 MHz,
CDCl3) ppm 11.7, 17.0, 18.6, 21.3, 21.9 (CH2), 22.5, 22.8, 23.8
(CH2), 24.1 (CH2), 27.2 (CH2), 28.0, 28.1 (CH2), 29.7, 32.4 (CH2),
35.0 (C-10), 35.7, 36.1 (CH2), 36.6 (CH2), 38.0 (CH2), 39.5 (CH2),
39.7 (CH2), 42.2 (C-13), 50.9, 56.1 (2C), 62.5 (C-5), 63.6 (C-6),
71.3 (C-3), 170.6 (CH3COO). 5R,6R-Epoxy-cholestan-3ꢀ-ol (1a,
58 mg, 77% yield) mp 141-142 °C (EtOH); lit.69 140-141.5 °C.
IR (film) 3398, 2942, 2867, 1466, 1379, 1036, 763 cm-1. 1H NMR
(300 MHz, CDCl3) ppm 0.61 (3H, s, 18-CH3), 0.86 and 0.87 (each
3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-CH3), 0.89 (3H, d, J ) 6.6 Hz,
21-CH3), 1.07 (3H, s, 19-CH3), 2.91 (1H, d, J ) 4.4 Hz, 6ꢀ-H),
3.92 (1H, tt, J ) 11.4, 11.4, 5, 5 Hz, 3R-H). 13C NMR (75 MHz,
CDCl3) ppm 11.8, 15.9, 18.6, 20.6 (CH2), 22.5, 22.8, 23.8 (CH2),
24.0 (CH2), 28.0, 28.0 (CH2), 28.8 (CH2), 29.8, 31.0 (CH2), 32.4

4016 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
CarValho et al.
(CH2), 34.8 (C-10), 35.7, 36.1 (CH2), 39.4 (CH2), 39.5 (CH2), 39.8
(CH2), 42.3 (C-13), 42.5, 55.8, 56.8, 59.3, 65.7 (C-5), 68.7. MS
(ESI) m/z (%): 403.6 (50) [M + H]+, 385.0 (100); positive mode,
m/z [M + Na]+ calcd for C27H46O2Na: 425.3390; found: 425.3386.
5ꢀ,6ꢀ-Epoxycholestan-3ꢀ-ol (1b). To a solution of 5,6-epoxy-
cholestan-3ꢀ-ol (130 mg, R/ꢀ ratio 78:22) in toluene (7 mL) and
vinyl acetate (2 mL), lipase AY (880 mg) was added and the
reaction shaken at 200 rpm at 45 °C. After 3 days, the enzyme
was filtered and the solvent evaporated. Flash chromatography
(petroleum ether, ethyl acetate 9:1 to 3:2, with 0.1% of Et3N v/v)
afforded in the following elution run a mixture of 5,6-epoxyc-
holestan-3ꢀ-yl acetate (2a,2b, 112 mg, R/ꢀ ratio 91:09, 78% yield)
and the pure 5ꢀ,6ꢀ-epoxycholestan-3ꢀ-ol (1b, 20.3 mg, 15.6%) mp
132-133 °C (EtOH); lit.70 131-132 °C. IR (film) 3392, 2936,
2867, 1471, 1379, 1268, 1059, 937, 757 cm-1. 1H NMR (300 MHz,
CDCl3) ppm 0.64 (3H, s, 18-CH3), 0.86 and 0.87 (each 3H, 2d, J
) 6.6 Hz, 26-CH3 and 27-CH3), 0.89 (3H, d, J ) 6.6 Hz, 21-CH3),
1.00 (3H, s, 19-CH3), 3.07 (1H, d, J ) 2.4 Hz, 6R-H), 3.70 (1H,
m, 3R-H). 13C NMR (75 MHz, CDCl3) δ ppm 11.7, 17.0, 18.6,
22.0 (CH2), 22.5, 22.8, 23.8 (CH2), 24.2 (CH2), 28.0, 28.1 (CH2),
29.7, 31.0 (CH2), 32.6 (CH2), 34.8 (C-10), 35.7, 36.1 (CH2), 37.2
(CH2), 39.5 (CH2), 39.8 (CH2), 42.2 (CH2), 42.2 (C-13), 51.3, 56.2,
56.2, 62.9 (C-5), 63.7, 69.4. MS (ESI) m/z (%): 403.6 (96) [M +
H]+, 384.8 (100), 292.5 (62), 254.0 (32), 114.5 (25). HRMS (ESI),
positive mode, m/z [M + Na]+ calcd for C27H46O2Na: 425.3390;
found 425.3396.
CH3 and 27-CH3), 0.90 (3H, d, J ) 6.6 Hz, 21-CH3), 1.00 (3H, s,
19-CH3), 3.12 (1H, d, J ) 1.6 Hz, 6R-H), 3.50 (1H, m, 7R-H),
3.70 (1H, tdd, J ) 11.6, 9.9, 4.5, 4.5 Hz, 3R-H). 13C NMR (126
MHz, CDCl3) δ ppm 11.8, 16.8, 18.8, 22.0 (CH2), 22.5, 22.8, 23.8
(CH2), 27.2 (CH2), 28.0, 28.5 (CH2), 30.9 (CH2), 34.2 (C-10), 35.7,
36.2 (CH2), 36.9 (CH2), 38.3, 39.5 (CH2), 39.6 (CH2), 41.7 (CH2),
43.0 (C-13), 49.8, 55.3, 55.5, 67.2 (C-5), 67.5, 69.1, 74.8. MS (ESI)
m/z (%): 419.6 (74) [M + H]+, 413.5 (26), 402.6 (100). HRMS
(ESI), positive mode, m/z [M + Na]+ calcd for C27H46O3Na:
441.3339; found: 441.3328.
5r,6r-Epoxycholestane-3ꢀ,7r-diol (5a) and 7r-Hydroxy-
5ꢀ,6ꢀ-epoxycholestan-3ꢀ-yl Acetate (6b). To a solution of 5,6-
epoxycholestane-3ꢀ,7R-diol (50 mg, ratio R/ꢀ 89:11) in toluene (2
mL) and vinyl acetate (0.3 µL), Novozym 435 (55 mg) was added
and the reaction shaken 200 rpm at 45 °C. After 24 h, the enzyme
was filtered and the solvent evaporated. Flash chromatography
(chloroform, methanol 97:3) afforded according to elution run the
pure 7R-hydroxy-5ꢀ,6ꢀ-epoxycholestan-3ꢀ-yl-acetate (6b, 5 mg,
1
9%). H NMR (300 MHz, CDCl3) δ ppm 0.66 (3H, s, CH3-18),
0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-CH3),
0.90 (3H, d, J ) 6.6 Hz, CH3-21), 1.02 (3H, s, CH3-19), 2.04 (3H,
s, 3ꢀ-CH3COO), 3.12 (1H, d, J ) 3.0 Hz, 6R-H), 4.07 (1H, t, J )
3.0, 3.0 Hz, 7ꢀ-H), 4.82 (1H, ddd, J ) 14.8, 9.7, 4.9 Hz, 3R-H).
5R,6R-Epoxy-cholestane-3ꢀ,7R-diol (5a, 43.5 mg, 87%) mp 182-
183 °C (EtOH); lit.72 183-185 °C. IR (film) 3317, 2930, 2867,
1
1471, 1373, 1041, 960, 874, 722, 664 cm-1. H NMR (300 MHz,
CDCl3) δ ppm 0.62 (3H, s, CH3-18), 0.86 and 0.86 (each 3H, 2d,
J ) 6.6 Hz, 26-CH3 and 27-CH3), 0.90 (3H, d, J ) 6.6 Hz, CH3-
21), 1.07 (3H, s, CH3-19), 3.24 (1H, d, J ) 4.7 Hz, 6ꢀ-H), 3.87
(1H, m, 7ꢀ-H), 3.92 (1H, tt, J ) 11.2, 11.2, 4.9, 4.9 Hz, 3R-H).
13C NMR (75 MHz, CDCl3) δ ppm 11.6, 16.0, 18.7, 20.3 (CH2),
22.5, 22.8, 23.6 (CH2), 24.4 (CH2), 28.0, 28.1 (CH2), 30.9 (CH2),
32.4 (CH2), 35.4 (C-10), 35.7, 36.1, 36.9 (CH2), 37.1, 38.8 (CH2),
39.4 (CH2), 39.5 (CH2), 41.9 (C-13), 48.8, 55.4, 62.7, 64.8, 68.7,
69.3 (C-5). MS (ESI) m/z (%): 419.6 (94) [M + H]+, 413.4 (42),
401.4 (100), 384.4 (35), 316.5 (28), 241.4 (35), 148.7 (20), 101.6
(17). HRMS (ESI), positive mode, m/z [M + Na]+ calcd for
C27H46O3Na: 441.3339; found: 441.3331.
5r,6r-Epoxycholestane-3ꢀ,7ꢀ-diol (3a) and 7ꢀ-Hydroxy-
5ꢀ,6ꢀ-epoxycholestan-3ꢀ-yl Acetate (4b). To a solution of 5,6-
epoxy-cholestan-3ꢀ,7ꢀ-diol (81 mg, ratio R/ꢀ 44:56) in toluene (8
mL) and vinyl acetate (1 mL), Novozym 435 (180 mg) was added
and the reaction shaken 200 rpm at 45 °C. After 24 h, the enzyme
was filtered and the solvent evaporated. Flash chromatography
(chloroform, methanol 99:1 to 95:5) afforded the 7ꢀ-hydroxy-5ꢀ,6ꢀ-
epoxycholestan-3ꢀ-yl-acetate (4b, 49 mg, 55%). IR (film) 3468,
1
2950, 2867, 1738, 1468, 1368, 1239, 1036, 754 cm-1. H NMR
(300 MHz, CDCl3) δ ppm 0.62 (3H, s, 18-CH3), 0.71(1H, m), 0.84
and 0.84 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-CH3), 0.88
(3H, d, J ) 6.5 Hz, 21-CH3), 0.99 (3H, s, 19-CH3), 2.01 (3H, s,
3ꢀ-CH3COO), 3.13 (1H, d, J ) 1.5 Hz, 6R-H), 3.49 (1H, dd, J )
8.7, 1.5 Hz, 7R-H), 4.75 (1H, tdd, J ) 11.9, 9.7, 4.8, 4.8 Hz, 3R-
H). 13C NMR (75 MHz, CDCl3) δ ppm 11.7, 16.8, 18.7, 21.2, 21.8
(CH2), 22.5, 22.8, 23.8 (CH2), 27.0 (CH2), 27.2 (CH2), 27.9, 28.5
(CH2), 34.3 (C-10), 35.6, 36.1 (CH2), 36.3 (CH2), 37.5 (CH2), 38.1,
39.4 (CH2), 39.4 (CH2), 42.9 (C-13), 49.4, 55.1, 55.4, 66.7 (C-5),
67.4, 70.9, 74.5, 170.4 (CH3COO). 5R,6R-Epoxy-cholestane-3ꢀ,7ꢀ-
diol (3a, 35.4 mg, 44% yield) mp 153-154 °C (EtOH); lit.71
153-155 °C. IR (film) 3381, 2948, 2867, 1466, 1378, 1041, 734
5r,6r-Epoxycholestane-3ꢀ,4ꢀ-diol (7a), 4ꢀ-Hydroxy-5ꢀ,6ꢀ-
epoxycholestan-3ꢀ-yl Acetate (13b) and 3ꢀ-Hydroxy-5ꢀ,6ꢀ-
epoxycholestan-4ꢀ-yl Acetate (14b). To a solution of 5,6 epoxy-
cholestane-3ꢀ,4ꢀ-diol (176 mg, ratio R/ꢀ 43:57) in toluene (18 mL)
and vinyl acetate (2 mL), Novozym 435 (1505 mg) was added and
the reaction shaken 200 rpm at 45 °C. Reaction was controlled by
TLC (petroleum ether, ethyl acetate 1:1, double development). After
7 days, the enzyme was filtered and the solvent evaporated. Flash
chromatography (petroleum ether-ethyl acetate, 4:1 to 1:1) yielded
accordingly to elution sequence 4ꢀ-hydroxy-5ꢀ,6ꢀ-epoxycholestan-
3ꢀ-yl acetate (13b, 60 mg, 31%), mp 155-156 °C (EtOH). IR (film)
1
cm-1. H NMR (300 MHz, CDCl3) δ ppm 0.63 (3H, s, 18-CH3),
0.84 and 0.85 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-CH3),
0.88 (3H, d, J ) 6.6 Hz, 21-CH3), 1.08 (3H, s, 19-CH3), 2.86 (1H,
s, 6ꢀ-H), 3.67 (1H, d, J ) 6.6 Hz, 7R-H), 3.86 (1H, tt, J ) 10.9,
10.9, 4.7, 4.7 Hz, 3R-H). 13C NMR (75 MHz, CDCl3) δ ppm 11.8,
16.1, 18.6, 20.5 (CH2), 22.5, 22.8, 23.8 (CH2), 25.2 (CH2), 28.0,
28.2 (CH2), 30.9 (CH2), 32.3 (CH2), 34.6 (C-10), 35.7, 36.1 (CH2),
39.3 (CH2), 39.4 (CH2), 39.7, 40.0, 42.7 (C-13), 55.5, 55.8, 62.7,
66.2 (C-5), 68.5, 70.3. MS (ESI) m/z (%): 419.2 (20) [M + H]+,
401.9 (14), 383.6 (15), 367.8 (18), 352.7 (28), 232.6 (11), 102.6
(63), 101.6 (100). HRMS (ESI), positive mode, m/z [M + Na]+
calcd for C27H46O3Na: 441.3339; found: 441.3335.
3524, 2955, 2868, 1717, 1454, 1376, 1242, 1044, 918, 760 cm-1
.
1H NMR (500 MHz, CDCl3) δ ppm 0.61 (1H, m), 0.63 (3H, s,
18-CH3), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.89 (3H, d, J ) 6.6 Hz, 21-CH3), 1.16 (3H, s, 19-CH3),
2.11 (3H, s, 3ꢀ-CH3COO), 3.22 (1H, dd, J ) 2.5, 1.0 Hz, 6R-H),
3.45 (1H, dd, J ) 3.2, 1.0 Hz, 4R-H), 4.82 (1H, ddd, J ) 11.6,
4.3, 3.2 Hz, 3R-H). 13C NMR (75 MHz, CDCl3) ppm 11.6, 17.9,
18.6, 21.1 (CH2), 21.1 (CH2), 21.3, 22.5, 22.8, 23.8 (CH2), 24.2
(CH2), 28.0, 28.1 (CH2), 29.6, 32.4 (CH2), 35.0 (C-10), 35.7, 36.1
(CH2), 36.8 (CH2), 39.5 (CH2), 39.5 (CH2), 42.1 (C-13), 51.6, 56.1,
56.2, 63.8 (C-6), 65.2 (C-5), 73.4, 74.9, 170.7 (CH3COO). MS (ESI)
m/z (%) 461.5 (100) [M + H]+, 442.5 (48), 417.7 (37), 391.9 (25),
367.5 (39), 349.2 (28), 325.9 (22). 3ꢀ-Hydroxy-5ꢀ,6ꢀ-epoxyc-
holestan-4ꢀ-yl acetate (14b, 46 mg, 24%) mp 153-154 °C (EtOH).
IR (film) 3527, 2953, 2867, 1716, 1454, 1379, 1239, 1047, 914
cm-1. 1H NMR (500 MHz, CDCl3) δ ppm 0.61 (1H, m), 0.62 (3H,
s, 18-CH3), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.89 (3H, d, J ) 6.6 Hz, 21-CH3), 1.12 (3H, s, 19-CH3),
2.15 (3H, s, 4ꢀ-CH3COO), 3.21 (1H, dd, J ) 2.2, 0.5 Hz, 6R-H),
3.78 (1H, ddd, J ) 11.4, 4.3, 3.5 Hz, 3R-H), 4.72 (dd, J ) 3.5, 0.5
Hz, 4R-H). 13C NMR (126 MHz, CDCl3) ppm 11.6, 17.3, 18.6,
21.1, 21.4 (CH2), 22.5, 22.8, 23.7 (CH2), 24.2 (CH2), 24.7 (CH2),
5ꢀ,6ꢀ-Epoxycholestane-3ꢀ,7ꢀ-diol (3b). To a solution of 5,6-
epoxy-cholestan-3ꢀ,7ꢀ-diol (60 mg, ratio R/ꢀ 44:56) in toluene (6
mL) and vinyl acetate (0.6 mL), lipase AY (53 mg) was added and
the reaction shaken 200 rpm at 45 °C. After 24 h the enzyme was
filtered and the solvent evaporated. Flash chromatography (petro-
leum ether, ethyl acetate 2:1 to 1:2) afforded the 7ꢀ-hydroxy-5,6-
epoxycholestan-3ꢀ-yl-acetate mixture (4a,4b, 35 mg, R/ꢀ ratio 79:
21, 53%) and the pure 5ꢀ,6ꢀ-epoxy-cholestane-3ꢀ,7ꢀ-diol (3b, 26.3
mg, 44% yield) mp 168-170 °C (EtOH); lit.72 170-171 °C. IR
1
(film) 3366, 2942, 2867, 1466, 1379, 1059, 763 cm-1. H NMR
(500 MHz, CDCl3) δ ppm 0.64 (3H, s, 18-CH3), 0.71 (1H, dt, J )
11.4, 11.4, 5.6 Hz), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4017
28.0, 28.1 (CH2), 29.6, 32.4 (CH2), 34.9 (C-10), 35.7, 36.1 (CH2),
37.1 (CH2), 39.4 (CH2), 39.6 (CH2), 42.1 (C-13), 51.7, 56.1, 56.1,
62.7 (C-5), 62.9 (C-6), 70.7 (C-7), 79.2 (C-3), 171.7 (CH3COO).
MS m/z (%) 461.4 (59) [M + H]+, 444.4 (69), 441.0 (100), 382.3
(70), 377.0 (44), 273.9 (63). HRMS (ESI), positive mode, m/z [M
+ Na]+ calcd for C29H48O4Na: 483.3444; found: 483.3437. 5R,6R-
Epoxycholestane-3ꢀ,4ꢀ-diol (7a, 74 mg, 42%) mp 205-206 °C
(EtOH); lit.73 195.5-198 °C. IR (film) 3356, 2942, 2867, 1454,
1373, 1047, 972, 763 cm-1. 1H NMR (500 MHz, CDCl3) ppm 0.61
(3H, s, 18-CH3), 0.86 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3
and 27-CH3), 0.88 (3H, d, J ) 6.6 Hz, 21-CH3), 1.22 (3H, s, 19-
CH3), 3.03 (1H, dd, J ) 4.2, 6ꢀ-H), 3.19 (1H, dd, J ) 3.5, 1.0 Hz,
4R-H), 3.86 (1H, ddd, J ) 12.0, 5.0, 3.5 Hz, 3R-H). 13C NMR
(126 MHz, CDCl3) ppm 11.9, 15.3, 18.7, 20.0 (CH2), 22.6, 22.8,
23.9 (CH2), 24.0 (CH2), 25.6 (CH2), 28.0, 28.1 (CH2), 28.3 (CH2),
30.0, 32.7 (CH2), 34.3 (C-10), 35.8, 36.2 (CH2), 39.4 (CH2), 39.5
(CH2), 42.45 (C-13), 43.4, 55.9, 57.0, 59.3, 64.7 (C-5), 70.1, 77.4.
MS (ESI) m/z (%): 419.7 (71) [M + H]+, 418.1 (36), 402.0 (17),
391.0 (22), 239.4 (25), 102 (100). HRMS (ESI), positive mode,
m/z [M + Na]+ calcd for C27H46O3Na: 441.3339; found: 441.3346.
4r,5r-Epoxycholestan-3ꢀ-ol (9a) and 4ꢀ,5ꢀ-Epoxycholestan-
3ꢀ-yl Acetate (10b). To a solution of 4,5-epoxy-3ꢀ-cholestanol
(110 mg, ratio R/ꢀ 21:79) in toluene (10 mL) and vinyl acetate
(1.3 mL), Novozym 435 (1434 mg) was added and the reaction
shaken 200 rpm at 45 °C. Reaction was controlled by TLC
(petroleum ether, ethyl acetate 3:1, double development). After 5
days, the enzyme was filtered and the solvent evaporated. Flash
chromatography (petroleum ether, ethyl acetate 10:1 to 3:1) afforded
in the following elution run the 4ꢀ,5ꢀ-epoxycholestan-3ꢀ-yl acetate
(10b, 92 mg, 76%). 1H NMR (300 MHz, CDCl3) ppm 0.68 (3H, s,
18-CH3), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.90 (3H, d, J ) 6.6 Hz 21-CH3), 1.04 (3H, s, 19-CH3),
2.10 (1H, s, 3ꢀ-CH3COO), 3.16 (1H, d, J ) 3.6 Hz, 4R-H), 5.13
(1H, m, 3R-H). 13C NMR (75 MHz, CDCl3) ppm 11.9, 18.4, 18.6,
21.2, 21.4 (CH2), 22.5 (CH2), 22.5, 22.8, 23.8 (CH2), 24.2 (CH2),
28.0, 28.1 (CH2), 29.6 (CH2), 30.3 (CH2), 31.2 (CH2), 35.2, 35.7,
35.8 (C-10), 36.1 (CH2), 39.5 (CH2), 39.7 (CH2), 42.5 (C-13), 49.0,
56.0, 56.2, 61.3, 66.8 (C-5), 69.0, 171.0 (CH3COO). 4R,5R-
Epoxycholestan-3ꢀ-ol (9a, 22 mg, 20%) mp 134-136 °C (EtOH),
lit.74 136-137 °C. IR (film) 3408, 2942, 2867, 1466, 1379, 960,
and 27-CH3), 0.89 (3H, d, J ) 6.6 Hz, 21-CH3), 1.12 (3H, s, 19-
CH3), 3.19 (1H, d, J ) 2.2 Hz, 6R-H), 3.35 (1H, d, J ) 3.6 Hz,
4R-H), 3.61 (1H, td, J ) 11.6, 4.5, 3.6 Hz, 3R-H). 13C NMR (126
MHz, CDCl3) ppm 11.7, 17.8, 18.7, 21.2 (CH2), 22.5, 22.8, 23.8
(CH2), 24.2 (CH2), 24.8 (CH2), 28.0, 28.1 (CH2), 29.6, 32.5 (CH2),
34.7 (C-10), 35.7, 36.1 (CH2), 36.9 (CH2), 39.5 (CH2), 39.6 (CH2),
42.2 (C-13), 51.5, 56.1, 56.2, 64.0, 65.5 (C-5), 71.2, 77.1. MS (ESI)
m/z (%): 419.8 (81) [M + H]+, 357.5 (20), 289.2 (25), 242.7 (24),
102.3 (100). HRMS (ESI), positive mode, m/z [M + Na]+ calcd
for C27H46O3Na: 441.3339; found: 441.3343.
5r,6r-Epoxy-7-norcholestan-3ꢀ-ol (11a). To a solution of
5R,6R-epoxy-7-norcholestan-3ꢀ-yl-acetate (12a, 45 mg) in toluene
(5 mL) and octanol (0.15 mL), lipase AY (450 mg) was added and
the reaction shaken 200 rpm at 45 °C. After 3 days, the enzyme
was filtered and the solvent evaporated. 5R,6R-Epoxy-7-norc-
holestan-3ꢀ-ol (11a, 39 mg, 96% yield) was isolated by flash
chromatography (petroleum ether, ethyl acetate 3:1 to 1:1) mp
138.5-139.5 °C (EtOH); lit.76 139-140 °C. IR (film) 3366, 2936,
1
2872, 1460, 1379, 1083, 757 cm-1. H NMR (600 MHz, CDCl3)
ppm 0.63 (3H, s, 18-CH3), 0.86 and 0.87 (each 3H, 2d, J ) 6.6
Hz, 26-CH3 and 27-CH3), 0.89 (3H, s, 19-CH3), 0.90 (3H, d, J )
6.6 Hz, 21-CH3), 3.26 (1H, s, 6ꢀ-H), 3.96 (1H, m, 3R-H). 13C NMR
(75 MHz, CDCl3) ppm 12.1, 15.4, 18.7, 20.9, 22.5, 22.8, 23.9, 24.2,
28.0, 28.5, 30.9, 31.2, 34.6, 35.7, 36.2, 38.8 (C-10), 39.4, 39.8,
42.5, 44.4 (C-13), 48.2, 50.6, 55.5, 60.4, 69.1 (C-5), 69.8. MS (ESI)
m/z (%): 389.7 (58) [M + H]+, 371.8 (100), 353.0 (28), 339.2 (11),
269.1 (11). HRMS (ESI), positive mode, m/z [M + Na]+ calcd for
C26H44O2Na: 411.3233; found 411.3229.
Chemical Deacylation of Epoxysterols. General Procedure
for Mild Acetoxy Hydrolysis of Epoxysteroids. 4ꢀ,5ꢀ-Epoxyc-
holestan-3ꢀ-ol (9b). To a solution of 4ꢀ,5ꢀ-epoxycholestan-3ꢀ-yl
acetate (10b, 40 mg, 0.090 mmol) in methanol (10 mL) was added
sodium carbonate (19 mg, 0.180 mmol), and the reaction mixture
was stirred at room temperature overnight. TLC analysis showed
the disappearance of 10b and the formation of a polar compound.
Then solvent was concentrated under vacuum and the residue
dissolved in dichloromethane, washed with water, and dried over
anhydrous sodium sulfate. Evaporation gave a white solid and flash
chromatography (petroleum ether, ethyl acetate 8:1 to 4:1) afforded
the pure 4ꢀ,5ꢀ-epoxycholestan-3ꢀ-ol (9b, 34.7 mg, 96%) mp 96-97
°C (EtOH); lit.74 96-97 °C. IR (film) 3415, 2936, 2867, 1466,
1379, 1036, 855 cm-1. 1H NMR (500 MHz, CDCl3) ppm 0.68 (3H,
s, 18-CH3), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and
27-CH3), 0.90 (3H, d, J ) 6.6 Hz, 21-CH3), 1.03 (3H, s, 19-CH3),
3.14 (1H, d, J ) 4.8 Hz, 4R-H), 4.06 (1H, dt, J ) 4.8, 4.8, 2.5 Hz,
3R-H). 13C NMR (126 MHz, CDCl3) ppm 11.9, 18.6, 19.1, 21.3
(CH2), 22.8, 23.8 (CH2), 24.3 (CH2), 25.9 (CH2), 26.0 (CH2), 28.0,
28.2 (CH2), 30.3 (CH2), 31.1 (CH2), 35.0, 35.7 (C-10), 36.1 (CH2),
36.1, 39.5 (CH2), 39.7 (CH2), 42.5 (C-13), 47.0, 56.1, 56.2, 63.7,
64.2, 69.0 (C-5). MS (ESI) m/z (%): 403.7 (51) [M + H]+, 372.1
(37), 369.1 (39), 354.4 (100), 296.6 (80), 288.7 (41), 206.8 (54),
72.9 (50), 59.2 (50). HRMS (ESI), positive mode, m/z [M + Na]+
calcd for C27H46O2Na: 425.3390; found: 425.3395.
Biology. General. HT29 and BJ cell lines were purchased from
ATCC, whereas LAMA-84 cells were acquired from DSMZ.
Roswell Park Memorial Institute medium with L-glutamine and
without sodium bicarbonate (RPMI 1640), resazurin, penicillin, and
streptomycin were obtained from Sigma-Aldrich Co. Fetal bovine
serum (FBS) and tripsin/EDTA (0.5 g/L) were obtained from Gibco.
MycoAlert was purchase from Cambrex Bioscience Verviers, Inc.
Cisplatin (Faulplatin, 100 mg/mL, Mayne Pharma, Portugal, Lda)
and doxorubicin (doxorubicin hydrochloride for injection, USP, 2
mg/mL, Faulding, Canada, Inc.) were kindly donated by the
Hospital Universita´rio de Coimbra. A stock solution of 0.1 mg/
mL of resazurin in Dulbecco’s phosphate buffer saline (PBS: 1.4
mM KH2PO4, 2.7 mM KCl, 4.3 mM Na2HPO4, 137 mM NaCl, pH
7.4) was prepared and used for the Alamar assay. Absorbance was
determined by a Multiscan Ex, Thermo Labsystem, 96-well reader
spectrometer.
1
757 cm-1. H NMR (300 MHz, CDCl3) δ ppm 0.68 (3H, s, 18-
CH3), 0.86 and 0.87 (each 3H, 2d, J ) 6.6 Hz, 26-CH3 and 27-
CH3), 0.90 (3H, d, J ) 6.5 Hz, 21-CH3), 1.11 (3H, s, 19-CH3),
2.90 (1H, d, J ) 1.0 Hz, 4ꢀ-H), 3.99 (1H, m, 3R-H). 13C NMR
(126 MHz, CDCl3) δ ppm 12.1, 17.7, 18.6, 20.9 (CH2), 22.6, 22.8,
23.8 (CH2), 24.2 (CH2), 27.0 (CH2), 28.0, 28.2 (CH2), 28.4 (CH2),
28.5 (CH2), 29.7 (CH2), 35.5, 35.5 (C-10), 35.8, 36.1 (CH2), 39.5
(CH2), 39.8 (CH2), 42.6 (C-13), 50.0, 55.7, 56.2, 64.7, 66.0, 67.4
(C-5). MS (ESI) m/z (%): 403.7 (86) [M + H]+, 399.9 (35), 394.7
(100), 384.8 (84), 370.9 (36), 367.2 (30), 326.2 (32), 311.6 (39),
288.4 (94). HRMS (ESI), positive mode, m/z [M + Na]+ calcd for
C27H46O2Na: 425.3390; found: 425.3381.
Enzymatic Deacylation of Epoxysterols. 5r,6r-Epoxycholes-
tane-3ꢀ,7r-diol (5a). To a solution of 7R-hydroxy-5R,6R-epoxy-
cholestan-3ꢀ-yl acetate (6a, 40 mg) in toluene (5 mL) and octanol
(0.15 mL), lipase AY (400 mg) was added and the reaction shaken
200 rpm at 45 °C. After 4 days, the enzyme was filtered and the
solvent evaporated. 5R,6R-Epoxy-cholestane-3ꢀ,7R-diol (5a, 34 mg,
94% yield) was isolated by flash chromatography (petroleum ether,
ethyl acetate 2:1 to 1:1).
5ꢀ,6ꢀ-Epoxycholestane-3ꢀ,4ꢀ-diol (7b). To a solution of 4ꢀ-
hydroxy-5ꢀ,6ꢀ-epoxycholestan-3ꢀ-yl acetate (13b, 30 mg) in
toluene (4 mL) and octanol (0.1 mL), Novozym 435 (90 mg) was
added and the reaction shaken 200 rpm at 45 °C. After 20 days,
the reaction was stopped, enzyme was filtered, and the solvent
concentrated under vacuum. Flash chromatography (petroleum
ether, ethyl acetate 4:1 to 2:1) afforded the 5ꢀ,6ꢀ-epoxycholestane-
3ꢀ,4ꢀ-diol (7b, 14.2 mg, 52% yield) mp 175.5-176.5 °C (EtOH);
lit.75 173-175 °C. IR (film) 3419, 2942, 2867, 1466, 1379, 914,
1
763 cm-1. H NMR (500 MHz, CDCl3) ppm 0.59 (1H, m), 0.63
Cell Lines and Culture Condition. Cytotoxicity of the com-
(3H, s, 18-CH3), 0.85 and 0.86 (each 3H, 2d, J ) 6.6 Hz, 26-CH3
pounds synthesized was assessed on HT29 (human colon adeno-

4018 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13
CarValho et al.
(5) Brown, A. J.; Jessup, W. Oxysterols and atherosclerosis. Atheroscle-
rosis 1999, 142, 1–28.
(6) Gill, S.; Chow, R.; Brown, A. J. Sterol regulators of cholesterol
homeostasis and beyond: The oxysterol hypothesis revisited and
revised. Prog. Lipid Res. 2008, 47, 391–404.
(7) Vejux, A.; Malvitte, L.; Lizard, G. Side effects of oxysterols:
cytotoxicity, oxidation, inflammation, and phospholipidosis. Braz.
J. Med. Biol. Res. 2008, 41, 545–556.
(8) Murphy, R. C.; Johnson, K. M. Cholesterol, Reactive Oxygen Species,
and the Formation of Biologically Active Mediators. J. Biol. Chem.
2008, 283, 15521–15525.
carcinoma; adherent epitheloid cell line), LAMA84 (human chronic
myeloid leukemia in blast crisis; suspended/slowly adherent cell
line), and the BJ noncancer cell line (human foreskin fibroblast;
adherent cell line). Cells were grown at 37 °C in a humidified
atmosphere of 95% air and 5% CO2, in RPMI 1640 medium with
0.2% NaHCO3 and supplemented with 10% of heat inactivated fetal
bovine serum (iFBS), 100 units/mL penicillin and 0.1 mg/mL
streptomycin (PS)-culture medium (CM).
Adherent cells were harvested using trypsin for 10 min at 37 °C
and exponentially growing cells were used throughout. Cells were
tested periodically for the presence of mycoplasma with the
mycoplasma detection kit MycoAlert.
(9) Naz, S.; Kerr, R. G.; Narayanan, R. New antiproliferative epoxysec-
osterols from Pseudopterogorgia americana. Tetrahedron Lett. 2000,
41, 6035–6040.
(10) Ahmed, A. F.; Hsieh, Y.-T.; Wen, Z.-H.; Wu, Y.-C.; Sheu, J.-H.
Polyoxygenated Sterols from the Formosan Soft Coral Sinularia
gibberosa. J. Nat. Prod. 2006, 69, 1275–1279.
(11) Luo, X.; Li, F.; Shinde, P. B.; Hong, J.; Lee, C.-O.; Im, K. S.; Jung,
J. H. 26,27-Cyclosterols and Other Polyoxygenated Sterols from a
Marine Sponge Topsentia sp. J. Nat. Prod. 2006, 69, 1760–1768.
(12) Misra, L.; Lal, P.; Chaurasia, N. D.; Sangwan, R. S.; Sinha, S.; Tuli,
R. Selective reactivity of 2-mercaptoethanol with 5[beta],6[beta]-
epoxide in steroids from Withania somnifera. Steroids 2008, 73, 245–
251.
(13) Chen, W.-Y.; Chang, F.-R.; Huang, Z.-Y.; Chen, J.-H.; Wu, Y.-C.;
Wu, C.-C. Tubocapsenolide A, a Novel Withanolide, Inhibits Prolif-
eration and Induces Apoptosis in MDA-MB-231 Cells by Thiol
Oxidation of Heat Shock Proteins. J. Biol. Chem. 2008, 283, 17184–
17193.
(14) Parker, R. E.; Isaacs, N. S. Mechanisms Of Epoxide Reactions. Chem.
ReV. 1959, 59, 737–799.
Cell Viability Assay. Stock solutions of the synthesized sterols
were prepared in dimethylsulfoxide (DMSO) and stored at -20
°C.
HT29 cells were seeded at density of 9 × 103 cells per well in
96-well culture plates, LAMA84 at 10 × 103 cells/well, and BJ at
2 × 103 cells/well in a total volume of 100 µL of CM. Cells were
incubated overnight and then exposed to different sterol concentra-
tions, containing DMSO, in 100 µL of CM. The maximal DMSO
concentration in the total medium volume was always lower than
0.6% (including controls). Compounds 1, 7, and 9 were dissolved
in a mixture of DMSO and tetrahydrofuran (THF) and given to
cells in a solvent mixture concentration lower than 0.7%. Compound-
induced cytotoxicity was quantified after 48 h of exposure by a
cell viability colorimetric redox assay using resazurin,48 which
measures the conversion of resazurin (blue-absorbance at 630 nm)
to its reduced form, resorufin (pink-absorbance at 540 nm) by viable
cells. After 48 h incubation, the medium was removed and replaced
by fresh medium, depleted in iFBS and SP, but containing 10% of
resazurin. Cells were further incubated for 2 h at 37 °C. Absorbance
was then measured at 540 and 630 nm with a plate spectropho-
tometer and cell viability (CV) calculated for each concentration
tested, according to the following formula:
(15) Szeliga, J.; Dipple, A. DNA Adduct Formation by Polycyclic Aromatic
Hydrocarbon Dihydrodiol Epoxides. Chem. Res. Toxicol. 1998, 11,
1–11.
(16) Zheng, J.; Cho, M.; Jones, A. D.; Hammock, B. D. Evidence of
Quinone Metabolites of Naphthalene Covalently Bound to Sulfur
Nucleophiles of Proteins of Murine Clara Cells after Exposure to
Naphthalene. Chem. Res. Toxicol. 1997, 10, 1008–1014.
(17) Greene, J. F.; Zheng, J.; Grant, D. F.; Hammock, B. D. Cytotoxicity
of 1,2-epoxynaphthalene Is Correlated with Protein Binding and in
Situ Glutathione Depletion in Cytochrome P4501A1 Expressing Sf-
21 Cells. Toxicol. Sci. 2000, 53, 352–360.
(18) Lesuisse, D.; Gourvest, J. F.; Hartmann, C.; Tric, B.; Benslimane,
O.; Philibert, D.; Vevert, J. P. Synthesis and evaluation of a new series
of mechanism-based aromatase inhibitors. J. Med. Chem. 1992, 35,
1588–1597.
CV(%) )
(Abs540D - Abs540R) - (Abs630D - Abs630R)
(Abs540C - Abs540R) - (Abs630C - Abs630R)
× 100 (1)
(19) Numazawa, M.; Yamada, K.; Watari, Y.; Ando, M. Improved Synthesis
and Molecular Modeling of 4ꢀ,19-Dihydroxyandrost-5-en-17-one, an
Excellent Inhibitor of Aromatase. Chem. Pharm. Bull. 2002, 50, 703–
705.
(20) Michne, W. F.; Schroeder, J. D.; Bailey, T. R.; Young, D. C.; Hughes,
J. V.; Dutko, F. J. Keto/enol epoxy steroids: a new structural class of
HIV-1 tat inhibitors. J. Med. Chem. 1993, 36, 2701–2702.
(21) Michne, W. F.; Schroeder, J. D.; Bailey, T. R.; Neumann, H. C.; Cooke,
D.; Young, D. C.; Hughes, J. V.; Kingsley, S. D.; Ryan, K. A.; Putz,
H. S.; Shaw, L. J.; Dutko, F. J. Keto/Enol Epoxy Steroids as HIV-1
Tat Inhibitors: Structure-Activity Relationships and Pharmacophore
Localization. J. Med. Chem. 1995, 38, 3197–3206.
(22) Zhao, K.; Wang, Y.; Han, L. 4,5-Epoxycholestane-3,6-diols: Templates
for generating the full set of eight cholestane-3,5,6-triol stereoisomers
in multigram scales, but not for a cholestane-3,4,6-triol. Steroids 2007,
72, 95–104.
(23) Pe´rez-Ornelas, V.; Cabeza, M.; Bratoeff, E.; Heuze, I.; Sa´nchez, M.;
Ram´ırez, E.; Naranjo-Rodr´ıguez, E. New 5[alpha]-reductase inhibitors:
in vitro and in vivo effects. Steroids 2005, 70, 217–224.
(24) Numazawa, M.; Oshibe, M. 6-Alkyl- and 6-Arylandrost-4-ene-3,17-
diones as Aromatase Inhibitors. Synthesis and Structure-Activity
Relationships. J. Med. Chem. 1994, 37, 1312–1319.
D, drug treated cells; C, control cells; R, resazurin control.
The inhibitory concentration in which 50% of the cells die (IC50)
was determined for each compound by plotting the CV observed
for each concentration versus the logarithmized concentration and
fitting a nonlinear regression curve. All experiments were performed
in triplicates and were repeated at least 3 times. Results were
expressed as the mean and are summarized in Table 4.
Acknowledgment. We thank Fundac¸a˜o para a Cieˆncia e
Tecnologia (FCT), Portugal, through POCI for financial support.
We acknowledge the “Laborato´rio de Espectroscopia de Massa
do Centro de Estudos Farmaceˆuticos da Universidade de
Coimbra”, Portugal, for providing data from the LC-MSn
(QITMS), integrated in the Portuguese National Mass Spec-
trometry Network (QNEM). J.F.S. Carvalho also thanks the
Fundac¸a˜o para a Cieˆncia e Tecnologia (FCT), Portugal, for a
grant (SFRH/BD/18263/2004).
(25) Numazawa, M.; Yoshimura, A. Synthesis and Biochemical Studies
of 19-Oxygenated Derivatives of 6R- and 6ꢀ-Methylandrostenediones
as Catalytic Probes for the Active Site of Aromatase. Biol. Pharm.
Bull. 2000, 23, 1059–1065.
(26) Salvador, J. A. R.; Silvestre, S. M.; Moreira, V. M. Catalytic
Epoxidation and syn-Dihydroxylation Reactions in Steroid Chemistry.
Curr. Org. Chem. 2008, 12, 492–522.
(27) Ball, P.; Zienelis, K.; Allen, L. Biocatalysis: Synthesis methods that
exploit enzymatic activities. Nature 2001, 409, 225–225.
(28) Koeller, K. M.; Wong, C.-H. Enzymes for chemical synthesis. Nature
2001, 409, 232–240.
(29) Hager, L. P.; Allain, E. J. Stereoselective epoxidation of alkenes by
chloroperoxidase. U.S. Patent 5358860, 1994.
References
(1) Ferderbar, S.; Pereira, E. C.; Apolina´rio, E.; Bertolami, M. C.; Faludi,
A.; Monte, O.; Calliari, L. E.; Sales, J. E.; Gagliardi, A. R.; Xavier,
H. T.; Abdalla, D. S. P. Cholesterol oxides as biomarkers of oxidative
stress in type 1 and type 2 diabetes mellitus. Diabetes Metab. Res.
ReV. 2007, 23, 35–42.
(2) Glass, C. K.; Saijo, K. Immunology: Oxysterols hold T cells in check.
Nature 2008, 455, 40–41.
(3) van Reyk, D. M.; Brown, A. J.; Hult’en, L. M.; Dean, R. T.; Jessup,
W. Oxysterols in biological systems: sources, metabolism and patho-
physiological relevance. Redox Rep. 2006, 11, 255–262.
(4) Olkkonen, V. M. New Functions for Oxysterols and Their Cellular
Receptors. Lipids Insights 2008, 2, 1–9.
(30) Lakner, F. J.; Hager, L. P. Chloroperoxidase-mediated asymmetric
epoxidation. Synthesis of (R)-dimethyl 2-methylaziridine-1,2-

Synthesis, Cytotoxicity, and SAR of Epoxysterols
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 13 4019
dicarboxylatesa potential alpha-methylamino acid synthon. Tetrahe-
dron: Asymmetry 1997, 8, 3547–3550.
(54) Goldstein, L. J. MDR1 gene expression in solid tumours. Eur. J.
Cancer 1996, 32, 1039–1050.
(55) Ridgway, N. D.; Dawson, P. A.; Ho, Y. K.; Brown, M. S.; Goldstein,
J. L. Translocation of oxysterol binding protein to Golgi apparatus
triggered by ligand binding. J. Cell Biol. 1992, 116, 307–319.
(56) Brown, A. J.; Leong, S. L.; Dean, R. T.; Jessup, W. 7-Hydroperoxy-
cholesterol and its products in oxidized low density lipoprotein and
human atherosclerotic plaque. J. Lipid Res. 1997, 38, 1730–1745.
(57) Kang, K. A.; Chae, S.; Lee, K. H.; Park, M. T.; Lee, S. J.; Lee, Y. S.;
Hyun, J. W. Cytotoxic Effect of 7ꢀ-Hydroxycholesterol on Human
NCI-H460 Lung Cancer Cells. Biol. Pharm. Bull. 2005, 28, 1377–
1380.
(58) Maier, G.; Bing, G.; Falken, U.; Wagner, E.; Unger, C. Antitumor
activity and induction of apoptosis by water-soluble derivatives of
7-beta-hydroxycholesterol in human colon carcinoma cell lines.
Anticancer Res. 1999, 19, 4251–4256.
(59) Roussi, S.; Gosse´, F.; Aoude´-Werner, D.; Zhang, X.; Marchioni, E.;
Geoffroy, P.; Miesch, M.; Raul, F. Mitochondrial perturbation,
oxidative stress and lysosomal destabilization are involved in 7-hy-
droxysitosterol and 7-hydroxycholesterol triggered apoptosis in human
colon cancer cells. Apoptosis 2007, 12, 87–96.
(60) Lordan, S. A.; Mackrill, J. J.; O’Brien, N. M. Involvement of Fas
Signalling in 7ꢀ-Hydroxycholesterol-and Cholesterol-5ꢀ,6ꢀ-Epoxide-
Induced Apoptosis. Int. J. Toxicol. 2008, 27, 279.
(61) Pirillo, A.; Zhu, W.; Roma, P.; Galli, G.; Caruso, D.; Pellegatta, F.;
Catapano, A. L. Oxysterols from oxidized LDL are cytotoxic but fail
to induce hsp70 expression in endothelial cells. FEBS Lett. 1999, 462,
113–116.
(62) O’Sullivan, A. J.; O’Callaghan, Y. C.; Woods, J. A.; O’Brien, N. M.
Toxicity of cholesterol oxidation products to Caco-2 and HepG2 cells:
modulatory effects of R- and γ-tocopherol. J. Appl. Toxicol. 2003,
23, 191–197.
(63) Ryan, L.; O’Callaghan, Y. C.; O’Brien, N. M. The role of the
mitochondria in apoptosis induced by 7ꢀ-hydroxycholesterol and
cholesterol-5ꢀ,6ꢀ-epoxide. Br. J. Nutr. 2005, 94, 519–525.
(64) Macdonald, I. A. Detection of bile salts with Komarowsky’s reagent
and group specific dehydrogenases. J. Chromatogr. 1977, 136, 348–
352.
(65) Kimura, M.; Muto, T. The Reactions of Cholesteryl Acetate with tert-
Butyl Hydroperoxide and Molybdenum Complexes. Chem. Pharm.
Bull. 1981, 29, 35–42.
(31) Farinas, E. T.; Alcalde, M.; Arnold, F. Alkene epoxidation catalyzed
by cytochrome P450 BM-3 139-3. Tetrahedron 2004, 60, 525–528.
(32) Rai, G. P.; Sakai, S.; Flo´rez, A. M.; Mogollon, L.; Hager, L. P. Directed
Evolution of Chloroperoxidase for Improved Epoxidation and Chlo-
rination Catalysis. AdV. Synth. Catal. 2001, 343, 638–645.
(33) Choi, W. J.; Choi, C. Y. Production of Chiral Epoxides: Epoxide
Hydrolase-catalyzed Enantioselective Hydrolysis. Biotechnol. Biopro-
cess. Eng. 2005, 10, 167–179.
(34) Cruz Silva, M. M.; Riva, S.; Sa´ e Melo, M. L. Highly selective lipase-
mediated discrimination of diastereomeric 5,6-epoxysteroids. Tetra-
hedron: Asymmetry 2004, 15, 1173–1179.
(35) Carvalho, J. F. S.; Silva, M. M. C.; Sa´ e Melo, M. L. Highly efficient
epoxidation of unsaturated steroids using magnesium bis(monoper-
oxyphthalate) hexahydrate. Tetrahedron 2009, 65, 2773–2781.
(36) Bjorkhem, I.; Diczfalusy, U. Oxysterols: Friends, Foes, or Just Fellow
Passengers? Arterioscler. Thromb. Vasc. Biol. 2002, 22, 734–742.
(37) Bodin, K.; Bretillon, L.; Aden, Y.; Bertilsson, L.; Broome, U.;
Einarsson, C.; Diczfalusy, U. Antiepileptic Drugs Increase Plasma
Levels of 4beta-Hydroxycholesterol in Humans. Evidence for Involve-
ment of Cytochrome P450 3A4. J. Biol. Chem. 2001, 276, 38685–
38689.
(38) Princen, H. M. G.; Post, S. M.; Twisk, J. Regulation of Bile Acid
Biosynthesis. Curr. Pharm. Des. 1997, 3, 59–84.
(39) Prunet, C.; Petit, J. M.; Ecarnot-Laubriet, A.; Athias, A.; Miguet-
Alfonsi, C.; Rohmer, J. F.; Steinmetz, E.; Ne´el, D.; Gambert, P.;
Lizard, G. High circulating levels of 7[beta]- and 7[alpha]-hydroxy-
cholesterol and presence of apoptotic and oxidative markers in arterial
lesions of normocholesterolemic atherosclerotic patients undergoing
endarterectomy. Pathol. Biol. 2006, 54, 22–32.
(40) Ryan, L.; Callaghan, Y. C. O.; Brien, N. M. O. Comparison of the
apoptotic processes induced by the oxysterols 7beta-hydroxycholesterol
and cholesterol-5beta,6beta-epoxide. Cell Biol. Toxicol. 2004, 20, 313–
323.
(41) Lathe, R. Steroid and sterol 7-hydroxylation: ancient pathways. Steroids
2002, 67, 967–977.
(42) Salvador, J. A. R.; Sa´ e Melo, M. L.; Neves, A. S. C. Copper-Catalysed
Allylic Oxidation of Delta5-Steroids by t-Butyl Hydroperoxide.
Tetrahedron Lett. 1997, 38, 119–122.
(43) Arsenou, E. S.; Koutsourea, A. I.; Fousteris, M. A.; Nikolaropoulos,
S. S. Optimization of the allylic oxidation in the synthesis of 7-keto-
[delta]5-steroidal substrates. Steroids 2003, 68, 407–414.
(44) Kumar, V.; Amann, A.; Ourisson, G.; Luu, B. Stereospecific Syntheses
of 7ꢀ-AND 7R-Hydroxycholesterols. Synth. Commun. 1987, 17, 1279–
1286.
(45) Rossiter, B. E.; Verhoeven, T. R.; Sharpless, K. B. Stereoselective
epoxidation of acyclic allylic alcohols. A correction of our previous
work. Tetrahedron Lett. 1979, 20, 4733–4736.
(46) Zhao, K.; Wang, Y.; Billington, D. C. Studies on stereocontrolled
epoxidations of bis-alicyclic alcohols in steroidal skeletons: preparation
of eight diastereomerically pure epoxides from cholest-4-en-3beta,
6beta-;-3beta,6alpha-;-3alpha,6beta-, and-3alpha,6alpha-diols. Tet-
rahedron: Asymmetry 2001, 12, 1211–1217.
(47) Cruz Silva, M. M.; Riva, S.; Sa´ e Melo, M. L. Regioselective enzymatic
acylation of vicinal diols of steroids. Tetrahedron 2005, 61, 3065–
3073.
(48) O’Brien, J.; Wilson, I.; Orton, T.; Pognan, F. Investigation of the
alamar blue (resazurin) fluorescent dye for the assessment of mam-
malian cell cytotoxicity. Eur. J. Biochem. 2000, 267, 5421–5426.
(49) Hamid, R.; Rotshteyn, Y.; Rabadi, L.; Parikh, R.; Bullock, P.
Comparison of alamar blue and MTT assays for high through-put
screening. Toxicol. in Vitro 2004, 18, 703–710.
(50) Cerda, S.; Wilkinson, J.; Broitman, S. Regulation of cholesterol
synthesis in four colonic adenocarcinoma cell lines. Lipids 1995, 30,
1083–1092.
(51) Ikonen, E. Cellular cholesterol trafficking and compartmentalization.
Nat. ReV. Mol. Cell Biol. 2008, 9, 125–138.
(52) Cummings, J.; Anderson, L.; Willmott, N.; Smyth, J. F. The molecular
pharmacology of doxorubicin in vivo. Eur. J. Cancer 1991, 27, 532–
535.
(66) Li, S.; Pang, J.; Wilson, W. K.; Schroepfer, J. Sterol synthesis.
Preparation and characterization of fluorinated and deuterated analogs
of oxygenated derivatives of cholesterol. Chem. Phys. Lipids 1999,
99, 33–71.
(67) Nickon, A.; Bagli, J. F. Reactivity and Geometry in Allylic Systems.
I. Stereochemistry of Photosensitized Oxygenation of Monoolefins.
J. Am. Chem. Soc. 1961, 83, 1498–1508.
(68) Saito, A.; Tanaka, A.; Oritani, T. Highly Diastereoselective Reduction
of Bicyclic and Steroidal Conjugated Enones with Dibal in THF-DME.
Synth. Commun. 1997, 27, 1969–1973.
(69) Chakravorty, P. N.; Levin, R. H. Studies on Cholesteryl Oxides1.
J. Am. Chem. Soc. 1942, 64, 2317–2322.
(70) Baxter, R. A.; Spring, F. S. ꢀ-Cholesterol oxide. J. Chem. Soc. 1943,
613–615.
(71) Romero, M. NotessReaction of NBS with Allylic Alcohols. J. Org.
Chem. 1957, 22, 1267–1268.
(72) Kawata, M.; Tohma, M.; Sawaya, T.; Kimura, M. A New Product of
Cholesterol by Metal-Free Autoxidation in Aqueous Dispersion. Chem.
Pharm. Bull. 1976, 24, 3109–3113.
(73) Fieser, L. F.; Goto, T.; Bhattacharyya, B. K. Identification of Ketone
104 as 3,4-Secocholestane-6-one-(3R,5R) (3ꢀ,4)-dioxide. J. Am. Chem.
Soc. 1960, 82, 1700–1706.
(74) Collins, D. J. Steroidal Rꢀ-epoxy-ketones. Part I. Rearrangement of
4R,5-epoxy-5a-cholestan-3-one and its 4ꢀ,5ꢀ-isomer by means of the
boron trifluoride-ether complex. J. Chem. Soc. 1959, 3919–3928.
(75) Cicala, G.; Curci, R.; Fiorentino, M.; Laricchiuta, O. Stereo- and
regioselectivities in the epoxidation of some allylic alcohols by the
dioxirane intermediate generated in the reaction of potassium caroate
with acetone. J. Org. Chem. 1982, 47, 2670–2673.
(76) Dauben, W. G.; Boswell, G. A.; Templeton, W.; McFarland, J. W.;
Berezin, G. H. The Configuration of B-Norsteroid Derivatives. J. Am.
Chem. Soc. 1963, 85, 1672–1676.
(53) Jemal, A.; Siegel, R.; Ward, E.; Hao, Y.; Xu, J.; Murray, T.; Thun,
M. J. Cancer Statistics, 2008. Can. Cancer J. Clin. 2008, 58, 71–96.
JM9003973