2,3,18,19-Dioxidosqualene stereoisomers: characterization and activity as inhibitors of purified pig liver 2,3-oxidosqualene-lanosterol cyclase

Full text of DOI:10.1021/jo9607020

J . Org. Chem. 1996, 61, 7603-7607
7603
Sch em e 1
2,3,18,19-Dioxid osqu a len e Ster eoisom er s:
Ch a r a cter iza tion a n d Activity a s In h ibitor s
of P u r ified P ig Liver
2,3-Oxid osqu a len e-La n oster ol Cycla se
J ose´-Luis Abad, Marisa Guardiola, J osefina Casas,
Francisco Sa´nchez-Baeza, and Angel Messeguer*
Department of Biological Organic Chemistry,
C.I.D. (C.S.I.C.), J . Girona, 18. 08034 Barcelona, Spain
Received April 17, 1996
inhibitor of this enzyme.15 Recently, improved substrate
and intermediate mimic inhibitors containing either
nitrogen16,17 or sulfur18 moieties have been described.
From these compounds, the analog bearing a sulfur atom
at C-19 in the squalene skeleton has resulted in the most
powerful inhibitor to date for rat liver cyclase (IC50 ) 0.8
nM).19 In this context, our approach was to explore the
OSLC potential inhibitory activity of substrate mimics
containing the epoxy moiety. We reasoned that this class
of compounds might be formed endogenously under stress
or unregulated oxidation conditions. Accordingly, the
preparation of internal oxidosqualenes and all possible
dioxidosqualenes as racemates and their evaluation as
OSLC inhibitors in rat liver microsomes were carried out.
We found that 2,3,18,19-dioxidosqualene (1) elicited the
highest activity as OSLC inhibitor (IC50 ) 0.11 µM).20,21
More recently, we reported that (6S,7S)-oxidosqualene
is a strong competitive inhibitor of purified pig liver
squalene epoxidase (SE, EC 1.14.99.7) (IC50 ) 6.7 µM;
Ki ) 2.7 µM). Furthermore, incubation of the enantiomer
of the above epoxy derivative, i.e., (6R,7R)-oxidosqualene
with purified SE led to the formation of dioxide
1-(3S,18R,19R) (Scheme 1) as major compound.22 In
view of these results, it was then interesting to identify
the specific stereoisomers of dioxidosqualene 1 which
were contributing to the OSLC inhibitory activity exhib-
ited by the racemic mixture.
In tr od u ction
2,3-Oxidosqualene-lanosterol cyclase (OSLC, EC
5.4.99.7) is the key enzyme of the ring cyclization reaction
suffered by (3S)-2,3-oxidosqualene to give lanosterol as
final steroid precursor in fungi and mammalian systems.1
The mechanism of this cyclization is still a subject of
debate, although current hypotheses are in favor of a
stepwise process, which proceeds through a serie of
conformationally rigid, partially cyclized carbocationic
intermediates. These studies have been supported by the
recent advances in the characterization and purification
of OSLC. Thus, several cyclases have been purified to
homogeneity from different sources2-6 and OSLC from
Candida albicans has been cloned and sequenced.7,8
More recently, the cloning of rat9,10 and human liver11
OSLC cDNA has been reported.
The availability of potent inhibitors of OSLC consti-
tutes also a valuable tool for unraveling specific features
of the cyclization mechanism. Another objective of the
search of efficient OSLC inhibitors seeks the development
of potential hypocholesteremic agents.12 As it occurs with
other enzyme systems related to the squalene biosyn-
thesis pathway, the fact that the OSLC step is beyond
the synthesis of other physiologically relevant terpenoid
compounds, such as farnesyl derivatives, dolichol and
ubiquinones, makes this strategy therapeutically attrac-
tive. On the other hand, it has been recently pointed out
that OSLC inhibitors can modulate sterol biosynthesis
by acting on HMG-CoA reductase by a feedback mecha-
nism involving the participation of oxysterols.13
To this aim, the present paper reports the preparation
and determination of the absolute configuration of the
different stereoisomers of dioxidosqualene 1, i.e.,
1-(3S,18S,19S), 1-(3S,18R,19R), 1-(3R,18S,19S), and
1-(3R,18R,19R) (Scheme 1), and their activities as
inhibitors of OSLC purified from pig liver. Results on
the determination of the kinetic parameters for the two
most active compounds and on their possible irreversible
interaction with OSLC are also reported.
Numerous OSLC inhibitors have elicited potent activ-
ity.14 Among them 29-methylidene-2,3-oxidosqualene
was reported as the first mechanism-based irreversible
(1) Abe, I.; Rohmer, M.; Prestwich, G. D. Chem. Rev. 1993, 93, 2189-
2206.
Resu lts a n d Discu ssion
(2) Saar, J .; Kader, J . C.; Poralla, K.; Ourisson, G. Biochim. Biophys.
Acta 1991, 1075, 93-101.
P r ep a r a tion a n d Deter m in a tion of th e Absolu te
Con figu r a t ion of t h e F ou r St er eoisom er s of Di-
(3) Abe, I.; Bai, M.; Xiao, X. Y.; Prestwich, G. D. Biochem. Biophys.
Res. Commun. 1992, 187, 32-38.
(4) Abe, I.; Sankawa, W.; Ebizuka, Y. Chem. Pharm. Bull. 1992, 40,
1755-1760.
(14) For a comprehensive review on OSLC inhibitors, see: Abe, I.;
Tomesch, J . C.; Wattanasin, S.; Prestwich, G. D. Nat. Prod. Rep. 1994,
11, 279-302.
(5) Corey, E. J .; Matsuda, S. P. T. J . Am. Chem. Soc. 1991, 113,
8172-8174.
(15) Xiao, X. Y.; Prestwich, G. D. J . Am. Chem. Soc. 1991, 113,
9673-9674.
(6) Ochs, D.; Tappe, C. H.; Gaertner, P.; Kellner, R.; Poralla, K. Eur.
J . Biochem. 1990, 194, 75-80.
(16) Hoshino, T.; Kobayashi, N.; Ishibashi, E.; Hashimoto, S. Biosci.
Biotech. Biochem. 1995, 59, 602-609.
(7) Buntel, C. J .; Griffin, J . H. J . J . Am. Chem. Soc. 1992, 114, 9711-
9713.
(17) Cattel, L.; Ceruti, M.; Balliano, G.; Viola, F.; Grosa, G.; Rocco,
F.; Brusa, P. Lipids 1995, 30, 235-246.
(8) Roessner, C. A.; Min, C.; Hardin, S. H.; Harris-Haller, L. W.;
McCollum, J . C.; Scott, A. I. Gene 1993, 127, 149-150.
(9) Abe, I.; Prestwich, G. D. Proc. Natl. Acad. Sci. U.S.A. 1995, 92,
9274-9278.
(10) Kusano, M.; Shibuya, M.; Sankawa, U.; Ebizuka, Y. Biol.
Pharm. Bull. 1995, 18, 195-197.
(18) Zheng, Y. F.; Dodd, D. S.; Oehlschlager, A. C.; Hartman, P. G.
Tetrahedron 1995, 51, 5255-5276.
(19) Zheng, Y. F.; Oehlschlager, A. C.; Georgopapadakou, N. H.;
Hartman, P. G.; Scheliga, P. J . Am. Chem. Soc. 1995, 117, 670-680.
(20) Abad, J . L.; Casas, J .; Sa´nchez-Baeza, F.; Messeguer, A. Bioorg.
Med. Chem. Lett. 1992, 2, 1239-1242.
(11) Baker, C. H.; Matsuda, S. P. T.; Liu, D. R.; Corey, E. J . Biochem.
Biophys. Res. Commun. 1995, 213, 154-160.
(12) Abe, I.; Tomesch, J . C.; Wattanasin, S.; Prestwich, G. D. Nat.
Prod. Rep. 1994, 11, 279-302.
(21) Abad, J . L.; Casas, J .; Sa´nchez-Baeza, F.; Messeguer, A. J . Org.
Chem. 1993, 58, 3991-3997.
(22) Abad, J . L.; Casas, J .; Sa´nchez-Baeza, F.; Messeguer, A. J . Org.
Chem. 1995, 60, 3648-3656.
(13) Spencer, T. A. Acc. Chem. Res. 1994, 27, 83-90.
S0022-3263(96)00702-5 CCC: $12.00 © 1996 American Chemical Society

7604 J . Org. Chem., Vol. 61, No. 21, 1996
Notes
Sch em e 2a
a
(a) (DHQD)2-PHAL/K3Fe(CN)6/K2CO3/CH3SO2NH2/K2OsO4/tBuOH/H2O/0 °C; (b) CH3SO2Cl/Et3N/CH2Cl2; (c) NaH/THF; (d) HClO4/
THF/H2O; (e) (S)-MTPA-Cl/Et3N/DMAP/CH2Cl2; (f) semipreparative reversed-phase HPLC; (g) LiAlH4, Et2O.
oxid osqu a len e 1. Our previous work on the prepara-
tion of the different internal oxidosqualene stereoisomers
used a strategy involving the formation of the (R)-MTPA
Mosher esters of the corresponding erythro internal
squalenediol mixtures. These esters were employed for
the chromatographic resolution of these mixtures and the
further determination of the absolute configuration of the
stereogenic centers present at the squalene skeleton for
reagent (DHQD)2-PHAL (Scheme 2). This reaction led
to the diastereomer mixture of tetraols 3 with (3R)
configuration. The derivatization of these tetraols with
(S)-MTPACl afforded the corresponding Mosher diesters
2 (one of the constituents of the second peak and the third
peak in the above HPLC profile). These diesters were
separated by semipreparative HPLC with good recover-
ies. Conversely, the low reaction rates and enantiomeric
excesses usually obtained with the (DHQ)2-PHAL re-
agent led us to use an alternative procedure for the
preparation of the two remaining tetraol stereoisomers
3 with (3S) configuration. Thus, the diastereomeric
mixture of tetraols 3-(3R) was converted into the corre-
sponding dioxidosqualenes 1-(3S) by mesylation and
treatment with base in 78% overall yield. The acid
hydrolysis of this mixture afforded the pair of tetraols 3
with (3S) configuration. The Mosher diesters 2 derived
from these tetraols were prepared and separated as
described above, although with poorer recoveries in this
case. Finally, the different dioxidosqualenes 1 were
obtained from the respective diester 2 in 55-60% overall
yields by treatment with LiAlH4 to give the tetraol,
followed by mesylation and ring closure. The stereo-
chemical purity of these dioxidosqualenes was deter-
mined by the conversion into their corresponding tetraols
3 followed by formation of the MTPA esters and analysis
1
each diastereomer. For this latter purpose, the H NMR
procedure developed by the group of Kakisawa23 was
adapted to the case of enantiomer mixtures from syn-
thetic origin. The final conversion of the above esters
into the stereochemically related oxidosqualenes with
high chemical and optical purities supported the validity
of this approach.22
Consequently, we decided to use a similar strategy for
synthesizing the four stereoisomers of dioxidosqualene
1. In this case, however, the weak interactions expected
between the stereocenter at C-3 and those at C-18 and
C-19 would anticipate a more difficult chromatographic
resolution of the diastereomeric mixtures of the corre-
sponding Mosher diesters 2 (Scheme 1). Thus, acid
hydrolysis of racemic dioxide 1 afforded the correspond-
ing tetraol racemate 3 which was converted into the
mixture of diastereomers 2 by treatment with (S)-
MTPACl following the general procedure previously
described.22 Unfortunately, all assays carried out to
resolve this mixture by reversed-phase HPLC led to the
obtention of three peaks in a 1:2:1 ratio, which indicated
the coelution of two diastereomers. Moreover, when
these coeluted diastereomers were treated with LiAlH4
and the resulting tetraols were rederivatized with (R)-
MTPACl, again a diastereomeric mixture not separable
by chromatographic means was obtained.
1
of these derivatives by HPLC, H NMR, or 19F NMR. A
racemization lower than 5% was observed for the conver-
sion of the chiral squalenetetraol into the bisepoxy
derivative and vice versa.
The absolute configuration of dioxidosqualenes 1 was
determined by adapting the method reported by the
group of Kakisawa to the case of stereoisomers mixtures
1
from synthetic origin.22 For this purpose, the H NMR
spectra for the four different Mosher diesters 2 were
registered, and the chemical shifts for all protons with
respect to each diastereomer were assigned. The chemi-
cal shifts differences (∆δ) between pairs of diastereomer
diesters from the same synthetic origin were not signifi-
Then we envisaged another approach involving the
asymmetric dihydroxylation of diol 4 with the Sharpless
(23) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J . Am.
Chem. Soc. 1991, 113, 4092-4096.

Notes
J . Org. Chem., Vol. 61, No. 21, 1996 7605
Sch em e 3. 1H NMR Ch em ica l Sh ift Differ en ces (in
p p m ) for All P osition s of th e Squ a len e Sk eleton
betw een th e F ou r Mosh er Diester s 2, P r ecu r sor s
of Dioxid osqu a len e Ster eoisom er s 1
Ta ble 1. IC50 a n d Ki Va lu es of P u r ified P ig Liver
2,3-Oxid osqu a len e-La n oster ol Cycla se In h ibition for
2,3,18,19-Dioxid osqu a len e Ster eoisom er s 1a
compdb
IC50 (µM)
0.12
Ki (µM)
rac-1
n.d.
1-(3R,18S,19S)
1-(3R,18R,19R)
1-(3S,18S,19S)
1-(3S,18R,19R)
0.37
0.22
0.071 (0.062)
0.037 (0.027)
n.d.
n.d.
0.076
0.021
a
For assay details and concentrations of inhibitors used, see
supporting information. Substrate concentration was 20 µM;
b
values in parentheses were determined by using 5 µM. The
enantiomeric excess determined for diepoxides 1-(3R,18S,19S),
1-(3R,18R,19R), 1-(3S,18S,19S), and 1-(3S,18R,19R) were 90%,
94%, 95%, and 95%, respectively.
cant for hydrogen atoms in the proximity of C-3 (Scheme
3). Conversely, the ∆δ values obtained for the hydrogen
atoms in the vicinity of C-18 and C-19 made possible the
assignment of the absolute configuration for these ste-
reocenters (Scheme 3, cf. a and b). On the other hand, in
addition to the stereochemical reference derived from the
(DHQD)2-PHAL-promoted asymmetric dihydroxylation
of the terminal double bond mentioned above (cf. Scheme
2), the absolute configuration of the stereocenters at C-3
was confirmed by comparing the ∆δ values from Mosher
diesters with the same configuration at C-18 and C-19.
Once the four Mosher diesters were fully characterized,
the absolute configuration of the respective dioxido-
squalenes 1 could be unequivocally inferred.
In h ibition of OSLC Activity. OSLC assays with
dioxidosqualenes 1-(3S,18S,19S), 1-(3S,18R,19R), 1-(3R,
18S,19S), and 1-(3R,18R,19R) were carried out employ-
ing the partially purified enzyme from pig liver and
following the procedure described by Duriatti et al.24
Previously, the inhibitory activity due to (3R)-2,3-oxido-
squalene was tested and found negligible, which permit-
ted the use of 2,3-oxidosqualene as substrate for the
OSLC activity assays. In this sense, internal oxides
(6S,7S)- and (6R,7R)-oxidosqualenes22 exhibited weak
inhibitory activities (IC50 ) 81 and 190 µM, respectively,
using 20 µM of substrate), which discarded that the high
activities shown below were due only to the presence of
the internal epoxy moiety.
F igu r e 1. Lineweaver-Burk plots of the inhibition of purified
pig liver OSLC by (a) 1-(3S,18S,19S) and (b) 1-(3S,18R,19R).
The concentrations of inhibitors used were 0 (squares), 0.05
(diamonds), 0.1 (circles), and 0.2 (triangles) µM, and 0 (squares),
0.025 (diamonds), 0.05 (circles), and 0.1 (triangles) µM,
respectively.
configuration, the most active compound was dioxide
1-(3S,18R,19R) (IC50 ) 27 nM using 5 µM of substrate).
This compound is within the best OSLC inhibitors
described to date.12,19 Dioxide 1-(3S,18S,19S) displayed
also a high inhibitory activity (IC50 ) 62 nM using 5 µM
of substrate). From the kinetic data obtained for this pair
of stereoisomers, dioxide 1-(3S,18R,19R) elicited a non-
competitive inhibition (Ki ) 21 nM), whereas dioxide
1-(3S,18S,19S) exhibited a linear-mixed pattern inhibi-
tion (Ki ) 76 nM) (Table 1, Figure 1). On the other hand,
these results support the inhibition data previously found
for the racemic mixture of dioxide 1 in rat liver mi-
crosomes, where it can be assumed that stereoisomers
with (3S) configuration contributed predominantly to the
inhibitory potency obtained in that case.21 Finally,
assays carried out as described by Bai and Prestwich25
showed that the OSLC inhibition caused by dioxido-
squalenes 1-(3S,18R,19R) and 1-(3S,18S,19S) was re-
versible.
The results shown above deserve some additional
comments. First, the comparison from the activities of
internal oxidosqualenes with those of dioxides 1 indicates
that both epoxy moieties are needed for eliciting a potent
OSLC inhibition, which confirmed our previous results
on rat liver microsomes.21 In addition, an exam of the
numerous OSLC inhibitors described to date points out
Conversely, dioxidosqualenes 1 elicited potent inhibi-
tory activities on OSLC (Table 1). As expected, those
stereoisomers with (3S) configuration were more active,
approximately one order of magnitude, than those with
(3R) configuration. From the pair of dioxides 1 with (3S)
(24) Duriatti, A.; Bouvier-Nave´, P.; Benveniste, P.; Schuber, F.;
Delprino, L.; Balliano, G.; Cattel, L. Biochem. Pharmacol. 1985, 34,
2765-2777.
(25) Bai, M.; Prestwich, G. D. Arch. Biochem. Biophys. 1992, 293,
305-313.

7606 J . Org. Chem., Vol. 61, No. 21, 1996
Notes
over MgSO4, and the purification procedures were carried out
by flash chromatography on silica gel. In some cases, the silica
gel was previously impregnated with Et3N to minimize decom-
position of squalene derivatives.
the importance of structural modifications in the vicinity
of the C18-C19 position of the squalene skeleton. Thus,
the most potent OSLC inhibitor reported so far bears a
sulfur atom at C-19,19 and the first mechanism-based
OSLC irreversible inhibitor presents a methylidene
moiety at C-29, which is the methyl group linked to
C-19.15 The case of dioxides 1-(3S,18R,19R) and
1-(3S,18S,19S) corroborates the above observation. Re-
garding the stereochemical features at the C18-C19
region, our results suggest that both epoxide configura-
tions seem to interact with the enzyme, although the
effect reached by the (R,R) stereoisomer leads to the
higher inhibitory activity. This observation may serve
for the future design of improved substrate-mimic inhibi-
tors. On the other hand, the fact that dioxides with (3R)
configuration showed inhibitory activities in the nano-
molar range could be partially due to the increased
polarity of these compounds in comparison with oxido-
squalenes, which would facilitate the interaction with the
enzyme, and to the presence of minor amounts of stereo-
isomers with (3S) configuration (cf. footnotes in Table 1).
The negative results shown by the irreversibility
assays performed on these dioxides suggest that no
covalent interaction with the enzyme occurs. In this
sense, it is conceivable that the presence of moieties more
reactive that these trisubstituted epoxides might facili-
tate the establishment of stronger interactions with
OSLC.
Asym m etr ic Dih yd r oxyla tion of 6,7-Dih yd r oxy-6,7-d i-
h yd r osqu a len e. Tetraol 3-(3R) was prepared by using the
procedure reported by Crispino and Sharpless (Tetrahedron Lett.
1992, 33, 4273-4274). Thus, treatment of 0.88 g (2 mmol) of
6(R*),7(S*)-d ih yd r oxy-6,7-d ih yd r osqu a len e (4)22 with the
mixture of reagents containing the ligand (DHQD)2-PHAL for
48 h at 0 °C led to a crude reaction mixture which was purified
by chromatography to give, among other tetraol derivatives, 0.05
g (5% yield) of compound 3-(3R). 3(R),18(R*),19(S*)-2,3,18,19-
tetr ah ydr oxy-2,3,18,19-tetr ah ydr osqu alen e 3-(3R): IR: 3415
(OH); 1H NMR: δ 5.30-5.02 (4 H), 3.36 (2 H), 2.34-1.88 (16 H
(2 OH)),1.69 (3 H), 1.60 (br, 3 H), 1.80-1.05 (4 H), 1.20 (s, 3 H),
1.17 (s, 3 H), 1.15 (s, 3 H); 13C NMR: δ 135.1, 134.8, 131.8, 125.1,
125.0, 124.6, 124.3, 78.3, 78.2, 74.6, 73.0, 39.6, 36.9, 36.8, 35.9,
29.6, 29.2, 28.1, 28.1, 26.3, 26.3, 25.7, 23.4, 23.3, 22.1, 17.7, 15.9,
15.9; MS (m/ z) (relative intensity): 478 (M, 32), 460 (M - 18,
10), 443 (M + 1 - 36, 100), 425 (M + 1 - 48, 80). Anal. Calcd.
for C30H54O4: C; 75.26; H; 11.37. Found: C; 75.17; H; 11.48.
Con ver sion of Tetr a ol 3-(3R) in to Tetr a ol 3-(3S). This
conversion was carried out following the procedure described
previously for the case of terminal squalenediols.22 In this case,
reaction of tetraol 3-(3R) (48 mg, 0.1 mmol) with CH3SO3Cl (18
µL, 0.22 mmol) and Et3N (60 µL, 0.40 mmol) in CH2Cl2 for 15
min at -10 °C afforded the expected bismesylate, which was
allowed to react with NaH (24 mg, 1 mmol) in THF solution, to
give after chromatography purification diepoxide 1-(3S) (35 mg,
78% overall yield). Treatment of a solution of this diepoxide (35
mg, 0.08 mmol) in 5 mL of a 4:1 mixture of THF:H2O with 10
µL of 60% HClO4 for 24 h at 20 °C, followed by purification of
the crude reaction mixture by chromatography, afforded tetraol
3-(3S) (28 mg, 74% yield).
Finally, the fact that internal oxidosqualenes have
been described in nature,26 and that (6R,7R)-oxidosqualene
is substrate of purified SE from pig liver to give precisely
dioxide 1-(3S,18R,19R) as major product,22 confers a
particular interest to the potential role of this inhibitor
in the steroidogenesis pathway. On the other hand,
research is on progress in our laboratory to find out the
presence or absence of OSLC-promoted cyclization prod-
ucts for the pair of dioxides 1 with (3S) configuration,
which would provide important information on the
inhibition mechanism of these compounds.
P r epar ation of MTP A Diester s 2-(3R,18S,19R), 2-(3R,18R,
19S), 2-(3S,18S,19R), a n d 2-(3S,18R,19S). These compounds
were obtained by using the procedure described previously for
the case of squalenediols.22 Thus, starting from tetraol 3-(3R)
(30 mg, 0.06 mmol), the expected mixture of diesters
2-(3R,18S,19R) and 2-(3R,18R,19S) (49 mg, 90% yield) was
isolated. This mixture was separated by semipreparative HPLC
(15 × 1 cm ODS-2 column, 10 µm, eluting with 79:21 CH3CN:
H2O at 3.6 mL/min), followed by chromatography of the collected
compound, to give 18 mg of each diastereomer. Diest er
2-(3R,18S,19R): [R]D ) +18.0 (c ) 1.5, 95% ee); 1H NMR: δ
7.70-7.52 (4 H), 7.47-7.32 (6 H), 5.22-4.94 (6 H), 3.59 (q, 3 H,
J ) 1.5 Hz), 3.57 (q, 3 H, J ) 1.5 Hz), 2.14-1.80 (14 H), 1.67 (s,
3 H), 1.59 (br, 6 H), 1.56 (d, 3 H, J ) 1 Hz), 1.52 (s, 3 H), 1.80-
1.10 (6 H), 1.22 (s, 3 H), 1.16 (s, 3 H), 1.12 (s, 3 H); 13C NMR:
δ 166.9, 166.2, 135.1, 133.8, 133.5, 132.3, 132.1, 129.6, 128.4,
Exp er im en ta l Section
Ap p a r a tu s. The HPLC analysis was performed by using a
Spherisorb ODS-2 (5 µm) column and eluting with CH3CN-H2O
mixtures at 1 mL/min. The gas chromatography-mass spec-
trometry analysis was performed with positive chemical ioniza-
tion (GC-MS-CI), using methane as ionization gas and a 30 m
HP-5 bonded phase capillary column (0.25 mm i.d.). The liquid
chromatography-thermospray-mass spectrometry analysis
(HPLC-TSP-MS) was carried out with a quadrupole apparatus
(direct flow injection with 50 mM HCO2NH4/CH3CN (50:50) at
1 mL/min; positive mode; TSP tip, 180 °C; TSP stem, 96 °C; and
TSP ion source, 250 °C). Optical rotations were determined at
25 °C in CHCl3 solution at the specified concentration (expressed
in g/L, 10 cm cell). The enantiomeric excess (ee) values were
calculated by HPLC analysis of the corresponding (R)-MTPA
diester derivatives.
127.6, 127.4, 125.3, 125.2, 124.2, 124.0, 123.4 (q, CF3, J F-C
)
287 Hz), 84.4 and 84.4 (q, C, J F-C ) 28 Hz), 82.4, 82.0, 74.1,
72.7, 55.5, 39.6, 36.9, 36.3, 35.7, 28.7, 28.3, 28.3, 28.1, 27.0, 26.7,
25.7, 23.7, 23.2, 21.8, 17.6, 16.0, 15.9, 15.8; 19F NMR: δ -71.23,
-71.26; HPLC-TSP-MS: 928 (M + 18), 911 (M + 1), 893 (M -
18 + 1), 694 (M - 217 + 1). Diester 2-(3R,18R,19S): [R]D
)
+27.0 (c ) 1.5, 95% ee); 1H NMR: δ 7.70-7.54 (4 H), 7.47-7.32
(6 H), 5.22-4.94 (6 H), 3.58 (q, 3 H, J ) 1 Hz), 3.55 (q, 3 H, J )
1 Hz), 2.20-1.80 (14 H), 1.68 (s, 3 H), 1.59 (br, 6 H), 1.52 (br, 6
H), 1.75-1.10 (6 H), 1.22 (s, 3 H), 1.17 (s, 3 H), 1.16 (s, 3 H); 13
C
NMR: δ 166.9, 166.7, 135.1, 133.8, 133.5, 132.2, 132.1, 132.0,
129.6, 128.4, 127.7, 127.6, 125.2, 125.2, 124.2, 124.1, 123.4 (q,
CF3, J F-C ) 287 Hz), 84.7 (q, C, J F-C ) 28 Hz), 82.4, 82.2, 74.2,
72.7, 55.5, 55.4, 39.6, 36.3, 36.0, 35.8, 28.7, 28.3, 28.2, 28.1, 27.0,
26.6, 25.7, 23.9, 23.7, 21.7, 17.6, 16.0, 15.9, 15.8; 19F NMR: δ
-71.06, -71.23; HPLC-TSP-MS: 928 (M + 18), 911 (M + 1),
893 (M - 18 + 18), 694 (M - 217 + 1).
By a similar procedure, tetraol 3-(3S) (28 mg, 0.58 mmol) was
converted into the mixture of diesters 2-(3S,18S,19R) and
2-(3S,18R,19S) (49 mg, 90% yield). This mixture was separated
by semipreparative HPLC and each collected compound purified
as indicated above to give 20 mg of diester 2-(3S,18S,19R) and
14 mg of diester 2-(3S,18R,19S). Since these compounds are
enantiomers of the diesters described above, i.e., 2-(3R,18R,
19S) and (3R,18S,19R), respectively, only their ee and [R]D
values are given. 2-(3S,18S,19R): [R]D ) +11.9 (c ) 1.5, 94%
ee). 2-(3S,18R,19S): [R]D ) +15.3 (c ) 1, 90% ee).
NMR Sp ectr a . The NMR spectra (1H NMR, 300 MHz; 13C
NMR, 75 MHz; 19F NMR, 282 MHz) were recorded with a 4
pretuned nucleus auto NMR probe. All spectra were performed
in freshly neutralized CDCl3 solutions, and chemical shifts are
given in ppm downfield from Si(CH3)4 for 1H, CDCl3 for 13C, and
1
CFCl3 (internal reference) for 19F. The standard H DQFCOSY
spectra recorded for the determination of the absolute configu-
ration of MTPA diesters 2 were obtained as previously de-
scribed.22
Com p ou n d s. Unless otherwise stated, organic solutions
obtained from the workup of crude reaction mixtures were dried
(26) De Napoli, L.; Fattorusso, E.; Magno, S.; Mayol, L. Phytochem-
istry 1982, 21, 782-784.

Notes
P r ep a r a tion of Ch ir a l Dioxid osqu a len es 1. These com-
J . Org. Chem., Vol. 61, No. 21, 1996 7607
oxidosqualene were 40 or 10 µM as indicated for each case. The
mixture (final volume 1 mL) was incubated for 60 min at 37 °C.
The enzymatic reaction was quenched by treatment with 6%
KOH in methanol (1 mL) for 60 min at 37 °C. Then, 24,25-
dihydrolanosterol was added as internal standard, and the
mixture was extracted with hexane (3 × 2 mL). The combined
extracts were evaporated to dryness, redissolved in 10:1 hexane/
MTBE mixture (3 × 50 µL), loaded onto a column fitted with
silica gel (0.75 g, 35-70 µm) containing 5% Et3N, freshly
prepared, and eluted with the same solvent mixture. After
discarding the first 8 mL, lanosterol and internal standard were
collected in a 5 mL fraction. The eluates were evaporated to
dryness, redissolved in hexane, and injected onto a GC system
(for conditions cf. ref 21). Incubations were performed by
duplicate, and a minimum of two experiments per point were
carried out. The IC50 values were determined by interpolation
from the respective plot of percent inhibition vs log [I]. The
concentration ranges used were: 0.005-1 µM for 1-(3S,18S,
19S), 0.01-0.1 µM for 1-(3S,18R,19R), 0.005-5 µM for
1-(3R,18S,19S) and 0.005-5 µM for 1-(3R,18R,19R). The Ki
values for dioxidosqualenes 1-(3S,18S,19S) and 1-(3S,18R,19R)
were determined from replots of slopes and intercepts of Line-
weaver-Burk double-reciprocal plot vs the inhibitor concentra-
tion. In these determinations the substrate concentrations were
3, 5, 10, 20, 25, and 30 µM, whereas those of inhibitors were 0,
0.05, 0.1 and 0.2 µM for 1-(3S,18S,19S) and 0, 0.025, 0.05 and
0.1 µM for 1-(3S,18R,19R).
Irreversibility assays were carried out by adapting the
procedure reported by Bai and Prestwich25 to our case. Thus,
100 µL of pig liver OSLC solution were incubated with the
corresponding inhibitor, i.e., 1 µM 1-(3S,18S,19S) or 0.5 µM
1-(3S,18R,19R), in buffer A (0.1 M Tris-HCl, 1 mM EDTA, 1
mM DTT and 0.2% Triton X, final volume 1 mL), in absence of
organic solvents for 1 h at 37 °C. The inhibitor was removed by
adsorption of the OSLC to a DEAE-Sepharose column preequili-
brated with buffer A. Elution with buffer A containing 25 mM
KCl (5 mL) followed by 200 mM KCl in buffer A (5 mL) permitted
the collection of a 1.5 mL fraction containing the OSLC activity
pounds were obtained from the respective MTPA diesters 2 by
using the procedure described previously for the preparation of
chiral oxidosqualenes.22 Thus, treatment of diester 2 with
LiAlH4 in Et2O followed by purification of the crude reaction
mixture by chromatography eluting with 3:1 hexane:AcOEt
afforded the corresponding tetraol 3. A sample of tetraol 3 was
isolated in 75% yield from the acid hydrolysis (HClO4/H2O/THF)
of racemic dioxidosqualene 1. 3: MS m/ z (relative intensity)
478 (M, 3), 461 (M - 18 + 1,10), 443 (M + 1 - 36, base peak).
Anal. Calcd for C30H54O4: C, 75.26; H, 11.37. Found: C, 75.17;
H, 11.48.
(3R,18R,19S)-2,3,18,19-Tet r a h yd r oxy-2,3,18,19-t et r a h y-
d r osqu a len e. This tetraol was isolated (12 mg, 84% yield)
starting from 27 mg of diester 2-(3R,18R,19S). (3-(3R,18R,
19S): [R]D ) +17.9 (c ) 1, 95% ee); 1H NMR: δ 5.30-5.02 (4
H), 3.37 (dd, 1 H, J 1 ) 11 Hz, J 2 ) 1 Hz), 3.36 (dd, 1 H, J 1
)
10.5 Hz, J 2 ) 2 Hz), 2.34-1.94 (14 H), 1.69 (s, 3 H), 1.62 (br, 9
H), 1.60 (s, 3 H), 1.90-1.10 (6 H), 1.20 (s, 3 H), 1.17 (s, 3 H),
1.15 (s, 3 H); 13C NMR: δ 135.1, 135.1, 134.8, 131.9, 125.1, 125.1,
124.6, 124.4, 78.3, 78.3, 74.7, 73.0, 39.6, 37.0, 36.8, 35.8, 29.6,
29.2, 28.2, 28.1, 26.4, 26.3, 25.7, 23.4, 23.3, 22.1, 17.7, 16.0, 15.9.
(3R,18S,19R)-2,3,18,19-Tet r a h yd r oxy-2,3,18,19-t et r a h y-
d r osqu a len e. This tetraol was isolated (12 mg, 81% yield)
starting from 29 mg of diester 2-(3R,18S,19R). 3-(3R,18S,
19R): [R]D ) +2.0 (c ) 1, 95% ee).
(3S,18R,19S)-2,3,18,19-Tet r a h yd r oxy-2,3,18,19-t et r a h y-
d r osqu a len e. This tetraol was isolated (8 mg, 80% yield)
starting from 19 mg of diester 2-(3S,18R,19S). 3-(3S,18R,
19S): [R]D ) +17.9 (c ) 1, 95% ee); 1H NMR: δ 5.30-5.02 (4
H), 3.37 (dd, 1 H, J 1 ) 11 Hz, J 2 ) 1 Hz), 3.36 (dd, 1 H, J 1
)
10.5 Hz, J 2 ) 2 Hz), 2.34-1.94 (14 H), 1.69 (d, 3 H, J ) 1 Hz),
1.62 (br, 9 H), 1.60 (d, 3 H, J ) 1 Hz), 1.90-1.10 (6 H), 1.20 (s,
3 H), 1.17 (s, 3 H), 1.15 (s, 3 H); 13C NMR: δ 135.1, 135.1, 134.8,
131.9, 125.1, 125.1, 124.6, 124.4 , 78.3, 78.2, 74.7, 73.0, 39.6,
37.0, 36,8, 35.8, 29.6, 29.2, 28.2, 28.1, 26.4, 26.3, 25.7, 23.4, 23.3,
22.1, 17.7, 16.0, 15.9, 15.9.
(3S,18S,19R)-2,3,18,19-Tet r a h yd r oxy-2,3,18,19-t et r a h y-
d r osqu a len e. This tetraol was isolated (6 mg, 83% yield)
starting from 14 mg of diester 2-(3S,18S,19R). 3-(3S,18S,
19R): [R]D ) -17.8 (c ) 1, 94% ee).
Conversion of each tetraol into the respective dioxidosqualene
stereoisomer was carried out following the general procedure
described previously.22 For the spectroscopic and analytical
characterization of 1 see ref 21.
after discarding 6.5 mL of eluate. The effect of buffer
A
containing 200 mM KCl on OSLC activity was previously
checked. All this chromatography process was carried out at 4
°C. The fraction containing the residual OSLC activity was then
incubated in the presence of (R,S)-2,3-oxidosqualene (40 µM) and
0.4% (v/v) isopropyl alcohol for 1 h at 37 °C. The quenching of
the enzymatic reaction, extraction, purification, and quantifica-
tion of formed lanosterol was carried out as described above. In
the control assays performed with no separation of enzyme and
inhibitor, the crude preparation obtained after the preincubation
was incubated in the presence of (R,S)-2,3-oxidosqualene (40 µM)
and 0.4% (v/v) isopropyl alcohol (final volume 1.5 mL buffer A)
for 1 h at 37 °C.
(3S,18S,19S)-2,3:18,19-Dioxid osqu a len e. This compound
was isolated (8 mg, 74% yield) starting from 12 mg of tetraol
3-(3R,18R,19S). 1-(3S,18S,19S): [R]D ) -3.9 (c ) 0.5, 95% ee).
(3S,18R,19R)-2,3:18,19-Dioxid osqu a len e. This compound
was isolated (8 mg, 78% yield) starting from 11 mg of tetraol
3-(3R,18S,19R). 1-(3S,18R,19R): [R]D ) 0.6 (c ) 0.5, 95% ee).
(3R,18S,19S)-2,3:18,19-Dioxid osqu a len e. This compound
was isolated (4 mg, 72% yield) starting from 6 mg of tetraol
3-(3S,18R,19S). 1-(3R,18S,19S): [R]D ) -0.4 (c ) 0.4, 90% ee).
(3R,18R,19R)-2,3:18,19-Dioxid osqu a len e. This compound
was isolated (5 mg, 74% yield) starting from 7.5 mg of tetraol
3-(3S,18S,19R). 1-(3R,18R,19R): [R]D ) 4.0 (c ) 0.5, 94% ee).
Assa y Meth od for OSLC. Partially purified OSLC was
obtained from pig liver following the procedures described by
Abe et al.3 The specific activity of the purified enzyme was 2059
nmol/h mg protein. The procedure used for the OSLC assays
was based on that reported by Duriatti et al.24 with minor
modifications.21 Briefly, isopropyl alcohol solutions of the
substrate and inhibitors were added to the test tubes (the alcohol
contents did not exceed 0.4% v/v of the overall test mixture),
followed by addition of 1 mL of 0.1 M Tris-HCl buffer (pH )
7.4) containing 30 µL of pig liver OSLC solution, 1 mM EDTA,
and Triton X-100 (final concentration 0.05% w/v). For determi-
nation of IC50 values final concentrations of substrate (R,S)-2,3-
Ack n ow led gm en t. Financial support from CICYT-
Generalitat de Catalunya (Grant QFN 92-4304) and a
fellowship from the Ministerio de Educacio´n y Ciencia
to one of us (J .L.A.) are acknowledged. We thank Alex
Aymerich for technical assistance.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures, including synthesis and characterization data of all
new compounds, 1H and 13C NMR chemical shifts assignment
for diesters 2 and OSLC inhibition plots (9 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O9607020