An in vitro assay for evaluation of small-molecule inhibitors of cholesterol absorption

Article abstract of DOI:10.1002/anie.200460348

Giving cholesterol the brush off: An intestinal brush border membrane vesicle assay has been devised for the convenient in vitro testing of small molecules for inhibition of cholesterol absorption. The assay was used to identify new nonhydrolyzable glycosides as potent cholesterol-absorption inhibitors and an oxazolidinone as an effective replacement of the β-lactam scaffold of ezetimibe (1).

Full text of DOI:10.1002/anie.200460348

Angewandte
Chemie
Medicinal Chemistry
inhibitors and an oxazolidinone as an effective replacement of
the b-lactam scaffold of ezetimibe (1).
An In Vitro Assay for Evaluation of Small-
The actual identity of the intestinal cholesterol transport
proteins is still actively debated,with evidence suggesting the
involvement of the scavenger receptors SR-BI[7e,f,8] and
CD36,[7f,9] the Niemann–Pick C1-like 1 protein,[10] and an
Annexin 2/Caveolin 1 complex.[11] Regardless of which pro-
teins are involved,prior work has demonstrated that choles-
terol absorption takes place in brush border membrane
vesicles made from human or animal small intestines.[7] We
thus set out to investigate whether brush border membrane
vesicles could be used in the development of an in vitro assay
for small-molecule cholesterol absorption inhibitors. For the
successful implementation of such an assay,we needed to
validate it by synthesizing and examining a number of known
ezetimibe analogues for which the corresponding in vivo data
were available.
Rabbit brush border membrane vesicles are prepared by
shearing the enterocyte brush border membrane region and
allowing its reassembly into vesicles with a diameter of
180 nm (Figure 1).[7a,12] Such a preparation can be carried out
conveniently in hours,and the vesicles produced can be stored
for months. In the experiments,cholesteryl ester is delivered
to the brush border membrane vesicles by incorporation into
egg phosphatidyl choline small unilamellar vesicles (25 nm
diameter) that act as donor particles.[7,13] Direct application of
the small-molecule candidates in 1% aqueous dimethylsulf-
oxide was complicated by their low solubility in water.
However,key to the development of the assay was the
discovery that these molecules,like the cholesteryl ester,
could also be incorporated into the phospholipid small
Molecule Inhibitors of Cholesterol Absorption**
Lisbet Kværnø, Tobias Ritter, Moritz Werder,
Helmut Hauser, and Erick M. Carreira*
Cardiovascular disease is the leading cause of death in the
Western industrialized world and claims more lives than the
next four causes combined.[1] A major risk factor for this
disease is elevated cholesterol levels.[2,3] The current predom-
inant therapy prescribes the use of statins,which lead to
inhibition of cholesterol biosynthesis in the liver.[3,4] However,
half of all patients undergoing current lipid-lowering treat-
ments fail to reach their cholesterol goals,[3,5] and thus a clear
need for the development of new cholesterol-lowering agents
remains. Ezetimibe (1,Scheme 1) is a recent exciting example
of a new class of drugs operating by a different mechanism
involving inhibition of intestinal cholesterol absorption.
Ezetimibe (1) was developed solely by using animal studies.[6]
The rapid identification and optimization of new inhibitors
would be greatly facilitated with an in vitro assay. In this
communication,we introduce the use of intestinal brush
border membrane vesicles[7] for the convenient in vitro testing
of small molecules for inhibition of cholesterol absorption.
We demonstrate the validity of the assay by comparison of the
observed in vitro efficacies with published in vivo data
(Scheme 1). Moreover,the assay allows us to identify new
nonhydrolyzable glycosides as potent cholesterol absorption
Scheme 1. Correlation between in vitro and in vivo activities with various
[*] L. Kværnø, T. Ritter, Prof. Dr. E. M. Carreira
Laboratorium für Organische Chemie
ETH Hönggerberg, HCI H335
8093 Zürich (Switzerland)
small molecules.
unilamellar vesicles.[14] At a fixed concentration of 9 mol%,
corresponding to a nominal concentration of 6 mm,the test
substrates were readily incorporated into the vesicles and
used in routine inhibition experiments. The cholesterol
transfer from donor particles to brush border membrane
vesicles was quantified by using radiolabeled [1a,2a(N)-
3H]cholesterol oleyl ether (37 Cimmolꢀ1).[15]
Fax: (+41)1-632-1328
E-mail: carreira@org.chem.ethz.ch
Dr. M. Werder, Prof. Emeritus Dr. H. Hauser
Institute of Biochemistry
ETH Hönggerberg
We selected ezetimibe (1),[16] b-lactam (ꢁ )-2,[17,18] and the
O-glycosides 3–5[19] as convenient benchmark substrates of
the in vitro assay,because in vivo results were available for
these compounds (Table 1). In the brush border membrane
vesicle assay with rabbit intestine,ezetimibe ( 1) was shown to
inhibit cholesteryl ester uptake by 16% when the convenient
9-mol% concentration in the vesicles was used. b-Lactam
8093 Zürich (Switzerland)
[**] This research was supported by a CTI grant (6813.2 BTS-LS) and
Lipideon AG. L.K. was supported by a fellowship from The Technical
University of Denmark. T.R. thanks the Fonds der Chemischen
Industrie for a KekulØ Fellowship.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2004, 43, 4653 –4656
DOI: 10.1002/anie.200460348
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4653

Communications
Scheme 2. Nonhydrolyzable glycosides evaluated in the brush border
membrane vesicle assay.
by Mitsunobu reactions of 10 with the appropriate primary
alcohols,as shown for cellobioside 9 in Scheme 3.
When 7–9 were evaluated in the brush border membrane
vesicle assay,the glucosides 7 and 8 displayed inhibitions of
13% and 15%,respectively (Scheme 2). The cellobioside 9
(28% inhibition) was clearly more active and was as active an
inhibitor as the corresponding O-cellobioside 5 (27% inhi-
bition). The activities of the nonhydrolyzable glycosides 7–9
indicate that glycosides of ezetimibe (1) are indeed capable of
being potent cholesterol absorption inhibitors.
Figure 1. Preparation of brush border membrane vesicles.
(ꢁ )-2 was inactive,a result consistent with the corresponding
reported low activity in vivo.[18] Furthermore,the same
correlation was observed for the O-glycosides as that reported
for in vivo studies,[19] wherein cellobioside 5 was significantly
more active than glucuronide 3 and
glucoside 4. In the development of
ezetimibe (1) by using animal stud-
ies,the benzylic hydroxy group was
shown to be of prime importance
for activity.[6c] Likewise in our
assay,removal of this hydroxy
functionality in 5 to give 6 resulted
in a drastic decrease in the in vitro
inhibitory activity (3% inhibition
for 6 versus 27% for 5). This
molecular editing experiment[21]
provides yet another correlation
with the in vivo data.
Table 1: Activities in the brush border membrane vesicle assay compared to in vivo data.
Inhibitor
In vivo
In vitro
inhibition
inhibition [%][b,c]
[mgkgꢀ1 dayꢀ1 [a]
]
(ꢁ)-2:
low activity[d][18]
2
1: R=H
0.04[6]
16
With a validated in vitro assay
in hand,we started to address two
key issues: 1) the role of glycosy-
lation in inhibitory activity and
2) the use of other ring scaffolds
to replace the b-lactam. In vivo,
ezetimibe (1) has been shown to be
rapidly glucuronidated in the intes-
tine to form glucuronide 3 before
entering the portal plasma.[22]
Rapid metabolic hydrolysis of the
O-glycosides 3–5 and subsequent
interconversion between ezetimibe
(1) and the glucuronide 3 cannot be
completely ruled out in the
reported animal experiments. Con-
sequently,we designed,synthe-
sized,and examined the structur-
ally related glycosides 7–9 with
3: R=
4: R=
5: R=
0.09[19]
0.08[19]
19
14
0.01[19]
27
3
6:
–
[a] Reported effective dose values for 50% reduction of liver cholesteryl ester level (ED50) in the seven-
day cholesterol-fed hamster model.[20] [b] Inhibition in the brush border membrane vesicle assay with
rabbit intestine at nominal concentrations of 6 mm. [c] Average standard deviations were ꢁ3%
inhibition. [d] At 50 mgkgꢀ1 dayꢀ1, no ED50 value was determined.
nonhydrolyzable
(Scheme 2). These were prepared
linkages[23]
4654
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 4653 –4656

Angewandte
Chemie
the appropriate side chains could effectively replace ezeti-
mibe (1).
In summary,we have used brush border membrane
vesicles in the first in vitro assay to evaluate small-molecule
cholesterol absorption inhibitors. In comparison with pub-
lished in vivo data of identical and closely related compounds,
consistent in vitro results were obtained. These indicate that
both ezetimibe (1) and a variety of ezetimibe glycosides are
all potent cholesterol absorption inhibitors. The salient
feature of the assay is that it permits rapid identification of
other useful ring scaffolds for the synthesis of novel choles-
terol absorption inhibitors,as exemplified by the oxazolidi-
none 16. It should thus be possible to replace tedious,
expensive,and lengthy animal experiments by convenient
evaluation in the brush border membrane vesicle assay when
investigating diverse libraries of small-molecule cholesterol
absorption inhibitors.
Scheme 3. a) 1,[16] Ac2O, NaOH, iPrOH, 97%; b) TBDMSCl, imid,
DMF, 91%; c) Al2O3 (neutral), 708C, 83%; d) 11,[24] 1,1’-(azodicarbon-
yl)dipiperidine, Bu3P, THF; e) H2, Pd(OH)2/C, EtOAc/EtOH, 21% over
two steps; f) HF·pyridine, pyridine, THF, 62%. Bn=benzyl,
DMF=N,N-dimethylformamide, imid=imidazole, TBDMS=tert-butyl-
dimethylsilyl, THF=tetrahydrofuran.
Received: April 16,2004
Keywords: cholesterol · inhibitors · medicinal chemistry ·
.
membranes
The in vitro assay also allowed us to conveniently examine
whether the b-lactam is an integral and essential pharmaco-
phore,as concluded in the early development of ezetimibe
(1),[18,25] or simply a ring scaffold to appropriately position the
required substituents. Thus,oxazolidinone 16,which closely
resembles ezetimibe (1) with respect to the positioning of the
ring substituents,[25] was synthesized in enantiomerically pure
form as outlined in Scheme 4. We were pleased to observe
that oxazolidinone 16 showed a similar activity (19%
inhibition) to ezetimibe (1; 16% inhibition) in the brush
border membrane vesicle assay. The notable activity for 16
suggests that an oxazolidinone ring scaffold substituted with
[1] Heart Disease and Stroke. Statistics—2003 Update. Available
from The American Heart Association at: http://www.ameri-
canheart.org/downloadable/heart/
10590179711482003HDSStatsBookREV7-03.pdf.
[2] T. Sudhop,K. von Bergmann, Drugs 2002, 62,2333.
[3] E. Bruckert, Cardiology 2002, 97,59.
[4] D. J. Maron,S. Fazio,M. F. Linton, Circulation 2000, 101,207.
[5] Euroaspire II Study Group, Eur. Heart J. 2001, 22,554.
[6] a) FDA Drug Approvals List [online]: http://www.fda.gov/cder/
foi/label/2002/21445lbl.pdf; b) J. W. Clader, J. Med. Chem. 2004,
47,1; c) S. B. Rosenblum,T. Huynh,A. Afonso,H. R. Davis,N.
Yumibe,J. W. Clader,D. A. Burnett, J. Med. Chem. 1998, 41,
973.
[7] a) H. Thurnhofer,H. Hauser, Biochemistry 1990, 29,2142; b) S.
Compassi,M. Werder,D. Boffelli,F. E. Weber,H. Hauser,G.
Schulthess, Biochemistry 1995, 34,16473; c) S. Compassi,M.
Werder,F. E. Weber,D. Boffeli,H. Hauser, Biochemistry 1997,
36,6643; d) G. Schulthess,S. Compassi,D. Boffelli,M. Werder,
F. E. Weber,H. Hauser, J. Lipid Res. 1996, 37,2405; e) H.
Hauser,J. H. Dyer,A. Nandy,M. A. Vega,M. Werder,E.
Bieliauskaite,F. E. Weber,S. Compassi,A. Gemperli,D.
Boffelli,E. Wehrli,G. Schulthess,M. C. Phillips,
Biochemistry
1998, 37,17843; f) M. Werder,C. H. Han,E. Wehrli,D.
Bimmler,G. Schulthess,H. Hauser,
11643.
Biochemistry 2001, 40,
[8] a) S. F. Cai,R. J. Kirby,P. N. Howles,D. Y. Hui,
J. Lipid Res.
2001, 42,902; b) L. Cai,E. R. M. Eckhardt,W. Shi,Z. Zhao,M.
Nasser,W. J. S. de Villiers,D. R. van der Westhuyzen, J. Lipid
Res. 2004, 45,253; c) S. W. Altmann,H. R. Davis,X. R. Yao,M.
Laverty,D. S. Compton,L. J. Zhu,J. H. Crona,M. A. Caplen,
L. M. Hoos,G. Tetzloff,T. Priestley,D. A. Burnett,C. D. Strader,
M. P. Graziano, Biochim. Biophys. Acta 2002, 1580,77; d) E.
Levy,D. Menard,I. Suc,E. Delvin,V. Marcil,L. Brissette,L.
Thibault,M. Bendayan, J. Cell Sci. 2004, 117,327; e) D.
Jourdheuil-Rahmani,M. Charbonnier,N. Domingo,F. Luccioni,
H. Lafont,D. Lairon, Biochem. Biophys. Res. Commun. 2002,
292,390; f) B. Play,S. Salvini,Z. Haikal,M. Charbonnier,A.
Scheme 4. a) 12,[26] 13,[27] (Cl3CO)2CO, Et3N, 08C!RT, 75%, d.r. =3:2;
b) THF/H2O, reflux, 50%; c) NaOMe, MeOH; d) (Cl3CO)2CO, iPr2NEt,
DMAP, CH2Cl2, ꢀ788C!238C, 73% over two steps; e) NaBH4, EtOH;
=
f) (COCl)2, DMSO, Et3N, CH2Cl2, ꢀ788C; g) pFC6H4COCH PPh3,
CH2Cl2, ꢀ788C!238C, 82% over three steps; h) H2, Pd/C, EtOH;
i) (R)-CBS,[28] BH3·Me2S, CH2Cl2, ꢀ20!08C; j) H2, Pd/C, EtOH, 57%
over three steps. (R)-CBS=R-tetrahydro-1-methyl-3,3-diphenyl-1H,3H-
pyrrolo(1,2-C)(1,3,2)oxazaborolidine, DMAP=4-dimethyl-
aminopyridine, DMSO=dimethylsulfoxide.
Harbis,M. Roussel,D. Lairon,D. Jourdheuil-Rahmani,
Bio-
chem. Biophys. Res. Commun. 2003, 310,446.
Angew. Chem. Int. Ed. 2004, 43, 4653 –4656
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4655

Communications
[9] A. van Bennekum,M. Werder,S. T. Thuahnai,C.-H. Han,D. L.
Williams,P. Wettstein,G. Schulthess,M. C. Phillips,H. Hauser,
Biochemistry 2004,submitted.
[10] S. W. Altmann,H. R. Davis,L. J. Zhu,X. R. Yao,L. M. Hoos,G.
Tetzloff,S. P. N. Iyer,M. Maguire,A. Golovko,M. Zeng,L. Q.
Wang,N. Murgolo,M. P. Graziano, Science 2004, 303,1201.
[11] E. J. Smart,R. A. De Rose,S. A. Farber, Proc. Natl. Acad. Sci.
USA 2004, 101,3450.
[12] H. Hauser,K. Howell,R. M. C. Dawson,D. E. Bowyer,
Bio-
chim. Biophys. Acta 1980, 602,567.
[13] Cholesteryl ester,instead of free cholesterol,was chosen for the
standard inhibition experiments,since passive diffusion is
negligible for cholesteryl ester.[7e]
[14] The incorporation into vesicles was shown to be quantitative for
representative inhibitors by gel filtration of the vesicles accord-
ing to established procedures (see G. Schulthess,S. Compassi,M.
Werder,C. H. Han,M. C. Phillips,H. Hauser,
Biochemistry
2000, 39,12623,and references therein) followed by reversed-
phase HPLC analysis.
[15] Radiolabeled cholesterol oleyl ether was used as a nonhydro-
lyzable analogue of cholesterol oleyl ester,since both have been
shown to exhibit identical brush border membrane uptake
kinetics.[7b]
[16] G. Z. Wu,Y. S. Wong,X. Chen,Z. X. Ding, J. Org. Chem. 1999,
64,3714.
[17] M. Browne,D. A. Burnett,M. A. Caplen,0L. Y. Chen,J. W.
Clader,M. Domalski,S. Dugar,P. Pushpavanam,R. Sher,W.
Vaccaro,M. Viziano,H. R. Zhao, Tetrahedron Lett. 1995, 36,
2555.
[18] J. W. Clader,D. A. Burnett,M. A. Caplen,M. S. Domalski,S.
Dugar,W. Vaccaro,R. Sher,M. E. Browne,H. R. Zhao,R. E.
Burrier,B. Salisbury,H. R. Davis, J. Med. Chem. 1996, 39,3684.
[19] W. D. Vaccaro,H. R. Davis, Bioorg. Med. Chem. Lett. 1998, 8,
313.
[20] a) D. A. Burnett,M. A. Caplen,H. R. Davis,R. E. Burrier,J. W.
Clader, J. Med. Chem. 1994, 37,1733; b) R. Schnitzerpolokoff,D.
Compton,G. Boykow,H. Davis,R. Burrier,
Comp. Biochem.
Physiol. Part A 1991, 99,665.
[21] a) A. M. Rouhi, Chem. Eng. News 2003, 81,104; for recent
examples of successful molecular editing of epothilones,see:
b) T. C. Chou,H. J. Dong,A. Rivkin,F. Yoshimura,A. E.
Gabarda,Y. S. Cho,W. P. Tong,S. J. Danishefsky, Angew. Chem.
2003, 115,4910; Angew. Chem. Int. Ed. 2003, 42,4761; c) K. C.
Nicolaou,P. K. Sasmal,G. Rassias,M. V. Reddy,K. H. Altmann,
M. Wartmann,A. OꢀBrate,P. Giannakakou, Angew. Chem. 2003,
115,3639; Angew. Chem. Int. Ed. 2003, 42,3515.
[22] M. van Heek,C. Farley,D. S. Compton,L. Hoos,K. B. Alton,
E. J. Sybertz,H. R. Davis, Br. J. Pharmacol. 2000, 129,1748.
[23] W. D. Vaccaro,R. Sher,H. R. Davis, Bioorg. Med. Chem. Lett.
1998, 8,35.
[24] S. J. Spak,O. R. Martin, Tetrahedron 2000, 56,217.
[25] S. Dugar,M. P. Kirkup,J. W. Clader,S. I. Lin,R. Rizvi,M. E.
Snow,H. R. Davis,S. W. McCombie, Bioorg. Med. Chem. Lett.
1995, 5,2947.
[26] B. A. Shinkre,V. G. Puranik,B. M. Bhawal,A. Deshmukh,
Tetrahedron: Asymmetry 2003, 14,453.
[27] T. Kambara,K. Tomioka, J. Org. Chem. 1999, 64,9282.
[28] a) E. J. Corey,R. K. Bakshi,S. Shibata, J. Am. Chem. Soc. 1987,
109,5551; b) E. J. Corey,R. K. Bakshi,S. Shibata,C. P. Chen,
V. K. Singh, J. Am. Chem. Soc. 1987, 109,7925.
4656
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 4653 –4656