Novel pyrrole- and 1,2,3-triazole-based 2,3-oxidosqualene cyclase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.07.131

Pyrrole- and 1,2,3-triazole-based 2,3-oxidosqualene cyclase (OSC) inhibitors 3 and 4 were discovered by conducting a virtual screening, a docking study based on the crystallographic structure of OSC, and biological assays. The hit rate of the assays was increased by establishing appropriate substructural filters in the virtual screening stage. Amide derivatives of 8 and 12 preserved the inhibitory activity of parent compound 3, which provided a reasonable starting point for further structure-activity-relationship (SAR) studies on related compounds.

Full text of DOI:10.1016/j.bmcl.2010.07.131

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5807–5810
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel pyrrole- and 1,2,3-triazole-based 2,3-oxidosqualene cyclase inhibitors
*
Takumi Watanabe , Yoji Umezawa, Yoshikazu Takahashi, Yuzuru Akamatsu
Institute of Microbial Chemistry, Tokyo, 3-14-23 Kamiosaki, Shinagawa-ku, Tokyo 141-0021, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Pyrrole- and 1,2,3-triazole-based 2,3-oxidosqualene cyclase (OSC) inhibitors 3 and 4 were discovered by
conducting a virtual screening, a docking study based on the crystallographic structure of OSC, and bio-
logical assays. The hit rate of the assays was increased by establishing appropriate substructural filters in
the virtual screening stage. Amide derivatives of 8 and 12 preserved the inhibitory activity of parent com-
pound 3, which provided a reasonable starting point for further structure–activity-relationship (SAR)
studies on related compounds.
Received 20 July 2010
Revised 28 July 2010
Accepted 29 July 2010
Available online 3 August 2010
Keywords:
2,3-Oxidosqualene cyclase inhibitor
Virtual screening
Ó 2010 Elsevier Ltd. All rights reserved.
Hypercholesterolemia
Structure–activity-relationship
Elevated plasma cholesterol levels are a major risk factor for
coronary artery disease; however, cholesterol biosynthesis inhibi-
tors can be used for the prevention of heart disease.1 To this end,
HMG-CoA reductase inhibitors such as statins are used clinically
for the treatment of hypercholesterolemia.2 However, high doses
of statins cause severe side effects, including myopathy, because
the enzyme catalyzes the rate limiting step of cholesterol synthe-
sis, depleting isoprenoids such as coenzyme Q.3 2,3-Oxidosqualene
cyclase (OSC, lanosterol synthase)4 which catalyzes the conversion
of 2,3-oxidosqualene to lanosterol is an alternative target for the
inhibition of cholesterol biosynthesis. Because the conversion of
2,3-oxidosqualene–lanosterol is downstream from HMG-CoA in
the biosynthesis of cholesterol, OSC inhibitors are not expected
to produce statin-derived adverse effects.
Various substances have shown inhibitory activity against
OSC,5–20 and representative compounds are listed in Figure 1.
Among active compounds, Ro 48-8071 (1)16 is one of the most po-
tent inhibitors. However, OSC inhibitors have not been approved as
clinical drugs for the treatment of hypercholesterolemia because of
toxicological issues such as skin and epididymal changes and cat-
aract formation. Alternatively, OSC inhibitors with unique struc-
tural scaffolds are expected to show cholesterol-lowering activity
without the toxicity.
To identify OSC inhibitors, docking study was performed in
which conformers in 3D small molecule database ZINC21 were
superimposed onto the crystal structure of 1 bound to human
OSC22 and optimized energetically. Thus, the results of the compu-
tational experiment were highly dependent on the quality of the
virtual library.
At the time when this screening started, ZINC contains the ener-
getically optimal structures of 3,300,000 compounds. Because
performing a docking study for every compound would be time-
consuming, the database was subdivided into 46 subsets based
on the supplier of the compounds, and five subsets (222,000 chem-
icals altogether) from the 46 subsets were selected for virtual
screening.
The selected five subsets were filtered according to several
parameters (À2 < x log P < 4, 150< molecular weight <350, hydro-
gen bond donor 63, hydrogen bond acceptor 66) to obtain a subset
of 110,507 substances. A docking study was performed on the se-
lected subsets using the CDOCKER application (Discovery Studio
2.1, Accelrys), where the virtual hits were selected based on two
Herein, we describe the discovery of pyrrole and 1,2,3-triazole-
based OSC inhibitors. By considering the crystal structure of 1
bound to OSC, a virtual screening based on the 3D structure of
commercially available small molecules was performed. Among
the resultant candidates, eight substances were found to display
OSC inhibitory activity.
* Corresponding author. Tel.: +81 3 3441 4173; fax: +81 3 3441 7589.
E-mail address: twatanabe@bikaken.or.jp (T. Watanabe).
Figure 1. Structure of OSC inhibitors, Ro 48-8071 (1) and LDAO (2).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.131

5808
T. Watanabe et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5807–5810
benzene rings of 1 interact with several of the aromatic amino acid
residues of the enzyme: Trp 192, Trp 230, His 232, Phe 521, and
Phe 696. The bromine atom also interacts with Ile 524, Tyr 237,
Figure 2. Interactions between 1 and human OSC.23
scoring functions. Specifically, hits were chosen for biological as-
says if they displayed a LigScore2 >7 and a Consensus >2 calculated
with Score Ligand Poses and Consensus Score applications (Discov-
ery Studio 2.1, Accelrys). The virtual assay yielded 157 hits and 40
compounds were purchased for the examination of OSC inhibitory
activity according to Ebizuka’s procedure.17 However, the results
indicated that the samples obtained from the virtual assay did
not inhibit OSC in vitro. Thus, by selecting substructures that are
expected to bind to the active site of OSC, we constructed a more
focused subset to increase the probability of hits.
Figure 3. OSC inhibitors identified during the virtual screening.
Table 1
LigScore 2 and OSC inhibitory activity (IC50
:
lM) of virtual hits and amide analogs
Compounds
LigScore 2
OSC inhibitory activity
3
4
8
12
7.58
6.99
7.36
7.38
nd
12.7
1.41
35.4
58.8
2.15
A co-crystallographic structure of 1 and human OSC (Fig. 2)
clearly demonstrates that the carboxyl group of Asp 455 forms a
charged hydrogen bond with the allylmethylamino group, and
the amide NH of Ile 338 interacts with the carbonyl oxygen of 1
by a hydrogen bond via adventitious water molecule. Two of the
LDAO (2)
Scheme 1. Virtual screening of OSC inhibitors.

T. Watanabe et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5807–5810
5809
Scheme 2. Reagents and conditions: (a) 4-methoxybenzylaniline, toluene, rt, 72 h, 98%; (b) KOH, aqueous EtOH, 70 °C, 7 h, 65%; (c) 4-dimethylaminomethylbenzylamine,
WSCÁHCl, HOBt, TEA, DMF, rt, 16 h, 26%. (d) DIBAL, THF, À78 °C to 0 °C, 3 h, 16%; (e) HN3, DEAD, PPh3, THF, rt, 2 h, 61%; (f) H2, 10% Pd/C, AcOEt, rt, 16 h, 85%; (g) 4-
dimethylaminomethylbenzoic acid, WSCÁHCl, HOBt, TEA, DMF, rt, 16 h, 26%.
and Cys 233. Based on the crystallographic information, a struc-
tural filter was established to construct a new subset for virtual
screening, as shown in Scheme 1. Each substructure possessed
the basic characteristic which was expected to make a charged
hydrogen bond with Asp 455 of OSC.
In the second virtual screening, 3D structures of the compounds
that were not included in ZINC were generated using Prepare Li-
gands application (Discovery Studio 2.1, Accelrys), and used in
the docking study. The molecular property filter was utilized as de-
scribed previously. In the second virtual screening, the threshold
for hits was changed to a LigScore2 >6 and a Consensus >1 because
only a few virtual hits were obtained when more stringent thresh-
olds were applied. The second virtual screening yielded 180 hits, of
which the 40 most readily available substances were purchased
and tested for OSC inhibitory activity.
above, we conducted a preliminary structure–activity-relationship
(SAR) study on 3.
Compound 3 contains a pyrrole and tertiary amino moieties
connected by a dimethyleneamino spacer. If the linker can be sur-
rogated by an amide, a variety of analogs can be synthesized easily
by coupling corresponding primary amines to carboxylic acids. As
shown in Scheme 2, the synthesis of the amide derivatives of 3 was
achieved by coupling pyrrolecarboxylic acid (6) or aminomethyl-
pyrrole (11) with 4-dimethylaminomethylbenzylamine or 4-dime-
thylaminomethylbenzoic acid to afford analogs
8
and 12.25
Subsequently, a docking study on the crystal structure of OSC
was performed with analogs 8 and 12, and LigScore2 of 7.36 and
7.38, respectively, was obtained. Indeed, compounds 8 and 12 dis-
played an IC50 for OSC that was comparable to that of parent com-
pound 3 (8: 35.4 lM, 12: 58.8 lM), and further SAR studies on
To evaluate the inhibitory activity of the compounds, LDAO (2)
was selected as a positive control.23 LDAO is an inhibitor that
displays an IC50 value for OSC inhibition that is approximately
130-times higher than that of 1. Considering experimental error,
compounds with a potency greater than 60% of the inhibitory
activity of 2 were defined to have significant biological activity.
OSC inhibitors that were structurally related to 1 and displayed
an IC50 comparable to that of 2 were obtained during a SAR study
for the development 1; thus the aforementioned threshold is
reasonable for a primary screening assay.24 OSC reactions were
monitored with radio-labeled 2,3-oxidosqualene according to the
previously reported procedure.17 In the assays, the sample concen-
amide derivatives are expected to provide more potent OSC inhib-
itors. A synthetic study and further evaluation of the biological
activity of the remaining six hits of the in vitro assay will be re-
ported in due course.
Acknowledgments
The authors thank Professor Dr. Yutaka Ebizuka (currently at
the National Institute of Health Science) and Dr. Masaaki Shibuya
(currently at Nigata University of Pharmacy and Applied Life
Sciences) at the University of Tokyo for providing Saccharomyces
cervisiae expressing human OSC and [14C] (3S)-2,3-oxidosqualene.
tration was set to 20 lM and 84% of the OSC reactions were sup-
pressed by 2.
References and notes
The results indicated that eight compounds, including 3 and 4,
(Fig. 3) inhibited OSC in vitro with an efficacy that was comparable
to that of 2; the inhibitory activity of 3 and 4 were 86% and 128% of
1. Jain, K. S.; Kathiravan, M. K.; Somani, R. S.; Shishoo, C. J. Bioorg. Med. Chem.
2007, 15, 4674.
2. Istvan, E. S.; Deisenhofer, J. Science 2001, 292, 1160.
3. Thompson, P. D.; Clarkson, P.; Karas, R. H. JAMA 2003, 289, 1681.
4. Abe, I.; Rohmer, M.; Prestwich, G. D. Chem. Rev. 1993, 93, 2189.
5. Barth, M. M.; Binet, J. L.; Thomas, D. M.; deFornel, D. C.; Samreth, S.; Schuber, F.
J.; Renaut, P. P. J. Med. Chem. 1996, 39, 2302.
the activity of 2. According to a reported protocol,17 the IC50
(lM)
of 2, 3, and 4 were determined to be 2.15, 12.7, and 1.41, respec-
tively, as shown in Table 1. Encouraged by the results described

5810
T. Watanabe et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5807–5810
6. Brown, G. R.; Hollinshead, D. M.; Stokes, E. S. E.; Waterson, D.; Clarke, D. S.;
Foubister, A. J.; Glossop, S. C.; McTaggart, F.; Mirrlees, D. J.; Smith, G. J.; Wood,
R. J. Med. Chem. 2000, 43, 4964.
7. Ceruti, M.; Balliano, G.; Rocco, F.; Milla, P.; Arpicco, S.; Cattel, L.; Viola, F. Lipids
2001, 36, 629.
8. Eisele, B.; Budzinski, R.; Muller, P.; Maier, R.; Mark, M. J. Lipid Res. 1997, 38,
564.
9. Fouchet, M. H.; Donche, F.; Martin, C.; Bouillot, A.; Junot, C.; Boullay, A. B.;
Potvain, F.; Magny, S. D.; Coste, H.; Walker, M.; Issandou, M.; Dodic, N. Bioorg.
Med. Chem. 2008, 16, 6218.
19. Shibuya, M.; Snider, B. B.; Sakano, Y.; Tomoda, H.; Omura, S.; Ebizuka, Y. J.
Antibiot. 2005, 58, 599.
20. Tanaka, R.; Sakano, Y.; Nagatsu, A.; Shibuya, M.; Ebizuka, Y.; Goda, Y. Bioorg.
Med. Chem. Lett. 2005, 15, 159.
21. See http://zinc.docking.org.; Irwin, J. J.; Shoichet, B. K. J. Chem. Inf. Model. 2005,
45, 177.
22. Thoma, R.; Schulz-Gasch, T.; D’Arcy, B.; Benz, J.; Aebi, J.; Dehmlow, H.; Hennig,
M.; Stihle, M.; Ruf, A. Nature 2004, 432, 118.
23. Cattel, L.; Ceruti, M.; Viola, F.; Delprino, L.; Balliano, G.; Duriatti, A.;
Bouviernave, P. Lipids 1986, 21, 31.
10. Hoshino, T.; Kobayashi, N.; Ishibashi, E.; Hashimoto, S. Biosci., Biotechnol.,
Biochem. 1995, 59, 602.
24. Guerry, P.; Jolidon, S.; Zurflueh, R. EP 410359, 1991.
25. All the new compounds showed satisfactory analytical data. The representatives
11. Lenhart, A.; Reinert, D. J.; Aebi, J. D.; Dehmlow, H.; Morand, O. H.; Schulz, G. E. J.
Med. Chem. 2003, 46, 2083.
are listed below: Compound 8: a colorless amorphous; IR (KBr)
mmax 3336, 2939,
1631, 1512, 1250, 1169, 1034, 840 cmÀ1 1H NMR (CDCl3, 600 MHz); d 7.31 (2H,
;
12. Mark, M.; Muller, P.; Maier, R.; Eisele, B. J. Lipid Res. 1996, 37, 148.
13. Oliaro-Bosso, S.; Taramino, S.; Viola, F.; Tagliapietra, S.; Ermondi, G.; Cravotto,
G.; Balliano, G. J. Enzyme Inhib. Med. Chem. 2009, 24, 589.
14. Stach, D.; Zheng, Y. F.; Perez, A. L.; Oehlschlager, A. C.; Abe, I.; Prestwich, G. D.;
Hartman, P. G. J. Med. Chem. 1997, 40, 201.
d, J = 8.3 Hz), 7.27 (2H, d, J = 8.3 Hz), 7.09 (2H, m), 6.98 (2H, m), 6.04 (1H, s), 6.00
(1H, t, J = 5.5 Hz), 4.59 (1H, d, J = 5.5 Hz), 3.86 (3H, s), 3.42 (2H, s), 2.336 (3H, s),
2.24 (6H, s), 1.96 (3H, s); 13C NMR (CDCl3, 150 MHz); d 165.8, 159.4, 138.0, 137.7,
134.5, 130.4, 129.4, 129.2, 128.9, 127.7, 114.4, 113.6, 104.0, 64.0, 55.5, 45.3, 42.9,
+
12.6, 12.2; HRMS (ESI-Orbitrap) m/z calcd for
C
24H29N3NaO2 [(M+Na)+]
15. Viola, F.; Brusa, P.; Balliano, G.; Ceruti, M.; Boutaud, O.; Schuber, F.; Cattel, L.
Biochem. Pharmacol. 1995, 50, 787.
414.2152, found: 414.2141. Compound 12: a colorless amorphous; IR (KBr)
mmax 3309, 2939, 1639, 1516, 1292, 1250, 1169, 1034 cmÀ1 1H NMR (CDCl3,
;
16. Morand, O. H.; Aebi, J. D.; Dehmlow, H.; Ji, Y. H.; Gains, N.; Lengsfeld, H.;
Himber, J. J. Lipid Res. 1997, 38, 373.
17. Sakano, Y.; Shibuya, M.; Matsumoto, A.; Takahashi, Y.; Tomoda, H.; Omura, S.;
Ebizuka, Y. J. Antibiot. 2003, 56, 817.
18. Sakano, Y.; Shibuya, M.; Yamaguchi, Y.; Masuma, R.; Tomoda, H.; Omura, S.;
Ebizuka, Y. J. Antibiot. 2004, 57, 564.
600 MHz); d 7.74 (2H, d, J = 8.2 Hz), 7.36 (2H, d, J = 8.2 Hz), 7.12 (2H, m), 6.97 (2H,
m), 6.14 (1H, br), 5.94 (1H, s), 4.46 (1H, d, J = 4.8 Hz), 3.86 (3H, s), 3.46 (2H, s), 2.25
(6H, s), 2.00 (3H, s), 2.00 (3H, s); 13C NMR (CDCl3, 150 MHz); d 167.0, 159.0, 142.2,
133.6, 131.4, 129.2, 129.1, 128.7, 126.93, 126.90, 114.7, 114.3, 106.2, 63.9, 55.5,
+
45.3, 36.8, 12.7, 10.6; HRMS (ESI-Orbitrap) m/z calcd for C24H29N3NaO2
[(M+Na)+] 414.2152, found: 414.2141.