Identification and optimization of tetrahydro-2H-3-benzazepin-2-ones as squalene synthase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.02.004

Novel squalene synthase inhibitors are disclosed. SAR and pharmacological profile of selected compounds are discussed.

Full text of DOI:10.1016/j.bmcl.2011.02.004

Bioorganic & Medicinal Chemistry Letters 21 (2011) 2554–2558
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and optimization of tetrahydro-2H-3-benzazepin-2-ones
as squalene synthase inhibitors
Nils Griebenow a, , Timo Flessner a, Anja Buchmueller b, Martin Raabe, Hilmar Bischoff b, Peter Kolkhof b
⇑
a Bayer Pharma AG, Global Drug Discovery, Medicinal Chemistry Wuppertal, Building 460, D-42096 Wuppertal, Germany
b Bayer Pharma AG, Global Drug Discovery, Cardiology Research, Building 500, D-42096 Wuppertal, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Novel squalene synthase inhibitors are disclosed. SAR and pharmacological profile of selected compounds
Received 3 December 2010
Revised 1 February 2011
Accepted 1 February 2011
Available online 17 February 2011
are discussed.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Squalene synthase
Inhibitor
SAR
Atherosclerosis
3-Benzazepin-2-ones
Atherosclerosis is by far the most important pathological
process in the development of ischemic cardiopathy and cerebral
infarction, which are the most common causes of morbidity and
mortality in industrialized countries.
Cl
N
O
Cl
OH
IC50 = 3.3 µM
Atherosclerosis is characterized by the accumulation of choles-
terol, mainly LDL-cholesterol in macrophages, leading to lesion for-
mation in large and medium-size arteries.1 These lesions may
develop into atheromatic plaques, which are prone to rupture
and cause clinical events such as heart attack and stroke. Currently,
treatment of atherosclerosis aims at reducing blood cholesterol
and triglyceride content. A temporary interruption of cholesterol
biosynthesis can lead to decreased plasma LDL-cholesterol
levels by removing LDL-cholesterol from the circulation through
up-regulation of hepatic LDL-receptors. Inhibitors of cholesterol
biosynthesis via inhibition of HMGCo-A (3-hydroxy-3-methylglut-
aryl-CoA) reductase are presently the most effective means for
blood LDL-cholesterol reduction.2 However, inhibition of HMGCo-
A reductase may block the formation of biologically necessary
isoprenoids. Squalene synthase inhibition seems to offer a prefera-
ble alternative because squalene synthase catalyzed reductive
dimerization of farnesyl pyrophosphate to squalene is the first spe-
cific step in cholesterol biosynthesis; thus, isoprenoid formation is
not affected. Furthermore, squalene synthase inhibitors have been
found to lower triglyceride levels in addition to the decrease of
plasma cholesterol.1c
O
rac-1
Figure 1. Initial lead structure 1. IC50 was determined for the racemic compound.
Relative stereochemistry was assigned based on the X-ray crystal structure analysis
of compound 19.
Therefore, squalene synthase inhibitors are a promising drug
class for the treatment of hyperlipidemia and atherosclerosis.
Various squalene synthase inhibitors have been evaluated over
the years as cholesterol lowering agents, with lapaquistat (TAK
475) advancing furthest in development. To date, the only squalene
synthase inhibitor to have made it into phase III clinical trials is
lapaquistat.3 Our interest was aroused by the identification of 1
as a moderately potent squalene synthase inhibitor (Fig. 1). The
IC50 values were determined by a biochemical assay measuring
the conversion of farnesyl pyrophosphate to squalene by squalene
synthase.4
The tetrahydro-2H-3-benzazepin-2-one scaffold was discov-
ered during a structure based design approach derived from known
inhibitors such as 4,1-benzoxazepine-3-acetic acid derivatives.5
⇑
Corresponding author. Tel.: +49 202 36 5866.
E-mail address: nils.griebenow@bayer.com (N. Griebenow).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.02.004

N. Griebenow et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2554–2558
2555
Cl
Cl
b
Cl
O
Cl
Cl
a
O
2
3
I
O
c
Cl
Cl
Br
O
O
Cl
Cl
O
d
NH
O
N
O
5
4
Cl
Cl
O
O
5
e
Cl
Cl
O
OH
4
N
N
3
2
1
O
O
rac-6
rac-1
Scheme 1. Synthetic route A. Reagents and conditions: (a) AlCl3, CH2Cl2, 0 °C, 1.5 h (31%); (b) TMSN3, H2SO4, CH2Cl2, rt, 1 h (29% pure isomer 4); (c) Cs2CO3, DMF, rt, 18 h
(51%); (d) tert-butyl-P4, ꢀ78 °C, 2 h, (53%, ds >99:1); (e) TFA/CH2Cl2 (v/v = 1:2), rt, 1.5 h (99%). For details see Ref. 6.
Compound 1 was built up in five linear steps (Scheme 1).6 Initially,
Friedel–Crafts reaction of (4-chlorophenyl)acetyl chloride and
1-chloro-2-vinylbenzene mediated by aluminum trichloride gave
the tetralone 3. Schmidt rearrangement7 of tetralone 3 yielded
the regioisomeric 1-benzazepin-2-one and 3-benzazepin-2-one 4
2-one core via the intermediate 1-phenyl-1,4-dihydro-3H-isochro-
men-3-ones by a ring-opening and recyclization reaction sequence
as first described by Busacca et al. in 1992 (Scheme 2).11
In detail, isochromen-3-one 7 was first reacted with ethanolic
HCl to give 8 in high yield. Introduction of the cyano group under
Lewis acid catalysis proceeded smoothly, providing cyanoester 9.
This species was then reduced to the crude amino ester with
NaBH4/CF3CO2H and thermally cyclized to target lactam 10 in
modest yield.
Our target intermediate isochromen-3-one 15 was built up in
seven linear steps as outlined in Scheme 3. The synthesis started
with a one pot reductive alkylation and decarboxylation to deliver
3-arylpropanoic acid 11.12 In detail, 4-chlorobenzaldehyde was re-
acted with Meldrum’s acid in the presence triethylammonium for-
mate at elevated temperature to yield 11, which was thermally
cyclized to indanone 12 upon treatment with polyphosphoric
acid.13
as
a 1:1 mixture of isomers which were separated by
chromatography.
N-Alkylation of tetrahydro-2H-3-benzazepin-2-one 4 with tert-
butyl bromoacetate in the presence of cesium carbonate as base
provided 5, which was subsequently alkylated at the benzylic posi-
tion with iso-butyl iodide mediated by the Schwesinger phospha-
zene base tert-butyl-P4.8,9 The trans-isomer was formed
exclusively (ds >99:1).10 Other bases such as sodium hydride and
lithium bis(trimethylsilyl)amide were not effective. Finally, the
tert-butyl ester
6 was cleaved with trifluoroacetic acid in
dichloromethane (v/v = 1:2) to provide racemic 1. However, the
introduction of important electron-rich phenyl substituents at the
5-position5a of the 3-benzazepin-2-one core was not feasible via
this synthetic route. In our hands, electron-rich styrenes, such as
1-methoxy-2-vinylbenzene or 1,2-dimethoxy-3-vinylbenzene,
yielded side-products during Friedel–Crafts reaction with phenyl-
acetyl chlorides. Thus, we envisaged to prepare the 3-benzazepin-
Indanone 12 was reacted with (2,3-dimethoxyphenyl)lithium to
form 6-chloro-1-(2,3-dimethoxyphenyl)indan-1-ol which was
subsequently converted to indene 13 by acid catalyzed dehydra-
tion with p-toluenesulfonic acid. The 2-carboxymethylphenyl ke-
tone 14 was obtained from indene 13 by oxidative cleavage with
a
b
c, d
Cl
CN
O
NH
O
O
O
OEt
O
OEt
10
7
8
9
Scheme 2. Reagents and conditions: (a) anhyd ethanolic HCl, 25 °C (99%); (b) TMSCN, cat. SnCl4, CH2Cl2, 30 min (91%); (c) 10 equiv NaBH4/CF3CO2H, 5 h, 25 °C; (d) PhCH3, 8 h
reflux (35% over two steps).

2556
N. Griebenow et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2554–2558
Cl
Cl
O
O
O
a
b
O
OH
+
O
H
O
11
O
O
O
O
c,d
e
Cl
O
Cl
O
Cl
12
13
14
O
OH
j, k
O
O
O
O
f, g
h, i
Cl
CN
Cl
O
O
O
O
15
16
O
O
O
O
O
O
O
l
m, n
Cl
OH
O
Cl
Cl
N
OEt
NH
O
N
O
O
17
18
rac-19
Scheme 3. Synthetic route B. Reagents and conditions: (a) NEt3, HCO2H, 95 °C, 2 h, (88%); (b) polyphosphoric acid, 100 °C, 1 h (60%); (c) veratrol, n-BuLi, ꢀ78 °C to rt, 3 h; (d)
p-toluenesulfonic acid, CH2Cl2, rt, 16 h (22% over two steps); (e) RuCl3ꢁH2O, NaIO4, hexane/ACN/H2O (2:2:3), rt, 16 h, (49%); (f) NaBH4, ethanol, reflux, 2 h; (g) 10% aq HCl,
40 °C, 1 h, (86% over two steps); (h) TMSCN, cat. I2, CH2Cl2, rt, 18 h; (i) TMSCl, MeOH, rt, 16 h, (54% over two steps); (j) NaBH4, CoCl2ꢁ6H2O, MeOH, rt, 1 h; (k) toluene, reflux,
16 h, (41% over two steps); (l) ethyl 2-bromoacetate, Cs2CO3, DMF, rt, 16 h (30%); (m) iso-butyliodide, tert-butyl-P4, ꢀ78 °C, 2 h, ds >99:1, (50%); (n) NaOH aq, THF, MeOH, rt,
16 h (99%).
sodium periodate and ruthenium(III) chloride.14 Reduction of the
ketone 14 with sodium borohydride followed by acid-promoted
lactonization of the thereby formed alcohol provided the target
isochromen-3-one 15.
In analogy to the reaction protocol of Busacca et al. we tried
ring-opening reaction of isochromen-3-one 15 with ethanolic
HCl. However, the starting material underwent decomposition un-
der these reaction conditions. Therefore, we investigated the direct
introduction of the cyano group. Inspired by the efficiency of the
iodine catalyzed cyanation of carbonyl compounds with trimethyl-
silyl cyanide,15 we considered to adopt this reaction protocol for
the ring-opening of 15. Indeed, iodine promoted cyanation of 15
with trimethylsilyl cyanide proceeded smoothly to give the ring-
opened species, which was treated with trimethylsilyl chloride
and methanol to yield the cyanoester 16.
This intermediate was then converted to the crude aminoester
with
a
cobalt(II) chloride-promoted sodium borohydride
reduction16 and thermally cyclized to lactam 17 in moderate yield.
Lactam 17 was elaborated further to the N- and C-alkylated 3-ben-
zazepin-2-one 19 as described for compound 4.
Figure 2. The molecular structure of the racemic compound 19 obtained by X-ray
diffraction analysis.
X-ray crystal structure analysis of racemic compound 19
revealed trans-orientation of the iso-butyl and the 2,3-dimethoxy-
phenyl substituent (Fig. 2).
Using the route outlined in Scheme 1 (synthetic route A) we
briefly explored variation at the C-7 position of the lead 1 (Table 1).
SAR at the C-7 position of the 3-benzazepin-2-one revealed that a
chlorine substituent is clearly favored over other substituents such
as methyl, fluorine and hydrogen (see 21, 22, 23). Migration of the
7-chloro atom in the fused benzene ring to the 6- and 8-positions
yielded compounds which had poor potencies (data not shown).
Modification at the C-1 position of the seven-membered ring had
a great influence on potency. The 1-benzyl compound 26 was

N. Griebenow et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2554–2558
2557
Table 1
SAR of the tetrahydro-2H-3-benzazepin-2-ones
4'
3'
R3
5'
2'
6'
O
R
6
R1
R4
5
1
7
8
N
3
2
R
9
O
R2
R1
R2
R3
R4
IC50
(l
M)
HLMb (%)
Sterol biosynthesisc (%)
a
Compd
Route
rac-1
A
A
A
A
A
A
A
A
A
A
B
B
Cl
Cl
Me
F
iso-Butyl
iso-Butyl
iso-Butyl
iso-Butyl
20-Cl
20-Cl
20-Cl
20-Cl
20-Cl
20-Cl
20-Cl
20-Cl
H
OH
OEt
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
3.3
60
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
rac-20
rac-21
rac-22
rac-23
rac-24
rac-25
rac-26
rac-27
rac-28
rac-29
rac-19
>20
>20
>20
>20
5.5
>20
>20
15.3
2.2
H
iso-Butyl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
2-Methylprop-2-en-1-yl
2-Hydroxy-2-methylpropyl
Benzyl
iso-Butyl
iso-Butyl
20-Br
iso-Butyl
iso-Butyl
20-OMe
20,30-diOMe
1.9
1.2
H
N
O
*
rac-30
rac-31
A
A
Cl
Cl
iso-Butyl
iso-Butyl
20-Cl
2.4
1.4
nd
70
nd
25
OH
O
OH
20-Cl
N
N
N
N
*
*
*
*
O
rac-32
rac-33
rac-34
A
B
B
Cl
Cl
Cl
iso-Butyl
iso-Butyl
iso-Butyl
20-Cl
0.93
0.37
0.50
53
2
25
nd
39
OH
20-OMe
20-OMe
O
72
OH
O
20,30-diOMe
20,30-diOMe
20,30-diOMe
0.45
>20
78
nd
89
nd
nd
47
N
N
N
rac-35
B
B
B
Cl
Cl
Cl
iso-Butyl
iso-Butyl
iso-Butyl
*
*
*
OH
O
(1S,5S)-36
(1R,5R)-37
OH
O
0.22
OH
a
b
c
Values are means of three experiments. For a detailed description of the biochemical assay see Ref. 4.
Human liver microsomal stability, % compound remaining after 60 min.
Inhibition of sterol biosynthesis in %. For details see Refs. 6,18.
completely devoid of any activity. In the series of C-alkyl deriva-
tives (1, 24, 25 and 26), iso-butyl compound 1 was found to be
the most potent inhibitor of squalene synthase. Biological data
for compounds having various substituents on the 5-phenyl ring
are also shown in Table 1. Compound 27 with no substituent at
the 20-position was significantly less potent than the 20-chloro ana-
logue 1.
Esters such as 20 were found to be inactive, whereas the piper-
idine derivative 33 showed improved potency but suffered from
inferior stability in human liver microsomes. The piperidine-4-car-
boxylic acid derivatives 32, 34 and 35, finally, combined good
potencies with markedly improved microsomal stabilities. Best
properties regarding potency and stability were displayed by com-
pound 35. It was therefore separated into its enantiomers 36 and
37 by chiral HPLC; only the (1R,5R)-isomer17 37 was found to be
potent against squalene synthase.
Compound 37 was progressed further to in vivo animal studies.
Inhibition of hepatic cholesterol biosynthesis was investigated in
NMRI-mice. After po administration of 3 mg/kg, 37 showed a 47%
reduction in sterol biosynthesis (Table 1).18
Replacement of the chloro atom at the 20-position with a bromo
and a methoxy group led to even more active compounds 28 and
29. Best potency was obtained by a 20,30-dimethoxy substitution,
prepared via synthetic route B (see compound 19). Other substitu-
tion patterns were found to be less promising (data not shown).
Finally SAR was investigated varying mainly the carboxylic acid
(R4) as outlined in Table 1. Amide formation was accomplished un-
der standard coupling conditions (EDC, HOBt, DIEA).
Inhibitor 37 exhibited nearly the same in vitro potency com-
pared with lapaquistat (37: IC50 = 0.22 lM; lapaquistat:

2558
N. Griebenow et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2554–2558
concn
incubated for 10 min at 37 °C. Subsequently, 100
with 200 l chloroform, 200 l methanol and 60 l 5 N sodium hydroxide and
adjusted to 2 mM squalene. An aliquot of the organic phase was mixed with
scintillation fluid and measured for beta-emission.
5
l
M) and 0.2 nM trans,trans-[1-3H]-farnesyl pyrophosphate and
l solution was extracted
IC50 = 0.21 l
M).19 Furthermore 37 has a higher stability in human
l
liver microsomes than the pharmacologically active metabolite of
lapaquistat T-91485l (37: 89% remaining after 60 min, T-91485:
58%remainingafter 60 min). Howeverlapaquistatshoweda 1.5-fold
higher reduction in sterol biosynthesis than 37 in NMRI-mice.20
In conclusion, we have identified a novel series of substituted
tetrahydro-2H-3-benzazepin-2-ones as potent squalene synthase
inhibitors. Two complementary synthetic routes have been estab-
lished, which allowed for the flexible decoration of the scaffold.
Lead optimization of 1 resulted in the identification of 37 with
favorable potency and microsomal stability suitable for in vivo
studies. Lead 37 showed the expected profile of a squalene syn-
thase inhibitor regarding sterol biosynthesis after oral administra-
tion. Further pharmacological studies of 37 will be reported in due
course.
l
l
l
5. (a) Miki, T.; Kori, M.; Fujishima, A.; Mabuchi, H.; Tozawa, R.; Nakamura, M.;
Sugiyama, Y.; Yukimasa, H. Bioorg. Med. Chem. 2002, 10, 385; (b) Miki, T.; Kori,
M.; Mabuchi, H.; Banno, H.; Tozawa, R.; Nakamura, M.; Itokawa, S.; Sugiyama,
Y.; Yukimasa, H. Bioorg. Med. Chem. 2002, 10, 401.
6. For experimental details and characterization of the compounds see:
Griebenow, N.; Flessner, T.; Raabe, M.; Bischoff, H. Ger. Offen. DE
102004006325, 2005.
7. (a) Schmidt, K. F. Angew. Chem. 1923, 36, 511–524; (b) Schmidt, K. F. Chem. Ber.
1924, 57, 704.
8. tert-Butyl-P4 = N000-tert-butyl-N,N0,N00-tris[tris(dimethylamino)phosphoranylidene]-
phosphorimidic triamide.
9. Schwesinger, R.; Schlemper, H. Angew. Chem., Int. Ed. Engl. 1987, 26, 1167.
10. The relative stereochemistry was assigned by analogy to the X-ray crystal
structure of 19. Diastereoselectivity was determined by LC/MS and 1H NMR.
11. Busacca, C. A.; Johnson, R. E. Tetrahedron Lett. 1992, 33, 165.
12. Tóth, G.; Kövér, K. E. Synth. Commun. 1995, 25, 3067.
13. Simchen, G.; Krämerer, W. Chem. Ber. 1969, 102, 3656.
14. McDonald, I. M.; Dunstone, D. J. PCT Int. Appl., WO 2006051312, 2006.
15. Yadav, J. S.; Reddy, B. V. S.; Reddy, M. S.; Prasad, A. R. Tetrahedron Lett. 2002, 43,
9703.
References and notes
1. (a) Carpender, K. I.; Taylor, S. E.; Ballantine, J. A.; Fussell, B.; Halliwell, B.;
Mitchinson, M. J. Biochim. Biophys. Acta 1993, 1167, 121; (b) Brown, A. J.;
Mander, E. L.; Gelissen, I. C.; Kritharides, L.; Dean, R. T.; Jessup, W. J. Lipid Res.
2000, 41, 226; (c) Kourounakis, A. P.; Charitos, C.; Rekka, E. A.; Kourounakis, P.
N. J. Med. Chem. 2008, 51, 5861.
2. Walley, T.; Folino-Gallo, P.; Stephens, P.; Van Ganse, E. Br. J. Clin. Pharmacol.
2005, 60, 543.
3. (a) Burnett, J. Curr. Opin. Investig. Drugs 2006, 7, 850; (b) Seiki, S.; Frishman, W.
H. Cardiol. Rev. 2009, 17, 70.
16. Satoh, T.; Suzuki, S.; Suzuki, Y.; Miyaji, Y.; Imai, Z. Tetrahedron Lett. 1969, 52,
4555.
17. Absolute stereochemistry at the C-5 position of the 3-benzazepin-2-one 37
was assigned by analogy to the stereoselective inhibitory potencies of 4,1-
5b
benzoxazepine derivatives described in Ref.
.
18. The compounds were administered orally to NMRI-mice (n = 8) one hour before
monitoring of the cholesterol synthesis started. Biosynthesis of labeled
cholesterol and its precursors was monitored after ip application of [14C]-
mevalonolactone. One hour after mevalonolactone application, livers were
removed and the amount of hepatic cholesterol synthesis was determined by
4. Squalene synthase catalyses the reductive condensation of farnesyl
pyrophosphate to squalene during cholesterol biosynthesis. Turnover of
trans,trans-[1-3H]-farnesyl pyrophosphate to [3H]-squalene by squalene
sterol extraction.
19. In-house data from the biochemical assay described in Ref.
synthase was determined under the following conditions: 0.03 lg recombinant
4
.
SQS, 1 mM NADPH, 5 mM DTT, 10% PBS, 10 mM sodium fluoride, 0.5 mM MgCl2,
pH 7.5. Test compounds were dissolved in DMSO and added in defined
concentrations. The assay was started by adding farnesyl pyrophosphate (final
20. After po administration of 3 mg/kg (37: 47% reduction, lapaquistat: 73%
reduction).