Discovery of potent, nonsystemic apical sodium-codependent bile acid transporter inhibitors (part 1)

Article abstract of DOI:10.1021/jm040215+

Elevated plasma levels of low-density lipoprotein (LDL) cholesterol are a major risk factor for atherosclerosis leading to coronary artery disease (CAD), which remains the main cause of mortality in Western society. We believe that by preventing the reabsorption of bile acids, a minimally absorbed apical sodium-codependent bile acid transporter (ASBT) inhibitor would lower the serum cholesterol without the potential systemic side effects of an absorbed drug. A series of novel benzothiepines (3R,3R′-2,3,4,5-tetrahydro-5-aryl-1- benzothiepin-4-ol 1,1-dioxides) were synthesized and tested for their ability to inhibit the apical sodium dependent bile acid transport (ASBT)-mediated uptake of [¹⁴C] taurocholate (TC) in H14 cells. A 3R,4R,5R/3S,4S,5S racemate was found to have greater potency than the other three possible racemates. Addition of electron-donating groups such as a dimethylamino substituent at the 7 position greatly enhanced potency, and incorporation of a long-chain quaternary ammonium substituent on the 5-phenyl ring was useful in minimizing systemic exposure of this locally active ASBT inhibitor while also increasing water solubility and maintaining potency. The reported results describe the synthesis and SAR development of this benzothiepine class of ASBT inhibitors resulting in an 6000-fold improvement in ASBT inhibition with desired minimal systemic exposure of this locally acting drug candidate.

Full text of DOI:10.1021/jm040215+

J. Med. Chem. 2005, 48, 5837-5852
5837
Discovery of Potent, Nonsystemic Apical Sodium-Codependent Bile Acid
Transporter Inhibitors (Part 1)
Samuel J. Tremont,*,‡ Len F. Lee,‡ Horng-Chih Huang,*,† Bradley T. Keller,§ Shyamal C. Banerjee,‡
Scott R. Both,‡ Andrew J. Carpenter,† Ching-Cheng Wang,† Danny J. Garland,† Wei Huang,† Claude Jones,‡
Kevin J. Koeller,‡ Steve A. Kolodziej,† James Li,† Robert E. Manning,† Matthew W. Mahoney,†
Raymond E. Miller,‡ Deborah A. Mischke,† Nigam P. Rath,# Theresa Fletcher,† Emily J. Reinhard,†
Michael B. Tollefson,† William F. Vernier,† Grace M. Wagner,‡ Steve R. Rapp,§ Judy Beaudry,§ Kevin Glenn,§
Karen Regina,§ Joe R. Schuh,§ Mark E. Smith,§ Jay S. Trivedi,| and David B. Reitz†
Department of Discovery Chemistry and Department of Cardiovascular Disease, Pharmacia, 700 Chesterfield Parkway W,
Chesterfield, Missouri 63017, Office of Science and Technology, Chemical Science Division, Pharmacia, 800 Lindbergh
Boulevard, Creve Coeur, Missouri 63167, Department of Pharmaceutical Sciences, Pharmacia, Skokie, Illinois, and Department
of Chemistry, University of Missouri, St. Louis, Missouri
Received December 14, 2004
Elevated plasma levels of low-density lipoprotein (LDL) cholesterol are a major risk factor for
atherosclerosis leading to coronary artery disease (CAD), which remains the main cause of
mortality in Western society. We believe that by preventing the reabsorption of bile acids, a
minimally absorbed apical sodium-codependent bile acid transporter (ASBT) inhibitor would
lower the serum cholesterol without the potential systemic side effects of an absorbed drug. A
series of novel benzothiepines (3R,3R′-2,3,4,5-tetrahydro-5-aryl-1-benzothiepin-4-ol 1,1-dioxides)
were synthesized and tested for their ability to inhibit the apical sodium dependent bile acid
transport (ASBT)-mediated uptake of [14C]taurocholate (TC) in H14 cells. A 3R,4R,5R/3S,4S,5S
racemate was found to have greater potency than the other three possible racemates. Addition
of electron-donating groups such as a dimethylamino substituent at the 7 position greatly
enhanced potency, and incorporation of a long-chain quaternary ammonium substituent on
the 5-phenyl ring was useful in minimizing systemic exposure of this locally active ASBT
inhibitor while also increasing water solubility and maintaining potency. The reported results
describe the synthesis and SAR development of this benzothiepine class of ASBT inhibitors
resulting in an 6000-fold improvement in ASBT inhibition with desired minimal systemic
exposure of this locally acting drug candidate.
Introduction
evidence for the clinical and arteriographic benefits of
lipid profile normilization by partial ileal bypass, and
longterm followup POSCH data demonstrate that lipid
profile normalization decereased overall mortality.4
Perhaps the most compelling results were the findings
from the CARE study (pravastatin), which indicated
that a 28% lowering of the plasma LDL cholesterol in
patients at risk for heart disease with “normal” total
cholesterol levels (∼210 mg/dL) decreased their risk of
CAD-related morbidity and mortality 24%.5 Results
from the CARE study suggests that therapeutic benefit
can be derived from treating patients that have choles-
terol levels previously considered to be on the high end
of acceptable.
A nonsystemic strategy of lowering cholesterol would
complement the existing HMG-CoA reductase inhibitors
(statins) and should also reduce the incidence of systemic-
related side effects. The bile acid resins are an example
of a nonsystemic LDL lowering therapy that have the
desired safety profile for treating people with moder-
ately elevated cholesterol, but they are underutilized
because of their high dosages and poor palatability. Our
studies began with the search for a more potent bile acid
sequestrant as an approach to a nonsystemic LDL
lowering therapy. When considering ways of improving
the in vivo efficiency of bile acid sequestrants, the ileal
bile active transporter appeared to be a prime suspect
Clinical data strongly indicate that elevated plasma
levels of low-density lipoprotein (LDL) cholesterol are
a major risk factor for atherosclerosis leading to coro-
nary artery disease (CAD), the number one cause of
morbidity and mortality in Western society.1 The im-
portance of lipid-lowering therapies has been well
documented in recent years, leading to revised treat-
ment guidelines in 2001 by the National Cholesterol
Education Program (NCEP) of the National Institutes
of Health that recommend a more aggressive approach
toward hypercholesterolemia.2
Results from recent clinical trials involving large
patient populations with CAD demonstrate that long-
term therapy with cholesterol-lowering agents has a
substantial positive impact on survival (lovastatin in
ASCAPS; pravastatin in West of Scotland study; sim-
vastatin in 4S study).3 Also The Program on the Surgical
Control of the Hyperlipidemiaa (POSCH) provided solid
* To whom correspondence should be addressed. For S.J.T.: phone,
(636) 391-8993; fax, (636) 247-6953; e-mail, sjtrem1@charter.net. For
H.-C.H.: phone (636) 247-6350; fax (636) 247-6953; e-mail, horng-
chih.huang@Pfizer.com.
‡ Chemical Science Division, Pharmacia.
† Department of Discovery Chemistry, Pharmacia.
§ Department of Cardiovascular Disease, Pharmacia.
# University of Missouri.
| Department of Pharmaceutical Sciences, Pharmacia.
10.1021/jm040215+ CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/06/2005

5838 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
1, R1 or R2 ) Ph) is outlined in Scheme 1. Another
approach, which integrates substituents on the ben-
zothiepine A ring, is outlined in Scheme 2. Scheme 3
summarizes a method for attaching a positively charged
quaternary ammonium group to the benzothiepine
backbone. Scheme 4 summarizes methods for incorpo-
rating substituents on the 5-phenyl ring and at the 7,
8, and 9 positions. In all methods (Schemes 1-4), a key
intermediate is the sulfone aldehyde.11 The cyclization
of this intermediate with KOtBu occurs with excellent
regioselectivity, giving only the cis products as a racemic
mixture.
Scheme 1 describes the synthesis of the potent class
of 7-amino substituted benzothiepines with further
functionalization of the 5-phenyl ring. Benzophenones
1 were treated with triethylsilane/triflic acid to give the
diphenylmethanes 2.14 Displacement of the chlorine
with lithium disulfide,15 followed by nucleophilic addi-
tion of the resulting lithium thiophenoxide intermedi-
ates (not shown) to the mesylate aldehydes, gave the
sulfide aldehydes 3. Selective oxidation of the sulfides
3 with mCPBA gave the sulfone intermediates 4 (X )
NO2).16 Low-pressure catalytic hydrogenation of the
nitro group gave the hydroxylamines 6 (X ) NHOH),
but higher pressures (100 psi) gave completely reduced
amine 4 (X ) NH2). Both reduction products were
amino-protected as carbamates 5 and 6 (X ) NOHBOC).
Under cyclization conditions, these amino-protected
aldehydes gave stereoselectively the desired benzothi-
epines 7 (X ) NOHBOC) and 8 (X ) NHCO2R) with
the hydroxyl and aryl groups in a cis relationship.
Deprotection conditions gave the free amine 8 (X ) NH2)
and hydroxylamine 7 (X ) NHOH) products. The free
amine products were further substituted at nitrogen via
alkylation conditions to give dialkylbenzothiepines and
trialkylaminobenzothiepines 8 (X ) NRR′, NRR′Me+I-).
Another useful approach that incorporates substitu-
ents at the aromatic A ring is outlined in Scheme 2.
Phenols 9 were treated with benzoyl chloride/boron
trichloride to give the benzophenones 10.17 Alkylation
of the hydroxy group with a dimethylthiocarbamoyl
chloride gave thiocarbamate esters 11. These esters
underwent thermal rearrangement, followed by base
hydrolysis, to give the desired thiophenols 12.18 Sulfur
alkylation and oxidation (as described for compounds 3
and 6, Scheme 1), followed by catalytic hydrogenation
of the benzophenone group, gave the sulfone aldehydes
14. Cyclization conditions gave stereoselectively the
benzothiepines 15 and 16. In an alternative approach
to thiophenols 12, 4-methoxythiophenol was treated
with benzonitrile.19
Scheme 3 describes the attachment of positively
charged groups to either the 5-phenyl ring or the A ring.
Methoxy substituted benzothiepines 17 were treated
with BBr3 to give hydroxybenzothiepines 18. Alkylation
of the hydroxy group gave the iodoalkoxybenzothiepines
19. Finally, reaction with a tertiary amine gave the
trialkylaminobenzothiepines 20.
Figure 1. Benzothiepines.
for the commercial products’ inefficiency;6 the active
transporter effectively removes the bile acids from the
resin via an irreversible sodium-dependent mechanism.
Because of this irreversible pathway, we decided to shift
our attention to small-molecule apical sodium-codepen-
dent bile acid transporter (ASBT) inhibitors, which
would offer a more direct route to remove bile acids from
the enterohepatic circulation via a nonsystemic mech-
anism. A palatable, potent, safe, and nonabsorbed ASBT
inhibitor would be an improved stand-alone medication
over current therapies to lower plasma LDL cholesterol.
In this first of two papers, we will describe our path
to a late lead that meets all requirements of a potent,
nonsystemic ASBT inhibitor. In part 2, we will describe
our strategy toward even more potent, nonsystemic
ASBT inhibitors, as well as optimization to acceptable
commercial candidates for clinical evaluation.
In Vitro Assay
Identification and selection of lead compounds that
inhibit ASBT-mediated bile acid transport were ac-
complished using a cell-based high-throughput screen.
A transfected baby hamster kidney cell line that con-
stitutively expresses human ASBT was established
early in the project (H-14) and has provided a high-
throughput system to initially evaluate all chemical
leads. ASBT inhibitory activity is assessed on the basis
of the ability of compounds to inhibit the cellular uptake
of 5 µM [14C]taurocholate during a 2 h incubation period.
Selectivity is tested in the same assay system using 5
µM [14C]alanine in place of taurocholate to determine
the effect on another cellular sodium-dependent cotrans-
porter. All compounds reported in this paper had greater
than 100-fold selectivity against the alanine assay.
Synthesis
When we began our search for ASBT inhibitors bile
acid analogues in 1990, bile acid oligomers,7 bile acid
dimers,8 and benzothiazepines9 were shown to be ASBT
inhibitors. The benzothiazepine class represented a new
class of inhibitors with improved in vitro activity over
the bile acid systems. Recent reports have shown that
some benzothiazepine analogues were quite potent in
a rat in vivo model.10 In this paper, we describe the
synthesis and pharmacological evaluation of substituted
benzothiepines11 (3R,3R′-2,3,4,5-tetrahydro-5-aryl-1-
benzothiepin-4-ol 1,1-dioxides, Figure 1). By synthetic
modification of the benzothiepine ring system, the in
vitro potency of our ASBT inhibitors have improved
more than 6000-fold as measured in the H-14 cell
culture assay system. Since we began our research, a
number of interesting inhibitor classes have been
developed.12,13
Another useful method for preparing the diphenyl-
methane intermediate directly from phenols is described
in Scheme 4. Phenols 21 were C-benzylated to give
diphenylmethanes 22. Following a similar transforma-
tion sequence described for compounds 15 and 16
We employed various routes in preparing substituted
benzothiepines. Detailed synthetic procedures for these
transformations are provided in the Experimental Sec-
tion. One of our initial approaches that we used to
incorporate substituents at the 5-phenyl ring (Figure

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5839
Scheme 1a
a Reagents: (a) Et3SiH, TfOH, CH2Cl2; (b) Li2S, DMSO, then MsOCH2C(R2)CHO; (c) mCPBA; (d) 100 psig of H2, Pd/C; (e) low pressure
H2, Pd/C; (f) K2CO3, R′OCOX; (g) KOtBu, THF; (h) HCl, dioxane; (i) EtSH, BF3OEt2, CH2Cl2; (j) RCOCl or MeSO2Cl, 4-methylmorpholine;
or metal, H2, RCHO; (k) MeI, CH3CN.
Scheme 2a
temic availability and increase water solubility without
decreasing potency. The SAR studies described in this
paper include (A) relative stereochemistry at the 3, 4,
and 5 positions, (B) aromatic substitution of both phenyl
rings, and (C) alkyl substitution at the 3 position.
Our early benzothiepine leads included compounds 34
and 35. To correlate the relative stereochemistry of
positions 3, 4, and 5 with biological activity, we prepared
and isolated four isomers as racemates (see Table 1).
From this preliminary study, we discovered that opti-
mum in vitro activity occurred when the phenyl, hy-
droxyl, and ethyl groups are all in the cis configuration
(34a). Further SAR at the C-3 position will be discussed
in a later section. Preliminary in vivo studies with 34a
indicated modest activity, and thus, further modifica-
tions were developed to increase potency.
Characterization of 34 and 35. Proton, COSY
(correlation spectroscopy), and NOESY (nuclear Over-
hauser effect spectrometry) NMR experiments and
X-ray crystallography were used to assign stereochem-
istry in compounds 34a,b and 35a,b. Proton NMR easily
distinguished the two cis compounds 34a and 35a from
the trans compounds 34b and 35b because the trans
compounds exhibit the typical trans three-bond cou-
plings for H4 and H5 and the cis compounds do not. In
1H NMR, compound 34a exhibits two doublets of triplets
(δ ) 1.65 and 2.24 ppm) with typical two-bond (J ≈ 14
Hz) and three-bond coupling constants (J ) 4 Hz),
indicative of the ring-adjacent methylene protons of the
butyl group. These same doublets of triplets exhibit
NOESY interactions with H4 and H5, which indicates
that H4, H5, and the butyl group are on the same side
of the benzothiepine ring. X-ray crystallography data
for 34a (Figure 2) confirms this configuration. Com-
pound 35a exhibits two doublets of quartets (δ ) 1.76
a Reagents: (a) PhCOCl, BCl3; (b) NaH, ClC(S)NMe2, DMF; (c)
PhOPh, D; (d) KOH, MeOH; (e) BuLi, PhCN, tetramethylethyl-
enediamine; (f) NaH, MsOCH2CRR′CHO; (g) mCPBA; (h) Pd/C,
H2; (i) KOtBu, THF.
(Scheme 2), these diphenylmethanes 22 were sequen-
tially converted to compounds 23-26.
Results and Discussion
In early studies, our primary focus was to increase
potency and reduce the number of stereogenic centers.
Later, we sought out structural features to limit sys-

5840 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
Scheme 3a
a Reagents: (a) BBr3 CHCl3; (b) I(CH2)6I or I(CH2CH2O)2CH2CH2I, K2CO3, THF; (c) NR3, CH3CN.
Scheme 4a
a Reagents: (a) NaH, YPhCH2X (X ) Cl, Br); (b) NaH, ClC(S)NMe2, DMF; (c) PhOPh, D; (d) KOH, MeOH; (e) NaH, MsOCH2C(R1)2CHO;
(f) mCPBA; (g) KOtBu, THF.
and 2.32 ppm), with typical two-bond (J ) 13.7 Hz) and
three-bond coupling constants (J ) 7.5 Hz), indicative
of the ethyl group’s methylene protons. These same
doublets of triplets exhibit NOE interactions with H4
and H5, which indicates that H4, H5, and the ethyl
group are on the same side of the benzothiepine ring.
In the 1H NMR spectrum of compound 35b, a multiplet
(δ ) 1.70 ppm, two protons) has a strong COSY
interaction with a methyl group triplet (δ ) 0.91 ppm),
which indicates it must represent the ethyl’s methylene.
This same multiplet has a strong NOESY interaction
with H5, which indicates the ethyl group is on the same
side of the benzothiepine ring as H5. Compound 34b’s
stereochemistry was assigned via lack of obvious NMR
distinctions and via the process of elimination.
In a search of more potent analogues, we prepared
various substituted benzothiepines (see Table 2) that
contained the same 3, 4, 5 configuration as compound
34a. From this SAR study, we found that electron-
donating groups in the 7 or 8 position increased in vitro
potency, with the 7-substituted dimethylamino analogue
being the most potent (IC50 ) 0.005 uM). For compari-
son the IC50 for chenodeoxycholic acid is 50 µM. These
results suggest that the dimethylamino group has the
right size and electronic properties for optimum activity.

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5841
Table 3. Removal of A Stereogenic Center
Table 1. Preliminary SAR of Parent Compounds (()-34 and
(()-35
compd
R1
R, R
X
IC50 (µM) Scheme
55
56
57
58
60
7-NH2 Me, Me
7-NH2 Et, Et
7-NH2 Pr, Pr
7-NH2 Bu, Bu
H
>50
4
0.23
0.054
1
1
1
1
4
compd
R1
R2
IC50 (µM)
synthesis
Scheme 2
H
H
H
(()-34a
(()-34b
(()-35a
(()-35b
H
OH
H
OH
H
0.28
4.8
OH
H
2
Scheme 2
7-Me
pentyl, pentyl 4’-F 0.318
OH
37% at 10 µM
Table 4. Comparison of Et, Bu with Bu, Bu and Further SAR
compd
R1, R2
R
X
IC50 (µM)
Scheme
61
62
39
58
45
64
65
66
67
68
Et, Bu
Bu, Bu
Et, Bu
Bu, Bu
Et, Bu
Bu, Bu
Et, Bu
Bu, Bu
Et, Bu
Bu, Bu
8-OMe
H
0.12
2
2
1
1
1
1
1
1
4
4
0.19
7-NH2
H
0.068
0.054
0.005
0.013
0.083
0.21
7-NMe2
7-N+Me3
7-Me
H
H
Figure 2. ORTEP of compound (()-34a.
Table 2. SAR by Substitution on Aromatic Rings of 34
4’-F
0.021
0.046
study (see Table 3) suggests that the dimethyl analogue
is too small to participate in binding the protein’s
hydrophobic pocket. Furthermore, chain length appears
proportional to in vitro potency up to the butyl group.
This suggests a size limitation in the hydrophobic
pocket. Our results also indicated the Bu, Bu system
(IC50 ) 0.054 µM) is as potent as the unsymmetrical
Et, Bu system (IC50 ) 0.068 uM). Although the Bu, Bu
analogue 58 still contains two chiral centers (positions
4 and 5), both are fixed simultaneously by the ste-
reospecific cyclization reaction that is utilized to form
the seven-member ring (see Scheme 1).
In general, the Bu, Bu analogues were moderately
less potent than the Et, Bu compounds (see Table 4).
Although the Et, Bu 7-dimethylamino system was the
most potent in this series at 0.005 µM, the Bu, Bu
system was also very potent (IC50 ) 0.013 uM). These
results encouraged us to focus on the Bu, Bu system in
the further optimization of our ASBT candidate. When
the dimethylamino group was alkylated to form the less
electron-donating quaternary ammonium salt, the ac-
tivity was reduced by approximately 15-fold for the Et,
Bu system (IC50 ) 0.083 uM) and 16-fold for the Bu,
Bu system (IC50 ) 0.21 uM).
compd
R
X
IC50 (µM)
Scheme
34a
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
27
28
7-H
7-OMe
8-OH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
0.28
0.055
0.09
16
2
2
2, 3
2, 3
1
8-OC10H19
7-NH2
7-NHOH
0.068
0.26
2.67
0.4
1
7-N(OH)CO2-tBu
7-NHCO2CH2Ph
7-NHCO2-tBu
7-NHC6H13
7-NMe2
1
1
1.3
1
1.67
0.005
0.09
0.3
1
1
7-NHCOCH3
7-NHCOC5H11
7-NHSO2CH3
7-SMe
7-pyrrolidinyl
7-Br
1
1
0.6
1
4-F
4-F
3’-OMe
3’-OMe
0.019
0.059
0.07
0.18
4
4
4
7-Ph
4
The larger alkyl amino groups and electron-withdraw-
ing groups decreased potency.
Next we investigated the possibility of replacing the
sterogenic center at the C-3 position, which seemed to
be critical in the benzothiazepine analogues with a
symmetrical diastereotopic dialkyl substitiuent. This
Substituent position and electron-donating properties
seem to significantly affect potency of the Bu, Bu series
(see Table 5). A dimethylamino group para to the
sulfone 76 was almost 20-fold more potent than when
the dimethylamino group is placed ortho to the sulfone
81. Interestingly, we saw a slight increase in potency

5842 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Table 5. Further SAR in the Bu, Bu Series
Tremont et al.
Table 7. The 0.24% Cholesterol-Fed Hamster Model [14 Days,
0.2% Dietary Admixture, ∼200 (mg/kg)/day]a
serum total
fecal bile acids (FBA) 24 h fecal
compd cholesterol (mg/dL) ((mmol/24 h)/100 g) weight (g)
control
61
143 ( 7
6.2 ( 0.8
2.3 ( 0.1
2.4 ( 0.04
126 ( 2a
11.9 ( 0.5a
control
39
129 ( 9
126 ( 7
6.1 ( 0.4
8.8 ( 1.1a
1.7 ( 0.1
1.6 ( 0.1
control
45
135 ( 4
97 ( 5a
8.0 ( 0.7
2.1 ( 0.2
2.1 ( 0.1
compd
R
X
IC50 (µM)
Scheme
13.7 ( 0.4a
control
195 ( 28
8.7 ( 0.6
3.5 ( 0.2
3.9 ( 0.3
64
73
74
75
76
80
81
88
89
90
68
93
7-NMe2
H
H
H
4’-OMe
4’-F
H
4’-F
4’-OMe
2’-Br
4’-F
0.013
1.9
1
1
1
1
4
4
4
4
4
4
4
4
97
141 ( 13a
16.2 ( 0.6
7-NHCH2Ph
7-NHCO2CH2Ph
7-NMe2
0.6
a P < 0.05, Students T-test.
0.007
0.01
0.12
0.18
0.002
0.55
0.05
0.046
0.18
7-NMe2
In Vivo Efficacy Studies of Benzothiepine Ana-
logues 39, 45, 61, and 97. Our lead ASBT inhibitors
were tested in a 14-day cholesterol-fed hamster model.
The primary measurements were fecal bile acids (FBA)
and serum cholesterol. Table 7 illustrates key results
from this study. All compounds chosen for this study
were active. For compounds 61 and 39 we believe 200
(mg/kg)/day is not a maximally effective dose because
of significant absorption of these compounds in the
proximal portion of the small intestine and the lower
potency of these compounds (120 and 68 nM, respec-
tively). However we believe this is a maximally effective
dose for compounds 45 and 97 because of their markedly
increased potency (5 and 3 nM, respectively) and the
fact that compound 97 exhibits little, if any, absorption
from the GI tract. This study showed significant in-
creases of FBA (140-190%), which is consistent with
inhibiting bile acid reabsorption at the apical bile acid
transporter site in the ileum. In response to the in-
creased excretion of bile acids, serum cholesterol levels
were significantly decreased. One of our earlier leads
61 (IC50 ) 120 nM) decreased cholesterol in the hamster
model by 12% while 45 (IC50 ) 5 nM) decreased total
cholesterol by 28%. Studies with 97 (IC50 ) 3 nM), which
was designed to be minimally absorbed and where
chemical complexity was reduced by removal of a chiral
center, indicated low systemic exposure and a 28%
decrease in total cholesterol in the hamster model. The
fact that this is a potent and minimally absorbed drug
further supports the hypothesis that inhibition at the
ASBT site in the ileum can lead to a decrease in plasma
cholesterol.
8-NMe2
9-NMe2
7-NMe2, 9-OMe
7-Br
7-OMe
7-Me
4’-F
4’-F
9-F
Table 6. Quarternary Ammonium Salts
compd
R+
IC50 (µM)
Scheme
95, F
96, F
97, G
-O(CH2)6N+Me3
21
18
2, 3
2, 3
1, 3
-(OCH2CH2)3N+Me3
-(OCH2CH2)3N+Et3
0.003
when the 7-dimethylamino system was further substi-
tuted in the 4′ position by a methyl ether group 75 (IC50
) 0.007 µM). Enhanced activity was also seen in the
7-dimethylamino system that was further substituted
in the 9 position by an electron-donating methoxy 88
(IC50 ) 0.002 µM). Larger amino groups (7 or 9 position)
were less active by about an order of magnitude. In
contrast, less electron-donating groups (e.g., Table 5,
last four entries) generally led to a decrease in activity
that seemed related to the substituent’s electron-donat-
ing ability and steric properties.
Conclusions
Because the site of action of this drug is located in
the distal region of the ileum, limiting systemic exposure
was thought to be advantageous. To limit systemic
exposure and increase water solubility, we employed a
series of analogues containing positively charged qua-
ternary ammonium substrates (see Table 6). When the
quaternary ammonium group was placed on a long
chain at the 8 position (95 and 96), a large decrease in
activity was observed. However, the charged group
minimally affected potency when attached at the 4′
phenyl position, especially when placed more than four
atoms from the phenyl ring. Compound 97 exhibited
The benzothiepine class of is one of the most potent
classes of inhibitors discovered, and it has been shown
in this paper that substituents with appropriate size
that donate electron density to the sulfone group have
a very positive effect on potency. Our SAR studies have
achieved increased potency (6000-fold from early lead)
and reduced chirality. We have also found a potent
water-soluble racemic analogue with low systemic ex-
posure: 97. Although analogue 97 is potent and mini-
mally absorbed, its physical properties (hygroscopic and
amorphous) are not optimal for a commercial candidate.
In our next studies we pursued further fine-tuning the
SAR of the substituted phenyl ring at the 5 position of
the benzothiepine ring with the goal of identifying a
single enantiomer analogue that is a potent, crystalline,
nonhygroscopic, and efficacious ASBT inhibitor with low
water solubility and fortunately was as potent (IC50
)
0.003 µM) as any of the uncharged analogues. Analogue
97 was shown to have less than 1% of the oral dose of
parent compound measurable in the systemic circulation
of rat.

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5843
alternating 12 h light and dark cycles. Hamsters were fed
Teklad 7001 rodent meal chow at libitum for 2 weeks before
the experimental studies began and the chow was switched
to Teklad 7001 rodent meal chow supplemented with 0.24%
cholesterol on day 1 of the 14 day experiment. Water was
continuously available to the animals. Hamsters were assigned
to groups in a manner that normalized each group by pre-
treatment body weights. Experimental compounds were ad-
ministered as a 0.2% dietary admixture in the meal chow. On
the basis fo average food consumption, the daily dose of
compound was calculated to be approximately 200 (mg/kg)/
day. Blood samples were collected after the 14 day treatment
period by cardiac puncture. Hamsters were anesthetized but
not fasted prior to blood collections. Fecal samples were
collected during a 48 h period at the end of the study (i.e.,
days 13-14).
Fecal Bile Acid Measurement.20 Fecal samples were
collected to determine the fecal bile acid (FBA) concentration
for each animal. The separate collections from each hamster
were weighed and homogenized with distilled water with a
Polytron tissue processor (Brinkman Instruments) to generate
a homogeneous slurry. Fecal homogenate (1.4 g) was extracted
with 2.6 mL of a solution containing tertiary butanol/distilled
water in the ratio of 2:0.6 [final concentration of 50% (v/v)
tertiary butanol] for 45 min in a 37 °C water bath and
subjected to centrifugation for 13 min at 2000g. The concen-
tration of bile acids (mmol/g homogenate) was determined
using a 96-well enzymatic assay system.1,2 Aliquots of the fecal
extracts (20 mL) were added to two sets of triplicate wells in
a 96-well assay plate. A standardized sodium taurocholate
solution and a standardized fecal extract solution (previously
made from pooled samples and characterized for its bile acid
concentration) were also analyzed for assay quality control.
Aliquots of 90 mM sodium taurocholate (20 mL) were serially
diluted to generate a standard curve containing 30-540 nmol/
well. A 230 mL aliquot of reaction mixture containing 1 M
hydrazine hydrate, 0.1 M pyrophosphate, and 0.46 mg/mL
NAD was added to each well. Subsequently, a 50 mL aliquot
of either 3a-hydroxysteroid dehydrogenase enzyme (HSD, 0.8
units/mL) or assay buffer (0.1 M sodium pyrophosphate) was
then added to one each of the two sets of triplicates. Following
60 min of incubation at room temperature, the optical density
at 340 nm was measured and the mean of each set of triplicate
samples was calculated. The difference in optical density (
HSD enzyme was used to determine the bile acid concentration
(mM) of each sample based on the sodium taurocholate
standard curve. The bile acid concentration of the extract
(mmol/g homogenate), the total weight of the fecal homogenate
(g), and the body weight of the hamsters (g) were used to
calculate the corresponding FBA concentration in (mmol/day)/
kg body weight for each animal. All reagents used for the assay
were obtained from Sigma Chemical Co., St. Louis, MO (HSD,
catalog no. H-1506; NAD, catalog no. N1636; sodium tauro-
cholate, catalog no. T-4009).
systemic exposure. These studies are described in part
2 of this series.
Experimental Section
In Vitro Assay of Compounds That Inhibit ASBT-
Mediated Uptake of [14C]Taurocholate (TC) in H14 Cells.
Baby hamster kidney cells (BHK) transfected with the cDNA
of human IBAT (H14 cells) are seeded in 96-well Top-Count
tissue culture plates at 60 000 cells/well for assays run within
24 h of seeding, 30 000 cells/well for assays run within 48 h,
and 10 000 cells/well for assays run within 72 h.
On the day of assay, the cell monolayer is gently washed
once with 100 mL of assay buffer (Dulbecco’s modified Eagle’s
medium with 4.5 g/L glucose plus 0.2% (w/v) fatty acid free
bovine serum albumin ((FAF)BSA)). To each well 50 mL of a
2-fold concentrate of test compound in assay buffer is added
along with 50 mL of 6 mM [14C]taurocholate in assay buffer
(final concentration of 3 mM [14C]taurocholate). The cell
culture plates are incubated 2 h at 37 °C prior to gentle
washing of each well twice with 100 mL of 4 °C Dulbecco’s
phosphate-buffered saline (PBS) containing 0.2% (w/v) (FAF)-
BSA. The wells are then gently washed once with 100 mL of
4 °C PBS without (FAF)BSA. To each 200 mL of liquid,
scintillation counting fluid is added. The plates are heat-sealed
and shaken for 30 min at room temperature prior to measuring
the amount of radioactivity in each well on a Packard Top-
Count instrument.
In Vitro Assay of Compounds That Inhibit Uptake of
[
14C]Alanine. The alanine uptake assay is performed in a
fashion identical to that of the taurocholate assay, with the
exception that labeled alanine is substituted for the labeled
taurocholate.
Bioavailability Assay. Compounds were initially screened
for absorption in the rat by comparing iv and oral dosing to
calculate % bioavailability. Compounds were solubilized in iv
compatible vehicles (saline with pH adjusted to attain clear
solutions) and dosed at 2-10 mg per kg of body weight.
Sprague Dawley rats were anesthetized with inactin, and the
femoral artery and vein were cannulated. Compounds were
administered via the femoral vein, and peripheral blood
samples were taken via the arterial cannulae at selected time
periods up to 8 h. Plasma concentrations of the compounds
were determined by quantitative HPLC with UV detection.
The lower limit of sensitivities typically ranged from 10 to 100
ng/mL. Clearance and AUC (area under the curve) for the
compounds were determined. A second set of rats (n ) 3 per
compound and route) was prepared as described above. An oral
dose at 10-20 mpk was administered, and peripheral plasma
samples were again administered via the arterial cannulae.
To rank compounds that were typically <1% and to dif-
ferentiate between these low levels of systemic bioavailability,
an alternative method was used. Instead of calculating sys-
tematic bioavailabilty using plasma levels from peripheral
blood, samples of portal plasma were also obtained. To
accomplish this, rats receiving the oral dose were prepared
with peripheral and portal access. Cannulae were inserted into
both the femoral artery for peripheral access as well as the
hepatic portal vein to obtain estimates of the total absorbed
drug without first-pass hepatic clearance. In addition, and to
add a final method to compare compounds that had low total
absorption, the common bile duct was cannulated and total
bile flow was collected for the duration of the experiment. The
AUC for portal plasma was then obtained to determine % drug
absorbed and total biliary excretion as % total dose was also
obtained. The fraction absorbed (F) is calculated by
Serum Lipid Measurements.21 Blood was collected at 14
and 28 days from nonfasted hamsters and put into serum
separator tubes. The blood cells were separated from the serum
by centrifugation at 2000g for 20 min, and the serum was
decanted. Total cholesterol was measured using a Cobas Mira
Classic clinical chemistry analyzer (Roche Diagnostic Systems,
Indianapolis, IN). This analyzer used the cholesterol oxidase
reaction to produce hydrogen peroxide, which was measured
colorimetrically. The cholesterol reagent (Roche Diagnostic
Systems) was reconstituted according to the package insert.
The reagent was calibrated using Roche calibrator serum.
Commercial bilevel quality control material was analyzed to
verify calibration and reagent performance (QC1, QC2; Bio-
Rad Laboratories, Irvine, CA). The cholesterol content of each
sample was measured at 500 nm and 37 °C in a calibrated
analyzer by mixing 3 mL of sample with 150 mL of cholesterol
reagent and 40 mL of water.
AUCpo(hepatic portal)
F )
AUCiv
Materials and Methods. Animal Handling, Dosing, and
Sample Collection. Male golden Syrian hamsters (126-147
g) were obtained from Charles Rivers Laboratories and were
single-housed in a constant temperature environment with
The unimate HDL direct cholesterol assay (DHDL, Roche
Diagnostics) was based on the absorbance of synthetic poly-
mers and polyanions onto the surface of lipoproteins. The

5844 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Table 8. HPLC Gradient
Tremont et al.
2-ethyl-2-(hydroxymethyl)hexanal, prepared according to the
procedure described in ref 22, while maintaining the reaction
temperature below 30 °C. The reaction mixture was stirred
at room temperature for 18 h, and the reaction was quenched
with dilute HCl and extracted with methylene chloride. The
methylene chloride extract was dried over MgSO4 and con-
centrated in vacuo to give 24.4 g of brown oil 98. 1H NMR
(CDCl3) δ 0.80 (t, J ) 7.5 Hz, 3H), 0.82 (t, J ) 7.2 Hz, 3H),
1.06-1.28 (m, 5H), 1.48-1.60 (m, 3H), 2.95 (s, 3H), 4.19 (ABq,
2H), 9.38 (s, 1H). HRMS (CI/M + NH4) calcd for C10H24O4SN:
254.1431; found, 254.1426.
Step 2. 2-((2-Benzoylphenylthio)methyl)-2-ethylhexa-
nal (99). A mixture of 31 g (0.144 mol) of 2-mercaptoben-
zophenone, 24.4 g (0.1 mol) of 2-ethyl-2-(mesyloxymethyl)-
hexanal (98), 14.8 g (0.146 mol) of triethylamine, and 80 mL
of 2-methoxyethyl ether was held at reflux for 24 h. The
reaction mixture was poured into 3 N HCl and extracted with
300 mL of methylene chloride. The methylene chloride layer
was washed with 300 mL of 10% NaOH, dried over MgSO4,
and concentrated in vacuo to remove 2-methoxyethyl ether.
The residue was purified by HPLC (10% EtOAc/hexane) to give
20.5 g (58%) of 99 as an oil.
Step 3. 2-((2-Benzoylphenylsulfonyl)methyl)-2-ethyl-
hexanal (100). To a solution of 9.0 g (0.025 mol) of compound
99 in 100 mL of methylene chloride was added 14.6 g (0.025
mol) of 50-60% MCPBA portionwise. The reaction mixture
was stirred at room temperature for 64 h and then was stirred
with 200 mL of 1 M potassium carbonate and filtered through
Celite. The methylene chloride layer was washed twice with
300 mL of 1 M potassium carbonate, once with 10% sodium
hydroxide, and once with brine. The insoluble solid formed
during washing was removed by filtration through Celite. The
methylene chloride solution was dried and concentrated in
vacuo to give 9.2 g (95%) of semisolid. A portion (2.6 g) of this
solid was purified by HPLC (10% ethyl acetate/hexane) to give
1.9 g of crystals of 100, mp 135-136 °C.
Step 4. 2-((2-Phenylmethylphenylsulfonyl)methyl)-2-
ethylhexanal (101). A solution of 50 g (0.13 mol) of crude
100 in 250 mL of methylene chloride was divided into two
portions and charged to two Fisher-Porter bottles. To each
bottle was charged 125 mL of methanol and 5 g of 10% Pd/C.
The bottles were pressurized with 70 psi of hydrogen, and the
reaction mixture was stirred at room temperature for 7 h
before being charged with an additional 5 g of 10% Pd/C. The
reaction mixture was again hydrogenated with 70 psi of
hydrogen for 7 h. This procedure was repeated one more time,
but only 1 g of Pd/C was charged to the reaction mxture. The
combined reaction mixture was filtered and concentrated in
vacuo to give 46.8 g of 101 as a brown oil.
time (min)
solution A (%)
solution B (%)
0
5
30
40
50
50
0
50
50
100
100
0
combined action of polymers, polyanions, and detergent solu-
bilizes cholesterol from HDL but transforms LDL, VLDL, and
chylomicrons into detergent-resistant forms. Solubilized cho-
lesterol was oxidized by the sequential enzymatic action of
cholesterol esterase and cholesterol oxidase. The H2O2 pro-
duced in this reaction was reacted with chromogens to form a
colored dye. The increase in absorbance at 550 nm was directly
proportional to the HDL cholesterol concentration of the
sample. The test was performed using a Cobas Mira Classic
clinical chemistry analyzer (Roche Diagnostic Systems). The
DHDL reagent (Roche Diagnostic Systems) was reconstituted
according to the package insert. The reagent was calibrated
using Roche HDL direct calibrator. Commercial bilevel quality
control material was analyzed to verify calibration and reagent
performance (Liquichek lipids control, levels 1 and 2, Bio-Rad
Laboratories, Irvine, CA). A 2.4 mL aliquot of serum was
analyzed in the presence of 240 mL of reagent 1, 80 mL of
reagent 2, and 5 mL of water at 37 °C and 550 nm wavelength.
Triglycerides were hydrolyzed by lipoprotein lipase to
glycerol and fatty acids. Glycerol was then phosphorylated to
glycerol 3-phosphate by adenosine 5-triphosphate (ATP) in a
reaction catalyzed by glycerol kinase (GK). The oxidation of
glycerol 3-phosphate was catalyzed by glycerolphosphate oxi-
dase (GPO) to form dihydroxyacetone phosphate and hydrogen
peroxide. The hydrogen peroxide reacted with 4-cholorophenol
and 4-aminophenazone in the presence of peroxidase to form
a quinoneimine complex, which was read at 490-550 nm. The
increase in absorbance was proportional to the concentration
of triglycerides in the sample. The test was performed using
a Cobas Mira Classic clinical chemistry analyzer (Roche
Diagnostic Systems). The triglyceride reagent (Roche Diag-
nostic Systems) was reconstituted according to the package
insert. The triglyceride content of each sample was measured
at 500 nm at 37 °C in a calibrated analyzer by mixing 4 mL of
sample with 300 mL of triglyceride reagent and 40 mL of
water.
General Comments. Chemicals were obtained from Ald-
rich Chemical Co. and were used without further purification.
1H and 13C NMR spectra were recorded on a Varian 300
spectrometer at 300 and 75 MHz, respectively. The 1H chemi-
cal shifts are reported in ppm downfield from Me4Si. The 13C
chemical shifts are reported in ppm relative to the center line
of CDCl3 (77.0 ppm). Melting points were recorded on a Buchi
510 melting point apparatus and are uncorrected. High-
resolution mass spectra were determined by Monsanto Ana-
lytical Sciences Center, and microanalyses were performed by
Atlantic Microlab Inc. Compound purity was checked by
microanalysis. For this manuscript only racemates were
evaluated. The following article, part 2 of this series, will
describe our research with single enantiomers. Biological data
were reported only for compound samples with greater than
95% purity as determined by elemental analysis, HPLC, and/
Step 5. (()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahy-
dro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (34a) and
(()-(3S,4R,5S)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phen-
yl-1-benzothiepin-4-ol 1,1-Dioxide (35a). To a solution of
27.3 g (73.4 mmol) of 101 in 300 mL of anhydrous THF cooled
to 2 °C with an ice bath was added 9.7 g (73.4 mmol) of 95%
potassium tert-butoxide. The reaction mixture was stirred for
20 min, the reaction was quenched with 300 mL of 10% HCl,
and the sample was extracted with methylene chloride. The
methylene chloride layer was dried over magnesium sulfate
and concentrated in vacuo to give 24.7 g of yellow oil.
Purification by HPLC (ethyl acetate/hexane) yielded 9.4 g of
recovered 101 in the first fraction, 5.5 g (20%) of 35a in the
second fraction, and 6.5 g (24%) of 34a in the third fraction.
1
or H NMR. HPLC data were obtained on a Spectra Physics
8800 chromatograph using a Beckman Ultrasphere C18 250
mm × 4.6 mm column. HPLC conditions were as follows:
detector wavelength ) 254 nm; sample size ) 10 µL; flow rate
) 1.0 mL/min; mobile phase ) (A) 0.1% aqueous trifluoroacetic
acid, (B) acetonitrile.
34a: solid; mp 160-161 °C. 1H NMR (CDCl3) δ 0.86 (t, J )
7.6 Hz, 3 H), 0.92 (t, J ) 6.8 Hz, 3 H), 1.0-1.56 (m, 6 H), 1.65
(dt, J ) 14.0, 4.8 Hz, 1 H), 2.24 (dt, J ) 13.6, 3.6 Hz, 1 H),
3.12 (ABq, J ) 16.8 Hz, 2 H), 4.21 (d, J ) 6.7 Hz, 1 H), 5.60
(s, 1 H), 6.79 (d, J ) 7.5, 1 H), 7.30-7.54 (m, 7 H), 8.14 (d, J
) 7.5 Hz, 1 H). MS (ES/M + H): 373. HRMS (CI/M + H) calcd
for C22H29O3S: 373.1837; found, 373.1839.
The HPLC gradient is shown in Table 8.
(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phen-
yl-1-benzothiepin-4-ol 1,1-Dioxide (34a) and (()-(3S,4S,5S)-
3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phenyl-1-benzothi-
epin-4-ol 1,1-Dioxide (35a). Step 1. (()-2-Ethyl-2-(methane-
sulfonyloxymethyl)hexanal (98). To a cold (10 °C) solution
of 12.6 g (0.11 mole) of methanesulfonyl chloride and 10.3 g
(0.13 mol) of triethylamine was added dropwise 15.8 g of (()-
1
35a: solid; mp 179-181 °C. H NMR (CDCl3) δ 0.85-0.96
(m, 6 H), 1.15-1.50 (m, 6 H), 1.76 (dq, J ) 13.7, 7.5 Hz, 1 H),
2.32 (dq, J ) 13.7, 7.5 Hz, 1 H), 3.13 (ABq, J ) 15.1 Hz, 2 H),
4.22 (d, J ) 6.4 Hz, 1 H), 5.60 (s, 1 H), 6.80 (m, 1 H), 7.30-

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5845
7.50 (m, 7 H), 8.10 (m, 1H). Anal. Calcd for C22H29O3S: C,
70.91; H, 7.58; S, 8.61. Found: C, 70.69; H, 7.58; S, 8.55.
C (0.65 g) as a crystalline product. Trituration with hexanes
gave pure B (0.142 g) as a white crystalline solid. The hexane
filtrate was concentrated in vacuo to give a mixture of 30% B
and 70% C.
Alternative Synthesis of 101. Step 1. 2-Mercapto-
diphenylmethane (102). To a 500 mL flask was charged 16
g (0.4 mol) of 60% sodium hydride oil dispersion. The sodium
hydride was washed twice with 50 mL of hexane. To the
reaction flask was charged 100 mL of DMF. To this mixture
was added a solution of 55.2 g (0.3 mol) of 2-hydroxydiphe-
nylmethane in 2000 mL of DMF in 1 h while the temperature
was maintained below 30 °C by an ice-water bath. After
complete addition of the reagent, the mixture was stirred at
room temperature for 30 min, then cooled with an ice bath.
To the reaction mixture was added 49.4 g (0.4 mol) of
dimethylthiocarbamoyl chloride at once. The ice bath was
removed, and the reaction mixture was stirred at room
temperature for 18 h before being poured into 300 mL of water.
The organic was extracted into 500 mL of toluene. The toluene
layer was washed successively with 10% sodium hydroxide and
brine and was concentrated in vacuo to give 78.6 g of a yellow
oil, which was 95% pure dimethyl O-2-benzylphenylthiocar-
bamate. This oil was heated at 280-300 °C in a Kugelrohhr
pot under house vacuum for 30 min. The residue was Kugel-
rohr distilled at 1 Torr (180-280 °C). The distillate (56.3 g)
was crystallized from methanol to give 37.3 g (46%) of the
rearranged product dimethyl S-2-benzylphenylthiocarbamate
as a yellow solid. A mixture of 57 g (0.21 mol) of this yellow
solid, 30 g of potassium hydroxide, and 150 mL of methanol
was stirred overnight and then was concentrated in vacuo. The
residue was diluted with 200 mL of water and extracted with
ether. The aqueous layer was made acidic with concentrated
HCl. The oily suspension was extracted into ether. The ether
extract was dried over magnesium sulfate and concentrated
in vacuo. The residue was crystallized from hexane to give 37.1
g (88%) of compound 102 as a yellow solid.
B: 1H NMR (CDCl3) δ 0.67 (m, 3H), 0.85-1.15 (m, 6H), 1.60
(m, 1H), 1.75 (m, 1H), 3.29 (s, 1H), 3.42 (qAB, JAB ) 12.0 Hz,
∆ν ) 24.0 Hz, 2H), 7.24 (m, 5H), 7.57 (d, J ) 7.2 Hz, 1H), 7.59
(s, J ) 7.2 Hz, 1H), 7.66 (d, 1H), 8.02 (d, 1H). MS (EI): 370.
C: 1H NMR (CDCl3) δ 0.67 (m, 3H), 0.90-1.00 (m, 3H),
1.25-1.50 (m, 4), 1.74 (m, 1H), 3.30 (s, 1H), 3.40 (qAB, JAB
)
12.0 Hz, ∆ν ) 24.0 Hz, 2H), 7.24 (m, 5H), 7.57 (d, J ) 7.2 Hz,
1H), 7.59 (s, J ) 7.2 Hz, 1H), 7.66 (d, 1H), 8.02 (d, 1H). MS
(EI): 370.
Step 3. (()-(3S,4S,5R)-3-butyl-3-ethyl-2,3,4,5-tetrahy-
dro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (34b) and
(()-(3S,4S,5S)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phen-
yl-1-benzothiepin-4-ol 1,1-Dioxide (35b). A mixture of 30%
B and 70% C (0.15 g, 0.40 mmol) was dissolved in methanol
(15 mL) in a 3 oz. Fischer-Porter bottle. A 10% Pd/C (0.10 g)
sample was added to the bottle, and it was pressurized to 70
psig of H2. After 5 h, the reaction mixture was filtered and
concentrated in vacuo. Purification by preparative HPLC
eluting with EtOAc/hexanes gave isolation of 34b and 35b.
1
34b: H NMR (CDCl3) δ 0.83 (t, J ) 7.8 Hz, 3H), 0.88 (t, J
) 7.8 Hz, 3H), 1.29 (m, 6H), 1.64 (m, 2H), 1.75 (d, J ) 4.7 Hz,
1H), 3.38 (ABq, J ) 15.5 Hz, 2H), 4.42 (dd, J ) 9.7, 3.9 Hz,
1H), 5.03 (d, J ) 9.7 Hz, 1H), 7.10 (d, J ) 7.8 Hz, 1H), 7.37-
7.43 (m, 7H), 8.08 (d, J ) 7.8 Hz, 1H). MS (EI): 372.
1
35b: H NMR (CDCl3) δ 0.78 (t, J ) 6.6 Hz, 3H), 0.91 (t, J
) 6.6 Hz, 3H), 1.06-1.35 (m, 6H), 1.70 (m, 2H), 1.80 (d, J )
3.9 Hz, 1H), 3.40 (ABq, J ) 15.5 Hz, 2H), 4.42 (dd, J ) 9.3,
3.9 Hz, 1H), 5.04 (d, J ) 9.3 Hz, 1H), 7.10 (d, J ) 6.9 Hz, 1H),
7.33-7.44 (m, 7H), 8.05 (d, J ) 6.9 Hz, 1H). MS (EI): 372.
Step 2. (()-2-((2-Phenylmethylphenylthio)methyl)-2-
ethylhexanal (103). A mixture of 60 g (0.3 mol) of compound
102, 70 g (0.3 mol) of 98, 32.4 g (0.32 mol) of triethylamine,
and 120 mL of 2-methoxyethyl ether was held at reflux for 6
h and concentrated in vacuo. The residue was triturated with
500 mL of water and 30 mL of concentrated HCl. The mixture
was extracted into 400 mL of ether. The ether layer was
washed successively with brine, 10% sodium hydroxide, and
brine and was dried over magnesium sulfate and concentrated
in vacuo. The residue (98.3 g) was purified by HPLC with
2-5% ethyl acetate/hexane as eluent to give 103 as a yellow
syrup.
Step 3. (()-2-((2-Phenylmethylphenylsulfonyl)methyl)-
2-ethylhexanal (101). To a solution of 72.8 g (0.21 mol) of
103 from step 2 in 1 L of CH2Cl2 cooled to 10 °C was added
132 g of 50-60% MCPBA in 40 min. The reaction mixture was
stirred for 2 h. An additional 13 g of 50-60% MCPBA was
added to the reaction mixture. The reaction mixture was
stirred for 2 h and filtered through Celite. The CH2Cl2 solution
was washed twice with 1 L of 1 M potassium carbonate and
then with 1 L of brine. The CH2Cl2 layer was dried (MgSO4)
and concentrated to give 76 g of compound 101 as a thick oil.
(()-(3S,4S,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phen-
yl-1-benzothiepin-4-ol1,1-Dioxide(34b)and(()-(3S,4R,5S)-
3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-phenyl-1-benzothi-
epin-4-ol 1,1-Dioxide (35b). Step 1. (()-3-Butyl-3-ethyl-
2,3-dihydro-5-phenylbenzothiepine (A). Compound A was
prepared according to the procedure described in ref 11.
Step 2. (()-(3S,4S,5R)-3-Butyl-3-ethyl-4,5-oxo-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepine 1,1-Dioxide (B) and
(()-(3S,4R,5S)-3-Butyl-3-ethyl-4,5-oxo-2,3,4,5-tetrahydro-
5-phenyl-1-benzothiepine 1,1-Dioxide (C). To a solution
of 1.3 g (4.03 mmol) sulfide olefin A from step 1 in 25 mL of
CH2Cl2 was added 5.0 g of 50-60% MCPBA portionwise. After
being stirred overnight, the reaction mixture was heated to
reflux for 3 h. The mixture was filtered, and the filtrate was
washed with 10% potassium carbonate (3 × 50 mL) and brine
(50 mL). The CH2Cl2 layer was dried (MgSO4), filtered, and
concentrated to 1.37 g of yellow oil. Purification by preparative
HPLC eluting with EtOAc/hexanes gave a mixture of B and
(()-(3S,4R,5R)-3-Butyl-3-ethyl-5-(3-methoxyphenyl)-
2,3,4,5-tetrahydro-7-bromobenzothiepin-4-ol 1,1-Dioxide
(27). Alkylation of 4-bromophenol with 3-methoxybenzyl chlo-
ride according to the procedure described in ref 23 gave
4-bromo-2-(3′-methoxybenzyl)phenol: mp 95.4-96.8 °C. 1H
NMR (CD3OD) δ 3.64 (s, 3H), 3.85 (s, 2H), 6.67-6.78 (m, 4H),
7.05-7.17 (m, 3H). MS (EI/M+): 293. This material was
converted to compound 27, mp 102-106 °C, and its (()-
(3S,4S,5S)-diastereomer, mp 97-101.5 °C, by a procedure
similar to the one for 34a described from synthesis of 102
above. The diastereomers were separated by HPLC (ethyl
acetate/hexane). 1H NMR (CDCl3) δ 0.9-1.1 (m, 6 H), 1.1-1.8
(m, 7 H), 2.28 (br t, J ) 13.8 Hz, 1 H), 3.21 (ABq, J ) 15.7 Hz,
2 H), 3.92 (s, 3H), 4.28 (d, J ) 7 Hz, 1 H), 5.58 (s, 1 H), 7.00
(d, J ) 8 Hz, 1 H), 7.10 (s, 2 H), 7.16 (d, J ) 8 Hz, 1 H), 7.46
(t, J ) 8 Hz, 1 H), 7.62 (d, J ) 8 Hz, 1 H), 8.04 (d, J ) 8 Hz,
1 H). MS (CI/M + H): 481. Anal. Calcd for C23H29BrO4S: C,
57.38; H, 6.07; S, 6.60. Found: C, 57.11; H, 6.01; S, 6.53.
(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-5-(3-
methoxyphenyl)-7-phenyl-1-benzothiepin-4-ol 1,1-Diox-
ide (28). To a mixture of 0.6 g (1.25 mmol) of 27 of 60 mL of
toluene and 25 mL of ethanol was added 0.18 g (1.55 mmol)
of phenylboronic acid followed by a solution of 0.38 g (2.7 mmol)
of potassium carbonate in 1 mL of water. To the above mixture
was added 0.1 g (0.087 mmol) of tetrakis(triphenylphosphinyl)-
palladium. The reaction mixture was stirred and heated at
90 °C for 1 h. To the above mixture was add another equal
portion of phenylboronic acid, potassium carbonate solution,
and tetrakis(triphenylphosphinyl)palladium. The mixture was
held at 90 °C for 4 h and cooled. The reaction mixture was
diluted with 100 mL of water and extracted with methylene
chloride (4 × 50 mL). The methylene chloride extracts were
washed with 10% sodium hydroxide (2 × 50 mL), dried over
MgSO4, and concentrated in a vacuum to give 0.6 g of solid.
Purification with HPLC on silica gel (4:1 hexane/EtOAc) gave
1
0.4 g of a white solid 28, mp 212-214 °C. H NMR (CDCl3) δ
0.9-1.1 (m, 6 H), 1.1-1.9 (m, 7 H), 2.28 (br t, J ) 13.8 Hz, 1
H), 3.14 (ABq, J ) 15.7 Hz, 2 H), 3.92 (s, 3H), 4.16 (d, J ) 7
Hz, 1 H), 5.62 (s, 1 H), 7.00 (d, J ) 8 Hz, 1 H), 7.16 (d, J ) 8
Hz, 1 H), 7.24 (d, J ) 8 Hz, 1 H), 7.4-7.5 (m, 7H), 7.66 (d, J

5846 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
) 8 Hz, 1 H), 8.25 (d, J ) 8 Hz, 1 H). MS (CI/M + H): 479.
HRMS (CI/M + H) calcd for C29H35O4S: 479.2262; found,
479.2256.
7.6 Hz, 2H), 7.44 (d, J ) 7.2 Hz, 1H), 7.60 (d, J ) 2.8 Hz, 1H).
MS (FAB/M + Li): 535. HRMS (ESI/M + H) calcd for
C32H49O4S: 529.3352; found, 529.3305. Anal. Calcd for
C32H48O4S: C, 72.68; H, 9.15; S, 6.06. Found: C, 71.27; H, 8.84;
S, 5.35.
(()-(3S,4R,5R)-7-Methoxy-3-butyl-3-ethyl-2,3,4,5-tet-
rahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (36).
Step 1. 2-Mercapto-5-methoxybenzophenone (104).
Amounts of 66.2 g of 4-methoxythiophenol, 360 mL of 2.5 N
n-butyllithium, and 105 g of tetramethylethylenediamine were
combined with 400 mL of cyclohexane. The mixture was cooled
to 0 °C, and a solution of 66.7 g of benzonitrile in 200 mL of
cyclohexane was added. The reaction mixture was stirred at
0 °C for 3 and 1 h at room temperature. A solution of 600 mL
of 6 N sulfuric acid was added, and the mixture was heated to
60 °C for 2 h. The organic phase was separated, and the
aqueous phase was neutralized with 10% aqueous NaOH.
Extraction with ethyl ether gave 73.2 g of brown oil, which
was Kugelrohr distilled to remove 4-methoxythiophenol and
gave 43.86 g of crude 104 as a brown oil.
Step2. 2-((2-Benzoyl-4-methoxyphenylthio)methyl)-2-
ethylhexanal (105). Reaction of 10 g (0.04 mol) of crude 104
with 4.8 g (0.02 mol) of mesylate 98 and 3.2 mL (0.23 mol) of
triethylamine in 50 mL of diglyme according to the procedure
for the preparation of 99 gave 10.5 g of crude product, which
was purified by HPLC (5% ethyl acetate/hexane) to give 1.7 g
(22%) of 105.
Step 3. 2-((2-Benzoyl-4-methoxyphenylsulfonyl)methyl)-
2-ethylhexanal (106). A solution of 1.2 g (3.1 mmol) of 105
in 25 mL of methylene chloride was reacted with 2.0 g (6.2
mmol) of 50-60% MCPBA according to the procedure for 100,
giving 1.16 g (90%) of 106 as a yellow oil.
Step 4. 2-((2-Phenylmethyl-4-methoxyphenylsulfonyl)-
methyl)-2-ethylhexanal (107). Hydrogenation of 1.1 g of 106,
according to the procedure for intermediate 101 from step 4
of 34a, gave 107 as a yellow oil (1.1 g).
Step 5. (()-(3S,4R,5R)-7-Methoxy-3-butyl-3-ethyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (36).
A solution of 1.1 g of 107, 0.36 g of potassium tert-butoxide,
and 25 mL of anhydrous THF was held at reflux for 2 h and
worked up as for compound 34a to give 1.07 g of a crude
product, which was purified by HPLC to give 40 mg of the
(()-(3S,4S,5S)-diastereomer as crystals, mp 153-154 °C and
90 mg (8%) of the (()-(3S,4R,5R)-diastereomer of 36, mp 136-
140 °C. 1H NMR (CDCl3) δ 0.9-1.1 (m, 6 H), 1.1-1.8 (m, 7 H),
2.22 (br t, J ) 13.8 Hz, 1 H), 3.10 (ABq, J ) 15.7 Hz, 2 H),
3.66 (s, 3H), 4.20 (s, 1 H), 5.52 (s, 1 H), 6.30 (d, J ) 2 Hz, 1 H),
6.81 (dd, J ) 8, J ) 2 Hz, 1 H), 7.3-7.5 (m, 5 H), 8.05 (d, J )
8 Hz, 1 H). Anal. Calcd for C23H30O4S: C, 67.94; H, 7.46; S,
6.98. Found: C, 68.08; H, 7.61; S, 8.01.
(()-(3S,4R,5R)-8-Hydroxy-3-butyl-3-ethyl-2,3,4,5-tet-
rahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (37).
Reaction of 61 with BBr3 according to the procedure described
in ref 24 gave compound 37. The resulting residue was
concentrated in vacuo, and the desired product was collected
as a white solid, mp 226.5-228.5 °C. HPLC: tR ) 17.7 min.
1H NMR (CDCl3) δ 0.86 (t, J ) 7.6 Hz, 3H), 0.92 (t, J ) 7.6
Hz, 3H), 1.14 (m, 1H), 1.26-1.64 (m, 6H), 2.22 (br t, J ) 10.0
Hz, 1H), 3.12 (ABq, 2H), 4.17 (s, 1H), 5.26 (br s, 1H), 5.51 (s,
1H), 6.65 (d, J ) 8.8 Hz, 1H), 6.86 (dd, J ) 8.8, 2.8 Hz, 1H),
7.35 (d, J ) 6.8 Hz, 1H), 7.42 (t, J ) 7.6 Hz, 2H), 7.48 (d, J )
7.2 Hz, 2H), 7.61 (d, J ) 2.8 Hz, 1H). MS (CI/M + Li): 395.
HRMS (CI/M + H) calcd for C22H29O4S: 389.1787; found,
389.1778. Anal. Calcd for C22H28O4S: C, 68.01; H, 7.26.
Found: C, 67.01; H, 7.28.
(()-(3S,4R,5R)-7-Amino-3-butyl-3-ethyl-2,3,4,5-tetrahy-
dro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (39). Com-
pound 42 (1.91 g, 3.66 mmol) and CH2Cl2 (5.0 mL) were
combined in a 100 mL round-bottom flask. The flask was
purged with N2 and fitted with a magnetic stirrer. Ethanethiol
(6.80 g, 109 mmol) and boron trifluoride etherate (5.20 g, 36.6
mmol) were added. After 96 h, water (50 mL) was added. After
10 min, the mixture was extracted with CHCl3 (3×). The
combined organic layers were dried (MgSO4), filtered, and
concentrated in vacuo to give compound 39 as a white solid
1
(1.50 g). H NMR (CD2Cl2) δ 0.85 (t, J ) 7.5 Hz, 3H), 0.92 (t,
J ) 7.2 Hz, 3H), 1.05-1.20 (m, 1H), 1.25-1.68 (m, 6H), 2.19
(m, 1H), 3.07 (ABq, 2H), 4.15 (s, 1H), 5.49 (s, 1H), 5.99 (d, J )
1.8 Hz, 1H), 6.56 (dd, J ) 8.7, 2.4 Hz, 1H), 7.34-7.50 (m, 6H),
7.81 (d, J ) 8.4 Hz, 1H). MS (CI/M + H): 388. HRMS (M +
H) calcd for C22H30O3SN: 388.1946; found, 388.1963. Anal.
Calcd for C22H29O3SN: C, 68.19; H, 7.54; N, 3.61; S, 8.27.
Found: C, 67.14; H, 7.44; N, 3.48; S, 8.12.
(()-(3S,4R,5R)-7-Hydroxyamino-3-butyl-3-ethyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (40).
Compound 41 (0.050 g, 0.099 mmol) and 4 N HCl in dioxane
(0.50 mL/2.0 mmol) were combined in a 4 mL vial. The vial
was purged with N2 and fitted with a magnetic stirrer. After
30 min, a solution of NaOAc (0.27 g, 2.0 mmol) in water (1.0
mL) and ethyl ether (1.0 mL) was added to the reaction vial.
After an additional 10 min, additional ethyl ether was added,
and the mixture was washed with water (6×). The organic
layer was dried (Na2SO4), filtered, and concentrated in vacuo
to give compound 40 (0.0397 g, 99%). 1H NMR (CDCl3) δ 0.83
(t, J ) 7.2 Hz, 3H), 0.90 (t, J ) 7.5 Hz, 3H), 1.05-1.20 (m,
1H), 1.22-1.65 (m, 6H), 1.18 (m, 1H), 3.08 (br q, 2H), 4.13 (s,
1H), 5.46 (s, 1H), 6.12 (s, 1H), 6.79 (s, 1H), 7.30-7.43 (m, 6H),
7.86 (s, 1H). Anal. Calcd for C22H29NO4S: C, 65.48; H, 7.24;
N, 3.47: S, 7.95. Found: C, 64.98; H, 7.15; N, 3.25; S, 7.65.
(()-[(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-hy-
droxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]hydrox-
ycarbamic Acid 1,1-Dimethylethyl Ester (41). Step 1.
2-((2-Phenylmethyl-4-hydroxyaminophenylsulfonyl)m-
ethyl)-2-ethylhexanal (113). Intermediate 110 (from step 3
of the synthesis of 42) (0.50 g, 1.2 mmol), ethanol (7.50 mL),
and 10% Pd/C (0.050 g) were combined in a 25 mL round-
bottom flask. The flask was purged with N2. Hydrogen gas was
bubbled through the reaction mixture for 30 min. The reaction
mixture was purged with N2, filtered through Celite washing
with ethanol and then ethyl ether, and concentrated in vacuo.
The resulting residue was dissolved in ethyl ether, dried (Na2-
SO4), filtered, and concentrated in vacuo to give compound 113.
1H NMR (CDCl3) δ 0.76 (t, J ) 7.5 Hz, 3H), 0.87 (t, J ) 7.2
Hz, 3H), 0.94-1.15 (m, 2H), 1.16-1.29 (m, 2H), 1.49-2.10 (m,
4H), 2.97 (s, 2H), 4.41 (s, 2H), 6.77 (s, 1H), 6.92 (d, J ) 9.0
Hz, 1H), 7.16-7.34 (m, 6H), 7.87 (d, J ) 8.4 Hz, 1H), 9.35 (s,
1H).
Step 2. 2-((2-Phenylmethyl-4-hydroxycarboxybuty-
laminophenylsulfonyl)methyl)-2-ethylhexanal (114). Com-
pound 113 (8.50 g, 0.0211 mol, step 1) and di-tert-butyl
dicarbonate (4.80 g, 0.0220 mol) were combined in a 100 mL
round-bottom flask. The flask was purged with N2 and fitted
with a magnetic stirrer. THF (25.0 mL) was added, and the
mixture was heated to 60 °C. After 6 h, the reaction mixture
was cooled to room temperature, and additional di-tert-butyl
dicarbonate (1.43 g, 0.0066 mol) was added. The mixture was
heated to 60 °C. After 16 h, the reaction mixture was cooled
to room temperature and concentrated in vacuo. The residue
was dissolved in EtOAc (40.0 mL) and washed with aqueous
1% HCl (2 × 25 mL) and aqueous 5% NaHCO3 (2 × 25 mL).
The organic layer was dried (MgSO4), filtered, and concen-
trated in vacuo. Purification by flash column chromatography
on silica gel eluting with 10% EtOAc/hexanes and concentrated
(()-(3S,4R,5R)-8-(1-Decyloxy)-3-butyl-3-ethyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (38).
Compound 37 was reacted with K2CO3 and 1-iododecane,
according to the procedure described in ref 25 to give compound
38. Purification via preparative HPLC (eluting with EtOAc/
hexane) gave the isolated product as a colorless oil, 38.
HPLC: tR ) 31.0 min. 1H NMR (CDCl3) δ 0.82-0.91 (m, 9H),
1.14-1.32 (m, 19H), 1.35-1.49 (m, 3H), 1.66-1.74 (m, 3H),
2.27 (m, 1H), 3.08 (ABq, 2H), 3.92 (t, J ) 6.4 Hz, 2H), 4.14 (d,
J ) 6.8 Hz, 1H), 5.47 (s, 1H), 6.63 (d, J ) 8.8 Hz, 1H), 6.85
(dd, J ) 8.4, 2.4 Hz, 1H), 7.30 (t, J ) 7.2 Hz, 1H), 7.37 (t, J )

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5847
in vacuo gave 114 (4.11 g, 39%). 1H NMR (CDCl3) δ 0.76 (t, J
) 7.2 Hz, 3H), 0.87 (t, J ) 7.2 Hz, 3H), 0.93-1.17 (m, 2H),
1.20-1.28 (m, 2H), 1.52 (s, 9H), 1.59-1.98 (m, 4H), 2.95 (s,
2H), 4.51 (s, 2H), 7.13-7.31 (m, 5H), 7.40 (s, 1H), 7.56 (d, J )
8.7 Hz, 1H), 8.01 (d, J ) 8.7 Hz, 1H), 9.35 (s, 1H). MS (FAB/M
- BOC): 401.
Step 3. [(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahy-
dro-4-hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]-
hydroxycarbamic Acid, 1,1-Dimethylethyl Ester (41).
Compound 114 (1.00 g, 1.98 mmol) was reacted with KOtBu
by a procedure similar to that used for compound 34a, step 5.
Purification by preparative HPLC eluting with 20% EtOAc/
hexanes and concentration in vacuo gave separation of 41
Step 3. 2-((2-Phenylmethyl-4-nitrophenylsulfonyl)m-
ethyl)-2-ethylhexanal (110). Compound 109 (52.30 g, 0.136
mol) and CHCl3 (1000 mL) were combined in a 2 L round-
bottom flask. The flask was purged with N2, fitted with a
magnetic stirrer, and cooled with an ice bath. mCPBA (72.20
g, 0.2719 mol) was added slowly, and the reaction mixture was
allowed to warm to room temperature. After 18 h, the reaction
mixture was cooled with an ice bath and filtered. The filtrate
was washed with 10% aqueous K2CO3 (900 mL), and the
aqueous washes were extracted with CHCl3 (3 × 250 mL). All
CHCl3 layers were combined, dried (MgSO4), filtered, and
concentrated in vacuo. Purification by precipitation from
ethanol and then crystallization from ethyl ether and concen-
trated in vacuo gave 110 (31.80 g, 56%). The mother liquor
was concentrated in vacuo and purified by preparative HPLC
eluting with 10% EtOAc/hexanes to give additional 110 (5.66
g, 10%). Combined yield: 66%. 1H NMR (CDCl3) δ 0.81 (t, J )
7.5 Hz, 3H), 0.88 (t, J ) 7.2 Hz, 3H), 0.96-1.20 (m, 2H), 1.22-
1.35 (m, 2H), 1.57-2.00 (m, 4H), 3.08 (ABq, 2H), 4.61 (s, 2H),
7.18 (d, J ) 6.9 Hz, 1H), 7.26-7.37 (m, 4H), 8.12 (d, J ) 1.5
Hz, 1H), 8.23 (d, J ) 2.1 Hz, 1H), 9.36 (s, 1H). MS (CI/M +
H): 418.
Step 4. 2-((2-Phenylmethyl-4-aminophenylsulfonyl)-
methyl)-2-ethylhexanal (111). Compound 110 (5.80 g, 0.0139
mol) and ethanol (180 mL) were combined in a Parr reactor.
Under a N2 atmosphere, 10% Pd/C (0.58 g) was added. The
reactor was subjected to 100 psig of H2 and heated to 55 °C.
After 17 h, the reactor was cooled to room temperature, and
the reaction mixture was filtered through Celite and concen-
trated until nearly dry to give a solution of the compound 111.
Note: The product will polymerize if concentrated to dryness.
1H NMR (CDCl3) δ 0.82 (t, J ) 7.5 Hz, 3H), 0.92 (t, J ) 7.2
Hz, 3H), 1.00-1.20 (m, 2H), 1.25-1.35 (m, 2H), 1.55-1.95 (m,
4H), 3.04 (ABq, 2H), 4.45 (s, 2H), 6.48 (d, J ) 1.8 Hz, 1H),
6.63 (dd, J ) 8.7, 2.1 Hz, 1H), 6.95-7.60 (m, 5H), 7.85 (d, J )
8.7 Hz, 1H), 9.40 (s, 1H). MS (CI/M + H): 388.
Step 5. 2-((2-Phenylmethyl-4-carboxybenzylaminophe-
nylsulfonyl)methyl)-2-ethylhexanal (112). A solution of
compound 111 (5.38 g, 0.0139 mol) in PhCH3 (17.4 g) and
additional PhCH3 (30 mL) were combined in a 500 mL round-
bottom flask. The reaction flask was purged with N2 and fitted
with a magnetic stirrer. Potassium carbonate (4.23 g, 0.0306
mol) and additional PhCH3 (20 mL) were added. A solution of
benzyl chloroformate (5.0 g, 0.29 mol) in PhCH3 (25 mL) was
added. After 17 h, the reaction flask was heated to 50 °C. After
an additional 1.5 h, more K2CO3 (0.96 g, 0.0069 mol) and
benzyl chloroformate (1.20 g, 0.007 03 mol) were added. After
4 h more, the reaction mixture was allowed to cool to room
temperature, and it was washed with water and 10% aqueous
K2CO3. The organic layer was dried (Na2SO4), filtered, and
concentrated in vacuo. Crystallization from PhCH3/hexanes
and concentrated in vacuo gave the desired product 112 plus
a minor amount of benzyl chloroformate (7.49 g). 1H NMR
(CDCl3) δ 0.77 (t, J ) 7.5 Hz, 3H), 0.87 (t, J ) 7.2 Hz, 3H),
0.95-1.20 (m, 2H), 1.22-1.30 (m, 2H), 1.52-1.93 (m, 4H), 3.00
(ABq, 2H), 4.47 (s, 2H), 5.19 (s, 2H), 6.92 (s, 1H), 7.17-7.40
(m, 10H), 7.54 (dd, J ) 9.0, 2.1 Hz, 1H), 7.96 (d, J ) 9.7 Hz,
1H).
(0.195 g, 20%) from its (3R,4R,5R)-diastereomer. HPLC: tR
)
21.2 min. 1H NMR (CDCl3) δ 0.85 (t, J ) 7.5 Hz, 3H), 0.92 (t,
J ) 7.2 Hz, 3H), 1.02-1.20 (m, 1H), 1.23-1.75 (m, 6H), 1.37
(s, 9H), 2.19 (m, 1H), 3.09 (ABq, 2H), 4.15 (s, 1H), 5.55 (s, 1H),
6.90 (s, 1H), 7.30-7.46 (m, 6H), 8.01 (d, J ) 8.7 Hz, 1H).
HRMS (FAB/M + H) calcd for C27H38NO6S: 504.2420; found,
504.2428.
(()-[(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-hy-
droxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]carbam-
ic Acid, Phenylmethyl Ester (42). Step 1. 2-Chloro-5-
nitrodiphenylmethane (108). 2-Chloro-5-nitrobenzophenone
(45.0 g, 0.172 mol) and CH2Cl2 (345.0 mL) were combined in
a 3 L round-bottom flask. The reaction flask was purged with
N2 and fitted with an addition funnel and magnetic stirrer.
The solution was cooled with ice bath. A solution of trifluo-
romethane sulfonic acid (51.62 g, 0.3440 mol) in CH2Cl2 (345.0
mL) was added dropwise to the reaction flask via the addition
funnel. After 5 min, a solution of triethylsilane (30.0 g, 0.258
mol) in CH2Cl2 (345.0 mL) was added dropwise to the reaction
flask via the addition funnel. After an additional 5 min, a
second solution of trifluoromethane sulfonic acid (51.62 g,
0.3440 mol) in CH2Cl2 (345.0 mL) was added dropwise to the
reaction flask via the addition funnel. After an additional 5
min, a solution of triethylsilane (30.0 g, 0.258 mol) in CH2Cl2
(345.0 mL) was added dropwise to the reaction flask via the
addition funnel. After the mixture was allowed to warm to
room temperature overnight, the reaction mixture was slowly
poured into cold saturated aqueous NaHCO3 (1600 mL). After
the mixture was stirred for 30 min, the layers were separated.
The aqueous layer was extracted with additional CH2Cl2 (2 ×
500 mL). All organic layers were combined, dried (MgSO4),
filtered, and concentrated in vacuo. Crystallization from
hexane gave the desired 108 (39.00 g, 92%). 1H NMR (CDCl3)
δ 4.19 (s, 2H), 7.20 (d, J ) 8.1 Hz, 1H), 7.23-7.35 (m, 4H),
7.54 (d, J ) 9.3 Hz, 1H), 8.03 (m, 2H). MS (CI/M + H): 248.
Step 2. 2-((2-Phenylmethyl-4-nitrophenylthio)methyl)-
2-ethylhexanal (109). Compound 108 (38.70 g, 0.156 mol),
lithium sulfide (7.18 g, 0.156 mol), and DMSO (1500 mL) were
combined in a dry 2 L round-bottom flask. The reaction flask
was purged with N2, equipped with magnetic stirrer and
condenser, and heated to 75 °C. After 15 h, the reaction
mixture was cooled to room temperature. A solution of
compound 98 (step 1 of the synthesis of 34a) (51.7 g, 0.219
mol) in DMSO (90 mL) was added, and the flask was heated
to 80 °C. After 48 h, the reaction mixture was cooled to room
temperature. Additional 98 (14.75 g, 0.0624 mol) in DMSO (20
mL) was added, and the flask was heated to 80 °C. After 63 h,
the reaction mixture was cooled to room temperature and
slowly added to 5% aqueous acetic acid (1900 mL). The
resulting mixture was divided into two portions, and each was
extracted with ethyl ether (4 × 700 mL). Organic extracts were
combined, dried (Na2SO4), filtered, and concentrated in vacuo.
The resulting oil was redissolved in ethyl ether, washed with
water, dried (Na2SO4), filtered, and concentrated in vacuo.
Purification by preparative HPLC eluting with 5% EtOAc/
hexanes and concentrated in vacuo gave compound 109 (40.22
Step 6. (()-[(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahy-
dro-4-hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]-
carbamic Acid, Phenylmethyl Ester (42) Compound 112
(7.49 g, 0.0144 mol) was reacted with KOtBu by a procedure
similar to that used for compound 34a, step 5. Purification by
preparative HPLC eluting with 15% EtOAc/hexanes and
concentrated in vacuo gave separation of compound 42 (2.05
1
g, 27%) from its (3R,4R,5R)-diastereomer. H NMR (CDCl3) δ
0.84 (t, J ) 7.2 Hz, 3H), 0.91 (t, J ) 6.9 Hz, 3H), 1.05-1.20
(m, 1H), 1.25-1.70 (m, 6H), 2.20 (m, 1H), 3.11 (ABq, 2H), 4.17
(s, 1H), 5.10 (s, 2H), 5.53 (s, 1H), 6.48 (d, J ) 2.10 Hz, 1H),
6.78 (s, 1H), 7.28-7.48 (m, 10H), 7.68 (dd, J ) 8.4, 1.2 Hz,
1H), 8.03 (d, J ) 8.7 Hz, 1H). HRMS (CI/M + H) calcd for
C30H36O5SN: 522.2314; found, 522.2287. Anal. Calcd for
C30H35O5SN: C, 69.07; H, 6.76; N, 2.68; S, 6.15. Found: C,
66.18; H, 6.60; N, 2.56; S, 5.83.
1
g, 67%). H NMR (CDCl3) δ 0.82 (t, J ) 10.5 Hz, 3H), 0.88 (t,
J ) 7.2 Hz, 3H), 1.03-1.23 (m, 2H), 1.24-1.35 (m, 2H), 1.62-
1.77 (m, 4H), 3.16 (s, 2H), 4.10 (s, 2H), 7.10-7.35 (m, 5H), 7.42
(d, J ) 8.7 Hz, 1H), 7.90 (d, J ) 2.4 Hz, 1H), 8.06 (dd, J ) 8.7,
2.4 Hz, 1H), 9.44 (s, 1H). %). MS (CI/M + H): 386.

5848 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
(()-[(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-
hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]-
carbamic Acid 1,1-Dimethylethyl Ester (43). Step 1.
2-((2-Phenylmethyl-4-carboxybenzylaminophenylsulfonyl)-
methyl)2-ethylhexanal (115). Compound 111 (from step 5
of the synthesis of 42) (0.47 g, 1.2 mmol) was reacted with
di-tert-butyl dicarbonate according to the procedure for inter-
mediate 114 (step 2 of the synthesis of 41). Purification by
preparative HPLC eluting with 15% EtOAc/hexanes and
magnetic stirrer. A solution of 4-methylmorpholine (0.029 g,
0.29 mmol) in CH2Cl2 (1.5 mL) was added, and the mixture
was cooled with an ice bath. A solution of acyl or mesyl chloride
(actual reagent used for each analogue given below) (0.29
mmol) in CH2Cl2 (1.5 mL) was added, and the reaction mixture
was allowed to warm to room temperature. After 17 h,
additional acyl or mesyl chloride (0.013 g, 0.097 mmol) and
4-methylmorpholine (0.010 g, 0.099 mmol) in CH2Cl2 (1.0 mL)
was added. After an additional 1 h, the reaction mixture was
diluted with ethyl ether and washed with water. The aqueous
wash was extracted with ethyl ether (2×). All organic layers
were combined, washed with aqueous 10% K2CO3 and brine,
dried (MgSO4), filtered, and concentrated in vacuo. Purification
by precipitation from ethyl ether/hexanes gave the desired
amide 46, 47, or 48.
1
concentrated in vacuo gave compound 115 (0.19 g, 32%). H
NMR (CDCl3) δ 0.76 (t, J ) 7.5 Hz, 3H), 0.86 (t, J ) 7.2 Hz,
3H), 0.95-1.16 (m, 2H), 1.20-1.29 (m, 2H), 1.49 (s, 9H), 1.55-
1.90 (m, 4H), 2.97 (ABq, 2H), 4.46 (s, 2H), 6.74 (s, 1H), 7.15-
7.32 (m, 6H), 7.54 (dd, J ) 8.7, 2.1 Hz, 1H), 7.94 (d, J ) 8.7
Hz, 1H), 9.33 (s, 1H). MS (FAB/M + Li): 494.2.
N-[(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-
hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]aceta-
mide (46). Reagent: acetyl chloride. Yield: 90%. Mp 148.0
Step 2. [(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahy-
dro-4-hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]-
carbamic Acid 1,1-Dimethylethyl Ester (43). Compound
115 (0.142 g, 0.290 mmol) was reacted with KOtBu by a
procedure similar to that used for compound 34a, step 5.
Purification by preparative HPLC eluting with 15% EtOAc/
hexanes and concentrated in vacuo gave separation of the
desired product 43 (0.0533 g, 38%) from its (3R,4R,5R)-
1
°C. HPLC: tR ) 12.9 min. H NMR (CDCl3) δ 0.85 (t, J ) 7.5
Hz, 3H), 0.91 (t, J ) 7.2 Hz, 3H), 1.05-1.15 (m, 1H), 1.23-
1.69 (m, 6H), 1.77 (d, J ) 6.6 Hz, 1H), 2.04 (s, 3H), 2.20 (m,
1H), 3.12 (ABq, 2H), 4.17 (d, J ) 6.9 Hz, 1H), 5.52 (s, 1H),
6.70 (s, 1H), 7.33-7.52 (m, 6H), 7.74 (d, J ) 8.1 Hz, 1H), 8.00
(d, J ) 8.7 Hz, 1H). HRMS (FAB/M + NH4) calcd for C24H35O4-
SN2: 447.2318; found, 447.2304. Anal. Calcd for C24H31O4SN:
C, 67.10; H, 7.27; N, 3.26; S, 7.46. Found: C, 65.51; H, 7.24;
N, 3.13; S, 7.11.
1
diastereomer, mp 140.0-150.0 °C. HPLC: tR ) 23.7 min. H
NMR (CDCl3) δ 0.84 (t, J ) 7.5 Hz, 3H), 0.90 (t, J ) 7.2 Hz,
3H), 1.05-1.20 (m, 1H), 1.22-1.65 (m, 6H), 1.44 (s, 9H), 2.19
(m, 1H), 3.09 (ABq, 2H), 4.15 (s, 1H), 5.52 (s, 1H), 6.31 (d, J )
1.8 Hz, 1H), 6.60 (s, 1H), 7.34-7.48 (m, 5H), 7.80 (d, J ) 8.7
Hz, 1H), 8.02 (d, J ) 8.7 Hz, 1H). MS (FAB/M + Li): 494.
HRMS (FAB/M + NH4) calcd for C27H41N2O5S: 505.2736;
found, 505.2731.
N-[(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-
hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]hex-
anamide (47). Reagent: hexanoyl chloride. Yield ) 84%. Mp
1
152.0 °C. HPLC: tR ) 23.3 min. H NMR (CDCl3) δ 0.84 (t, J
) 6.6 Hz, 3H), 0.90 (t, J ) 7.2 Hz, 3H), 1.02-1.18 (m, 1H),
1.20-1.68 (m, 14H), 1.73 (d, J ) 6.6 Hz, 1H), 2.15 (m, 1H),
2.23 (t, J ) 7.8 Hz, 3H), 3.11 (ABq, 2H), 4.16 (d, J ) 6.3 Hz,
1H), 5.53 (s, 1H), 6.64 (d, J ) 1.5 Hz, 1H), 7.31-7.48 (m, 6H),
7.85 (d, J ) 8.4 Hz, 1H), 8.02 (d, J ) 8.7 Hz, 1H). HRMS
(FAB/M + NH4) calcd for C28H43N2O4S: 503.2944; found,
503.2931. Anal. Calcd for C28H39O4SN: C, 69.24; H, 8.09; N,
2.88; S, 6.60. Found: C, 68.24; H, 8.02; N, 2.84; S, 6.50.
N-[(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-
hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]-
methanesulfonamide (48). Reagent: methanesulfonyl chlo-
ride. Yield ) 33%. Mp 265.0 °C. HPLC: tR ) 14.3 min. 1H NMR
(CDCl3) δ 0.84 (t, J ) 7.5 Hz, 3H), 0.92 (t, J ) 7.2 Hz, 3H),
1.09-1.22 (m, 1H), 1.25-1.65 (m, 6H), 2.25 (m, 1H), 2.84 (s,
3H), 3.08 (s, 2H), 4.09 (s, 2H), 5.49 (s, 1H), 6.52 (d, J ) 2.1
Hz, 1H), 7.25-7.31 (m, 2H), 7.38 (t, J ) 7.2 Hz, 2H), 7.50 (d,
J ) 7.2 Hz, 2H), 7.93 (d, J ) 8.7 Hz, 1H), 9.10 (s, 1H). MS
(FAB/M + Li): 472. HRMS (FAB/M + NH4) calcd for
C23H35N2O5S2: 483.1987; found, 483.1976. RP-HPLC: 5-90%
MeOH/H2O/0.1% TFA, 96.7% pure; 5-90% CH3CN/H2O/0.1%
TFA, 97.1% pure; average, 96.9%.
(()-(3S,4R,5R)-3-Butyl-3-ethyl-5-(4-fluorophenyl)-2,3,4,5-
tetrahydro-7-(methylthio)-1-benzothiepin-4-ol 1,1-Diox-
ide (49). Step 1. (()-(3S,4R,5R)-3-Butyl-3-ethyl-5-(4-fluo-
rophenyl)-2,3,4,5-tetrahydro-7-fluoro-benzothiepin-4-
ol 1,1-Dioxide (Intermediate 69). Alkylation of 4-fluorophenol
with 4-fluorobenzyl chloride according to the procedure de-
scribed in ref 23 gave 4-fluoro-2-(4′-fluorobenzyl)phenol. This
material was converted to compound 69 by a procedure similar
to that described for compound 34a, mp 134.5-139 °C, and
its (()-(3S,4S,5S)-diastereomer, mp 228-230 °C. HPLC (ethyl
acetate/hexane) gave separation of the diastereomers. 13C NMR
(CDCl3) δ 6.63, 13.77, 13.84, 23.0, 24.7, 27.6, 29.0, 44.1, 46.1,
56.4, 75.5, 113.3 (d, J ) 22.1 Hz), 115.7 (d, J ) 21.4 Hz), 118.7
(d, J ) 24.0 Hz), 129.3 (d, J ) 9.55 Hz), 130.4 (d, J ) 8.0 Hz),
135.6 (d, J ) 3.1 Hz), 138.1 (d, J ) 3.4 Hz), 143.4 (d, J ) 7.3
Hz), 161.8 (d, J ) 246.4 Hz), 165.1 (d, J ) 254.0 Hz).
(()-(3S,4R,5R)-7-(1-Hexylamino)-3-butyl-3-ethyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (44).
Compound 39 (0.10 g, 0.26 mmol), Ru3(CO)12 (0.010 g, 0.0078
mmol), hexanal (0.052 g, 0.52 mmol), and ethanol (50.0 mL)
were combined in a Parr reactor. The reactor was purged with
H2 and fitted with an overhead stirrer. The reactor was loaded
with 540 psig of H2. An amount of 60 psig of CO/H2 was added,
and the reactor was heated to 125 °C. After 18 h, the reactor
was cooled to room temperature, and its contents were filtered
and concentrated in vacuo. Purification by preparative HPLC
eluting with 15% EtOAc/hexanes and concentrated in vacuo
gave compound 44 (0.0757 g, 62%), mp 154.0-160.0 °C. 1H
NMR (CDCl3) δ 0.82-0.93 (m, 9H), 1.20-1.64 (m, 15H), 2.20
(m, 1H), 3.10 (m, 4H), 4.17 (s, 1H), 5.50 (br s, 1H), 6.37 (br s,
1H), 6.98 (br s, 1H), 7.33-7.48 (m, 5H), 7.95 (br s, 1H). MS
(CI/M + H): 472. HRMS (FAB/M + H) calcd for C28H42NO3S:
472.2885; found, 472.2866. Anal. Calcd for C28H41O3SN: C,
71.30; H, 8.76; N, 2.79; S, 6.80. Found: C, 70.16; H, 8.82; N,
2.88; S, 6.60.
(()-(3S,4R,5R)-7-(Dimethylamino)-3-butyl-3-ethyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (45).
Compound 39 (0.200 g, 0.516 mmol), methanol (4.0 mL), THF
(3.0 mL), aqueous 37% formaldehyde (0.400 mL/5.34 mmol),
and a solution of Raney nickel (0.200 g) in ethanol (0.200 mL)
were combined in a 1 oz. Fischer-Porter bottle. The bottle was
purged with N2, loaded to 40 psig of H2, and heated to 40 °C.
After 2 h, the reaction mixture was allowed to cool to room
temperature, filtered through Celite washing with ethyl ether,
and concentrated in vacuo. The residue was redissolved in
ethyl ether, washed with brine, dried (MgSO4), filtered, and
concentrated in vacuo to give compound 45 (0.200 g, 93%). 1H
NMR (CDCl3) δ 0.83 (t, J ) 7.5 Hz, 3H), 0.89 (t, J ) 7.5 Hz,
3H), 1.04-1.17 (m, 1H), 1.23-1.65 (m, 6H), 2.20 (m, 1H), 2.77
(s, 6H), 3.06 (ABq, 2H), 4.14 (d, J ) 7.8 Hz, 1H), 5.53 (s, 1H),
5.91 (d, J ) 2.4 Hz, 1H), 6.49 (dd, J ) 9.0, 2.7 Hz, 1H), 7.29-
7.41 (m, 5H), 7.49 (d, J ) 7.2 Hz, 1H), 7.88 (d, J ) 9.0, 1H).
MS (FAB/M + Li): 422. HRMS (M+) calcd for C24H33O3SN:
415.2181; found, 415.2187; C, 69.36; H, 8.0; N, 3.37; S, 7.1.
Found: C, 69.21; H, 8.08; N, 3.28; S, 7.58.
Step 2. A mixture of 0.68 g (1.66 mmol) of intermediate 69,
0.2 g (5 mmol) of sodium methanethiolate, and 15 mL of
anhydrous DMF was stirred at room temperature for 16 days.
The reaction mixture was diluted with ether and washed with
water and brine and dried over MgSO4. The ether solution was
concentrated in vacuo. The residue was purified by HPLC (20%
General Procedure for Preparation of Compounds
46-48. In a typical procedure, compound 39 (0.194 mmol) and
CH2Cl2 (4.0 mL) were combined in a 25 mL round-bottom flask.
The reaction flask was purged with N2 and fitted with a

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5849
ethyl acetate in hexanes). The first fraction was impure (()-
(3S,4S,5S)-diastereomer. The second fraction was compound
49, mp 185-186.5 °C. 1H NMR (CDCl3) δ 0.88-1.04 (m, 6 H),
1.1-1.8 (m, 7 H), 2.26 (br t, J ) 12.0 Hz, 1 H), 2.38 (s, 3 H),
3.18 (ABq, J ) 15.1 Hz, 2 H), 4.20 (d, J ) 6 Hz, 1 H), 5.58 (s,
1 H), 6.62 (s, 1 H), 7.14-7.26 (m, 3 H), 7.5-7.6 (m, 2 H), 8.04
(d, J ) 9 Hz, 1 H). 13C NMR (CDCl3) δ 162.99 (d, J ) 246.5
Hz), 147.66, 140.08 (d, J ) 1.2 Hz), 139.17 (d, J ) 3.4 Hz),
136.75, 131.63 (d, J ) 8 Hz), 129.04, 128.47, 116.93 (d, J )
21.2 Hz), 76.91, 57.76, 47.16, 45.22, 30.21, 28.95, 25.91, 24.25,
15.54, 15.10, 7.98. GC-MS (CI/M + H): 437.
(()-(3S,4R,5R)-3-Butyl-3-ethyl-5-(4-fluorophenyl)-2,3,4,5-
tetrahydro-7-(1-pyrrolidinyl)-1-benzothiepin-4-ol 1,1-
Dioxide (50). The amine is pyrrolidine. A mixture of 0.53 g
(1.30 mmol) of intermediate 69 (from step 1 of synthesis of
49) and 50 equiv of amine was held at reflux for 2 h. The
reaction mixture was diluted with ether, washed with water
and brine, and dried over MgSO4. The ether solution was
concentrated in vacuo. The residue was crystallized from ether/
hexanes to give compound 50, mp 174.5-177 °C. 1H NMR
(CDCl3) δ 0.8-1.0 (m, 6 H), 1.1-1.8 (m, 11 H), 2.26 (br t, J )
12.6 Hz, 1 H), 2.9-3.2 (m, 6 H), 3.81 (d, J ) 7 Hz, 2 H), 4.11
(d, J ) 7.2 Hz, 1 H), 5.46 (s, 1 H), 5.82 (s, 1 H), 6.04 (d, J )
7.8, 1 H), 7.18 (t, J ) 7.8 Hz, 2 H), 7.58 (t, J ) 6.6 Hz, 2 H),
7.74 (d, J ) 7.8 Hz, 1 H). 13C NMR (acetone-d6) δ 6.4, 13.5,
14.8, 23.2, 24.8, 25.0, 27.8, 44.1, 46.5, 47.0, 57.1, 75.8, 108.0,
114.5, 115.2 (d, J ) 21.2 Hz), 126.3, 128.2, 130.9 (d, J ) 8
Hz), 140.4 (d, J ) 3.4 Hz), 141.0, 150.5, 161.7 (d, J ) 243.8
Hz). MS (CI/M + H): 460.
General Procedure for Compounds 55-60. In a typical
procedure, 0.100 mol of 2-alkyl-2-(hydroxymethyl)alkanal,
prepared by reaction of formaldehyde with 2-alkylalkanal,
according to the procedure described in ref 22, was added
dropwise to a cold (10 °C) solution of 12.6 g (0.11 mol) of
methanesulfonyl chloride and 10.3 g (0.13 mol) of triethy-
lamine while maintaining the reaction temperature below 30
°C. The reaction mixture was stirred at room temperature for
18 h, the reaction was quenched with dilute HCl, and the
sample was extracted with methylene chloride. The methylene
chloride extract was dried over MgSO4 and concentrated in
vacuo to give the 2-alkyl-2-(methanesulfonylmethyl)alkanal as
a brown oil. This brown oil was used in the synthesis of
compounds 55-58 by a procedure similar to the one described
for compound 34a above. See each compound below for
purification methods and specific alkanal used.
(()-(4R,5R)-7-Amino-3,3-dimethyl-2,3,4,5-tetrahydro-5-
phenyl-1-benzothiepin-4-ol 1,1-Dioxide (55). The reagent
was 2-methylpropanal. The residue was concentrated in vacuo,
and the desired product was recrystallized from EtOAc/CH2-
Cl2/hexane as a white solid, mp 185-186 °C. 1H NMR (CDCl3)
δ 1.12 (s, 3H), 1.49 (s, 3H), 3.14 (qab, Jab) 15, 81.9 Hz, 2H),
3.28 (d, J ) 15 Hz, 1H), 4.00 (s, 1H), 5.30 (s, 1H), 5.51 (s, 1H),
5.97 (s, 1H), 6.56 (dd, J ) 6.5, 2.1 Hz), 7.31-7.52 (m, 5H),
7.89 (d, J ) 8.4 Hz, 1H). MS (FAB/M + H): 332. HRMS (EI)
calcd for C18H21NO3S: 331.1242; found, 331.1232.
2.13-2.21 (m, 1H), 3.09 (d, J ) 15.3 Hz, 2H), 3.45 (t, J ) 6.6
Hz, 1H), 3.70 (t, J ) 6.6 Hz, 1H), 4.15 (m, 2H), 5.50 (s, 1H),
5.94 (s, 1H), 6.73 (dd, J ) 8.4, 2.4 Hz, 1H), 7.32-7.48 (m, 6H),
7.86 (d, J ) 9.0 Hz, 1H). MS (FAB/M + H): 388.
(()-(4R,5R)-7-Amino-3,3-dibutyl-2,3,4,5-tetrahydro-5-
phenyl-1-benzothiepin-4-ol 1,1-Dioxide (58). The reagent
was 2-butylhexanal. The residue was concentrated in vacuo,
and the desired product was isolated via preparative HPLC
(eluting with EtOAc/hexane) as a white solid, mp 93.4-94.6
°C. 1H NMR (CDCl3) δ 0.92 (m, 6H), 1.03-1.70 (m, 12H), 2.21
(m, 1H), 3.09 (qAB, JAB ) 18.0 Hz, ∆ν ) 37.8 Hz, 2H), 3.96 (bs,
2H), 4.14 (d, J ) 7.4 Hz), 5.51 (s, 1H), 5.94 (s, 1H), 6.56 (d, J
) 8.7 Hz, 1H), 7.41-7.53 (m, 6H), 7.87 (d, J ) 8.4 Hz, 1H).
HRMS (ES+) calcd for C24H34NO3S: 416.2259; found, 416.2261.
Anal. Calcd for C24H33NO3S‚0.2H2O: C, 68.77; H, 8.03; N, 3.34.
Found: C, 68.83; H, 8.33; N, 3.01.
(()-(4R,5R)-7-Methyl-3,3-dipentyl-2,3,4,5-tetrahydro-5-
(4-fluorophenyl)-1-benzothiepin-4-ol 1,1-Dioxide (60). The
reagent was 2-pentylheptanal. Alkylation of 4-methylphenol
with 4-fluorobenzyl chloride according to the procedure de-
scribed in J. Chem. Soc. 1958, 2431 gave 4-methyl-2-(4′-
1
fluorobenzyl)phenol. H NMR (CDCl3) δ 2.25 (s, 3H), 3.92 (s,
2H), 6.68 (d, J ) 8.0 Hz, 1H), 6.90-6.94 (m, 4H), 7.17 (d, J )
5.2 Hz, 1H), 7.19 (d, J ) 6.0 Hz, 1H). MS (EI/M+): 216. HRMS
(EI/M+) calcd for C14H13FO: 216.0950; found, 216.0941. This
phenol and the 2-pentylheptanal were employed in a procedure
similar to the one for compounds 55-58 described above to
give compound 60. The residue was concentrated in vacuo, and
the desired product was isolated via preparative HPLC (eluting
with EtOAc/hexane) as a white solid 60: mp 85-87°C.
1
HPLC: tR ) 35.1 min. H NMR (CDCl3) δ 0.85-1.05(m, 6H),
1.10-1.95 (m, 15H), 2.28 (dt, J ) 13.2, 3.3 Hz, 1H), 2.35 (s,
3H), 3.12 (ABq, 2H), 4.18 (d, J ) 6.9 Hz, 1H), 5.54 (s, 1H),
6.68 (s, 1H), 7.18-7.26 (m, 3H), 7.50 (t, J ) 6.9 Hz, 2H), 8.0
(d, J ) 2.7 Hz, 1H). MS (CI/M + H): 461. HRMS (CI/M +
NH4) calcd for C27H41FNO3S: 478.2791; found, 478.2794.
(()-(3S,4R,5R)-8-Methoxy-3-butyl-3-ethyl-2,3,4,5-tet-
rahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (61).
The substitution of 3-methoxythiophenol for 4-methoxythiophe-
nol in the procedure for compound 36, steps 1-5, gave 61. The
desired product was isolated via preparative HPLC (eluting
with EtOAc/hexane) as a white solid, mp 125.5-126.5 °C.
HPLC: tR ) 24.1 min. 1H NMR (CDCl3) δ 0.86 (t, J ) 7.5 Hz,
3H), 0.92 (t, J ) 7.2 Hz, 3H), 1.10-1.66 (m, 7H), 2.23 (dt, J )
13.2, 3.3 Hz, 1H), 3.12 (ABq, 2H), 3.83 (s, 3H), 4.18 (d, J ) 6.9
Hz, 1H), 5.54 (s, 1H), 6.68 (d, J ) 8.7 Hz, 1H), 6.91 (dd, J )
8.7, 3.0 Hz, 1H), 7.34 (t, J ) 6.9 Hz, 1H), 7.42 (t, J ) 6.9 Hz,
2H), 7.47 (t, J ) 7.2 Hz, 2H), 7.66 (d, J ) 2.7 Hz, 1H). MS
(CI/M + H): 403. HRMS (CI/M + H) calcd for C23H31O4S:
403.1943; found, 403.1939.
(()-(4R,5R)-8-Methoxy-3,3-dibutyl-2,3,4,5-tetrahydro-
5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (62). Compound
62 was prepared by a similar procedure to that described for
61, substituting 2-butylhexanal for 2-ethylhexanal. The result-
ing residue was concentrated in vacuo, and the desired product
62 was isolated via preparative HPLC (eluting with EtOAc/
hexane) as a white foam. 1H NMR (CDCl3) δ 0.82-0.98 (m,
6H), 1.1-1.65 (m, 12H), 2.25 (td, J ) 15, 3 Hz, 1H), 3.12 (ABq,
JAB ) 16.5 Hz, 2H), 3.83 (s, 3H), 4.18 (d, J ) 7.5 Hz, 1H), 5.3
(s, 1H), 6.68 (d, J ) 8.4 Hz, 1H), 6.92 (dd, J ) 8.4, 3 Hz, 1H),
7.3-7.5 (m, 5H), 7.66 (d, J ) 3 Hz, 1H). HRMS calcd for
C25H35O4S: 431.2256; found, 431.2281.
(()-(4R,5R)-7-Amino-3,3-diethyl-2,3,4,5-tetrahydro-5-
phenyl-1-benzothiepin-4-ol 1,1-Dioxide (56). The reagent
was 2-ethylbutanal. The residue was concentrated in vacuo,
and the desired product was isolated via preparative HPLC
(eluting with EtOAc/hexane) as a white solid, mp 245.8-246.8
1
°C. H NMR (CDCl3) δ 0.84 (t, J ) 6 Hz, 3H), 0.92 (t, J ) 6
Hz, 3H), 1.36 (m, 1H), 1.54 (m, 1H), 1.71 (m, 1H), 2.28 (m,
1H), 3.09 (qAB, JAB ) 15.0 Hz, ∆ν ) 42.4 Hz, 2H), 3.97 (bs,
2H), 4.15 (d, J ) 8 Hz, 1H), 5.49 (s, 1H), 5.95 (s, 1H), 6.54 (d,
J ) 1H), 7.29-7.53 (m, 6H), 7.88 (d, J ) 8.2 Hz, 1H). HRMS
(ES/M + H) calcd for C20H26NO3S: 360.1633; found, 360.1632.
Anal. Calcd for C20H25NO3S: C, 66.82; H, 7.01; N, 3.90. Found
C, 66.54; H, 7.20; N, 3.69.
(()-(4R,5R)-7-(Dimethylamino)-3,3-dibutyl-2,3,4,5-tet-
rahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (64).
Compound 64 was prepared by a similar procedure to that
described for 45, substituting 2-butylhexanal for 2-ethylhexa-
nal. The resulting residue was concentrated in vacuo, and the
desired product was isolated via preparative HPLC (eluting
(()-(4R,5R)-7-Amino-3,3-dipropyl-2,3,4,5-tetrahydro-5-
phenyl-1-benzothiepin-4-ol 1,1-Dioxide (57). The reagent
was 2-propylpentanal. The residue was concentrated in vacuo,
and the desired product was isolated via preparative HPLC
(eluting with EtOAc/hexane) as a white solid. 1H NMR (CDCl3)
δ 0.85-0.95 (m, 6H), 1.42-1.80 (m, 4H), 1.92-1.98 (m, 1H),
1
with EtOAc/hexane) as a white solid, mp 161.5-162.2 °C. H
NMR (CDCl3) δ 0.90 (t, J ) 6.3 Hz, 6H), 1.05-1.54 (m, 11H),
1.63 (m, 1H), 2.24 (m, 1H), 2.78 (s, 6H), 3.05 (qAB, JAB ) 15
Hz, ∆ν ) 46.7 Hz, 2H), 4.18 (m, 1H), 5.53 (s, 1H), 5.93 (s, 1H),
6.94 (d, J ) 8.5 Hz, 1H), 7.27-7.42 (m, 3H), 7.45 (d, J ) 8.3

5850 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
Hz, 2H), 7.87 (d, J ) 8.7 Hz, 1H). HRMS (ES+) calcd for
in vacuo, and the desired product was isolated via preparative
HPLC (eluting with EtOAc/hexane) as a white solid, mp 95.5-
97.6 °C. 1H NMR (CDCl3) δ 0.91 (m, 6H), 1.02-1.52 (m, 12H),
1.63 (m, 1H), 2.23 (m, 1H), 3.12 (qAB, JAB ) 18.0 Hz, ∆ν ) 36.1
Hz, 2H), 4.18 (d, J ) 7.0 Hz, 1H), 5.13 (s, 1H), 5.53 (s, 1H),
6.43 (s, 1H), 6.65 (s, 1H), 7.29-7.52 (m, 10H), 7.74 (d, J ) 8.7
Hz, 1H), 8.03 (d, J ) 8.4 Hz, 1H). HRMS (ES+/M + H) calcd
for C32H40NO5S: 550.2627; found, 550.2643.
C26H38NO3S: 444.2572; found, 444.2563.
(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-hy-
droxy-N,N,N-trimethyl-5-phenyl-1-benzothiepin-7-amin-
ium Iodide 1,1-Dioxide (65). Reaction of compound 45 with
MeI in CH3CN, according to the procedure described in
Tetrahedron 1993, 49 (46), 10733-10738, gave 65. The residue
was concentrated in vacuo, and the desired product was
recrystallized from acetonitrile/ethyl ether as a yellow solid,
mp 120-124 °C. 1H NMR (CD3OD) δ 0.83 (t, J ) 8.2 Hz, 3H),
0.9 (t, J ) 8.2 Hz, 3H), 1.04-1.6 (m, 6H), 1.78 (br t, J ) 14.4
Hz, 1H), 2.07 (br t, J ) 14.4 Hz, 1H), 3.18 (q, J ) 16.8 Hz,
2H), 3.46 (s,9H), 4.16 (s, 1H), 5.46 (s, 1H), 7.06 (s, 1H), 7.33-
7.55 (m, 5H), 7.95 (d, J ) 7.5 Hz, 1H), 8.2 (d, J ) 7.5 Hz, 1H).
HRMS (M+) calcd for C25H36NO3S: 430.2416; found, 430.2417.
(()-(4R,5R)-3,3-Dibutyl-2,3,4,5-tetrahydro-4-hydroxy-
N,N,N-trimethyl-5-phenyl-1-benzothiepin-7-aminium Io-
dide 1,1-Dioxide (66). Compound 66 was prepared by a
similar procedure to that described for 65, substituting 2-bu-
tylhexanal for 2-ethylhexanal. The resulting residue was
concentrated in vacuo, and the desired product was isolated
via preparative HPLC (eluting with EtOAc/hexane) as a white
solid 66, mp 146.5-148.0 °C. 1H NMR (CD3OD) δ 0.91 (m, 6H),
1.02-1.55 (m, 11H), 1.88 (m, 1H), 2.12 (m, 1H), 3.15-3.35 (m,
2H), 3.49 (s, 9H), 4.20 (s, 1H), 5.51 (s, 1H), 7.10 (s, 1H), 7.37-
7.59 (m, 5H), 8.01 (d, J ) 8.7 Hz, 1H), 8.24 (d, J ) 8.4 Hz,
1H). HRMS (FAB/M+) calcd for C27H40NO3S: 458.2729; found,
458.2743.
(()-(3S,4R,5R)-7-Methyl-3-butyl-3-ethyl-2,3,4,5-tetrahy-
dro-5-(4-fluorophenyl)-1-benzothiepin-4-ol 1,1-Dioxide
(67). Compound 67 was prepared by a procedure similar to
that described for 60, substituting 2-ethylhexanal for 2-pen-
tylheptanal. The resulting residue was concentrated in vacuo,
and the desired product was isolated via preparative HPLC
(eluting with EtOAc/hexane) as a white solid, mp 164-165 °C.
HPLC: tR ) 26.4 min. 1H NMR (CDCl3) δ 0.85 (t, J ) 7.5 Hz,
3H), 0.95 (t, J ) 7.2 Hz, 3H), 1.10-1.80 (m, 7H), 2.28 (dt, J )
13.2, 3.3 Hz, 1H), 2.35 (s, 3H), 3.12 (ABq, 2H), 4.18 (d, J ) 6.9
Hz, 1H), 5.54 (s, 1H), 6.68 (s, 1H), 7.18-7.26 (m, 3H), 7.50 (t,
J ) 6.9 Hz, 2H), 8.10 (d, J ) 2.7 Hz, 1H). MS (CI/M + H):
405. HRMS (CI/M + NH4) calcd for C23H33NFO3S: 422.2165;
found, 422.2142.
(()-(4R,5R)-7-Methyl-3,3-dibutyl-2,3,4,5-tetrahydro-5-
(4-fluorophenyl)-1-benzothiepin-4-ol 1,1-Dioxide (68). Com-
pound 68 was prepared by a procedure similar to that
described for 60, substituting 2-butylhexanal for 2-pentylhep-
tanal. The residue was concentrated in vacuo, and the desired
product was isolated via preparative HPLC (eluting with
EtOAc/hexane) as a white solid, mp 145-147 °C. HPLC: tR )
31.5min. 1H NMR (CDCl3) δ 0.85-1.05 (m, 6H), 1.15-1.80 (m,
11H), 2.28 (dt, J ) 13.2, 3.3 Hz, 1H), 2.35 (s, 3H), 3.12 (ABq,
2H), 4.18 (d, J ) 6.9 Hz, 1H), 5.54 (s, 1H), 6.68 (s, 1H), 7.18-
7.26 (m, 3H), 7.50 (t, J ) 6.9 Hz, 2H), 8.10 (d, J ) 2.7 Hz,
1H). MS (CI/M + H): 433. HRMS (CI/M + NH4) calcd for
C25H37NFO3S: 450.2478; found, 450.2471.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-meth-
oxyphenyl)-7-(dimethylamino)-1-benzothiepin-4-ol 1,1-
Dioxide (75). 2-Chloro-5-nitro-4′-methoxybenzophenone was
prepared according to the procedure described in Chem.
Pharm. Bull. 1997, 45 (9), 1470-1474. This material was
converted to compound 75 by a procedure similar to that
described for 45, substituting 2-butylhexanal for 2-ethylhexa-
nal. The residue was concentrated in vacuo, and the desired
product was isolated via preparative HPLC (eluting with
1
EtOAc/hexane) as a white solid. H NMR (CDCl3) δ 0.90 (m,
6H), 1.05-1.45 (m, 10H), 1.63 (m, 1H), 2.21 (m, 1H), 2.80 (s,
6H), 3.06 (qAB, JAB ) 16.2 Hz, ∆ν ) 45.0 Hz, 2H), 3.83 (s, 3H),
4.10 (d, J ) 7.5 Hz, 1H), 5.46 (s, 1H), 5.98 (d, J ) 0.9 Hz, 1H),
6.50 (d, J ) 8.7 Hz, 1H), 6.94 (d, J ) 8.7 Hz, 2H), 7.42 (d, J )
8.7 Hz, 2H), 7.87 (d, J ) 8.7 Hz, 1H). HRMS (FAB+/M + H)
calcd for C27H40NO4S: 474.2678; found, 474.2675. Anal. Calcd
for C27H39NO4S: C, 68.48; H, 8.30; N, 2.96. Found C, 68.89;
H, 8.41; N, 2.86.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-fluo-
rophenyl)-7-(dimethylamino)-1-benzothiepin-4-ol 1,1-
Dioxide (76). Compound 70 was reacted with HNMe2 by a
procedure similar to that described for 50, to give compound
76. The residue was concentrated in vacuo, and the desired
product was isolated via preparative HPLC (eluting with
EtOAc/hexane) as a white solid, mp 170.5-171.0 °C. 1H NMR
(CDCl3) δ 0.95-1.02 (m, 6H), 1.15-1.58 (m, 10H), 1.74 (m, 1H),
2.29 (m, 1H), 2.90 (s, 6H), 3.16 (qAB, JAB ) 14.4 Hz, ∆ν ) 51.0
Hz, 2H), 4.19 (d, J ) 8.7 Hz, 1H), 5.62 (s, 1H), 5.98 (s, 1H),
6.60 (d, J ) 8.7 Hz, 1H), 7.17 (d, J ) 7.5 Hz, 1H), 7.20 (d, J )
7.5 Hz, 1H), 7.58 (m, 2H), 7.99 (d, J ) 8.7 Hz, 1H). MS (EI):
461.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-phenyl-
8-(dimethylamino)-1-benzothiepin-4-ol 1,1-Dioxide (80).
The phenol was 3-fluorophenol. Benzyl chloride was used, and
the amine was dimethylamine. Alkylation of a substituted
phenol with a substituted benzyl chloride, according to the
procedure described in ref 23 gave a substituted diphenyl-
methane. With this diphenylmethane and with substitution
of 2-butylhexanal for 2-ethylhexanal, compound 80 was pre-
pared by a procedure similar to that described for 50. The
residue was concentrated in vacuo, and the desired product
was isolated via preparative HPLC (eluting with EtOAc/
hexane) as a white solid. 1H NMR (CDCl3) δ 0.91 (m, 6H),
1.07-1.78 (m, 11H), 2.24 (m, 1H), 2.98 (s, 6H), 3.11 (qAB, J )
15.3 Hz, ∆ν ) 37.6 Hz, 2H), 4.14 (d, J ) 6.6 Hz, 1H), 5.49 (s,
1H), 6.60 (d, J ) 8.7 Hz, 1H), 6.76 (m, 1H), 7.30-7.56 (m, 7H).
HRMS (FAB/M + H) calcd for C26H38NO3S: 444.2572; found,
444.2565.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-fluo-
rophenyl)-9-(dimethylamino)-1-benzothiepin-4-ol 1,1-
Dioxide (81). Compound 93 was reacted with dimethylamine,
using a procedure similar to that for compound 50, to give
compound 81. The residue was concentrated in vacuo, and the
desired product was isolated via recrystallization from ethyl
ether/hexane as a colorless solid, mp 159.0-160.0 °C. 1H NMR
(CDCl3) δ 0.86-0.94 (m, 6H), 1.12-1.62 (m, 11H), 2.30 (m, 1H),
3.00 (m, 8H), 4.08 (d, J ) 7.8 Hz, 1H), 5.70 (s, 1H), 6.29 (d, J
) 7.8 Hz, 1H), 6.88 (d, J ) 7.8 Hz, 1H), 7.01 (d, J ) 8.1 Hz,
1H), 7.04 (d, J ) 8.1 Hz, 1H), 7.15 (t, J ) 7.9 Hz, 1H), 7.43
(m, 2H). MS (EI): 461. Anal. Calcd for C26H36FNO3S: C, 67.65;
H, 7.86; N, 3.03. Found: C, 67.16; H, 7.91; N, 3.02.
(()-(4R,5R)-7-(Phenylmethylamino)-3,3-dibutyl-2,3,4,5-
tetrahydro-5-phenyl-1-benzothiepin-4-ol 1,1-Dioxide (73).
Compound 73 was prepared by a procedure similar to that
described for 45, substituting benzaldehyde for formaldehyde
and substituting 2-butylhexanal for 2-ethylhexanal. The resi-
due was concentrated in vacuo, and the desired product was
isolated via preparative HPLC (eluting with EtOAc/hexane)
as a white solid, mp 86.8-88.1 °C. 1H NMR (CDCl3) δ 0.85 (t,
J ) 6 Hz, 6H), 0.98-1.66 (m, 12H), 2.16 (m, 1H), 3.04 (qAB
,
JAB ) 15 Hz, ∆ν ) 41.3 Hz, 2H), 4.08 (s, 1H), 4.12 (s, 1H),
5.44 (s, 1H), 5.84 (s, 1H), 6.42 (d, J ) 9 Hz, 1H), 7.12 (d, J )
7.8 Hz, 2H), 7.16-7.26 (m, 10H), 7.83 (d, J ) 8.3 Hz, 1H).
HRMS (ES+/M + H) calcd for C31H40NO3S: 506.2729; found,
506.2714.
[(()-(4R,5R)-3,3-Dibutyl-2,3,4,5-tetrahydro-4-hydroxy-
1,1-dioxido-5-phenyl-1-benzothiepin-7-yl]carbamic Acid,
Phenylmethyl Ester (74) Compound 74 was prepared by a
similar procedure to that described for 42, substituting 2-bu-
tylhexanal for 2-ethylhexanal. The residue was concentrated
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-meth-
oxyphenyl)-7-(dimethylamino)-9-methoxy-1-benzothiepin-
4-ol 1,1-Dioxide (88). The phenol was 2 methoxy-4-fluorophe-
nol, benzyl chloride was used, and the 4-methoxybenzyl
chloride amine was dimethylamine. Alkylation of a substituted

Part 1: Bile Acid Transporter Inhibitors
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18 5851
phenol with a substituted benzyl chloride, according to the
procedure described in ref 23, gave a substituted diphenyl-
methane. Using this diphenylmethane and substituting 2-bu-
tylhexanal for 2-ethylhexanal, compound 88 was prepared by
a procedure similar to that described for 50. The residue was
concentrated in vacuo, and the desired product was isolated
via preparative HPLC (eluting with EtOAc/hexane) as a white
solid. 1H NMR (CDCl3) δ 0.87-0.91 (m, 6H), 1.10-1.63 (m,
11H), 2.80 (s, 6H), 3.18 (qAB, JAB ) 15.0 Hz, ∆ν ) 40.5 Hz,
2H), 3.82 (s, 3H), 3.88 (s, 3H), 4.08 (s, 1H), 5.65 (s, 1H), 6.12
(s, 1H), 6.90 (d, J ) 8.4 Hz, 2H), 7.38 (d, J ) 8.4 Hz, 2H), 7.90
(d, J ) 9.0 Hz, 1H). MS for C28H41O5NS: 516.4. RP-HPLC:
5-90% MeOH/H2O/0.1% TFA, 91.5% pure; 5-90% CH3CN/
H2O/0.1% TFA, 92.4% pure; average, 92.0%. HRMS (ESI/M +
H) calcd for C28H42NO5S: 504.2785; found, 504.2778.
General Procedure for Compounds 89, 90, and 93.
Alkylation of a substituted phenol with a substituted benzyl
chloride, according to the procedure described in ref 23 gave
a substituted diphenylmethane. With this diphenylmethane
and with substitution of 2-butylhexanal for 2-ethylhexanal,
compounds 89, 90, and 93 were prepared by a procedure
similar to that described for 34a. See each compound below
for purification methods and specific phenol and benzyl
chloride used.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(2-bro-
mophenyl)-7-bromo-1-benzothiepin-4-ol 1,1-Dioxide (89).
The phenol was 4-bromophenol, and benzyl chloride was
2-bromobenzyl chloride. The residue was concentrated in
vacuo, and the desired product was isolated via preparative
HPLC (eluting with EtOAc/hexane) as a white solid. 1H NMR
(CDCl3) δ 0.91 (t, J ) 7.2 Hz, 3H), 0.92 (t, J ) 7.2 Hz, 3H),
1.08 (m, 1H), 1.20-1.47 (m, 9H), 1.70 (m, 1H), 2.20 (m, 1H),
3.14 (qAB, JAB ) 15.6 Hz, ∆ν ) 45.6 Hz, 2H), 4.15 (d, J ) 6.0
Hz, 1H), 5.99 (s, 1H), 6.81 (s, 1H), 7.24-7.28 (m, 2H), 7.49 (t,
J ) 8.0 Hz), 7.55 (dd, J ) 8.4, 2.0 Hz, 1H), 7.65 (d, J ) 8.0 Hz,
1H), 7.83 (d, J ) 7.6 Hz, 1H), 7.98 (d, J ) 8.4 Hz, 1H). HRMS
(CI/M + H) calcd for C24H31O3SBr2: 557.0361; found, 557.0352.
Anal. Calcd for C24H30O3SBr2: C, 51.63; H, 5.42; S, 5.74; Br,
28.62. Found: C, 51.10; H, 5.47; S, 5.61; Br, 29.10.
95. The residue was concentrated in vacuo, and the desired
product was isolated by precipitation from CH3CN/ethyl ether
as a white solid. 1H NMR (CDCl3) δ 0.89 (m, 6H), 1.12 (m,
1H), 1.23-1.80 (m, 14H), 2.18 (m, 1H), 3.10 (ABq, 2H), 3.33
(s, 9H), 3.46 (m, 2H), 3.99 (br s, 2H), 4.16 (s, 1H), 5.44 (s, 1H),
6.64 (d, J ) 7.8 Hz, 1H), 6.85 (m, 1H), 7.32 (d, J ) 6.0 Hz,
1H), 7.39 (t, J ) 6.6 Hz, 2H), 7.46 (m, 2H), 7.58 (s, 1H). MS
(FAB/M + H): 530.2.
2-[2-[2-[[(()-(3S,4R,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahy-
dro-4-hydroxy-1,1-dioxido-5-phenyl-1-benzothiepin-8-yl]-
oxy]ethoxy]ethoxy]-N,N,N-trimethylethanaminium Io-
dide (96). Compound 37 was reacted with K2CO3 and 1,2-
bis(2-iodoethoxy)ethane and then trimethylamine, using a
procedure similar to that described for 95, to give compound
96. The residue was concentrated in vacuo, and the desired
product was isolated by precipitation from CH3CN/ethyl ether
as a white solid, mp 101.0-105.0 °C. HPLC: tR ) 8.1 min. 1H
NMR (CDCl3) δ 0.80 (t, J ) 7.2 Hz, 3H), 0.85 (t, J ) 6.9 Hz,
3H), 1.06 (m, 2H), 1.20-1.40 (m, 3H), 1.45-1.65 (m, 2H), 2.12
(br t, J ) 12.6 Hz, 1H), 2.31 (d, J ) 6.3 Hz, 1H), 3.10 (ABq,
2H), 3.30 (s, 9H), 3.62 (s, 4H), 3.76 (m, 4H), 3.88 (br s, 2H),
4.09 (br t, J ) 3.9 Hz, 2H), 4.14 (br d, J ) 6.0 Hz, 1H), 5.36 (s,
1H), 6.63 (d, J ) 8.7 Hz, 1H), 6.86 (dd, J ) 8.7, 3.0 Hz, 1H),
7.28 (d, J ) 7.2 Hz, 1H), 7.35 (t, J ) 7.2 Hz, 2H), 7.46 (d, J )
7.2 Hz, 2H), 7.56 (d, J ) 2.7 Hz, 1H). MS (ES/M+): 562.2.
HRMS (ES/M+) calcd for C31H48O6SN: 562.3187; found,
562.3202. Anal. Calcd for C31H48O6SNI: C, 53.99; H, 7.01; N,
2.03. Found: C, 53.00; H, 6.97; N, 2.39.
N-[2-[2-[2-[4-[(()-(4R,5R)-3,3-Dibutyl-7-(dimethylamino)-
2,3,4,5-tetrahydro-4-hydroxy-1,1-dioxido-1-benzothiepin-
5-yl]phenoxy]ethoxy]ethoxy]ethoxy]-N,N,N-triethylamin-
ium Iodide (97). Compound 75 was reacted with BBr3, using
a procedure similar to that described for 37, to give (4â,5â)-
3,3-dibutyl-4-hydroxy-7-dimethylamino-5-(4′-hydroxyphenyl)-
2,3,4,5-tetrahydrobenzothiepine 1,1-dioxide 124. 1H NMR
(CDCl3) δ 0.90 (m, 6H), 1.04-1.18 (m, 1H), 1.20-1.46 (m, 9H),
1.63 (m, 1H), 2.19 (m, 1H), 3.07 (ABq, J ) 14.7 Hz, ∆ν ) 44.1
Hz, 2 H), 3.66-3.79 (m, 5H), 3.87 (m, 2H), 4.06-4.18 (m, 3H),
5.46 (s, 1H), 5.98 (d, J ) 3.0 Hz, 1H), 6.49 (dd, J ) 9.9, 2.7
Hz, 1H), 6.95 (d, J ) 8.1 Hz, 2H), 7.40 (d, J ) 8.1 Hz, 2H),
7.87 (d, J ) 9.3 Hz, 1H). MS (CI/M + Li): 749. Compound
124 was reacted with K2CO3 and 1,2-bis(2-iodoethoxy)ethane
and then triethylamine, using a procedure similar to that
described for 95, to give compound 97. The residue was
concentrated in vacuo, and the desired product was isolated
by precipitation from CH3CN/ethyl ether as a white solid, mp
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-fluo-
rophenyl)-7-methoxy-1-benzothiepin-4-ol 1,1-Dioxide (90).
The phenol was 4-methoxyphenolbenzyl chloride, and 4-fluo-
robenzyl chloride was used. The residue was concentrated in
vacuo, and the desired product was isolated via preparative
HPLC (eluting with EtOAc/hexane) as a white solid, mp 167-
1
170°C. HPLC: tR ) 30.5 min. H NMR (CDCl3) δ 0.85-1.05
1
131-133 °C. H NMR (CDCl3) δ 0.81-0.95 (m, 6H), 1.0-1.7
(m, 6H), 1.15-1.80 (m, 11H), 2.28 (dt, J ) 13.2, 3.3 Hz, 1H),
3.75 (s, 3H), 3.12 (ABq, 2H), 4.21 (s, 1H), 5.56 (s, 1H), 6.88
(dd, J ) 7 Hz, 2.6 Hz, 1H), 7.18 (t, J ) 7.0 Hz, 1H), 7.51 (t, J
) 6.9 Hz, 2H), 8.09 (d, J ) 2.7 Hz, 1H). MS (CI/M + H): 449.
HRMS (CI/M + NH4) calcd for C25H37FO4SN: 466.2427; found,
466.2417. Anal. Calcd for C25H33FO4S: C, 66.94; H, 7.40; S,
7.14. Found: C, 65.35; H, 7.06; S, 7.04.
(m, 20H), 2.17 (br t, J ) 13.1 Hz, 1H), 2.78 (s, 6H), 3.03 (q, J
) 20 Hz, 2H), 3.5 (q, J ) 7.5 Hz, 6H), 3.7 (s, 6H), 3.81 (t, J )
5.0 Hz, 2H), 3.97 (t, J ) 5.0 Hz, 2H), 4.04 (d, J ) 7.5 Hz, 1H),
4.1 (t, J ) 5.0 Hz, 2H), 5.42 (s, 1H), 5.89 (d, J ) 5.2 Hz, 1H),
6.45 (d, J ) 7.5 Hz, 1H), 6.86 (d, J ) 7.5 Hz, 2H), 7.35 (d, J )
7.5 Hz, 2H), 7.8 (d, J ) 7.5 Hz, 1H). HRMS (M - I) calcd for
C38H63N2O6S: 675.4407; found, 675.4368. Anal. Calcd for
C38H63N2O6SI: C, 56.85. H, 7.91. N, 3.49. I, 15.80. Found: C,
55.74. H, 7.83. N, 3.47. I, 15.49.
(()-(4R,5R)-3,3-Dibutyl-3-2,3,4,5-tetrahydro-5-(4-fluo-
rophenyl)-9-fluoro-1-benzothiepin-4-ol 1,1-Dioxide (93).
The phenol was 4-fluorophenol, and the benzyl chloride was
4-fluorobenzyl chloride. The residue was concentrated in vacuo,
the desired product was isolated via preparative HPLC (eluting
Supporting Information Available: X-ray crystallo-
graphic information for 34a. This material is available free of
charge via the Internet at http://pubs.acs.org.
1
with EtOAc/hexane) as a colorless solid. H NMR (CDCl3) δ
0.86-0.96 (m, 6H), 1.10-1.62 (m, 11H), 2.16 (m, 1H), 3.25 (qAB
,
References
JAB ) 15.0 Hz, ∆ν ) 58.3 Hz, 2H), 4.13 (s, 1H), 5.65 (s, 1H),
6.56 (d, J ) 5.1 Hz, 1H), 7.00-7.10 (m, 3H), 7.32 (m, 1H), 7.53
(m, 1H). MS (EI) for C24H30F2O3S: 436. Anal. Calcd for
C24H30F2O3S: C, 66.03; H, 6.92; S, 8.70. Found: C, 65.93; H,
7.01; S, 7.53.
(1) Stark, R. M. Review of the major intervention trials of lowering
coronary artery disease risk through cholesterol reduction. Am.
J. Cardiol. 1996, 78, 13-19.
(2) Anonymous. Executive summary of the third report of the
National Cholesterol Education Program (NCEP) expert panel
on detection, evaluation, and treatment of high blood cholesterol
in adults (adult treatment panel iii). JAMA, J. Am. Med. Assoc.
2001, 285 (May 16), 2486-2497.
(3) Shepherd, J.; Cobbe, S. M.; Ford, I.; Isles, C. G.; Lorimer, A. R.;
et al. Prevention of coronary heart disease with pravastatin in
men with hypercholesterolemia. N. Engl. J. Med.1995, 333,
1301-1307.
(4) Buchwald, H.; Williams, S. E.; Matts, J. P.; Nguyen, P. A.; Boen,
J. R. Overall mortality in the program on the surgical control of
the hyperlipidemias. J. Am. Coll. Surgeons 2002, 195 (Septem-
ber), 327-331.
6-[[(()-(3S,5R)-3-Butyl-3-ethyl-2,3,4,5-tetrahydro-4-hy-
droxy-1,1-dioxido-5-phenyl-1-benzothiepin-8-yl]oxy]-
N,N,N-trimethyl-1-hexanaminium Iodide (95). Compound
37 was reacted with K2CO3 and 1,6-diiodohexane according
to the procedure described in Org. Synth. 1955, 3 (Collective
Vol.), 140, to give (4â,5â)-3-butyl-3-ethyl-4-hydroxy-8-(6′-io-
dohexyloxy)-5-phenyl-2,3,4,5-tetrahydrobenzothiepine 1,1-
dioxide, which was reacted with trimethylamine, using a
procedure similar to that described for 50, to give compound

5852 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 18
Tremont et al.
(5) Sacks, F. M.; Pfeffer, M. A.; Moye, L. A.; Rouleau, J. L.;
Rutherford, J. D.; et al. The effect of pravastatin on coronary
events after myocardial infarction in patients with average
cholesterol levels. Cholesterol and recurrent events trial inves-
tigators. N. Engl. J. Med. 1996, 335 (October 3), 1001-1009.
(6) Stedronsky, E. R. Interaction of bile acids and cholesterol with
non-systemic agents having hypocholesterolemic properties.
Biochim. Biophys. Acta 1994, 1210, 255-287.
(7) Kramer, W.; Muellner, S.; Gutweiler, M.; Kroggel, M. (Hoechst
A.-G., Germany). Polymers and oligomers of bile acids, process
for their production and their use as medicines. Eur. Pat. Appl.
DE 91-4142379, EP 549967, 1993.
(15) Baganz, H.; Peissker, H. Water-soluble sulfonamides. I. p-
Nitrobenzenesulfenyl- and p-nitrobenzenesulfonylglycine deriva-
tives. Arch. Pharm. 1956, 289, 262-268.
(16) Luly, J. R.; Yi, N.; Soderquist, J.; Stein, H.; Cohen, J.; et al. New
inhibitors of human renin that contain novel Leu-Val replace-
ments. J. Med. Chem. 1987, 30, 1609-1616.
(17) Piccolo, O.; Filippini, L.; Tinucci, L.; Valoti, E.; Citterio, A. Boron
trichloride-catalyzed ortho carbonylation of phenols: synthesis
of 2-hydroxyarenecarboxamides and ketones. Tetrahedron 1986,
42, 885-891.
(18) Wolfers, H.; Kraatz, U.; Korte, F. New route to 1-alkylthio-3,5-
dihydroxybenzenes (monothiophloroglucinol-S-alkyl ethers). Syn-
thesis 1975, 43-44.
(19) Katritzky, A. R.; Kirichenko, K.; Ji, Y.; Prakash, I. Novel
syntheses of 2,3-disubstituted benzothiophenes. Chem. Hetero-
cycl. Compds. 2002, 38, 156-164 (Translation of Khim. Geterot-
sikl. Soedin.).
(20) van der Meer, R. H. d. V. a. J. F. C. G. t-Butanol Extraction of
Feces: A Rapid Procedure for Enzymatic Determination of Fecal
Bile Acids. In Cholesterol Metabolism in Health; Ponsen and
Looyen: Wageninged, 1985; pp 113-119.
(21) Turley, S. D.; Dietschy, J. M. Re-evaluation of the 3R-hydrox-
ysteroid dehydrogenase assay for total bile acids in bile. J. Lipid
Res. 1978, 19, 924-928.
(22) Nerdel, F. D. F. a. H.-J. L. Quatare salze von B-amino-aldhyden
und b-jod-aldehyde. Chem. Ber. 1965, 98, 728-734.
(23) Hickinbottom, F. M. E. a. W. J. The C-benzylation of phenols
by use of sodium hydride. J. Chem. Soc. 1958, 2431-2432.
(24) Benton, F. L.; Dillon, T. E. Cleavage of ethers with boron
bromide. I. Some common ethers. J. Am. Chem. Soc. 1942, 64,
1128-1129.
(8) Wess, G.; Kramer, W.; Enhsen, A.; Glombik, H.; Baringhaus,
K.-H.; et al. Specific inhibitors of ileal bile acid transport. J. Med.
Chem. 1994, 37, 873-875.
(9) Root, C.; Smith, C. D.; Winegar, D. A.; Brieaddy, L. E.; Lewis,
M. C. Inhibition of ileal sodium-dependent bile acid transport
by 2164U90. J. Lipid Res. 1995, 36, 1106-1115.
(10) Lewis, M. C.; Brieaddy, L. E.; Root, C. Effects of 2164U90 on
ileal bile acid absorption and serum cholesterol in rats and mice.
J. Lipid Res. 1995, 36, 1098-1105.
(11) Lee, L. F.; Banerjee, S. C.; Huang, H.-c.; Li, J. J.; Miller, R. E.;
et al. (G. D. Searle and Co.). Preparation of substituted 5-aryl-
benzothiepines as ileal bile acid transport and taurocholate
uptake inhibitors. U.S. Ser. No. 109,551; U.S. 99-275463
19990324, 2001; U.S. Ser. No. 831,284, 2004.
(12) Feng, W.; Hou, J. Progress in the study of ileal Na+/bile acid
transporter and its inhibitors. Shengli Kexue Jinzhan 2002, 33,
215-219.
(13) Booker, M. a. A. Cynthia ileal bile acid transport inhibitors as
potential hypocholesterolemic agents. Curr. Opin. Cardiovasc.,
Pulm. Renal Invest. Drugs 2000, 2, 208-215.
(14) Olah, G. A.; Arvanaghi, M.; Ohannesian, L. Synthetic methods
and reactions. 126. Trifluoromethanesulfonic acid/triethylsi-
lane: a new ionic hydrogenation reagent for the reduction of
diaryl and alkyl aryl ketones to hydrocarbons. Synthesis 1986,
770-772.
(25) Hornig, E. C. Organic synthesis of n-butoxynitorbenzene. Org.
Synth. 1955, 3 (Collective Volume), 140-141.
JM040215+