Discovery of a highly potent, nonabsorbable apical sodium-dependent bile acid transporter inhibitor (GSK2330672) for treatment of type 2 diabetes

Article abstract of DOI:10.1021/jm400459m

The apical sodium-dependent bile acid transporter (ASBT) transports bile salts from the lumen of the gastrointestinal (GI) tract to the liver via the portal vein. Multiple pharmaceutical companies have exploited the physiological link between ASBT and hepatic cholesterol metabolism, which led to the clinical investigation of ASBT inhibitors as lipid-lowering agents. While modest lipid effects were demonstrated, the potential utility of ASBT inhibitors for treatment of type 2 diabetes has been relatively unexplored. We initiated a lead optimization effort that focused on the identification of a potent, nonabsorbable ASBT inhibitor starting from the first-generation inhibitor 264W94 (1). Extensive SAR studies culminated in the discovery of GSK2330672 (56) as a highly potent, nonabsorbable ASBT inhibitor which lowers glucose in an animal model of type 2 diabetes and shows excellent developability properties for evaluating the potential therapeutic utility of a nonabsorbable ASBT inhibitor for treatment of patients with type 2 diabetes.

Full text of DOI:10.1021/jm400459m

Article
pubs.acs.org/jmc
Discovery of a Highly Potent, Nonabsorbable Apical Sodium-
Dependent Bile Acid Transporter Inhibitor (GSK2330672) for
Treatment of Type 2 Diabetes
Yulin Wu, Christopher J. Aquino, David J. Cowan, Don L. Anderson, Jeff L. Ambroso, Michael J. Bishop,
Eric E. Boros, Lihong Chen, Alan Cunningham, Robert L. Dobbins, Paul L. Feldman, Lindsey T. Harston,
Istvan W. Kaldor, Ryan Klein, Xi Liang, Maggie S. McIntyre, Christine L. Merrill, Kristin M. Patterson,
Judith S. Prescott, John S. Ray, Shane G. Roller, Xiaozhou Yao, Andrew Young, Josephine Yuen,
and Jon L. Collins*
GlaxoSmithKline Research & Development, Five Moore Drive, Research Triangle Park, North Carolina 27709, United States
ABSTRACT: The apical sodium-dependent bile acid trans-
porter (ASBT) transports bile salts from the lumen of the
gastrointestinal (GI) tract to the liver via the portal vein.
Multiple pharmaceutical companies have exploited the
physiological link between ASBT and hepatic cholesterol
metabolism, which led to the clinical investigation of ASBT
inhibitors as lipid-lowering agents. While modest lipid effects
were demonstrated, the potential utility of ASBT inhibitors for treatment of type 2 diabetes has been relatively unexplored. We
initiated a lead optimization effort that focused on the identification of a potent, nonabsorbable ASBT inhibitor starting from the
first-generation inhibitor 264W94 (1). Extensive SAR studies culminated in the discovery of GSK2330672 (56) as a highly
potent, nonabsorbable ASBT inhibitor which lowers glucose in an animal model of type 2 diabetes and shows excellent
developability properties for evaluating the potential therapeutic utility of a nonabsorbable ASBT inhibitor for treatment of
patients with type 2 diabetes.
of both circulating peptides is to initiate signals in beta cells of
the pancreas to enhance glucose-stimulated insulin secretion,
which, in turn, controls glucose concentrations in the
bloodstream.3−5
INTRODUCTION
■
The World Health Organization estimates that 3.4 million
people died from diabetes in 2004 and projects that worldwide
deaths from this debilitating disease will increase by two-thirds
between 2008 and 2030. Approximately 90% of all cases of
diabetes are classified as type 2 diabetes, a condition that arises
when the pancreas produces insufficient insulin and/or tissues
become insulin resistant, resulting in excessively high blood
sugar levels which can eventually lead to severe microvascular
and macrovascular complications as well as death.1
Current guidelines for treatment of type 2 diabetes outline
lifestyle modification including weight loss and increased
physical activity as an initial approach.2 However, this approach
predominantly fails, and patients will require pharmacological
therapies that can include metformin, sulfonylureas, thiazolidi-
nediones, insulin, dipeptidyl peptidase-4 (DPP-4) inhibitors,
and glucagon-like-1 (GLP-1) analogues. While effective in
subsets of patients with type 2 diabetes, these agents have
limitations and side effects. Clearly, type 2 diabetes remains a
significant unmet medical need.
Bile salts have been shown to enhance secretion of peptides
from the gastrointestinal tract that enhance glucose-stimulated
insulin secretion from the pancreatic beta cells. In particular,
bile salts have been shown to promote secretion of GLP-1 from
colonic L cells in a vascularly perfused rat colon model as well
as GLP-1, peptide YY (PYY), and neurotensin in a vascularly
perfused rat ileum model.6,7 In humans, infusion of
deoxycholate into the sigmoid colon produces a rapid and
marked dose responsive increase in plasma PYY and entero-
glucagon concentrations.8 Novel mechanisms that can increase
ileal and colonic bile salt concentrations would be expected to
induce secretion of GLP-1 and other peptides from the distal
gastrointestinal tract, providing a means to elicit an antidiabetic
effect.
Bile acids are synthesized from cholesterol in the liver then,
upon conjugation, released from the gallbladder into the
proximal small intestine after each meal to facilitate digestion of
nutrients, in particular fat, lipids, and lipid-soluble vitamins.9−11
Conjugated bile acids then descend down the small intestine to
the distal ileum, where 90% are reabsorbed into enterocytes via
The incretin peptides GLP-1 and glucose dependent
insulinotropic peptide (GIP) are secreted by L and K
enteroendocrine cells, respectively, from the gastrointestinal
tract into the bloodstream following ingestion of nutrients. This
important physiological response serves as the primary
signaling mechanism between nutrients in the gastrointestinal
tract and peripheral organs. Upon secretion, one of the effects
Received: March 28, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
the apical sodium-dependent bile acid transporter (ASBT; also
known as the ileal bile acid transporter, iBAT). Pharmacological
disruption of bile acid reabsorption with an inhibitor of ASBT
has been shown to increase the conversion of hepatic
cholesterol to bile acids to compensate for fecal loss of bile
acids.12 While ASBT inhibitors have been explored in humans
as lipid-lowering agents (Chart 1),13 the physiological
Thus, we set out to identify a potent, highly soluble,
nonabsorbable inhibitor of ASBT for evaluation in a preclinical
animal model of type 2 diabetes and ultimately humans.
CHEMISTRY
■
Compound 116 served as the key starting material for the lead
optimization program because kilogram quantities of the
compound were available in our chemical stores (Scheme 1).
Not surprisingly, regioselective demethylation of the C8-
methoxy (see Scheme 1, compound 1 for numbering) proved
to be very challenging; however, highly reproducible, scalable
conditions were ultimately identified to affect clean demethy-
lation of 1. Pretreatment of 1 with HCl gas followed by
addition of AlCl3 provided a 1:1 mixture of phenol 2 and the
regioisomeric C7-phenol in 96% yield. Separation of the
regioisomers by chiral column chromatography provided
intermediate phenol 2 in approximately 40% overall yield.
Several key synthetic intermediates were prepared from 2
(Scheme 1). Conversion of 2 to the corresponding triflate 3
was uneventful. Pd-catalyzed carbonylation in the presence of
methanol provided methyl ester 4, which was subjected to basic
hydrolysis using LiOH to give the corresponding C8-carboxylic
acid. Triflate 3 was also converted to C8-nitrile following
subjection to ZnCN2 in the presence of Pd2(dba)3.17 DiBAL-H
mediated reduction of 6 followed by hydrolysis of the
intermediate imine generated aldehyde 7. Nitrile 6 was also
reduced to aminomethyl intermediate 8 upon exposure to H2
atmosphere at 40 psi in the presence of 10% Pd/C.
Chart 1. Chemical Structures of Select ASBT Inhibitors That
Have Advanced to Clinical Development
consequences of increasing the concentrations of bile salts in
the colon with an ASBT inhibitor on glucose homeostasis has
not been extensively explored. Data from exploratory studies in
preclinical animal models of type 2 diabetes with the ASBT
inhibitor 264W94 (1) or other inhibitors supports this strategy
as a potential target for treatment of this debilitating
disease.14,15 Unfortunately, 1 is not an ideal clinical tool for
testing this hypothesis in humans because the compound is
extensively metabolized following oral administration, an
undesirable profile which did not lead to maximal concen-
trations of 1 at the primary site of action in the distal ileum.
A small number of C8-phenolic ethers were prepared via
base-catalyzed alkylation of 2 with haloalkyl esters, haloalkyl
sulfonic acids, haloalkyl phosphonates, dimethyl α-(4-
chlorobenzenesulfonyl)methylphosphonate,18 or 1,2-dibromo-
ethane followed by bromide displacement with sulfite or
trialkylphosphonate (Scheme 2). Subsequent LiOH or TMSBr-
mediated deprotection of the alkyl esters or alkyl phosphonates,
respectively, provided compounds 9−12 and 14−15.
Triflate intermediate 3 proved to be a key intermediate for
preparation of analogues containing alkyl and amide-linked
acidic functionalities at C8 (Scheme 3). Heck reaction of 3 with
a
Scheme 1. Preparation of Key Intermediates
a
Reagents and conditions: (a) HCl(g), AlCl3, DCE; (b) Tf2O, pyridine, DCM; (c) Pd2(dba)3, dppf, CO, MeOH, 70 °C; (d) LiOH, THF, MeOH,
H2O; (e) ZnCN2, Pd2(dba)3, dppf, Zno, DMF, 80 °C; (f) DiBAL-H, toluene, DCM; (g) H2, 10% Pd/C, EtOH, 40 psi.
B
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2. General Synthetic Route to Oxygen-Linked Analogues
a
Reagents and conditions: (a) RX, K2CO3, DMSO, 70−100 °C; (b) 1,2-dibromoethane, 1 M KOH, Bu4NBr, DMSO, 100 °C; (c) Na2SO3, EtOH,
100 °C; (d) 4 N HCl, 1,4-dioxane, DCM or LiOH, THF, MeOH, H2O; (e) TMSBr, DCM; (f) (EtO)2P(O)CH2OSO2-(4-Cl-Ph), NaHMDS, THF,
DMSO.
a
Scheme 3
a
Reagents and conditions: (a) ethyl acrylate, Pd(PPh3)2Cl2 Et3N, DMF, 120 °C; (b) H2, 10% Pd/C, EtOH; (c) LiOH, THF, MeOH, H2O; (d)
RNH2, HATU, DIPEA, DCM or EDC, DMAP, DCM; (e) BH3-THF, THF; (f) PPh3, CBr4, THF; (g) Na2SO3, EtOH, H2O, Δ; (h) MsCl, Et3N,
DCM; (i) NaCN, DMSO, 60 °C; (j) 37% aq HCl, 100 °C; (k) HP(O)(OEt)2, NaH, THF; (l) TMSBr, DCM.
ethyl acrylate in the presence of palladium catalyst at elevated
temperature followed by hydrogenation of the intermediate
ethyl cinnamate and saponification with LiOH provided
compound 16. Subjection of carboxylic acids 5 and 16 to
substituted amines and HATU with base or EDC with DMAP
led directly to amide analogues 24−25 and 27. Amide
analogues possessing an ester moiety were saponified with
LiOH to give derivatives containing a terminal carboxylic acid
(compounds 23, 26, and 32−35). Reduction of 16 with BH3-
THF provided alcohol 16a, which was converted to compound
19 upon bromination then reaction with Na2SO3. Mesylate
16b, derived from 16a using standard conditions, was reacted
with NaCN then subjected to acidic hydrolysis to give
carboxylic acid 17. In a similar fashion, the corresponding
phosphonic acid 22 was also prepared from 16b upon
conversion to a phosphonate ester followed by deprotection.
Additional amide and amine-linked analogues containing
acidic functionalities were readily prepared from benzylamine
intermediate 8 (Scheme 4). Compound 28 was prepared from
8 upon exposure to methyl chlorooxoacetate and Et3N followed
by saponification of the methyl ester. Acylation of 8 with
haloalkyl acid chlorides in the presence of base led to
intermediate 8b, which was subsequently treated with
Na2SO3 at elevated temperature to give sulfonic acid analogues
29−30. Phosphonic acid analogue 31 was synthesized in a
similar fashion using P(OEt)3 as nucleophile followed by
deprotection with TMSBr. Intermediate 8b was also treated
with KI and a substituted amine to give, following base-
catalyzed saponification, dicarboxylic acid analogue 36. Amine-
linked analogues containing one (47) or two (54) terminal
sulfonic acid moieties were generated upon heating in the
presence of bromoethylsulfonic acid sodium salt, and similarly,
C
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 4
a
Reagents and conditions: (a) ROCOCOCl, Et3N, DCM; (b) LiOH, THF, H2O; (c) Cl(CH2)xCOCl, pyridine, DCM; (d) Na2SO3, EtOH, H2O;
(e) P(OEt)3, 135 °C; (f) TMSBr, DCM; (g) R1NH2, KI, K2CO3, DMF, 60 °C; (h) Na+−O3SCH2CH2Br, DMF, 70 °C; (i) paraformaldehyde,
HP(O)(OEt)2, THF or toluene, 70 °C.
the corresponding analogues containing one (48) or two (55)
phosphonic acids were synthesized following condensation with
formaldehyde in the presence of diethyl phosphate, then
TMSBr-catalyzed deprotection.19
Reductive amination of aldehyde 7 using several substituted
aminocarboxylic and aminophosphonic acid starting materials
under standard reaction conditions followed by deprotection
(LiOH, HCl, or TMSBr as appropriate) provided compounds
37−46, 49−51, and 56 in a highly efficient manner (Scheme
5). Mannich reaction of 7 with diethyl malonate in the presence
of piperidine followed by base-catalyzed Michael addition of
diethyl malonate then acidic hydrolysis led to the formation of
compound 52. Condensation of 7 with tetraethyl
methanediylbis(phosphonate) in the presence of NaH followed
by Pd0-catalyzed olefin hydrogenation then deprotection
provided compound 21.
Two-step conversion of methyl ester intermediate 4 to
bromomethyl moiety (4a) was achieved upon reduction with
DiBAL-H to the alcohol followed by conversion to the bromide
upon addition to a premixed solution of PPh3, Br2, and
imidazole at 0 °C (Scheme 6).20 Importantly, use of low
temperature was critical for minimizing degradation of the
sensitive benzyl bromide. Bromide 4a was aminated upon
heating in the presence of diethyl iminodiacetate then
saponified to give compound 53. Compounds 18 and 20
were readily synthesized from 4a upon exposure to Na2SO3 and
P(OEt)3, respectively, with deprotection of the phosphonate
ester being achieved via acidic hydrolysis.
a
Scheme 5
RESULTS AND DISCUSSION
■
All compounds were initially evaluated in a series of assays
beginning with a 3H-taurocholate uptake assay in HEK293 cells
expressing human ASBT. In general, compounds showing
potencies comparable to or better than 1 were evaluated for
cellular permeability using a MDR1-MDCK assay. Compounds
with low to medium permeability (Papp < 15 nm/s) were
subsequently evaluated in vivo in fed male SD rats for
determination of maximum portal vein and systemic blood
parent drug concentrations (Cmax) following a single oral dose
at 10 mg/kg. The data from profiling of potential ASBT
inhibitors in these assays are summarized in Table 1. Consistent
with previous data, 1 was a modestly potent ASBT inhibitor
with good cellular permeability but showed unexpectedly low
portal vein drug levels (Cmax = 100 ng/mL) in vivo, which may
be due to extensive intestinal and hepatic first pass metabolism
(unpublished data). These results and other unpublished data
led us to initiate a lead optimization effort to identify a
nonabsorbable derivative of 1 with improved ASBT potency,
a
Reagents and conditions: (a) RNH2, NaHB(OAc)3, AcOH, DCE;
(b) LiOH, THF, MeOH, H2O; (c) 4 N HCl, 1,4-dioxane; (d) TMSBr,
DCM; (e) piperidine, diethylmalonate, toluene, 100 °C; (f) diethyl
malonate, NaOEt, EtOH; (g) 37% HCl, Δ; (h) (EtO)2P(O)CH2P-
(O)(OEt)2, NaH, THF; (i) H2, 10% Pd/C, EtOH.
D
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 6
a
Reagents and conditions: (a) DiBAL-H, toluene; (b) PPh3, Br2, imidazole, DCM, 0 °C; (c) diethyl iminodiacetate, CH3CN, 65 °C; (d) LiOH,
THF, H2O; (e) Na2SO3, 1,4-dioxane, H2O; (f) P(OEt)3, toluene; (g) 6 N HCl, EtOH, Δ.
Table 1. C(8) Oxygen-Linked and Carbon-Linked ASBT Inhibitor Potency and Rat Portal Vein Drug Concentrations
a
b
b
compd
X
R
hASBT IC50 (nM)
180 55
MDCK Papp (nm/s)
portal Cmax (ng/mL)
systemic Cmax (ng/mL)
1
O
Me
107
10
100
1458
NT
220
289
−
19
26
−
7
9
O
HO2CCH2−
34
13
4
5
1
3
10
11
12
13
14
15
16
17
18
19
20
21
22
O
HO2C(CH2)3−
HO3S(CH2)2−
HO3S(CH2)3−
H2O3PCH2−
H2O3P(CH2)CH2−
H2O3P(CH2)3−
HO2CCH2−
HO2C(CH2)2−
HO3S−
HO3S(CH2)2−
H2O3P−
303
12
O
O
50 25
900 150
44 28
8
1
c
O
NT
−
16
7
O
8
10
383
342
9
82
O
133 60
17 19
260
NT
NT
170
NT
14
CH2
CH2
CH2
CH2
CH2
CH2
CH2
−
−
3
2
4
2
1
1
1
4
4
7
−
3
23
51
7
H2O3PCH2−
H2O3P(CH2)2−
12
NT
−
−
51 14
7
NT
a
b
c
IC50 values are expressed as the mean standard error of ≥2 replicates. dosed via oral gavage at 10 mg/kg. NT or − means not tested.
decreased cellular permeability, improved metabolic stability in
the gastrointestinal tract, and low (<50 ng/mL) portal vein
parent drug levels when dosed orally at 10 mg/kg.
improvement in ASBT inhibition potency and low cellular
permeability. Further evaluation of 9 in a rat portal vein study
revealed the compound to be highly absorbed into the portal
vein as parent or the corresponding glucuronide with drug
levels surpassing 1 μM. Insertion of two additional methylene
groups between the acid and methoxy carbon as in 10 led to
further improvement in potency, however, the additional
lipophilicity led to a significant increase in cellular permeability
in the MDCK assay. Analogues containing more acidic sulfonic
and phosphonic acid moieties were targeted because these
modifications would lead to a higher percentage of sulfonate or
phosphonate anion in the intestine, which would be predicted
to decrease passive cellular permeability. While the direct
phosphonic acid analogue (13) of 9 showed significantly
reduced ASBT potency, analogues with extended alkyl chains
containing terminal sulfonic or phosphonic acids (compounds
11−12, 14−15) showed improved ASBT potency and
decreased cellular permeability relative to 1 but did not show
significantly reduced portal vein drug concentrations. Overall,
while improved potencies were observed with several
We initiated our optimization efforts by exploring the SAR of
the 5-aryl and 7,8-dimethoxy moieties, however, our focus
quickly shifted to modifications of the 7- and 8-methoxy
substituents because highly potent compounds with low cellular
permeability were rapidly identified. While a full description of
the SAR of the 7 and 8-dimethoxy substituents is beyond the
scope of this article, data with analogues containing structural
modifications of the 8 position clearly revealed that this
position was more tolerant of structural modifications and led
to compounds with superior ASBT inhibitory activity.
Numerous 8-substituted derivatives of 1 were synthesized
containing a range of functional groups to explore whether
basicity, acidity, lipophilicity, ionization state, and conforma-
tional rigidity impacted ASBT potency and cellular perme-
ability. However, compounds with acidic moieties consistently
showed potent ASBT inhibition and, in general, low to
moderate cellular permeability (Table 1). Substitution of the
8-methoxy carbon with a carboxylic acid (9) led to a 6-fold
E
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. C(8) Carbon-Linked ASBT Inhibitor Potency, Permeability, And Rat Portal Vein Drug Concentrations
a
b
b
compd
R
hASBT IC50 (nM)
32 21
MDCK Papp (nm/s)
portal Cmax (ng/mL)
systemic Cmax (ng/mL)
c
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
HO2CC(Me)2CH2NHCO−
HO3SCH2NHCO−
HO3S(CH2)2NHCO-
HO2CCH2NHCO(CH2)2−
HO3SCH2NHCO(CH2)2−
HO2CCONHCH2−
HO3SCH2CONHCH2−
HO3S(CH2)2CONHCH2−
H2O3PCH2CONHCH2−
(HO2CCH2)2NCO−
(HO2CCH2)2CHNHCO−
(HO2CCH2)2CHNHCO(CH2)2−
(HO2CCH2)2NCO(CH2)2−
(HO2CCH2)2CHNHCH2CONHCH2−
HO2C(CH2)3NHCH2−
HO2C(CH2)2NHCH2−
HO2CC(Me)2NHCH2−
(L)-N-Pro-CH2−
43
7
NT
−
4
11
2
1
92
132
NT
NT
87
122
−
1
4
38
−
−
−
26
−
−
−
−
−
−
−
−
−
−
−
−
−
−
24
1
276 39
167 25
12
7
6
9
1
1
NT
19
NT
3
128 102
249 29
2540
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
5
−
−
−
−
−
−
−
−
−
−
−
−
12
27
224
7
97 61
>3000
728 44
>3000
1287
367 25
190
8
>3000
(D)-N-Pro-CH2−
(L)-α-N-Lys-CH2−
(L)-N-Thr-CH2−
(L)-N-Ser-CH2−
751
1458
425 248
59 34
HO2CCH2NHCH2−
4
1
27
10
10
NT
NT
7
HO2CCH2N(Me)CH2−
HO3S(CH2)2NHCH2−
H2O3PCH2NHCH2−
H2O3P(CH2)2NHCH2−
(D)-N-Asp-CH2−
41 33
54
1
6
0
>3000
2698 2225
55 11
−
−
−
−
1
11
93
333
56
NT
−
(L)-N-Asp-CH2−
47
6
6
1
(HO2CCH2)2CH−
>3000
245 21
365 79
>3000
5
19
43
−
−
0
(HO2CCH2)2NCH2−
(HO3SCH2CH2)2NCH2−
(H2O3PCH2)2NCH2−
(HO2CCH2)2CHNHCH2−
4
8
NT
7
42
3
4
a
b
c
IC50 values are expressed as the mean standard error of ≥2 replicates. Dosed via oral gavage at 10 mg/kg. NT or − means not tested.
analogues, almost all of the analogues were highly absorbed
from the GI tract into the portal vein.
(21 vs 13, 22 vs 14), and of these, compound 20 showed low
oral absorption (portal Cmax = 14 ng/mL) when tested in a rat
portal vein study. Of all of the 8-oxygen and carbon analogues,
compound 20 had the best overall profile and was selected for
further evaluation in more advanced in vitro and in vivo assays
(vide infra).
Compounds containing a lipophilic linker between the aryl
ring and a terminal acid moiety were highly potent ASBT
inhibitors with highly variable cellular permeability; however,
portal vein drug levels for the majority of these derivatives were
higher than the targeted threshold. Our strategy for maintaining
potency while decreasing cellular permeability was to conserve
the terminal acid group but incorporate more polar atoms
within the 8-linker region. Toward this end, a limited number
of compounds containing terminal carboxylic, sulfonic, and
phosphonic acids with benzamide, benzylamide, and prop-
anamide linkers were synthesized (Table 2). ASBT potency was
Early in the course of our optimization effort, data emerged
from exploratory intestinal stability studies with several C8
oxygen-linked compounds showing the presence of the
corresponding 8-phenol (compound 2) as a metabolite. As a
result, the optimization effort shifted focus to evaluation of
terminal acid analogues containing a methylene instead of the
8-oxygen. With carboxylic acid analogues, this modification led
to a 2−6-fold improvement in ASBT potency (compounds 16
vs 9 and 17 vs 10) but did not decrease cellular permeability. In
contrast, two sulfonic acid analogues (18, 19) were highly
potent and showed low permeability, however, portal vein drug
levels for 18 were higher than the targeted threshold of 50 ng/
mL. Structurally similar phosphonic acid compounds (20, 21,
22) showed equivalent or better ASBT potency and cellular
penetration as compared to the oxygen-containing analogues
F
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Effects of Nonabsorbable ASBT Inhibitors in Rodent ASBT Inhibition, Intestinal Stability, and Fecal Bile Acid
Excretion Assays
a
b
c
d
compd
R
mouse/rat ASBT IC50
rat fecal drug recovery (%)
rat intestinal contents stability (h)
mFBA excretion ED50
e
f
1
OMe
3/22
3
87
h
0.13 (100)
20
44
45
47
50
56
H2O3PCH2
0.5/0.9
>12
8
0.12 (92)
0.012 (47)
0.017 (98)
2.7 (73)
g
(L)-N-Ser-CH2−
NT
63
HO2CCH2NHCH2−
HO3S(CH2)2NHCH2−
(D)-N-Asp-CH2−
0.3/0.2
0.9/2.3
NT
92
>12
>12
NT
>12
100
100
98
>1 (70)
(HO2CCH2)2CHNHCH2−
2.1/1.9
0.022 (94)
a
b
IC50 values are in nM and are expressed as the mean of ≥2 replicates (variation <25%). Percentage of parent compound in rat feces following a 10
mg/kg oral dose. Data are expressed as t1/2 of parent compound. Mouse fecal bile acid excretion (mFBA) data are in mg/kg, numbers in
parentheses are relative efficacy versus compound 1. The corresponding phenol metabolite (R = OH) was detected at increasing concentrations at
2, 4, and 24 h. Active metabolites of compound 1 undergo enterohepatic recirculation in mice. NT means not tested.
c
d
e
f
g
above 100 nM for compounds containing propanamide (26,
27) or benylamide (30, 31) linkers. In contrast, compounds
with a decreased overall linker length such as benzamides (23−
25) and benzylamides (28, 29) were highly potent ASBT
inhibitors with low to moderate cellular permeability.
Unfortunately, exemplars (24, 25, 28, 29) with good ASBT
potency and low cellular permeability showed high oral
absorption in the rat portal vein assay. In an attempt to
decrease portal vein drug levels, a small set of amide analogues
(32−36) with more polar dicarboxylic acids were evaluated,
however, the majority were very weak ASBT inhibitors, leading
us to abandon this approach.
The cumulative data suggested the distance between the
terminal acidic functionality and the aryl ring was critical for
potent inhibition of ASBT, as compounds with a total of 6 or
more contiguous atoms (26, 27, 34−36) at the 8 position led
to weak inhibition of ASBT whereas compounds with 4 or 5
(for example compounds 24, 28, 33) contiguous atoms were
potent ASBT inhibitors. Recognizing this general trend, our
strategy shifted toward synthesis and evaluation of several
substituted amino acid-derived analogues containing 4−5
contiguous atoms with an amine as one of these atoms. Our
rationale for insertion of an amine was to explore whether this
basic atom would impact ASBT activity and, more importantly,
decrease cellular permeability and oral absorption (Table 2,
38−49). β-Aminoglycine (38) and α-aminoisobutyryl deriva-
tives (39) showed moderate to weak inhibition of ASBT while
proline-derived analogues with reduced conformational flexi-
bility (40, 41) were very weakly active. Consistent with SAR
from our initial exploratory studies, substitution with a basic
lysine (42) led to very weak ASBT inhibition whereas neutral
threonine (43) and serine (44) analogues were active, with the
latter being a potent ASBT inhibitor with low cellular
permeability and portal vein drug levels. Removal of the α-
amino acid side chain to give a glycine analogue (45) resulted
in a highly potent analogue with very low portal vein and
systemic drug levels despite showing moderate permeability in
the MDCK assay. Not surprisingly, we did not observe a linear
correlation between the in vitro MDCK permeability and in
vivo portal vein drug level assays. While the MDCK assay did
effectively identify compounds with moderate to high cellular
permeability, data from the in vivo assay was viewed as more
physiologically relevant for making decisions about compound
progression. N-Methylation of 45 to give compound 46 led to a
potent ASBT inhibitor, however, portal vein drug levels
increased substantially despite low cellular permeability in the
MDCK assay. A β-aminosulfonic acid analogue (47) showed
potent ASBT inhibitory activity as well as low cellular
permeability and oral absorption, whereas structurally related
α and β-aminophosphonic acid derivatives (48, 49) were
significantly less potent than the corresponding sulfonic acid
analogue (47), a trend that was also observed with amide-
linked analogues 29 versus 31. Two diastereomeric aspartic
acid-derived analogues (50, 51) potently inhibited ASBT and
showed low cellular permeability, however, only the D-
diastereomer showed low oral absorption when tested in
vivo. The data with diacids 50 and 51 prompted us to explore
other diacid-containing analogues 52−56. Not surprisingly,
diacid 52 lacking an amine did not inhibit ASBT. The addition
of an acetic acid moiety to glycine analogue 45 to give
aminodicarboxylic acid 53 resulted in a weakly potent ASBT
inhibitor. Similar data was also observed with a structurally
related aminodisulfonic analogue (54), whereas the respective
aminodiphosphonic analogue (55) was inactive. Speculating
that the weak activities were due to the presence of a tertiary
amine (see 46), a methylene was inserted adjacent to the amine
to give compound 56, resulting in potent inhibition of ASBT
and very low oral absorption in vivo in the rat. Of all the
analogues in Table 2, compounds 44, 45, 47, 50, and 56, along
with compound 20 in Table 1, met our key criteria of potent
ASBT inhibitory activity and low oral absorption in vivo to
warrant further evaluation as potential development com-
pounds.
Prior to evaluation in a rodent model of type 2 diabetes, the
lead compounds were evaluated for mouse and rat ASBT
activity, stability in the gastrointestinal tract as measured by GI-
tract luminal contents and fecal parent drug recovery assays,
and efficacy in a mouse fecal bile acid excretion assay (Table 3).
Compound 1 more potently inhibited mouse and rat versus
human ASBT but was unstable when tested in the rat fecal drug
recovery and intestinal contents stability assays. Despite this
instability, 1 did increase fecal bile acid excretion in mice with
G
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
○
Figure 1. Effects of compound 20 on portal vein plasma bile acids (BA), total GLP-1 (tGLP-1), PYY, and GIP. Compound 20 ( , 1 mg/kg) or
●
vehicle ( ) was administered orally in portal vein cannulated male SD rats at 0800 (time 0) without fasting. Portal vein blood samples were collected
at indicated time points up to 8 h. Plasma BA, tGLP-1, PYY, and GIP were measured as described in methods. *p < 0.05, **p < 0.01, ***p < 0.001.
an ED50 of 0.13 mg/kg, which can be attributed to extensive
first pass metabolism generating active metabolites that
undergo enterohepatic recirculation. Compound 20 potently
inhibited mouse and rat ASBT and was very stable in the GI
luminal contents and fecal parent drug recovery assays.
Consistent with blocking bile acid reuptake in the ileum, 20
increased fecal bile acid excretion in mice with an EC50 of 0.12
mg/kg. In contrast, compound 44 was relatively unstable in the
GI tract, with only 63% of parent drug being recovered in the
fecal recovery assay. While potent in the mouse fecal bile acid
excretion assay, this reduced intestinal stability led to an
approximately 50% reduction in total amount of excreted fecal
bile acids as compared to 20 and 1. Compound 45 potently
inhibited mouse and rat ASBT and showed excellent stability in
the rat GI luminal contents stability and fecal parent drug
that inhibition of bile acid reuptake in the distal small intestine
would increase portal vein concentrations of GLP-1 and PYY.
Oral dosing of 20 at 10 mg/kg to portal vein cannulated lean
SD rats was followed by sampling of portal vein blood for bile
acids, total GLP-1, PYY, and GIP concentrations at specified
time points. As can be seen in Figure 1, inhibition of ASBT led
to a decrease in portal vein concentration of bile acids from 1 to
2 h postdosing relative to vehicle treated animals. Interestingly,
this time frame is consistent with published transit times of
small molecules through the small intestine in rats.21 In contrast
to portal vein bile acid concentrations, portal vein total GLP-1
and PYY concentrations increased with ASBT inhibitor-treated
animals from 3 to 8 h postdosing whereas the concentrations of
a separate upper GI-secreted peptide, GIP, did not change
relative to vehicle treated animals. Agonists of the G-protein
coupled receptor TGR5 have been linked to GLP-1 secretion,22
which could be an alternative mechanism for increasing portal
vein GLP-1 concentrations. However, 52 exemplars from the
1,4-benzothiazepine series did not show activity when tested in
a TGR5 assay up to 10 μM, suggesting that direct TGR5
activation is not the primary mechanism of action for this
chemical series of ASBT inhibitors. These data are consistent
with our hypothesis that inhibition of ASBT with a non-
absorbable compound could stimulate GLP-1 and PYY
secretion from the distal intestines.
Encouraged by the results from the proof of mechanism
studies, we next set out to compare the effects of compounds
20, 45, and 56 on fecal bile acid secretion and glucose
homeostasis in the Zucker Diabetic Fatty (ZDF) rat, an
established preclinical animal model of type 2 diabetes.
Treatment of 8-week-old male ZDF rats with the three
compounds at the indicated doses twice daily for 14 days and
determination of fecal bile acid concentration over the last 24 h
of the dosing period provided the data that is summarized in
Figure 2. As expected, all three compounds stimulated bile acid
excretion to a maximum of ∼7−10 fold, however, in general,
compounds 45 and 56 showed a higher fold increase in fecal
bile acid secretion at the lowest dose (0.05 mg/kg) as
recovery assays. When tested in mice, 45 potently (ED50
=
0.012 mg/kg) increased fecal bile acid excretion and showed
equal or greater percent maximum efficacy as compared to 20
and 1. Despite good mouse and rat ASBT potency and GI
stability, compound 47 was significantly less potent in the
mouse fecal bile acid excretion assay as compared to
compounds 20 and 45. While rodent ASBT activity and
stability data in the GI luminal contents assay were not
collected with compound 50, significantly weaker mouse fecal
bile excretion activity prompted us to abandon this compound
as well as compound 47. The dicarboxylic acid derivative 56
showed potent mouse and rat ASBT activity and was stable in
the in vitro and in vivo GI stability assays. Moreover, 56
potently increased fecal bile acid excretion in mice with activity
comparable to compounds 20 and 45. Overall, these data
indicated that compounds 20, 45, and 56 were stable in the
rodent GI tract and potently induced fecal bile acid excretion in
vivo in mice, leading us to select these three compounds for
mechanistic and efficacy studies in vivo in lean rats and Zucker
Diabetic Fatty (ZDF) rats, respectively.
Prior to in vivo efficacy studies, compound 20 was selected as
an exemplar of a potent, selective, nonabsorbable ASBT
inhibitor for proof of mechanism studies to test our hypothesis
H
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
improvement in glucose homeostasis, maximum efficacy was
comparable with each compound at identical doses which did
not permit the identification of a clearly superior compound.
The cumulative data with compounds 20, 45, and 56
suggested that each compound was a viable clinical candidate;
however, data from preclinical development assays would be
required for selection of a development candidate. The three
compounds were tested in a battery of preclinical development
assays including equilibrium aqueous solubility, in vitro
biotransformation, general selectivity, salt selection, crystal-
linity, stability, hygroscopicity, and reactive metabolite assays.
All three compounds showed acceptable aqueous solubility
(≥200 μg/mL) and were highly selective when screened
against a broad panel of unrelated targets. However, a reactive
metabolite was observed with compound 20 when tested in a
glutathione trapping assay and a crystalline salt derivative of 45
proved to be thermally unstable and extremely hygroscopic,
properties that would slow or possibly preclude further
development of these two compounds. In contrast, the
zwitterionic, nonhygroscopic, crystalline salt form of 56 showed
good aqueous solubility at pH 7.4 (>7 mg/mL), excellent
thermal stability, and did not generate reactive or human-
specific metabolite, characteristics that suggested 56 was a
superior development compound versus 20 and 45.
Figure 2. Effects of ASBT inhibitors at the indicated doses on fecal bile
acid concentrations in ZDF rats. Male ZDF rats were treated orally
twice a day with vehicle, compound 20 (open bar), compound 56
(gray bar), or compound 45 (black bar) for two weeks. Twenty-four h
fecal samples were collected on day 14 and extracted for fecal bile acid
(BA) measurement. Data are presented as relative fold change from
vehicle groups (dashed line). All treatment groups were statistically
different from vehicle groups (p < 0.001).
compared to 20, a trend that was also observed at the next
three higher doses (0.10, 0.50, 1.0 mg/kg). The effects of the
three compounds on glucose homeostasis in ZDF rats were
determined in two separate in vivo studies in which each
compound was given orally at doses ranging from 0.05 to 10
mg/kg (compounds 20 and 45) or 0.001 to 10 mg/kg
(compound 56) twice daily for 14 days. Plasma glucose,
hemoglobin A1c (HbA1c), insulin, and total GLP-1 concen-
trations were determined at the end of the study (Figure 3).
Treatment with the three nonabsorbable ASBT inhibitors led
to a 1.30−1.64% reduction in HbA1c, a greater than 50%
reduction in nonfasted plasma glucose to below 200 mg/dL,
and statistically significant higher plasma insulin, and,
importantly, greater total GLP-1 relative to vehicle treated
animals. While all three compounds showed a significant
In preparation for safety assessment studies, compound 56
was profiled in rat, dog, and monkey PK studies to determine if
low oral absorption translated across multiple preclinical
species. Consistent with our previous data, oral dosing of a
solution of 56 at 10 mg/kg to fed SD rats did not lead to any
detectable (LLQ = 1 ng/mL) levels of parent compound in the
systemic circulation. Similarly, very low systemic levels (Cmax
=
6 ng/mL) of 56 were observed following oral dosing to male
beagle dogs at 10 mg/kg, and importantly, 56 could not be
detected in the systemic circulation following oral dosing of
cynomolgous monkeys at a very high oral dose (3 mg/kg). The
Figure 3. Effects of ASBT inhibitors on blood HbA1c, plasma glucose, insulin, and total GLP-1 in 8-week-old male ZDF rats. Rats were treated orally
○
▲
△
twice a day with compound 20 ( ), 45 ( ), or 56 ( ) at the indicated doses for two weeks in two separate studies. Blood samples were collected
on day 14 from tail vein. Blood HbA1c, plasma glucose, insulin, and total GLP-1 were measured as described in methods. *p < 0.05 vs vehicle.
I
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
gaseous HCl then treated with AlCl3 (200 g, 1.5 mol) in one portion.
The reaction mixture was stirred while slowly cooling to 25 °C then
stirred for 2.5 h. The reaction mixture was added to an ice−H2O
mixture with vigorous stirring. The biphasic mixture was treated with 1
N HCl (∼200 mL) then the phases were separated after stirring for 30
min. The organic phase was isolated, washed twice with dilute HCl
(1.5 L H2O/∼200 mL 1 N HCl), dried over MgSO4, filtered, and
concentrated to dryness to give a regioisomeric mix of 7-OH and 8-
OH products (185 g, 458 mmol, 96% yield) as a white foam. 1H NMR
indicated a 47:53 mix of 7/8-phenols, respectively. The regioisomers
were separated by chiral chromatography [stationary phase (CSP)-
cellulose tris(3,5-dichlorophenylcarbamate) polymer immobilized on
silica (CHIRALPAK IC), DCM and 2-propanol (98/2 v/v) as mobile
data from these PK studies suggested that 56 would show very
low or no oral absorption in humans at pharmacologically
relevant oral doses.
CONCLUSIONS
■
In summary, numerous analogues were designed to probe the
effects of molecular size, polarity, lipophilicity, and ionization
state on ASBT inhibition, cellular permeability, and oral
absorption. Compounds containing acidic moieties at C8 of
the 1,1-dioxido-1,4-benzothiazepine core structure emerged as
highly potent ASBT inhibitors but showed highly variable
cellular permeability and oral absorption, the latter being
determined by measuring portal vein parent drug concen-
trations after oral dosing. Three lead compounds (20, 45, 56)
were identified from this optimization effort which displayed
potent ASBT inhibition and very low oral absorption in rats as
well as robust efficacy in an animal model of type 2 diabetes.
Further characterization of the three compounds in develop-
ability assays culminated in the identification of compound 56
as a highly potent, nonabsorbable ASBT inhibitor with excellent
aqueous solubility, selectivity, and developability properties for
evaluation in safety studies and ultimately humans. Compound
56 will be a valuable clinical tool for exploring the therapeutic
utility of a nonabsorbable ASBT inhibitor for treatment of
patients with type 2 diabetes.
1
phase] to give 72.5 g of 2 as a white solid. H NMR (400 MHz,
DMSO-d6) δ ppm 0.75 (t, J = 7.1 Hz, 3 H), 0.79 (t, J = 7.5 Hz, 3 H),
0.95−1.28 (m, 4 H), 1.31−1.53 (m, 2 H), 1.64−1.76 (m, 1 H), 1.98−
2.10 (m, 1 H), 2.43 (d, J = 9.8 Hz, 1 H), 3.06 (d, J = 14.8 Hz, 1 H),
3.42 (s, 3 H), 3.49 (d, J = 14.8 Hz, 1 H), 5.85 (d, J = 9.8 Hz, 1 H), 6.04
(s, 1 H), 7.24−7.35 (m, 1 H), 7.35−7.48 (m, 5 H), 9.72 (s, 1 H).
(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phenyl-
2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl trifluoromethane-
sulfonate (3). To a DCM solution of 2 (105 g, 26 mmol) and
pyridine (27.3 mL, 34 mmol) cooled to 0 °C under nitrogen was
added Tf2O (57.1 mL, 34 mmol) slowly over 30 min while
maintaining an internal temperature between 5 and 10 °C. Upon
complete addition, the reaction was stirred until TLC and LCMS
indicated complete conversion to product. H2O (250 mL) was slowly
added to the mixture, and the mixture stirred for 10 min, after which
the layers separated. The aqueous layer was extracted an additional
time with DCM, and the combined organics washed with 10% HCl
and brine, then dried (Na2SO4), filtered, and concentrated to half
volume. Hexanes was added until the solution became turbid, and
crystallization began to occur. The solids were then filtered from the
solution to give 3 (134.6 g, 95%) as a white solid, which was used
without further purification. 1H NMR (400 MHz, CDCl3) δ ppm
0.76−0.96 (m, 6 H), 1.03−1.40 (m, 4 H), 1.40−1.67 (m, 4 H), 1.77−
1.97 (m, 1 H), 2.10−2.28 (m, 1 H), 3.07 (d, J = 14.8 Hz, 1 H), 3.48
(d, J = 14.9 Hz, 1 H), 3.62 (s, 3 H), 6.09 (s, 1 H), 6.34 (s, 1 H), 7.32−
7.50 (m, 5 H), 7.96 (s, 1 H). ES-LCMS m/z 536 (M + H)+.
EXPERIMENTAL SECTION
■
General Protocols. All chemical reagents were purchased and
1
used as received. H NMR spectra were recorded on a Varian Unity-
300 or Varian Unity Plus-400. Chemical shifts are expressed in parts
per million (ppm, δ units). Coupling constants are in units of hertz
(Hz). Splitting patterns describe apparent multiplicities and are
designated as s (singlet), d (doublet), t (triplet), q (quartet), m
(multiplet), or br (broad).
Unless otherwise noted, the LCMS system used to determine purity
was a UPLC analysis was conducted on a Waters Acquity system with
BEH C18, 2 mm × 50 mm, 1.7 μm column at 40 °C 95% H2O, 5%
MeCN to 99% MeCN in 1.1 min, holding at 100% MeCN for 40 s.
Water contained 0.2% v/v formic acid. MeCN contained 0.15% v/v
formic acid. The flow rate was 1 mL/min with 5 μL of solution
injected. Mass spectra were recorded utilizing electrospray ionization
(ESI) or atmospheric pressure chemical ionization (APCI) switching
between positive and negative modes with DAD scanning from 210 to
350 nm. All compounds tested were of ≥95% purity.
Methyl (3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-5-phenyl-
2,3,4,5-tetrahydro-1,4-benzothiazepine-8-carboxylate 1,1-Di-
oxide (4). To a reaction tube containing 3 (0.103 g, 0.192 mmol),
Pd2(dba)3 (0.011 g, 0.012 mmol), dppf (0.013 g, 0.024 mmol), and
NEt3 (0.040 mL, 0.288 mmol) in DMF (1 mL) was bubbled in carbon
monoxide gas for a period of 15 min. Anhydrous MeOH (0.039 mL,
0.962 mmol) was added, and the reaction vessel was sealed under a
carbon monoxide atmosphere then heated at 70 °C overnight. The
reaction was cooled to room temperature, and the contents diluted
with Et2O then poured into H2O. The layers were separated, the
aqueous extracted one additional time with Et2O, and the combined
organics washed with brine, dried (Na2SO4), filtered, and concen-
trated. The crude material was chromatographed on silica gel using
Normal phase chromatography was accomplished using Teledyne
Isco equipment using prepacked silica columns.
Reverse phase HPLC for purification of final compounds was
accomplished on a Gilson 845Z series prep HPLC system using a C18
Sunfire 30 mm × 50 mm or 30 mm × 150 mm, 5 μm column using
the gradient described with 0.05% TFA in H2O. The flow rate was 60
or 45 mL/min and the product was collected based on UV detection
at 220 or 254 nm. In some cases, basic conditions were established
using the same Gilson system and a C18 XBridge 30 mm × 50 mm or
30 mm × 150 mm, 5 μm column using the gradient described with
0.2% NH4OH in the H2O. The flow rate was 60 or 45 mL/min and
again the product was collected on UV detection at 220 or 254 nm.
HRMS. High resolution LC-MS (top) and MS-MS spectra. Spectra
were acquired on a Waters qTOF Premiere mass spectrometer
operating in W mode (resolving power ∼15000). A sample volume of
0.2 μL was introduced by flow injection from a Waters nanoAcquity
UPLC. Mobile phase A: H2O + 0.1% formic acid. Mobile phase B:
MeOH + 0.1% formic acid. Wash: 50:50 H2O:MeOH + 0.1% formic
acid. Flow: 0.02 mL/min, 40% A, 60% B. Lock mass: 0.7 mg leucine
enkephalin in 500 mL MeOH. Collision gas: argon.
1
hexanes/EtOAc to give 4 (0.080 g, 91%) as a a white solid. H NMR
(400 MHz, CDCl3) δ ppm 0.75−0.96 (m, 6 H), 0.99−1.36 (m, 4 H),
1.36−1.59 (m, 5 H), 1.86 (ddd, J = 14.4, 12.0, 4.4 Hz, 1 H), 2.07−2.23
(m, 1 H), 3.02 (d, J = 14.8 Hz, 1 H), 3.45 (d, J = 14.9 Hz, 1 H), 3.57
(s, 3 H), 3.86 (s, 3 H), 6.08 (s, 1 H), 6.27 (s, 1 H), 7.29−7.48 (m, 5
H), 8.52 (s, 1 H). ES-LCMS m/z 446 (M + H)+.
(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-5-phenyl-2,3,4,5-tet-
rahydro-1,4-benzothiazepine-8-carboxylic Acid 1,1-Dioxide
(5). H2O (0.333 mL), MeOH (0.333 mL), and THF (1 mL) were
added to 4 (0.046 g, 0.103 mmol) along with LiOH (0.013 g, 0.310
mmol), and the mixture stirred for 2 h at 25 °C. The reaction mixture
was concentrated to half volume, then 6 N HCl was added. The
mixture was extracted with EtOAc (2×), washed with brine, dried
(Na2SO4), filtered, and concentrated in vacuo to give 5 (0.039 g,
1
86%). H NMR (400 MHz, CDCl3) δ ppm 0.67−0.99 (m, 6 H),
(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-5-phenyl-2,3,4,5-tet-
rahydro-1,4-benzothiazepin-8-ol-1,1-dioxide (2). A solution of 1
(200 g, 479 mmol) (see ref 16) in DCE (1 L) was saturated with
0.99−1.64 (m, 7 H), 1.87 (t, J = 11.8 Hz, 1 H), 2.08−2.29 (m, 1 H),
3.09 (d, J = 14.8 Hz, 1 H), 3.46 (d, J = 14.9 Hz, 1 H), 3.70 (br s, 3 H),
J
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
6.11 (br s, 1 H), 6.33 (br s, 1 H), 7.30−7.60 (m, 5 H), 8.73 (br s, 1
H). ES-LCMS m/z 432 (M + H)+.
organics were combined, dried over MgSO4, filtered, and concentrated
to give an additional batch of 9 (0.28 g, 0.61 mmol, 16% yield) as a
white foam. H NMR (400 MHz, DMSO-d6) δ ppm 0.95−1.28 (m,
1
(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-5-phenyl-2,3,4,5-tet-
rahydro-1,4-benzothiazepine-8-carbonitrile-1,1-dioxide (6). A
mixture of 3 (0.535 g, 1.0 mmol), Pd2(dba)3 (0.055 g, 0.06 mmol),
dppf (0.066 g, 0.12 mmol), zinc powder (0.004 g, 0.06 mmol), and
Zn(CN)2 (0.117 g, 0.99 mmol) in DMF (10 mL) was stirred at 25 °C
for 15 min under a stream of nitrogen. The reaction was heated to 80
°C and stirred for 8 h. After the mixture was cooled to 25 °C, DMF
was removed in vacuo, Et2O was added, and the organics were washed
with 2 N NH4OH followed by brine. The organics were dried
(Na2SO4), filtered, and concentrated. The residue was triturated with
hexanes/EtOAc, and a white solid was collected by filtration to give 6
(0.386 g, 92%). 1H NMR (400 MHz, CDCl3) δ ppm 0.73−0.97 (m, 6
H), 0.98−1.62 (m, 8 H), 1.77−1.92 (m, 1 H), 2.08−2.24 (m, 1 H),
3.01 (d, J = 14.9 Hz, 1 H), 3.45 (d, J = 14.9 Hz, 1H), 3.62 (s, 3 H),
6.08 (br s, 1 H), 6.26 (s, 1 H), 7.32−7.49 (m, 5 H), 8.28 (s, 1 H). ES-
LCMS m/z 413 (M + H)+.
(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-5-phenyl-2,3,4,5-tet-
rahydro-1,4-benzothiazepine-8-carbaldehyde 1,1-Dioxide (7).
Compound 6 (7.42 g, 17.99 mmol) dissolved in DCM (150 mL) was
treated with 1 M DIBAL-H in toluene (36.0 mL, 36.0 mmol) at 0 °C
with stirring for 1 h after which time LCMS indicated complete
conversion. The reaction mixture was poured into an ice/1 N HCl
mixture and stirred vigorously for 1 h. The organic phase was isolated,
dried over MgSO4, filtered, and concentrated to give a yellow solid.
The crude product was dissolved in 80 mL hot EtOAc to which was
added hexanes (250 mL) until cloudy. The mixture was allowed to
cool slowly to ambient temperature and then cooled in an ice bath.
The resultant precipitate was filtered off, washed with cold 20%
EtOAc/hexanes, and air-dried to give 7 (4.00 g, 53.5% yield) as a white
solid. The mother liquor was concentrated to dryness, and the residue
was purified on silica gel eluting with 20−40% EtOAc/hexanes to give
additional product as a white solid. 1H NMR (400 MHz, DMSO-d6) δ
ppm 0.74 (t, J = 7.0 Hz, 3 H), 0.81 (t, J = 7.4 Hz, 3 H), 0.96−1.29 (m,
4 H), 1.33−1.54 (m, 2 H), 1.68−1.81 (m, 1 H), 2.01−2.13 (m, 1 H),
2.80 (d, J = 9.8 Hz, 1 H), 3.13 (d, J = 15.0 Hz, 1 H), 3.58 (s, 3 H), 3.63
(d, J = 15.0 Hz, 1 H), 5.98 (d, J = 9.8 Hz, 1 H), 6.27 (s, 1 H), 7.32−
7.40 (m, 1 H), 7.40−7.49 (m, 4 H), 8.23 (s, 1 H), 10.27 (s, 1 H). LC-
MS (ES+) m/z 416 [M + H].
{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}amine
(8). A mixture of 6 (468 mg, 1.13 mmol) and 10% Pd/C (60.4 mg,
0.567 mmol) in EtOH (20 mL) was added 37% aqueous HCl (0.279
mL, 3.40 mmol) and subjected to H2 atmosphere at 40 psi overnight
then filtered. The residue was purified via HPLC (eluting with MeCN/
H2O with 0.05% TFA-H2O and 0.05% TFA-MeCN) to give 8 (621
mg, 79%, TFA salt) as a white solid. 1H NMR (CDCl3) δ ppm 0.90 (t,
J = 7.2 Hz, 3H), 0.97−1.15 (m, 1H), 1.14−1.38 (m, 3H), 1.44−1.75
(m, 2H), 1.80−1.99 (m, 1H), 2.14−2.38 (m, 1H), 3.29 (d, J = 15.0
Hz, 1H), 3.38−3.48 (m, 1H), 3.55 (s, 3H), 3.98- 4.12 (m, 1H), 4.12−
4.27 (m, 1H), 6.17 (s, 1H), 6.27 (s, 1H), 7.31−7.57 (m, 5H), 7.98 (s,
1H), 8.07−8.46 (br s, 2H). LC-MS (ES+) m/z 417 (M + H)+.
2-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)acetic
Acid (9). To a solution of 2 (1.24 g, 3.07 mmol) in DMSO (10 mL)
was added methyl bromoacetate (0.319 mL, 3.38 mmol) and
potassium carbonate (0.85 g, 6.15 mmol). The reaction mixture was
stirred for 2 h at 50 °C. The reaction mixture was diluted with EtOAc,
washed twice with water and once with brine, dried over MgSO4,
filtered, and concentrated to dryness to give the intermediate ester as a
clear oil. The material was dissolved in THF (30 mL) and treated with
1 M aqueous LiOH (30.7 mL, 30.7 mmol) at 25 °C for 16 h, at which
time LCMS indicated complete conversion to the acid. The reaction
mixture was concentrated to remove the THF, washed once with
ether, and acidified with 1 N HCl with stirring and cooling in an ice
bath until a heavy precipitate formed. The mixture was stirred for 30
min at ice bath temperature, filtered, and the cake washed once with
dilute aqueous HCl to give 9 (1.19 g, 2.58 mmol, 84% yield) as a white
solid. The mother liquor was extracted four times with ether and the
4H), 1.31−1.53 (m, 2H), 1.65−1.78 (m, 1H), 1.97−2.09 (m, 1H),
2.55 (d, J = 9.46 Hz, 1H), 3.03 (d, J = 15.04 Hz, 1H), 3.44 (s, 3H),
0.76 (m, 6H), 3.53 (d, J = 15.04 Hz, 1H), 4.70 (s, 2H), 5.88 (d, J =
7.95 Hz, 1H), 6.10 (s, 1H), 7.27−7.44 (m, 6H), 13.12 (br s, 1H),
exchangeable protons (COOH @ 13.12 and NH @ 2.55) were no
longer visible by 1H NMR upon exposure to D2O. LC-MS (ES−) m/z
460 [M − 1]. LC-MS (ES+) m/z 462 [M + H]. HRMS (ES+) calcd for
C24H31NO6S (M + 1) m/z 462.1949, found 462.1950.
4-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)-
butanoic Acid (10). Compound 2 (50 mg, 0.124 mmol) was
dissolved in DMSO (0.5 mL) and treated with 1,1-dimethylethyl 4-
bromobutanoate (27.6 mg, 0.124 mmol) and K2CO3 (51.4 mg, 0.372
mmol) at 100 °C for 2 h, at which time LCMS indicated complete
conversion to the intermediate ester. The reaction mixture was cooled
to ambient temperature and then treated with excess water to
precipitate the intermediate. The supernatant was decanted off, and
the solid residue was washed twice with water and pumped dry to give
the ester. The residue was diluted with DCM (1 mL) and treated with
4 N HCl in dioxane (2 mL, 8.00 mmol) at 25 °C for 2 h, after which
time LCMS indicated complete conversion. The reaction mixture was
concentrated to dryness and purified by RP-HPLC (Sunfire 30 mm ×
150 mm, C18, 5 um, Waters column; 10−100% CH3CN + 0.05% TFA
in H2O + 0.05% TFA over 10 min at 50 mL/min) to give 10 (41 mg,
1
0.083 mmol, 67% yield). H NMR (400 MHz, DMSO-d6) δ ppm
0.67−0.86 (m, 6H), 0.95−1.30 (m, 4H), 1.33−1.56 (m, 2H), 1.64−
1.84 (m, 1H), 1.87−1.98 (m, 2H), 1.98−2.16 (m, 1H), 2.36 (t, J =
7.31 Hz, 2 H), 3.04−3.19 (m, 1H), 3.44 (br s, 3H), 3.49−3.63 (m,
1H), 3.96−4.07 (m, 2H), 5.81−6.05 (m, 1H), 6.05−6.21 (m, 1H),
7.22−7.56 (m, 6H), COOH proton not observed. LC-MS (ES−) m/z
488 [M − 1]. LC-MS (ES+) m/z 490 [M + H]. HRMS (ES+) calcd for
C26H35NO6S (M + 1) m/z 490.2263, found 490.2263.
2-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)-
ethanesulfonic Acid (11). A solution of 2 (250 mg, 0.620 mmol)
and 1,2-dibromoethane (1.16 g, 6.20 mmol) in THF was treated with
1 M KOH (9.29 mL, 9.29 mmol) and tetrabutylammonium bromide
(60 mg, 0.186 mmol) at 100 °C for 16 h in a sealed tube apparatus,
after which time LCMS indicated complete conversion to the
intermediate bromide. The reaction mixture was partitioned between
EtOAc and water, and the organic phase was isolated and dried over
MgSO4. The mixture was filtered, and the filtrate was concentrated to
dryness and pumped dry overnight to give the intermediate as an
amber oil. The residue was dissolved in EtOH (6 mL) and added to a
solution of sodium sulfite (78 mg, 0.620 mmol) in water (6 mL). The
mixture was heated at 100 °C for 16 h with stirring, after which time
LCMS indicated complete conversion to the desired sulfonic acid. The
reaction mixture was concentrated to dryness and the residue was
purified by RP-HPLC (30 mm × 150 mm Sunfire column at 50 mL/
min; CH3CN and H2O + 0.05% TFA solvent system, solvent gradient
consisted of 10−100 over 8 min, 100 to 100 to 10 min, sample
collection at λmax) to give 11 (210 mg, 0.410 mmol, 66% yield) as a
1
white foam. H NMR (DMSO-d6) δ ppm 0.75−0.91 (m, 6H), 0.89−
1.06 (m, 1H), 1.15−1.39 (m, 3H), 1.48−1.63 (m, 1H), 1.81−2.09 (m,
2H), 2.39−2.59 (m, 1H), 2.96 (td, J = 7.6, 2.0 Hz, 2H), 3.51 (s, 3H),
3.54 (d, J = 15.6 Hz, 1H), 4.01 (d, J = 15.6 Hz, 1H), 4.26 (dd, J = 8.2,
7.0 Hz, 2H), 6.37 (s, 1H), 6.22 (s, 1H), 7.45−7.66 (m, 6H), 2
exchangeable protons not observed. LCMS (ES−) m/z 510 [M − 1].
LC-MS (ES+) m/z 512 [M + H]. HRMS (ES+) calcd for
C24H33NO7S2 (M + 1) m/e 512.1777, found 512.1777.
3-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)-
propane-1-sulfonic Acid (12). Compound 2 (50 mg, 0.124 mmol)
was dissolved in DMSO (0.5 mL) and treated with 3-bromo-1-
propanesulfonic acid sodium salt (30.8 mg, 0.136 mmol) and K2CO3
(34.2 mg, 0.248 mmol) at 100 °C for 2 h, after which time LCMS
indicated complete conversion. The reaction mixture was diluted with
K
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
water, acidified with 1 N HCl, and extracted with EtOAc (3×). The
organic layers were combined, dried over MgSO4, filtered, and
concentrated to dryness to give 12 (22 mg, 0.042 mmol, 34% yield) as
a clear glass. The majority of the product, however, remained in the
aqueous phase. The aqueous layer was stored overnight in the
refrigerator, and the resultant precipitate was filtered off, washed with
water, and air-dried to give a second batch 12 (21 mg, 0.040 mmol,
32% yield) as a white solid. 1H NMR (400 MHz, DMSO-d6 with D2O)
δ ppm 0.70−0.86 (m, 6H), 0.88−1.03 (m, 1H), 1.13−1.34 (m, 3H),
1.48 (dd, J = 13.75, 7.09 Hz, 1H), 1.69−1.83 (m, 1H), (1.84−1.95 (m,
1H), 2.01 (qd, J = 6.88, 6.66 Hz, 2H), 2.25−2.40 (m, 1H), 2.62 (t, J =
7.41 Hz, 2H), 3.37 (d, J = 15.26 Hz, 1H), 3.46 (s, 3H), 3.68−4.23 (m,
3H), 6.09 (br s, 1H), 6.28 (s, 1H), 7.36−7.60 (m, 6H), exchangeable
protons were not observed. LC-MS (ES−) m/z 524 [M − 1]. LC-MS
(ES+) m/z 526 [M + H]. HRMS (ES+) calcd for C25H35NO7S2 (M +
1) m/z 526.1934, found 526.1933.
((((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phenyl-
2,3,4,5-tetrahydrobenzo[f][1,4]thiazepin-8-yl)oxy)methyl)-
phosphonic Acid (13). Step 1: A solution of 2 (100 mg, 0.248
mmol) in DMSO (3 mL) was treated with 1 M NaHMDS in THF
(0.260 mL, 0.260 mmol) at ambient temperature for 20 min.
[Bis(ethyloxy)phosphoryl]methyl 4-chlorobenzenesulfonate (140 mg,
0.408 mmol) was added to the reaction mixture, and the mixture was
stirred at 25 °C for 16 h, after which time LCMS indicated ∼60%
conversion to the desired product. The reaction mix was heated at 60
°C for 3 days, after which time LCMS indicated no discernible
improvement. The reaction mixture was partitioned between EtOAc
and water, and the organic phase was isolated, dried over MgSO4,
filtered, and concentrated to dryness. The residue was purified on silica
gel eluted with 40−100% EtOAc/hexanes to give an intermediate
phosphonate as a clear oil. LC-MS (ES+) m/z 554 [M + H]. 1H NMR
(400 MHz, DMSO-d6) δ ppm 0.76 (t, J = 7.00 Hz, 3H), 0.80 (t, J =
7.42 Hz, 3H), 0.97−1.29 (m, 4H), 1.20−1.28 (m, 6H), 1.32−1.53 (m,
2H), 1.67−1.80 (m, 1H), 1.97−2.10 (m, 1H), 2.55 (d, J = 9.6 Hz, 1
H), 3.06 (d, J = 14.8 Hz, 1H), 3.45 (s, 3H), 3.56 (d, J = 15.1 Hz, 1H),
4.10 (m, 4H), 4.40−4.56 (m, 2H), 5.91 (d, J = 9.6 Hz, 1H), 6.12 (s, 1
H), 7.28−7.36 (m, 1H), 7.36−7.48 (m, 4H), 7.60 (s, 1H).
phosphite (5.26 mL, 30.1 mmol) and heated at 100 °C for 16 h with
stirring, after which time LCMS indicated ∼20% conversion. The
temperature was increased to 130 °C for 16 h, after which time LCMS
indicated complete conversion to the desired product. The reaction
mixture was concentrated to dryness, and the residue was purified on
silica gel eluting with 60−100% EtOAc/hexanes to give an
intermediate phosphonate (112 mg, 0.196 mmol, 79% yield) as a
1
white foam. H NMR (400 MHz, DMSO-d6) δ ppm 0.76 (t, J = 7.05
Hz, 3 H), 0.80 (t, J = 7.42 Hz, 3 H), 0.96−1.30 (m, 4 H), 1.23 (t, J =
6.87 Hz, 6 H), 1.32−1.53 (m, 2 H), 1.66−1.79 (m, 1 H), 1.98−2.11
(m, 1 H), 2.21−2.37 (m, 2 H), 2.51−2.56 (m, 1 H), 3.11 (d, J = 14.83
Hz, 1 H), 3.43 (s, 3 H), 3.55 (d, J = 14.83 Hz, 1 H), 3.94−4.10 (m, 4
H), 4.10−4.26 (m, 2 H), 5.90 (d, J = 9.52 Hz, 1 H), 6.11 (s, 1 H),
7.25−7.36 (m, 1 H), 7.36−7.44 (m, 4 H), 7.46 (s, 1 H).
Step 2: A solution of intermediate phosphonate (111.5 mg, 0.196
mmol) in DCM (1 mL) was treated with TMSBr (120 mg, 0.786
mmol) at 23 °C for 16 h, after which time LCMS indicated complete
conversion. The reaction mixture was concentrated to dryness and
purified by RP-HPLC (basic conditions) to give 14 (37 mg, 0.073
mmol, 37% yield) as a white solid. 1H NMR (400 MHz, DMSO-d6) δ
ppm 0.71−0.84 (m, 6 H), 0.97−1.28 (m, 4 H), 1.32−1.53 (m, 2 H),
1.66−1.91 (m, 3 H), 1.96−2.09 (m, 1 H), 2.51−2.55 (m, 1 H), 3.11
(d, J = 14.83 Hz, 1 H), 3.42 (s, 3 H), 3.53 (d, J = 15.04 Hz, 1 H),
4.02−4.17 (m, 2 H), 5.90 (d, J = 9.67 Hz, 1 H), 6.08 (s, 1 H), 7.31 (m,
J = 8.65, 4.27, 4.27, 4.08 Hz, 1 H), 7.40 (d, J = 4.51 Hz, 4 H), 7.44 (s,
1 H) (acidic protons not observed). LC-MS (ES−) m/z 510 [M − 1].
LC-MS (ES+) m/z 512 [M + H]. HRMS (ES+) calcd for
C24H40N3O7PS (M + 1) m/z 512.1872, found 512.1871.
(3-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)propyl)-
phosphonic Acid (15). Prepared as in compound 14. The crude
material was purified by RP-HPLC to give 15 (61 mg, 0.117 mmol,
1
59% yield) as a white foam. H NMR (400 MHz, DMSO-d6) δ ppm
0.69−0.93 (m, 6 H), 0.97−1.33 (m, 4 H), 1.34−1.57 (m, 2 H), 1.58−
1.71 (m, 2 H), 1.71−1.85 (m, 1 H), 1.84−1.98 (m, 2 H), 1.99−2.17
(m, 1 H), 3.00−3.26 (m, 1 H), 3.46 (s, 3 H), 3.52−3.71 (m, 1 H),
3.99−4.15 (m, 2 H), 5.70−6.05 (m, 1 H), 6.05−6.42 (m, 1 H), 7.01−
7.85 (m, 6 H) (exchangeables were not observed). LC-MS (ES−) m/z
524 [M − 1]. LC-MS (ES+) m/z 526 [M + H]. HRMS (ES+) calcd for
C25H36NO7PS (M + 1) m/z 526.2028, found 526.2024.
Step 2: A solution of the intermediate phosphonate (77.9 mg, 0.141
mmol) in DCM (3 mL) was treated with TMSBr (100 mg, 0.653
mmol) at 23 °C for 16 h, after which time LCMS indicated complete
conversion. The reaction mixture was concentrated to dryness, chased
with MeOH, and purified by RP-HPLC (30 mm × 150 mm Sunfire
column at 50 mL/min. CH3CN and H2O + 0.05% TFA were used as
solvent system. The samples were collected at λmax. Gradient consisted
of: 10−100 over 8 min, 100 to 100 to 10 min) to give 13 [70 mg,
3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]propanoic
Acid (16). Step 1: In a sealed tube, a solution of 3 (3.5 g, 6.53 mmol)
in DMF (30 mL) was treated with Et3N (4.55 mL, 32.7 mmol),
followed by ethyl acrylate (3.56 mL, 32.7 mmol) and bis-
(triphenylphosphine)palladium(II) chloride (0.459 g, 0.653 mmol).
The reaction mixture was stirred at 120 °C overnight, cooled to 25 °C,
and partitioned between H2O and EtOAc. The organic layer was
washed with saturated brine, dried (Na2SO4), filtered, and
concentrated under reduced pressure. Purification on silica gel
(EtOAc/hexanes = 1:6 to 2:1) to give a crude cinnamate ester (3.06
g, 94%) as a white solid. 1H NMR (CDCl3): δ ppm 0.81 (t, J = 7.0 Hz,
3H), 0.88 (t, J = 7.4 Hz, 3H), 1.03−1.21 (m, 2H), 1.28−1.35 (m, 3H),
1.38−1.54 (m, 3H), 1.76−1.93 (m, 1H), 2.08−2.26 (m, 1H), 3.02 (d, J
= 14.9 Hz, 1H), 3.44 (d, J = 14.9 Hz, 1H), 3.55 (s, 3H), 4.18−4.30 (m,
2H), 6.06 (d, J = 6.8 Hz, 1H), 6.19 (s, 1H), 6.56 (d, J = 16.0 Hz, 1H),
7.30−7.48 (m, 5H), 7.85 (d, J = 16.2 Hz, 1H), 8.21 (s, 1H). LC-MS
(ES+) m/z 486 (M + H)+.
Step 2: A mixture of cinnamate ester (0.27g, 0.556 mmol) and 10%
Pd/C (0.012 g) in EtOH (10 mL) at 25 °C was hydrogenated under
an atmosphere of H2 at 1 atm overnight. The reaction mixture was
filtered through diatomaceous earth and washed with EtOH. The
filtrate was concentrated under reduced pressure to give an
intermediate ester (182 mg, 57%) as a clear oil. ES-LCMS m/z 488
(M + H)+.
Step 3: To a solution of ester (172 mg, 0.353 mmol) in a 1:1:2
mixture of THF/MeOH/H2O (12 mL) was added lithium hydroxide
(84 mg, 3.53 mmol). The reaction mixture was stirred at room
temperature overnight then partially concentrated under reduced
pressure to remove organic solvents. The resulting aqueous layer was
1
0.141 mmol, 57% yield (two steps)] as a white solid. H NMR (400
MHz, DMSO-d6) δ ppm 0.68−0.90 (m, 6 H), 0.97−1.32 (m, 4 H),
1.34−1.62 (m, 2 H), 1.67−1.88 (m, 1 H), 1.95−2.25 (m, 1 H), 2.43−
2.63 (m, 1 H), 2.97−3.24 (m, 1 H), 3.47 (s, 3 H), 3.52−3.72 (m, 1 H),
4.03−4.29 (m, 2 H), 5.80−6.02 (m, 1 H), 6.04−6.24 (m, 1 H), 7.42
(br s, 5 H), 7.60 (br s, 1 H) (acidic protons not observed). LC-MS
(ES−) m/z 496 [M − 1]. LC-MS (ES+) m/z 498 [M + H]. HRMS
(ES+) calcd for C23H32NO7PS (M + 1) m/z 498.1715, found
498.1716.
(2-(((3R,5R)-3-Butyl-3-ethyl-7-methoxy-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydrobenzo[f ][1,4]thiazepin-8-yl)oxy)ethyl)-
phosphonic Acid (14). Step 1: To a solution of 1,2-dibromoethane
(0.466 g, 2.478 mmol) and DIPEA (0.087 mL, 0.496 mmol) in THF
(5 mL) was added 2 (100 mg, 0.248 mmol), and the mixture was
stirred for 1 h, after which time LCMS indicated no reaction. K2CO3
(0.068 g, 0.496 mmol) was added to the reaction mixture, and the
mixture was stirred for an additional 16 h, after which time LCMS still
showed no reaction. Then 1 M KOH (3.0 mL, 3.00 mmol),
tetrabutylammonium bromide (24 mg, 0.074 mmol), and additional
1,2-dibromoethane (1.07 g, 5.70 mmol) were added to the reaction
mixture and stirred for 3 days, after which time LCMS indicated good
conversion to the intermediate bromide. The reaction mixture was
partitioned between EtOAc and water, and the organic phase was
isolated, dried over MgSO4, filtered, and concentrated to dryness to
give the intermediate as a clear oil. The oil was dissolved in triethyl
L
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
then acidified to pH 1−2 with 1 N HCl and extracted with DCM. The
organic layer was washed with saturated brine, dried (Na2SO4),
filtered, and concentrated under reduced pressure to give 16 (149 mg,
methanesulfonic Acid Trifluoroacetate Salt (18). Step 1: To a
DCM (2 mL) solution of 4 (0.275 g, 0.617 mmol) at 0 °C under
nitrogen was added a 1 M solution of DIBAL-H in toluene (1.3 mL,
1.30 mmol). The reaction was warmed to 25 °C and stirred for 1.5 h
then MeOH was added followed by H2O. The reaction mixture was
concentrated then redissolved in EtOAc. The layers were separated,
and the aqueous layer was extracted with EtOAc, then the combined
organics were washed with brine, dried (Na2SO4), filtered, and
concentrated. Chromatography on silica using hexanes/EtOAc
provided an intermediate alcohol (0.238 g, 91%). 1H NMR (400
MHz, CDCl3) δ ppm 0.76−0.93 (m, 6 H), 1.04−1.36 (m, 6 H), 1.36−
1.59 (m, 6 H), 1.76−1.93 (m, 1 H), 2.05−2.24 (m, 2 H), 3.02 (d, J =
14.8 Hz, 1 H), 3.42 (d, J = 14.8 Hz, 1 H), 3.54 (s, 3 H), 4.65 (qd, J =
13.2, 6.3 Hz, 2 H), 6.07 (br s, 1 H), 6.17 (s, 1 H), 7.29−7.50 (m, 5 H),
8.03 (s, 2 H). ES-LCMS m/z 418 (M + H)+.
1
87%) as a white solid. H NMR (CDCl3) δ ppm 0.81 (t, J = 7.0 Hz,
3H), 0.88 (t, J = 7.3 Hz, 3H), 0.98−1.14 (m, 1H), 1.13−1.38 (m, 5H),
1.77−1.97 (m, 1H), 2.09−2.39 (m, 1H), 2.48−2.66 (m, 2H), 2.78−
2.95 (m, 2H), 3.41 (d, J = 15.0 Hz, 1H), 3.50 (s, 3H), 5.97−6.32 (m,
2H), 7.41 (d, J = 7.0 Hz, 5H), 7.85 (s, 1H). LC-MS (ES−) m/z 458
(M − 1)−. HRMS (ES+) calcd for C25H33NO5S (M + 1) m/z
460.2159, found 460.2158.
3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-1-propanol
(16a). To an ice-cold solution of 16 (100 mg, 0.22 mmol) in THF (5
mL) was added BH3-THF complex (0.653 mL, 0.653 mmol). The
reaction mixture was stirred at 25 °C overnight then quenched by
dropwise addition of MeOH. The reaction mixture was stirred for 30
min then concentrated under reduced pressure. The residue was taken
up with MeOH and evaporated under reduced pressure again. The
residue was purified using silica gel (MeOH/DCM = 0:100 to 10:90)
to give 16a (96 mg, 94%) as a colorless oil. 1H NMR (CDCl3) δ ppm
0.80 (t, J = 7.0 Hz, 3H), 0.86 (t, J = 7.3 Hz, 3H), 0.99−1.52 (m, 14H),
1.69−1.92 (m, 3H), 2.02−2.25 (m, 1H), 2.56−2.75 (m, 2H), 3.00 (d, J
= 14.8 Hz, 1H), 3.40 (d, J = 14.8 Hz, 1H), 3.48 (s, 3H), 3.58 (t, J = 6.2
Hz, 2H), 6.01 (s, 1H), 6.11 (s, 1H), 7.26−7.48 (m, 5H), 7.84 (s, 1H).
LC-MS (ES+) m/z 446 (M + H)+.
Step 2: Imidazole (0.153 g, 0.563 mmol) was dissolved in DCM (2
mL), and the solution was cooled to 0 °C. Triphenylphosphine (0.295
g, 1.126 mmol) was added, followed by bromine (0.058 mL, 1.126
mmol). A solution of alcohol (0.235g, 0.563 mmol) in DCM (1 mL)
was added slowly at 0 °C. The reaction was stirred at 0 °C for 2 h
followed by addition of aqueous Na2SO3, and the resulting mixture
separated. The organic layer was dried (Na2SO4), filtered through a
silica pad, and concentrated to give a thick oil which solidified upon
1
standing to give an intermediate bromide (0.265 g, 84%). H NMR
(400 MHz, CDCl3) δ ppm 0.73−0.95 (m, 6 H), 1.02−1.36 (m, 5 H),
1.38−1.65 (m, 5 H), 1.76−1.92 (m, 1 H), 2.08−2.23 (m, 1 H), 3.02
(d, J = 14.8 Hz, 1 H), 3.43 (d, J = 14.8 Hz, 1 H), 3.57 (s, 3 H), 4.38−
4.56 (m, 2 H), 6.06 (s, 1 H), 6.16 (s, 1 H), 7.28−7.45 (m, 5 H), 8.05
(s, 1 H). LC-MS m/z 480 (M + H)+. LC-MS 482 (M + H + 2)+.
Step 3: To a solution of bromide (0.100 g, 0.21 mmol) in 1,4-
dioxane (1 mL) was added a solution of sodium sulfite (131 mg, 1.04
mmol) in H2O (1 mL). The mixture was stirred under reflux overnight
then concentrated under vacuum, and the crude reaction mixture was
washed with 1 N HCl. The supernatant was decanted, leaving a white
gummy solid which was triturated with DCM and hexanes to give a
solid that was collected via filtration. Purification was accomplished
using the Agilent prep-HPLC (C18 packing with MeCN, H2O w/0.1%
TFA as the mobile phase) to give 18 (83 mg, 66%) as a
4-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]butanoic Acid
Trifluoroacetate Salt (17). Step 1: To an ice-cold solution of 16a
(250 mg, 0.561 mmol) in DCM (10 mL) was added Et3N (0.235 mL,
1.683 mmol) and MsCl (0.048 mL, 0.62 mmol). The reaction mixture
was stirred at 25 °C overnight and partitioned between H2O and
DCM. The organic layer was washed with saturated brine, dried
(Na2SO4), filtered, and concentrated under reduced pressure to give
1
an intermediate mesylate (280 mg, 91%) as a light-yellow solid. H
NMR (CDCl3) δ ppm 0.80 (t, J = 6.8 Hz, 3H), 0.86 (t, J = 7.4 Hz,
3H), 0.97−1.34 (m, 8H), 1.35−1.61 (m, 4H), 1.73−2.27 (m, 4H),
2.46−2.78 (m, 4H), 2.85−3.07 (m, 4H), 3.31−3.43 (m, 1H), 3.47 (s,
3H), 4.04−4.26 (m, 2H), 6.01 (d, J = 8.0 Hz, 1H), 6.11 (s, 1H), 7.27−
7.52 (m, 5H), 7.82 (s, 1H).
1
trifluoroacetate salt. H NMR (400 MHz, DMSO-d6) δ ppm 0.62−
Step 2: To a solution of the above mesylate (90 mg, 0.172 mmol) in
DMSO (5 mL) was added NaCN (17 mg, 0.344 mmol). The reaction
mixture was stirred at 60 °C for 60 h and directly partitioned between
H2O and EtOAc. The organic layer was washed with saturated brine,
dried (Na2SO4), filtered, and concentrated under reduced pressure.
Purification using silica gel (EtOAc/hexanes = 10:90 to 1:1) afforded
an intermediate nitrile (76 mg, 95%) as a white solid. 1H NMR
(CDCl3) δ ppm 0.80 (t, J = 7.0 Hz, 3H), 0.86 (t, J = 7.4 Hz, 3H),
0.99−1.35 (m, 4H), 1.37−1.51 (m, 2H), 1.75−1.98 (m, 3H), 2.08−
2.21 (m, 1H), 2.29 (t, J = 7.2 Hz, 2H), 2.64−2.80 (m, 2H), 3.01 (d, J =
14.6 Hz, 1H), 3.40 (d, J = 14.8 Hz, 1H), 6.02 (s, 1H), 6.12 (s, 1H),
7.26−7.48 (m, 5H), 7.81 (s, 1H). LC-MS (ES+) m/z 455 (M + H)+.
Step 3: A mixture of nitrile (40 mg, 0.088 mmol) and 37% HCl (4
mL) was stirred at 100 °C overnight, cooled to 25 °C, and
concentrated under reduced pressure. MeOH was added to transfer
the residue to a flask, and the solution was concentrated under reduced
pressure. Purification by RP-HPLC (eluting with MeCN/H2O with
0.05% TFA-H2O and 0.05% TFA-MeCN) afforded trace amount of
desired acid with the major product being the corresponding methyl
ester of the acid, which was subsequently hydrolyzed in the presence
of excess lithium hydroxide. Purification by RP-HPLC (eluting with
MeCN/H2O with 0.05% TFA-H2O and 0.05% TFA-MeCN) afforded
17 (10 mg, 18%, TFA salt) as a white solid. 1H NMR (CDCl3) δ ppm
0.86 (t, J = 7.0 Hz, 3H), 0.96 (t, J = 7.4 Hz, 4H), 1.26−1.47 (m, 3H),
1.62−2.25 (m, 5H), 2.26−2.38 (m, 1H), 2.40−2.66 (m, 2H), 2.83−
2.97 (m, 1H), 3.45 (d, J = 15.4 Hz, 1H), 3.65 (d, J = 15.6 Hz, 1H),
6.32 (s, 1H), 6.59 (s, 1H), 7.41−7.65 (m, 5H), 7.87 (s, 1H). LC-MS
(ES+)m/z 474 (M + H)+. HRMS (ES+) calcd for C26H35NO5S (M +
1) m/z 474.2316, found 474.2314.
1.13 (m, 7 H), 1.33 (br s, 3 H), 1.60 (br s, 1 H), 2.02 (br s, 2 H),
3.14−4.46 (m, 6 H + H2O), 6.33 (br s, 2 H), 7.59 (br s, 5 H), 8.17 (br
s, 1 H). LC-MS (ES+) m/z 482 (M + H)+. HRMS (ES+) calcd for
C23H31NO6S2 (M + 1) m/z 482.1669, found 482.1671.
3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-1-propanesul-
fonic Acid (19). Step 1: To a solution of compound 16a (84.2 mg,
0.189 mmol) in THF (5 mL) was added triphenylphosphine (99 mg,
0.378 mmol) and CBr4 (125 mg, 0.378 mmol). The reaction mixture
was stirred at 25 °C overnight then partitioned between H2O and
DCM. The organic layer was washed with saturated brine, dried
(Na2SO4), filtered, and concentrated under reduced pressure.
Purification using silica gel (EtOAc/hexanes = 10:90 to 50:50)
afforded an intermediate bromide (78 mg, 80%) as a clear oil. ES-
LCMS m/z 508 (M + H)+.
Step 2: To a solution of bromide (70 mg, 0.138 mmol) in a 1:1
mixture of EtOH/H2O (10 mL) was added sodium sulfite (868 mg,
6.88 mmol). The reaction mixture was stirred at reflux overnight,
cooled to 25 °C, and partially concentrated under reduced pressure to
remove the organic solvents. The aqueous layer was then acidified to
pH 1 with 1 N HCl. The aqueous layer was extracted with DCM. The
combined organic layers were washed with saturated brine, dried
(Na2SO4), filtered, and concentrated under reduced pressure.
Purification on silica gel (MeOH/DCM = 0:100 to 20:80) afforded
1
19 (27.2 mg, 38%) as a white solid. H NMR (CDCl3) δ ppm 0.47−
2.26 (m, 12H), 2.33−3.63 (m, 14H), 6.06 (br s, 2H), 7.16−7.60 (m,
5H), 7.78 (br s, 1H). LC-MS (ES+) m/z 510 (M + H)+. HRMS (ES+)
calcd for C25H35NO6S2 (M + 1) m/z 510.1988, found 510.1984.
{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
phosphonic Acid Hydrochloride Salt (20). Step 1: Prepared from
[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-
M
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
bromomethyl intermediate (see compound 18, step 2) (0.920 g, 1.92
mmol) upon reaction with triethyl phosphate (0.352 mL, 2.01 mmol)
in toluene (10 mL) to give an intermediate phosphonate ester.
Step 2: The phosphonate ester from the previous step was dissolved
in 2 mL of 6 N HCl and 1 mL EtOH then heated at reflux for 6 h. The
reaction was cooled then concentrated. The residue was dissolved in
EtOH then concentrated to give 20 (0.86 g, 85%) as a hydrochloride
salt. 1H NMR (400 MHz, MeOH-d4) δ ppm 0.89 (t, J = 6.88 Hz, 3 H),
0.98 (t, J = 7.41 Hz, 4 H), 1.23−1.46 (m, 3 H), 1.64 (dd, J = 14.15,
7.32 Hz, 1 H), 1.97 (dd, J = 14.15, 7.12 Hz, 2 H), 2.68 (br s, 1 H),
2.99−3.12 (m, 2 H), 3.16−3.25 (m, 1 H), 3.45 (d, J = 15.51 Hz, 1 H),
3.58 (s, 3 H), 3.79 (d, J = 15.51 Hz, 1 H), 6.36 (s, 1 H), 6.46 (s, 1 H),
7.57 (s, 5 H), 8.04 (d, J = 2.34 Hz, 1 H). LC-MS (ES+) m/z 482 (M +
H)+. HRMS (ES+) calcd for C23H32NO6PS (M + 1) m/z 482.1766,
found 482.1769.
{2-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]ethyl}-
phosphonic Acid Bis-ammmonium Salt (21). Step 1: Tetraethyl
methanediylbis(phosphonate) (645 mg, 2.24 mmol) dissolved in THF
(7 mL) was treated with NaH (83 mg, 2.075 mmol, 60% dispersion in
oil) with stirring at 25 °C for 30 min. Compound 7 (300 mg, 0.722
mmol) was added, and the mixture was stirred for 1 h at 25 °C, after
which time LCMS indicated complete conversion. The mixture was
quenched with H2O and partitioned between EtOAc and brine. The
organic phase was dried over MgSO4, filtered, and concentrated to an
oil. The residue was purified on silica gel eluting with 40−100%
EtOAc/hexanes to give an intermediate phosphonate ester (342 mg,
86% yield) as a clear oil. LC-MS (ES+) m/z 550 [M + H].
7.39−7.60 (m, 5H), 7.90 (s, 1H). LC-MS(ES+) m/z 510 (M + H)+.
HRMS (ES+) calcd for C25H36NO6PS (M + 1) m/z 510.2078, found
510.2079.
3-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]carbonyl}-
amino)-2,2-dimethylpropanoic Acid (23). Step 1: To a DMF (0.5
mL) solution of 5 (0.025 g, 0.058 mmol) and ethyl 3-amino-2,2-
dimethylpropanoate23 (0.009 g, 0.064 mmol) at 0 °C was added
HATU (0.024 g, 0.064 mmol) followed by DIEA (0.012 mL, 0.070
mmol). The reaction was stirred for 10 min then warmed to 25 °C.
After 1 h, the reaction was concentrated to half volume, and H2O (3
mL) was added. The precipitant was filtered and dried and then used
without further purification in the next step.
Step 2: The material was dissolved in THF (0.5 mL), and H2O
(0.25 mL) was added along with excess LiOH (7 mg, 0.174 mmol).
The reaction was stirred at 25 °C for 3 h then concentrated. The
residue was purified by silica chromatography using DCM/MeOH to
give 23 (20 mg, 63%). 1H NMR (400 MHz, CDCl3) δ ppm 0.71−0.96
(m, 6 H), 0.99−1.65 (m, 12 H), 1.83 (br s, 1 H), 2.18 (br s, 1 H), 3.11
(br s, 1 H), 3.31−3.70 (m, 7 H), 6.08 (br s, 1 H), 6.21 (br s, 1 H),
7.27−7.52 (m, 5 H), 8.13 (br s, 1 H), 8.71 (s, 1 H). LC-MS (ES+) m/z
531 (M + H)+. HRMS (ES+) calcd for C28H38N2O6S (M + 1) m/z
531.2530, found 531.2529.
({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]carbonyl}-
amino)methanesulfonic Acid (24). Prepared in analogous fashion
to compound 23 (step 1) via HATU amide-coupling using
aminomethanesulfonic acid (0.013 g, 0.116 mmol) and 5 (50 mg,
0.12 mmol) to give compound 24 (45 mg, 73%). 1H NMR (400 MHz,
CDCl3) δ ppm 0.63−0.93 (m, 6 H), 0.93−1.32 (m, 5 H), 1.45 (br s, 2
H), 1.80 (br s, 1 H), 2.14 (br s, 1 H), 3.22 (br s, 1 H), 3.43−4.72 (m,
6 H + H2O), 6.02 (br s, 1 H), 6.29 (br s, 1 H), 7.47 (d, J = 3.62 Hz, 5
H), 8.28 (partially resolved t, J = 5.72 Hz, 1 H), 8.47 (s, 1 H). LC-MS
(ES+) m/z 525 (M + H)+. HRMS (ES+) calcd for C24H32N2O7S2 (M +
1) m/z 525.1724, found 525.1729.
Step 2: The ester (240 mg, 0.44 mmol) dissolved in EtOH (20 mL)
was treated with 10% Pd/C (46.5 mg, 0.44 mmol) under a balloon of
H2 at 25 °C overnight. The catalyst was filtered off, and the filtrate was
concentrated to dryness to give the intermediate phosphonate as a
clear oil. The residue was dissolved in DCM (10 mL) and treated with
TMSBr (0.227 mL, 1.75 mmol) for 16 h at 23 °C with stirring. The
reaction mixture was concentrated to dryness then purified by RP-
HPLC (30 mm × 100 mm XBridge column, MeCN/H2O containing
0.2% NH4OH buffer, 10 to 80 to 100% over 8 min) to give 21 (88 mg,
2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]carbonyl}-
amino)ethanesulfonic Acid Trifluoroacetate (25). Prepared in an
analogous fashion to compound 23 (step 1) via HATU amide-
coupling using taurine (0.014 g, 0.116 mmol) and compound 5 (50
mg, 0.116 mmol). Purification via RP-HPLC (30 mm × 150 mm H2O
Sunfire C18 column) using MeCN/H2O 10−100% with 0.1% TFA as
1
41% yield) bis-ammonium salt as a white solid. H NMR (400 MHz,
DMSO-d6) δ ppm 0.70−0.85 (m, 6 H), 0.97−1.30 (m, 4 H), 1.29−
1.64 (m, 4 H), 1.67−1.81 (m, 1 H), 2.00−2.12 (m, 1 H), 2.58 (d, J =
9.8 Hz, 1 H), 2.63−2.79 (m, 2 H), 3.05 (d, J = 14.9 Hz, 1 H), 3.39 (s,
3 H), 3.50 (d, J = 14.9 Hz, 1 H), 5.92 (d, J = 9.6 Hz, 1 H), 6.05 (s, 1
H), 7.26−7.36 (m, 1 H), 7.36−7.46 (m, 4 H), 7.65 (s, 1 H)
(phosphonic acid protons not observed). LC-MS (ES+) m/z 496 [M +
H]. HRMS (ES+) calcd for C24H34NO6PS (M + 1) m/z 496.1922,
found 496.1923.
{3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]propyl}-
phosphonic Acid Trifluoroacetate Salt (22). Step 1: To a solution
of diethylphosphite (119 mg, 0.859 mmol) in THF (5 mL) was added
NaH (27.5 mg, 0.687 mmol). The reaction mixture was stirred for 30
min, followed by dropwise addition of mesylate (90 mg, 0.172 mmol)
(see compound 17, step 1) in THF (5 mL). The reaction mixture was
stirred at 25 °C overnight then concentrated under reduced pressure.
The residue was partitioned between 1 N HCl and EtOAc. The
organic layer was washed with saturated brine, dried (Na2SO4),
filtered, and concentrated under reduced pressure. Purification using
silica gel (EtOAc/hexanes = 10:90 to 100:0) afforded an intermediate
phosphonate ester (55 mg, 53%) as a white solid. ES-LCMS m/z 566
(M + H)+.
1
mobile phase over 8 min provided 25 (15 mg, 19%). H NMR (400
MHz, MeOH-d4) δ ppm 0.91 (t, J = 7.13 Hz, 3 H), 1.02 (partially
resolved t, J = 7.48 Hz, 4 H), 1.43 (d, J = 6.55 Hz, 3 H), 1.66 (s, 1 H),
2.13 (br s, 2 H), 2.77 (br s, 1 H), 2.91−3.02 (m, 2 H), 3.55 (d, J =
15.83 Hz, 1 H), 3.72 (s, 3 H), 3.75−3.83 (m, 2 H), 4.02 (d, J = 15.83
Hz, 1 H), 6.47 (s, 1 H), 6.68 (s, 1 H), 7.48−7.72 (m, 5 H), 8.74 (s, 1
H). LC-MS (ES+) m/z 539 (M + H)+. HRMS (ES+) calcd for
C25H34N2O7S2 (M + 1) m/z 539.1883, found 539.1886.
N-{3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-
propanoyl}glycine (26). Step 1: Prepared in an analogous fashion to
compound 23 using compound 16 (50 mg, 0.109 mmol), DIPEA
(0.095 mL, 0.544 mmol), HATU (124 mg, 0.326 mmol), and glycine
methyl ester hydrochloride (19.39 mg, 0.218 mmol) in DCM (4 mL).
Purification on silica gel (MeOH/DCM = 0:100 to 10:90) afforded an
intermediate methyl ester (45 mg, 76%) as a clear oil. ES-LCMS m/z
531 (M + H)+. The methyl ester was saponified in an analogous
fashion to compound 23. Step 2: Purification on silica gel
(MeOH:DCM = 0:100 to 10:90) afforded 26 (21 mg, 47%) as a
white solid. 1H NMR (CDCl3) δ ppm 0.49−0.93 (m, 13H), 1.76−1.96
(m, 1H), 2.27−2.66 (m, 3H), 2.70−2.90 (m, 1H), 3.02 (d, J = 6.8 Hz,
1H), 3.14−3.33 (m, 1H), 3.34−3.44 (m, 1H), 3.49 (s, 3H), 3.89 (br s,
2H), 6.02−6.37 (m, 3H), 7.31−7.62 (m, 5H), 7.83 (s, 1H). LC-MS
(ES+) m/z 515 (M − H)−. HRMS (ES+) calcd for C27H36N2O6S (M +
1) m/z 517.2374, found 517.2372.
Step 2: To a solution of phosphonate ester (50 mg, 0.088 mmol) in
DCM (5 mL) was added TMSBr (0.115 mL, 0.884 mmol). The
reaction mixture was stirred at 25 °C overnight then concentrated
under reduced pressure. Purification by RP-HPLC (eluting with
MeCN/H2O with 0.05% TFA-H2O and 0.05% TFA-MeCN) afforded
1
22 (25 mg, 43%, TFA salt) as a white solid. H NMR (MeOH-d4) δ
ppm 0.86 (t, J = 7.0 Hz, 3H), 0.95 (t, J = 7.4 Hz, 3H), 0.99−1.13 (m,
1H), 1.20−1.45 (m, 3H), 1.48−1.72 (m, 3H), 1.75−2.10 (m, 4H),
2.51−2.66 (m, 1H), 2.73 (t, J = 7.4 Hz, 2H), 3.37 (d, J = 15.6 Hz, 1H),
3.55 (s, 3H), 3.79 (d, J = 15.6 Hz, 1H), 6.32 (s, 1H), 6.42 (s, 1H),
({3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-
propanoyl}amino)methanesulfonic Acid (27). Prepared in an
analogous fashion to compound 23 (step 1) using compound 16 (50
N
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
mg, 0.109 mmol), EDC (62.6 mg, 0.326 mmol), DMAP (39.9 mg,
0.326 mmol), and aminomethanesulfonic acid (12.1 mg, 0.109 mmol)
in a 2:1 mixture of DCM/THF (6 mL). The reaction mixture was
stirred at 25 °C overnight then partitioned between H2O and DCM.
The organic layer was washed with saturated brine, dried (Na2SO4),
filtered, and concentrated under reduced pressure. Purification with
HPLC (eluting with MeCN/H2O with 0.05% TFA-H2O and 0.05%
TFA-MeCN) afforded 27 (11 mg, 18%, TFA salt) as a white solid. 1H
NMR (CDCl3) δ ppm 0.63−0.99 (m, 6H), 1.31 (m, 3H), 1.69−1.97
(m, 2H), 2.12−2.52 (m, 2H), 2.58−2.80 (m, 2H), 2.92 (br s, 1H),
3.21−3.50 (m, 2H), 3.67 (s, 3H), 3.85 (d, J = 15.2 Hz, 2H), 6.51 (d, J
= 8.1 Hz, 2H), 7.35−7.77 (m, 5H), 7.85 (s, 1H). LC-MS (ES+) m/z
553 (M + H)+. HRMS (ES+) calcd for C26H36N2O7S2 (M + 1) m/z
553.2039, found 553.2042.
({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)(oxo)acetic Acid (28). Step 1: A solution of compound 8 (77
mg, 0.19 mmol) in DCM (924 μL) at 0 °C was treated with Et3N
(56.7 μL, 0.407 mmol) followed by methyl chloro(oxo)acetate (18.7
μL, 0.203 mmol). The reaction was stirred for 30 min then warmed to
25 °C and stirred for an additional 1 h. LCMS indicated 92:8/
prod:SM. Added additional acid chloride (2 μL) then stirred at 25 °C
for 2 h. The reaction was diluted with DCM and H2O, the layers
separated, and the aqueous layer extracted with DCM (3×). The
combined organic layers were dried (MgSO4), filtered, and
concentrated in vacuo to give a methyl ester (93 mg, 100% yield)
which was used as is in the next step. 1H NMR (CDCl3) δ 0.81 (t, J =
7.0 Hz, 3H), 0.88 (t, J = 7.4 Hz, 3H), 1.03−1.58 (m, 6H), 1.76−1.90
(m, 1H), 2.06−2.22 (m, 1H), 2.96−3.10 (m, 2H), 3.41 (d, J = 14.8
Hz, 1H), 3.54 (s, 3H), 3.87 (s, 3H), 4.48 (d, J = 6.0 Hz, 2H), 6.06 (s,
1H), 6.18 (s, 1H), 7.28−7.54 (m, 6H), 7.96 (s, 1H). LC-MS (ES+) m/
z 503 [M + H].
Step 2: A solution of methyl ester (93 mg, 0.185 mmol) in THF
(694 μL) and H2O (231 μL) at 0 °C was treated with 1 M LiOH (204
μL, 0.204 mmol). The reaction was stirred at 0 °C for 60 min then
warmed to 25 °C and stirred for 1 h. The reaction was diluted with
EtOAc and acidified with dilute HCl. The layers were separated, and
the aqueous layer was extracted with EtOAc (3×). The combined
organic layers were washed with brine, dried (MgSO4), filtered, and
concentrated in vacuo. The residue was triturated with hexanes/Et2O
then a white solid was collected by filtration. 28 (79 mg, 87% yield)
was obtained as a white solid. 1H NMR (DMSO-d6) δ ppm 0.46−0.95
(m, 6H), 0.94−1.61 (m, 6H), 1.62−1.88 (m, 1H), 1.95−2.26 (m, 1H),
3.03 (br s, 1H), 3.48 (m 4H), 4.16−4.38 (m, 2H), 5.92 (br s, 1H),
6.17 (br s, 1H), 7.27−7.65 (m, 5H), 7.75 (br s, 1H), 9.37 (br s, 1H).
LC-MS (ES+) m/z 489 [M + H]. HRMS (ES+) calcd for
C25H32N2O6S (M + 1) m/z 489.2062, found 489.2059.
2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8- yl]methyl}-
amino)-2-oxoethanesulfonic Acid (29). Step 1: To an ice-cold
solution of compound 8 (50 mg, 0.120 mmol) in DCM (6 mL) was
added pyridine (0.097 mL, 1.20 mmol) and chloroacetyl chloride
(0.048 mL, 0.60 mmol). The reaction mixture was stirred at 25 °C
overnight then concentrated under reduced pressure to give a light-
yellow oil. The intermediate chloroacetamide (55 mg, 92%) was used
in the next step without further purification. 1H NMR (CDCl3) δ ppm
0.80 (t, J = 7.0 Hz, 3H), 0.86 (t, J = 7.4 Hz, 3H), 0.99−1.60 (m, 6H),
1.69−1.90 (m, 1H), 2.09−2.24 (m, 1H), 3.02 (s, 2H), 3.37 (s, 1H),
3.53 (s, 3H), 4.05 (d, J = 7.0 Hz, 3H), 4.44 (d, J = 6.1 Hz, 2H), 6.05
(s, 1H), 6.17 (s, 1H), 6.91−7.13 (m, 1H), 7.27−7.45 (m, 5H). LC-MS
(ES+) m/z 494 (M + H)+.
6.28 (br s, 1H), 7.16−7.68 (m, 5H), 7.94 (s, 1H). LC-MS (ES+) m/z
539 (M + H)+. HRMS (ES+) calcd for C25H34N2O7S2 (M + 1) m/z
539.1889, found 539.1886.
3-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)-3-oxo-1-propanesulfonic Acid Ammonium Salt (30).
Step 1: To an ice-cold solution of 8 (150 mg, 0.233 mmol) in DCM
(10 mL) was added pyridine (0.132 mL, 1.629 mmol) and 3-
chloropropanoyl chloride (0.045 mL, 0.465 mmol). The reaction
mixture was stirred at 25 °C for 4 h then partitioned between H2O and
DCM. Purification via silica gel (MeOH/DCM = 0:100 to 3:97)
afforded an intermediate chloropropanamide (105 mg, 89%) as a
colorless oil. ES-LCMS m/z 508 (M + H)+.
Step 2: To a solution of chloropropanamide (150 mg, 0.296 mmol)
in a 1:1 mixture of EtOH/H2O (10 mL) was added Na2SO3 (186 mg,
1.479 mmol). The reaction mixture was stirred at 60 °C overnight,
cooled to 25 °C, acidified with AcOH, and concentrated under
reduced pressure. Purification by RP-HPLC (eluting with MeCN/
H2O with 0.5% ammonium hydroxide in H2O) afforded 30 (27 mg,
1
15%, ammonium salt) as a white solid. H NMR (MeOH-d4) δ ppm
0.78 (t, J = 6.8 Hz, 3H), 0.86 (t, J = 7.4 Hz, 3H), 1.03−1.32 (m, 4H),
1.32−1.46 (m, 1H), 1.48−1.66 (m, 1H), 1.67−1.85 (m, 1H), 2.08−
2.31 (m, 1H), 2.57−2.73 (m, 2H), 2.95−3.11 (m, 3H), 3.44 (d, J =
14.8 Hz, 1H), 3.51 (s, 3H), 4.18−4.37 (m, 2H), 6.04 (s, 1H), 6.20 (s,
1H), 7.13−7.55 (m, 5H), 7.86 (s, 1H), 8.31−8.49 (m, 1H). LC-MS
(ES+) m/z 553 (M + H)+. HRMS (ES+) calcd for C26H36N2O7S2 (M +
1) m/z 553.2042, found 553.2042.
[2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)-2-oxoethyl]phosphonic Acid Trifluoroacetate Salt (31).
Step 1: A mixture of intermediate chloroacetamide (see compound 29,
step 1) (82 mg, 0.166 mmol) and triethyl phosphite (2.9 mL, 16.63
mmol) was stirred at 135 °C overnight, cooled to 25 °C, and
concentrated under reduced pressure. The crude diethyl phosphonate
(99 mg, 98%) was used in the next step without further purification.
ES-LCMS m/z 595 (M + H)+.
Step 2: To a solution of intermediate diethyl phosphonate (99 mg,
0.166 mmol) in DCM (5 mL) was added TMSBr (0.065 mL, 0.499
mmol), and the reaction mixture was stirred at 25 °C overnight.
Additional TMSBr (0.216 mL, 1.66 mmol) was added, and the
reaction mixture was stirred overnight at 25 °C. The reaction mixture
was then concentrated under reduced pressure and treated with H2O.
Purification by RP-HPLC (eluting with MeCN/H2O with 0.05% TFA-
H2O and 0.05% TFA-MeCN) afforded 31 (69 mg, 63%, TFA salt) as a
1
gray−white solid. H NMR (MeOH-d4) δ ppm 0.89 (t, J = 6.8 Hz,
3H), 0.99 (t, J = 7.4 Hz, 3H), 1.22−1.49 (m, 3H), 1.56−1.75 (m, 1H),
1.96−2.13 (m, 2H), 2.57−2.87 (m), 3H), 3.61 (s, 3H), 3.66 (br s,
1H), 3.85 (d, J = 15.4 Hz, 1H), 4.24 (d, J = 14.1 Hz, 1H), 4.43 (d, J =
14.8 Hz, 1H), 6.42 (s, 1H), 6.51 (s, 1H), 7.35−7.75 (m, 5H), 8.01 (s,
1H). LC-MS (ES+) m/z 539 (M + H)+. HRMS (ES+) calcd for
C25H35N2O7PS (M + 1) m/z 539.1980, found 539.1981.
2,2′-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]carbonyl}-
imino)diacetic Acid Hydrochloride (32). Prepared in analogous
fashion to compound 23 via HATU amide-coupling using diethyl
iminodiacetate (0.022 g, 0.116 mmol) and compound 5 (0.050 g,
0.116 mmol) followed by LiOH (10 mg, 0.24 mmol) hydrolysis
conditions to give 32 (45 mg, 63%). 1H NMR (400 MHz, DMSO-d6)
δ ppm 0.64−0.92 (m, 6 H), 0.95−1.63 (m, 7 H), 1.66−1.91 (m, 1 H),
2.08 (br s, 1 H), 3.01−3.53 (m, 4 H + H2O), 3.86 (br s, 2 H), 4.15 (br
s, 2 H), 5.97 (br s, 1 H), 6.15 (s, 1 H), 7.22−7.60 (m, 5 H), 7.73 (s, 1
H), 12.76 (br s, 2 H). LC-MS (ES+) m/z 547 (M + H)+. HRMS (ES+)
calcd for C27H34N2O8S (M + 1) m/z 547.2115, found 547.2114.
3-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]carbonyl}-
amino)pentanedioic Acid (33). Prepared in analogous fashion to
compound 23 via HATU amide-coupling using dimethyl 3-amino-
pentanedioate (81 mg, 0.463 mmol) and compound 5 (100 mg, 0.232
mmol) followed by LiOH (10 mg, 0.24 mmol) hydrolysis in 1:1:1
mixture of MeOH/H2O/THF (15 mL). Purification with HPLC
(medium column, Gilson 845Z preparation system; Sunfire C18 5 μM,
Step 2: To a solution of the intermediate chloroacetamide (55 mg,
0.112 mmol) in a 1:1 mixture of EtOH/H2O (6 mL) was added
Na2SO3 (141 mg, 1.115 mmol). The reaction mixture was stirred at 80
°C overnight, cooled to 25 °C, acidified with AcOH to pH 3−4, and
concentrated under reduced pressure. Purification using silica gel
(MeOH/DCM = 0:100 to 30:70) afforded 29 (32 mg, 52%) as an off-
white solid. 1H NMR (MeOH-d4) δ ppm 0.81 (t, J = 6.6 Hz, 3H), 0.89
(t, J = 7.4 Hz, 3H), 1.01−1.94 (m, 7H), 2.06−2.49 (m, 1H), 3.43−
3.59 (m, 4H), 3.63−3.78 (m, 2H), 4.39 (br s, 2H), 6.06 (br s, 1H),
O
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
30 mm × 150 mm. Method, 10−100% MeCN/H2O with 0.05% TFA
over 10 min, 45 mL/min, λ = 220, 254 nm) afforded 33 as a light-
yellow solid (61 mg, 38% yield) as a TFA salt. H NMR (400 MHz,
1H), 7.34−7.66 (m, 5H), 7.95 (s, 1H). LC-MS (ES+) m/z 604 (M +
H)+. HRMS (ES+) calcd for C30H41N3O8S (M + 1) m/z 604.2697,
found 604.2693.
1
4-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)butanoic Acid Trifluoroacetate Salt (37). Step 1: To a
solution of 7 (200 mg, 0.481 mmol) in 1,2-dichloroethane (5 mL) was
added methyl 4-aminobutanoate (111 mg, 0.722 mmol) and AcOH
(0.276 mL, 4.81 mmol). The reaction mixture was stirred at 25 °C
overnight, treated with NaHB(OAc)3 (255 mg, 1.20 mmol), and
stirred for 1 h. The reaction mixture was treated with aqueous sodium
carbonate solution and extracted with DCM. The organic layer was
washed with H2O and saturated brine, dried (Na2SO4), filtered, and
concentrated under reduced pressure. Purification using silica gel
(EtOAc/hexanes = 20:80 to 60:40) afforded an intermediate ester
(210 mg, 83%) as a light-yellow oil. ES-LCMS m/z 517 (M + H)+.
Step 2: To a solution of ester (210 mg, 0.406 mmol) in a 1:1:1
mixture of THF/MeOH/H2O (3 mL) was added LiOH (10 mg, 0.406
mmol). The reaction mixture was stirred at 25 °C overnight then
concentrated under reduced pressure. Purification by RP-HPLC
(eluting with MeCN/H2O with 0.05% TFA-H2O and 0.05% TFA-
MeCN) afforded 37 (46 mg, 44%) trifluoroacetate salt as a white solid.
1H NMR (DMSO-d6) δ ppm 0.76 (t, J = 6.6 Hz, 3H), 0.81 (t, J = 7.2
Hz, 3H), 0.96−1.32 (m, 4H), 1.35−1.61 (m, 2H), 1.67−1.90 (m, 3H),
1.99−2.15 (m, 1H), 2.33 (t, J = 7.2 Hz, 2H), 2.88−3.01 (m, 2H), 3.09
(d, J = 15.0 Hz, 1H), 3.49 (s, 3H), 3.63 (d, J = 14.8 Hz, 1H), 4.16 (br
s, 2H), 5.97 (s, 1H), 6.18 (s, 1H), 7.30−7.56 (m, 5H), 8.09 (s, 1H),
8.67 (br s, 2H). LC-MS (ES+) m/z 503 (M + H)+. HRMS (ES+) calcd
for C27H38N2O5S (M + 1) m/z 503.2584, found 503.2580.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-2-
methylalanine Trifluoroacetate Salt (39). Prepared in a similar
fashion to 37 using methyl 2-amino-2-methylpropanoate. Purification
by RP-HPLC (eluting with MeCN/H2O with 0.05% TFA-H2O and
0.05% TFA-MeCN) afforded 39 (102 mg, 66%, TFA salt) as a white
solid. 1H NMR (DMSO-d6) δ ppm 0.77 (t, J = 6.8 Hz, 3H), 0.82 (t, J
= 7.2 Hz, 3H), 0.98−1.31 (m, 4H), 1.35−1.49 (m, 1H), 1.70−1.86 (m,
1H), 2.00−2.23 (m, 1H), 2.46−2.55 (m, 1H), 3.13 (d, J = 14.8 Hz,
1H), 3.49 (s, 3H), 3.64 (d, J = 15.0 Hz, 1H), 4.14 (d, J = 17.6 Hz, 2H),
6.00 (s, 1H), 6.18 (s, 1H), 7.28−7.55 (m, 5H), 8.10 (s, 1H). LC-MS
(ES+) m/z 503 (M + H)+. HRMS (ES+) calcd for C27H38N2O5S (M +
1) m/z 503.2581, found 503.2580.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-D-
aspartic Acid Trifluoroacetate Salt (50). Prepared in a similar
fashion to 37 using dimethyl D-aspartate. Purification by RP-HPLC
(eluting with MeCN/H2O with 0.05% TFA-H2O and 0.05% TFA-
MeCN) afforded 50 (11 mg, 37%, TFA salt) as a white solid. 1H NMR
(MeOH-d4) δ ppm 0.83 (t, J = 6.8 Hz, 3H), 0.91 (t, J = 7.4 Hz, 3H),
1.04−1.41 (m, 4H), 1.43−1.57 (m, 1H), 1.58−1.78 (m, 1H), 1.80−
1.96 (m, 1H), 2.92−3.14 (m, 2H), 3.58−3.67 (m, 4H), 4.13−4.24 (m,
1H), 4.31−4.45 (m, 2H), 6.18 (s, 1H), 6.42 (s, 1H), 7.33−7.54 (m,
5H), 8.10 (s, 1H). LC-MS (ES+) m/z 533 (M + H)+. HRMS (ES+)
calcd for C27H36N2O7S (M + 1) m/z 533.2324, found 533.2321.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-L-
aspartic Acid Trifluoroacetate Salt (51). Prepared in a similar
fashion to 37 using dimethyl L-aspartate. Purification by RP-HPLC
(eluting with MeCN/H2O with 0.05% TFA-H2O and 0.05% TFA-
MeCN) afforded 51 (25 mg, 89%, TFA salt) as a white solid. 1H NMR
(MeOH-d4) δ ppm 0.81 (t, J = 6.8 Hz, 3H), 0.91 (t, J = 7.4 Hz, 3H),
1.03−1.40 (m, 4H), 1.42−1.58 (m, 1H), 1.62−1.77 (m, 1H), 1.79−
1.95 (m, 1H), 2.38 (d, J = 4.1 Hz, 1H), 2.88−3.13 (m, 2H), 3.23 (d, J
= 15.0 Hz, 1H), 3.55−3.66 (m, 4H), 4.19 (dd, J = 6.8, 4.3 Hz, 1H),
4.38 (d, J = 7.6 Hz, 2H), 6.17 (s, 1H), 6.40 (s, 1H), 7.32−7.56 (m,
5H), 8.10 (s, 1H). LC-MS (ES+) m/z 533 (M + H)+. HRMS (ES+)
calcd for C27H36N2O7S (M + 1) m/z 533.2324, found 533.2321.
3-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)pentanedioic Acid (56). Prepared in a similar fashion to 37
using diethyl 3-aminopentanedioate (see below). Following saponfi-
DMSO-d6) δ ppm 0.68−0.89 (m, 6H), 0.95−1.34 (m, 4H), 1.34−1.61
(m, 2H), 1.66−1.88 (m, 1H), 1.97−2.23 (m, 1H), 2.52−2.61 (m, 4H),
3.06−3.20 (m, 1H), 3.47−3.70 (m, 4H), 4.46−4.63 (m, 1H), 5.86−
6.05 (m, 1H), 6.17−6.28 (m, 1H), 7.24−7.55 (m, 5H), 8.36 (s, 1H),
8.40−8.55 (m, 1H). LC-MS (ES+) m/z 561 (M + H)+. HRMS (ES+)
calcd for C28H36N2O8S (M + 1) m/z 561.2273, found 561.2271.
3-({3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-
propanoyl}amino)pentanedioic Acid Trifluoroacetate Salt (34).
Prepared in analogous fashion to compound 23 using 16 (100 mg,
0.218 mmol), dimethyl 3-aminopentanedioate (81 mg, 0.463 mmol),
DIPEA (0.190 mL, 1.088 mmol), and HATU (165 mg, 0.435 mmol)
in DCM (6 mL). Saponification with LiOH (10 mg, 0.24 mmol) in
1:1:1 mixture of MeOH/H2O/THF (15 mL). Purification by RP-
HPLC (eluting with MeCN/H2O with 0.05% TFA-H2O and 0.05%
1
TFA-MeCN) afforded 34 (92 mg, 60%, TFA salt) as white solid. H
NMR (MeOH-d4) δ ppm 0.87 (t, J = 7.0 Hz, 3H), 0.96 (t, J = 7.4 Hz,
3H), 1.33−1.39 (m, 3H), 1.49−1.67 (m, 1H), 1.86−2.13 (m, 2H),
2.37−2.46 (m, 2H), 2.48−2.58 (m, 4H), 2.60−2.75 (m, 1H), 2.78−
3.03 (m, 2H), 3.46 (d, J = 15.6 Hz, 2H), 3.57 (s, 3H), 3.81 (d, J = 15.6
Hz, 1H), 4.33−4.48 (m, 1H), 6.37 (s, 1H), 6.44 (s, 1H), 7.48−7.62
(m, 5H), 7.90 (s, 1H). LC-MS (ES+) m/z 589 (M + H)+. HRMS
(ES+) calcd for C30H40N2O8S (M + 1) m/z 589.2581, found 589.2584.
2,2′-({3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-
5-phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]-
propanoyl}imino)diacetic Acid (35). Step 1: Prepared in a similar
fashion to compound 23 using compound 16 (50 mg, 0.109 mmol),
diethyl 2,2′-iminodiacetate (41.2 mg, 0.218 mmol), EDC (62.6 mg,
0.326 mmol), and DMAP (39.9 mg, 0.326 mmol) in a 2:1 mixture of
DCM/THF (6 mL). Purification using silica gel (EtOAc/hexanes =
10:90 to 50:50) afforded an intermediate diester (45 mg, 65%) as a
clear oil. ES-LCMS m/z 631 (M + H)+. Saponification of intermediate
diethyl diester (45 mg, 0.071 mmol) as in compound 23, step 2 using
LiOH (85 mg, 3.55 mmol) in a 1:1:1 mixture of THF/MeOH/H2O (9
mL). Purification with HPLC (eluting with MeCN/H2O with 0.05%
TFA-H2O and 0.05% TFA-MeCN) afforded 35 (23 mg, 53%, TFA
1
salt) as a white solid. H NMR (CDCl3) δ ppm 0.55−1.45 (m, 8H),
1.69−2.09 (m, 3H), 2.31−2.48 (m, J = 14.1 Hz, 1H), 2.50−2.95 (m,
2H), 2.97−3.13 (m, 1H), 3.23−3.70 (m, 11H), 3.72−3.88 (m, J = 8.6
Hz, 2H), 3.96−4.25 (m, 2H), 6.42 (s, 1H), 6.54 (br s, 1H), 7.37−7.59
(m, 3H), 7.61−7.75 (m, 2H), 7.84 (s, 1H). LC-MS (ES+) m/z 575 (M
+ H)+. HRMS (ES+) calcd for C29H38N2O8S (M + 1) m/z 575.2428,
found 575.2427.
3-{[2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)-2-oxoethyl]amino}pentanedioic Acid Trifluoroacetate
Salt (36). Step 1: To a solution intermediate chloroacetamide (see
compound 29, step 1) (55 mg, 0.112 mmol) and diethyl 3-
aminopentanedioate (45.3 mg, 0.223 mmol) in DMF (3 mL) was
added K2CO3 (61.7 mg, 0.446 mmol) and KI (74.1 mg, 0.446 mmol).
The reaction mixture was stirred at 60 °C overnight, cooled to 25 °C,
and partitioned between H2O and EtOAc. The organic layer was
washed with saturated brine, dried (Na2SO4), filtered, and
concentrated under reduced pressure. Purification using silica gel
(MeOH/DCM = 0:100 to10:90) afforded an intermediate diester (75
mg, 100%) as a yellow oil. ES-LCMS m/z 660 (M + H)+.
Step 2: To a solution of the diester (74 mg, 0.112 mmol) in a 1:1:1
mixture of THF/MeOH/H2O (6 mL) was added LiOH (134 mg, 5.60
mmol). The reaction mixture was stirred at 25 °C overnight, acidified
with AcOH, and concentrated under reduced pressure. Purification by
RP-HPLC (eluting with MeCN/H2O with 0.05% TFA-H2O and
0.05% TFA-MeCN) afforded 36 (20 mg, 21%, TFA salt) as a white
solid. 1H NMR (MeOH-d4) δ ppm 0.86 (t, J = 6.8 Hz, 3H), 0.93 (t, J
= 7.4 Hz, 3H), 0.97−1.13 (m, 1H), 1.24−1.45 (m, 3H), 1.49−1.72 (m,
1H), 1.77−2.11 (m, 2H), 2.45−2.72 (m, 2H), 2.85 (d, J = 5.7 Hz,
3H), 3.46 (d, J = 15.4 Hz, 1H), 3.58 (s, 3H), 3.74 (d, J = 15.4 Hz, 1H),
3.82−4.00 (m, 3H), 4.41 (d, J = 3.3 Hz, 2H), 6.30 (s, 1H), 6.45 (s,
P
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
cation, the reaction was concentrated under reduced pressure then
dissolved in H2O and MeCN. The solution was acidified to pH 4−5
with AcOH, partially concentrated to remove MeCN under reduced
pressure, and left to stand for 30 min. A white precipitate was collected
by filtration then dried under reduced pressure at 50 °C overnight to
proline Ammonium salt (41). Prepared in a similar fashion to 38
using 1,1-dimethylethyl D-prolinate. Cleavage of the t-butyl ester was
accomplished using TFA. Purification by RP-HPLC (30 mm × 100
mm XBridge column; MeCN + 0.1% NH4OH and H2O + 0.1%
NH4OH were used as the solvent system; 10−70% over 8 min) to give
1
1
41 ammonium salt (333 mg, >99% yield) as a white solid. H NMR
give 56 as a white solid in >90% yield. H NMR (MeOH-d4) δ ppm
(400 MHz, DMSO-d6) δ ppm 0.75 (t, J = 7.0 Hz, 3 H), 0.81 (t, J = 7.4
Hz, 3 H), 0.95−1.31 (m, 4 H), 1.34−1.56 (m, 2 H), 1.67−2.14 (m, 5
H), 2.36−2.47 (m, 1 H), 2.58−2.81 (m, 1 H), 3.06 (d, J = 15.0 Hz, 1
H), 3.14−3.27 (m, 1 H), 3.35−3.47 (m, 1 H), 3.49 (s, 3 H), 3.62 (d, J
= 15.0 Hz, 1 H), 4.17−4.30 (m, 1 H), 4.36 (d, J = 13.1 Hz, 1 H), 4.50
(d, J = 13.1 Hz, 1 H), 5.96 (br s, 1 H), 6.17 (s, 1 H), 7.31−7.55 (m, 5
H), 8.09 (s, 1 H), 9.84 (br s, 1 H). LC-MS (ES−) m/z 513 [M − 1].
LC-MS (ES+) m/z 515 [M + H]. HRMS (ES+) calcd for
C28H38N2O5S (M + 1) m/z 515.2584, found 515.2580.
0.78 (t, J = 7.0 Hz, 3H), 0.87 (t, J = 7.4 Hz, 3H), 1.05−1.48 (m, 5H),
1.48−1.63 (m, 1H), 1.69−1.84 (m, 1H), 2.12−2.27 (m, 1H), 2.52−
2.73 (m, 4H), 3.09 (d, J = 14.8 Hz, 1H), 3.47 (d, J = 14.8 Hz, 1H),
3.58 (s, 3H), 3.60−3.68 (m, 1H), 4.25 (s, 2H), 6.06 (s, 1H), 6.29 (s,
1H), 7.27−7.49 (m, 5H), 8.05 (s, 1H). 13C NMR (500 MHz, DMSO-
d6) δ ppm 7.5, 13.9, 22.5, 24.5, 30.1, 32.4, 37.5, 43.2, 51.2, 54.3, 55.2,
56.8, 63.1, 110.3, 124.5, 126.9, 127.7, 128.1, 128.7, 131.8, 142.2, 147.1,
159.9, 172.7. LC-MS (ES+) m/z 547 (M + H)+. HRMS (ES+) calcd for
C28H39N2O7S (M + H+) m/e 547.2478, found 547.2474.
C28H38N2O7S requires C, 61.52%; H, 7.01%; N, 5.12%; S, 5.87%.
Found: C, 61.49%; H, 6.98%; N, 5.08%; S, 5.86%.
Preparation of diethyl 3-aminopentanedioate: To a solution of β-
glutamic acid (500 mg, 3.40 mmol) in EtOH (10 mL) was added
thionyl chloride (0.992 mL, 13.59 mmol) dropwise. The reaction
mixture was stirred at 25 °C overnight and concentrated under
reduced pressure. The residue was partitioned between DCM and
saturated potassium carbonate solution. The organic layer was washed
with saturated brine, dried (Na2SO4), filtered, and concentrated under
reduced pressure to give the title compound (702 mg, 97%) as a
colorless oil. 1H NMR (CDCl3) δ ppm 1.09−1.29 (m, 6H), 2.24−2.60
(m, 4H), 3.46−3.74 (m, 1H), 4.04−4.27 (m, 4H).
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-N-
methyl-β-alanine Trifluoroacetate Salt (38). Step 1: To a solution
of 7 (200 mg, 0.481 mmol) in 1,2-dichloroethane (10 mL) was added
1,1-dimethylethyl β-alaninate (175 mg, 0.963 mmol) and AcOH
(0.276 mL, 4.81 mmol). After 1 h of stirring, NaHB(OAc)3 (255 mg,
1.203 mmol) was added to the reaction mixture. The reaction mixture
was stirred at 25 °C for 1 h then treated with an aqueous sodium
carbonate solution. The layers were separated, and the aqueous layer
was extracted with DCM. The combined organic layers were washed
with H2O and saturated brine, dried (Na2SO4), filtered, and
concentrated under reduced pressure to afford an intermediate ester
(210 mg, 80%) as an oil. ES-LCMS m/z 559 (M + H)+.
N2-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-L-
lysine Ammonium Salt (42). Prepared in a similar fashion to 38
using 1,1-dimethylethyl N6-{[(1,1-dimethylethyl)oxy]carbonyl}-L-lysi-
nate·HCl. The crude material was purified by RP-HPLC (30 mm ×
100 mm XBridge column, MeCN/H2O containing 0.2% NH4OH
buffer, 10 to 80 to 100% over 10 min) to give 42 (297 mg, 38% yield)
1
ammonium salt as a pink foam. H NMR (400 MHz, DMSO-d6) δ
ppm 0.75 (t, J = 7.0 Hz, 3 H), 0.80 (t, J = 7.4 Hz, 3 H), 0.97−1.60 (m,
12 H), 1.67−1.80 (m, 1 H), 1.99−2.14 (m, 1 H), 2.54 (d, J = 9.8 Hz, 1
H), 2.64−2.78 (m, 3 H), 3.06 (d, J = 14.9 Hz, 1 H), 3.41 (s, 3 H), 3.50
(dd, J = 15.0, 9.3 Hz, 2 H), 3.65 (d, J = 15.0 Hz, 1 H), 5.94 (d, J = 9.6
Hz, 1 H), 6.06 (s, 1 H), 7.32 (dq, J = 8.5, 4.2 Hz, 1 H), 7.41 (d, J = 4.5
Hz, 4 H), 7.99 (s, 1 H) (4 exchangeable protons not observed). LC-
MS (ES+) m/z 546 [M + H]. HRMS (ES+) calcd for C29H43N3O5S
(M + 1) m/z 546.2998, found 546.3002.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-L-
threonine Ammonium Salt (43). Prepared as in 38 using 1,1-
dimethylethyl L-threoninate. The crude product mixture was purified
by RP-HPLC (30 mm × 100 mm XBridge column, MeCN/H2O
containing 0.2% NH4OH buffer, 10 to 70 to 100% over 10 min) to
give 43 (188 mg, 60% yield) ammonium salt as a white solid. 1H NMR
(400 MHz, DMSO-d6) δ ppm 0.75 (t, J = 7.0 Hz, 3 H), 0.80 (t, J = 7.4
Hz, 3 H), 0.97−1.28 (m, 4 H), 1.16 (d, J = 6.4 Hz, 3 H), 1.33−1.54
(m, 2 H), 1.68−1.79 (m, 1 H), 2.00−2.12 (m, 1 H), 2.59 (d, J = 9.8
Hz, 1 H), 2.96 (d, J = 4.7 Hz, 1 H), 3.06 (d, J = 14.9 Hz, 1 H), 3.44 (s,
3 H), 3.54 (d, J = 14.9 Hz, 1 H), 3.67−3.93 (m, 3 H), 5.94 (d, J = 9.4
Hz, 1 H), 6.09 (s, 1 H), 7.29−7.37 (m, 1 H), 7.37−7.46 (m, 4 H), 8.03
(s, 1 H) (3 exchangeable protons not observed). LC-MS (ES+) m/z
519 [M + H]. HRMS (ES+) calcd for C27H38N2O6S (M + 1) m/z
519.2532, found 519.2529.
Step 2: To the ester (210 mg, 0.289 mmol) was added 4 N HCl in
dioxane (3.61 mL, 14.46 mmol). The reaction mixture was stirred at
25 °C overnight then concentrated under reduced pressure.
Purification by RP-HPLC (eluting with MeCN/H2O with 0.05%
TFA-H2O and 0.05% TFA-MeCN) afforded 38 (143 mg, 82%, TFA
1
salt) as a white solid. H NMR (CDCl3) δ ppm 0.86 (t, J = 6.8 Hz,
3H), 0.92 (t, J = 7.2 Hz, 3H), 1.00−1.17 (m, 1H), 1.18−1.46 (m, 3H),
1.51−1.68 (m, 1H), 1.68−1.82 (m, 1H), 1.89−2.10 (m, 1H), 2.35 (br
s, 1H), 2.79 (br s, 2H), 3.25 (br s, 2H), 3.33−3.52 (m, 2H), 3.62 (s,
3H), 4.12−4.42 (m, 2H), 6.24 (s, 1H), 6.35 (s, 1H), 7.35−7.62 (m,
5H), 8.13 (s, 1H). LC-MS (ES+) m/z 503 (M + H)+. HRMS (ES+)
calcd for C27H38N2O5S (M + 1) m/z 503.2581, found 503.2580.
1-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-L-
proline Ammonium Salt (40). Prepared in a similar fashion to 38
using 1,1-dimethylethyl L-prolinate HCl. Purification by RP-HPLC (30
mm × 100 mm XBridge column; MeCN + 0.2% NH4OH and H2O +
0.2% NH4OH were used as the solvent system; 10−70% over 8 min,
100−100% to 10 min) to give 40 (260 mg, 84% yield) ammonium salt
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-L-
serine Bis-ammonium Salt (44). Prepared as in 38 using 1,1-
dimethylethyl L-serinate·HCl. The crude product was purified by RP-
HPLC (30 mm × 100 mm XBridge column, MeCN/H2O containing
0.2% NH4OH buffer, 10 to 70 to 100% over 8 min) to give 44 (194
1
mg, 64% yield) bis-ammonium salt as a white solid. H NMR (400
MHz, DMSO-d6) δ ppm 0.75 (t, J = 7.0 Hz, 3 H), 0.80 (t, J = 7.4 Hz, 3
H), 0.96−1.29 (m, 4 H), 1.32−1.54 (m, 2 H), 1.66−1.82 (m, 1 H),
1.98−2.13 (m, 1 H), 2.62 (d, J = 9.6 Hz, 1 H), 3.06 (d, J = 15.0 Hz, 1
H), 3.16 (t, J = 5.2 Hz, 1 H), 3.45 (s, 3 H), 3.56 (d, J = 14.9 Hz, 1 H),
3.65 (qd, J = 11.1, 5.2 Hz, 2 H), 3.79−3.94 (m, 2 H), 5.02 (br s, 1 H),
5.94 (d, J = 9.6 Hz, 1 H), 6.10 (s, 1 H), 7.28−7.37 (m, 1 H), 7.37−
7.47 (m, 4 H), 8.02 (s, 1 H) (2 exchangables not observed). LC-MS
(ES+) m/z 505 [M + H]. HRMS (ES+) calcd for C26H36N2O6S (M +
1) m/z 505.2372, found 505.2372.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
glycine (45). Prepared as in 38 using 1,1-dimethylethyl glycinate.
Recrystallization from CH3CN provided 45 (5.0 g, 82%) as a white
solid. 1H NMR (MeOH-d4) δ ppm 0.78 (t, J = 6.8 Hz, 3H), 0.87 (t, J
= 7.4 Hz, 3H), 1.01−1.34 (m, 4H), 1.34−1.48 (m, 1H), 1.49−1.64 (m,
1H), 1.67−1.86 (m, 1H), 2.12−2.31 (m, 1H), 3.08 (d, J = 15.0 Hz,
1H), 3.42 (s, 2H), 3.50 (d, J = 14.8 Hz, 1H), 3.58 (s, 3H), 4.22 (s,
1
as a white solid. H NMR (400 MHz, DMSO-d6) δ ppm 0.75 (t, J =
7.0 Hz, 3 H), 0.80 (t, J = 7.4 Hz, 3 H), 0.96−1.29 (m, 4 H), 1.32−1.53
(m, 2 H), 1.64−1.94 (m, 4 H), 1.99−2.16 (m, 2 H), 2.62 (d, J = 10.0
Hz, 1 H), 2.98−3.12 (m, 2 H), 3.26−3.37 (m, 2 H), 3.44 (s, 3 H), 3.55
(d, J = 15.0 Hz, 1 H), 3.82 (d, J = 14.1 Hz, 1 H), 3.89−4.00 (m, 1 H),
5.94 (d, J = 9.2 Hz, 1 H), 6.10 (s, 1 H), 7.27−7.37 (m, 1 H), 7.42 (d, J
= 4.5 Hz, 4 H), 7.99 (s, 1 H) (COOH proton not observed). LC-MS
(ES+) m/z 513.5 [M + H]. HRMS (ES+) calcd for C28H38N2O5S (M
+ 1) m/z 515.2583, found 515.2583.
1-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-D-
Q
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
2H), 6.06 (s, 1H), 6.30 (s, 1H), 7.25−7.50 (m, 5H), 8.06 (s, 1H). LC-
MS (ES+) m/z 475 (M + H)+. HRMS (ES+) calcd for C25H34N2O5S
(M + 1) m/z 475.2270, found 475.2267.
509 [M − 1]. LC-MS (ES+) m/z 511 [M + H]. HRMS (ES+) calcd
for C24H35N2O6PS (M + 1) m/z 511.2034, found 511.2032.
[2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)ethyl]phosphonic Acid Trifluoroacetate Salt (49). Step 1:
To a solution of 7 (50 mg, 0.120 mmol) in 1,2-dichloroethane (3 mL)
was added diethyl (2-aminoethyl)phosphonate (109 mg, 0.602 mmol).
The reaction mixture was stirred for 3 h then treated with
NaHB(OAc)3 (128 mg, 0.602 mmol). The reaction mixture was
stirred at 25 °C for 2 h then treated with H2O followed by 1 N HCl.
The resulting mixture was concentrated under reduced pressure.
Purification by RP-HPLC (eluting with MeCN/H2O with 0.05% TFA-
H2O and 0.05% TFA-MeCN) afforded an intermediate diethyl
aminophosphonate (70 mg, 71%) as a white solid. ES-LCMS m/z
581 (M + H)+.
Step 2: To a solution of the diethyl aminophosphonate (70 mg,
0.121 mmol) in DCM (3 mL) was added TMSBr (0.065 mL, 0.499
mmol). The reaction mixture was stirred at 25 °C overnight.
Additional TMSBr (0.16 mL, 1.21 mmol) was added. The reaction
mixture was stirred overnight, concentrated under reduced pressure,
and treated with H2O. Purification by RP-HPLC (eluting with MeCN/
H2O with 0.05% TFA-H2O and 0.05% TFA-MeCN) afforded 49 (51
mg, 55%, TFA salt) as a white solid. 1H NMR (MeOH-d4) δ ppm 0.81
(t, J = 7.0 Hz, 3H), 0.90 (t, J = 7.4 Hz, 3H), 1.00−1.38 (m, 4H), 1.38−
1.55 (m, 1H), 1.59−1.75 (m, 1H), 1.76−2.06 (m, 3H), 2.23−2.40 (m,
1H), 3.14−3.24 (m, 3H), 3.57−3.65 (m, 4H), 4.12−4.35 (m, 2H),
6.15 (s, 1H), 6.39 (s, 1H), 7.30−7.53 (m, 5H), 8.09 (s, 1H). LC-MS
(ES+) m/z 525 (M + H)+. HRMS (ES+) calcd for C25H37N2O6PS (M
+ 1) m/z 525.2185, found 525.2188.
N-{[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-N-
methylglycine Ammonium Salt (46). Prepared as in 38 using 1,1-
dimethylethyl N-methylglycinate·HCl. The residue was purified by RP-
HPLC (30 mm × 100 mm XBridge column; acetonitrile + 0.2%
NH4OH and H2O + 0.2% NH4OH were used as the solvent system;
10% to 70% over 8 min, 100% to 100% to 10 min) to give 46 (130 mg,
1
44% yield) ammonium salt as a white solid. H NMR (400 MHz,
DMSO-d6) δ ppm 0.75 (t, J = 7.0 Hz, 3 H), 0.80 (t, J = 7.4 Hz, 3 H),
0.96−1.28 (m, 4 H), 1.32−1.53 (m, 2 H), 1.68−1.80 (m, 1 H), 1.99−
2.11 (m, 1 H), 2.30 (s, 3 H), 2.62 (d, J = 9.4 Hz, 1 H), 3.07 (d, J =
14.9 Hz, 1 H), 3.24 (s, 2 H), 3.42 (s, 3 H), 3.54 (d, J = 14.9 Hz, 1 H),
3.69 (s, 2 H), 5.93 (d, J = 8.2 Hz, 1 H), 6.10 (s, 1 H), 7.33 (dq, J = 8.5,
4.2 Hz, 1 H), 7.41 (d, J = 4.3 Hz, 4 H) 7.98 (s, 1 H) (COOH proton
not observed). LC-MS (ES+) m/z 489 [M + H]. HRMS (ES+) calcd
for C26H36N2O5S (M + 1) m/z 489.2422, found 489.2423.
2-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)ethanesulfonic Acid Ammonium Salt (47). To a solution
of 8 (100 mg, 0.240 mmol) in DMF (4 mL) was added 2-
bromoethane sulfonic acid sodium salt (25.3 mg, 0.120 mmol). The
reaction mixture was stirred at 70 °C overnight, cooled to 25 °C,
acidified with TFA, and concentrated under reduced pressure.
Purification by RP-HPLC (eluting with MeCN/H2O with 0.05%
TFA-H2O and 0.05% TFA-MeCN), followed by further purification
with RP-HPLC (eluting with MeCN/H2O with 0.5% ammonium
hydroxide in H2O) afforded 47 (28 mg, 54%, ammonium salt) as a
3-[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]pentanedioic
Acid (52). Step 1: To a solution of 7 (109 mg, 0.262 mmol) in
toluene (3 mL) was added diethyl malonate (0.048 mL, 0.315 mmol)
and piperidine (0.013 mL, 0.131 mmol). The reaction mixture was
stirred at 100 °C overnight, cooled to 25 °C, and concentrated under
reduced pressure. Purification via SiO2 chromatography (EtOAc/Hex
= 10:90 to 80:20) afforded an intermediate diethyl propanedioate (146
mg, 90% pure, 90%) as a light-yellow oil. ES-LCMS m/z 558 (M +
H)+.
1
1
white solid. H NMR (MeOH-d4) δ ppm H NMR (MeOH-d4) δ
ppm 0.78 (t, J = 6.6 Hz, 3H), 0.87 (t, J = 7.4 Hz, 3H), 1.05−1.32 (m,
4H), 1.34−1.49 (m, 1H), 1.49−1.70 (m, 1H), 1.71−1.88 (m, J = 5.1
Hz, 1H), 2.09−2.31 (m, 1H), 2.96−3.15 (m, 3H), 3.30−3.38 (m, 2H),
3.49 (d, J = 15.0 Hz, 1H), 3.58 (s, 3H), 4.20 (s, 2H), 6.06 (s, 1H), 6.30
(s, 1H), 7.19−7.54 (m, 5H), 8.03 (s, 1H). LC-MS (ES+) m/z 525 (M
+ H)+. HRMS (ES+) calcd for C25H36N2O6S2 (M + 1) m/z 525.2093,
found 525.2093.
[({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
amino)methyl]phosphonic Acid Trifluoroacetate Salt (48). Step
1: Compound 8 (45.8 mg, 0.110 mmol) was combined with
paraformaldehyde (3.5 mg, 0.11 mmol) and diethyl phosphite
(0.014 mL, 0.11 mmol) in THF (4 mL) and stirred for 16 h at 75
°C, after which time all solvent was gone and LCMS indicated <20%
conversion. The THF was replaced and additional diethyl phosphite
(0.043 mL, 0.330 mmol) was added, and the mixture was heated an
additional 24 h at 75 °C, after which time LCMS indicated ∼50%
conversion. Additional diethyl phosphite (0.150 mL, 1.16 mmol) was
added and heating was continued for 24 h, after which time LCMS
indicated 70% completion. Additional paraformaldehyde (1.7 mg,
0.055 mmol) was added and heating was continued for 24 h, after
which time LCMS indicated complete conversion of starting material
to mono- and disubstituted products. The mixture was concentrated to
dryness and purified on silica gel eluting with 0−20% MeOH/DCM to
give an intermediate diethyl phosphonate (36 mg, 58% yield) as a clear
oil. LC-MS (ES+) m/z 567 [M + H].
Step 2: The diethyl phosphonate (36 mg, 0.064 mmol) dissolved in
DCM (1 mL) was treated with TMSBr (100 mg, 0.65 mmol) at 25 °C
for 18 h. The mixture was concentrated to dryness, and the crude
product was purified by RP-HPLC (30 mm × 75 mm , C18, 5 um
H2O Sunfire column at 55 mL/min; MeCN + 0.05% TFA and H2O +
0.05% TFA were used as the solvent system; 10−100 over 8 min, 100
to 100 to 10 min) to give 48 (6 mg, 18% yield) trifluoroacetate salt as
a white glass. 1H NMR (400 MHz, DMSO-d6) δ ppm 0.75 (t, J = 7.0
Hz, 3 H), 0.81 (t, J = 7.4 Hz, 3 H), 0.95−1.32 (m, 4 H), 1.33−1.56
(m, 2 H), 1.68−1.80 (m, 1 H), 1.98−2.12 (m, 1 H), 2.63−2.73 (m, 1
H), 3.00−3.13 (m, 2 H), 3.48 (s, 3 H), 3.57−3.66 (m, 2 H), 4.14−4.30
(m, 2 H), 5.86−6.02 (m, 1 H), 6.15 (s, 1 H), 7.31−7.50 (m, 5 H), 8.10
(s, 1 H) (3 exchangable protons not observed). LC-MS (ES−) m/z
Step 2: To a solution of diethyl malonate (0.096 mL, 0.628 mmol)
in EtOH (3 mL) was added sodium ethoxide (0.254 mL, 0.681
mmol). The reaction mixture was stirred at 25 °C for 20 min and
treated with a solution of the intermediate diethyl malonate (146 mg,
0.262 mmol) in EtOH (2 mL). The reaction mixture was stirred at 25
°C overnight, acidified with AcOH to pH 3−4, and partially
concentrated under reduced pressure to remove the organic solvents.
The residue was partitioned between water and DCM. The organic
layer was washed with saturated brine, dried, (Na2SO4), filtered, and
concentrated under reduced pressure. Purification using silica gel
(EtOAc/hexanes from 1:5 to 2:1) afforded an intermediate tetraethyl
1,1,3,3-propanetetracarboxylate (120 mg, 84% pure, 53%) as a
1
colorless oil. H NMR (400 MHz, CDCl3) δ ppm 0.66−0.92 (m,
6H), 0.97−1.10 (m, 6H), 1.17−1.27 (m, 8H), 1.34−1.54 (m, 2H),
1.74−1.90 (m, 1H), 2.04 (d, J = 14.7 Hz, 1H), 2.07−2.23 (m, 2H),
2.90 (d, J = 14.9 Hz, 1H), 3.37 (d, J = 14.9 Hz, 1H), 3.86−4.03 (m,
4H), 4.04−4.20 (m, 6H), 4.43 (br s, 1H), 5.97 (s, 1H), 6.08 (s, 1H),
7.26−7.48 (m, 5H), 7.97 (s, 1H). LC-MS (ES+) m/z 718 (M + H)+.
Step 3: A mixture of the tetraethyl 1,1,3,3-propanetetracarboxylate
(120 mg, 0.167 mmol) and 37% HCl (3 mL) was stirred at reflux for
20 h. LC-MS showed no starting material left but instead a mixture of
triacid and diacid. Additional 37% HCl (3 mL) was added, and the
reaction mixture was heated at reflux overnight. The reaction mixture
was cooled to 25 °C then concentrated under reduced pressure.
Purification with HPLC (eluting with MeCN/water with 0.05% TFA-
H2O and 0.05% TFA-MeCN) afforded 52 (76 mg, 95% pure, 68%) as
1
a white solid. H NMR (400 MHz, DMSO-d6) δ ppm 0.68−0.92 (m,
6H), 0.95−2.26 (m, 7H), 2.59 (d, J = 7.2 Hz, 4H), 3.46 (s, 3H), 3.55−
3.82 (m, 2H), 5.58−6.41 (m, 1H), 6.98−7.66 (m, 5H), 7.78 (br s,
1H), 12.14 (br s, 2H). LC-MS (ES+) m/z 518 (M + H)+. HRMS
(ES+) calcd for C27H35NO2S (M + 1) m/z 518.2211, found 518.2212.
R
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
2,2′-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
imino)diacetic Acid Hydrochloride Salt (53). Step 1: Prepared
from bromomethyl intermediate (see compound 18, step 2) (77 mg,
0.16 mmol) by heating a MeCN solution of diethyl iminodiacetic acid
(33 mg, 0.18 mmol) at 65 °C for 4 h then concentrating to a thick oil.
Chromatography on silica using hexanes/EtOAc provided a clear oil.
Step 2: H2O and THF were added to the product from step 1 along
with LiOH (15 mg, 0.36 mmol), and the mixture stirred for 2 h at 25
°C to hydrolyze the diacid. The reaction mixture was concentrated to
half volume, and 1 N HCl added to acidify the reaction contents then
the organics were extracted with DCM. The combined organics were
washed with brine, dried (Na2SO4), filtered, and concentrated. The
residue was triturated (DCM/EtOAc), filtered, and dried. The material
was redissolved in DCM/EtOAc/MeOH then concentrated to a solid
and kept under N2 overnight to dry. 53 was obtained as a white
(ES+) calcd for C25H38N2O9P2S (M + 1) m/z 605.1852, found
605.1853.
Method for Determination of Human, Mouse, and Rat ASBT
Inhibition. In preparation for measurement of bile acid uptake into
cells expressing ASBT, HEK293 cells were cultured in DMEM/F12
supplemented with 10% FBS. Twenty-four h prior to running an
experiment, cells were harvested when at a confluence of 80−90%.
Cells were seeded in poly D-lysine coated plates at 50000 cells per well,
and ASBT Bacmam virus was added such that each well contains 3.67
× 106 pfu (73.4 pfu/cell). Each assay plate was covered with Breathe
Easy Seal and placed in an incubator for 24 h to allow expression of
the transporter.
On the day of the uptake experiment, 10 mM HEPES was added to
Hank’s Balanced Salt Solution, and the pH was adjusted to 7.4 with
TRIS (HBSSH). The assay buffer was prepared by adding 100 pM
[3H]-taurocholate and 10 μM cold taurocholate to room temperature
HBSSH. A separate washing buffer was prepared by adding 10 μM
cold taurocholate to HBSSH (∼30 mL per assay plate) and placed on
ice. Using 100% DMSO, 8-point, 3-fold dilution curves for each test
compound was prepared starting at 200 μM. Similarly, an 8-point dose
response curve was prepared of the control compound 1 starting at 1.8
mM. Drug plates were created by adding 3 μL of each concentration to
a v-bottom 96-well plate then diluted 60-fold with 177 μL of assay
buffer. Plates were removed from the incubator and allowed to cool to
25 °C. Media was aspirated, and wells were washed once with 300 μL
of HBSSH. Then 50 μL of each dose response curve concentration
was added in triplicate by column to the assay plates, reserving column
10 for control (assay buffer + 1.67% DMSO) and columns 11 and 12
for the control compound. Plates were incubated at ambient
temperature for 90 min then the plates were aspirated then washed
1× with 300 μL of wash buffer. Then 220 μL of Microscint 20 was
added to each well, and the plates were sealed. The amount of [3H]-
taurocholate in cell lysate was quantitated using a microplate
scintillation counter on the following day.
1
hydrochloride salt (45 mg, 47% over 2 steps). H NMR (400 MHz,
DMSO-d6) δ ppm 0.62−0.90 (m, 6 H), 0.98−1.56 (m, 8 H), 1.75 (br
s, 1 H), 2.08 (br s, 1 H), 2.51 (br s, 14 H), 3.07 (d, J = 14.83 Hz, 1 H),
3.16−3.59 (m, 26 H), 3.81 (br s, 2 H), 5.95 (br s, 1 H), 6.09 (s, 1 H),
7.16−7.55 (m, 5 H), 8.03 (s, 1 H), 12.36 (br s, 2 H). LC-MS (ES+) m/
z 533 (M + H)+. HRMS (ES+) calcd for C27H36N2O7S (M + 1) m/z
533.2321, found 533.2321.
2,2′-({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-
phenyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}-
imino)diethanesulfonic Acid Trifluoroacetate Salt (54). To a
solution of 8 (25 mg, 0.060 mmol) in DMF (2 mL) was added 2-
bromoethane sulfonic acid sodium salt (127 mg, 0.600 mmol). The
reaction mixture was stirred at 70 °C overnight, cooled to room
temperature, acidified to pH 3−4 with AcOH, and concentrated under
reduced pressure. Purification by RP-HPLC (eluting with MeCN/
H2O with 0.05% TFA-H2O and 0.05% TFA-MeCN) afforded 54 (13
1
mg, 24%, TFA salt) as a white solid. H NMR (MeOH-d4) δ ppm
0.71−1.17 (m, 7H), 1.22−1.53 (m, 3H), 1.57−1.80 (m, 1H), 1.85−
2.27 (m, 2H), 2.60−2.85 (m,1H), 3.41−3.83 (m, 8H), 3.95 (d, J =
15.8 Hz, 1H), 4.40 (d, J = 13.5 Hz, 1H), 4.72 (d, J = 13.3 Hz, 1H),
6.53 (s, 1H), 6.68 (s, 1H), 7.42−7.80 (m, 5H), 8.23 (s, 1H). LC-MS
(ES+) m/z 633 (M + H)+. HRMS (ES+) calcd for C27H40N2O9S3 (M +
1) m/z 633.1974, found 633.1974.
Percent inhibition of uptake was determined using the following
formula at each drug concentration: 100 × (1 − ((T1 − C2)/(C1 −
C2))); where T1 is average cpm for the test compound, C1 is average
cpm observed in the absence of any added inhibitor, and C2 is average
cpm observed in the presence of a substance known to elicit 100%
inhibition of uptake (30 μM control compound). IC50s can be
generated using the formula, y = (Vmax × xn)/(Kn + xn).
[({[(3R,5R)-3-Butyl-3-ethyl-7-(methyloxy)-1,1-dioxido-5-phe-
nyl-2,3,4,5-tetrahydro-1,4-benzothiazepin-8-yl]methyl}imino)-
dimethanediyl]bis(phosphonic acid) Tetraammonium Salt
(55). Step 1: Compound 8 (155 mg, 0.37 mmol) was combined
with TsOH (cat.) in toluene (20 mL) and concentrated to dryness to
azeotrope off any residual H2O. Paraformaldehyde (11 mg, 0.37
mmol) and toluene (5 mL) were added, and the mixture was heated at
75 °C with vigorous stirring for 30 min, after which time diethyl
phosphite (0.048 mL, 0.37 mmol) was added. After 30 min, THF (20
mL) was added to the mixture, and the resultant homogeneous
reaction mixture was stirred overnight at 75 °C. The mixture was
concentrated to dryness and purified by RP-HPLC (30 mm × 150 mm
Sunfire column under acidic conditions; MeCN + 0.05% TFA and
H2O + 0.05% TFA were used as the solvent system; 30−100% over 10
min, 100−100% to 12 min) to give an intermediate tetraethyl
bis(phosphonate) (50 mg, 19% yield) as a yellow oil. LC-MS (ES+)
m/z 717 [M + H].
Method for Determination of MDCK Permeability. Passive
permeability was measured in vitro using stably transfected human
Multi-Drug Resistance 1−Madin−Darby Canine Kidney (hMDR1-
MDCK, Netherlands Cancer Institute) cells incubated under
conditions relevant to intestinal absorption. Briefly, hMDR1-MDCK
cells were seeded at 6.6 × 105 cells/well onto 12-well polycarbonate
Transwells filter membranes with 0.4 μm pore size (Corning
(Corning, NY)) and maintained in Dulbecco’s Modified Eagle’s
Media containing 10% fetal bovine serum (DMEM-FBS) at 37 °C in
an atmosphere of 5% CO2 and 95% relative humidity. After three days,
media was removed from both the apical and basolateral chambers and
replaced with transport buffer (HBSS containing 25 mM glucose and
25 mM HEPES) containing the P-gp inhibitor GF120918A at a final
concentration of 2 μM. After a 30 min equilibration, the transport
buffer was removed from the apical chambers and replaced with fasted-
state simulated intestinal fluid (FaSSIF) containing 3 μM test
compound, 2 μM GF120918A, 25 mM glucose, and 250 μM Lucifer
Yellow CH. Next, the transport buffer was removed from the
basolateral chambers and replaced with transport buffer containing 1%
(w/v) human serum albumin and 2 μM GF120918A.
After 60 min incubation at 37 °C, samples were collected from the
apical (donor) and basolateral (receiver) compartments and added to
acetonitrile (1:1 and 1:2 (v/v), respectively). Receiver samples were
then centrifuged and the supernatants were removed and analyzed by
LC-MS/MS.
The final DMSO concentration in all dose solutions was 0.3% (v/v).
Each treatment was performed in duplicate. Propranolol, a high
permeability marker compound, and amprenavir, a marker compound
for P-gp activity, were included in separate wells as controls for the
Step 2: The tetraethyl bis(phosphonate) (100 mg, 0.14 mmol)
dissolved in DCM (5 mL) was treated with TMSBr (171 mg, 1.12
mmol) at 25 °C overnight. The mixture was concentrated to dryness,
and the crude product was purified by RP-HPLC (30 mm × 100 mm
XBridge column under basic conditions; MeCN + 0.1% NH4OH and
H2O + 0.1% NH4OH were used as the solvent system; 10% to 55%
over 6 min, 100% to 100% to 8 min) to give 55 (45 mg, 54% yield) as
1
an ammonium salt. H NMR (400 MHz, DMSO-d6) δ ppm 0.75 (t, J
= 6.9 Hz, 3 H), 0.80 (t, J = 7.5 Hz, 3 H), 0.97−1.29 (m, 4 H), 1.32−
1.55 (m, 2 H), 1.68−1.81 (m, 1 H), 1.98−2.14 (m, 1 H), 2.58−2.77
(m, 5 H), 3.09 (d, J = 14.8 Hz, 1 H), 3.41 (s, 3 H), 3.52 (d, J = 15.0
Hz, 1 H), 3.75−3.96 (m, 2 H), 5.94 (d, J = 8.6 Hz, 1 H), 6.08 (s, 1 H),
7.25−7.37 (m, 1 H), 7.36−7.49 (m, 4 H), 7.90 (s, 1 H) (phosphonic
acid protons not observed). LC-MS (ES+) m/z 605 [M + H]. HRMS
S
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
assay. Cell monolayer integrity was assessed by measuring Lucifer
yellow transport via a fluorescence-based assay.
Fecal Collection in Rat. Male ZDF rats arrived at seven weeks of
age ( 3 days). After a one-week acclimation period, rats were assigned
to different treatment groups (n = 6−8/group) based upon baseline
glucose/vehicle (0.5% hydroxypropyl methylcellulose (HPMC), 0.1%
Tween80); one vehicle group for each compound) and six doses (0.05,
0.1, 0.5, 1, 5, and 10 mg/kg) of compounds 20, 45, and 56. All
treatments were given via oral gavage twice a day. Fecal samples were
collected for 24 h on day 7 of treatment.
Metabolic Effects in ZDF Rats. Male ZDF rats arrived at seven
weeks of age ( 3 days). After a one-week acclimation period, rats were
anesthetized with isoflurane (Abbott Laboratories, IL) and tail-vein
blood samples were collected at 0900 without fasting. To ensure
balanced treatment groups, ZDF rats were assigned to six treatment
groups based upon baseline glucose/vehicle (0.5% hydroxypropyl
methylcellulose (HPMC), 0.1%Tween80) and selected doses of
compounds (0.05, 0.1, 0.5, 1, 5, and 10 mg/kg or 0.001, 0.01, 0.1,
1, and 10 mg/kg for compounds 20 and 45 or 56, respectively). All
treatments were given via oral gavage twice a day, and animals were
followed for two weeks with blood samples collected from tail vein on
day 14 at 0900 without fasting. Plasma samples were stored at −80 °C
for further analyses.
Measurement of Clinical Chemistry Parameters. Plasma glucose
and bile acids in fecal extraction were measured using the Olympus
AU640 clinical chemistry analyzer (Beckman Coulter, Irving, TX).
Glucose test reagents were manufactured by Beckman Coulter. Bile
acids reagents were manufactured by Trinity Biotech. HbA1c was
measured by the Primus Affinity Ultra2 HPLC system using Primus
Affinity Assay reagents (Primus Diagnostics, Kansas City, MO).
Insulin, total GLP-1 (tGLP-1), PYY, and GIP were assayed using the
Meso Scale Discovery (MSD) assay kits. Total Glp-1 was assayed by
the MSD Total Glp-1 assay kit and analyzed on an MSD Sector
Imager 6000 (Meso Scale Discovery, Gaithersburg, MD).
Fecal Bile Acid Extraction. Fecal samples were air-dried for five
days and extracted in methanol−KOH (300 mM) at 60 °C for 24 h.
Fecal extract was then mixed with 150 mM Mg2SO4 (1:1). After
centrifugation, the supernatant was saved and submitted for bile acids
measurement as described above.
Rat Luminal Contents Stability Assay. A 10% (w/v)
homogenate of the luminal contents from rat cecum and colon in
phosphate buffered saline (PBS, pH 7.4) was prepared as follows.
Two male SD rats (Charles River Laboratories, Raleigh, NC) were
fasted overnight and euthanized by CO2 asphyxiation, followed by
exsanguination. The large intestine and cecum were removed from
both animals and cut lengthwise. The luminal contents were removed,
pooled into a preweighed 50 mL conical tube, diluted with PBS (10
mL/g sample weight), and gently mixed by inversion. The
homogenate was placed on wet ice until use.
Test compounds (10 μM final concentration) were added to a 4 mL
glass screwcap vial containing 3 mL of the homogenate of the luminal
contents from rat cecum and colon. Immediately after the addition of
test compound, the vial was gently mixed and 3 × 100 μL aliquots
were removed (t = 0) and placed into a 96-deepwell plate containing
400 μL of stopping solution (80% acetonitrile/20% methanol). Next,
the glass vial was purged under a gentle stream of nitrogen gas for
approximately 30 s, capped, and placed in a 37 °C shaking water bath.
At t = 2, 4, and 24 h, 3 × 100 μL aliquots were removed from the vial
and placed into a 96-deepwell plate containing 400 μL of stopping
solution. The vial was purged under a gentle stream of nitrogen gas for
approximately 30 s, capped, and placed in a 37 °C shaking water bath
after each time point. Samples were covered and stored at −10 °C
until LC-MS/MS analysis.
Rat Oral Absorption Assay. Male Sprague−Dawley (SD) rats
(271−303 g; Charles River Laboratories, Raleigh, NC) were housed
with free access to standard chow (PMI 5002 block chow) and water,
unless otherwise noted. Animals for intravenous treatment groups
were surgically implanted with a jugular and femoral vein cannula.
Animals for oral treatment groups were surgically implanted with a
jugular and portal vein cannula. Food was withheld from rats overnight
prior to dosing and was returned at approximately 4 h postdose.
Oral treatment groups received test compounds formulated as a
homogeneous suspension in 0.5% HPMC/0.1% Tween via oral gavage
at a dose of 10 mg/kg. Blood samples were collected from both jugular
and portal vein cannulae at 0.25, 0.5, 1, 2, 4, and 8 h postdose. Plasma
samples were prepared and stored at −70 °C until analysis.
Data Analysis. All of the in vivo studies described above were
analyzed as described in this paragraph.
Values are given as mean SEM. The data were analyzed in JMP
7.0.0 (SAS Institute, 174 Cary, NC) using one-way ANOVA with
prespecified contrasts to compare each group to the appropriate
control group (fecal BA data) or two-way ANOVA with Bonferroni
post-tests. P values <0.05 were considered to indicate a significant
difference between treatment groups. Dose−responses were fit to four-
parameter logistic functions (GraphPad Prism v5.04, San Diego, CA).
Rat Fecal Drug Recovery. Fecal Recovery. Male SD rats (Charles
River Laboratories, Raleigh, NC) were administered test compounds
formulated as a homogeneous suspension in 0.1% HPMC:0.5% Tween
via oral gavage at a dose of 10 mg/kg. Fecal samples were collected
across the following intervals: 0−6, 6−12, 12−24, 24−36, 36−48, 48−
60, and 60−72 h postdose. After each collection interval, the samples
were capped and stored at −70 °C until analysis. Prior to analysis, the
samples were diluted with 5 volumes of 20% EtOH:80% H2O, soaked
overnight at 10 °C, and then homogenized using a Polytron hand-held
homogenizer. The homogenates were extracted with 3 volumes of
acetonitrile and then centrifuged for 15 min at 2304g and 4 °C.
Aliquots of each acetonitrile supernatant was transferred to clean 96-
well plates and diluted with an equal volume of water. Drug
concentrations were quantified via LC-MS/MS.
Animals for Efficacy Studies. Male C57BL/6J mice were purchased
from Jackson Laboratories (Bar Harbor, Maine) at seven weeks old
and fed with normal rodent diet (Purina 5001, Harlan Teklad,
Indianapolis, IN). Male Zucker Diabetic Fatty (ZDF/GmiCrl-fa/fa)
mice were purchased from Charles River (Raleigh, NC) and had free
access to rodent food (Purina 5008, Harlan Teklad, Indianapolis, IN).
All animals were housed under controlled conditions (12/12 light−
dark cycle, 24 °C and 50% relative humidity). All procedures
performed were in compliance with the Animal Welfare Act and
U.S. Department of Agriculture regulations and were approved by the
GlaxoSmithKline Animal Care and Use Committee.
AUTHOR INFORMATION
■
Corresponding Author
*email: Jon.L.Collins@gsk.com, tel 919-483-5694.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors would like to thank Erin D. Swiger for purification
support as well as Mike Lewandowski and Iris Paulus for
generation of HRMS data, and the many other GlaxoSmithK-
line scientists who made ancillary contributions to the work
that is described in this manuscript.
ABBREVIATIONS USED
■
Fecal Collection in Mice. Male C57BL/6J mice were dosed with
vehicle (0.5% hydroxypropyl methylcellulose (HPMC), 0.1%
Tween80) or six doses (0.0001, 0.001, 0.01, 0.1, 1, and 10 mg/kg)
of compounds at 0700 and 1500 for one day, and fecal samples were
collected for 24 h (0700−0700). Animals were used for up to five
studies with one week washout between studies.
ASBT, apical sodium-dependent bile acid transporter; BA, bile
acids; DPP-4, dipeptidyl peptidase-4; GLP-1, glucagon-like-1;
GIP, glucose dependent insulinotropic peptide; HEK293,
human embryonic kidney 293; iBAT, ileal bile acid transporter;
HbA1c, hemoglobin A1c; hMDR1-MDCK, human multidrug
T
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(18) Cornforth, J.; Wilson, J. R. H. A general reagent for O-
phosphonomethylation of phenols. J. Chem. Soc., Perkin Trans. 1 1994,
14, 1897−1900.
resistance 1−Madin−Darby canine kidney; PYY, peptide YY;
tGLP-1, total GLP-1; ZDF, Zucker Diabetic Fatty
(19) Fields, E. K. The synthesis of esters of substituted amino
phosphonic acids. J. Am. Chem. Soc. 1952, 74, 1528−1531.
(20) Stefely, J. A.; Palchaudhuri, R.; Miller, P. A.; Peterson, R. J.;
Moraski, G. C.; Hergenrother, P. J.; Miller, M. J. N-((1-Benzyl-1H-
1,2,3-triazol-4-yl)methyl)arylamide as a new scaffold that provides
rapid access to antimicrotubule agents: synthesis and evaluation of
antiproliferative activity against select cancer cell lines. J. Med. Chem.
2010, 53, 3389−3395.
(21) Brown, N. J.; Rumsey, R. D. E.; Read, N. W. Adaptation of
hydrogen analysis to measure stomach to caecum transit time in the
rat. Gut 1987, 28, 849−854.
(22) Thomas, C.; Gioiello, A.; Noriega, L.; Strehle, A.; Oury, J.;
Rizzo, G.; Macchiarulo, A.; Yamamoto, H.; Mataki, C.; Pruzanski, M.;
Pellicciari, R.; Auwerx, J.; Schoonjans, K. TGR5-mediated bile acid
sensing controls glucose homeostasis. Cell Metab. 2009, 10, 167−177.
(23) Dalmer, O.; Diehl, C.; Pieper, H. Derivatives of tertiary amino
aliphatic acids. (Merck & Co., Inc., Rahway, NJ). U. S. Patent
US2,370,015, February 20, 1945.
REFERENCES
■
(1) Recent facts about diabetes can be found via the Internet at the
World Health Organization website: Diabetes, Fact Sheet no. 312:
World Health Organization: Geneva, March, 2013; http://www.who.
int/mediacentre/factsheets/fs312/en/.
(2) Inzucchi, S. E.; Bergenstal, R. M.; Buse, J. B.; Diamant, M.;
Ferrannini, E.; Nauck, M.; Peters, A. L.; Tsapas, A.; Wender, R.;
Matthews, D. Management of hyperglycemia in type 2 diabetes: a
patient-centered approach. Position statement of the American
Diabetes Association (ADA) and the European Association for the
Study of Diabetes (EASD). Diabetes Care 2012, 35, 1364−1379.
(3) Todd, J. R.; Bloom, S. R. Incretins and other peptides in the
treatment of diabetes. Diabetic Med. 2007, 24 (3), 223−232.
(4) Holst, J. J.; Vilsbøll, T.; Deacon, C. F. The incretin system and its
role in type 2 diabetes mellitus. Mol. Cell. Endocrinol. 2009, 297 (1−2),
127−136.
(5) Creutzfeldt, W. The entero-insular axis in type 2 diabetes
incretins as therapeutic agents. Exp. Clin. Endocrinol. Diabetes 2001,
109 (Suppl. 2), S288−S303.
(6) Plaisancie, P.; Dumoulin, V.; Chayvialle, J.-A.; Cuber, J.-C.
̇
Luminal glucagon-like peptide-1(7−36) amide-releasing factors in the
isolated vascularly perfused rat colon. J. Endocrinol. 1995, 145 (3),
521−526.
(7) Dumoulin, V.; Moro, F.; Barcelo, A.; Dakka, T.; Cuber, J.-C.
Peptide YY, glucagon-like peptide-1, and neurotensin responses to
luminal factors in the isolated vascularly perfused rat ileum.
Endocrinology 1998, 139 (9), 3780−3786.
(8) Adrian, T. E.; Ballantyne, G. H.; Longo, W. E.; Bilchik, A. J.;
Graham, S.; Basson, M. D.; Tierney, R. P.; Modlin, I. M. Deoxycholate
is an important releaser of peptide YY and enteroglucagon from the
human colon. Gut 1993, 34 (9), 1219−1224.
(9) Hofmann, A. F. The enterohepatic circulation of bile acids in
mammals: form and functionas. Front. Biosci. 2009, 14, 2584−2598.
(10) Roberts, M. S.; Magnusson, B. M.; Burczynski, F. J.; Weiss, M.
Enterohepatic circulation: physiological, pharmacokinetic and clinical
implications. Clin. Pharmacokinetics 2002, 41 (10), 751−790.
(11) Schneider, B. L. Intestinal bile acid transport: biology,
physiology, and pathophysiology. J. Pediatr. Gastroenterol. Nutr.
2001, 32, 407−417.
(12) Root, C.; Smith, C. S.; Sundseth, S. S.; Pink, H. M.; Wilson, J.
G.; Lewis, M. C. Ileal bile acid transporter inhibition, CYP7A1
induction, and antilipemic action of 264W94. J. Lipid. Res. 2002, 43,
1320−1330.
(13) Kramer, W.; Glombik, H. Bile acid reabsorption inhibitors
(BARI): novel hypolipidemic drugs. Curr. Med. Chem. 2006, 13, 997−
1016.
(14) Chen, L.; Yao, X.; Young, A.; McNulty, J.; Anderson, D.; Liu, Y.;
Nystrom, C.; Croom, D.; Ross, S.; Collins, J.; Rajpal, D.; Hamlet, K.;
Smith, S.; Gedulin, B. Inhibition of apical sodium-dependent bile acid
transporter as a novel treatment for diabetes. Am. J. Physiol. Endocrinol.
Metab. 2012, 302, E68−E76.
(15) Lundasen, T.; Andersson, E.-M.; Snaith, M.; Lindmark, H.;
̊
̋
Lundberg, J.; Ostlund-Lindqvist, A.-M.; Angelin, B.; Rudling, M.
Inhibition of intestinal bile acid transporter slc10a2 improves
triglyceride metabolism and normalizes elevated plasma glucose levels
in mice. PLoS ONE 2012, 7 (5), 1−14.
(16) Brieaddy, L. E. Hypolipidemic 1,4-benzothiazepen-1,1-dioxides.
(Wellcome Foundation, London). Patent WO9605188, February 22,
1996.
(17) Ghosh, M.; Christopher, K.; Zhang, X.; Tarrant, J. G.; Lee, C.-
W.; Kaiser, B.; Zablocki, M.-M.; Chenard, B. L.; Doller, D.; Yamaguchi,
Y.; Wustrow, D. J.; Li, G. Aminopiperidines and related compounds a
MCH modulators and their preparation, pharmaceutical compositions
and use in the treatment of metabolic, feeding and sexual disorders.
(Neurogen Corporation) Patent WO2008016811, July 25, 2007.
U
dx.doi.org/10.1021/jm400459m | J. Med. Chem. XXXX, XXX, XXX−XXX