Pyrrole[2,3-d]azepino compounds as agonists of the farnesoid X receptor (FXR)

Article abstract of DOI:10.1016/j.bmcl.2009.07.148

Pyrrole[2,3-d]azepines have been identified as potent agonists of the farnesoid X receptor (FXR). Based on the planar X-ray crystal structure of WAY-362450 1 in the ligand binding domain and molecular modeling studies, non-planar reduced compounds were de

Full text of DOI:10.1016/j.bmcl.2009.07.148

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5289–5292
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Pyrrole[2,3-d]azepino compounds as agonists of the farnesoid X receptor (FXR)
a
a
b
John F. Mehlmann a, Matthew L. Crawley a, , Joseph T. Lundquist IV , Ray J. Unwalla , Douglas C. Harnish ,
*
Mark J. Evans b, Callain Y. Kim a, Jay E. Wrobel a, Paige E. Mahaney a
a Chemical Sciences, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
b Cardiovascular and Metabolic Disease Research, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Pyrrole[2,3-d]azepines have been identified as potent agonists of the farnesoid X receptor (FXR). Based on
the planar X-ray crystal structure of WAY-362450 1 in the ligand binding domain and molecular model-
ing studies, non-planar reduced compounds were designed which led to agonists that exhibit high aque-
ous solubility and retain moderate in vitro potency.
Received 24 July 2009
Revised 29 July 2009
Accepted 30 July 2009
Available online 3 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
FXR
Nuclear hormone receptor
Rational design
The farnesoid X receptor (FXR) is a nuclear hormone receptor
expressed in the liver, kidney and intestine that is activated by bile
acids.1,2 Upon activation, FXR binds to DNA as a heterodimer with
the retinoid X receptor (RXR).3 This binding event regulates the
expression of a number of genes and proteins involved in bile acid
and cholesterol homeostasis (CYP7A1, SHP, IBABP, BSEP, and Apo
A-I)4, triglyceride synthesis5, and lipogenesis (SREBP-1c and Apo
C-III).6 Early FXR agonists reported in the literature include
GW4064,7 fexaramine, fexarene, and fexarine.8 Although these
compounds were shown to activate FXR, they suffered from poor
ADME properties. These reports demonstrated that agonism of
FXR lowers triglyceride levels in animals9 and the mechanism of
action suggests a potential to reduce total cholesterol levels as
well.10
Recently, WAY-362450 1 (EC50 = 5 nM in a transient transfec-
tion assay using hFXR on an ECREx7-TK-Luc construct in CV-1
cells), which contains a novel azepino[4,5-b]indole core structure
was disclosed as a potent and selective FXR agonist that advanced
to clinical trials (Fig. 1).11 When 1 was dosed in LDL receptor
knockout (LDLR KO) mice, statistically significant reductions in
both cholesterol and triglyceride levels were observed; however,
the poor aqueous solubility of 1 necessitated the use of non-aque-
ous vehicles such as corn oil for oral dosing. The limited solubility
denced by its high melting point of 174–175 °C, and may therefore
contribute to its poor aqueous solubility.
To potentially improve the solubility of 1, we examined com-
pounds in which the indole ring was replaced with a 2-cyanopyr-
role moiety. The use of
a pyrrole[2,3-d]azepino core would
decrease both the cLog P and the molecular weight of the 1 analogs.
In addition, to disrupt the planarity of the core, we investigated
analogs in which the azepine olefin was reduced, therefore dimin-
ishing the planarity of the molecule and the potential for tight
crystal packing.
The synthesis of the pyrrole[2,3-d]azepino analogs began with
the formylation of 1H-pyrrole-2-carboxylic acid methyl ester 2
using dichloromethoxymethane and aluminum chloride to yield
compound 3. Treatment of aldehyde 3 with dimethyl amine and
sodium triacetoxyborohydride followed by treatment with methyl
iodide provided the quaternary ammonium salt 4 which was re-
acted with sodium cyanide to give the nitrile 5. After Boc protec-
tion of the pyrrole nitrogen, the acetonitrile was dimethylated,
and the protecting group was removed to afford intermediate 6.
Reduction of the nitrile using Raney Nickel provided the primary
amine which was Boc protected prior to hydrolysis of the methyl
of
1 could be attributed, in part, to its high lipophilicity
O
F
(cLog P = 5.30). In addition, a small molecule crystal structure of
1 revealed the exceedingly planar nature of the molecule that
may result in a high degree of crystal packing, additionally evi-
N
F
N
O
H
O
1
* Corresponding author. Tel./fax: +1 484 865 9097.
E-mail address: crawlem@wyeth.com (M.L. Crawley).
Figure 1. The azepino[4,5-b]indole WAY-362450 (1).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.148

5290
J. F. Mehlmann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5289–5292
Table 1
ester with sodium hydroxide to form compound 7. Conversion of
the carboxylic acid to the nitrile was accomplished over two steps,
namely, treatment with thionyl chloride followed by trifluoroace-
tic anhydride. Removal of the Boc group with 6 N HCl provided
hydrochloride salt 8 which underwent a one-pot Pictet–Spengler
cyclization-rearrangement reaction to provide the desired pyr-
role[2,3-b]azepino core 9. Subsequent acylation of the azepine
nitrogen with various acyl chlorides furnished targets 10a–h
(Scheme 1).
The compounds were evaluated using an FXR functional assay
that measured their ability to activate the gal4-hFXR-LBD fusion
protein in HEK293 cells. Maximal efficacy was reported as a per-
centage of the maximal efficacy observed in this assay with
GW4064. FXR functional potency and efficacy data for this series
of pyrrole[2,3-b]azepine analogs are shown in Table 1. All eight
of the pyrrole[2,3-d]azepino analogs were either equipotent to 1
or exhibited an improvement in potency. Indeed, the direct ana-
logue of 1, 10a (EC50 = 6.2 nM) was three times more potent than
the indoleazepine lead and the 3-fluorophenyl compound 10b
(EC50 = 2.5 nM) was, at the time, the most potent FXR agonist pre-
pared in our laboratories. Unfortunately, despite the significantly
reduced clog P, none of the analogues exhibited solubility in aque-
ous pH 7.4 buffer.12
FXR functional activity and aqueous solubility for the pyrrole[2,3-d]azepino analogs
O
N
R
NC
N
H
O
O
Eff.b (%)
clog P
Sol.c
(lg/mL)
a
Compd
R
EC50 (nM)
10a
10b
10c
10d
10e
10f
10g
10h
1
3,4-Di-F-phenyl
3-F-Phenyl
4-F-Phenyl
6.3
2.5
5.0
6.5
35
12
8.5
16
121
120
119
124
116
122
94
3.89
3.79
3.79
4.91
3.18
4.36
4.59
3.35
5.30
BLDd
BLD
BLD
BLD
BLD
BLD
BLD
BLD
BLD
Cyclohexyl
4-Cyanophenyl
3-Cl-Phenyl
3-CF3-Phenyl
2-Thienyl
114
130
N/A (Figure 1)
16
a
Induction of FXR measured in HEK293 cells stably co-transfected with an
expression plasmid for an FXR-LBD-gal4 DNA binding fusion protein and a Luc12
luciferase reporter gene construct.
b
Efficacy = [maximal fold induction of test compound/maximal fold induction of
GW4064 Â 100].
c
Solubility as measured in an aqueous pH 7.4 buffer using
instrument and software.
a pION PSR4s
d
BLD = Below the limit of detection.
To gain insight into the pharmacokinetic parameters of these
analogs, compound 10a was orally dosed to C57 mice (30 mg/kg
in 2% Tween80/0.5% methylcellulose, an aqueous formulation in
which the compound had adequate solubility, i.e., >0.10 mg/mL).
However, no detectable plasma concentration of the compound
was measured. This result underscores the limited solubility of
the pyrrole[2,3-d]azepino analogs (10a–h) and their low absorp-
tion in vivo.
The previously disclosed crystal structure of 1 (PDB code 3FLI)
bound to the FXR ligand-binding domain was used for docking
studies to understand the binding mode of 2-cyanopyrrole and
shed further light on residues which are important for ligand rec-
ognition within the binding site. As shown in Figure 1, the docked
binding mode13 of 10a is similar to that of 1 in that ligand recogni-
tion is achieved through a combination of a specific hydrogen bond
and large number of hydrophobic interactions observed between
the ligand and FXR residues (Fig. 2).
bonyl oxygen of the amide group was in position to accept a hydro-
gen bond from the hydroxyl group of the Tyr373 residue while an
intramolecular hydrogen bond was seen between the indole NH of
the pyrrole and the ester carbonyl group (though it had an impro-
per dihedral angle for N–H–O–C of ꢀ11° out of plane). This internal
hydrogen bond rigidified this template further and allowed the iso-
propyl ester group to orient towards the narrow hydrophobic
pocket surrounded by Ile339 residue.
Previous SAR studies suggested that a longer alkyl chain
branched or linear in this hydrophobic Ile339 cavity was critical
for ligands potency. We hypothesized that the reduction of the vi-
nyl amide bond of 10a would twist the azepine ring into a non-pla-
nar conformation and result in a loss of the intramolecular
hydrogen bond. This conformational change would also effect
how the isopropyl group was positioned into the narrow hydro-
phobic cavity.
The seven-membered azepine ring was oriented in a slightly
puckered conformation, that is, in a twist-chair with the C2 group
twisted out of the plane with respect to the pyrrole ring. The car-
To test this hypothesis and further explore the solubility of this
series of compounds, analogues were prepared that lacked the
O
H
CN
N
i
ii, iii
iv
O
O
O
O
I
N
H
N
H
N
H
N
H
O
O
O
O
3
4
5
2
CN
xi, xii, xiii
viii, ix, x
v, vi, vii
NHBoc
HO
O
N
H
N
O
O
H
7
6
O
xiv, xv
N
xvi
NH
R
NH2-HCl
NC
NC
N
NC
N
H
N
H
H
O
O
O
O
8
10a-h
9
Scheme 1. Reagents and conditions: (i) dichloromethyl methyl ether, AlCl3, CH2Cl2, 0 °C, 50%; (ii) Me2NH, Na(OAc)3BH, THF, 20 °C; (iii) MeI, THF, 20 °C; (iv) NaCN, DMF,
120 °C, 55% over 3 steps; (v) Boc2O, Et3N, DMAP, THF, 20 °C; (vi) NaH, MeI, DMF, 0–20 °C; (vii) TFA, DCM, 20 °C, 77% over 3 steps; (viii) H2, Raney Ni, MeOH, NH4OH, THF,
20 °C; (ix) Boc2O, Et3N, DMAP, THF, 20 °C; (x) NaOH, CH3CN, 70 °C, 72% over 3 steps; (xi) SOCl2, CH2Cl2, 0–20 °C then NH3, Et2O, 20 °C; (xii) TFAA, py, 0–20 °C; (xiii) 6 N HCl,
20 °C, 49% over 3 steps; (xiv) isopropylbromopyruvate, (CH3)2CHOH, CH3CN, 80 °C; (xv) Py, DMAP, 80 °C, 62% over 2 steps; (xvi) RC(O)Cl, Et3N, CH3CN, 20 °C.

J. F. Mehlmann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5289–5292
5291
Table 2
FXR functional activity and aqueous solubility for the reduced pyrrole[2,3-d]azepino
analogs
O
N
R
NC
N
H
O
O
EC50 (nM) Eff.b (%) cLog P Sol.c
(
lg/mL)
a
Compd Structure
11a
11b
11c
11d
11e
11f
11g
11h
11i
3,4-Di-F-phenyl
3-F-Phenyl
4-F-Phenyl
Cyclohexyl
4-Cyanophenyl
3-Cl-Phenyl
3-CF3-Phenyl
4-Cl-Phenyl
280
1300
640
1900
5700
1300
2700
4100
115
84
85
127
91
103
114
115
112
3.67
3.57
3.57
3.74
2.96
4.14
4.36
4.14
1.34
31
>100
>100
>100
21
17
10
5
>100
4-Tetrahydropyran 3300
Figure 2. The docked binding mode of 10a (in yellow) in the binding pocket of the
hFXR X-ray co-crystal structure (from a complex with 1). Only critical residues
within the binding site have been shown for clarity. Hydrogen bonds are shown by
dotted lines (white intramolecular, yellow intermolecular).
a
Induction of FXR measured in HEK293 cells stably co-transfected with an
expression plasmid for an FXR-LBD-gal4 DNA binding fusion protein and a Luc12
luciferase reporter gene construct.
b
Efficacy = [maximal fold induction of test compound/maximal fold induction of
GW4064 Â 100].
c
Solubility as measured in an aqueous pH 7.4 buffer using
instrument and software.
a pION PSR4s
azpine olefin. To access these derivatives, the pyrrole[2,3-b]azepi-
no core 9 was treated with sodium cyanoborohydride and subse-
quently acylated (Scheme 2) to yield compounds 11a–i as
racemic mixtures. The in vitro activity and aqueous solubility of
these analogs is shown in Table 2.
A distinct improvement in solubility was observed among all
nine of the analogues versus the unsaturated compounds. Four of
the reduced compounds (11b–d, 11i) were soluble in aqueous
media at levels greater than 100 lg/mL. Clearly, the reduction of
the azepine olefin led to a dramatic improvement in solubility that
may be attributed to the disruption of the planar nature of the
compounds. However, a significant loss of potency was seen when
the azepine olefin was removed. The 3,4-difluorophenyl analogue
11a proved to be the most potent (EC50 = 280 nM) and the 4-fluo-
rophenyl analogue 11c (EC50 = 640 nM) was the only other ana-
logue to possess an EC50 under 1000 nM.
Further docking studies carried out on the saturated analog 11a
revealed that although one of the enantiomers (i.e., S) was able to
fit and orient the isopropyl group in a similar position as seen in
the X-ray structure of 1, the azepine ring was in an energetically
unfavorable conformation, that is, twist-boat (Fig. 3). The ester car-
bonyl group was further twisted out of plane and resulted in loss of
the intramolecular hydrogen bond (as it had an improper dihedral
angle for N–H–O–C of ꢀ38° out of plane). A conformational search
on the unbound ligand 11a revealed that none of the low energy
conformers within a ꢀ3 kcal energy window, favored a ligand-
Figure 3. The docked binding mode of 11a (in white) in the binding pocket of the
hFXR X-ray co-crystal structure (from a complex with 1). Only critical residues
within the binding site have been shown for clarity. Hydrogen bonds are shown by
dotted yellow lines. The ester carbonyl group is now twisted out of plane and not in
position to form hydrogen bond with NH of pyrrole.
bound-like conformation of 10a, that is, azepine ring in
a
Furthermore, by disrupting the planarity of the core azepine ring
system, we have identified the first aqueous soluble analogues of
the series, albeit with an associated loss of potency.
twist-chair conformation and the presence of an intramolecular
hydrogen bond. We believe that this high conformational energy
cost for 11a was responsible for the reduction in ligand potency.
However, we cannot rule out that other factors such as ligand
desolvation energy and electronic effects could also play a role.
In conclusion, we have discovered a novel class of pyrrole[2,3-
d]azepines with low nanomolar in vitro potency as FXR agonists.
Acknowledgment
The authors would like to acknowledge the contributions of the
Wyeth Discovery Analytical Chemistry (DAC) group in the analysis
of these compounds.
O
References and notes
N
O
NH
O
i, ii
R
1. (a) Lefebvre, P.; Cariou, B.; Lien, F.; Kuipers, F.; Staels, B. Physiol. Rev. 2009, 89,
147; (b) Forman, B. M.; Goode, E.; Chen, J.; Oro, A. E.; Bradley, D. J.; Perlmann,
T.; Noonan, D. J.; Burka, L. T.; McMorris, T.; Lamph, W. W.; Evans, E. M.;
Weinberger, C. Cell 1995, 81, 687–693.
2. (a) Makishima, M.; Okamoto, A. Y.; Repa, J. J.; Tu, H.; Learned, R. M.; Luk, A.;
Hull, M. V.; Lustig, K. D.; Mangelsdorf, D. J.; Shan, B. Science 1999, 284, 1362; (b)
Parks, D. J.; Blanchard, S. G.; Bledsoe, R. K.; Chandra, G.; Consler, T. G.; Kliewer,
NC
NC
N
N
H
H
O
O
11a-i
9
Scheme 2. Reagants and conditions: (i) NaCNBH4, AcOH, 20 °C, 61%; (ii) RC(O)Cl,
Et3N, CH3CN, 20 °C, 80%.

5292
J. F. Mehlmann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5289–5292
S. A.; Stimmel, J. B.; Willson, T. M.; Zavacki, A. M.; Moore, D. D.; Lehmann, J. M.
Science 1999, 284, 1365.
9. Bell, G. D.; Lewis, B.; Petrie, A.; Dowling, R. H. Br. Med. J. 1973, 3, 520.
10. Lambert, G.; Amar, M. J. A.; Guo, G.; Brewer, H. B.; Gonzalez, F. J.; Sinal, C. J. J.
Biol. Chem. 2003, 278, 2563.
11. (a) Flatt, B.; Martin, R.; Wang, T. L.; Mahaney, P. E.; Murphy, B.; Gu, X.-H.;
Foster, P.; Li, J.; Pircher, P.; Petrowski, M.; Schulman, I.; Westin, S.; Wrobel, J.;
Yan, G.; Bischoff, E.; Daige, C.; Mohan, R. J. Med. Chem. 2009, 52, 904; (b) Busch,
B.; Flatt, B. T.; Gu, X.-H.; Martin, R.; Mohan, R.; Wang, T.-L.; Wu, J. H. U.S. Patent
Application 0054634/A1, 2005.
12. (a) Di, L.; Kerns, E. Int. J. Pharm. 2006, 317, 54; (b) Kerns, E.; Di, L. Drug Like
Properties: Concepts, Structure Design and Methods; Elsvier, 2008. p 526.
13. (a) Glide, version 5.0, Schrodinger, LLC, New York, NY, 2008.; (b) Friesner, R. A.;
Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.; Repasky, M. P.;
Knoll, E. H.; Shelley, M.; Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S.; Glide
J. Med. Chem. 2004, 47, 1739; (c) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.;
Beard, H. S.; Frye, L. L.; Pollard, W. T.; Banks, J. L.; Glide J. Med. Chem. 2004, 47,
1750.
3. (a) Pellicciari, R.; Costantino, G.; Fiorucci, S. J. Med. Chem. 2005, 48, 5383; (b)
Rizzo, G.; Renga, B.; Antonelli, E.; Passeri, D.; Pellicciari, R.; Fiorucci, S. Mol.
Pharmacol. 2005, 68, 551.
4. Sinal, C. J.; Tohkin, M.; Miyata, M.; Ward, J. M.; Lambert, G.; Gonzalez, F. J. Cell
2000, 102, 731.
5. Claudel, T.; Staels, B. t.; Kuipers, F. Arterioscler. Thromb. Vasc. Biol. 2005, 25,
2020.
6. Hirokane, H.; Nakahara, M.; Tachibana, S.; Shimizu, M.; Sato, R. J. Biol. Chem.
2004, 279, 45685.
7. Maloney, P. R.; Parks, D. J.; Haffner, C. D.; Fivush, A. M.; Chandra, G.; Plunket, K.
D.; Creech, K. L.; Moore, L. B.; Wilson, J. G.; Lewis, M. C.; Jones, S. A.; Willson, T.
M. J. Med. Chem. 2000, 43, 2971.
8. Nicolaou, K. C.; Evans, R. M.; Roecker, A. J.; Hughes, R.; Downes, M.; Pfefferkorn,
J. A. Org. Biomol. Chem. 2003, 1, 908.