Discovery and optimization of orally active cyclohexane-based prolylcarboxypeptidase (PrCP) inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.09.094

The synthesis, SAR, binding affinities and pharmacokinetic profiles are described for a series of cyclohexane-based prolylcarboxypeptidase (PrCP) inhibitors discovered by high throughput screening. Compounds show high levels of ex vivo target engagement in mouse plasma 20 h post oral dose.

Full text of DOI:10.1016/j.bmcl.2013.09.094

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6228–6233
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and optimization of orally active cyclohexane-based
prolylcarboxypeptidase (PrCP) inhibitors
a
a
a
a
John S. Debenham a, , Thomas H. Graham , Andreas Verras , Yong Zhang , Matthew J. Clements ,
Jeffrey T. Kuethe b, Christina Madsen-Duggan a, Wensheng Liu a, Urmi R. Bhatt c, Dunlu Chen c,
Qing Chen d, Margarita Garcia-Calvo c, Wayne M. Geissler c, Huaibing He d, Xiaohua Li d,
JeanMarie Lisnock e, Zhu Shen c, Xinchun Tong d, Elaine C. Tung d, Judyann Wiltsie e, Suoyu Xu d,
Jeffrey J. Hale a, Shirly Pinto c, Dong-Ming Shen a
⇑
a Department of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065-0900, USA
b Department of Process Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065-0900, USA
c Department of Metabolic Disorders, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065-0900, USA
d Department of Drug Metabolism, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065-0900, USA
e Department of In Vitro Sciences, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065-0900, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis, SAR, binding affinities and pharmacokinetic profiles are described for a series of cyclohex-
ane-based prolylcarboxypeptidase (PrCP) inhibitors discovered by high throughput screening. Com-
pounds show high levels of ex vivo target engagement in mouse plasma 20 h post oral dose.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 9 September 2013
Accepted 30 September 2013
Available online 8 October 2013
Keywords:
PrCP
Prolylcarboxypeptidase
Serine protease
Inhibitor
Obesity
Prolylcarboxypeptidase (PrCP) is a lysosomal carboxypeptidase.
This serine protease cleaves the amide bond between a C-terminal
amino acid and a proline residue (i.e., peptide-Pro-Xxx-OH), and
degrades angiotensins II and III,1 plasma prekallikrein2 and
in body weight. The KO mice showed a 1% decrease in body weight
indicating that some off-target activity was in play as well, but
these results were promising.7 Compound 1 had a low bioavailabil-
ity of 13% with a clearance of 23 mL/min/kg. Compound 2 was
designed to reduce structural complexity of 1 by the removal of
one of the chiral centers of the original lead. Potency was main-
tained, but at the expense of bioavailability moving from 13% for 1
to just 4% with 2. In order to evaluate it in a similar PrCP WT/KO
mouse weight loss model a subcutaneous constant infusion pump
was required. Dosing for 7 days at 12 mg/kg resulted in weight loss
that was equivalent between WT and KO animals again indicating
an off-target activity was in play.8 The third tool evaluated 3 was
designed to differentiate itself by removal of the dichlorobenzimi-
dazole substituted pyrrolidine component of the previous series.
Additionally, this compound had considerably more CNS exposure,
something that was postulated to be required for enhanced weight
loss effects past what was observed in the first study. Like 2, 3 had
inadequate oral exposure requiring it to be dosed with a continu-
ous infusion pump. At 30 mg/kg non-mechanism based weight
loss was again observed withsignificant skin relatedAEs at theinjec-
tion site due to compound precipitation around the point of
infusion.9a
a
-melanocyte stimulating hormone (a
-MSH).3 It is found in many
tissues including liver, kidney, pancreas, heart, brain, adipose,
hypothalamus and gut and extracellularly in plasma and urine.4
PrCP has been implicated in several biological functions including
cardiovascular,5 inflammation,6 and food intake regulation.3 The
previous lack of small molecule tools to interrogate PrCP function
in vivo has made it difficult to confirm these findings and also
assess its potential utility as a therapeutic target.
Our laboratories have recently disclosed several small molecule
tools to evaluate PrCP inhibition for the modulation of feeding
behavior and weight gain in mouse obesity models (Fig. 1).7–9
The first tool compound our laboratories identified was 1. When
evaluated in a 5 day PrCP wild-type (WT) knock-out (KO) mouse
weight loss study at 100 mg/kg, the WT showed a 4.9% decrease
⇑
Corresponding author. Tel.: +1 908 740 5497; fax: +1 908 740 3132.
E-mail address: john_debenham@merck.com (J.S. Debenham).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.09.094

J. S. Debenham et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6228–6233
6229
Cl
Cl
O
O
O
N
N
N
N
R2
R2
R2
R2
R1
R1
NH
HN
5
6
N
H
N
H
1
O
h,m PrCP IC50 1.0, 2.0 nM
h,m PrCP IC50 43, 27 nM
(SSR or RRS)
NH2
h,m PrCP
IC50 2.5,1.4 nM
(SSS or RRR)
Cl
Cl
2
N
N
N
h,m PrCP IC50 1.4, 0.8 nM
N
HN
O
O
N
N
N
N
R2
R2
R2
R1
R1
N
R2
N
O
7
8
H
H
Cl
H
N
N
h PrCP IC50 192 nM
(SSS or RRR)
h PrCPIC50 642 nM
(SSR or RRS)
3
h,m Pr
O
CP IC50 1.5, 5.2 nM
Figure 3. Stereochemical deconvolution of PrCP activity. R1 = 4-chlorobenzyl,
R2 = phenyl.
Figure 1. Early PrCP lead compounds; h = human, m = mouse.
Our objective evolved to find a structurally distinct scaffold that
could generate tools with improved oral exposure suitable for
in vivo biological evaluation in mice and that could then serve as
a platform for further lead optimization. A high throughput screen
(HTS) of the Merck sample collection was carried out. One of the
hits obtained was isomeric mixture 4 as trans cinnamide-derived
racemates and a mixture of diastereomers at the benzylamine sub-
stitution (Fig. 2). It showed 200 nM IC50 human (h) PrCP activity
despite being a mixture of stereoisomers. In order to determine
the biological properties of the most active isomer an intermediate
used in the preparation of 4 was resolved by chiral HPLC and all the
trans isomers were prepared (Fig. 3). The isomer bearing an axially
oriented benzylamino group 5 showed potent activity for human
(h) and mouse (m) PrCP activity at 2.5 and 1.4 nM IC50, respec-
tively, with the other isomers showing 10–300-fold less activity.
With the serendipitous 5 in hand, its off target activity was evalu-
ated to benchmark potential strengths and shortcomings of the
new scaffold. Compound 5 was shown not to be a significant inhib-
O
O
O
N
R3
R3
a-b
c
N
O
OH
O
OTMS
9
10
Bn
R3
O
R3
O
O
e-h
O
O
d
N
O
Bn
O
Bn
O
11
12
R3
j-l
O
13 R4=H
i
OR4
O
14 R4=Me
O
R3
R3
itor of the cytochrome p450 enzymes (IC50 in
CYP2C9>50, and CYP2D6 20. Human PXR activity was also minimal
with an EC50 of >15 M. The ion channel activity of 5 was less opti-
lM): CYP3A4 25,
O
m
R5
OH
l
18
mal. In the hERG potassium channel assay the activity was deter-
mined to be about 3 nM binding IC50. Not surprisingly, based on
the hydrophobic nature of 5 (LogP 8.04),10 the compound had
essentially unmeasurable free fraction in both human and mouse
plasma. Compound 5 was screened broadly against a customized
panel of radioligand binding and enzymatic assays to examine
other potential off target activites.11 Of note, it showed activity
at human CB1R of 345 nM IC50. CB1R activity is known to strongly
influence feeding behavior and body weight changes across a wide
variety of species including humans. As a tool compound to inter-
rogate biological activity including feeding behavior, this CB1
N
N
O
•HCl
O
15
16
R6
R6
n-o
N
Ar
NH2
+
MgBr
Ar
18
17
Scheme 1. Methods to prepare compounds of general structure 16. Reagents and
conditions: (a) oxalyl chloride, CH2Cl2, rt; (b) n-BuLi, (R)-4-benzyloxazolidin-2-one,
THF, À78 °C; (c) diethylaluminum chloride, CH2Cl2, 0 °C to rt, 40–50% steps a–c; (d)
2,2-dimethylpropane-1,3-diol, p-TsOH, toluene, reflux, THF; (e) LiOH, H2O2, H2O,
THF, 0 °C to rt; (f) HCl (aq), 60 °C; (g) dicyclohexylamine, MTBE, rt; (h) NaHSO4 (aq),
rt, steps d–h 81%; (i) (CH3)3SiCHN2, benzene:MeOH (7:1) P99%; (j) morpholine
(2 equiv), Ti(O-iPr)4 (2 equiv), CH2Cl2 (0.4 M), rt, then NaCNBH3 (8 equiv), MeOH,
HCl (8 equiv) À78 °C to rt, slowly (6–7 h), ꢀ89% isomer 16, ꢀ10% alternate isomer
(not shown); (k) LiOH, H2O, 70 °C, 24 h; (l) acetone, H2O, HCl, P99%; (m) HATU,
DIEA, DMA, amine, rt, 7–70%; (n) THF, 0 °C or rt; (o) MeOH or THF/MeOH, NaBH4,
25–80%.
O
N
activity would need to be minimized in order to allow a more
direct interpretation of observed in vivo activity.
In order to access structurally optimal diastereomers an asym-
metric synthesis of the 4-oxo-2-phenylcyclo-hexanecarboxylic
acid core 13 was devised employing the methods of Evans et al.12
h PrCP IC50 198 nM
(racemic trans)
N
H
4
Cl
Figure 2. PrCPi screening hit.

6230
J. S. Debenham et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6228–6233
Table 1
PrCP inhibition and binding affinities of compounds at human and mouse PrCP inhibition and human hERG (IKr) binding expressed as IC50 (nM)
R3
O
R5
19
N
O
a
Compound
R4
PrCP(h, m)b IC50, nM
hERG IC50, nM
R3 =H
20
29, 180
170
N
H
21
2.5, 12
1.2, 6.6
17, 170
530
290
NA
N
H
N
22 (CB1 = 8600)
23
N
N
H
24
25
440, NA
12, 320
NA
N
H
2200
F
N
H
26
480, NA
3.0, 44
NA
NA
F
N
H
27
Cl
a
R3 = Br unless otherwise indicated.
Values are based on one or two experiments, each in triplicate.
b
(Scheme 1). Starting from commercially available substituted phe-
nylprop-2-enoic acids and the (4R)-4-benzyl-1,3-oxazolidin-2-one
chiral auxiliary, oxazolidinone dienophile 10 was prepared, and
subsequently subjected to a Diels–Alder cycloaddition reaction to
form the core 4-oxo-2-phenylcyclohexanecarboxylate 11. In order
to remove the chiral auxiliary, compounds such as 11 are typically
treated with aqueous LiOH and H2O2. However, the presence of the
ketone was not compatible with those conditions, and required
protection as the cyclic acetal 12. Once protected, the auxiliary
on 12 was hydrolyzed. Subsequent deprotection of the acetal with
HCl afforded the free acid 13. It was found that this intermediate
was readily purified by formation of the dicyclohexylamine salt
in EtOAc. Following the salt break of purified 13 with NaHSO4,
the acid was protected as the methyl ester. Reductive amination
of the ketone was carried out with morpholine mediated by
Ti(O-i-Pr)4 at room temperature for 15 h in CH2Cl2 followed by
subsequent addition of NaCNBH3 in MeOH at À78 °C and then
methanolic HCl at À78 °C. The reaction mixture was gradually
warmed to rt over 7 h to afford the desired R amine isomer in a
9:1 ratio with the less active (S) isomer as the minor product.
When the reaction was warmed more quickly, over 3–4 h, for
example, the selectivity was observed to drop giving the preferred
isomer in a reduced 4–5:1 selectivity. The protection of 13 as the
methyl ester facilitated the isolation and purification of the inter-
mediate amino ester by silica gel flash chromatography in bulk.
Hydrolysis of the ester followed by activation of the acid for amide
coupling with HATU and subsequent reaction with the various
amines in the presence of DIEA afforded the final compounds of
general formula 16. When the amines employed in the amide cou-
pling reaction of Scheme 1 are not commercially available biaryl or
aryl-heteroaryl, the methods of Terrasson et al. were employed in
their construction.13 In the first step, a substituted aryl Grignard
reagent was reacted with substituted benzonitrile in THF. Once
the initial addition was complete the solvent was either exchanged
with MeOH or the reaction is diluted with MeOH so that the reduc-
tion of the intermediate imine can be completed with NaBH4 pro-
viding the amine that was the coupling partner to acid 15.
Human and mouse PrCP enzyme inhibition was determined
using a standard protocol that was previously described.7,8 In sin-
gle enantiomer cases the listed data are for the eutomers, which

J. S. Debenham et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6228–6233
6231
Table 2
Binding affinities of compounds at human and mouse PrCP and human hERG (IKr)
expressed as IC50 (nM)
Br
O
R5
N
28
O
Compound
R4
PrCP(h, m)a IC50, nM
hERG IC50, nM
Cl
F
29
2.6, 3.4
39
N
H
Figure 4. Fragment fractional polar surface area versus hERG IC50. Fractional PSA
for fragmented molecules. Biaryl (right portion) fragments are indicated in blue.
Substituted cyclohexyl (left portion) fragments are indicated in red. IC50 values are
calculated in an MK499 displacement assay.
Cl
30
0.66, 1.4
100
N
H
F
F
N
N
31
32
0.60, 3.0
8400
1900
N
H
0.18, 0.74
N
H
Cl
N
N
N
F
33
34
35
2.2, 10
1500
690
N
H
Cl
Cl
Cl
Cl
F
2.7, 8.6
N
H
Figure 5. Hydrophobicity map and overlay with sertindole. hERG residues are
indicated with a ribbon model colored by sequence position. Compound 29 is
represented with grey carbons and sertindole is represented with green carbons.
Hydrophobicity maps 2 Å around the ligand are indicated in light green.
0.079, 0.28
590
are typically 10–100-fold more potent than the corresponding dis-
tomers. Our initial goal was to evaluate structures of reduced lipo-
philicity. It was anticipated that as compound polarity increased
we would see free fraction move to measureable amounts, and also
potential decreased CB1 activity. The 4-chlorobenzyl domain of 6
was exchanged with several amines14 and it was found that
morpholine maintained potency, while decreasing MW and
lipophilicity. Table 1 shows a variety of analogs that were prepared
exploring reduction of complexity and molecular weight from lead
compound 5 which had a MW of 549 and a LogP of 8.04.15 Com-
pound 20 shows the move from the diphenyl azetidine amide to
the benzhydryl amide and removal of the ortho methyl group of
5. 20 lost about 12-fold activity for human PrCP and about 129-fold
for mouse activity. Installation of a para bromo group 21 on the
phenyl at the 2-position brought back the human activity of 5
and mouse activity was within ninefold of 5 at 12 nM. Moreover
hERG activity was attenuated 177-fold from 3 to 530 nM IC50. N-
methylation of the amide nitrogen of 21 was detrimental to both
mouse and human activity with 23 displaying 7 and 14-fold loss
of potency, respectively. Comparing diphenyl azetidine amide 22
N
H
N
N
N
36
0.95, 1.7
180
N
H
Cl
Cl
Cl
N
37
38
4.6, 14
1000
1200
N
H
N
36, 210
N
H
a
Values are based on one or two experiments, each in triplicate; Data are for the
more potent single enantiomer; absolute stereochemistry was not determined.

6232
J. S. Debenham et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6228–6233
(100 nM), it still had an improved hERG/PrCP activity ratio of
150:1 versus the non-selective 5 which was about 1:1.
As stated earlier, lead compound 5 possessed very potent hERG
activity. In order to evaluate compound features and properties
contributing to hERG ion channel binding, we calculated several
physical properties in Pipeline Pilot including polar surface area
(PSA) and fractional PSA (PLP v. 7.5.2, Accelrys). To explore the fea-
tures of our ligands effecting hERG affinity, compounds were split
into two fragments at the amide bond. Fragments were then neu-
tralized and hydrogens were added. This resulted in fragments that
contain the biaryl portion (right) and fragments that contain the
more polar substituted cyclohexyl portion (left). PSA was calcu-
lated for all fragments using the 2D method of Ertl implemented
in PLP.16 Figure 4 illustrates a clear dependence of hERG binding
on fractional PSA, particularly on the biaryl portion of the mole-
cules. Compounds with increased polarity proportional to their
size in the biaryl region of the molecule, display significantly less
hERG affinity. Small modifications resulting in an approximately
10% increase in polarity on the right side alone can attenuate hERG
binding affinity by greater than 100-fold. While there may be some
hERG binding dependency on the left side of the molecule, the
structural diversity in this region does not allow such conclusions.
To elucidate atomic features that drive hERG activity we also
generated a putative binding mode to hERG homology models. Ini-
tially we used an in house structure generated from a KcsA tem-
plate containing a high potency hERG inhibitor MK499.17 This
homology model is of the closed form of the channel, wherein
the pore channel is narrowed and thought to be in an inactive
state. Ligand docking was done with Glide18 using a grid prepared
with decreased Van der Waals radii of 0.8 to allow for an increase
in the number of poses; however, given the size of the active site
relative to our compounds we were unable to generate a satisfac-
tory binding mode. Furthermore, alignment of our compounds
with MK499 was poor due to discrepancies between aromatic
and cation features in the ligands hypothesized to be important
in binding (data not shown).
Figure 6. Predicted binding mode of 29. Close up of 29 and select residues within
4.5 Å of the ligand.
Br
F
h,m PrCP IC50 16.7, 1.2 nM
hERG IC50 780 nM
O
O
N
39
N
N
F
Figure 7. Non-morpholino analog 39.
Table 3
Pharmacokinetic profiles of select compounds
Compd
%F
t
(h)
AUCNpo
M h kg/mg)
PPBa
½
To overcome these issues we used coordinates for the hERG
channel homology model from the Åqvist group that was built
using a KvAP structural template (pdbid 2A0L).19 This structure
was based on the open form of the channel and features docked
poses for ligands that share significant 3D shape to our series.20
Compound 29 was docked into the hERG homology model using
Glide with default settings. We found a very similar binding mode
to the sertindole ligand (Fig. 5). Representation of the grid surface
using a hydrophobicity surface in Maestro indicated that the aryl
(
l
37
31
32
34
39
21
36
30
35
5.2
6.9
8.9
1.2
2.0
1.4
1.5
6.5
3.1
4.9
7.6
2.9
0.010
0.019
0.016
0.049
0.59
0.19
0.46
0.92
0.56
—
2.1
1.9
—
17
32
36
50
77
90
—
3.2
1.8
0.5
1.9
groups occupy
a lipophilic region consistent with the PSA
a
Plasma protein binding data reported as % unbound in 100% C57BL/6 mouse
dependent SAR. The biaryl groups lie in a region bordered by
hydrophobic and aromatic residues including F656 and F652
plasma.
with 5, the presence of the morpholine amine and bromo phenyl
group allow retention of activity and almost a 100-fold shift in
hERG activity to 290 nM IC50. Moreover the human CB1R activity
attenuates 25-fold from 345 nM IC50 to 8600 nM. Cumyl amide
24 showed a 176-fold loss of human PrCP activity relative to
screening lead 5. However removal of one of the cumyl methyl
groups and substitution with either para F or Cl returned most of
the human activity of lead 5 with the S amide enantiomers. Of note
was the R enantiomer 26 that showed a 40-fold drop in activity.
Also, of the active S enantiomers the mouse potency was shifted
14–26-fold less active than human versus 5 that had no such shift.
At this point it was clear the extra aryl group was providing en-
hanced mouse potency. Table 2 shows more elaborated benzhydryl
derivatives both with achiral and chiral amides. Parahalogenated
29 and 30 showed more parity between human and mouse with
the difluoro 30 showing this first significant improvement from
the original lead. While 30 has an unfavorable hERG activity
Table 4
Mouse ex vivo target engagement (TE) assay results
Compd p.o. dosea (mg/kg) [compd]plasma
(nM)
Ki (100%)b (nM) TEc (%)
30
36
39
35
35
30
30
30
30
10
1025
1837
2883
390
70.5
158
19.0
1.10
1.10
90.9
92.9
99.9
P100
98.2
163
a
Fasted lean male C57 mice (n = 3) were dosed by oral gavage, samples collected
at 20 h post-dose, citrate treated blood (1:3 blood/0.1 M disodium citrate) was
prepared for LC/MS analysis by protein precipitation with acetonitrile.
b
Calculated Ki in 100% mouse plasma. The inhibition of PrCP in the presence of
25% plasma was measured and extrapolated to 100% plasma using data from 0%
plasma in the primary assay. The assay used heat-inactivated plasma to minimize
the activity of non-PrCP proteases.
c
Inhibition was measured in 100% mouse plasma that was heat-inactivated to
minimize the protease activity.

J. S. Debenham et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6228–6233
6233
2. (a) Shariat-Madar, Z.; Mahdi, F.; Schmaier, A. H. Blood 2004, 103, 4554; (b)
Shariat-Madar, Z.; Mahdi, F.; Schmaier, A. H. J. Biol. Chem. 2002, 277, 17962; (c)
Moreira, C. R.; Schmaier, A. H.; da Mahdi, F.; Motta, G.; Nader, H. B.; Shariat-
Madar, Z. FEBS Lett. 2002, 523, 167.
3. (a) Jeong, J. K.; Diano, S. Trends Endocrin. Met. 2013, 24, 61; (b) Jeong, J. K.;
Szabo, G.; Raso, G. M.; Meli, R.; Diano, S. Am. J. Physiol. Endocrinol. Metab. 2012,
302, E1502; (c) Diano, S. Front. Neuroendocrin. 2011, 32, 70; (d) Wallingford, N.;
Perroud, B.; Gao, Q.; Coppola, A.; Gyengesi, E.; Liu, Z.-W.; Gao, X.-B.; Diament,
A.; Haus, K. A.; Shariat-Madar, Z.; Mahdi, F.; Wardlaw, S. L.; Schmaier, A. H.;
Warden, C. H.; Diano, S. J. Clin. Invest. 2009, 119, 2291; (e) Rabey, F. M.;
Gadepalli, R. S. V. S.; Diano, S.; Cheng, Q.; Tabrizian, T.; Gailani, D.; Rimoldi, J.
M.; Shariat-Madar, Z. Curr. Med. Chem. 2012, 19, 4194.
4. (a) Skidgel, R. A.; Erdos, E. G. Immunol. Rev. 1998, 161, 129; (b) Kumamoto, K.;
Stewart, T. A.; Johnson, A. R.; Erdos, E. G. J. Clin. Invest. 1981, 67, 210; (c) Yang,
H. Y. T.; Erdos, E. G.; Chiang, T. S. Nature 1968, 218, 1224; (d) Miller, J. J.;
Changaris, D. G.; Levy, R. S. Life Sci. 1991, 48, 1529.
5. (a) Adams, G. N.; LaRusch, G. A.; Stavrou, E.; Zhou, Y.; Nieman, M. T.; Jacobs, G.
H.; Cui, Y.; Lu, Y.; Jain, M. K.; Mahdi, F.; Shariat-Madar, Z.; Okada, Y.; D’Alecy, L.
G.; Schmaier, A. H. Blood 2011, 117, 3929; (b) Mallela, J.; Yang, J.; Shariat-
Madar, Z. Int. J. Biochem. Cell Biol. 2009, 41, 477; (c) Wang, L.; Feng, Y.; Zhang, Y.;
Zhou, H.; Jiang, S.; Niu, T.; Wei, J. L.; Xu, X.; Xu, X. P.; Wang, X. Am. J. Obstet.
Gynecol. 2006, 195, 162.
6. (a) Chajkowski, S. M.; Mallela, J.; Watson, D. E.; Wang, J.; McCurdy, C. R.;
Rimoldi, J. M.; Shariat-Madar, Z. Biochem. Biophys. Res. Commun. 2011, 405, 338;
(b) Ngo, M. L.; Mahdi, F.; Kolte, D.; Shariat-Madar, Z. J. Inflamm. 2009, 6, 3.
7. Zhou, C.; Garcia-Calvo, M.; Pinto, S.; Lombardo, M.; Feng, Z.; Bender, K.; Pryor,
K. D.; Bhatt, U. R.; Chabin, R. M.; Geissler, W. M.; Shen, Z.; Tong, X.; Zhang, Z.;
Wong, K. K.; Sinha Roy, R.; Chapman, K. T.; Yang, L.; Xiong, Y. J. Med. Chem.
2010, 53, 7251.
8. Shen, H. C.; Ding, F. X.; Zhou, C. Y.; Xiong, Y. S.; Verras, A.; Chabin, R. M.; Xu, S.
Y.; Tong, X. C.; Xie, D.; Lassman, M. E.; Bhatt, U. R.; Garcia-Calvo, M.; Geissler,
W.; Shen, Z.; Chen, D. L.; SinhaRoy, R.; Hale, J. J.; Tata, J. R.; Pinto, S.; Shen, D.-
M.; Colletti, S. L. Bioorg. Med. Chem. Lett. 2011, 21, 1299.
9. (a) Graham, T. H.; Shen, H. C.; Liu, W.; Xiong, Y.; Verras, A.; Bleasby, K.; Bhatt, U.
R.; Chabin, R. M.; Chen, D.; Chen, Q.; Garcia-Calvo, M.; Geissler, W. M.; He, H.;
Lassman, M. E.; Zhu, S.; Tong, X.; Tung, E.; Xie, D.; Xu, S.; Colletti, S. L.; Hale, J. J.;
Tata, J. R.; Pinto, S.; Shen, D.-S. Bioorg. Med. Chem. Lett. 2012, 22, 658; (b)
Graham, T. H.; Liu, W.; Verras, A.; Reibarkh, M.; Bleasby, K.; Bhatt, U. R.; Chen,
Q.; Garcia-Calvo, M.; Geissler, W. M.; Gorski, J.; He, H.; Lassman, M. E.; Lisnock,
J.; Li, Z.; Zhu, S.; Tong, X.; Tung, E.; Xie, D.; Xu, S.; Xiao, J.; Hale, J. J.; Pinto, S.;
Shen, D.-S. Bioorg. Med. Chem. Lett. 2012, 22, 2818.
(Fig. 6). L650 further contributes to the lipophilicity in this region.
One clear hydrogen bond is evidently made from S649 to the mor-
pholine oxygen. SAR supporting this interaction is evidenced by
the decreased potency of 39 which is unable to make this
contact.21
These observations that increased fractional polar surface area
directly related to decreased hERG affinity corresponded well with
our desire to reduce the hydrophobicity of lead compound 5.
Examining some of the more polar analogs in Table 2 showed that
the unsubstituted pyridine analogs 31 and 32 both showed a
marked decrease in hERG affinity to 8400 and 1900 nM, respec-
tively, while also maintaining or slightly increasing potency (32).
For 31 and 32 the hERG to hPrCP ratio has increased to over
10,000:1. Pyrimidine and pyrazine analogs had the most fractional
polar surface area, but also lost considerable PrCP activity.
The compound of greatest interested to us in Table 2 was 35,
which showed exceptional human and mouse potency of 0.079
and 0.28 nM IC50, respectively. While the 590 nM IC50 hERG activ-
ity would indicate the compound would have questionable value
for human development it would not present a problem as a tool
for the use in mouse since hERG does not play a major role in car-
diac repolarization in mice like it does in humans. Substituted pyr-
idine or pyridine isomer analogs 33, 34, and 36 showed good parity
between human and mouse activity of between 1 and 10 nM IC50
.
Compound 39 in Figure 7 shows the effect of moving the basic
amine further from the cyclohexane ring. While potency can be
maintained for mouse at 1 nM the parity between mouse and hu-
man decreases with 14-fold less activity at human PrCP.
Several compounds were selected for further profiling in mouse.
In terms of their pharmacokinetic profiles (Table 3), pyrazine 37,
and unsubstituted pyridines 31, 32, and 3-chloro substituted pyri-
dine 34 showed poor bioavailability at 5–17% with low oral expo-
sure. Considerable improvements in %F were noted as the
heterocycle was replaced with phenyl 39, 21, and 30 or 2-substi-
tuted pyridine 36. 2-Fluropyridine 35 had the highest bioavailabil-
ity at 90% with good oral exposure and 1.9% unbound fraction in
mouse plasma. A select group of these compounds were subse-
quently evaluated for ex vivo target engagement (TE) using meth-
ods described by our laboratories previously (Table 4).22 Here
compounds were dosed orally and plasma samples were collected
20 h post dosing to assess TE. A single 30 mg/kg oral dose was able
to achieve ꢀ90% TE in plasma using compounds 30 and 36. The im-
proved Ki values of 39 and 35 in 100% mouse plasma allowed them
to achieve complete (35 at 30 mg/kg) or near complete target
engagement (35 at 10 mg/kg or 39 at 30 mg/kg). Structurally dif-
ferentiated from our earlier leads 1–3, compounds 35 and 39 main-
tain high levels of TE and can be orally administered. Based on the
above results, these cyclohexane based PrCP inhibitors could be
used as tools to further elucidate the potential utility and biological
actions of PrCP inhibition in vivo.
10. Value calculated with ChemBioDraw version 12.
11. Discovery Pharmacology services provided by MDS Pharma Services, Taipei,
Taiwan.
12. (a) Evans, D. A.; Chapman, K. T.; Hung, D. T.; Kawaguchi, A. T. Angew. Chem., Int.
Ed. Engl. 1987, 26, 1184–1186; (b) Evans, D. A.; Chapman, K. T.; Bisaha, J. J. Am.
Chem. Soc. 1988, 110, 1238–1256.
13. Terrasson, V.; Marque, S.; Scarpacci, A.; Prim, D. Synthesis 2006, 11, 1858–1862.
14. This work is not presented herein.
15. Calculated with ChemBioDraw Ultra 12.
16. Ertl, P.; Rohde, B.; Selzer, P. J. Med. Chem. 2000, 43, 3714.
17. Mitcheson, J. S.; Chen, J.; Lin, M.; Culberson, J. C.; Sanguinetti, M. C. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 12329–12333.
18. Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.;
Repasky, M. P.; Knoll, E. H.; Shaw, D. E.; Shelley, M.; Perry, J. K.; Francis, P.;
Shenkin, P. S. J. Med. Chem. 2004, 47, 1739.
19. Stary, A.; Wacker, S. J.; Boukharta, L.; Zachariae, U.; Karimi-Nejad, Y.; Åqvist, J.;
Vriend, G.; de Groot, B. L. Chem. Med. Chem. 2010, 5, 455.
20. Boukharta, L.; Keränen, H.; Stary-Weinzinger, A.; Wallin, G.; Groot, B.; Åqvist, J.
Biochemistry 2011, 50, 6146.
21. For synthetic preparation see: Debenham, J.S. Shen, D.-M. Graham, T.H.
Clements, M.J. Zhang, Y. Int. Pat. Appl. WO 2011146300, 2011.
22. The target engagement assay was conducted by oral dosing of mice (n = 3) and
collection of tissue samples at 20 h post-dose. Plasma was heat-treated to
eliminate endogenous protease activity. Fresh recombinant PrCP was added
and the conversion of an exogenous substrate was monitored by fluorescence
in a similar manner to the in vitro primary assay. Additional details can be
found in: Graham, T. H.; Liu, W.; Verras, A.; Reibarkh, M.; Bleasby, K.; Bhatt, U.
R.; Chen, Q.; Garcia-Calvo, M.; Geissler, W. M.; Gorski, J.; He, H.; Lassman, M. E.;
Li, X.; Shen, Z.; Tong, X.; Tung, E. C.; Wiltsie, J.; Xie, D.; Xu, S.; Xiao, J.; Hale, J. J.;
Pinto, S.; Shen, D.-M. Bioorg. Med. Chem. Lett. 2012, 22, 2818.
References and notes
1. (a) Odya, C. E.; Marinkovic, D. V.; Hammon, K. J.; Stewart, T. A.; Erdos, E. G. J.
Biol. Chem. 1978, 253, 5927; (b) Yang, H. Y. T.; Erdos, E. G.; Chiang, T. S.; Jenssen,
T. A.; Rodgers, J. G. Biochem. Pharmacol. 1970, 19, 1201.