Synthesis and biological activity of 5-aryl-4-(4-(5-methyl-1H-imidazol-4-yl)piperidin-1-yl)pyrimidine analogs as potent, highly selective, and orally bioavailable NHE-1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.06.077

A series of potent inhibitors of the sodium hydrogen exchanger-1 (NHE-1) is described. Structure-activity relationships identified the 3-methyl-4-fluoro analog 9t as a highly potent (IC₅₀ = 0.0065 μM) and selective (NHE-2/NHE-1 = 1400) non-acylguanidine NHE-1 inhibitor. Pharmacokinetic studies showed that compound 9t has an oral bioavailability of 52% and a plasma half life of 1.5 h in rats. Because of its promising potency, selectivity, and a good pharmacokinetic profile, compound 9t was selected for further studies.

Full text of DOI:10.1016/j.bmcl.2006.06.077

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4796–4799
Synthesis and biological activity of 5-aryl-4-(4-(5-methyl-1H-
imidazol-4-yl)piperidin-1-yl)pyrimidine analogs as potent,
highly selective, and orally bioavailable NHE-1 inhibitors
a
a,
a
Karnail S. Atwal,a,* Steven V. O’Neil, Saleem Ahmad, Lidia Doweyko,
*
Mark Kirby,b Charles R. Dorso,b Gamini Chandrasena,c Bang-Chi Chen,a
Rulin Zhaoa and Robert Zahlera
aDepartment of Discovery Chemistry, Bristol-Myers Squibb, Pharmaceutical Research Institute,
PO Box 4000, Princeton, NJ 08543-5400, USA
bDepartment of Cardiovascular Biology, Bristol-Myers Squibb, Pharmaceutical Research Institute,
PO Box 4000, Princeton, NJ 08543-5400, USA
cDepartment of Drug Metabolism, Bristol-Myers Squibb, Pharmaceutical Research Institute,
PO Box 4000, Princeton, NJ 08543-5400, USA
Received 13 April 2006; revised 22 June 2006; accepted 23 June 2006
Available online 25 July 2006
Abstract—A series of potent inhibitors of the sodium hydrogen exchanger-1 (NHE-1) is described. Structure–activity relationships
identified the 3-methyl-4-fluoro analog 9t as a highly potent (IC50 = 0.0065 lM) and selective (NHE-2/NHE-1 = 1400) non-acylgua-
nidine NHE-1 inhibitor. Pharmacokinetic studies showed that compound 9t has an oral bioavailability of 52% and a plasma half life
of 1.5 h in rats. Because of its promising potency, selectivity, and a good pharmacokinetic profile, compound 9t was selected for
further studies.
Ó 2006 Elsevier Ltd. All rights reserved.
Sodium hydrogen exchanger-1 (NHE-1) has attracted
considerable attention over the past several years.1 This
is largely because NHE-1 inhibitors have the potential
to treat myocardial ischemia, a leading cause of death
in the western world.2 Several NHE-1 inhibitors have
been evaluated preclinically and clinically for indications
such as myocardial infarction, ischemic heart disease,
and angina. The results of a phase II/III clinical trial
of cariporide (1) in patients with acute coronary syn-
drome were less than encouraging.3 The lack of positive
clinical data for cariporide (1) may be due to the design
of the trial or the modest potency (IC50 = 3.5 lM) of the
compound; however, the results from clinical trials on a
follow-up compound, eniporide (2, IC50 = 0.4 lM), have
also been disappointing.4
acylguanidine group has the potential to undergo cleav-
age under metabolic conditions, releasing guanidine, a
known cause of toxicity.
In our efforts to discover more potent, efficacious, and
safe NHE-1 inhibitors, we sought to investigate com-
pounds with an acylguanidine surrogate group.5 Several
imidazole containing NHE-1 inhibitors, represented by
3, had previously been reported in the literature.6 Eval-
uation of these compounds showed only modest potency
O
O
NH2
NH2 MeO2S
MeO2S
N
N
NH2
NH2
Me
N
Me
Me
Eniporide, 2
Cariporide, 1
The salient feature of most known NHE-1 inhibitors is
an acylguanidine functionality. We believed that the
Ar
H
N
H
R2
N
N
N
N
N
N
N
R1
Me
NH
Me
Keywords: NHE; NHE-1; Sodium hydrogen exchanger.
*
Corresponding authors. Tel.: +1 609 252 6955; fax: +1 609 252
6804; e-mail: saleem.ahmad@bms.com
Me
O
4
3
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.077

K. S. Atwal et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4796–4799
4797
Me
Me
Me
a
a, b
NH
NH
NH
9o
Cbz
N
O
HN
Br
+
N
N
OMe
N
N
N
5
11
Br
Me
Me
c
b
c
NH
NH
5
N
Cl
+
N
Cl
N
9q
N
NH2
N
N
N
Cl
6
9p
7
Me
Br
e
Scheme 2. Reagents and conditions: (a) 4-methoxybenzenediazonium
tetrafluoroborate; (b) H2/PtO2/ethanol; (c) Ac2O/Py.
NH
d
N
N
N
N
F
OMe
Me
8
Me
ly the morpholine derivative 9c was prepared from 9d by
heating it in neat morpholine at 80 °C. Compound 9d
could be prepared via condensation of 4,6-dichloro-5-
phenylpyrimidine (12) with piperidinylimidazole 5 in
diglyme in the presence of potassium carbonate at
150 °C for 30 min (Scheme 3). Compounds 9e–o were
prepared via Suzuki coupling of the corresponding chlo-
ropyrimidinylimidazole 13 (prepared via coupling of
compound 5 with the corresponding dichloropyrimi-
dine) with an appropriate arylboronic acid (Scheme 3).
Methylation of the imidazole 9o with potassium carbon-
ate and methyl iodide in acetone gave predominantly the
N3-methylated product 9r. Compounds 9s–v were syn-
thesized by displacement of the chloride of 7 by the
appropriate nucleophile (MeOÀ, morpholine, and
iPrOÀ) prior to Suzuki coupling (Scheme 3).
NH
N
N
N
N
OMe
9t
Scheme 1. Reagents and conditions: (a) KOtBu/neat/140 °C; (b) H2/
Pd–C/EtOH; (c) K2CO3/acetonitrile; (d) NaOMe/MeOH; (e) 4-fluoro-
3-methylphenylboronic acid/Pd(Ph3P)4/Na2CO3/DMF/80 °C.
for NHE-1 (3, IC50 = 0.25 lM). Using piperidinylimi-
dazole as a core template, we systematically introduced
a variety of groups at the 1-position of piperidine ring as
well as the imidazole group. This paper describes our ef-
forts to optimize piperidinylimidazoles with arylpyrimi-
dine substituents (4).
Synthesis of the nitrile 9w and the amide 9x was carried
out as outlined in Scheme 4. The nitrile 14 was prepared
by heating the chloride 7 with tert-butylcyanoacetate in
THF followed by decarboxylation (CF3COOH, 60 °C).
The bromide 14 was converted to 9w via Suzuki cou-
pling in 80% yield. Hydrolysis of the nitrile 9w (2%
NaOH and MeOH) and coupling of the resulting acid
with isopropylamine provided the amide 9x (78%).
Synthesis of the various analogs described herein started
with the readily available piperidinylimidazole core 5 as
outlined in Scheme 1. Thus treatment of N-cbz-piperi-
done with excess 4-methylimidazole (5 equiv) in the
presence of potassium tert-butoxide at 140 °C (melt)
afforded the corresponding 4-hydroxypiperidine deriva-
tive in 40% yield after silica gel chromatography. This
material could be readily converted to compound 5 (iso-
lated as HCl salt) via catalytic hydrogenation using 10%
palladium on carbon. Condensation of 5 with the tri-
halopyrimidine 6 provided compound 7 in 80% yield
(Scheme 1). The chlorine in 7 was displaced with a meth-
oxy group (NaOMe/MeOH) to provide the 2-methoxy
analog 8 in 81% yield. The pyrimidinyl bromide 8 was
then converted to the biaryl derivative 9t via Suzuki
reaction in 94% yield. Thus, all three halides in 6 were
conveniently displaced in a stepwise fashion to provide
the requisite compound.
The NHE-1 and NHE-2 activities were assessed as de-
scribed.7 These experiments were carried out in AP1 cell
line expressing human NHE isoforms. This cell line has
no endogenous NHE activity. The IC50 values were
determined by measuring the ability of the compounds
to inhibit 50% of the sodium dependent recovery of
pH following imposed acidosis. Using this protocol,
Ph
Cl
N
Cl
a
+
9d
5
N
Synthesis of the aminoimidazole 9p is summarized
in Scheme 2. The imidazole analog 9o (see Scheme 3
for synthesis of 9o) was treated with 4-meth-
oxybenzenediazonium tetrafluoroborate to give the
diazo compound 11, which upon hydrogenation over
platinum oxide provided the desired product 9p in
62% overall yield. Acetylation of 9p with acetic anhy-
dride in pyridine gave the N-acetyl compound 9q.
12
Cl
Me
b
NH
9e - 9o
N
N
N
N
13
c, b
9s - 9v
7
The ethoxy analog 9b was prepared from the corre-
sponding chloro analog 9d by treatment with sodium
ethoxide in ethanol at room temperature (80%). Similar-
Scheme 3. Reagents and conditions: (a) K2CO3/diglyme/150 °C; (b)
arylboronic acids/Pd(Ph3P)4/Na2CO3/DMF/80 °C; (c) NaOMe or
NaOiPr or morpholine.

4798
K. S. Atwal et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4796–4799
Replacing a pyrimidine proton in 9a with an ethoxy
Me
Br
group (compound 9b) did not have an effect on the
potency. However, significant activity was lost when
the ethoxy group in 9b was replaced by a morpholino
group (9c). Compound 9b, which contains electron
releasing ethoxy group, has similar potency to the ana-
log with an electron withdrawing chloro substituent
(9d). The potency appears to be sensitive to the size of
the R1 group, as shown by comparison of the morpho-
lino analog 9c with compounds having relatively smaller
substituents at this position (9a, 9b, and 9d).
a, b
NH
c
N
N
7
N
N
NC
14
F
Me
d, e
Me
9x
NH
N
N
N
N
NC
9w
While introduction of a chlorine atom at the 4-position
of the pendant phenyl ring (compound 9e) had no effect
on potency, introduction of a 3-chloro on the phenyl
ring (9f, IC50 = 0.061 lM) resulted in a nearly 6-fold
improvement in potency. The 2-chloro analog (9g,
IC50 = 0.061 lM) showed potency similar to the 3-chlo-
ro compound 9f. The 3- and 2-methoxy analogs (9h and
9i) showed no significant improvement in potency rela-
tive to the starting compound 9a. The lack of potency
of the methylenedioxy analog 9j may reflect a sterically
restricted binding site. While the compounds with
3-methyl (9m, IC50 = 0.027 lM), and 2-chloro groups
Scheme 4. Reagents and conditions: (a) tert-butyl cyanoacetate/NaH/
THF/heat/2 days; (b) TFA/60 °C; (c) 4-fluoro-3-methylphenylboronic
acid/Pd(Ph3P)4/Na2CO3/DMF/80 °C; (d) NaOH/MeOH; (e) isopro-
pylamine/carbonyldiimidazole/triethylamine.
the NHE-1 IC50 of cariporide (1) and eniporide (2) were
measured as 3.5 and 0.4 lM, respectively. The IC50 of
the imidazole analog 3 was 0.25 lM, while that of the
baseline compound 9a was 0.52 lM. Starting from 9a,
we optimized the SAR by elaborating the aryl, the
pyrimidine, and the imidazole rings.
Table 1. NHE-1 inhibitory activity of the various imidazole derived analogs
R6
R5
5
R3
R1
6
R4
N
4
1
N
N
N
N
3
2
R2
Me
R1
R2
R3
R4
R5
R6
IC50 (lM)
a
Compound
9a
9b
9c
9d
9e
9f
9g
9h
9i
H
OEt
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
—
—
—
H
H
H
0.52
0.82
H
H
H
H
1-Morpholino
Cl
H
H
H
H
8.6
0.58
H
H
H
H
H
H
H
H
4-Cl
3-Cl
2-Cl
3-OMe
2-OMe
(3,4)-OCH2O–
2-Me
4-Me
3-Me
4-F
0.35
H
H
H
H
0.061
0.061
0.35
H
H
H
H
H
H
H
H
H
H
H
H
0.94
9j
H
H
H
H
45.5% at 1 lM
0.25
1.0
9k
9l
H
H
H
H
H
H
H
H
9m
9n
9o
9p
9q
9r
9s
9t
9u
9v
9w
9x
1
H
H
H
H
0.027
0.021
0.041
0.0054
1.9
H
H
H
3-Me
3-Cl
3-Cl
3-Cl
3-Cl
3-Cl
3-Me
3-Me
3-Me
3-Me
3-Me
—
H
H
H
4-F
4-F
H
H
NH2
NH-Ac
H
H
H
4-F
4-F
H
H
11
H
–OMe
–OMe
1-Morpholine
–OiPr
H
4-F
4-F
0.015
0.0065
0.0079
0.0074
0.0046
0.087
3.5
H
H
H
H
4-F
4-F
H
H
H
–CH2CN
CH2C(O)NHiPr
H
4-F
4-F
H
H
Cariporide
Eniporide
—
—
—
—
—
—
—
—
—
2
—
—
0.40
0.25
3
—
a Each value is an average of at least two determinations.

K. S. Atwal et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4796–4799
4799
Table 2. NHE-2 activities of compound 9t, cariporide (1), and
eniporide (2)
oride(1).Ithasagoodoralbioavailability(52%)andmod-
est plasma half life (1.5 h). Compound 9t was thus selected
for further studies.
Compound
NHE2 (IC50, lM)
NHE2/NHE1
9t
1
9
62
17
1400
18
2
43
Acknowledgments
We greatly appreciate the support of Bristol-Myers
Squibb Department of Analytical Sciences and Depart-
ment of Chemical Synthesis for continuous support of
this work.
(9g, IC50 = 0.061 lM) showed good potency, com-
pounds with 2- and 4-methyl groups (9k and 9l, respec-
tively) were significantly less active. The presence of a 4-
fluoro group adjacent to the 3-methyl (9n) or 3-chloro
(9o) groups had no significant impact on potency.
Introduction of an amino group at the 2-position (R4) of
the imidazole of 9o gave 9p, possessing 8-fold greater
potency over the parent. Acetylation of the amino group
eroded potency (9q, IC50 = 1.9 lM), as did N-methyla-
tion (R3 = Me) of the imidazole nitrogen (9r,
IC50 = 11 lM).
References and notes
1. Karmazyn, M.; Gan, X. T.; Humphreys, R. A.; Yoshida,
H.; Kusumoto, K. Circ. Res. 1999, 85, 777.
2. Aronson, P. S. Annu. Rev. Physiol. 1985, 47, 545.
3. (a) Scholz, W.; Jessel, A.; Albus, U. J. Thromb. Throm-
bolysis 1999, 8, 61; (b) Karmazyn, M. Expert Opin.
Investig. Drugs 2000, 9, 1099.
4. (a) Baumgarth, M.; Beier, N.; Gericke, R. J. Med. Chem.
1997, 40, 2017; (b) Zeymer, U.; Suryapranata, H.; Mon-
assier, J. P.; Opolski, G.; Davies, J.; Rasmanis, G.;
Linssen, G.; Tebbe, U.; Schroder, R.; Tiemann, R.;
Machnig, T.; Neuhaus, K. L. J. Am. Coll. Cardiol. 2001,
38, 1644.
5. Ahmad, S.; Ngu, K.; Combs, D. W.; Wu, S. C.; Weinstein,
D. S.; Liu, W.; Chen, B. C.; Chandrasena, G.; Dorso, C.
R.; Kirby, M.; Atwal, K. S. Bioorg. Med. Chem. Lett.
2004, 14, 177.
6. (a) Cremer, G.; Hoornaert, C. WO 99/01435; (b)
Lorrain, J.; Briand, V.; Favennec, E.; Duval, N.;
Grosset, A.; Janiak, P.; Hoornaert, C.; Cremer, G.;
Latham, C.; O’Connor, S. E. Br. J. Pharmacol. 2000,
131, 1188.
7. Ahmad, S.; Doweyko, L. M.; Dugar, S.; Grazier, N.; Ngu,
K.; Wu, S. C.; Yost, K. J.; Chen, B.-C.; Gougoutas, J. Z.;
DiMarco, J. D.; Lan, S.-J.; Gavin, B. J.; Chen, A. Y.;
Dorso, C. R.; Serafino, R.; Kirby, M.; Atwal, K. S. J.
Med. Chem. 2001, 44, 3302.
Further modifications were made at the 2-position (R2) of
the pyrimidine ring. Introduction of a 2-methoxy group to
compound 9o improved NHE-1 inhibitory activity by 2-
fold (compound 9s, IC50 = 0.015 lM). Similarly, intro-
duction of a methoxy group to the 3-methyl-4-fluorophe-
nyl analog 9n resulted in ca. 3-fold enhancement in
potency (9t, IC50 = 0.0065 lM). Introduction of other
substituents to the 2-position of the pyrimidine, such as
morpholino (9u, IC50 = 0.0079 lM), O-isopropoxy (9v,
IC50 = 0.0074 lM), and cyanomethyl (9w, IC50
=
0.0046 lM), retained nanomolar potency (see Table 1).
Most compounds described herein showed good selec-
tivity for NHE-1 over NHE-2.8 Table 2 displays NHE-
2 activities for 9t, cariporide, and eniporide. The IC50
for inhibition of NHE-2 activity was 9 lM for 9t. Thus,
compound 9t is one of the most selective (1400-fold)
NHE-1 inhibitors known and, accordingly, may possess
a significantly improved safety profile.9
8. Wang, Z.; Orlowski, J.; Schull, G. E. J. Biol. Chem. 1993,
268, 11925.
To further differentiate between these highly potent
NHE-1 inhibitors, we determined their oral bioavailabil-
ities in rats. Compounds with IC50 values <0.020 lM
were initially evaluated in a coarse rat PK study. Based
on the data from the coarse PK studies, compound 9t
was selected for further evaluation in a full rat PK
study.10 After a single oral or intraarterial dose,
compound 9t had an oral bioavailability of 52% and a
plasma half life of 1.5 h in rats.
9. Schultheis, P. J.; Clarke, L. L.; Meneton, P.; Harline, M.;
Boivin, G. P.; Stemmermann, G.; Duffy, J. J.; Doetsch-
man, T.; Miller, M. L.; Shull, G. E. J. Clin. Invest. 1998,
101, 1243.
10. The test compounds were administered individually to
male rats as a solution in PEG–EtOH–H2O (1:1:1). Rats
were dosed orally (n = 3 and 10 lmol/kg) or intraarterially
(n = 3 and 5 lmol/kg) for full PK studies. Serial samples
were withdrawn at 0, 5, 10, 20, and 40 min (ia) and at 1, 2,
4, 6, 8, 12, and 24 h (po) after dosing. Plasma was
prepared from each sample by centrifugation and analyzed
by valid LC/MS/MS procedure on a reverse-phase C18
column using 0.02 M ammonium acetate, pH 5.1/acetoni-
trile (65:5, v/v) as a mobile phase. Pharmacokinetic
parameters were calculated by linear regression to deter-
mine the equation of the biexponential curve which best fit
the data.
In summary, we have identified compound 9t, with excel-
lent NHE-1 inhibitory activity (IC50 = 0.0065 lM) and
significantly greater selectivity for NHE-1 over NHE-2
(1400-fold) than either cariporide (1) or eniporide (2). In
addition, 9t is 60-fold more potent against NHE-1 than
eniporide (2) and nearly 500-fold more potent than carip-