Discovery and in vitro/in vivo studies of tetrazole derivatives as Kv1.5 blockers

Article abstract of DOI:10.1016/j.bmcl.2006.09.021

A novel class of tetrazole-derived Kv1.5 blockers is disclosed. In in vitro studies, several compounds had IC₅₀s ranging from 180 to 550 nM. In vivo studies indicated that compounds 2f and 2j increased right atrial ERP about 40% without affecting ventricular ERP.

Full text of DOI:10.1016/j.bmcl.2006.09.021

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6213–6218
Discovery and in vitro/in vivo studies of tetrazole derivatives
as Kv1.5 blockers
Shengde Wu,* Andrew Fluxe, Jim Sheffer, John M. Janusz, Benjamin E. Blass, Ron White,
Chris Jackson, Richard Hedges, Michael Murawsky, Bin Fang, Gina M. Fadayel,
Michelle Hare and Laurent Djandjighian
Procter & Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason-Montgomery Road, Mason, OH 45040, USA
Received 11 August 2006; revised 7 September 2006; accepted 8 September 2006
Available online 28 September 2006
Abstract—A novel class of tetrazole-derived Kv1.5 blockers is disclosed. In in vitro studies, several compounds had IC50s ranging
from 180 to 550 nM. In vivo studies indicated that compounds 2f and 2j increased right atrial ERP about 40% without affecting
ventricular ERP.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Atrial fibrillation (AF) is the most common chronic
arrhythmia1 and increases the risk of stroke and overall
mortality. Treatment of this ailment remains problemat-
ic.2 Most current drug therapies target the hERG potas-
sium ion channel which is present in both the atria and
the ventricles. Blockade of hERG leads to QT prolonga-
tion and increases the incidence of the serious ventricu-
lar arrhythmia, torsade de pointes.
herein replacement of the thiazolidinone scaffold with
a tetrazole scaffold 2 (Fig. 1).
The synthesis of these compounds begins with conver-
sion of appropriate acid derivatives 3 to the correspond-
ing acid chlorides 4. These acid chlorides 4 are then
treated in situ with aryl ethyl amines to generate the
amides 5. The amides 5 are then reacted with PCl5 and
TMSN3 to provide desired products 2(a–n) in good yield
(Scheme 1).
One strategy for the development of safe, effective atrial
antiarrhythmic drugs involves blockade of repolarizing
ion channels that are found predominantly or only in
the atria. The Kv1.5 potassium channel is an atrial-se-
lective ion channel which underlies the ultra-rapid de-
layed rectifier K+ current, IKur. This current is a
major repolarizing current in human atria and is not
found in human ventricles.3 Thus, Kv1.5 is an attractive
molecular target for treatment of atrial fibrillation or
atrial flutter.4 Significant efforts have been made to iden-
tify novel blockers of Kv1.5.5–19 Icagen described the
thiazolidinone derivatives 1 as potent and selective
Kv1.5 blockers.11 In a preceding publication, we report-
ed that the ring sulfur was rapidly metabolized by oxida-
tion. To overcome this metabolic liability, we describe
A variety of analogs were prepared (Table 1). Three ana-
logs, 2a–2c, with n = 0 were prepared and evaluated
in vitro. While both 2a and 2b showed modest block
of Kv1.5 at 1 lM, 2c had greater activity and an IC50
value of 549 nM.20 Since the planar tetrazole ring re-
stricts the orientation of the ‘lower’ aromatic ring, we
prepared tetrazoles 2d and 2e (n = 1, m = 0) with a
one-carbon spacer to allow additional conformational
O
N
N
N
N
N
S
R1
(CH2)
n
R1
m
R2
R2
Keywords: Tetrazoles; Kv1.5 blocker; Atrial antiarrhythmic; In vivo
efficacy.
1
2
*
Corresponding author. Tel.: +1 607 335 2340; fax: +1 607 335
2010; e-mail: wu.s.3@pg.com
Figure 1. Comparison of thiazolidinone and tetrazole derivatives.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.09.021

6214
S. Wu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6213–6218
H
HO
O
Cl
O
n
O
n
N
b
a
R1
(CH2)
(CH2)
n
m
m
(CH2)
m
R2
R2
R2
3
4
5
N N
N
N
c, d
R1
n=0, 1
m=0,2,3,4,5
(CH2)
n
m
R2
2 (a-n)
Scheme 1. Reagent and conditions: (a) SOCl2, rt, 1 h; (b) aryl ethyl amines, THF, rt, 4 h; (c) PCl5, CH2Cl2, À5 ꢁC, 2 h; (d) TMSN3, CH2Cl2, rt,
overnight.
Table 1. In vitro inhibitory activity of tetrazoles against the Kv1.5 channel
Compound
Structure
% Block of Kv1.5 at 1lM
Kv1.5 IC50 (lM)
N
N
N
N
2a
36
MeO
Me
Me
N
N
N
N
37
MeO
2b
tBu
N
N
N
N
2c
72
0.55
MeO
Cl
Cl
N
N
N
N
2d
75
0.33
Me
Me
MeO
N
N
N
N
2e
60
MeO
tBu

S. Wu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6213–6218
6215
Table 1 (continued)
Compound
Structure
% Block of Kv1.5 at 1lM
Kv1.5 IC50 (lM)
N
N
N
N
2f
90
20
89
96
94
72
81
0.33
MeO
Me
N
N
N
N
2g
2h
2i
MeO
MeO
MeO
MeO
Cl
N
N
N
N
0.40
Me
Me
N
N
N
N
N
N
N
N
2j
0.18
Me
Me
N
N
N
N
2k
0.48
2
Me
N
N
N
N
21
2
Cl
N
N
N
N
2m
74
MeO
Cl
N
N
N
N
2n
47
F
Cl

6216
S. Wu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6213–6218
Table 2. In vivo anesthetized pig results for compounds 2c, 2d, 2f, and 2j
Compound
AERP prolongation
VERP prolongation
10 mg/kg 30 mg/kg
dp/dtmax change
BP change
10 mg/kg
30 mg/kg
10 mg/kg
30 mg/kg
10 mg/kg
30 mg/kg
2c
2d
2f
2j
6
3
3%
7%
16 7%
23 10%
44 5%
37 7%
6
0
5
9
7%
3%
4%
4%
6
0
3
4
7%
5%
3%
4%
À24 1%
À26 5%
À1 7%
À10 l%
À49 4%
À32 18%
À14 14%
À12 15%
À6 5%
À11 13%
À5 3%
À42 4%
À21 10%
À4 1%
19 4%
17 2%
16 5%
7
4%
freedom in the ‘lower’ aromatic ring. Both compounds
are better Kv1.5 blockers compared to 2a and 2b, and
compound 2d exhibits a good potency with an IC50 val-
ue of 331 nM.
RAERP
200
175
150
125
100
44%
4
RAERP 10
RAERP 30
n=3, SEM
Following this finding, we then installed a cyclopropyl
ring on the linker to the ‘lower’ aromatic ring to provide
compounds 2f, 2j, and 2m bearing the 4-methoxyphen-
ethyl group in common. The in vitro results demonstrat-
ed that installation of this cyclopropyl ring increased the
% block of Kv1.5. Compound 2d has an IC50 value of
331 nM while the cyclopropyl ring-derived analog 2j is
about twice as potent with an IC50 value of 177 nM.
Additionally, the substituents on the ‘lower’ phenyl ring
have an impact on the activity of block Kv1.5 in these
cyclopropyl ring-derived analogs. The 3,4-di-Me analog
2j is more active than the 4-Me analog 2f and the 4-Cl
analog 2m.
17% 2
100
150
200
250
300
Heart Rate (bpm)
200
175
150
125
100
VERP
VERP 10
VERP 30
3% 3
n=3, SEM
Further ring size variations indicated that analogs 2f, 2h
and 2i with three-, five-, and six-membered ring were ac-
tive and potent Kv1.5 blockers. In contrast, analog 2g
with a four-membered ring showed significantly reduced
blockade of Kv1.5 versus 2m. These results demonstrate
that activity is sensitive to the size of the cyclic rings.
Furthermore, it is clear that compound 2n with the elec-
tron-withdrawing group on the ‘upper’ phenyl ring
(R1 = F) is significantly less active against Kv1.5 (47%)
compared to compound 2m (74%) with electron-donat-
ing group (R1 = OMe). Finally, compounds 2k and 2l
show that the linker to the phenethyl group can be
extended and still retain good Kv1.5 blockade.
100
150
200
250
300
Heart Rate (bpm)
Figure 2. Anesthetized pig AERP and VERP changes of 2f at different
heart rates.
Kv1.5 block over hERG22 (48-fold) and L-type calci-
um23 (69-fold) channels consistent with atrial selectivity
observed in vivo.
Compounds 2c, 2d, 2f, and 2j were selected for in vivo
evaluation in the anesthetized mini-pig (n = 3) which
was used to investigate the effects on the atrial and ven-
tricular effective refractory period (AERP and VERP).21
Compounds were administered via 15 min iv infusions
of 10 and 30 mg/kg doses. As shown in Table 2, com-
pounds 2c and 2d provided small increases in AERP
at a 10 mg/kg dose but larger increases at a 30 mg/kg
dose. No changes in VERP were noted at either dose.
However, reductions in contractility (dp/dtmax) and/or
blood pressure were noted at both doses. Fortunately,
improvements in both AERP prolongation and hemo-
dynamic side effects were obtained with the addition of
the cyclopropyl ring on the linker to the ‘lower’ aromatic
ring. Compounds 2f and 2j showed dose-dependent
increases of AERP and no effect on VERP. Compound
2f had the best overall profile showing minimal effects
on blood pressure or contractility. Figure 2 shows a typ-
ical dose–response curve for AERP prolongation. Addi-
tionally, compound 2f also showed good selectivity for
In conclusion, we have discovered a novel class of tetra-
zole derivatives that are potent blockers of the Kv1.5
channel. Analogs 2f and 2j, which contained the cyclo-
propyl ring on the linker to the ‘lower’ aromatic ring,
displayed very good in vitro blockade of Kv1.5 as well
as atrial-selective prolongation of ERP in vivo. The
compound 2f also lacked hemodynamic side effects.
The atrial-selective ERP prolongation in pigs is consis-
tent with the atrial-selective presence of IKur in hu-
man.24 These findings support the potential for Kv1.5
channel blockers to provide atrial-selective antiarrhyth-
mic drugs to treat AF.
Acknowledgments
We gratefully acknowledge Professor Dr. David
Williams at Indiana University for the preparation of

S. Wu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6213–6218
6217
fused onto the cells and the cells are pulsed every 5 s until
no further changes in current are evident at a given
compound concentration. Inhibition was measured at the
end of the 1 s pulses and expressed relative to controls.
Concentration–response curves are generated for appro-
priate compounds utilizing at least four concentrations
and an n = 3. Curve fitting and IC50 estimating were done
using Graphpad software (Ver. 4).
several key intermediates. We would also like to thank
Dr. Jeffrey Ares for support.
References and notes
1. Kannel, W. B.; Wolf, P. A.; Benjamin, E. J.; Levy, D. Am.
J. Cardiol. 1998, 82, 2N.
21. In vivo assay: Vehicle: Compounds are dissolved to a final
concentration of 20–50 mg/ml, first in dimethyl acetamide
(DMAC) then adding the balance of propylene glycol 200
(PEG200) for a ratio of 20% DMAC/80% PEG200. Mini
Swine: Minipigs of the Hanford or Sinclair strain weighing
15–30 kg are anesthetized with an IM injection of
ketamine/xylazine followed, if needed, by 1–1.5% isoflu-
rane to allow introduction of a venous catheter into the
vena cava in the neck. Following incubation, IV pento-
barbital is given and anesthesia is maintained via further
IV boluses given during the study. Two electrode-tipped
catheters are introduced via the jugular, one into the right
atrium and the other into the right ventricle, and are used
to pace the heart when needed. The carotid artery is
cannulated and a pressure transducer-tipped catheter
advanced into the left ventricle to track pressure develop-
ment in the LV. An incision in the groin is used to access
the femoral artery and vein. The artery is cannulated to
monitor arterial pressure at the lower aorta and the vein is
cannulated with an electrode-tipped catheter advanced
into the right atrium to detect signal propagation. The
arterial pressure, ECG, LV pressure, atrial electrogram,
body temperature, and exhaled Pco2 are monitored
continuously. When the surgical preparation is stable,
baseline effective refractory periods (ERPs) are deter-
mined. The ERP is found by repeatedly pacing the heart
chamber with current 2.5· the capture threshold at a given
rate for 13 beats followed by a 3 s pause. With each pacing
train the time interval between the 12th and 13th stimulus
is shortened until the 13th does not cause signal propa-
gation in the tissue. The interval that does not result in a
propagated signal in 4 out of 5 repetitions is the ERP.
Hearts are paced at rates of 150, 200, 240, and 300 beats
per minute from the right atrium and the right ventricle.
Compound is infused IV over 15 min, ERP determinations
are made starting at the 12th minute of the infusion, and
the animal is allowed to stabilize for about 15 min before
another dose is given. After the final dose the ERPs are
determined every 15 min until the values are back at
baseline.
22. HERG currents are recorded by the whole cell mode of
patch clamp electrophysiology as described by Hamill
et al.25 HERG is stably overexpressed in HEK cells.
Microelectrodes are pulled from borosilicate glass
(TW150) and heat polished (tip resistance, 1.5–3 mega-
ohms). The external solution is standard Tyrode’s solu-
tion. The internal (microelectrode) solution contained:
110 mM KCl, 5 mM K2APT, 5 mM K4BAPT, 1 mM
MgCl2, and 10 mM Hepes, adjusted to pH 7.2 with KOH.
Command potentials are applied for 2 s to +20 mV from a
holding potential of À80 mV using Axon software
(pClamp 8.1) and hardware (Axopatch 1D, 200B). Tail
currents are generated by returning to À40 mV for 2 s.
Compounds are prepared as 10–20 mM DMSO stocks and
diluted to appropriate test concentrations. After stable
currents are achieved, compounds are perfused onto the
cells and the cells are pulsed every 20 s until no further
changes in current are evident at a given compound
concentration. Inhibition of HERG is measured at the
peak of the tail currents and expressed relative to controls.
Initial HERG activity is estimated by single point deter-
2. (a) Pratt, C. M.; Moye, L. A. Am. J. Cardiol. 1990, 65,
20B; (b) Waldo, A. L.; Camm, A. J.; DeRuyter, H.;
Friedman, P. L.; MacNeil, D. J.; Pauls, J. F.; Pitt, B.;
Pratt, C. M.; Schwartz, P. J.; Veltri, E. P. Lancet 1996,
348, 7; (c) Tomaselli, G. Heart Drug 2001, 1, 183.
3. (a) Wang, Z.; Fermini, B.; Nattel, S. Circ. Res. 1993, 73,
1061; (b) Feng, J.; Wible, B.; Li, G. R.; Wang, Z.; Nattel,
S. Circ. Res. 1997, 80, 572.
4. Purerfellner, H. Curr. Med. Chem. CV& H Agents 2004, 2,
79.
5. Lynch, J., Jr.; Swanson, R. J.; Fermini, B. PCT Int. Appl.
1998, WO 98/18475. US Patent, 5969017, 1999.
6. Lynch, J., Jr.; Swanson, R. J.; Fermini, B. PCT Int. Appl.
1998, WO 98/18476. US Patent, 5935945, 1999.
7. Stump, G. L.; Wallace, A. A.; Regan, C. P.; Lynch, J. J.,
Jr. J. Pharmacol. Exp. Ther. 2005, 315, 1362.
8. Castle, N. A.; Hughes, P. F.; Mendoza, J. S.; Wilson, J.
W.; Amato, G.; Beaudoin, S.; Gross, M.; Mcnaughton-
Smith, G. PCT Int. Appl. 1997, WO 98/04521. US Patent,
6083986, 2000.
9. Gross, M.; Castle, N. A. PCT Int. Appl. 1999, WO 99/
37607.
10. Gross, M.; Beaudoin, S.; Reed, A. D. PCT Int. Appl.
2001, WO 01/46155 A1. US Patent, 6458794 B2, 2002.
11. Castle, N. A.; Gross, M.; Mendoza, J. S. PCT Int. Appl.
1999, WO 99/62891. US Patent, 6174908 B1, 2001. US
Patent, 6395730 B1, 2002.
12. Peukert, S.; Brendel, J.; Hemmerle, H.; Kleemann, H.-W.
PCT Int. Appl. 2002, WO 02/48131 A1, WO 02/44137 A1,
WO 02/446162 A1.
13. Peukert, S.; Brendel, J.; Pirard, B.; Bruggemann, A.;
Below, P.; Kleemann, H.-W.; Hemmerle, H.; Schmidt, W.
J. Med. Chem. 2003, 46, 486.
14. Brendel, J.; Schmidt, W.; Below, P. PCT Int. Appl., 2001,
WO 01/25189.
15. Brendel, J.; Pirard, B. US Patent, 0193422 A1, 2002.
16. Brendel, J.; Bohme, T.; Peukert, S.; Kleemann, H.-W. US
Patent, 0114499 A1, 2003.
17. Wirth, K. J.; Paehler, T.; Rosenstein, B.; Knobloch, K.;
Maier, T.; Frenzel, J.; Brendel, J.; Busch, A. E.; Bleich, M.
Cardiovasc. Res. 2003, 60, 298.
18. Peukert, S.; Brendel, J.; Pirard, B.; Strubing, C.; Klee-
mann, H.-W.; Bohme, T.; Hemmerle, H. Bioorg. Med.
Chem. Lett. 2004, 14, 2823.
19. Wu, S.; Janusz, J. 227th ACS National Meeting, Anaheim,
CA. March 28, 2004.
20. Kv1.5 currents are recorded by the whole cell mode of
patch clamp electrophysiology. Kv1.5 is stably overex-
pressed in either HEK or LTK-cells. Microelectrodes are
pulled from borosilicate glass (TW150) and heat polished
(tip resistance, 1.5–3 megaohms). The external solution is
standard Tyrode’s solution. The internal (microelectrode)
solution contained: 110 mM KCl, 5 mM K2APT, 5 mM
K4BAPTA, 1 mM MgCl2, and 10 mM Hepes, adjusted to
pH 7.2 with KOH. Command potentials are applied for
1 s to +60 mV from a holding potential of À70 mV using
Axon software (pClamp 8.1) and hardware (Axopatch 1D,
200B). Compounds are prepared as 10–20 mM DMSO
stocks and diluted to appropriate test concentrations.
After stable currents are achieved, compounds are per-

6218
S. Wu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6213–6218
minations run at 10 lM. Concentration–response curves
then placed in the center position of the FLIPR 1(Fluo-
rometric Imaging Plate Reader, Molecular Devices Cor-
poration). Cell monolayers in each well are simultaneously
illuminated at 488 nm with an Argon ion laser, and
fluorescence emission is monitored using a 510–570 nm
bandpass filter and a cooled CCD camera. To depolarize
the plasma membrane and activate L-type calcium chan-
nels, 50 lL per well of 20 mM KCl (final concentration) is
dispensed simultaneously to all 96 wells using the FLIPR’s
automatic 96-well pipettor. Fluorescence measurements
are captured for 5 min following KCl addition. Calcium
influx, expressed as % control, is calculated for each
concentration of test compound and concentration–re-
sponse curves and IC50 values are generated using
GraphPad Prism 4.0.
are generated for appropriate compounds utilizing at least
four concentrations and an n = 3. Curve fitting and IC50
estimating were done using Graphpad software (Ver. 4).
23. HL-1 cells expressing endogenous L-type calcium channels
are removed from culture flasks using trypsin, plated on
fibronectin/gelatin-coated, clear-bottommed, black-walled
96-well microplates in Claycomb media (JRH Biosciences
#51800) containing 10% fetal bovine serum, 4 mM L-
glutamine, and 10 lM norepinephrine, and grown to
confluency overnight. The next day, growth medium is
aspirated from confluent cell monolayers and replaced
with 100 lL per well Tyrode’s solution (in mM: 130 NaCl,
4 KCl, 1.8 CaCl2, 1.0 MgCl2, 20 Hepes, and 10 glucose,
pH 7.35) and 50 lL per well FLIPR Calcium Assay kit,
component A (#R-8033, Molecular Devices Corporation)
and incubated for 60 min in a 5% CO2 37 ꢁC incubator.
Fifty microliters per well test compounds is added to the
plates and further incubated for 15 min in a 5% CO2 37 ꢁC
incubator. All final solutions contain the anion exchange
inhibitor, probenecid (2.5 mM). The 96-well plates are
24. Unpublished results, Steve Houser, Temple University: On
treatment in vitro with ICA-32, Icagen’s prototypical
thiazolidinone Kv1.5 inhibitor, both human and pig atrial
myocytes showed similar increases in action potential
duration.
25. Hamill et al. Pflugers Arch. 1981, 391, 85.