Pyrano-[2,3b]-pyridines as potassium channel antagonists

Article abstract of DOI:10.1016/j.bmcl.2008.03.026

The design and synthesis of a series of highly functionalized pyrano-[2,3b]-pyridines is described. These compounds were assayed for their ability to block the I_Kur channel encoded by the gene hKV1.5 in patch-clamped L-929 cells. Six of the compounds in this series showed sub-micromolar activity, the most potent being 4-(4-ethyl-benzenesulfonylamino)-3-hydroxy-2,2-dimethyl-3,4-dihydro-2H-pyrano[2,3b]-pyridine-6-carboxylic acid ethyl-phenyl-amide with an IC₅₀ of 378 nM.

Full text of DOI:10.1016/j.bmcl.2008.03.026

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2714–2718
Pyrano-[2,3b]-pyridines as potassium channel antagonists
Heather J. Finlay,a,* John Lloyd,a Michael Nyman,b, Mary Lee Conder,b
Tonya West,b Paul Levesqueb and Karnail Atwala
aDepartment of Discovery Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute,
PO Box 5400, Princeton, NJ 08543-5400, USA
bDepartment of Biology, Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ 08543-5400, USA
Received 8 November 2007; revised 4 March 2008; accepted 6 March 2008
Available online 14 March 2008
Abstract—The design and synthesis of a series of highly functionalized pyrano-[2,3b]-pyridines is described. These compounds were
assayed for their ability to block the IKur channel encoded by the gene hKV1.5 in patch-clamped L-929 cells. Six of the compounds
in this series showed sub-micromolar activity, the most potent being 4-(4-ethyl-benzenesulfonylamino)-3-hydroxy-2,2-dimethyl-3,4-
dihydro-2H-pyrano[2,3b]-pyridine-6-carboxylic acid ethyl-phenyl-amide with an IC50 of 378 nM.
ꢀ 2008 Elsevier Ltd. All rights reserved.
Ventricular and atrial cardiac arrhythmias1,2 affect a
significant proportion of the general population. The
most common form of sustained cardiac arrhythmia
is atrial fibrillation (AF) which affects approximately
2.2 million adults in the US.3 The incidence of AF in-
creases significantly with age; <1% of 50–59 years old
were diagnosed with AF compared with 7–13% of the
octogenarian population.4 Concomitant with an aging
population, the prevalence of AF is increasing.5 Atrial
fibrillation is characterized by the rapid and irregular
contraction of the atria, which can lead to a higher inci-
dence of stroke6,7 via blood stasis and higher mortality
in patients with congestive heart failure.8 Conversion to
normal sinus rhythm may be achieved through invasive
AV node ablation and pacemaker insertion,9 catheter
based ablation10 or administration of antiarrhythmic
drugs.11,12
results in an increase in action potential duration.15
Three of the most important outward potassium cur-
rents are IKs (slowly activating) IKr (rapidly activat-
ing) and IKur17 (ultra rapidly activating). IKs and IKr are
present in the human atrium and ventricle, but IKur is at-
rial-specific.18
16
17
Currently approved class III agents, for example, D,L-
sotalol, amiodarone, dofetilide and ibutilide inhibit IKr
and have the potential liability of being proarrhythmic
in the ventricle.19,20 Ventricular arrhythmias, such as
torsades de pointe, are potentially fatal and limit initial
administration of several of these drugs to an in-patient
environment.21 Drugs which selectively target inhibition
of IKur may be safer since they prolong action potential
duration in the atrium, without delaying ventricular
repolarization and, therefore, should be without the sub-
sequent risk of ventricular arrhythmia.22,23
Potassium channels are transmembrane protein chan-
nels which selectively allow potassium ions across the
plasma membrane.13 The movement of ions across cell
membranes mediates many biological processes includ-
ing regulation of action potential duration in cardiac
cells.14 Potassium currents are essential for cardiac cell
repolarization and blocking of these outward currents
Early efforts in our program had identified a series of
potent and selective indane and tetraline24 based IKur
inhibitors. These series had low oral bioavailability
and lacked metabolic stability (e.g., 1, 2, Fig. 1). Our
goal was to investigate alternate related chemotypes
and determine the viability of those series showing po-
tency for IKur. A series of benzopyrans25 (e.g., 3,
Fig. 1) were investigated which led to synthesis and eval-
uation of a series of pyrano-[2,3b]-pyridine26 based
inhibitors which are described in this letter.
Keywords: Atrial fibrillation; Potassium channel; KV1.5; Pyrano-[2,3-
b]-pyridine.
*
Corresponding author. Tel.: +1 609 818 3734; e-mail: heather.finlay
@bms.com
The substitution at position 4 on the benzopyran moiety
had been optimized as the 4-ethyl-phenyl-sulfonamide.25
Present address: Exelixis, Inc., 4757 Nexus Centre Drive, San Diego,
CA, USA.
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.026

H. J. Finlay et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2714–2718
2715
a,b,c
Br
O
O
S
HN
H
N
N
O
N
O
O
7
8
OH
O
d
O
O
e,f
O
1
EtO
EtO
O
S
N
O
N
O
O
9
HN
O
g,h,i
N
O
S
N
O
HN
O
OH
HO
2
N
O
10
j
O
S
O
HN
O
O
OH
R
O
S
F
O
HN
O
OH
N
H
R =
N
O
3
4
5
N
HN
11
N
N
Scheme 1. Reagents and conditions: (a) Br2, acetic acid, 80 ꢁC, 50%
yield; (b) n-BuLi, À78 ꢁC, 2-butenal, 52% yield; (c) HBr (48%
aqueous), acetic acid, 100 ꢁC, 53% yield; (d) t-BuLi, À78 ꢁC, ethyl-
chloroformate, 54% yield; (e) NBS, DMSO aqueous, 100% yield; (f)
KOH, THF, 100% yield; (g) NH4OH, EtOH, 80 ꢁC, 75% yield; (h) 4-
ethyl-phenylsulfonyl chloride, TEA, DCM, 73% yield; (i) LiOH,
aqueous acetone, 65 ꢁC, 59% yield; (j) EDCI, HOAt, 2-fluoro
benzylamine, 43% yield.
Br
F
6
NH
Figure 1. Examples of early chemotypes.
Thus, this moiety was initially incorporated into the
pyrano-[2,3b]-pyridine series. The 2H-pyrano-[2,3b]-
pyridine core was synthesized as described previously.27
The resulting 6-bromo-2,2-dimethyl-2H-pyrano-[2,3b]-
pyridine (7) was converted to the corresponding ethyl
ester (8) by halogen metal exchange and quench with
ethylchloroformate (Scheme 1).28 Epoxidation was
accomplished via the bromohydrin.29,30 Closure of the
bromohydrin to the epoxide was unsuccessful using
the standard NaH/DMF conditions, due to difficulty
in extracting the epoxide from the DMF solution.
Alternatively, the epoxide (9) was formed in quantita-
tive yield by treatment of the bromohydrin with
KOH in THF and subsequently converted to the race-
mic trans amino alcohol using ammonium hydroxide in
ethanol (Scheme 1).29 We had demonstrated previously
that the absolute stereochemistry at positions 3 and 4
was not critical for potency in the benzopyran ana-
logs25 and thus, racemic trans amino alcohol was used
throughout this study. Standard sulfonylation condi-
tions, followed by hydrolysis yielded the 6-carboxylic
acid (10).31 For the benzopyran series, we were able
to utilize fluoro-N,tetramethylformamidinium hexa-
fluorophosphate32 for coupling example amines to the
6-carboxylic acid, (10) but this reagent was not success-
ful for coupling in the pyrano-[2,3b]-pyridine series.
Individual compounds were coupled using standard
EDCI/HOAt conditions33 (e.g., 11,34 Scheme 1) in
yields from 30% to 93%. Product amides with addi-
tional basic amines (e.g., 19) were purified by retention
on an SCX cartridge followed by elution with ammo-
nia/MeOH solution. Non-basic products were purified
by reverse phase preparative HPLC.
To flush out SAR at the 6-position, synthesis of a par-
allel solution phase amidation library furnished 42 di-
verse amides, examples of which are shown in Table 1.
From this first library, it was clear that initial SAR for
this series was divergent to that observed for the benz-
opyran series. Compound 21 was identified as a weak
inhibitor from this library and a second 2-step solu-
tion phase library was prepared to determine SAR
around the N-ethylated benzylamine and aniline
amides. N-Ethylated benzylamines and anilines were
synthesized in parallel by acylation of the correspond-
ing amine with acetyl chloride and subsequent
reduction with LAH at 70 ꢁC (Scheme 2). The crude
N-ethylamines were cautiously quenched with MeOH
and purified using automated C18 chromatography.
N-Ethyl benzylamines were coupled to acid (10) using
standard EDCI/HOAt conditions33 and N-ethyl ani-
lines were coupled to acid (10) using PyBrOP, triethyl-
amine in acetonitrile at 80 ꢁC.35
Product amides were purified by reverse-phase prepara-
tive HPLC. Additional 34 N-ethyl amide analogs were
synthesized in parallel via this route and examples are
included in Table 1.

2716
H. J. Finlay et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2714–2718
Table 1. Examples of amides from parallel solution phase coupling
F
F
a,b
N
H
NH2
O
S
F
F
O
HN
O
OH
R
Scheme 2. Reagents and conditions: (a) acetyl chloride, TEA, rt, 100%
yield; (b) LAH, THF, 70 ꢁC, 52% yield.
N
O
All compounds synthesized were characterized in L-929
cells that stably expressed human KV1.5. Compounds
were initially tested at 1 lM concentration using volt-
age clamp techniques.36,37 The % inhibition was mea-
sured and is reported as an average value from
testing in triplicate. For those compounds with >50%
inhibition at 1 lM, the IC50 for block of KV1.5 current
was subsequently measured. Compounds with % inhibi-
tion at 1 lM and subsequent IC50 determinants are
shown in Table 2.
N
12
23
24
HN
O
N
Cl
13
N
N
F
In the benzopyran series, amides 3–6 were among the
most potent compounds. However, in the pyrano-
[2,3b]-pyridines series, the corresponding analogs 12,
21, 17, and 18, respectively, were significantly less po-
tent. Overall, the most potent pyrano-[2,3b]-pyridine
identified in this study was N-ethyl amide 28 with an
IC50 value of 378 nM.
N
14
15
25
26
N
N
O
O
O
NH
N
N
N
In continued efforts to improve the aqueous solubility of
this series, basic amines, hydroxyl groups and heterocy-
cles were incorporated into the amide functionality (e.g.,
14, 19, and 20), but polar groups and heterocycles were
not tolerated at the amide position.
16
17
18
19
20
21
22
27
28
29
30
31
32
The most potent compound identified from the first li-
brary synthesis was benzylamide 21. The des-fluoro ben-
zylamide direct analog, 22 and the corresponding N-
methyl benzylamide 16 were less potent. Direct N-ethyl
benzylamide analogs of 21 showed some improvement
in potency (e.g., 23 and 24), with incorporation of aryl
group substituents. However, the SAR at this position
was narrow, for example, methoxy substitution was tol-
erated at the para position, but not at the ortho or meta
positions (23 vs 26 and 27). Additionally, para methoxy
substitution was tolerated, but not para methyl substitu-
tion (23 vs 25). In the aniline series, N-ethyl substitution
was also required for potency (28 vs 17 and 13). Addi-
tional efforts to explore the substitution on the aryl moi-
ety did not result in improved potency and SAR proved
to be divergent from that observed with the benzyla-
mides (e.g., 29 vs 23). The most potent aryl substituted
anilines had meta substituents (30 and 31) but these ana-
logs were 2-fold less potent than unsubstituted N-ethyl
aniline lead compound 28.
N
N
Br
N
NH
N
NH
NH
HN
Cl
N
O
N
S
N
NH
N
F3C
F
Some compounds in this series demonstrated signifi-
cantly improved equilibrium solubility in aqueous buffer
over the corresponding benzopyran amide direct ana-
logs (e.g., 4 had aqueous solubility in pH 6.5 buffer of
0.010 mg/mL compared to 21 with solubility 0.067 mg/
mL).38 The most potent pyrano-[2,3b]-pyridine, 28 had
an aqueous solubility of 0.108 mg/mL. However, due
to the generally reduced potency compared to the benz-
opyran series, the potential metabolic liabilities of the
F
N
F
N
N

H. J. Finlay et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2714–2718
Table 2. IC50 inhibition results for compounds
2717
Compound
Inhibition of current
in L-929 cells %
Inhibition of current
in L-929 cells
Compound
Inhibition of current
in L-929 cells %
Inhibition of current
a
in L-929 cells IC50 (lM)
a
inhibition at 1 lM
IC50 (lM)
inhibition at 1 lM
1
2
—
—
—
—
87
73
4
0.050
0.046
0.060
0.172
0.281
0.316
—
19
20
21
22
23
24
25
26
27
28
29
30
31
32
31
3
—
—
3
38
20
70
54
15
35
23
89
22
64
58
52
—
—
4
5
0.605
0.615
1.56
2.09
—
6
11
12
13
14
15
16
17
18
15
30
2
—
—
—
—
0.378
—
17
12
12
8
—
—
0.649
0.776
0.912
—
a Inhibition is measured in duplicate at 9 concentrations and the mean values were used to calculate IC50 values.
aniline functionality39 and the lack of clear SAR, further
22. Nattel, S. Nature 2002, 415, 219.
23. Knobloch, K.; Brendel, J.; Peukert, S.; Rosenstein, B.;
Busch, A. E.; Wirth, K. J. Naunyn Schmiedebergs Arch.
Pharmacol. 2002, 366, 482.
efforts were focused on investigation of alternate
chemotypes.
24. Castle, N. A.; Hollinshead, S. P.; Hughes, P. F.; Mendoza,
J. S.; Wilson, J. W.; Wilson, J. W.; Amato, G.; Beaudoin,
S. Patent Application WO 9804521, 1998.
References and notes
25. Lloyd, J.; Atwal, K. A.; Finlay, H. J.; Nyman, M.; Hyunh,
T.; Bhandaru, R.; Kover, A.; Schmidt, J.; Vacarro, W.;
Levesque, P.; Conder, M.; West, T. Bioorg. Med. Chem.
Lett. 2007, 17, 3271.
1. Denes, P.; Gillis, A. M.; Pawitan, Y.; Kammerling, J. M.;
Wilhelmsen, L.; Salerno, D. M. Am. J. Cardiol. 1991, 68,
887.
2. Go, A. S.; Hylek, E. M.; Philips, K. A.; Chang, Y.;
Henault, L. E.; Selby, J. V.; Singer, D. E. J. Am. Med.
Assoc. 2001, 285, 2370.
26. Lloyd, J.; Finlay, H. J.; Vaccaro, W.; Atwal, K. A.; Gross,
M. F.; Spear, K. L. Patent Application WO 0012077,
2000.
3. Feenberg, W. M.; Blackshear, J. L.; Laupais, A.; Kron-
mal, R.; Hart, J. Arch. Intern. Med. 1995, 155, 469.
4. Ryder, M. K.; Benjamin, E. J. Am. J. Cardiol. 1999, 84,
131.
27. Evans, J. M.; Stemp, G. Synth. Commun. 1988, 18, 1111.
28. Bargar, T. M.; Dulworth, J. K.; Kenny, M. T.; Massad,
R.; Daniel, J. K.; Wilson, T.; Sargant, R. N. J. Med.
Chem. 1986, 29, 1590.
5. Wolf, P. A.; Benjamin, E. J.; Belanger, A. J.; Kannel, W. B.;
Levy, D.; D’Agostino, R. B. Am. Heart J. 1996, 131, 790.
6. Hart, R. G.; Halperin, J. L. Stroke 2001, 32, 803.
7. Wolf, P. A.; Abbott, R. D.; Kannel, W. B. Stroke 1991,
22, 983.
29. Freshly recrystallized NBS was used for hydro bromina-
tion since the presence of HBr resulted in significant
amounts of the 3,4-dibromide.
30. Burrell, G.; Cassidy, F.; Evans, J. M.; Lightowler, D.;
Stemp, G. J. Med. Chem. 1990, 33, 3023.
8. Behan, S.; Zahawi, Z.; Goldbourt, U.; Reicher-Reiss, H.
Eur. Heart J. 1992, 13, 45.
9. Williamson, B. D.; Man, K. C.; Daoud, E., et al. N. Eng.
J. Med. 1994, 33, 910.
31. Patel, D. V.; VanMiddlesworth, F.; Donabauer, J.;
Gannett, P.; Sih, C. J. J. Am. Chem. Soc. 1986, 108, 15.
32. Carpino, L. A.; El-Faham, A. J. Am. Chem. Soc. 1995,
117, 5401.
10. Cox, J. L.; Schuessler, R. B.; Lappas, D. G. Ann. Surg.
1996, 224, 267.
11. Nattel, S.; Khairy, P.; Roy, D.; Thibault, B.; Guerra, P.;
Talajic, M.; Dubuc, M. Drugs 2002, 62, 2377.
12. Gilligan, D. M.; Ellenbogen, K. A.; Epstein, A. Am. J.
Med. 1996, 101, 413.
33. Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
34. All new compounds exhibited satisfactory spectroscopic
and/or analytical properties: example data for 11 Scheme
1.40
35. Coste, J.; Frerot, E.; Jouin, P. J. Org. Chem. 1994, 59,
2437.
13. Yellen, G. Nature 2002, 419, 35.
36. Po, S.; Roberds, S.; Snyders, D. J.; Tamkun, M. N.;
Bennett, P. B. Circ. Res. 1993, 72, 1326.
37. Snyders, D. J.; Tamkun, M. N.; Bennett, P. B. J. Gen.
Physiol. 1993, 101, 513.
38. A saturated solution in aqueous (buffered at pH 6.5) was
prepared and the solution sonicated at room temperature
for 18–24 h and then centrifuged. The supernatant was
analyzed by LC using a standard of the parent compound
in methanol to calibrate.
39. Kugler-Steigmeier, M. E.; Friederich, U.; Graf, U.; Lutz,
W. K.; Maier, P.; Schlatter, C. Mutat. Res. Fundam. Mol.
Mech. Mutagen. 1989, 211, 279.
14. Keating, M. T.; Sanguinetti, M. C. Cell 2001, 104, 569.
15. Synder, D. J. Cardiovasc. Res. 1999, 42, 377.
16. Barhanin, J.; Lesage, F.; Guillemore, E.; Fink, M.;
Lazdunski, M.; Romey, G. Nature 1996, 384, 78.
17. Wang, Z.; Fermini, B.; Nattel, S. Circ. Res. 1993, 73, 1061.
18. Li, G.-R.; Feng, J.; Wang, Z.; Fermini, B.; Nattel, S. Circ.
Res. 1996, 78, 903.
19. Curran, M. E.; Splawski, I.; Timothy, K. W.; Vincent,
G. M.; Green, E. D.; Keating, M. T. Cell 1995, 80,
795.
20. Marban, E. Nature 2002, 415, 213.
21. DeBruin, M. L.; Hoer, A. W.; Leukens, H. G. Am. J.
Cardiol. 2003, 91, 59.
40. Example data for 11, Scheme 1: 1H NMR (400 MHz,
CDCl3) 8.66 (1H, s), 8.42 (1H, s), 7.76 (2H, d, J = 8.2 Hz),

2718
H. J. Finlay et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2714–2718
7.62 (1H, br s), 7.29 (2H, d, J = 8.1 Hz), 7.23 (2H, m), 7.06
CDCl3) À76.25 (s). LC–MS retention time: 3.58 min
(100%) YMC ODS S5 4.6 · 50 mm, 4 min gradient 10%
MeOH/90% H2O À0.1% TFA to 90% MeOH/10% H2O
À0.1% TFA. 4 mL/min flow rate, k = 220 nM. [M+1]
514.20.
(1H, dd, J = 7.6 Hz), 7.03 (1H, dd, J = 7.6 Hz), 4.56 (2H,
d, J = 5.2 Hz), 4.29 (1H, d, J = 9.1 Hz), 3.71 (1H, d,
J = 9.7 Hz), 2.68 (2H, q, J = 7.6 Hz), 1.48 (3H, s), 1.25
(3H, s), 1.23 (3H, t, J = 7.6 Hz). 19F NMR (376 MHz,