New fatty acid oxidation inhibitors with increased potency lacking adverse metabolic and electrophysiological properties

Article abstract of DOI:10.1016/j.bmcl.2003.09.093

New inhibitors of palmitoylCoA oxidation were synthesized based on a structurally novel lead, CVT-3501 (1). Investigation of structure-activity relationships was conducted with respect to potency of inhibition of cardiac mitochondrial palmitoylCoA oxidation and metabolic stability. Potent and metabolically stable analogues 33, 42, and 43 were evaluated in vitro for cytochrome P450 inhibition and potentially adverse electrophysiological effects. Compound 33 was also found to have favorable pharmacokinetic properties in rat.

Full text of DOI:10.1016/j.bmcl.2003.09.093

Bioorganic & Medicinal Chemistry Letters 14 (2004) 549–552
New fatty acid oxidation inhibitors with increased potency lacking
adverse metabolic and electrophysiological properties
a,
a
a
a
b
Dmitry O. Koltun, * Timothy A. Marquart, Kevin D. Shenk, Elfatih Elzein, Yuan Li,
c
b
b
b
c
c
Marie Nguyen, Suresh Kerwar, Dewan Zeng, Nancy Chu, Daniel Soohoo, Jia Hao,
b
c
c
b
Victoria Y. Maydanik, David A. Lustig, Khing-Jow Ng, Heather Fraser
a
and Jeffery A. Zablocki
aDepartment of Bioorganic Chemistry, 3172 Porter Dr., Palo Alto, CA 94304, USA
bDepartment of Drug Research and Pharmacological Sciences, 3172 Porter Dr., Palo Alto, CA 94304, USA
cDepartment of Pre-Clinical Development, and CV Therapeutics, 3172 Porter Dr., Palo Alto, CA 94304, USA
Received 21 July 2003; accepted 16 September 2003
Abstract—New inhibitors of palmitoylCoA oxidation were synthesized based on a structurally novel lead, CVT-3501 (1). Investi-
gation of structure–activity relationships was conducted with respect to potency of inhibition of cardiac mitochondrial palmi-
toylCoA oxidation and metabolic stability. Potent and metabolically stable analogues 33, 42, and 43 were evaluated in vitro for
cytochrome P450 inhibition and potentially adverse electrophysiological effects. Compound 33 was also found to have favorable
pharmacokinetic properties in rat.
#
2003 Elsevier Ltd. All rights reserved.
Fatty acids and glucose are the two major sources of
energy utilized by cardiac muscle. Compounds that
decrease fatty acid oxidation and subsequently increase
the rate of glucose oxidation are of potential therapeutic
to optimize the potency with respect to palmitoylCoA
oxidation inhibition and improve metabolic stability.
Synthesis of analogues of 1 is outlined in Scheme 1.
Commercially available 5-hydroxybenzothiazole was
reacted with (S)-epichlorohydrin under standard condi-
1À3
importance for the treatment of ischemic heart disease.
5
(
R)-CVT-3501 (1, Fig. 1) was found to be an inhibitor of
tions to provide the desired (R)-epoxide 2 in 60% yield,
6
1
4
oxidation of 1-[ C]-palmitoylCoA, a substrate for long-
chain fatty acid oxidation (palmitoylCoA oxidation
inhibitor) in cardiac mitochondrial preparations isolated
>99% ee. The epoxide was reacted with Boc-piper-
azine and the Boc-group was removed. The resulting
piperazine 3 underwent coupling with corresponding
chloroacetamides in a library format under solution-
4
from rat heart (IC : 2.5 mM). It was reasonably meta-
5
0
7
,8
9
bolically stable with 50% remaining after 30 min incu-
bation with human isolated liver S9 preparations.
phase or resin-mediated conditions.
Initially all the compounds were screened at 30 mM for
their activities in palmitoylCoA oxidation inhibition
assay. If the compounds inhibited palmitoylCoA oxida-
tion >50% their IC50 values were determined, other-
wise % inhibition is reported. Liver S9 stability has been
determined for all compounds that had IC50 values
under 30 mM. Results are presented in Table 1.
We conducted a study to investigate the structure–activity
relationships in the aryl amide portion of our lead in order
The change of substitution in the aromatic ring of the
amide portion (4–12), insertion of methylene (13 and
Figure 1. Structure of (R)-CVT-3501.
14) or saturation (15 and 16) did not increase potency.
Introduction of a methyl group in the benzyl position
(20–24) showed clear preference for (S)-configuration
*
Corresponding author. Fax: +1-650-858-0390; e-mail: dmitry.kol-
tun@cvt.com
0
960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.09.093

550
D. O. Koltun et al. / Bioorg. Med. Chem. Lett. 14 (2004) 549–552
and led to a potent compound (24, IC : 150 nM). The
5
maintaining high potency to inhibit palmitoylCoA oxi-
dation and liver S9 stability. Introduction of a basic
nitrogen atom (27–31) reduced the potency, but a non-
basic nitrogen was tolerated (indole 32 and carbazole
33). Compound 33 also had favorable stability (62%).
Partial reduction of the aromatic system maintained
potency in some cases but reduced liver S9 stability (34–
0
stability of 24, however was very low (14%). Homo-
benzylic substitution was not favorable (25, IC :
5
0
1
7 mM). Naphthalene analogue 26 was very potent
(
IC : 105 nM) and reasonably stable (50%), but its
5
0
further evaluation was not possible because of low
aqueous solubility.
40). In all cases, analogues with (S)-configuration in the
aryl amide portion had higher inhibitory activity than
(R).
A set of naphthalene analogues (27–40) was synthesized
in an attempt to improve aqueous solubility while
Table 1. PalmCoA inhibition activity and liver S9 stability of mono- and bicyclic amides
Substituent
Activitya
Substituent
Activitya
(stability)b
Substituent
Activitya
(stability)b
Substituent
Activitya
(stability)b
(
stability)b
1
4
5
6
7
8
2.5 mM (50%)
15
16
8.6 mM (37%)
5.1 mM (29%)
23% @ 30 mM
27
28
29
30
31
32
4.7 mM (2%)
4.7 mM (12%)
39
40
41
42
43
44
77 nM (2%)
30 mM (2%)
10 mM (47%)
c
c
42% @ 30 mM (N/D) 17
40% @ 30 mM
39% @ 30 mM (N/D)
180 nM (58%)
(
N/D)c
(N/D)c
7.6 mM (28%)
1.7 mM (51%)
6.0 mM (29%)
18
19
20
4.4 mM (23%)
16% @ 30 mM
(
N/D)c
40% @ 30 mM
5.7 mM (11%)
460 nM (63%)
(
N/D)c
c
710 nM (10%)
860 nM (35%)
34% @ 30 mM (N/D)
c
9
1
1
1
1
1
a
4.0 mM (25%)
22 mM (34%)
35 mM (41%)
21
22
23
24% @ 30 mM (N/D) 33
850 nM (62%)
610 nM (2%)
45
46
47
360 nM (15%)
330 nM (18%)
2.0 mM (64%)
0
1
2
3
4
8.5 mM (10%)
8.2 mM (45%)
150 nM (14%)
17 mM (27%)
105 nM (50%)
34
35
36
37
38
190 nM (15%)
c
35% @ 100 mM 24
50% @ 30 mM (N/D)
2.2 mM (1%)
c
(
N/D)
3.9 mM
25
26
c
2.0 mM (16%)
27 mM (N/D)
Activity is expressed as either IC50 or % inhibition at listed concentration as obtained from the rat heart mitochondrial PalmCoA assay. Values are presented as
average of two or three measurements.
b
c
Stability is defined as % remaining after 30 min incubation with human liver S9 (n=2, 2 mg/mL protein, NADH added).
Not determined.

D. O. Koltun et al. / Bioorg. Med. Chem. Lett. 14 (2004) 549–552
551
Table 2. Summary of metabolic and electrophysiological properties of 33, 42, and 43 compared to the initial lead 1
a
Compd
Cytochrome P450 inhibition assay (IC50, mM)
Electrophysiological effects at 3 mM
ÁH–Vb ÁCPPb
CYP1A2
CYP3A4
CYP2C9
ÁS–Hb
1
>100
>100
23
39
>100
>100
3.0
0
1.0
3.0
5.0
4.0
0
0
0
0.5
À8.0
0.7
À0.4
À4.3
À6.6
33
33
42
43
c
c
c
6.0
31
37
7.9
24
30
a
b
c
In guinea pig isolated hearts, average of n=2.
Change in S–H and H–V intervals was measured in ms, change in coronary perfusion pressure was measured in mm Hg.
Measured at 10 mM.
Table 3. Comparison of pharmacokinetic properties of 1 and 33 in
rats
1
33
POa
IVb
POa
IVb
C
AUC (0–24 h), (ng h/mL)
t1=2 (h)
max (ng/mL)
26.1
61.1
1.2
26.3
230
2.6
313
1.4
1315
3.0
t
Cl (mL/min/kg)
Vdss (L/kg)
F (%)
max (h)
0.83
4–6
p
54.5
6.2
13.5
3.4
19.5
17.5
Scheme 1. Reagents and conditions: (a) K
chlorohydrine, reflux, 24 h, 60%; (b) Boc-piperazine, CH
2
CO
3
, acetone, (S)-epi-
Cl , reflux,
, rt, 24 h, 100%; (d) conditions A
a
Dosed at 2 mg/kg, n=3.
Dosed at 1 mg/kg, n=3.
2
2
b
24 h, 96%; (c) 50% TFA in CH
2
Cl
2
ꢀ
(solution-phase): R-NHCOCH
conditions B (resin-mediated): R-NHCOCH
ethane/DMF, 70 C, 24 h, then PS-NCO and PS-trisamine, 70 C, 24 h.
2
Cl, (i-Pr)
2
NEt, DMF, 70 C, 24 h or
Cl, PS-DIEA, dichloro-
2
ꢀ
ꢀ
measured at 3 or even 10 mM (33) concentrations. The
effects of 33, 42, and 43 on coronary vasodialation are
expected to be minimal based on small changes in
ÁCPP observed at concentrations greater than IC50 for
their respective palmCoA activity.
A set of analogues (41–47) was synthesized with a focus
on bulky biphenyl and phenoxyphenyl amides that
mimic the tricyclic carbazole system. While ortho-sub-
stitution led to inactive analogues (41 and 44), meta-
and para- phenyl- and phenoxyphenyl-analogues (42, 43,
Analogues 33, 42, and 43 were also screened for human
cytochrome P450 inhibition and compared with parent
1
2À15
4
5, and 46) all had potencies of less than 500 nM. Liver
analogue 1.
Similarly to 1, compound 33 showed no
S9 stabilities were higher in biphenyl analogues 42 (58%)
and 43 (63%) and lower in the phenoxyphenyl analogues
45 (15%) and 46 (18%) which could be alleviated by
introduction of 4-chloro substituent (47, 64%). This
or only mild cytochrome P450 inhibition, while com-
pounds 42 and 43 were moderately potent inhibitors of
CYP1A2 and CYP3A4, respectively.
however resulted in a 6-fold decrease in potency.
Comparison of in vivo pharmacokinetic properties of 1
and 33 is shown in Table 3. While having oral bioavail-
ability similar to 1, 33 demonstrated slower clearance
and longer elimination half-life than compound 1
(Table 3) in rats.
The best analogues from our study (33, 42, and 43) were
screened for potential effects in isolated guinea pig heart
and compared with parent analogue 1 (Table 2). The
electrophysiological (EP) parameters measured were AV
nodal conduction times including the S–H interval (sti-
mulus to His-bundle conduction time) and H–V interval
In conclusion, we have identified a new fatty acid oxi-
dation inhibitor 33 with increased potency, desirable
metabolic stability, low cytochrome P450 inhibition
potential, and lack of adverse electrophysiological
effects. In addition, 33 had favorable in vivo pharma-
cokinetic properties. The ability of 33 to alter cardiac
energy substrate oxidation has yet to be determined.
(
His-bundle to ventricular conduction time) measured
from His-bundle electrogram. Changes in coronary
perfusion pressure (CPP), were also monitored. Pro-
longation of the S–H interval has been interpreted to
suggest inhibition of calcium dependent currents
whereas prolongation of the H–V interval has been
interpreted to suggest inhibition of sodium dependent
currents. Hence, the effect of compounds on these
parameters may provide indirect information about its
Acknowledgements
1
0,11
potential effects on membrane ion currents.
respect, properties of all three compounds 33, 42, and
3 presented a substantial improvement over 1 when
In that
We would like to thank Dr. Brent Blackburn, Dr. Luiz
Belardinelli, and Dr. Kwan Leung for valuable input
and discussions.
4

5
52
D. O. Koltun et al. / Bioorg. Med. Chem. Lett. 14 (2004) 549–552
DIEA resin (200 mg) was added, and the vial shaken at
References and notes
ꢀ
7
200 mg each of PS-Isocyanate resin and PS-Trisamine
0 C overnight. After cooling to room temperature,
1
2
3
4
5
. Schofield, R. S.; Hill, J. A. Am. J. Cardiovasc. Drugs
ꢀ
2
001, 1, 23.
resin were added and shaken at 70 C overnight. After
. McCormack, J. G.; Stanley, W. C.; Wolff, A. A. Gen.
Pharm. 1998, 30, 639.
. McCormack, J. G.; Barr, R. L.; Wolff, A. A.; Lopaschuk,
G. D. Circulation 1996, 93, 135.
. Nedergaard, J.; Cannon, B. Methods Enzymol. 1979, 55,
cooling to room temperature, contents of the vial were
transferred into a frit-fitted syringe, filtered, resins washed
with additional dichloroethane repeatedly. After concen-
1
tration on Speedvac the crude mixture was purified with
semi-preparative reverse-phase HPLC (ACN/water). Pro-
ducts were isolated in 5–70% yields with purities exceed-
ing 95% (as determined by analytical HPLC) in most
cases. Structures of all final products were supported by
3.
. Shiratsuchi, M.; Kawamura, K.; Akashi, T.; Ishihama,
H.; Nakamura, M.; Takenaka, F. Chem. Pharm. Bull.
1
1
987, 35, 3691.
ESI MS and H NMR. All resin-bound reagents were
obtained from Argonaut Technologies.
6
. Enantiomeric excess has been determined by chiral HPLC
ꢀ
(Chiralcel OD column, 40 C, 89.1% hexanes/9.9%
EtOH/1% Et NH).
10. Lazzara R.; Scherlag G. J.; Belardinelli L. 2002. Atrio-
ventricular Conduction. In Foundations of Cardiac
Arrhythmias: Basic Concepts and Clinical Approaches;
Spooner, P. M, Rosen, M. R., Eds.; Marcel Dekker; New
York, Basel; p 265.
11. Compounds were formulated as x HCl salts by treatment
with excess HCl to assist solubility and administered with
5% DMSO and 0.9 M NaCl.
12. Tassaneeyakul, W.; Birkett, D.; Veronese, M.; McManus,
M.; Tukey, R.; Quattrochi, L.; Gelboin, H.; Miners, J. J.
Pharmacol. Exp. Ther. 1993, 265, 401.
13. Scmitz, G.; Lepper, H.; Estler, C. J. Chromatogr. 1993,
620, 158.
2
7
8
. Oshiro, Y.; Sakurai, Y.; Tanaka, T.; Kikuchi, T.; Hirose,
T.; Tottori, K. J. Med. Chem. 1991, 34, 2014.
. Solution-phase coupling procedure (conditions A): chlor-
oacetamide (0.14 mmol) and monosubstituted piperazine
.
(
4
(
7
solution in DMF, 0.17 mmol, 0.6 mL) were placed into a
-mL vial and diluted with 1.8 mL DMF. Hunig’s base
0.28 mmol) was added via syringe and the vial shaken at
0 C overnight. Products were isolated in 5–70% yields
with purities exceeding 95% (as determined by analytical
HPLC) in most cases. Structures of all final products were
ꢀ
1
supported by ESI MS and H NMR.
. Resin-mediated coupling procedure (conditions B): chlor-
oacetamide (0.14 mmol) and monosubstituted piperazine
9
14. Leemann, T.; Transom, C.; Dayer, P. Life Sci. 1993, 52,
29.
(
4
solution in DMF, 0.17 mmol, 0.6 mL) were placed into a
-mL vial and diluted with 1.8 mL dichloroethane. PS-
15. Wienkers, L.; Steenwyck, R.; Pearson, P. Drug Metab.
Dispos. 1995, 23, 383.