Potent Kv1.3 inhibitors from correolide - Modification of the C18 position

Article abstract of DOI:10.1016/j.bmcl.2004.10.058

Correolide (1) was converted to a new series of tetracyclic Kv1.3 blockers 2. SAR for this class of compounds in two functional assays, Rb_Kv and human T cell proliferation, is presented herein. Kv1.3, the voltage-gated potassium channel in human T cells, represents a new target for treating immunosuppression and autoimmune diseases. Correolide (1), a pentacyclic natural product, is a potent and selective Kv1.3 channel blocker. Simplification of correolide via removal of its E-ring generates enone 4, whose modification produced a new series of tetracyclic Kv1.3 blockers. The structure-activity relationship for this class of compounds in two functional assays, Rb_Kv and human T cell proliferation, is presented herein. The most potent analog 43 is 15-fold more potent than correolide as inhibitor of human T cell proliferation.

Full text of DOI:10.1016/j.bmcl.2004.10.058

Bioorganic & Medicinal Chemistry Letters 15 (2005) 447–451
Potent Kv1.3 inhibitors from correolide—modification
of the C18 position
Jianming Bao,a,* Shouwu Miao,a Frank Kayser,a Andrew J. Kotliar,a Robert K. Baker,a
George A. Doss,a John P. Felix,b Randal M. Bugianesi,b Robert S. Slaughter,b
Gregory J. Kaczorowski,b Maria L. Garcia,b Sookhee N. Ha,c Laurie Castonguay,c
Gloria C. Koo,d Kashmira Shah,d Marty S. Springer,d Mary Jo Staruch,d
William H. Parsonsa and Kathleen M. Rupprechta
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Ion Channels, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Molecular Systems, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dDepartment of Immunology Research, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 14 September 2004; revised 15 October 2004; accepted 18 October 2004
Available online 11 November 2004
Abstract—Kv1.3, the voltage-gated potassium channel in human T cells, represents a new target for treating immunosuppression
and autoimmune diseases. Correolide (1), a pentacyclic natural product, is a potent and selective Kv1.3 channel blocker. Simplifi-
cation of correolide via removal of its E-ring generates enone 4, whose modification produced a new series of tetracyclic Kv1.3
blockers. The structure–activity relationship for this class of compounds in two functional assays, Rb_Kv and human T cell pro-
liferation, is presented herein. The most potent analog 43 is 15-fold more potent than correolide as inhibitor of human T cell
proliferation.
Ó 2004 Elsevier Ltd. All rights reserved.
The voltage-gated potassium channel, Kv1.3, sets the
resting membrane potential of human T lymphocytes.1,2
Blockade of Kv1.3 channels causes membrane depolari-
zation and this leads to attenuation of the influx of
extracellular calcium that occurs upon activation of
the T cell receptor complex. Since this increase in cal-
cium concentration is necessary for T cell proliferation,
a selective Kv1.3 channel blocker may represent a new
class of immunosuppressant agent.3–5 Levels of Kv1.3
channels are significantly and specifically up-regulated
in pathogenic myelin-reactive T cells from patients with
multiple sclerosis (MS) and Kv1.3 inhibitors prevent
and reverse the symptoms of the disease in animal mod-
els of MS. These data suggest that Kv1.3 inhibitors will
have utility in MS and other autoimmune disorders.6
Correolide 1, a pentacyclic nor-triterpenoid isolated
from the plant Spachea correa, is a selective, bioavail-
able inhibitor of the Kv1 family of potassium channels
and inhibits human T cell proliferation in a dose-
dependent manner.7–9 CorreolideÕs complexity and lim-
ited supply warrant attempts toward simplification of
the molecule with the goal of preparing synthetic ana-
logs. The results of SAR studies at C18 of enone 4 in
the Kv1.3 functional assay (Rb_Kv) that measures
depolarization-induced 86Rb+ efflux from CHO cells sta-
bly transfected with the human Kv1.3 channel,8 and in
human T cell9 proliferation are presented herein.10–12
The conversion of 1 to oxepin 2 and the excision of the
E-ring via a double-retroaldol fragmentation reaction of
the C20 ketone 3 to tetracyclic enone 4 were described in
an earlier publication (Scheme 1).13
Keywords: Potassium channel; Kv1.3; Correolide; Ion channel blocker;
T cell; Immunosuppressant.
This process eliminates five stereocenters but the simpli-
fication is accompanied by a significant drop in potency.
The challenge now is to restore potency.
*
Corresponding author. Tel.: +1 732 594 6650; fax: +1 732 594
5350; e-mail: bao_jianming@merck.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.10.058

448
J. Bao et al. / Bioorg. Med. Chem. Lett. 15 (2005) 447–451
29
20
21
19
H
O
H
O
22
CO2Me
E
E
a, b
13
13
CO2Me
H
H
OH
16
H
H
OH
16
OAc
OAc
O
A
OAc
OAc
OAc
OAc
O
O
AcO
OAc
AcO
OAc
1 correolide
Rb_Kv IC50 = 100 nM
2 IC50 = 10 nM
c, d
O
T-Cell (anti-CD3) IC50 =184 nM
O
H
H
O
E
18
13
H
H
e
13
CO2Me
H
H
OH
OAc
OAc
O
OAc
OAc
O
OAc
AcO
OAc
AcO
OAc
4 IC50 = 5.6 µM
3 IC50 = 35 nM
Scheme 1. Reagents and conditions: (a) LiHAl(OtBu)3, CH2Cl2, 0°C–rt, 18h, 97%; (b) BF3–Et2O, Et3SiH, CH2Cl2, rt, 0.5h, 88%; (c) OsO4, NMMO;
(d) NaIO4, 95%, two steps; (e) LiCl, 125°C, DMSO, 5h, 66%.
Table 1. C18 SAR in Rb_Kv assay8
Models of 4 suggest that carbanion addition to the C18
carbonyl would occur from the a-face of the tetracycle
R
H
18
since approach from the b-face would be blocked by
H
H
the C14 methyl group. In the product of this addition,
the hydroxyl group (Fig. 1) would have the same orien-
tation as the C18 hydroxy group of correolide and the
carbon chain would occupy the same space as the E-
ring.
OH
OAc
O
OAc
AcO
OAc
Compd
R
IC50 (nM)
1814
Addition of Grignard reagents to enone 4 at C18 was
stereospecific. The stereochemistry of the adducts was
confirmed to be as predicted by extensive NMR studies,
which showed a strong NOE between the C18 hydroxy
proton and the C14 methyl group. It was also deter-
mined that there was no epimerization at H13 during
the reaction.
6
7
143
664
21
8
9
10
11
2000
380
Ph
Ph
Two series of analogs were prepared from Grignard
addition reactions and their data are listed in Table 1.
In the alkyl series, the 2-methyl-2-pentenyl analog 9 is
the most potent. In the second series, aryl groups are at-
tached to C18 via tethers of varying lengths. That com-
pound 12, in which the aryl group is attached to C18 by
a 2-carbon tether, is the most potent is consistent with
Ph
12
26
13
14
150
270
Ph
Ph
the SAR of the alkyl series. It is also consistent with
computer modeling studies, which show the ethylene
group of compound 12 overlapping well with C19 and
C20 of correolide (Fig. 2). Compound 12 is 4-fold more
potent than correolide in the Rb_Kv assay.
In the models presented in Figure 2, the phenyl group of
12 occupies the same space as the epoxide and methyl
ester groups on C21 and C22 of correolide. Therefore,
a series of substituted phenyl analogs were prepared.
The activities of these analogs (12, 16–24) are listed in
Table 2. In general, they are more potent than correolide
in the Rb_Kv assay; substituents at the para-position of
the phenyl ring reduce potency (24). In the T cell assay,
Figure 1. Nucleophilic addition to enone 4.

J. Bao et al. / Bioorg. Med. Chem. Lett. 15 (2005) 447–451
449
O
H
H
H
H
Ph
a
A
+
OH
TMS
H
H
OAc
OAc
O
AcO
OAc
OAc
O
OAc
AcO
OAc
4
26
b
H
H
OH
H
H
OH
H
H
Figure 2. Superposition of correolide and 12.
+
OAc
OAc
OAc
O
O
OAc
AcO
OAc
AcO
OAc
Table 2. C18 SAR in Rb_Kv and T cell9 assays (IC5, nM)
28
27
Scheme 2. Reagents and conditions: (a) TiCl4, À40°C, CH2Cl2, 12h,
80%; (b) WilkinsonÕs catalyst, H2 (50psi), THF, 23h, rt, 90%.
R
H
18
cell assay and equivalent to correolide. This result is
consistent with modeling studies indicating that the dou-
ble bond at 26 overlaps well with the C20–C29 alkene of
correolide (Fig. 3).
H
H
OH
OAc
O
OAc
AcO
OAc
Compd
R
Rb_Kv
T cell
Ratioa
The double bond of the C18 substituent in 26 was selec-
tively reduced with hydrogen in presence of WilkinsonÕs
catalyst to provide 27 and 28 in a 5:1 ratio. The stereo-
chemistry of the newly generated methyl center in 27
and 28 was assigned based on independent synthesis
(Scheme 3).
12
16
17
18
19
20
21
22
23
24
H
2-F
26
74
87
41
26
31
14
34
24
130
460
245
1840
194
428
—
0.40
0.75
0.10
0.95
0.43
—
2-CF3
2-Me
2-OH
2-OAc
2-OMe
2-OEt
3-OMe
4-OMe
142
80
1.30
2.30
0.14
—
The SAR data of the compounds with substitutions in
the ethylene tether are summarized in Table 3. Although
compound 25 is a 1:1 mixture of diastereomers, its activ-
ity suggests that any substitution at the carbon next to
C18 has a detrimental effect on the Rb_Kv activity (25).
1314
—
a IC50 ratio of correolide to compound (bigger is more potent).
Substitution at the carbon next to the phenyl affects
the orientation of the C18 substituent. Models show that
the phenyl group and olefin of 26 overlap closely with
2-alkoxy analogs 21 and 22 are the most potent.
Although certain compounds are more potent than cor-
reolide in the Rb_Kv assay (i.e., 12, 19 and 23), this
potency does not translate into inhibition of T cell
proliferation. The differences may be to due to the differ-
ences in incubation times used in the assays. Since com-
pounds access the channel from the inside of the cell,
parameters such as cell penetration can contribute to
the apparent compound affinity in the short incubation
time of the Rb_Kv assay, but they may not be as impor-
tant in the T cell assay, where compounds are present
for longer time.
A double bond was introduced at the carbon next to the
phenyl in 12 to mimic the C20–C29 double bond of cor-
reolide. Lewis acid catalyzed addition of 2-phenylallyl-
silane to enone 4 occurs from the less-hindered face to
afford 26 (Scheme 2), which has the same C18 configura-
tion as the Grignard addition products and correolide.
Compound 26 is 3.5-fold more potent than 12 in the T
Figure 3. Superposition of correolide and 26.

450
J. Bao et al. / Bioorg. Med. Chem. Lett. 15 (2005) 447–451
R2
Bn
N
R1
a, b, c, d
e, f
R2
R2
O
HO2C
O
R1
39
37
O
38
Br
g
R2
R1
H
h
R2
H
H
OH
R1
A
OAc
OAc
MgBr
O
28, 41-47
40
AcO
OAc
Scheme 3. Reagents and conditions: (a) t-BuCOCl, Et3N, À78°C, THF; (b) (R)-4-benzyl-2-oxazolidinone, BuLi, À78°C, THF; (c) NaN(TMS)2,
À78°C, THF; (d) R1I; (e) LAH, rt, THF; (f) CBr4, P P 3h, rt, ether; (g) Mg, ether; (h) 4, 70°C, THF.
Table 3. C18 SAR in Rb_Kv and T cell assays (IC50, nM)
potent blockers of the Kv1.3 channel in both the Rb_Kv
and the T cell proliferation assays. The (R)-Me com-
pound 27 is significantly less potent. These results sug-
R
H
gest that (S)-Me and vinyl groups at the carbon next
to the phenyl enhance the T cell potency. The potency
enhancement of the (S)-Me group is not diminished by
the presence of an (R) substituent (28 vs 29).
H
H
OH
OAc
O
OAc
AcO
OAc
Compd
R
Rb_Kv
26
T cell
460
T cell ratioa
0.4
Analogs containing the two potency enhancing substi-
tution patterns (substitution at the ortho-position of
phenyl and a vinyl group at the carbon next to the
phenyl) were then prepared (Table 4). Hybrid analogs
with lipophilic phenyl substituents are potent inhibitors
of T cell proliferation. The most potent compound in
this series, thioanisole 32, is 12.5-fold more potent than
correolide in the T cell proliferation assay. The two po-
tency enhancing substitution patterns have a synergistic
effect on inhibition of T cell proliferation.
Ph
12
25
26
200
19
—
—
Ph
Ph
131
1.4
Ph
Ph
Ph
27
28
29
120
12
613
102
184
0.3
1.8
1.0
Phenethyl analogs (41–47) with (S)-substitution on the
tether were prepared according to Scheme 3. Substituted
19
a IC50 ratio of correolide to compound.
Table 4. C18 SAR in Rb_Kv and T cell assays (IC50, nM)
R
H
H
H
OH
OAc
O
OAc
AcO
OAc
Compd
R
Rb_Kv
T cell
T cell ratioa
26
30
31
32
33
34
35
36
H
Et
19
44
131
25.5
51.1
14.7
31.7
—
1.4
7.2
3.6
12.5
5.8
—
OMe
SMe
SEt
11
14
17
SOMe
SOMe
SO2Me
145
259
352
Figure 4. Superposition of 26, 27 and 28.
—
—
—
—
the phenyl group and (S)-Me of 28 (Fig. 4).Vinyl analog
26, (S)-Me analog 28 and gem-dimethyl analog 29 are
a IC50 ratio of correolide to compound.

J. Bao et al. / Bioorg. Med. Chem. Lett. 15 (2005) 447–451
451
Table 5. C18 SAR in Rb_Kv and T cell assays (IC50, nM)
Acknowledgements
We would like to thank Dr. Mike Goetz for providing
correolide and Ms. Amy Bernick for MS support.
R2
R1
H
References and notes
H
H
OH
1. Lin, C. S.; Boltz, R. C.; Blake, J. T.; Nguyen, M.; Talento,
A.; Fischer, P. A.; Springer, M. S.; Sigal, N. H.; Slaughter,
R. S.; Garcia, M. L.; Kaczorowski, G. J.; Koo, G. C. J.
Exp. Med. 1993, 177, 637.
2. Leonard, R. J.; Garcia, M. L.; Slaughter, R. S.; Reuben,
J. P. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 10094.
3. Kaczorowski, G. J.; Koo, G. C. Perspect. Drug Discovery
Des. 1994, 2, 233.
4. Lewis, R. S.; Cahalan, M. D. Ann. Rev. Immunol. 1995, 13,
623.
5. Cahalan, M. D.; Chandy, K. G. Curr. Opin. Biotechnol.
1997, 8, 749.
OAc
O
OAc
AcO
OAc
Compd
R1
R2
H
Rb_Kv
T cell
T cell ratioa
28
41
42
43
44
45
46
47
Me
Me
Me
Me
Et
12
19
28
37
28
12
21
48
102
25.6
36.8
12.3
29.2
—
1.8
7.2
5.0
15
OMe
Et
OEt
OMe
Et
6.3
—
Et
Allyl
Allyl
OMe
Et
32.9
—
5.6
—
6. Candy, K. G.; Wulff, H.; Beeton, C.; Pennington, M.;
Gutman, G. A.; Cahalan, M. D. Trends Pharmcol. Sci.
2004, 25, 280.
a IC50 ratio of correolide to compound.
7. Goetz, M. A.; Hensens, O. D.; Zink, D. L.; Borris, R. P.;
Morales, F.; TamayoCastillo, G.; Slaughter, R. S.; Felix,
J.; Ball, R. G. Tetrahedron Lett. 1998, 39, 2895.
8. Felix, J. P.; Bugianesi, R. M.; Schmalhofer, W. A.; Borris,
R. P.; Robert, B.; Goetz, M. A.; Hensens, O. D.; Bao, J.;
Kayser, F.; Parsons, W. H.; Rupprecht, K.; Garcia, M. L.;
Kaczorowski, G. J.; Slaughter, R. S. Biochemistry 1999,
38, 4922.
9. Koo, G. C.; Blake, J. T.; Shah, K.; Staruch, M. J.;
Dumont, F.; Wunderler, D.; Sanchez, M.; McManus, O.
B.; Fischer, P.; Boltz, R. C.; Goetz, M. A.; Baker, R.; Bao,
J.; Kayser, F.; Rupprecht, K. M.; Parsons, W. H.; Tong,
X. C.; Ita, I. E.; Pivnichny, J.; Vincent, S.; Cunningham,
P., Jr.; Hora, D., Jr.; Kaczorowski, G. J.; Springer, M. S.
Cell. Immunol. 1999, 197, 99.
10. Schmalhofer, W. A.; Bao, J.; McManus, O. B.; Green, B.;
Matyskiela, M.; Wunderler, D.; Bugianesi, R. M.; Felix, J.
P.; Hanner, M.; Linde-Arias, A.; Ponte, C. G.; Velasco,
L.; Koo, G.; Staruch, M. J.; Miao, S.; Parsons, W.;
Rupprecht, K.; Slaughter, R. S.; Kaczorowski, G. J.;
Garcia, M. L. Biochemistry 2002, 41, 7781.
11. Miao, S.; Bao, J.; Garcia, M. L.; Goulet, J. L.; Hong, X.;
Kaczorowski, G. J.; Kayser, F.; Koo, G. C.; Kotliar, A.;
Schmalhofer, W. A.; Shah, K.; Sinclair, P. J.; Slaughter,
R. S.; Springer, M. S.; Staruch, M. J.; Tsou, N. N.; Wong,
F.; Parsons, W. H.; Rupprecht, K. M. Bioorg. Med. Chem.
Lett. 2003, 13, 1161.
phenyl acetic acid 37 was attached to (R)-4-benzyl-2-
oxazolidinone via a mixed anhydride. Stereoselective
alkylation of the resulting oxazolidinone provided 38.
Reduction of 38 with LAH and bromination of the
resulting alcohol afforded bromide 39. Treatment of 39
with Mg generated Grignard reagent 40 and addition
of 40 to enone 4 gave compounds 41–47.
Data for these compounds are presented in Table 5.
Increasing the size of the (S)-substituent had little effect
on activity in either assay. Ethoxy analog 43 is the most
potent of this series, with activity 15-fold greater than
correolide in the T cell proliferation assay. As in the
vinyl series, combination of the two potency enhancing
substitution patterns (substitution at the ortho-position
of phenyl and an alkyl group at the carbon next to the
phenyl) has a synergistic effect on inhibition of T cell
proliferation.
In summary, several series of tetracyclic analogs have
been synthesized. These compounds are structurally
simpler than correolide. Two substitution patterns that
enhance the compoundsÕ inhibition of T cell prolifera-
tion were identified. A synergistic effect in potency was
found in compounds that incorporate these substitution
patterns. The most potent compound 43 is as much as
15-fold more potent than correolide as inhibitor of
human T cell proliferation.
12. The T cell proliferation assay was performed with human
peripheral blood T cells stimulated with anti-CD3 anti-
body (Research Diagnostics, Flanders, NJ).
13. Bao, J.; Baker, R. K.; Doss, G. A.; Kayser, F.; Kotliar, A.;
Miao, S.; Parsons, W. H.; Rupprecht, K. M. Org. Lett.
2002, 4, 1871.