A three-residue, continuous binding epitope peptidomimetic of ShK toxin as a Kv1.3 inhibitor

Article abstract of DOI:10.1016/j.bmcl.2005.05.014

The ShK toxin is a polypeptide that blocks the Kv1.3 potassium channel in T-lymphocytes and has been identified as a potential therapeutic for multiple sclerosis. ShK is well characterised in terms of structure and binding, offering an attractive target for the design of structural and functional mimetics. Building on our previous success in developing rationally designed peptidomimetics of ShK, we report a novel mimetic of the K22-Y23-R24 residues of the peptide. The mimetic was shown to inhibit the Kv1.3 channel with moderate activity.

Full text of DOI:10.1016/j.bmcl.2005.05.014

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3193–3196
A three-residue, continuous binding epitope peptidomimetic
of ShK toxin as a Kv1.3 inhibitor
Andrew J. Harvey,a Robert W. Gableb and Jonathan B. Baella,*
aThe Walter and Eliza Hall, Institute of Medical Research, Biotechnology Centre, 4 Research Avenue,
La Trobe R&D Park, Bundoora 3086, Australia
bSchool of Chemistry, University of Melbourne, Vic. 3010, Australia
Received 14 April 2005; revised 27 April 2005; accepted 3 May 2005
Abstract—The ShK toxin is a polypeptide that blocks the Kv1.3 potassium channel in T-lymphocytes and has been identified as a
potential therapeutic for multiple sclerosis. ShK is well characterised in terms of structure and binding, offering an attractive target
for the design of structural and functional mimetics. Building on our previous success in developing rationally designed peptido-
mimetics of ShK, we report a novel mimetic of the K22–Y23–R24 residues of the peptide. The mimetic was shown to inhibit the
Kv1.3 channel with moderate activity.
Ó 2005 Elsevier Ltd. All rights reserved.
Although the voltage-gated potassium ion channel,
Kv1.3, occurs as a monomer throughout the body, the
potential therapeutic value of Kv1.3 arises from its pres-
ence as a homotetramer in selected cell types, most nota-
bly in T-cells.1 Kv1.3 is involved in setting the
membrane potential in T-cells. Blockade of Kv1.3
causes depolarisation across the membrane, disabling
the cellꢀs ability to increase intracellular calcium levels
to a point sufficient for proliferation.2,3 Thus, blockade
of the Kv1.3 ion channel has been identified as a poten-
tial means for immunosuppression.4
derivatives of khellinone and molecules based around
a psoralen core.7–11
The ShK peptide is a component of the venom of the
sun anemone (Stichodactyla helianthus) that blocks the
Kv1.3 ion channel in the picomolar range.12 Being a
peptide, ShK itself has limited promise as a therapeutic.
ShK and its synthetic analogue, ShK-Dap22, have lim-
ited oral bioavailability and have short half-lives within
the body,13 although ongoing efforts to modify ShK aim
to improve its pharmacokinetics and are meeting with
some success.14 The structure of ShK has been deter-
mined by NMR and its binding interactions with
Kv1.3 have been mapped. These data, together, allow
for the design of peptidomimetics that have more
drug-like properties than the peptide. We have an inter-
est in the design of binding-epitope mimetics,15–17 and
have previously reported a discontinuous binding epi-
tope mimetic of ShK.18 Here, we report a novel, contin-
uous binding epitope mimetic of ShK.
Recent findings show that myelin-reactive T-cells from
multiple sclerosis patients express extremely high levels
of the Kv1.3 channel.5 Such T-cells do not proliferate
when treated with specific blockers of Kv1.3. These find-
ings were applied to the treatment of autoimmune dis-
ease by showing that ShK-Dap22, a selective Kv1.3
blocker, ameliorates the symptoms of adoptive experi-
mental autoimmune encephalomyelitis (EAE), an ani-
mal model for multiple sclerosis.6 Links between
Kv1.3 blockade and immunosuppressive activity have
also been established for small molecule blockers of
Kv1.3 including natural product correolide, various
Information about the binding interactions between
Kv1.3 and ShK has been provided by two studies.19,20
Alanine scanning from the latter study (Table 1) re-
vealed that an elliptical ꢁpatchꢀ comprised the binding
surface and a series of thermodynamic cycles showed
that this patch sat deep in the extracellular vestibule of
Kv1.3. The pharmacophore comprises both discontinu-
ous (R11 and F27) and continuous (K22, Y23 and R24)
Keywords: Peptidomimetic; Indole; Kv1.3; Ion channel; Multiple
sclerosis.
*
Corresponding author. Tel.: +61 3 93452108; fax: +61 3 93452211;
e-mail: jbaell@wehi.edu.au
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.014

3194
A. J. Harvey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3193–3196
Table 1. Approximate decrease in free energy of binding, DDGobs (kJ/
mol), to Kv1.3 of single residue substituted analogues of ShK relative
to wild type ShK20
ShK analogue
DDGobs (kJ/mol)
R11A
K22A
Y23A
R24A
F27A
4
8
8
11
3
binding elements. We have previously disclosed a ration-
ally designed R11–K22–Y23 ShK mimetic, which binds
to Kv1.3 with an affinity/Kd of 95 lM.18 In the work dis-
cussed herein, we target K22, Y23 and R24. Residues
K22 and Y23 have been retained as targets for the
new mimetic since this diad appears as a common ele-
ment across Kv binding peptide toxins and consistently
rates as an important binding feature to Kv1.3.1 We
chose R24 as the third residue to target because, as
shown in Table 1, it might interact more strongly with
Kv1.3 than does R11 and could thus give rise to a more
potent mimetic. Also, the contiguous nature of the sur-
face of the K22–Y23–R24 binding epitope would allow
for a smaller mimetic that would correspondingly be
more drug-like than our previously reported R11–
K22–Y23 mimetic. Furthermore, choice of a continuous
epitope is attractive because the peptide backbone can
aid in the design of the peptidomimetic scaffold.
In designing the peptidomimetic, the a–b bond vectors
of the K22 and Y23 side chains in ShK were set as
constraints. These vectors were chosen so that the
peptidomimetic side chains would have at least the same
degree of rotational freedom as in the peptide. Various
molecular scaffolds were trailed during the interactive
design process to meet the K22 and Y23 bond vector
constraints.21 Scaffold requirements included ease of
synthesis, versatility and drug-likeness. An N-alkylated
indole-7-carboxamido scaffold appeared to satisfy the
criteria. Furthermore, such an indole suitably deriva-
tised at the 4-position would conveniently reach into
the space sampled by R24 in ShK. The indole 1 is an
example of the type of compound afforded by this pro-
cess (Fig. 1). The methyleneguanidinyl group attached
to C4 mimics R24 and the aminobutyl group mimics
the K22 residue. It has been shown that Y23F analogues
of ShK have Kv1.3 activity in the same order of magni-
tude as wild type ShK.19 Correspondingly, we have
found that the phenolic OH of Y23 in ShK mimetics
is not necessary for activity.18 Thus for ease of synthesis,
a phenethyl substituent on the indole nitrogen was used
to mimic the side chain of ShK Y23F rather than ShK
itself.
Figure 1. (a) The design rationale for 1: a superimposition of the active
residues of ShK with the minimised conformation of the proposed
peptidomimetic. The K22–Y23–R24 residues of ShK, taken from the
NMR structure of the peptide, are shown with cyan on the backbone
and orange on the Lys and Tyr side chains. (b) The indole
peptidomimetic labelled with the ShK side chains to be mimicked.
The synthesis is premised on the preparation of a sym-
metrically 4,7-disubstituted indole 6 and the subsequent
differentiation of the equivalent functional groups. The
key intermediate 6 was synthesised according to the
methodology of Jones et al. wherein a 4,7-disubstituted
indole is prepared from two successive Diels–Alder reac-
tions of a vinylpyrrole with two equivalents of a dieno-
phile followed by a retro-Diels–Alder reaction.22 Thus,
Scheme 1. Synthesis of intermediate 6: (a) PhCH2CH2Br, tetrabutyl-
ammonium bromide, Cs2CO3, DMF, 80 °C, 9 h, 76%. (b) Ph3PCH3Br,
NaH, THF, 70 °C, 2 h, 93%. (c) Methyl propiolate, hydroquinone,
PhMe, dark, 80 °C, 66 h, 68%. (d) LiOH, MeOH/H2O, rt, 16 h. (e)
BH3 Æ THF, THF, rt, 4 h. (f) PPh3, DIAD, DPPA, THF, rt, 16 h, 25%
over three steps.

A. J. Harvey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3193–3196
3195
BocHN
HN
NBoc
N
3
N
3
a
b
c,d
e
N
N
6
Ph
Ph
N
1
Ph
HO
O
HN
O
HN
O
7
8
9
NHBoc
NHBoc
Scheme 2. Synthesis of peptidomimetic 1: (a) LiOH, MeOH/H2O, 90 °C, 5 h, 84%. (b) NH2(CH2)4NHBoc, CDI, THF, rt, 16 h, 37%. (c) PPh3, H2O,
THF, rt, 16 h, 76%. (d) CF3SO2N@C(NHBoc)2, NEt3, DCM, rt, 24 h, 32%. (e) Trifluoroacetic acid, DCM, rt, 30 min, 90%.
pyrrole-2-carboxaldehyde 2 was N-alkylated with phen-
ethylbromide to give 3 in 76% yield by the returned
starting material (Scheme 1).23 The aldehyde underwent
a Wittig reaction to furnish the vinyl pyrrole 4. The reac-
tion of 4 with two equivalents of methyl propiolate in
the presence of hydroquinone gave intermediate 5.
The methyl ester attached to C4 was preferentially
hydrolysed with lithium hydroxide. The resulting mix-
ture was treated with borane tetrahydrofuran complex
1
to give a mixture of regioisomeric alcohols (6:1 by H
NMR) comprising predominantly the desired alcohol.
Mitsunobu conditions (PPh3, diphenylphosphorylazide
(DPPA), diisopropylazodicarboxylate (DIAD)) were
employed for the generation of the azide 6, which was
isolated pure after chromatography in 25% yield over
three steps.
The relatively hindered second methyl ester was cleaved
with heating in the presence of lithium hydroxide giving
an acid, which was then coupled using 1,10-carbonyl-
Figure 2. Overlay of the crystal structure of 7 with the K22-Y23-R24
residues from ShK.
diimidazole (CDI) with N-Boc-1,4-diaminobutane to af-
ford the amide 7 (Scheme 2). The azide was reduced
using Staudinger conditions and the resulting amine
was treated with a guanidinylating agent to give 13.24
Deprotection of 13 gave the target compound 1 as a
trifluoroacetate salt.
ing murine Kv1.3.10 The peptidomimetic gave an EC50
of 75 lM based on preliminary data. This promising re-
sult is on par with our previously reported R11–K22–
Y23 mimetic (EC50 = 95 lM),18 further confirming the
validity of this approach to it de novo designed three-
point peptidomimetics based on side-chain bond con-
straints. In the future, using chemistry amenable to mul-
tiple parallel synthesis, such as the elegant methodology
of Li et al.,25 we aim to build a library around the indole
scaffold to improve on the current activity.
A crystal structure of 7 was obtained to support the
structure arrived at by molecular modelling.ꢂ Com-
pound 7 is a 4,7-disubstituted N-phenethylindole and
as such comprises the essential elements of the pepti-
domimetic scaffold. The crystal structure of 7 overlaid
with the K22–Y23–R24 residues from the NMR struc-
ture of ShK shows high atom coincidence (see Fig. 2).
The level of Kv1.3 blockade by 1 was determined by
patch-clamp methods on L929 cells, which were express-
Acknowledgments
The authors thank Assoc. Prof. Heike Wulff for per-
forming the electrophysiological experiments.
ꢂ Crystallographic data (excluding structure factors) for the structures
in this letter have been deposited with the Cambridge Crystallo-
graphic Data Centre as supplementary publication number CCDC
266815. Copies of the data can be obtained free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
[fax: +44 (0)1223 336033 or e-mail: deposit@ccdc.cam.ac.uk].
References and notes
1. Kem, W. R.; Pennington, M. W.; Norton, R. S. Perspect.
Drug Discovery Des. 1999, 16, 111.
2. Cahalan, M. D.; Chandy, K. G. Curr. Opin. Biotechnol.
1997, 8, 749.

3196
A. J. Harvey et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3193–3196
3. Chandy, K. G.; Cahalan, M.; Pennington, M.; Norton, R.
S.; Wulff, H.; Gutman, G. A. Toxicon 2001, 39, 1269.
13. Norton, R. S.; Pennington, M. W.; Wulff, H. Curr. Med.
Chem. 2004, 11, 3041.
4. Chandy, K. G.; Wulff, H.; Beeton, C.; Pennington, M.;
Gutman, G. A.; Cahalan, M. D. Trends Pharmacol. Sci.
2004, 25, 280.
5. Wulff, H.; Calabresi, P. A.; Allie, R.; Yun, S.; Pennington,
M.; Beeton, C.; Chandy, K. G. J. Clin. Invest. 2003, 111,
1703.
6. Beeton, C.; Wulff, H.; Barbaria, J.; Clot-Faybesse, O.;
Pennington, M.; Bernard, D.; Cahalan, M. D.; Chandy,
K. G.; Beraud, E. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
13942.
7. Felix, J. P.; Bugianesi, R. M.; Schmalhofer, W. A.;
Borris, R.; Goetz, M. A.; Hensens, O. D.; Bao, J. M.;
Kayser, F.; Parsons, W. H.; Rupprecht, K.; Garcia, M.
L.; Kaczorowski, G. J.; Slaughter, R. S. Biochemistry
1999, 38, 4922.
14. Beeton, C.; Pennington, M. W.; Wulff, H.; Singh, S.;
Nugent, D.; Crossley, G.; Khaytin, I.; Calabresi, P. A.;
Chen, C.; Gutman, G. A.; Chandy, K. G. Mol. Pharmacol.
2005, 67, 1369.
15. Baell, J. B.; Forsyth, S. A.; Gable, R. W.; Norton, R. S.;
Mulder, R. J. J. Comput. Aid. Mol. Des. 2001, 15, 1119.
16. Baell, J. B.; Duggan, P. J.; Forsyth, S. A.; Lewis, R. J.;
Lok, Y. P.; Schroeder, C. I. Bioorg. Med. Chem. 2004, 12,
4025.
17. Baell, J. B.; Duggan, P. J.; Lok, Y. P. Aust. J. Chem. 2004,
57, 179.
18. Baell, J. B.; Harvey, A. J.; Norton, R. S. J. Comput. Aid.
Mol. Des. 2002, 16, 245.
19. Pennington, M. W.; Mahnir, V. M.; Krafte, D. S.; Zayden-
berg, I.; Byrnes, M. E.; Khaytin, I.; Crowley, K.; Kem, W.
R. Biochem. Biophys. Res. Commun. 1996, 219, 696.
20. Rauer, H.; Pennington, M.; Cahalan, M.; Chandy, K. G.
J. Biol. Chem. 1999, 274, 21885.
21. Sybyl 7.0 Tripos Associates, Missouri. Software package
was used both for graphics and minimization Tripos
forcefield, default minimization and convergence settings,
with Gasteiger–Huckel atomic charges and distance-
dependent dielectric.
8. Bao, J. M.; Miao, S. W.; Kayser, F.; Kotliar, A. J.; Baker,
R. K.; Dossa, G. A.; Felix, J. P.; Bugianesi, R. M.;
Slaughter, R. S.; Kaczorowski, G. J.; Garcia, M. L.; Ha,
S. N.; Castonguay, L.; Koo, G. C.; Shah, K.; Springer, M.
S.; Staruch, M. J.; Parsons, W. H.; Rupprecht, K. M.
Bioorg. Med. Chem. Lett. 2005, 15, 447.
9. Coghlan, M. J.; Carroll, W. A.; Gopalakrishnan, M.
J. Med. Chem. 2001, 44, 1627.
10. Baell, J. B.; Gable, R. W.; Harvey, A. J.; Toovey, N.;
Herzog, T.; Hansel, W.; Wulff, H. J. Med. Chem. 2004, 47,
2326.
11. Vennekamp, J.; Wulff, H.; Beeton, C.; Calabresi, P. A.;
Grissmer, S.; Hansel, W.; Chandy, K. G. Mol. Pharmacol.
2004, 65, 1364.
22. Jones, R. A.; Marriott, M. T. P.; Rosenthal, W. P.;
Sepulvedaarques, J. J. Org. Chem. 1980, 45, 4515.
23. Attempted alkylation of pyrrole with phenethylbromide in
the presence of various bases afforded mixtures of the
desired N-alkylated pyrrole and styrene.
24. Feichtinger, K.; Zapf, C.; Sings, H. L.; Goodman, M.
J. Org. Chem. 1998, 63, 3804.
25. Li, L. H.; Martins, A. Tetrahedron Lett. 2003, 44, 5987.
12. Tudor, J. E.; Pallaghy, P. K.; Pennington, M. W.; Norton,
R. S. Nat. Struct. Biol. 1996, 3, 317.