A new class of blockers of the voltage-gated potassium channel Kv1.3 via modification of the 4- or 7-position of khellinone

Article abstract of DOI:10.1021/jm050839v

The voltage-gated potassium channel Kv1.3 constitutes an attractive target for the selective suppression of effector memory T cells in autoimmune diseases. We have previously reported the natural product khellinone, 1a, as a versatile lead molecule and id

Full text of DOI:10.1021/jm050839v

J. Med. Chem. 2006, 49, 1433-1441
1433
A New Class of Blockers of the Voltage-Gated Potassium Channel Kv1.3 via Modification of the
4- or 7-Position of Khellinone
Andrew J. Harvey,*,† Jonathan B. Baell,*,† Nathan Toovey,† Daniel Homerick,‡ and Heike Wulff*,‡
The Walter and Eliza Hall Institute of Medical Research Biotechnology Centre, 4 Research AVenue, La Trobe R&D Park, Bundoora 3086,
Australia, and Department of Medical Pharmacology and Toxicology, UniVersity of California, DaVis, California 95616
ReceiVed August 24, 2005
The voltage-gated potassium channel Kv1.3 constitutes an attractive target for the selective suppression of
effector memory T cells in autoimmune diseases. We have previously reported the natural product khellinone,
1a, as a versatile lead molecule and identified two new classes of Kv1.3 blockers: (i) chalcone derivatives
of khellinone, and (ii) khellinone dimers linked through the 6-position. Here we describe the multiple parallel
synthesis of a new class of khellinone derivatives selectively alkylated at either the 4- or 7-position via the
phenolic OH and show that several chloro, bromo, methoxy, and nitro substituted benzyl derivatives inhibit
Kv1.3 with submicromolar potencies. Representative examples of the most potent compounds from each
subclass, 11m (5-acetyl-4-(4′-chloro)benzyloxy-6-hydroxy-7-methoxybenzofuran) and 14m (5-acetyl-7-(4′-
bromo)benzyloxy-6-hydroxy-4-methoxybenzofuran), block Kv1.3 with EC50 values of 480 and 400 nM,
respectively. Both compounds exhibit moderate selectivity over other Kv1-family channels and HERG, are
not cytotoxic, and suppress human T cell proliferation at low micromolar concentrations.
Introduction
thermore, recent studies with autoreactive T cells from MS
patients showed that these cells expressed much higher levels
of Kv1.3 than T cells from controls, and that their proliferation
was inhibited by selective Kv1.3 blockade with ShK toxin under
conditions that did not affect normal peripheral blood T cells.
A subsequent study on postmortem brain sections from MS
patients identified Kv1.3high T cells in active MS plaques,13
demonstrating that Kv1.3high TEM cells are indeed actively
involved in the disease process of MS.
Human T cells express two types of potassium channels, the
voltage-gated potassium channel Kv1.3 and the calcium-
activated potassium channel KCa3.1 (also known as IKCa1,
SK4), both of which are critically involved in membrane
potential regulation and calcium signaling during T cell
activation.1-4 Engagement of the T-cell receptor by an antigen-
presenting cell during an immune response initiates a calcium
flux into the T cell that raises the cytosolic calcium concentration
into the micromolar range and that ultimately results in cytokine
secretion and T-cell proliferation.3,5 This crucial calcium influx
is only possible if the T cell can keep its membrane potential
negative through a counterbalancing potassium efflux through
Kv1.3 and/or KCa3.1.1-4 Blockade of these two potassium
channels depolarizes the T cell membrane and thus inhibits
calcium signaling and consequently prevents T-cell activation.1,4-6
While it was initially supposed that lymphocyte potassium
channel blockade would lead to a general immunosuppression
similar to the calcineurin inhibitor cyclosporin, it has recently
become apparent that Kv1.3 and KCa3.1 are differentially
expressed in different T and B cell subsets and that Kv1.3
blockers preferentially affect terminally differentiated effector
memory T (TEM) cells.7,8 Kv1.3 blockade has therefore been
proposed as a new therapeutic approach for the treatment of
autoimmune diseases such as multiple sclerosis (MS), type-1
diabetes, rheumatoid arthritis, psoriasis, and chronic-graft-
versus-host diseases where TEM cells are involved in the
pathogenesis.4,9 The best evidence that a Kv1.3-based approach
will be effective currently exists for the case of MS: the Kv1.3-
blocking peptides kaliotoxin, ShK, and its more selective
derivative ShK(L5) prevent and treat experimental autoimmune
encephalomyelitis in rats, an animal model of MS.10-12 Fur-
While peptide toxins, most notably the toxin ShK, derived
from the anemone Stichodactyla helianthus, and its various
analogues, were valuable tools for elucidating the structure and
function of Kv1.3,14 they have the disadvantage of requiring
parenteral administration when used as drugs. A number of
groups in academia and industry4,15 have therefore actively
searched for small molecule Kv1.3 blockers since Kv1.3 was
cloned in 199316 and since cell lines stably expressing the
channel have become available. Peptidomimetics of the active
residues of ShK exhibited moderate levels of Kv1.3 blockade
(EC50 75-95 µM in whole-cell patch-clamp) and are noteworthy
as examples of the few successful de novo designed mimet-
ics.17,18 A more potent small molecule Kv1.3 blocker is the
natural product correolide (Figure 1A), which inhibited Rb+
efflux through the Kv1.3 channel with an EC50 of 86 nM19 and
blocked the channel with an EC50 of 110 nM in patch-clamp
experiments.20 In proof of correolide’s immunosuppressive
potential, scientists at Merck showed that it inhibited T-cell
proliferation and attenuated delayed-type hypersensitivity in
miniswine.21 However, development of correolide as a thera-
peutic was hampered by its lack of selectivity over other Kv1-
family channels22 and its molecular complexity that precluded
a cost-effective total synthesis.15,23 Other more recent examples
of small molecule Kv1.3 blockers include the cyclohexyl-
substituted benzamides23 (Figure 1A), which were identified in
a high-throughput screen and which exhibited EC50 values as
low as 50 nM, and the 5-phenylalkoxypsoralen Psora-4 (EC50
3 nM, Figure 1A), which was developed out of the natural
product 5-methoxypsoralen and which currently constitutes the
most potent small molecule Kv1.3 blocker.24
* To whom correspondence should be addressed. (J.B.) Phone: +61 3
9345 2108 Fax: +61 3 9345 2211. E-mail: jbaell@wehi.edu.au. (H.W.)
Phone: 530 754 6135 Fax: 530 752 7710. E-mail: hwulff@ucdavis.edu.
(A.H.) Phone: +61 3 9345 2103. Fax: +61 3 9345 2211. E-mail:
aharvey@wehi.edu.au.
† The Walter and Eliza Hall Institute of Medical Research Biotechnology
Centre.
‡ University of California.
10.1021/jm050839v CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/28/2006

1434 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
HarVey et al.
Figure 1. Figure 1. (A) Structures of correolide, a representative cyclohexyl benzamide and the 5-phenylalokoxypsoralen Psora-4. (B) Structures
of khellinone 1a and khellin 1b. The 7-position in khellinone corresponds to the 9-position in khellin as indicated in Table 1.
Table 1. Synthesis of Monoalkylated Khellinone Analogues and Their
In a previous report, we identified the natural product
khellinone 1a (Figure 1B) as a versatile lead molecule for the
development of selective Kv1.3 blockers.25 While khellinone
itself inhibited Kv1.3 with an EC50 of 45 µM, its facile
conversion to chalcones or to 6-linked dimers afforded respec-
tive jumps in activity to 0.3 and 0.7 µM. The khellinone
chalcones in particular were seen as an interesting class because
they are the only small molecule Kv1.3 blockers to show
selectivity over both the Kv1.2 and Kv1.5 channels. Khellinone
offers more room for development beyond the existing series.
Here we describe further efforts to derivatize this versatile
starting material, this time with a view to developing structure-
activity relationships around the 4- and 7-positions.
Blockade of Kv1.3a
Chemistry
R1 (7)
R2 (4)
Me
Me
Me
Me
Me
Me
Et
compoundc
EC50 (µM)
We report here a versatile approach to the preparation of 4-
and/or 7-alkoxy analogues of khellinone. Limited derivatization
at these positions in khellinone and the related khellin has been
previously reported in the literature. An early study from 1952
described the formation of 4-alkoxy khellins by the selective
demethylation of khellin with acid and subsequent alkylation.26
A restricted number of compounds with C1-C4 alkoxy sub-
stituents in the 4-position in combination with any alkoxy
substitution at the 7-position were afforded in a six-step
synthesis of khellinone derivatives from pyrogallol.27 Other
workers bis-demethylated khellin to dinorkhellin either in a
single step using magnesium iodide in ether,28 or in two steps
through the khellin quinone,29,30 followed by alkylation to give
4- or 9-alkoxy-substituted khellins.3 The methodologies to obtain
dinorkhellin were either low yielding or unnecessarily time-
consuming. We improved the synthesis of dinorkhellin and
developed a route to selectively alkylate this diphenol to develop
a general and reliable approach for the introduction of diversity
at the 4- and 7-positions of khellinone. Once the route was in
hand we tailored the methodology for multiple parallel synthesis
(MPS) application in order to synthesize a small library of
asymmetrically substituted khellinone derivatives.
Me
Et
Pr
Pn
Bn
2-Npm*
Me
Me
Me
Me
Me
Me
1a
5a
5b
5c
5d
5e
5f
5g
5h
5i
45
10
8
3
1.2
5
7
5
4
2.5
8
Pr
Pn
Bn
1-Npmb
2-Npmb
5j
5k
5
a Reagents: (i) BBr3, DCM, -78 °C; (ii) Me2SO4 or RX, K2CO3, DMF;
(iii) RX, Cs2CO3, DMF; (iv) NaOH, H2O, EtOH. b Npm ) naphthylmethyl.
c Each compound was tested 2-3 times at three concentrations. Values for
the EC50 were determined by fitting the Hill equation to the reduction of
area under the K+ current curve. Standard deviations were between 5 and
15%.
hydrogen-bonded 4-hydroxyl group.29 Cesium carbonate in
dimethylformamide provided sufficiently strong conditions for
alkylation at the 4-position. Finally, the khellin analogues 4a-k
were subjected to alkaline hydrolysis to afford the khellinone
analogues 5a-k. The order of alkylation was verified using
NOE NMR experiments on 5f, showing through-space proximity
between the 4-ethoxy group and the acetyl group (see Supporting
Information).
The synthesis of the 4-alkoxy and 7-alkoxy khellinones was
adapted to MPS by performing the reactions on a 24-block MPS
reactor and simplifying the workup of the alkylation reactions
to a protocol of acidification, filtration, and desiccation. By this
As a first step khellin (1b) was treated with boron tribromide
to afford the diphenol (see Table 1). Selective alkylation of 2
with 1 equiv of alkylating agent in the presence of potassium
carbonate gave the 9-alkyl-4-hydroxykhellin analogues 3a-f.
Selectivity at the 9-position was obtained because the pKa of
the free phenol is lower than the pKa of the intramolecularly

Potassium Channel KV1.3 Blockers
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1435
Scheme 1. Synthesis of the Chalcone 6 and the Dimer 8 of the 4-Ethyl Analogue 5fa
a Reagents: (i) 3-hydroxybenzaldehyde, NaOH, H2O; (ii) 1,6-dibromohexane, K2CO3, DMF. *The synthesis of 7 and 9 is described elsewhere.25
Table 2. Synthesis and Channel Blocking Activity Kv1.3 of Substituted
method we generated a library of compounds and investigated
the structure activity relationships around the phenyl ring on
the 4- or 7-positions (Table 1) by whole-cell patch-clamp.
The khellinone analogue 5f was subjected to further modi-
fications. From it chalcone 6 was readily prepared in 67% yield
by Claisen-Schmidt condensation and the dimer 8 in 90% yield
by alkylation of the phenolic OH with a bis-haloalkane (Scheme
1). The six-carbon alkyl linker was chosen for the dimer because
it was shown that this length was optimal for activity against
Kv1.3 in the previously reported khellinone dimer series.25,31
4-benzyl Ethersa
Results and Discussion
compdb
R
EC50 (µM) compdb
R
EC50 (µM)
SAR of 4- and 7-Substituted Khellinones on Kv1.3. We
screened all newly synthesized compounds by manual whole-
cell patch-clamp on L929 cells stably expressing Kv1.3. We
chose this technique instead of a faster Rb+-flux or binding
assay32 despite its low throughput and its extreme labor-intensity,
because whole-cell patch-clamp is considered the “gold-
standard” for testing ion channel modulators33 and allows to
study channel inhibition under physiological conditions. Rb+
flux assays normally work with high concentrations of extra-
cellular potassium. This prevents the inactivation of Kv1.334
and leads to an underestimation of the potency of compounds
that exhibit a use- and/or state-dependent block. Binding assays
have the disadvantage that they only detect whether a compound
binds to the Kv1.3 protein and do not provide information on
channel function.
Table 1 lists the effects that varying the 7-substituent has on
Kv1.3 blocking activity (5a-e). The potency improved mark-
edly with any alkyl group larger than methyl, giving rise to
around a 10-20-fold increase depending on the length of the
alkyl chain. The benzyl ether 5d and the pentyl ether 5c were
the most effective blockers of Kv1.3 in this series with EC50
values of 1.2 µM and 3 µM, respectively. Interestingly, the
activity trend observed in the analogues where the 4-substituent
has been changed (5f-k) mirrored the trend for the 7-substituted
compounds (5a-e).
Since the benzyl-substituted compounds 5d and 5i proved to
be the most potent compounds of the 4- and 7-substituted
khellinone series, and because many benzyl halides are com-
mercially available, we selected 5d and 5i as templates to test
the effect that varying the physicochemical properties of
substituents has on channel blocking activity. As shown in Table
2, a variety of ortho and meta substituents did not produce any
significant changes in potency relative to the unsubstituted
benzyl ether 5i with the exception of the 3-methoxy- and the
3-trifluoromethoxy-substituted 11k and 11j, which were ap-
proximately 4-fold more potent. A clear outlier here is the
3-carboxy 11h, which is not measurably active. Since the
carboxy group is negatively charged at physiological pH,
inactivity could be either due to poor membrane permeability
resulting in little or no compound reaching the binding site, or
intolerance of such a polar group in the binding site, or a
combination of these two effects.
11a
11b
11c
11d
11e
11f
11g
11h
11i
2-F
2-Cl
2-Me
2-Ph
3-F
1.7
1.2
2.0
2.0
1.2
1.4
1.0
11m
11n
11o
11p
11q
11r
11s
4-Cl
4-Br
0.48
0.50
0.51
0.70
1.5
0.90
8.8
0.50
no effectc
1.8
4-CF3
4-iPr
4-tBu
3-Cl
4-Ph
3-Me
3-COOH
3-NO2
3-OCF3
3-OMe
4-F
4-OBn
4-COPh
4-COOH
4-OCF3
4-OMe
>10
11t
1.4
11u
11v
11w
11j
11k
11l
0.55
0.68
1.0
1.5
a Reagents: (i) RX, Cs2CO3, DMF; (ii) NaOH, H2O, EtOH. Compounds
11a to 11w were prepared as described in General Procedure E (Supporting
Information). b Each compound was tested two times at 4-5 concentrations.
Values for the EC50 were determined by fitting the Hill equation to the
reduction of area under the K+ current curve. Standard deviations were
between 5 and 15%. The Hill coefficient was 1.2 for all compounds. c Tested
at 10 µM.
The para position could also accommodate a variety of
different substituents and yielded several compounds that were
2-5 times more potent than the unsubstituted benzyl ether 5i,
the most potent of which included 11m with a 4-chloro
substituent, 11n with a 4-bromo substituent, and 11o with a
4-trifluoromethyl. Each of these compounds blocked Kv1.3 with
EC50 values of around 0.5 µM. Bulky groups in 11q (4-tBu),
11r (4-Ph), and 11t (4-COPh) could all be tolerated but the
longest and most bulky group of 4-OBn in 11s gave rise to a
noticeable drop in activity (EC50 8.8 µM), suggesting that there
is a limit to the size of the para substituent that can be
accommodated. The inactivity of the p-carboxy compound 11u
could arise for the same reasons as those for the inactive
m-carboxy compound 11h discussed above.
Table 3 lists the effects that varying the substitution on the
7-benzyl ether has on Kv1.3-blocking activity. A number of
different substituents could be tolerated at the ortho, meta, and
para positions and the para position could also accommodate
bulky groups in 14q and 14s but less so the bulky 4-OBn group
of 14r. The most active compounds in this series bore the same
substitution as the most active analogues from the 7-benzyl ether
series, including the p-chloro derivative 14l, the para-bromo
derivative 14m, and the p-trifluoromethyl 14n, each with an
EC50 close to 0.4 µM.

1436 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
HarVey et al.
Table 3. Synthesis and Channel Blocking Activity Kv1.3 of Substituted
Table 4. Dialkylated Khellinone Analoguesa
7-Benzyl Ethersa
compound
EC50 (µM)
16a
16b
6.0a
1.2b
compdb
R
EC50 (µM) compdb
R
EC50 (µM)
a Hill coefficient of 2. b Hill coefficient of 1.
14a
14b
14c
14d
14e
14f
14g
14h
14i
2-F
2-Cl
2-Me
2-Ph
3-F
2.5
0.7
1.9
1.1
0.75
0.60
1.2
14l
4-Cl
4-Br
0.42
0.40
0.42
1.0
5.8
1.2
14m
14n
14o
14p
14q
14r
14s
14t
Table 5. Kv1.3 Blockade by Chalcones and Dimers
4-CF3
4-iPr
compound
EC50 (µM)
4-tBu
6
2.5
0.7
0.8
0.7
3-Cl
4-Ph
7a
8
3-Me
3-COOH
3-NO2
3-OCF3
4-F
4-OBn
4-COPh
4-COOH
4-OCF3
6.3
no effectc
0.60
0.45
0.70
0.50
no effectc
0.60
9a
14j
14k
14u
a Compounds 7 and 9 were reported previously.25
a Reagents: (i) RX, K2CO3, DMF; (ii) MeI, Cs2CO3, DMF; (iii) NaOH,
H2O, EtOH. Compounds 14a to 14u were prepared as described in General
Procedure F (Supporting Information). b Each compound was tested two
times at 4-5 concentrations. Values for the EC50 were determined by fitting
the Hill equation to the reduction of area under the K+ current curve.
Standard deviations were between 5 and 15%. The Hill coefficient was 1.8
for all compounds. c Tested at 10 µM.
be expected to increase potency. However the lack of change
in potency does not negate the flip/invert theory.
Our previous report showed that chalcone derivatives of
khellinone gave marked increases in activity relative to khelli-
none.25 The 3′-hydroxychalcone 7 (Scheme 1), for example,
blocked Kv1.3 with an EC50 of 0.7 µM, corresponding to a 60-
70-fold increase over khellinone (EC50 of 45 µM). The chalcone
6 was therefore synthesized from 5f, which was already more
than 6-fold more potent than khellinone with an EC50 of 7 µM,
and tested against Kv1.3 to determine whether the sharp gains
in activity conferred by chalcones of khellinone would map onto
the alkoxykhellinone analogues. However, with an EC50 of 2.5
µM the chalcone 6 only showed a 3-fold increase in activity
over its parent 5f (Table 5). Similarly, we have previously
reported that khellinone dimers such as 9 are more than 50-
fold more potent than khellinone itself in blocking Kv1.3.
However, dimerization of the more potent khellinone analogue
5f resulted in dimer 8 (Table 5) that exhibited a more modest
10-fold increase in potency relative to parent molecule 5f. These
results together suggest that these separated activity-enhancing
modifications to khellinone, that is, increasing the size of the
4-alkyl ether combined with chalcone formation or dimerization,
are not necessarily additive when combined in one molecule.
In summary, khellinone analogues with any ether larger than
methyl in the 4- and 7- positions showed markedly increased
activity over khellinone. From a library of monosubstituted
benzyl ethers at the 4- or 7-positions, optimal potency of
between 0.4 and 0.5 µM was observed for a number of
examples. One compound from each subclass, namely the
p-chlorobenzyl ether in the 4-position (11m) and p-bromobenzyl
ether in the 7-position (14m), was selected for further testing
on their mechanism of block, selectivity, and antiproliferative
function.
Figure 2. Pseudo-symmetry in the khellinone analogues.
Generally, we found that the SAR of the 7-benzyl ethers
paralleled the SAR for the 4-substituents, not just for the ortho-,
meta- and para-substituted benzyl compounds, but also for the
alkyl and naphthylmethyl compounds. To explain this phenom-
enon, we considered two possibilities relating to the binding
mode of these compounds. First, there may be a unique binding
pocket that accommodates the benzyl ether or alkyl ether group
(Figure 2 (a)), and the pseudo-symmetry of the 5-acetyl-6-
hydroxybenzofuran scaffold allows the 4-substituted compounds
to occupy this pocket similarly to 7-substituted compounds in
Figure 2d after flipping (Figure c) and inverting (Figure 2b)
the scaffold. The pseudo-symmetry of the compounds arises
because the furan and hydrogen-bonded salicyl groups can be
considered as bioisosteric. If the compounds cannot flip/invert
in the channel to exploit their pseudo-symmetry, an alternate
explanation has the scaffold locked in a single orientation, with
pseudosymmetry in the tetrameric Kv1.3 channel protein
offering two similar binding sites in the channel pore corre-
sponding to both the 4- and 7-positions. Compounds 16a and
16b were prepared to test these theories. The potency of
blockade of these analogues (Table 1, 16a, EC50 ) 6 µM and
16b, EC50 ) 1.2 µM) is comparable to their unsymmetrically
alkylated counterparts (5a, EC50 ) 10 µM, 5f, EC50 ) 7 µM
and 5d, EC50 ) 1.2 µM, 5i EC50 ) 2.5 µM), suggesting that
the latter hypothesis is probably false because an additional
substituent, able to access an appropriate binding site, would
Nature of Block. While both the 4- and the 7-series
compounds blocked Kv1.3 with comparable potencies and SAR,
they markedly differed in their mechanism of block. Figure 3
shows exemplary experiments performed with 11m and 14m.
We first applied a 500-ms depolarizing pulse to +40 mV to
elicit a control current from cells stably transfected with Kv1.3
and then clamped the membrane back to -80 mV. After 2 min,
a time interval which is sufficient to allow all Kv1.3 channels
to recover from their characteristic C-type inactivation,35,36 we
perfused blocking concentrations of the sea anemone toxin ShK,

Potassium Channel KV1.3 Blockers
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1437
Table 6. Selectivity of 7, 11m, 14m, and 16b over Other K+ Channelsa
7b
11m
14m
16b
channel
EC50 (µM)
EC50 (µM)
EC50 (µM)
EC50 (µM)
Kv1.1
Kv1.2
Kv1.3
Kv1.4
Kv1.5
HERG
1.7
>50
0.7
nd
12
1.7
2.5
0.5
4.8
0.7
5.0
1.2
1.4
0.4
4.3
0.9
9.0
7.0
10
1.2
>20
5.0
nd
nd
a EC50 values for blockade of other K+ channels. Compounds were tested
2-3 times at 3-5 different concentrations (standard deviations 2-15%).
nd: not determined. b Compound 7 was reported previously.25
that two blocker molecules bind to one channel molecule in a
cooperative manner. This observation is in accordance with two
binding sites for the 7-series on Kv1.3, one in the open state
and an additional site in the C-type inactivated state. These two
different stoichiometries of interaction are not necessarily
inconsistent with our flip/invert model. We hypothesize that the
first binding site in the open state of the channel binds both the
4- and the 7-series compounds but that the second site in the
C-type inactivated state recognizes khellinones only in the
inverted orientation as presented by 7-benzyl ether khellinones
and that therefore only the 7-series compounds can occupy this
second site. However, without mapping their binding sites on
the Kv1.3 channel protein we are currently not able to more
definitively interpret these different mechanisms of block by
compounds from the 4- and 7-series and it is certainly possible
to explain our results differently.
Figure 3. (A) Pulse protocol for the whole-cell patch-clamp experi-
ments on Kv1.3 shown in B, C, D, and E. After a control current had
been elicited the membrane was held at -80 mV for 2 min before
blockers were perfused and for another 8 min before the next
depolarizing pulse. Afterward pulses were applied every 60 s until no
further increase in effect was observed. (B) Effect of 100 pM ShK.
(C) Effect of 5 µM 11m. (D) Effect of 5 µM 14m. (E) Effect of 10
nM Psora-4.
Selectivity. As representative examples of the most active
compounds from each subclass 11m and 14m were tested for
selectivity over other Kv1-family channels and over the cardiac
K+ channel HERG (Kv11.1). As shown in Table 6, both 11m
and 14m exhibited a very similar selectivity profile: 3-fold
selectivity over Kv1.1, 5-fold over Kv1.2, 10-fold over Kv1.4,
2-fold over Kv1.5, and 10 to 20-fold over HERG. Interestingly,
this selectivity profile is very different from the previously
described chalcone derivatives of khellinone, such as 7 (Scheme
1),25 which exhibited close to 20-fold selectivity over the cardiac
potassium channels Kv1.5 and had no effect on the neuronal
potassium channel Kv1.2 (Table 6). The symmetrically alkylated
16b exhibited a similar selectivity profile to that of 11m and
14m with poor selectivity over Kv1.2 and Kv1.5 but was slightly
more selective over Kv1.1 (6-fold) and Kv1.4 (>20-fold). In
their lack of selectivity over Kv1.5 the 4- and 7-substituted
khellinones 11m and 14m and the symmetric 16b thus more
closely resemble the alkoxypsoralen Psora-424 and the nor-
triterpene correolide22 than the structurally related khellinone
chalcones.25 A reason for these differences in selectivity could
be that binding of the phenyl ring in the 5-position-derivatized
khellinone chalcones orients the furosalicyl scaffold differently
in the channel pores relative to the 4- or 7-substituted analogues
11m and 14m.
11m, 14m, or the alkoxypsoralen Psora-4 onto the closed
channel ensemble and then gave the compounds another 8 min
to diffuse to their respective binding sites (see pulse protocol
in Figure 3A). When we then opened the channels with another
depolarizing pulse, we observed very different kinetics of block
for the four blockers. ShK immediately showed full block and
had no effect on the inactivation kinetics because it can bind to
the channel in both the closed and the open state (Figure 3B).
Compound 11m in contrast closely resembled the open channel
blocker verapamil37,38 in its kinetics of block and in its
accelerating effect on Kv1.3 inactivation (Figure 3C). Like the
peptide ShK, 11m exhibited its full blocking potency on the
first pulse and its effect did not increase with subsequent
depolarizing pulses. Similar results were obtained with other
4-series compounds such as 11b, 11f, 11o, and 11r (data not
shown). Surprisingly, the 7-series compounds as exemplified
by 14m (Figure 3D) differed from 4-series compounds and
displayed blocking kinetics that are “a mixture” between the
open channel block of 4-series and the blockade of the C-type
inactivated state by Psora-424 (Figure 3E). Similar to 11m, 14m
greatly accelerated the inactivation of Kv1.3 but in contrast to
11m, its blocking effect increased considerably with subsequent
depolarizing pulses until it reached a steady-state block after
five pulses. Other 7-series compounds such as 14b, 14f, and
14s displayed the same blocking kinetics (data not shown).
For the 4-series compounds, the observed Hill coefficient of
1.2 (Table 2) indicates a 1:1 interaction between blocker and
channel molecule and is consistent with a single binding-site
in the open state of the channel. In contrast, the 7-series
compounds exhibit a Hill coefficient of 1.8 (Table 3), suggesting
Inhibition of Proliferation. To determine if 11m and 14m
indeed exhibit immunosuppressive activity as expected from
their submicromolar inhibition of Kv1.3, we tested their effect
on the anti-CD3 antibody stimulated proliferation of human
peripheral blood T cells. As shown in Figure 4, both com-
pounds suppressed T cell proliferation with an EC50 of 2.5 µM.
To ensure that this suppression of proliferation was not caused
by unspecific cytotoxicity, we tested both compounds against
Jurkat E6-1 and MEL cells and found that at concentrations
ranging from 250 nM to 20 µM neither compound significantly
reduced cell viability as measured by trypan blue exclusion after
48 h.

1438 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
HarVey et al.
Figure 4. Inhibition of T-cell proliferation. Effect of increasing concentrations of 11m and 14m on the anti-CD3 antibody stimulated [3H]-
thymidine incorporation of human T cells. Control: 41 500 ( 1500 counts.
moisture sensitive experiments were performed in oven-dried
glassware under an atmosphere of nitrogen. Anhydrous dimethyl-
formide was purchased from Aldrich. Petroleum ether describes a
mixture of hexanes in the bp range 30-50 °C.
General Procedure A. A suspension of 3f (1.0 equiv, 0.2 M),
cesium carbonate (2.0 equiv), and alkyl or benzyl halide (generally
1.2 equiv) in anhydrous dimethylformamide was stirred under
nitrogen at room temperature until completion as determined by
TLC (typically 16 h). The reaction mixture was then diluted with
ethyl acetate and washed twice with 2 M HCl and brine, dried over
MgSO4, and concentrated in vacuo. The crude product was purified
by flash chromatography.
General Procedure B. A suspension of 2 (1.0 equiv, 0.25 M),
potassium carbonate (2.0 equiv) and alkyl or benzyl halide (1.1
equiv) in anhydrous dimethylformamide was stirred under nitrogen
at 60 °C until completion as determined by TLC (typically 2-4
h). The reaction mixture was then diluted with ethyl acetate and
washed twice with 2 M HCl and brine, dried over MgSO4, and
concentrated in vacuo. The crude product was purified by flash
chromatography.
General Procedure C. A suspension of the phenol (1.0 equiv,
0.15 M), cesium carbonate (5.0 equiv) and iodomethane (5.0 equiv)
in anhydrous dimethylformamide was stirred under nitrogen at room
temperature for 16 h. The reaction mixture was partitioned over 2
M HCl and ethyl acetate and the organic phase was extracted twice
with 2 M HCl then washed with brine. After drying over MgSO4
and evaporation, the residue was either purified by flash chroma-
tography or used directly in the following step.
General Procedure D. To a solution of the khellin analogue
(0.01-0.1 M) in ethanol (2 parts) at reflux was added slowly 3 M
NaOH (1 part), and the resulting solution was stirred at 70 °C for
3 h. The reaction mixture was concentrated in vacuo to one-third
of the original volume and then acidified to pH 2 by Universal
indicator with 2 M HCl. If precipitation occurred, the solid was
collected by filtration washing with water. The solid was then dried
in the presence of P2O5. If precipitation did not occur, the acidified
reaction mixture was extracted twice with ethyl acetate and the
pooled organic layers were washed with brine, dried over MgSO4,
and concentrated in vacuo.
4,9-Dihydroxy-7-methyl-furo[3,2-g]chromen-5-one (2). To a
suspension of 1 (10.4 g, 40 mmol) in dichloromethane (100 mL)
at -78 °C boron tribromide (1 M in dichloromethane, 80 mL, 80
mmol) was added over 75 min. The reaction mixture was allowed
to warm to room temperature and then stirred for 16 h. On cooling
the reaction mixture to 0 °C, the reaction was quenched with water
by dropwise addition at first, but at an increasing rate until 50 mL
had been added. The reaction mixture was concentrated in vacuo
to remove the dichloromethane, and the resulting suspension was
filtered, washing with water. The orange solid was heated as a
suspension in propan-2-ol (200 mL) at 70 °C for 20 min, during
which time the solid became pale yellow. The suspension was
filtered and the solid was dried to afford 2 (8.4 g, 90%) as a pale
yellow solid: Mp 283-284 °C (iPrOH) lit.40 285-287 °C (EtOH);
1H NMR (DMSO-d6) δ: 2.42 (s, 3H), 6.19 (s, 1H), 7.05 (d, J )
2.1 Hz, 1H), 8.02 (d, J ) 2.1 Hz, 1H), 9.84 (br s, 1H), 12.9 (s,
1H); 13C NMR (DMSO-d6) δ: 20.2, 104.4, 105.1, 107.0, 112.9,
123.3, 141.8, 145.7, 146.3, 148.2, 168.5, 183.9; MS (ES+) m/z
233.2 (M + H+).
Conclusion
We reported here a new class of khellinone-based blockers
of the voltage-gated potassium channel Kv1.3. These were
obtained by replacing either the 4-methoxy or the 7-methoxy
groups in khellinone with a substituted benzyloxy group via
selective synthetic manipulation starting from khellin. The SAR
around the benzyl ring at the 4-position of khellinone paralleled
the SAR around the benzyl ring at the 7-position of khellinone.
However, taken together this SAR was relatively shallow,
demonstrating that the binding site for this class of compounds
on Kv1.3 must be relatively large and can accommodate
compounds of varying size. One of the more potent 4-benzyloxy
derivatives, 11m, contained a p-chloro substituent in the benzyl
ring and had an EC50 of 0.48 µM, while an representative potent
compound from the 7-benzyloxy series, 14m, with a p-bromo
substituent in the benzyl ring and had an EC50 of 0.40 µM.
Intriguingly, the two compound series were found to possess
quite different mechanisms of blockade. Compounds from the
4-series accelerated inactivation, displayed full blocking potency
on the first depolarizing pulse and a Hill coefficient close to 2.
In contrast, 7-series compounds, while also accelerating inac-
tivation, showed a use-dependent blockade and a Hill coefficient
close to unity. Both 11m and 14m, which were tested as
representatives of the 4- and 7-series, inhibited the proliferation
of human T cells with an EC50 of 2.5 µM and were not
appreciably cytotoxic at concentrations of up to 20 µM. Both
compounds were slightly to moderately selective for Kv1.3 over
a panel of other ion channels tested, with the exception of Kv1.5
where selectivity was marginal. Both compounds were 10-20-
fold selective for Kv1.3 over the cardiac potassium channel
HERG.
Experimental Section
Melting points were determined in open capillary tubes using
an Electrothermal melting point apparatus (Model 9200) and are
uncorrected. Multiple parallel synthesis was performed on a Mettler
Toledo Autochem MiniBlock XT with a 6 × 4 reaction grid. Mass
spectra (MS) were recorded on a Finnigan LQC Advantage MAX
mass spectrometer. High-resolution mass spectra (HRMS) were
recorded on a Kratos MS80 RFA mass spectrometer at the
University of Canterbury, New Zealand. Elemental analyses were
performed for carbon, hydrogen, and chlorine at the University of
Otago, New Zealand. Proton nuclear magnetic resonance (1H NMR)
were performed on a Bruker Avance DRX 300 (300 MHz) with
the solvents indicated, and carbon nuclear magnetic resonance (13C
NMR) were performed on a Bruker Avance DRX 300 (75 MHz)
with the solvents indicated. Chemical shifts were reported in parts
per million (ppm) on the δ scale and were referenced to the
appropriate solvent peaks: CDCl3 referenced to CHCl3 at δH 7.24
ppm and CDCl3 at δC 77.0 ppm; acetone-d6 referenced to
(CD3)(CHCD2)CO at δH 2.17 ppm and (CD3)2CO at δC 29.2 ppm.
Analytical thin-layer chromatography (TLC) was performed on
Merck silica gel 60F254 aluminum-backed plates. Flash chroma-
tography was performed on Merck silica 60 following the guidelines
given by Still et al.39 using ACS analytical grade solvents. All

Potassium Channel KV1.3 Blockers
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1439
4-Hydroxy-9-methoxy-7-methyl-furo[3,2-g]chromen-5-one (3f).
To a solution of 2 (2.32 g, 10 mmol) in acetonitrile (45 mL) were
added K2CO3 (2.76 g, 20 mmol) and dimethyl sulfate (1.00 mL,
10.5 mmol), and the reaction mixture was stirred at 80 °C for 90
min. The reaction was quenched by addition of 10% aqueous citric
acid solution (30 mL) followed by 15 min stirring at room
temperature. The reaction mixture was then concentrated to half
volume, and the resulting suspension was partitioned over chloro-
form (100 mL) and 10% citric acid solution (100 mL). The phases
were separated, and the aqueous phase was extracted with chloro-
form (2 × 100 mL) to give 3f (2.33 g, 95%) as a yellow solid:
67%) as a deep red solid: Mp 153-155 °C; 1H NMR (CDCl3) δ:
1.39 (t, J ) 6.9 Hz, 3H), 4.10 (s, 3H), 4.22 (q, J ) 6.9 Hz, 2H),
6.83 (m, 1H), 6.90 (d, J ) 7.8 Hz, 1H), 7.11 (s, 1H), 7.26 (m, 2H),
7.51 (m, 1H), 7.76 (d, J ) 15.6 Hz, 1H), 7.93 (d, J ) 15.6 Hz,
1H); 13C NMR (CDCl3) δ: 15.3, 60.6, 71.0, 105.0, 112.1, 113.3,
114.5, 117.3, 121.0, 127.3, 129.2, 129.8, 136.4, 142.4, 143.8, 149.5,
151.4, 152.7, 155.7; MS (ES+) m/z 355.2 (M + H+); HRMS (EI)
calcd C20H18O6 (M+) 354.1103, found 354.1106.
1,6-Bis(5-acetyl-4-ethoxy-7-methoxybenzofuran-6-yloxy)hex-
ane (8). A suspension of 5f (20 mg, 0.080 mmol), 1,6-dibromo-
hexane (10 mg, 0.042 mmol), and potassium carbonate (11 mg,
0.080 mmol) in dimethylformamide (0.5 mL) was stirred at 70 °C
for 6 h, and the reaction mixture was diluted with 10% aqueous
citric acid (20 mL) and extracted with ethyl acetate (3 × 10 mL).
The pooled organic layers were dried over MgSO4 and concentrated
in vacuo to give a residue that was purified by flash chromatography
eluting with ethyl acetate/hexane (gradient from 1:9 to 1:4) to afford
1
Mp 201-202 °C lit.41 204 °C (EtOH); H NMR (DMSO-d6) δ:
2.46 (s, 3H), 4.04 (s, 3H), 6.30 (s, 1H), 7.13 (d, J ) 2.1 Hz, 1H),
8.09 (d, J ) 2.1 Hz, 1H); 13C NMR (CDCl3) δ: 20.7, 61.8, 104.7,
105.6, 107.8, 113.8, 125.9, 114.9, 145.4, 149.9, 151.0, 167.4, 184.2;
MS (ES+) m/z 247.2 (M + H+).
4-Ethoxy-9-methoxy-7-methyl-furo[3,2-g]chromen-5-one (4f).
Compound 3f (110 mg, 0.45 mmol) and ethyl iodide (0.35 mL,
4.5 mmol) were treated as described under General Procedure A
to afford 4f (110 mg, 90%) as a colorless solid: Mp 104-106 °C
lit.26 109-110 °C (EtOH/H2O); 1H NMR (CDCl3) δ: 1.43 (t, J )
10.5 Hz, 3H), 2.35 (s, 3H), 4.16 (m, 5H), 6.03 (s, 1H), 6.93 (d, J
) 3.2 Hz), 7.58 (d, J ) 3.2 Hz); 13C NMR (CDCl3) δ: 15.6, 20.0,
61.3, 71.2, 105.2, 110.4, 120.5, 129.9, 145.4, 146.0, 147.0, 148.5,
164.0, 178.1, 196.2; MS (ES+) m/z 275.1 (M + H+).
5-Acetyl-4-ethoxy-6-hydroxy-7-methoxybenzofuran (5f). Com-
pound 4f (49 mg, 0.20 mmol) was treated as described under
General Procedure D to afford 5f (32 mg, 75%) as a colorless
solid: Mp 94-95 °C; lit.42 93-95 °C (EtOH/H2O); 1H NMR
(CDCl3) δ: 1.52 (t, J ) 6.9 Hz, 3H), 2.72 (s, 3H), 4.05 (s, 3H),
4.40 (q, J ) 6.9 Hz, 2H), 6.84 (d, J ) 2.4 Hz, 1H), 7.49 (d, J )
2.4 Hz, 1H), 13.08 (s, 1H); 13C NMR (CDCl3) δ: 15.2, 33.0, 60.6,
69.2, 110.5, 110.7, 123.1, 128.4, 143.3, 150.4, 151.9, 153.1, 205.0;
MS (ES+) m/z 251.0 (M + H+).
1
7 (21 mg, 90%) as a colorless solid: Mp 134-135 °C; H NMR
(CDCl3) δ: 1.49 (m, 4H), 1.76 (m, 4H), 2.53 (s, 3H), 4.06 (m,
10H), 4.20 (q, J ) 7.2 Hz, 4H), 6.81 (d, J ) 2.1 Hz, 2H), 7.56 (d,
J ) 2.1 Hz, 2H); 13C NMR (CDCl3) δ: 15.6, 25.7, 29.7, 30.1,
32.8, 61.1, 69.9, 75.2, 105.1, 117.1, 125.3, 134.2, 143.1, 144.5,
144.9, 148.6; MS (ES+) m/z 583.3 (M + H+); Anal. (C32H38O10):
C, H.
5-Acetyl-4-(4′-chloro)benzyloxy-6-hydroxy-7-methoxybenzo-
furan (11m). Compound 3f (27 mg, 0.11 mmol) was treated as
described under General Procedure E (Supporting Information) to
afford 11m (20 mg, 44% over two steps) as a yellow solid: Mp
106-108 °C; 1H NMR (CDCl3) δ: 2.61 (s, 3H), 4.06 (s, 3H), 5.27
(s, 2H), 6.75 (d, J ) 2.1 Hz, 1H), 7.38 (m, 4H), 7.49 (d, J ) 2.1
Hz, 1H), 12.94 (s, 1H); 13C NMR (CDCl3) δ: 33.3, 61.0, 75.2,
105.4, 111.4, 111.8, 129.0, 129.1, 129.4, 134.5, 134.6, 144.1, 150.0,
152.0, 153.3, 205.1; MS (ES-) m/z 345.3 (M - H+); Anal.
(C19H17O5Cl): C, H, Cl.
9-Benzyloxy-4-Hydroxy-7-methyl-furo[3,2-g]chromen-5-
one (3d). Compound 2 (348 mg, 1.5 mmol) and benzyl bromide
(214 µL, 1.8 mmol) were treated as described under General
Procedure B to afford 3d (370 mg, 77%) as a colorless solid: Mp
110-112 °C lit43 114 °C; 1H NMR (CDCl3) δ: 2.33 (s, 3H), 5.25
(s, 2H), 6.00 (s, 1H), 6.97 (d, J ) 2.2 Hz, 1H), 7.32 (m, 3H), 7.45
(m, 2H), 7.59 (d, J ) 2.2 Hz, 1H); 13C NMR (CDCl3) δ: 20.5,
76.1, 104.7, 105.5, 107.6, 113.7, 124.3, 128.3, 128.4, 136.8, 144.9,
146.0, 150.2, 151.5, 167.3, 184.1; MS (ES+) m/z 323.1 (M + H+).
9-Benzyloxy-4-methoxy-7-methyl-furo[3,2-g]chromen-5-one
(4d). Compound 3d (300 mg, 0.93 mmol) was treated as described
under General Procedure C to afford 4d (290 mg, 93%) as a
5-Acetyl-7-(4′-bromo)benzyloxy-6-hydroxy-4-methoxybenzo-
furan (14m). Compound 2 (59 mg, 0.26 mmol) was treated as
described under General Procedure F (Supporting Information) to
afford 14m (31 mg, 31% over three steps) as a yellow solid: Mp
106-107 °C; 1H NMR (CDCl3) δ: 2.70 (s, 3H), 4.12 (s, 3H), 5.18
(s, 2H), 6.86 (d, J ) 2.4 Hz, 1H), 7.37-7.46 (m, 5H); 13C NMR
(CDCl3) δ: 33.2, 60.4, 74.0, 105.7, 110.6, 110.6, 121.9, 127.2,
129.9, 131.4, 136.5, 143.8, 152.0, 152.7, 153.9, 205.2; MS (ES+)
m/z 390.8, 392.8 (M + H+); HRMS (EI) calcd C18H15O5Cl (M+)
346.0608, found 346.0614.
4,9-Diethoxy-7-methyl-furo[3,2-g]chromen-5-one (15a). A sus-
pension of 2 (174 mg, 0.75 mmol), cesium carbonate (731 mg,
2.25 mmol), and ethyl iodide (182 µL, 2.25 mmol) was stirred at
room temperature for 6 h, then diluted with ethyl acetate (30 mL)
and washed with 2 M HCl (3 × 15 mL) and brine (15 mL). The
resulting solution was dried over MgSO4 and concentrated in vacuo
to afford 15a (205 mg, 95%) as a brown solid: Mp 90 °C; lit.28 94
°C (petroleum ether); 1H NMR (CDCl3) δ: 1.46 (m, 6H), 2.37 (s,
3H), 4.20 (q, J ) 7.0 Hz, 2H), 4.41 (q, J ) 7.1 Hz, 2H), 6.05 (s,
1H), 6.95 (d, J ) 2.2 Hz, 1H), 7.59 (d, J ) 2.2 Hz, 1H).
4,9-Dibenzyloxy-7-methyl-furo[3,2-g]chromen-5-one (15b). A
suspension of 2 (174 mg, 0.75 mmol), cesium carbonate (731 mg,
2.25 mmol), and benzyl bromide (267 µL, 2.25 mmol) was stirred
at 60 °C for 3 h, then diluted with ethyl acetate (30 mL) and washed
with 2 M HCl (3 × 15 mL) and brine (15 mL). This solution was
dried over MgSO4 and concentrated in vacuo to give a residue that
was purified by flash chromatography, eluting with ethyl acetate/
petroleum ether (3:7), to afford 15b (263 mg, 85%) as a colorless
1
colorless solid: Mp 98-100 °C; H NMR (CDCl3) δ: 2.30 (s,
3H), 4.04 (s, 3H), 5.32 (s, 2H), 6.02 (s, 1H), 6.99 (d, J ) 2.2 Hz,
1H), 7.31 (m, 3H), 7.47 (m, 2H), 7.60 (d, J ) 2.2 Hz, 1H); 13C
NMR (CDCl3) δ: 19.6, 61.9, 75.5, 104.9, 110.2, 113.2, 118.7,
127.8, 128.0, 136.5, 145.2, 147.4, 147.6, 149.3, 163.6, 177.9; MS
(ES+) m/z 337.1 (M + H+); Anal. (C20H16O5): C, H.
5-Acetyl-7-benzyloxy-6-hydroxy-4-methoxybenzofuran (5d).
Compound 4d (200 mg, 0.59 mmol) was treated as described under
General Procedure D, and the crude product was purified by flash
chromatography eluting with ethyl acetate/petroleum ether (1:9) to
afford 5d (90 mg, 48%) as a yellow solid: Mp 78-80 °C; 1H NMR
(CDCl3) δ: 2.68 (s, 3H), 4.09 (s, 3H), 5.23 (s, 2H), 6.84 (d, J )
2.3 Hz, 1H), 7.32 (m, 3H), 7.60 (d, J ) 2.3 Hz, 1H), 7.54 (m, 2H);
13C NMR (CDCl3) δ: 32.9, 60.0, 74.5, 105.4, 110.1, 110.2, 127.1,
127.6, 127.9, 137.2, 143.4, 151.5, 152.5, 153.6, 204.9; MS (ES+)
m/z 312.9 (M + H+); Anal. (C18H16O5): C, H.
3-(4-ethoxy-6-hydroxy-7-methoxybenzofuran-5-yl)-1-(3-
hydroxyphenyl)-3-oxopropene (6). Compound 5f (15 mg, 0.060
mmol) and 3-hydroxybenzaldehyde (9.0 mg, 0.075 mmol) were
suspended in 2 M NaOH (0.5 mL) and stirred at 80 °C for 3 h.
The reaction mixture was diluted with 10% aqueous citric acid (10
mL) and then extracted with ethyl acetate (2 × 10 mL). The pooled
organic layers were dried over MgSO4, filtered, and concentrated
in vacuo. The resulting residue was purified by flash chromatog-
raphy eluting with ethyl acetate/hexane (1:4) to afford 6 (14 mg,
1
solid: Mp 128-130 °C; H NMR (CDCl3) δ: 2.34 (s, 3H), 5.17
(s, 2H), 5.36 (s, 2H), 6.05 (s, 1H), 6.71 (d, J ) 1.9 Hz, 1H), 7.33-
7.35 (m, 6H), 7.45-7.76 (m, 5H); 13C NMR (CDCl3) δ: 20.0, 75.8,
77.4, 105.3, 110.5, 114.2, 120.8, 128.0, 128.2, 128.3, 128.7, 128.9,
136.8, 137.3, 145.5, 146.2, 147.8, 149.2, 164.0, 178.2; MS (ES+)
m/z 413.2 (M + H+); Anal. (C26H20O5): C, H.
5-Acetyl-4,7-diethoxy-6-hydroxybenzofuran (16a). Compound
15a (144 mg, 0.50 mmol) was treated as described in General
Procedure D, and the crude product was purified by flash chro-

1440 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4
HarVey et al.
(3) Cahalan, M. D.; Wulff, H.; Chandy, K. G. Molecular Properties and
Physiological Roles of Ion Channels in the Immune System. J. Clin.
Immunol. 2001, 21, 235-252.
matography, eluting with ethyl acetate/petroleum ether (1:4), to
furnish 16a (73 mg, 55%) as a yellow solid: Mp 72-74 °C; H
1
NMR (CDCl3) δ: 1.36 (t, J ) 7.0 Hz, 2H), 1.46 (t, J ) 7.0 Hz,
2H), 2.70 (s, 3H), 4.22 (q, J ) 7.0 Hz, 2H), 4.36 (q, J ) 7.0 Hz,
2H), 6.79 (d, J ) 2.3 Hz, 1H), 7.42 (d, J ) 2.3 Hz, 1H); 13C NMR
(CDCl3) δ: 15.4, 15.5, 33.3, 68.8, 69.4, 105.8, 110.7, 110.8, 143.5,
150.7, 152.8, 153.8, 205.2; MS (ES+) m/z 265.1 (M + H+); Anal.
(C14H16O5): C, H.
(4) Chandy, K. G.; Wulff, H.; Beeton, C.; Pennington, M.; Gutman, G.
A. et al. Potassium Channels as Targets for Specific Immunomodu-
lation. Trends Pharmacol. Sci. 2004, 25, 280-289.
(5) Lewis, R. S. Calcium Signaling Mechanisms in T lymphocytes. Annu.
ReV. Immunol. 2001, 19, 497-521.
(6) Hess, S. D.; Oortgiesen, M.; Cahalan, M. D. Calcium Oscillations
in Human T and Natural Killer Cells depend on Membrane Potential
and Calcium Influx. J. Immunol. 1993, 150, 2620-2633.
(7) Wulff, H.; Calabresi, P. A.; Allie, R.; Yun, S.; Pennington, M. et al.
The Voltage-Gated Kv1.3 K+ Channel in Effector Memory T cells
as New Target for MS. J. Clin. InVest. 2003, 111, 1703-1713.
(8) Wulff, H.; Knaus, H. G.; Pennington, M.; Chandy, K. G. K+ Channel
Expression During B-cell Differentiation: Implications for Immuno-
modulation and Autoimmunity. J. Immunol. 2004, 173, 776-786.
(9) Beeton, C.; Chandy, K. G. Potassium Channels, Memory T cells and
Multiple Sclerosis. Neuroscientist 2005, in press.
5-Acetyl-4,7-benzyloxy-6-hydroxybenzofuran (16b). Com-
pound 15b (124 mg, 0.30 mmol) was treated as described in General
Procedure D, and the crude product was purified by flash chro-
matography, eluting with ethyl acetate/petroleum ether (1:9), to
1
furnish 16b (45 mg, 41%) as a yellow solid: Mp 74-76 °C; H
NMR (CDCl3) δ: 2.60 (s, 3H), 5.26 (s, 2H), 5.30 (s, 2H), 6.75 (d,
J ) 2.3 Hz, 1H), 7.25-7.78 (m, 11H); 13C NMR (CDCl3) δ: 33.4,
74.9, 105.6, 111.4, 111.6, 118.1, 127.8, 128.0, 128.3, 128.6, 128.8,
136.0, 137.4, 144.0, 150.6, 152.7, 153.8, 205.3; MS (ES+) m/z 389.3
(M + H+); Anal. (C24H20O5): C, H.
(10) Beeton, C.; Barbaria, J.; Giraud, P.; Devaux, J.; Benoliel, A. et al.
Selective Blocking of Voltage-Gated K+ Channels Improves Ex-
perimental Autoimmune Encephalomyelitis and Inhibits T Cell
Activation. J. Immunol. 2001, 166, 936-944.
(11) Beeton, C.; Wulff, H.; Barbaria, J.; Clot-Faybesse, O.; Pennington,
M. et al. Selective Blockade of T Lymphocyte K+ Channels
Ameliorates Experimental Autoimmune Encephalomyelitis, a Model
for Multiple Sclerosis. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 13942-
13947.
(12) Beeton, C.; Pennington, M. W.; Wulff, H.; Singh, S.; Nugent, D. et
al. Targeting Effector Memory T cells with a Selective Peptide
Inhibitor of Kv1.3 Channels for Therapy of Autoimmune Diseases.
Mol. Pharmacol. 2005, 67, 1369-1381.
(13) Rus, H.; Pardo, C. A.; Hu, L.; Darrah, E.; Cudrici, C. et al. The
Voltage-Gated Potassium Channel Kv1.3 is Highly Expressed on
Inflammatory Infiltrates in Multiple Sclerosis Brain. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 11094-11099.
(14) Norton, R. S.; Pennington, M. W.; Wulff, H. Potassium Channel
Blockade by the Sea Anemone Toxin ShK for the Treatment of
Multiple Sclerosis and Other Autoimmune Diseases. Curr. Med.
Chem. 2004, 11, 3041-3052.
(15) Wulff, H.; Beeton, C.; Chandy, K. G. Potassium Channels as
Therapeutic Targets for Autoimmune Disorders. Curr. Opin. Drug
DiscoVery DeV. 2003, 6, 640-647.
(16) Grissmer, S.; Dethlefs, B.; Wasmuth, J. J.; Goldin, A. L.; Gutman,
G. A. et al. Expression and Chromosomal Localization of a
Lymphocyte K+ Channel gene. Proc. Natl. Acad. Sci. U.S.A. 1990,
87, 9411-9415.
(17) Baell, J. B.; Harvey, A. J.; Norton, R. S. Design and Synthesis of
Type-III Mimetics of ShK Toxin. J. Comput.-Aided Mol. Des. 2002,
16, 245-262.
(18) Harvey, A. J.; Gable, R. W.; Baell, J. B. A Three-Residue, Continuous
Binding Epitope Peptidomimetic of ShK Toxin as a Kv1.3 Inhibitor.
Bioorg. Med. Chem. Lett. 2005, 15, 3193-3196.
(19) Felix, J. P.; Bugianesi, R. M.; Schmalhofer, W. A.; Borris, R.; Goetz,
M. A. et al. Identification and Biochemical Characterization of a
Novel Nortriterpene Inhibitor of the Human Lymphocyte Voltage-
Gated Potassium Channel, Kv1.3. Biochemistry 1999, 38, 4922-
4930.
(20) Ghanshani, S.; Wulff, H.; Miller, M. J.; Rohm, H.; Neben, A. et al.
Up-regulation of the IKCa1 Potassium Channel During T-cell
Activation: Molecular Mechanism and Functional Consequences. J.
Biol. Chem. 2000, 275, 37137-37149.
(21) Koo, G. C.; Blake, J. T.; Shah, K.; Staruch, M. J.; Dumont, F. et al.
Correolide and Derivatives are Novel Immunosuppressants Blocking
the Lymphocyte Kv1.3 Potassium Channels. Cell. Immunol. 1999,
197, 99-107.
(22) Hanner, M.; Schmalhofer, W. A.; Green, B.; Bordallo, C.; Liu, J. et
al. Binding of Correolide to Kv1 Family Potassium Channels. J. Biol.
Chem. 1999, 274, 25237-25244.
Electrophysiology. The effectiveness of the generated com-
pounds in blocking Kv1.3 was assayed on L929 cells stably
expressing mKV1.3.44 All experiments were conducted in the whole-
cell configuration of the patch-clamp technique with a holding
potential of -80 mV. Currents were recorded in normal Ringer
solution (160 mM NaCl, 4.5 mM KCl, 2 mM CaCl2, 1 mM MgCl2,
10 mM HEPES, pH 7.4, 290-310 mOsm) with an internal pipet
solution containing 134 mM KF, 2 mM MgCl2, 10 mM HEPES,
10 mM EGTA (pH 7.2, 290-310 mOsm). If currents exceeded 2
nA 60-80% series resistance compensation was used. Depolarizing
pulses to 40 mV were applied every 45 s for 500 ms. Kv.1.1, Kv1.2,
Kv1.5, Kv1.4, and HERG currents were recorded from cells stably
expressing these currents as previously described.44-46 EC50 values
were determined by fitting the reduction of area under the current
curve to the Hill equation.
Cytotoxicity Assay. Jurkat E61 and MEL cells were seeded at
5 × 105 cells/mL in twelve-well plates. Compounds were added at
concentrations of 2.5, 10, and 20 µM in a final DMSO concentration
of 0.1%, which was found not to affect cell viability. After 48 h
the cells in each well were well mixed and resuspended and the
number of trypan blue positive cells in three aliquots from each
well determined under a light microscope. The test was repeated
twice.
T-Cell Proliferation. Peripheral blood mononuclear cells were
isolated from the blood of healthy volunteers with the help of a
density gradient. Cells were washed and seeded at 2 × 105 cells
per well in medium (RPMI 1640 supplemented 10% fetal calf
serum, 2 mM glutamine, 1 mM sodium pyruvate, 1% nonessential
amino acids, 100 units/mL penicillin, 100 µg/mL streptomycin, and
50 µM â-mercaptoethanol) in flat-bottom 96-well plates (final
volume 200 µL). Cells preincubated with compound (30 min) and
then stimulated with 5 ng/mL anti-CD3 antibody (Biomeda) for
48 h. [3H]-Thymidine (1 µCi per well) was added for the last 6 h.
Cells were harvested onto glass fiber filters, and radioactivity was
measured in a scintillation counter. All experiments were done in
triplicate. Results are reported as normalized for maximum [3H]-
thymidine incorporation for controls.
Acknowledgment. The authors wish to thank Dr. Sanka-
narayanan for suggestions on the manuscript.
(23) Schmalhofer, W. A.; Bao, J.; McManus, O. B.; Green, B.; Matyskiela,
M. et al. Identification of a New Class of Inhibitors of the Voltage-
Gated Potassium Channel, Kv1.3, with Immunosuppressant Proper-
ties. Biochemistry 2002, 18, 7781-7794.
(24) Vennekamp, J.; Wulff, H.; Beeton, C.; Calabresi, P. A.; Grissmer,
S. et al. Kv1.3 Blocking 5-Phenylalkoxypsoralens: a New Class of
Immunomodulators. Mol. Pharmacol. 2004, 65, 1364-1373.
(25) Baell, J. B.; Gable, R. W.; Harvey, A. J.; Toovey, N.; Herzog, T. et
al. Khellinone Derivatives as Blockers of the Voltage-Gated Potas-
sium Channel Kv1.3: Synthesis and Immunosuppressive Activity.
J. Med. Chem. 2004, 47, 2326-2336.
(26) Abu-Shady, H.; Soine, T. O. Experiments with Khellin. I. The
Preparation of Desmethylkhellin and Some of its Derivatives. J. Am.
Pharm. Assoc. 1952, 41, 325-327.
Supporting Information Available: Experimental details. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Lin, C. S.; Boltz, R. C.; Blake, J. T.; Nguyen, M.; Talento, A. et al.
Voltage-gated Potassium Channels Regulate Calcium-dependent
Pathways Involved in Human T Lymphocyte Activation. J. Exp. Med.
1993, 177, 637-645.
(2) Lewis, R. S.; Cahalan, M. D. Potassium and Calcium Channels in
Lymphocytes. Annu. ReV. Immunol. 1995, 13, 623-653.

Potassium Channel KV1.3 Blockers
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 4 1441
(27) Gammill, R. B. Preparing 7-Alkoxybenzofurans, Intermediates Used
Therein, and Furochromones used Therefrom. Eur. Pat. Appl. EP
94769, 1983, p 18.
(28) Schonberg, A.; Sina, A. On Visnagin and Khellin and Related
Compounds - a Simple Synthesis of Chromone. J. Am. Chem. Soc.
1950, 72, 3396-3399.
(39) Still, W. C.; Kahn, M.; Mitra, A. Rapid Chromatographic Technique
for Preparative Separations With Moderate Resolution. J. Org. Chem.
1978, 43, 2923-2925.
(40) Murti, V. V. S.; Seshadri, T. R. Nuclear Oxidation in Flavones and
Related Compounds. XXIII. A Synthesis of Khellin. Proc. Ind. Acad.
Sci, Sect. A 1949, 30A, 107-113.
(29) Renato, S.; DiPaco, G. Ann. Chim. (Rome) 1958, 48, 1205-1209.
(30) Fourneau, J. P. Ann. Pharm. Fr. 1953, 11, 685-695.
(31) Baell, J. B.; Gable, R. W.; Harvey, A. J. 1,6-Bis(5-acetyl-4,7-
dimethoxybenzofuran-6-yloxy)hexane. Acta Crystallogr. Sect. E
2004, 60, O996-O997.
(41) Asker, W.; Shalaby, A.; Zayed, S. Reactions with Mercaptans 0.5.
Action of Aromatic Thiols on Furocoumarins, Furochromones, and
2-Aralkylidene-3(2H)-Thianaphthenone-1,1-Dioxides. J. Org. Chem.
1958, 23, 1781-1783.
(42) Musante, C.; Stener, A. Gazz. Chim. Ital. 1956, 86, 297-315.
(43) Martin, M.; Cautain, M.; Sado, M.; Zuckerka. F.; Fourneau, J. P. et
al. Synthesis and Beta-Adrenolytic Properties of Some Amino
alcohols of Fluoro 3,2-G Chromone. Eur. J. Med. Chem. 1974, 9,
563-570.
(44) Grissmer, S.; Nguyen, A. N.; Aiyar, J.; Hanson, D. C.; Mather, R. J.
et al. Pharmacological Characterization of Five Cloned Voltage-Gated
K+ Channels, Types Kv1.1, 1.2, 1.3, 1.5, and 3.1, Stably Ex-
pressed in Mammalian Cell Lines. Mol. Pharmacol. 1994, 45, 1227-
1234.
(45) Wulff, H.; Miller, M. J.; Haensel, W.; Grissmer, S.; Cahalan, M. D.
et al. Design of a Potent and Selective Inhibitor of the Intermediate-
Conductance Ca2+-Activated K+ Channel, IKCa1: A Potential
Immunosuppressant. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 8151-
8156.
(32) Bennett, P. B.; Guthrie, H. R. Trends in Ion Channel Drug
Discovery: Advances in Screening Technologies. Trends Biotechnol.
2003, 21, 563-569.
(33) Wang, X.; Li, M. Automated Electrophysiology: High Throughput
of Art. Assay Drug DeV. Technol. 2003, 1, 695-708.
(34) Levy, D. I.; Deutsch, C. Recovery from C-Type Inactivation is
Modulated by Extracellular Potassium. Biophys. J. 1996, 70, 798-
805.
(35) Cahalan, M. D.; Chandy, K. G.; DeCoursey, T. E.; Gupta, S. A
Voltage-Gated Potassium Channel in Human T Lymphocytes. J.
Physiol. (London) 1985, 358, 197-237.
(36) Panyi, G.; Sheng, Z.; Deutsch, C. C-type Inactivation of a Voltage-
Gated K+ Channel Occurs by a Cooperative Mechanism. Biophys.
J. 1995, 69, 896-903.
(37) DeCoursey, T. E. Mechanism of K+ Channel Block by Verapamil
and Related Compounds in Rat Alveolar Epithelial Cells. J. Gen.
Physiol. 1995, 106, 745-779.
(46) Bardien-Kruger, S.; Wulff, H.; Arieff, Z.; Brink, P.; Chandy, K. G.
et al. Characterization of the Human Voltage-Gated Potassium
Channel Gene, KCNA7, a Candidate Gene For Inherited Cardiac
Disorders, and its Exclusion as a Cause of Progressive Familial Heart
Block I (PFHBI). Eur. J. Human Gen. 2002, 10, 36-43.
(38) Robe, R. J.; Grissmer, S. Block of the Lymphocyte K+ Channel
mKv1.3 by the Phenylalkylamine Verapamil: Kinetic Aspects of
Block and Disruption of Accumulation of Block by a Single Point
Mutation. Br. J. Pharmacol. 2000, 131, 1275-1284.
JM050839V