A Rational Design of a Selective Inhibitor for Kv1.1 Channels Prevalent in Demyelinated Nerves That Improves Their Impaired Axonal Conduction

Article abstract of DOI:10.1021/acs.jmedchem.6b01262

K⁺ channels containing Kv1.1 α subunits, which become prevalent at internodes in demyelinated axons, may underlie their dysfunctional conduction akin to muscle weakness in multiple sclerosis. Small inhibitors were sought with selectivity for the culpable hyper-polarizing K⁺ currents. Modeling of interactions with the extracellular pore in a Kv1.1-deduced structure identified diaryldi(2-pyrrolyl)methane as a suitable scaffold with optimized alkyl ammonium side chains. The resultant synthesized candidate [2,2′-((5,5′(di-p-topyldiaryldi(2-pyrrolyl)methane)bis(2,2′carbonyl)bis(azanediyl)) diethaneamine·2HCl] (8) selectively blocked Kv1.1 channels (IC₅₀ ≈ 15 μM) recombinantly expressed in mammalian cells, induced a positive shift in the voltage dependency of K⁺ current activation, and slowed its kinetics. It preferentially inhibited channels containing two or more Kv1.1 subunits regardless of their positioning in concatenated tetramers. In slices of corpus callosum from mice subjected to a demyelination protocol, this novel inhibitor improved neuronal conduction, highlighting its potential for alleviating symptoms in multiple sclerosis.

Full text of DOI:10.1021/acs.jmedchem.6b01262

Subscriber access provided by University of Newcastle, Australia
Article
A Rational Design of a Selective Inhibitor for Kv1.1 Channels Prevalent
in Demyelinated Nerves that Improves their Impaired Axonal Conduction
Ahmed Al-Sabi, Declan Daly, Patrick Hoefer, Gemma K. Kinsella, Charles P. E. Metais,
Mark Pickering, Caroline Herron, Seshu Kumar Kaza, Kieran J. Nolan, and J Oliver Dolly
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01262 • Publication Date (Web): 22 Feb 2017
Downloaded from http://pubs.acs.org on February 23, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A Rational Design of a Selective Inhibitor for Kv1.1
Channels Prevalent in Demyelinated Nerves that
Improves their Impaired Axonal Conduction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ahmed Al-Sabi†, Declan Daly‡, Patrick Hoefer‡, Gemma K. Kinsella§, Charles Metais†, Mark
Pickering║, Caroline Herron┴, Seshu Kumar Kaza†, Kieran Nolan‡* and J. Oliver Dolly†*
†International Centre for Neurotherapeutics, Dublin City University, Glasnevin, Dublin 9, Ireland
‡School of Chemical Sciences, Dublin City University, Glasnevin, Dublin 9, Ireland
§School of Food Science and Environmental Health, College of Sciences and Health, Dublin
Institute of Technology, Cathal Brugha Street, Dublin 1, Ireland
║UCD School of Medicine, University College Dublin, Ireland
┴School of Biomolecular and Biomed Science, Conway Institute, Belfield, Dublin 4, Ireland
ABSTRACT: K+ channels containing Kv1.1 α subunits, which become prevalent at internodes in
demyelinated axons, may underlie their dysfunctional conduction akin to muscle weakness in
multiple sclerosis. Small inhibitors were sought with selectivity for the culpable hyperꢀpolarising
K+ currents. Modelling of interactions with the extracellular pore in a Kv1.1 deducedꢀstructure
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 42
1
2
3
4
5
6
7
8
9
identified a diaryldiꢀ(2ꢀpyrrolyl)methane (DPM) as a suitable scaffold, with optimised alkyl
ammonium side chains. The resultant synthesised candidate [2,2’ꢀ((5,5’(diꢀpꢀtopyldiaryldiꢀ(2ꢀ
pyrrolyl)methane)bis(2,2’carbonyl)bis (azanediyl)) diethaneamine.2HCl] (8) selectively blocked
Kv1.1 channels (IC50 ~15 µM) recombinantly expressed in mammalian cells, induced a positive
shift in the voltageꢀdependency of K+ current activation and slowed its kinetics. It inhibited
preferentially channels containing 2 or more Kv1.1 subunits, regardless of their positioning in
concatenated tetramers. In slices of corpus callosum from mice subjected to a demyelination
protocol, this novel inhibitor improved neuronal conduction, highlighting its potential for
alleviating symptoms in multiple sclerosis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
K+ channels are important for regulating membrane excitability which, in turn, affects signal
transmission. Voltageꢀgated K+ channels of the Shaker subfamily (Kv1) play vital roles in
neurons by controlling cell excitability and synaptic transmission. These channels are involved in
some human diseases,1 with their vital functions being altered by mutation (e.g. certain forms of
epilepsy2 or truncation (e.g. Episodic ataxia type I).3 Kv1 channels purified from mammalian
brain, using selective blockers, αꢀdendrotoxin (αDTX) or DTXk, are large (M.W. ~400 k)
sialoglycoprotein complexes,4 consisting of four poreꢀforming α subunits and four
cytoplasmicallyꢀassociated auxiliary β proteins.5, 6 When heterologously expressed, each of the
major genes encoding α subunits [Kv1.1ꢀ1.6]7ꢀ9 and Kv1.710 yields a homoꢀtetrameric channel
with distinct biophysical and pharmacological profiles. Further diversity can be created in vivo
by heteroꢀpolymerisation, but only a subset of the possible oligomeric combinations has been
isolated from mammalian brain,11ꢀ14 suggesting that their synthesis and/or assembly are
ACS Paragon Plus Environment

Page 3 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
restricted. Kv1.2 is the most prevalent in neuronal membranes where a portion occurs as a homoꢀ
tetramer and the majority is heteromerised with other Kv1 α subunits;11, 14 interestingly, there is a
preponderance of the less abundant Kv1.1 subunit in oligomers with Kv1.2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kv1 channels, exposed upon demyelination of axons in patients suffering from multiple sclerosis
(MS), contribute to abnormal propagation of nerve signals and resultant debilitating muscle
weakness.15 Although aminopyridines can inhibit the latter channels, such therapy is only
beneficial in the shortꢀterm; also, their blockade of other K+ channel types results in serious offꢀ
target effects, including seizures.15 Our observation of Kv1.1ꢀcontaining channels being
expressed ectopically in optic nerve demyelinated axons from mice fed cuprizone,16 mimicking
changes in MS, has indicated the subtypes present. In fact, the abnormal conductivity induced
could be nearꢀnormalised by attenuating the Kv1.1ꢀmediated currents with DTXk, a selective and
potent blocker of Kv1.1 homomeric channels in the picomolar range (see Table 1).16 Thus, Kv1.1
is a promising target for extracellular inhibitors to potentially ameliorate such symptoms in
demyelinated conditions, like MS.
Table 1. A list of selected mammalian Kv1.1channel blockers and their IC50 values
Name of the blocker
4ꢀAP
IC50 values (µM)
89–10015
3017, 18
20917
Tetraethlammonium
Flecainide
Diltiazem
14417
Nifedipine
9617
Capsaicin
2917
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 42
1
2
3
4
5
6
7
8
9
15 (this study)
2ꢀ([4ꢀ 1319
8
5,10,15,20ꢀtetrakis
tertꢀbutyl benzamido] ethyl
carbamate) porphyrin
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DTX
0.0217
MgTX
DTXk
0.0001413
0.00003–0.000615,16
More information, see the IUPHAR Compendium of VoltageꢀGated Ion Channels. (2016):
http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=538#ChannelBlockers and
http://www.guidetopharmacology.org/GRAC/ObjectInteractionsDisplayForward?objectId=538&familyId=81&type
=Channel%20blockers&structures=y. Notice that some of these blockers have crossꢀreactivity with other Kv1
channels
.
In search of low molecular weight blockers, a recombinant concatenation technology18, 20 was
utilized to express various Kv1 channel genes as single openꢀreading frames, after transfecting
human embryonic kidney (HEK)ꢀ293 cells. Such a strategy allows preꢀdetermination of not just
the combinations of α subunits but, also, their actual arrangements in the tetrameric channels
delivered to the plasmalemma. This is exemplified by the observed fastꢀinactivation of Kv1.4ꢀ
containing heteromeric channels only when the Kv1.6 α subunit, with its Nꢀterminal inactivation
prevention domain, is placed adjacent to Kv1.4 which possesses an Nꢀterminal inactivation
ball.18 In this way, we established the importance of the stoichiometry and positioning of α
subunits for interactions of inhibitory tetraꢀethylammonium with Kv1.1ꢀ and 1.2ꢀcontaining
concatenated tetramers that mimic those in brain.20, 21 Herein, channels composed of tandemꢀ
linked α subunits were utilised as representative targets for the development of new, small and
ACS Paragon Plus Environment

Page 5 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
selective inhibitors of Kv1.1. To aid their design, a homology model of rat Kv1.1 tetramer was
developed for molecular docking, based on the crystallographic structure of the mammalian
closelyꢀrelated Kv1.2.22 In this regard, it is noteworthy that symmetricallyꢀsubstituted porphyrins
bearing alkyl amino groups (cationicꢀcharged at physiological pH) can tightly bind a KcsAꢀ
Kv1.3 channel.23, 24 We recently reported that 4 porphyrin derivatives, with different alkyl
ammonium side chains (in length and geometry) inhibit Kv1.1 and/or 1.2 channels expressed in
mammalian cells.19 Now, a new generation based on DPM is described, with various
substitutions bearing alkyl ammonium sideꢀchains, which have been screened against
recombinantlyꢀgenerated Kv1 channels of known subunit structures. The resultant data were
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
used
to
design
a
novel,
small
inhibitor,
[2,2’ꢀ((5,5’ꢀ(diꢀpꢀtopyldiaryldiꢀ(2ꢀ
pyrrolyl)methane)bis(2,2’ꢀcarbonyl)bis(azanediyl)) diethaneamine.2HCl] (8), which proved
selective for Kv1.1ꢀcontaining channels found in rodent demyelinated axons.16
RESULTS
Rational design of a new selective inhibitor for Kv1.1 channels. In designing selective
blockers, their size and hydrophobicity are vital criteria because if made too small, they would
enter deep into the inner pore region which is conserved in all K+ channels and, similar to 4ꢀ
aminopyridine (4ꢀAP), selectivity would be lost. Although we have reported that various
substituted porphyrin derivatives are selective blockers of neuronal Kv1 channels,19 due to their
phototoxicity and high molecular weights, these are nonꢀideal therapeutic candidates25. However,
informed by structureꢀactivity relationships of the porphyrins and using molecular modelling,
valuable pharmacophore information was realised that afforded the development of a more
suitable lead structure. The goal was to rationally design an inhibitor which is 1) large enough to
avoid entering the deep inner pore region of the Kv1 channels, 2) target the essential amino acid
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 42
1
2
3
4
5
6
7
8
9
residues in the vicinity of their selectivity filter/inner turret region and 3) and lack absorption in
the visible region of the solar spectrum, thereby precluding phototoxicity. The scaffold of the
DPM derivatives, outlined in Scheme 1, was modelled using the same homology structure of
Kv1.1 as for the porphyrins. Geometric quality of the backbone conformations and energy
profiles of the modelled structures fall well within the restrictions established for reliable
structures.26ꢀ28 For refining the side chain of amino acids in the pore region of Kv1.1 channel,
flexible docking was performed with the Autodock protocol.29
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthetic route employed to prepare the DPMꢀbased K+ channel inhibitors 6 ꢀ 8. Three steps were used
to build the compounds from the DPM scaffold 1 (i) 4ꢀdimethylaminopyridine (DMAP), trichloroacetic acid,
dichloromethane (DCM); (ii) Nꢀboc protected diamine, triethylamine (TEA), DCM at room temperature; (iii) 4M
HCl in dioxane, DCM.
ACS Paragon Plus Environment

Page 7 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The alky ammonium sideꢀchains of 2, 3 and 4 carbon lengths in tetraphenyl prophyrin scaffold
revealed compounds with selective inhibition for currents mediated by Kv1.1 and 1.2 channels.19
Encouraged by these outcomes, we introduced these same sideꢀchains into the DPM scaffold to
give 6ꢀ8; note that the DPM scaffold represents half the size of a C4ꢀtetraphenyl porphyrin and
does not absorb visible light.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The predicted interactions of 6 with Kv1 α subunits in the Kv1.1 channel model (Figure 1a) via
hydrogen bonds (HB) are: Glu353 (two); Asp361 (one); Tyr375 (one); Gly376 (one); Tyr379
(two); Pro380 (one) and Val381 (one). Hydrophobic contacts include: Met378 (one) and Tyr379
(one). The modelling showed that the docked molecules interact with the outer region of the
channel (Figure 1a) mimicking that of DTXk, a renowned highꢀaffinity and absolutely specific
inhibitor of Kv1.1.16 This mechanism of binding differs from the characteristics of 4ꢀAP but can
participate in HB and πꢀπ stacking interactions. These advantageous features highlight that DPM
scaffold has promise as the basis for a new lead molecule.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. A docking model of Kv1.1 channel with either of two DPMs. Side views of the docking of 6 (a) and 7 (b)
(in yellow) to the extracellular pore region of Kv1.1 channel. Notice that the former interacts with the outer turret
region slightly offꢀcentre of the pore, while the latter binds more centrally, protruding deep enough to be close to the
selectivity filter of Kv1.1. The 4 Kv1.1 subunits are depicted with red, orange, blue and green colours for
clarification. Images were generated using PyMOL (The PyMOL Molecular Graphics System). Representative
electrophysiologicallyꢀrecorded current traces (lower panels) reveal that 6 caused limited inhibition of Kv1.1 and 1.3
[8 ± 2%; 8 ± 3%, (n = 3), respectively], with even lower reductions of Kv1.2, 1.4 and 1.6 currents [3 ± 1%; 3 ±
0.3%, 4 ± 1%, (n = 3), respectively]. More extensive blockade of Kv1.1, 1.3 and 1.4 was caused by 7 [35 ± 3%; 32 ±
2%, 16 ± 6%, (n = 3), respectively], with minimal decreases in Kv1.2 and 1.6 [6 ± 1%; 5 ± 2%, (n = 3),
ACS Paragon Plus Environment

Page 9 of 42
Journal of Medicinal Chemistry
1
2
3
4
respectively]. (c) Histograms showing the extents of inhibition of the different Kv1 channels by 6 and 7, reveals the
5
6
superiority of the latter over the former.
7
8
9
Similar channel interactions with the Kv1.1 α subunits were observed for 7 (Figure 1b) i.e. HB
with: Glu353 (one); Gly374 (all); Tyr375 (two); Gly376 (two); Asp377 (two); Tyr379 (one);
Pro380 (one) and Val381 (one). Again, hydrophobic contacts involve Tyr379 (two subunits).
Hence, both compounds were predicted to contact Glu353 of the outer turret and a range of
residues in the inner turret. Moreover, the arm of 7 was expected to be positioned slightly deeper
into the inner turret (e.g. making contact with Gly374 of all subunits) than 6. Predicted
interactions of 8 with Kv1.1 α subunits in the channel (Figure 2a) involve HB with: Val373
(two) and Gly374 (three) through the peptide bond; Tyr375 (three); Gly376 (two); Asp377 (two)
and Tyr379 (two). Note that these residues reside in the inner pore, and includes piꢀpi stacking
interactions with Tyr379 of Kv1.1 being key. Guided by the outcomes from these modelling
studies, this new generation of 6ꢀ8 was prepared with various substitutions, bearing alkyl
ammonium sideꢀchains, and screened against expressed recombinant Kv1 channels of known
subunit structures.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. 8 Showed preferential selectivity for Kv1.1 channel. (a) A closeꢀup view illustrates modelling of the
docking of 8 (in yellow) with the α subunits of monomeric Kv1.1 channel. This docking is seen in the upper panel
and after horizontal 180° rotation (bottom panel). Interacting amino acids from each α subunit of the 4 channels are
represented in orange, red, blue and green colours. Notice the interaction of sideꢀchains in 8 are mainly with 2
subunits. Images were generated using PyMOL. (b) Representative current traces from Kv1.1 channel in the absence
(black) and presence (green) of 10 ꢀM 8 showing inhibition which was relieved on washing. Currents were evoked
at 20 mV voltage steps from ꢀ90 mV holding potential. (c) Recordings demonstrate a lack of significant effect of 8
on Kv1.2, 1.4 and 1.6 [1 ± 0.5% (n = 3); 5 ± 1% (n = 4), 1 ± 0.5%, (n = 3), respectively], but some inhibition of
Kv1.3 [15 ± 5%, (n = 3)]. Histogram confirms the preferential inhibition by 8 of Kv1.1 [40 ± 3%, (n = 9)] over the
other channels tested.
ACS Paragon Plus Environment

Page 11 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Preparation of 6ꢀ8. Their synthesis (Scheme 1) first involved preparing a DPM scaffold by
condensation of pyrrole and 4,4ꢀdimethylbenzophenone in the presence of boron
trifluorodietherate [BF3(OEt)2] to yield compound 1 in 50% yield. Introduction of the required
carboxyl groups into compound 1 at the 2 position of pyrrole was achieved in quantitative yields
by electrophilic diꢀsubstitution of 1 with trichloroacetic anhydride (TClAA) in the presence of a
catalyst DMAP to generate compound 2. The latter was converted to the NꢀBoc protected
precursors 3ꢀ5, by stirring 2 with the appropriate mono NꢀBoc protected alkyl diamine and TEA
at room temperature overnight, to give compounds 3ꢀ5 in quantitative yield without need for
further purification. Deprotection of compound 3ꢀ5 was achieved using 4M HCl in dioxane to
give protonated derivatives 6ꢀ8, in quantitative yields. The identity and purity of 6ꢀ8 were
confirmed by highꢀresolution mass spectrometry and NMR (given in Supporting Information). It
should be noted that attempts were made to prepare various DPM derivatives containing various
substituents in the the phenyl groups, these attempts failed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Relative inhibition by 6ꢀ8 of Kv1.1ꢀ and Kv1.2ꢀcontaining channels. To measure the
reactivity of 6ꢀ8 with Kv1 channels, those containing the major α subunits found in mammalian
11, 14
brain
were stably expressed in HEKꢀ293 cells and their K+ currents recorded
electrophysiologically, using wholeꢀcell voltage patchꢀclamp. Representative K+ current traces,
in the absence or presence of 10 ꢀM 6 (Scheme 1), revealed limited inhibition of Kv1.1 and
Kv1.3, with negligible effects on Kv1.2, 1.4 or 1.6 channels (Figure 1c). 7 (Scheme 1), which
possesses a shorter alkyl chain than 6, gave much more inhibition of Kv1.1 at 10 ꢀM (35 ± 3 %,
n = 3) and a lesser extent for Kv1.3 (16 ± 6 %, n = 3) channels whilst being ineffective towards
Kv1.2 or 1.6 homomers (Figure 1c). Unexpectedly, the Kv1.4ꢀmediated current was
considerably reduced by 10 ꢀM 7 (32 ± 2 %, n = 3) (Figure 1c).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 42
1
2
3
4
5
6
7
8
9
The final candidate, 8 (Scheme 1), which has the shortest chain, caused a preferential block (40
± 3 %, n = 9) of Kv1.1 homomeric channel which could be reversed upon washing (Figure 2b,
d). Its selectivity was highlighted by a lack of effect on Kv1.2, 1.4 or 1.6, though some inhibition
(~20%) of Kv1.3 was apparent (Figure 2c, d). During some recordings in the presence of DPM
compounds, HEK cells were subjected to a UV source for less than a minute. No functional
damage was detected under such conditions demonstrating that these do not act as
photosensitisers.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analysis of different heteroꢀtetramers of tandemꢀlinked Kv1.1 and 1.2 stably expressed in HEKꢀ
293 cells yielded varying susceptibilities to the blocker. Decreasing the copy number of Kv1.1 in
those concatamers lowered the extent of inhibition by 10 ꢀM 8 (Figure 3a): Kv(1.1)4 > Kv(1.1)3ꢀ
1.2 > Kv(1.1)2ꢀ(1.2)2 with Kv(1.1)ꢀ(1.2)3 and Kv(1.2)4 being insensitive (Figure 3a, b).
ACS Paragon Plus Environment

Page 13 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Concatenated Kv1.1ꢀcontaining channels displayed different sensitivities to 8. (a) Representative current
traces from Kv1.1/1.2ꢀcontaining tetrameric channels in the absence (black) and presence (green) of 10 ꢀM 8. It
partially inhibited these containing 2, 3 or 4 copies of Kv1.2 subunits, while proving ineffective against Kv1.2
homoꢀtetramer or Kv1.2ꢀ1.2ꢀ1.1ꢀ1.2 channels. (b) Doseꢀresponse curves for Kv(1.1)4 (○) and Kv1.1 (■) demonstrate
similar sensitivity, followed by Kv1.1ꢀ1.2ꢀ1.1ꢀ1.1 (▲) and Kv1.1ꢀ1.1ꢀ1.2ꢀ1.2 (□), while Kv1.2ꢀ1.2ꢀ1.1ꢀ1.2 (△) and
Kv(1.2)4 (●) were nonꢀsusceptible. Some of the error bars fall within the data points. IC50 values are summarised in
Table 2. (c) Summary of equal inhibition by 8 of tetrameric channels containing 2 copies of Kv1.1 in different
positioning within the concatamers [Kv1.2ꢀ1.1ꢀ1.2ꢀ1.1 = 15 ± 1% (n = 6); Kv1.1ꢀ1.1ꢀ1.2ꢀ1.2 = 13 ± 2% (n = 4);
Kv1.2ꢀ1.2 ꢀ1.1ꢀ1.1 = 15 ± 2%, (n = 5)].
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 42
1
2
3
4
5
6
7
8
9
The doseꢀdependencies for blockade revealed that the IC50 values for 8 decreased from 14 ꢀM
for Kv(1.1)4 to 54 and 57 ꢀM upon introducing 1 and 2 copies of Kv1.2 (Table 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. IC50 values for inhibition by 8 of Kv1.1 homomer compared to these for various
concatenated tetramers of Kv1.1, Kv1.2 and botha.
Channel
IC50
Hill slope
N
[µM]
Kv1.1
16 ± 1
14 ± 1
*54 ± 7
*57 ± 15
>100
1.4 ± 0.2
1.5 ± 0.2
1.5 ± 0.1
1.3 ± 0.2
ꢀꢀꢀ
8
8
7
5
6
4
Kv(1.1)4
Kv(1.1)3ꢀ1.2
Kv(1.1)2ꢀ(1.2)2
Kv1.1ꢀ(1.2)3
Kv(1.2)4
>100
ꢀꢀꢀ
aResults are represented as means ± S.E.M. *Values are significant compared to Kv(1.1)4, P <
0.05, unpaired tꢀtest.
Notably, altering the positions of Kv1.1 and 1.2 within the concatamers did not influence their
blockade by 10 ꢀM 8 (Figure 3c). Interestingly, inhibition of Kv1.1 channels by 8 was
associated with significant slowing (~10ꢀfold) of activation (Figure 3a, 4a). The time constants
(τactivation) for activation of Kv1.1 currents were slowed from 1.9 (± 0.15 ms, n = 4) to 20.3 (± 3.3
ms, n = 4), p < 0.005 (Figure 4a) for the homomer, and 2.2 (± 0.1 ms, n = 5) to 19.5 (± 3.4 ms, n
= 4), p < 0.001 for the Kv1.1 concatamer (data not shown). Likewise, τactivation values for the
heteromeric channels were increased by 10 ꢀM 8 for channels in which the number of Kv1.1
ACS Paragon Plus Environment

Page 15 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
subunits was raised to 2 in Kv(1.1)2ꢀ(1.2)2 [from 2.8 ± 0.4 ms, n = 6 to 18 ± 2 ms, n = 5] and 3 in
Kv(1.1)3ꢀ1.2 [from 3.1 ± 0.2 ms, n = 5 to 14 ± 2 ms, n = 5], derived from the representative
current traces shown (Figure 3a). Moreover, the gV relationship of Kv1.1 channel was altered
by 10 ꢀM 8 from a halfꢀmaximal value for activation (V1/2) of ꢀ27 (± 1 mV, n = 7) to 11 (± 1
mV, n = 4) (Figure 4b), a significant shift of ~40 mV towards positive potentials. Collectively,
these findings unveil notable inhibition by 8 of Kv1ꢀcontaining K+ channels and compounded
modulation of the biophysical properties of Kv1.1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Activation kinetics of the K+ current mediated by Kv1.1 channel are slowed and the threshold potential is
raised by 8. (a) Time constants of activation (τactivation) for homomeric Kv1.1 before (ꢀ) and after application of 10
ꢀM 8 (ꢁ), determined by fitting the current traces with a monoꢀexponential function. Notice a significant slowing in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 42
1
2
3
4
5
6
7
8
9
the timeꢀcourse of activation [from 1.9 ± 0.15 ms (n = 4) to 20.3 ± 3.3 ms (n = 4); (p < 0.001)]. (b) Conductanceꢀ
voltage relationship of the steadyꢀstate currents for homomeric Kv1.1 (○) showing a significant shift towards more
positive potentials induced by 8 (●) [from ꢀ27 ± 1 mV (n = 7) to 11 ± 1 mV (n = 4); (p < 0.001)]. Some of the error
bars fall within the data points.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 8 raises the neuronal excitability of demyelinated fibres in brain slices.
Excitability of demyelinated fibres in the corpus callosum (CC) of cuprizoneꢀtreated mice was
increased by 8. Luxol staining revealed that myelination of the CC from mice fed a cuprizoneꢀ
containing diet was substantially reduced (data not shown), consistent with our previous report.16
In accord with the distinct peaks of compound action potentials (cAPs) reported for CC having
been attributed to mixed myelinated (N1) and unmyelinated (N2) axons,30, 31 the cAPs recorded
there (Figure 5a) show the expected N1 and N2 waves (Figure 5b).
ACS Paragon Plus Environment

Page 17 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 5. 8 Improves the excitability of demyelinated fibres in CC slices of mice fed a diet containing cuprizone.
(a) Schematic showing the positioning of stimulating and recording electrodes in a brain coronal slice to record CC
cAPs. (b) Examples of cAPs recorded in slices from control and cuprizoneꢀtreated mice. The myelinated N1 (1) and
the unmyelinated N2 (2) waves of the CC cAPs are shown. Tetradotoxin (TTX) sensitivity of cAPs was recorded in
the same slice taken from a cuprizoneꢀtreated mouse. (c) The I/O curve shifted to the left in slices taken from
cuprizoneꢀtreated mice (n = 6 mice, red curve) when compared to control animals (n = 4 mice, green curve). **, ***
indicate significant differences when comparing cuprizoneꢀtreated group vs. control group. (d) Representative
recordings of cAPs in the same slice from cuprizoneꢀtreated mice before (baseline) and the increase in excitation
(dashed lines) after application for 10 min of 1 and 10 µM 8 (a significant difference is seen between the dashed
lines, p<0.05).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The extraꢀcellular recordings (Figure 5) of cAPs correspond to the sequential activation of
myelinated and unmyelinated fibres (Figure 5b); blockade of both by TTX30 verified their
identities (Figure 5b). It is noteworthy that after cuprizone treatment the amplitude of the N1
wave was drastically reduced whereas that of the N2 increased (Figure 5b), consistent with
demyelination having occurred. Under control conditions, increasing the electrical stimulus
intensity produced an incremental increase in the amplitude of the N2 wave, in all slices from
control and cuprizoneꢀtreated mice (Figure 5c). A leftward shift of the input/output curve was
noted in the cuprizone group, as expected for demyelinated fibers (Figure 5c). Moreover, there
was a significant increase in the normalized N2 wave in slices from the cuprizoneꢀtreated relative
to control animals at stimulus intensities of 2 and 3 volts, with p < 0.001 and < 0.01,
respectively, derived from two way ANOVA test. Initial experiments on control slices showed
that neither 8 nor 4ꢀAP exerted any significant effect on the N2 amplitude relative to the baseline
(with respective values for 1 and 10 µM 8 being 100.9 ± 2.8 % and 102.6 ± 1.9 %; and for 1 and
10 µM 4ꢀAP 105.3 ± 7.3 % and 105.7 ± 4 %, with n = 4 and df = 3, p > 0.05 for all) is
consistent with published findings.32 On the other hand, in slices from cuprizoneꢀtreated mice, 1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 42
1
2
3
4
5
6
7
8
9
µM 8 slightly increased the amplitude of N2 (Figure 5d), though the difference was not
significant from the baseline (108.2 ± 4.3 %, n = 6, with t = 1.93, df = 5, p > 0.05). However,
application of 10 µM 8 increased the amplitude of the N2 significantly relative to the baseline
(127.8 ± 9 % with n = 6, t = 3.10, df = 5 and p < 0.05) (Figure 5d). In the cuprizoneꢀtreated
animal group, no significant difference was observed in the N2 amplitude after perfusing the
slices with 1 or 10 µM 4ꢀAP when compared to 8 (Supporting Figure 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DISCUSSION
These data show that 8 interacts with at least 2 Kv1.1 subunits in heteromeric channels and the
effect becomes accentuated by increasing their number. Molecular modelling of the porphyrins19
yielded valuable pharmacophore information on their channel binding affinities, especially the
types of interactions that could be crucial for inhibition of Kv1.1. The resultant data on the
porphyrins suggested that the lack of selectivity for particular channel subtypes could arise from
the large number of interactions with conserved residues present in all Kv1 channels. As a first
consideration of the necessary selectivity for Kv1.1 channels, size of the scaffold to which the
sideꢀarm alkyl ammonium groups are attached needed to be reduced. Rational drug design for
many voltageꢀgated ion channels has been hampered by a paucity of adequate crystal structures,
and the challenge of achieving inhibitors exhibiting subtype selectivity. In this work, a successful
combination of synthetic chemistry, molecular modelling and electrophysiology have led to 8, a
DPM molecule bearing two C2ꢀalkyl ammonium sideꢀchains.
The specific alignment of the backbone residues in the selectivity filter region of Kv1 channels is
essential for their conductance. As previously reported by Ader et al., conformational alterations
occurring in the HB network of the selectivity filter has a direct influence on K+ current gating.23
ACS Paragon Plus Environment

Page 19 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The residues that are located in this region are Val373, Gly374, Tyr375, Gly376 and Asp377. A
logical design of 6ꢀ8 was undertaken and the derivatives assembled with the aid of molecular
docking to disrupt the HB network of these residues deep in the pore. Residue Tyr379 in the
inner turret region of Kv1.1 was a key interaction to target due to its distinction from the other
Kv1 channels (e.g. Val in Kv1.2; His in Kv1.3; Lys in Kv1.4; and Tyr in Kv1.6). Interestingly,
the His in Kv1.3 may play a similar role in ligand binding to Tyr 379 of Kv1.1, partially
explaining the level of inhibition by 7 and 8 compared to other Kv1 subtypes. Compound 8,
which exhibited the best selectivity of the three compounds screened, possesses groups with the
potential to create HB, salt bridges and πꢀπ stacking between the residues of interest. Compound
8 (Figure 2a) binds to the outer region of the channel with its reactive groups reaching the pore.
This is the type of binding that is required to mimic the striking specification of DTXk,
establishing that this scaffold is a promising constituent for new lead molecules. Although
predominant cardiac Kv channels, such as Kv1.5 and hERG, were not tested, the pore alignment
of these channels with Kv1 members indicates that 8 might not interact with these channels (see
Supporting Figure 1). The key residue (Tyr379) of Kv1.1 is replaced by Ser431 in hERG channel
(in addition to other dissimilar residues at the selectivity filter region) and Arg489 in Kv1.5
channels. These findings support the notion that 8 should not exert adverse effects on cardiac
function. Also, Kv2.1 gives residue alignment similar to that of Kv1.2 at the outer turret region
but with unique residue difference at the key location (Thr380, see Supporting Figure 1); hence,
no inhibition is expected of Kv2.1 channel by 8.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The functional properties of 8, measured electrophysiologically, accord with those predicted
from in silico analysis. This compound is devoid of reactivity with Kv1.2, 1.4 and 1.6 channels
and preferentially inhibits Kv1.1 but with less blockade of 1.3 channel. Evaluation of this
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 42
1
2
3
4
promising molecule on Kv1.1/1.2ꢀheteromeric channels was possible by utilising a cloning
5
6
strategy that affords expression of 4 α subunits as singleꢀchain proteins on the plasmalemma,
7
8
9
21
with the compositions and positions of their subunits preꢀdetermined.20,
Tetrameric
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
combinations that should correspond to those predicted to occur in brain11, 14 were selected for
investigation, including those Kv1.1ꢀenriched channels observed in demyelinated axons.16
Selectivity of 8 for inhibiting Kv1.1 channel arise, in large part, from acting extracellularly,
namely with the inner turret regions of Kv1.1 before occluding the channel deep in its vestibule.
This explains its limited crossꢀreactivity with Kv1.3, which has accessible serines closely
resembling those lining the outer turret of Kv1.1 and only one residue difference (His404, see
Supporting Figure 1) at the key location.19 For comparison, 5,10,15,20ꢀtetrakis 2ꢀ([4ꢀtertꢀbutyl
benzamido] ethyl carbamate) porphyrin bearing similar alkyl ammonium sideꢀchains to that of 8
was selective in blocking Kv1.1 and 1.419, two subunits predominantly present in CNS, with IC50
values of 13 and 21 ꢀM, respectively16. Advantageously, 8 shows an equivalent inhibition of
Kv1.1 but lacks reactivity towards Kv1.4. On the other hand, Kv1.1 and 1.2 channels share
similar low susceptibility to 4ꢀAP due to their conserved inner vestibule region where it
interacts.8, 33 Furthermore, the selective inhibition and slowing of activation kinetics of the Kv1.1
channel highlights the dual advantages of 8. In comparison, the above mentioned pophyrin also
showed this dual effect towards Kv1.1 channel but induced a smaller modification of the
activation kineitcs 19. Bagchi et al., demonstrated that in demyelinated axons of mouse optic
nerve an increase in ectopic expression of Kv1.1 subunit contributes to abnormal cAPs.16 These
cAPs decrease the voltage threshold (activate at more negative potentials) and accelerate the
activation kinetics, thereby, perturbing the axonal propagation of electrical signals; accordingly,
it was found that this dysfunction can be partially overcome by the Kv1.1ꢀselective inhibitor,
ACS Paragon Plus Environment

Page 21 of 42
Journal of Medicinal Chemistry
1
2
DTXk.16 Although such an avid toxin cannot be considered as a potential neurotherapeutic, its
beneficial effects in vitro warranted the development of more suitable smaller inhibitors. Thus, it
is encouraging that 8 offers a double advantage in preferentially inhibiting Kv1.1ꢀdependent
currents, including slowing of their activation and shifting the conductanceꢀvoltage relationship
of Kv1.1 to more positive potentials. Such a beneficial modulatory action could underly its
observed normalisation of the cAPs in demyelinated axons, a discovery of relevance to pathoꢀ
physiology of MS (see below). Its effect on the activation kinetics could be explained by
interaction deep in the pore region of Kv1.1 channels. These interactions are in the vicinity of a
similar innerꢀpore region found in Kv7.2 channel which can be blocked by ((E)ꢀNꢀ[(S)ꢀ1ꢀ(4ꢀ
Cyclopropylmethylꢀ3,4ꢀdihydroꢀ2Hꢀ1,4ꢀbenzoxazinꢀ6ꢀyl)ꢀethyl]ꢀ3ꢀ(2ꢀfluoroꢀphenyl)ꢀNꢀ
methylacrylamide) and its activation kinetics slowed.34 Experiments on mutants of Kv7.2
channel revealed that the methyl group on the amide nitrogen of the aboveꢀnoted blocker binds
within the “retigabine binding site”. This inner pocket of the channel was shown to interact with
components of the selectivity filter in an open state manner and this leads to slowing of the
channels’ activation kinetics.35 A similar mechanism might apply to 8; however, further work
would have to be done to clarify this notion. Furthermore, 8 is an extracellular blocker
impermeable to intact cell membranes (measured logP value of ꢀ1.05 which is highly
unfavourable for membrane diffusion) that can distinguish between Kv1.1 and the much more
prevalent 1.2 channels. In contrast, the internal blocker, 4ꢀAP, is readily membrane permeable,
interacts with both Kv1.1 and 1.2 channels with similar potency and can cross the bloodꢀbrain
barrier. In summary, 8 advantageously exhibits a dual action in selectively inhibiting K+ currents
which would be expected to reduce the unwanted hyperpolarisation caused by homomeric Kv1.1
and Kv1.1ꢀcontaining channels found to be enriched after demyelination of optic nerve axons.16
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 42
1
2
As lesions in the CC occur in over 90% of MS cases,36 this is a pertinent structure for studying
demyelination and its functional consequences. cAP recordings in the CC revealed two
downward phases (N1 and N2), corresponding to activation of the myelinated and unmyelinated
axons.30 31 Both components were abolished by TTX (see Figure 5b). In the cuprizoneꢀtreated
group, the amplitude of the N1 wave was lower with the N2 magnitude being greater. Similarly,
the observed leftwards shift of the input/output curve for the amplitude of the N2 wave in slices
from the experimental mice indicates that the relative number of unmyelinated fibres had
increased. In animal models of demyelination, it is known that demyelinated/dysmyelinated
axons become initially exposed and that lateral diffusion of juxtaꢀparanodal and nodal voltageꢀ
gated ion channels can extend into the formerly myelinated internodal regions.32, 37 Such
reorganization partially contributes to the loss of saltatory conduction and slowing of action
potential propagation in white matter axons.38 In addition, after cuprizoneꢀinduced
demyelination, we observed in earlier work that Kv1.1 makes a more prominent contribution
than Kv1.2 subunits in regulating the excitability and conductivity of optic nerve.16 It is
noteworthy that Kv1.1 and 1.2 channels are abundantly expressed in the CC of mice,39 activate at
potentials near the threshold for action potential generation and are known to regulate firing
patterns in various regions of the CNS.40 Herein, the new specific Kv1.1 blocker developed was
found electrophysiologically to have an IC50 at least ten times lower than that of 4ꢀAP15 (c.f.
Table 1). As our results also established that 8 increases the N2 amplitude in slices from
cuprizoneꢀtreated mice, it may trigger the activation of other channels and elevate the overall
excitability of the demyelinated fibres in the CC, through inhibition of Kv1.1; this could offer a
substantial advantage for ameliorating the symptomatic phase of MS. The supporting data from
the in vitro recording of selected Kv1.1 currents, with its limited data acquisition in comparasion
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 23 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with the high pulseꢀfrequency of the cAP recordings, showed that 8 inhibition of Kv1.1 current, a
dominant component of Kv currents, leads to an excitation of demyelinated nerves that would be
therapeutically desireable.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Despite its therapeutic use, 4ꢀAP treatment causes many sideꢀeffects from nausea to
paraesthesia,41 as well as being a potent convulsant; therefore, its utilization in patients is limited.
Unlike 4ꢀAP, 8 is more specific and potent blocker of recombinant Kv1.1, and devoid of effect
on Kv1.2. As the inflammatory component in MS pathophysiology is of prime importance, the
noted inhibition of Kv1.3 by 8 might have an additional beneficial effect because this channel is
known to activate T cells and is highly expressed in the inflammatory infiltrates in MS;
moreover, Kv1.3 is a functional marker of activated effector memory T cells in experimental
allergic encephalomyelitis and in myelinꢀspecific T cells derived from the peripheral blood of
MS patients.
CONCLUSIONS
In summary, a newlyꢀdesigned small inhibitor, 8, is proven to be selective for Kv1.1ꢀcontaining
channels found in rodent demyelinated axons.16 Furthermore, cAP recordings inCC taken from
cuprizoneꢀtreated mice revealed that this novel compound increases the excitability of the
demyelinated fibres. Hence, 8 could have therapeutic potential for the symptomatic phase of MS.
EXPERIMENTAL SECTION
Computational methods. The crystal structure of the rat Kv1.2 channel (PDB, 2A79) has
previously been used as a template for building homology models of Kv1 homoꢀtetramers and
42ꢀ47
heteroꢀtetramers.23,
Protein structures were downloaded from the protein databank
(www.rcsb.org). Accelrys Discovery Studio 3.5 was used to prepare the protein structures (add
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 42
1
2
3
4
5
6
missing atoms, correct connectivity, correct names, etc.) For ligands, we employed Accelrys
Discovery Studio 3.5 to enumerate tautomers, stereoisomers and conformations.
7
8
9
Homology modelling. In the case of rat Kv1.1 channel, a homology model was developed and
validated, as previously described.19 In brief, the protein sequence for Kv1.1 was aligned to the
template of the known structure (PDB: 2A79) 20 using Accelrys Discovery Studio 3.5, and 100
protein structures were built for the selected alignment. The best model was selected using a
combination of the Modeller discrete optimized protein energy (DOPE) score and a collection of
protein assessment tools.26ꢀ28 Tetramer models were developed using PyMOL and the biological
assembly of the Kv1.1/1.2 channel structures.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molecular docking. AutoDock 4, based on a Lamarckian genetic algorithm method was utilised
for molecular docking.48, 49 Ligands were allowed to be flexible and flexible residues for the
protein were defined as residues 375ꢀ381 in each monomer.19 A cubic grid of
140 Å × 140 Å × 140 Å around the ion pore was constructed using the Autogrid program, with a
grid point step of 0.375 Å. Default parameters were implemented in Autodock4 with 50 docking
runs, while ga_num_evals was set to 25,000,000 and ga_num_generations fixed at 27,000. To
generate input files, the AutoDockTools program was used.29 The conformations showing a
lower free energy of binding for each ligand were further analyzed.
Synthesis of 1ꢀ8.
Diꢀpꢀtolyldiꢀ(2ꢀpyrrolyl)methane (1). A 250 mL 2ꢀnecked round bottomed flask was charged
with 1.31 g of 4,4'ꢀdimethylbenzophenone (6.18 mmol). Anhydrous methanol (50 mL) was
added and the mixture magnetically stirred under an argon atmosphere. When the 4,4'ꢀ
dimethylbenzophenone was fully dissolved, freshly distilled pyrrole (1.07 mL, 15.5 mmol) was
ACS Paragon Plus Environment

Page 25 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
added dropwise. After 5 minutes, BF3.(OEt)2 was added (1.07 mL, 8.6 mmol) and the reaction
mixture allowed to stir for 5 days at room temperature. The precipitate formed was filtered and
washed with cold methanol to give a white solid (1.01g, 50% yield).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz) δ (DMSOꢀd6) 10.2 (2H, s, pyrroleꢀNH) 7.1 (2H, d, phenylꢀH) 6.8 (2H, s,
phenylꢀH) 6.7 (2H, m, pyrroleꢀH) 5.9 (2H, m, pyrroleꢀH) 5.6 (2H, m, pyrroleꢀH) 2.3 (6H, s,
13
TolylꢀCH3). C NMR (100 MHz, DMSOꢀd6) δ 143.6, 135.7, 135.1, 129.0, 127.8, 117.9, 108.7,
105.9, 54.7, 20.3.
5,5'ꢀ(Diꢀpꢀtolyldipyrromethane)bis(2ꢀtrichlorocarbonyl) (2). To a 25 mL 2ꢀneck round
bottomed flask, 336 mg of 1 (1 mmol) and 24 mg of DMAP (0.2 mmol) were added, stirred and
placed under an argon atmosphere. Anhydrous DCM (10 mL) was then added and the mixture
cooled to 0oC. To the stirring solution, trichloroacetic anhydride (457 µL, 2.5 mmol) was applied
dropwise and the reaction allowed to stir for an additional 2 hours at room temperature. Then, the
reaction mixture was transferred to a small separating funnel and the organic phase washed twice
(2 x 10 mL water), followed by (2 x 10 mL) NaHCO3 and dried over MgSO4. The DCM was
removed to give 2 as a white solid in quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 12.1 (2H, s, pyrroleꢀNH) 7.29 (2H, d, pyrroleꢀH) 7.28 (2H, d,
13
phenylꢀH) 6.9 (2H, d, phenylꢀH) 6.1 (2H, m, pyrroleꢀH) 2.3 (6H, s, tolylꢀCH3). C NMR (100
MHz, DMSOꢀd6) δ 171.8, 145.8, 140.4, 136.4, 129.0, 128.5, 122.5, 120.8, 113.9, 95.2, 55.9, 20.3
General procedure for preparing NꢀBoc protected DPM from compounds 3ꢀ5 (described
below)
To a 50 mL round bottom flask, 400 mg (0.65 mmol) of 2 were added into 5 mL of DCM, and
the mixture allowed to stir for 5 mins at room temperature, under an argon atmosphere. Nꢀboc
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 42
1
2
3
4
5
6
7
8
9
ethylenediamine (250 µL; 1.6 mmol) was then added to the mixture followed by the dropwise
addition of anhydrous triethylamine (360 µL; 2.6 mmol). The reaction was stirred at room
temperature overnight, the resultant precipitate was collected by suction filteration and washed
twice with 25 mL aliquots of DCM before being dried under vacuum to give the NꢀBoc
derivatives in quantitative yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diꢀtertꢀbutyl
(((5,5'ꢀ(diꢀpꢀtolyldiaryldiꢀ(2ꢀpyrrolyl)methane)bis(2,2'ꢀcarbonyl))
bis(azanediyl))bis(butaneꢀ4,1ꢀdiyl))dicarbamate (3). The general procedure was used with 400
mg (0.65 mmol) of 2 and 290 µL (1.6 mmol) of Nꢀboc 1, 4 butanediamine to give 3 in
quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 10.7 (2H, s, pyrrole NH) 8.0 (2H, t, J= 5.6Hz, amide NH) 7.1
(4H, d, J = 8.0 Hz, phenylꢀH) 6.8 (4H, t, J = 8.0 Hz, phenylꢀH+ 2H amideꢀNH overlapping) 6.6
(2H, d, J = 3.6 Hz, pyrroleꢀH) 5.6 (2H, d, J = 3.6 Hz, pyrroleꢀH) 3.2 (4H, q, J = 6.0 Hz, CH2) 2.9
(4H, q, J = 6.0 Hz, CH2) 2.3 (6H, s, tolꢀCH3) 1.41 (8H, m, CH2) 1.4 (18H, s, BocꢀCH3). 13C
NMR (100 MHz, DMSOꢀd6) δ 160.3, 155.7, 142.0, 138.7, 135.9, 129.2, 128.2, 127.1, 110.8,
109.7, 77.4, 55.2, 55.0, 39.4, 38.3, 28.3, 27.2, 26.9, 20.6
Diꢀtertꢀbutyl
(((5,5'ꢀ(diꢀpꢀtolyldiaryldiꢀ(2ꢀpyrrolyl)methane)bis(2,2'ꢀ
carbonyl))bis(azanediyl))bis(propaneꢀ3,1ꢀdiyl))dicarbamate (4). The general procedure was
used with 400 mg (0.65 mmol) of 2 and 270 µL (1.6 mmol) of Nꢀboc 1, 3 propanediamine to
give 4 in quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 10.8 (2H, s, pyrrole NH) 8.0 (2H, t, J = 5.3 Hz, amide NH)
7.1 (4H, d, J = 8.2 Hz, phenylꢀH) 6.8 (4H, t, J = 8.2 Hz, phenylꢀH+ (2H) amideꢀNH overlapping)
6.6 (2H, d, J = 3.6 Hz, pyrroleꢀH) 5.7 (2H, d, J = 3.6 Hz, pyrroleꢀH) 3.2 (4H, q, J = 6.4 Hz, CH2)
ACS Paragon Plus Environment

Page 27 of 42
Journal of Medicinal Chemistry
1
2
13
3
4
5
6
7
8
9
2.9 (4H, q, J = 6.4 Hz, CH2) 2.3 (6H, s, tolꢀCH3) 1.6 (4H, p, CH2) 1.4 (18H, s, BocꢀCH3). C
NMR (100 MHz, DMSOꢀd6) δ 160.4, 155.6, 141.9, 138.7, 135.9, 129.2, 128.2, 126.9, 110.8,
109.7, 77.5, 55.2, 54.9, 37.7, 36.2, 29.9, 28.3, 20.6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diꢀtertꢀbutyl(((5,5'ꢀ(diꢀpꢀtolyldiaryldiꢀ(2ꢀpyrrolyl)methane) bis(2,2'carbonyl))
bis(azanediyl))bis(ethaneꢀ2,1ꢀdiyl)) dicarbamate (5). The general procedure was used with
400 mg (0.65 mmol) of 2 and 257 µL (1.6 mmol) of Nꢀboc 1, 2 ethaneanediamine to give 5 in
quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 10.8 (2H, s, pyrrole NH) 8.0 (2H, t, J = 5.3 Hz, amide NH)
7.1 (4H, d, J = 8.2 Hz, phenylꢀH) 6.9 (2H, m, amideꢀNH) 6.8 (4H, d, J = 8.2 Hz, phenylꢀH) 6.4
(2H, d, J = 3.3 Hz, pyrroleꢀH) 5.7 (2H, d, J = 3.3 Hz, pyrroleꢀH) 3.2 (4H, q, J = 6.0 Hz, CH2) 3.0
(4H, q, J = 6.0 Hz, CH2) 2.3 (6H, s, tolꢀCH3) 1.4 (18H, s, BocꢀCH3). 13C NMR (100 MHz,
DMSOꢀd6) δ 160.6, 155.7, 141.9, 138.9, 135.9, 129.2, 128.2, 126.9, 110.9, 109.9, 77.7, 55.2,
54.9, 39.5, 38.7, 28.3, 20.6.
General procedure for the deprotection of compound 3ꢀ5 to give 6ꢀ8
To a 25 mL round bottomed flask, the boc protected diaryldiꢀ(2ꢀpyrrolyl)methanes (100 mg)
were added. The flask was then placed under an argon atmosphere and anhydrous DCM (10 mL)
was added with stirring. After cooling the reaction mixture to 0oC, 4M HCl (1 mL) in dioxane
was added dropwise and allowed to stir overnight at room temperature. The formed precipitate
was then collected by suction filtration and washed with DCM to give white solids.
4ꢀ(5ꢀ((5ꢀ((4ꢀAminobutyl)carbamoyl)ꢀ1Hꢀpyrrolꢀ2ꢀyl)diꢀpꢀtolylmethyl)ꢀ1Hꢀpyrroleꢀ2ꢀ
carboxamido)butanꢀ1ꢀaminium chloride (6). The general procedure was followed using 3
(100 mg) to yield 6 as a white solid in quantitative yield.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 42
1
2
1H NMR (400 MHz) δ (DMSOꢀd6) 11.0 (2H, s, pyrroleꢀNH) 8.1 (2H, t, J = 8.0 Hz, amideꢀNH)
7.9 (6H, brs, NH3) 7.1 (4H, d, J = 8.0 Hz, phenylꢀH) 6.8 (4H, d, J = 8.0 Hz, phenylꢀH) 6.7 (2H,
m, pyrroleꢀH) 5.7 (2H, m, pyrroleꢀH) 3.2 (4H, q, J = 5.8 Hz, CH2) 2.8 (4H, q, J = 5.8 Hz, CH2)
2.3 (6H, s, CH3), 1.7 (8H, p, CH2)13C NMR (100 MHz, DMSOꢀd6 ) δ 160.3, 141.9, 138.8, 135.9,
129.2, 128.2, 127.0, 110.9, 109.8, 55.2, 38.5, 37.8, 26.4, 24.6, 20.6. MALDIꢀHRMS Calculated
(M): 554.3369 Observed:(M+1): 555.4175
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3,3'ꢀ((5,5'ꢀ(Diꢀpꢀtolylmethylene)bis(1Hꢀpyrroleꢀ5,2ꢀdiylꢀ2ꢀcarbonyl))bis(azanediyl))
bis(propanꢀ1ꢀaminium) chloride (7). The general procedure was followed using 4 (100 mg) to
give 7 as a white solid in quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 11.0 (2H, s, pyrroleꢀNH) 8.2 (2H, t, J = 5.2 Hz, amideꢀNH)
7.9 (6H, s, NH3) 7.1 (4H, d, J = 8.4 Hz, phenylꢀH) 6.8 (4H, d, J = 8.4 Hz, phenylꢀH) 6.7 (2H, d, J
= 3.6 Hz, pyrroleꢀH) 5.7 (2H, d, J = 3.6 Hz, pyrroleꢀH) 3.2 (4H, q, J = 6.4 Hz, CH2) 2.8 (4H, q, J
= 6.4 Hz, CH2) 2.3 (6H, s, CH3), 1.7 (4H, p, CH2)13C NMR (100 MHz, DMSOꢀd6) δ 160.7,
141.9, 139.0, 135.9, 129.2, 128.2, 126.8, 111.0, 110.0, 55.2, 36.8, 35.5, 27.6, 20.6.MALDIꢀ
HRMSꢀ Calculated (M+) 526.3056 Observed (M+1):527.3686; (M+Na):549.3571.
2,2'ꢀ((5,5'ꢀ(Diꢀpꢀtolylmethylene)bis(1Hꢀpyrroleꢀ5,2ꢀdiylꢀ2ꢀcarbonyl))bis(azanediyl))
bis(ethanꢀ1ꢀaminium) chloride (8). The general procedure was followed using 5 (100 mg) to
yield 8 as a white solid in quantitative yield.
1H NMR (400 MHz) δ (DMSOꢀd6) 11.0 (2H, s, pyrroleꢀNH) 8.3 (2H, t, J = 5.4 Hz, amideꢀNH)
7.9 (6H, br s, NH3) 7.1 (4H, d, J = 8.2 Hz, phenylꢀH) 6.8 (4H, d, J = 8.2 Hz, phenylꢀH) 6.7 (2H,
m pyrroleꢀH) 5.7 (2H, m, pyrroleꢀH) 3.4 (4H, q, J = 6.0 Hz, CH2) 2.9 (4H, q, J = 6.0 Hz CH2) 2.3
(6H, s, CH3)13C NMR (100 MHz, DMSOꢀd6) δ 160.9, 141.7, 139.2, 135.9, 129.1, 128.1, 126.5,
ACS Paragon Plus Environment

Page 29 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
110.94, 110.2, 55.2, 38.8, 36.5, 20.5.MALDIꢀHRMS: Calculated (M+) 498.2743 Observed
(M+1) 499.3296 m/z; (M+Na) 521.3234
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molecular biology
DNA constructs. cDNAs for rat Kv1.1. 1.2, 1.3, 1.4 and 1.6 were kindly provided by Professor
Olaf Pongs (Institute for Neural Signal Transduction, University of Hamburg, Germany).
Concatenation of four α Kv1.1 and/or Kv1.2 subunits as single open reading frames was
accomplished,50 using an interꢀsubunit linker derived from the untranslated regions of the
Xenopus βꢀglobin gene (GenBank® accession number J00978), as reported previously.20, 21
Expression of Kv1 channels. HEKꢀ293 cells (American Type Culture Collection) were
prepared stably expressing Kv1.1, 1.2, 1.3, 1.4, 1.6 or tandemꢀlinked Kv1.1ꢀ and 1.2ꢀ
homotetramers; additionally, other concatenated DNA constructs encoding Kv1.1ꢀ1.2ꢀ1.1ꢀ1.1,
Kv1.1ꢀ1.1ꢀ1.2ꢀ1.2, Kv1.2ꢀ1.2ꢀ1.1ꢀ1.1, Kv1.2ꢀ1.1ꢀ1.2ꢀ1.1 or Kv1.2ꢀ1.2ꢀ1.1ꢀ1.2 were transiently
transfected into HEKꢀ293 cells, using Polyfect reagent (Qiagen, Crawley, UK) or Transꢀ293®
transfection reagent (Mirus Bio LLC, Madison, WI, USA).
Wholeꢀcell voltage clamp recordings and data analysis. This was performed as previously
outlined,19, 21 except where specified. In the conventional patch clamp system [EPC10 amplifier
(HEKA Elektronik, Lambrecht/Pfalz, Germany)], the recording pipette was filled with an
internal solution that contained (in mM): 95 KF, 30 KCl, 1 CaCl2, 1 MgCl2, 11 EGTA, 10
HEPES, 2 K2ATP (pH 7.2 with KOH), and having fireꢀpolished tips of resistance between 2ꢀ5
MΩ. The external (bath) medium contained (in mM) 135 NaCl, 5 KCl, 2 CaCl2, 2 MgCl2 and
5 Hepes, and 10 sucrose (pH 7.4 with NaOH). Wholeꢀcell currents were measured at a holding
ACS Paragon Plus Environment