You are on page 1of 10

Pharmacological Reports

https://doi.org/10.1007/s43440-021-00304-5

ARTICLE

Thallium‑sensitive fluorescent assay reveals loperamide as a new


inhibitor of the potassium channel Kv10.1
Arlet Loza‑Huerta1 · Edgar Milo1 · Arturo Picones1 · Arturo Hernández‑Cruz1,2 · Enoch Luis1,3 

Received: 5 April 2021 / Revised: 21 June 2021 / Accepted: 23 June 2021


© Maj Institute of Pharmacology Polish Academy of Sciences 2021

Abstract
Background  Ion channels have been proposed as therapeutic targets for different types of malignancies. One of the most
studied ion channels in cancer is the voltage-gated potassium channel ether-à-go-go 1 or Kv10.1. Various studies have shown
that Kv10.1 expression induces the proliferation of several cancer cell lines and in vivo tumor models, while blocking or
silencing inhibits proliferation. Kv10.1 is a promising target for drug discovery modulators that could be used in cancer
treatment. This work aimed to screen for new Kv10.1 channel modulators using a thallium influx-based assay.
Methods  Pharmacological effects of small molecules on Kv10.1 channel activity were studied using a thallium-based fluo-
rescent assay and patch-clamp electrophysiological recordings, both performed in HEK293 stably expressing the human
Kv10.1 potassium channel.
Results  In thallium-sensitive fluorescent assays, we found that the small molecules loperamide and amitriptyline exert a
potent inhibition on the activity of the oncogenic potassium channel Kv10.1. These results were confirmed by electrophysi-
ological recordings, which showed that loperamide and amitriptyline decreased the amplitude of Kv10.1 currents in a dose-
dependent manner. Both drugs could be promising tools for further studies.
Conclusions  Thallium-sensitive fluorescent assay represents a reliable methodological tool for the primary screening of
different molecules with potential activity on Kv10.1 channels or other ­K+ channels.

Keywords  Kv10.1 potassium channel · Thallium-sensitive fluorescent assay · Electrophysiology · Oncogenic channel

Abbreviations Introduction
AMI Amitriptyline
Eag1 Human ether-à-go-go potassium channel 1 Potassium ion channels are the most diverse family of ion
HEK-WT HEK293 wild-type cells channels in mammalian cells. These transmembrane proteins
HEK-Kv10.1 HEK293 cells stably expressing the human regulate many fundamental biological processes, including
Kv10.1 potassium channel generating action potentials, heart rate, electrolytic balance,
LP Loperamide and cell division [1–3]. Given the diverse physiological roles
SR SR 33805 oxalate of ­K+ channels, they are associated with pathological con-
ditions, and consequently, are considered ideal targets for
drug discovery.
* Enoch Luis Human ether-à-go-go potassium channels (Eag1 or
enoch@ifc.unam.mx; enokluis@gmail.com Kv10.1, encoded by the KCNH1 gene) belong to voltage-
1 gated potassium channels. Topologically, each subunit
Laboratorio Nacional de Canalopatías, Instituto de Fisiología
Celular, Universidad Nacional Autónoma de México, contains six transmembrane segments (S1 to S6), with
Circuito Exterior s/n, C.U. 04510 Mexico City, Mexico the N- and C-terminal located in the intracellular side.
2
Departamento de Neurociencia Cognitiva, Instituto de The S1–S4 segments constitute the voltage-sensor domain.
Fisiología Celular, Universidad Nacional Autónoma de S5–S6 form the selective potassium pore, and the assem-
México, Circuito Exterior s/n, C.U. 04510 Mexico City, bly of four identical subunits forms a functional Kv10.1
Mexico channel [4–6]. Opening of the channel produces a K ­ +
3
Cátedras CONACYT ‑ Instituto de Fisiología Celular, efflux that hyperpolarizes the resting membrane poten-
Universidad Nacional Autónoma de México, Circuito tial of cells. In healthy individuals, Kv10.1 expression is
Exterior s/n, C.U. 04510 Mexico City, Mexico

13
Vol.:(0123456789)
A. Loza‑Huerta et al.

mainly restricted to some central brain neurons [7–9], with Thallium‑sensitive fluorescent assay
practically no expression in other regions of the human
body. However, an abnormal overexpression of Kv10.1 Flat-bottom 96-well plates (781611, BrandTech) were
channels has been observed in approximately 70% of solid plated with HEK-WT or HEK-Kv10.1 at a density of
tumors [10–12], making them a promising target for drug 30,000 cells/well in 100 µl of supplemented DMEM and
discovery modulators that would be used in cancer treat- incubated for 18–24 h at 37 °C in a 5% C ­ O2 incubator.
ment. The molecular mechanisms involved in this atypical The thallium-sensitive fluorescent assay was performed
expression in cancer cells are not completely clear. A vast using the FLIPR Potassium assay explorer kit (R8222,
number of publications have associated Kv10.1 overex- Molecular Devices) following the manufacturer’s proto-
pression with the malignant properties of many cancer cols. Briefly, supplemented DMEM on the 96-well plates
cells. In 1999, Luis Pardo showed for the first time that was removed and replaced with 50 µl of DMEM and an
Kv10.1 channels could transform the typical phenotype of equal volume of loading buffer, and then incubated at
mammalian cells into a malignant cancer-like phenotype, 37 °C for 60 min; loading buffer contained: 1X Hank’s
which is characterized by a high rate of proliferation and balanced salt solution (14065-056, Gibco; pH 7.4), 20 mM
metabolism. Furthermore, the transplant of cells express- HEPES, and the component A (R7578) and C (R7579) of
ing Kv10.1 channels into immunodeficient mice induced the FLIPR Potassium explorer kit. Next, the molecules
the development of solid tumors [13]. Pharmacological were pre-incubated (at a final concentration of 50 µM)
inhibition of Kv10.1 channels using small molecules or with the cells in 96-well plates for 15 min at 37 °C in a
antibodies has shown antitumor efficacy [14–20]. How- 5% ­CO2 incubator. After pre-incubation, 96-well plates
ever, only non-specific inhibitors of Kv10.1 have been were transferred to the microplate reader FlexStation3
described. Therefore, research on the Kv10.1 has focused (Molecular Devices) controlled with the SoftMax Pro 7
on searching for new and selective modulators. software. Thallium-sensitive dye was excited at 485 nm,
Cell-based fluorescent assays have frequently been and the emitted fluorescence was recorded at 515 nm. Data
extensively used to search and develop new modulators were acquired at 3-s intervals for 201 s. The experimental
of ion channels [21]. For example, to study the functional protocol consisted of 21 s of basal fluorescence record-
activity of ­K+-selective ion channels, most of the fluores- ings, after which 50 µl of the stimulus solution (2 mM
cence-based assays use (1) the permeability of thallium ­Tl2SO4/30 mM ­K2SO4) was added, and the fluorescence
­(Tl+) through open ­K+ channels, and (2) a thallium-sen- signal was recorded for another 180 s. FlexStation3 reads
sitive fluorescent dye [22–24]. Here, we use the ­FLIPR® the first column of the plate before moving to the next one;
Potassium Assay (Molecular Devices) to identify new every column has a positive control well (without pre-
Kv10.1 channel modulators that have not been previously incubated molecules) used to follow the time-dependent
described. We found that loperamide and amitriptyline variation among columns, and a negative control well
(drugs with known action on calcium channels) exert a (stimulated with 2  mM ­T l 2SO 4/0  mM ­K 2SO 4) used to
potent inhibitory effect on Kv10.1 activity. correct (subtract) and determine maximal fluorescence
signal of each well (Supplemental Fig.  1A–B). Each
well’s amplitude was normalized to the positive control
well (without pre-incubated molecules) of each column.
Material and methods The assay quality was estimated by Z′ factor, and only
plates with a Z′ of > 0.4 were deemed acceptable (data not
Culture of cells shown) [25].

HEK293 wild-type cells (HEK-WT) (CRL-1573, ATCC)


were cultured in Dulbecco’s Modified Eagle Medium Electrophysiological experiments
(DMEM) (12800–017, Gibco) containing 10% (v/v) heat-
inactivated fetal bovine serum (26140087, Gibco) and 1% For patch-clamp recordings, cells were plated in 3.8 c­ m2
Penicillin/Streptomycin (15140122, Gibco). HEK293 cells TC-Treated Multiple Well Plates (CLS3513, Sigma) at
stably expressing the human Kv10.1 potassium channel 200,000 cells/well. After 18–48 h, cells were trypsinized
(HEK-Kv10.1) (generously provided by Dr. Walter Stüh- and re-plated on 18 mm circular glass coverslips and kept
mer) were cultured in DMEM containing 10% fetal bovine in the incubator for 1–3 h. Coverslips were transferred to
serum, 1% Penicillin/Streptomycin, and as selection anti- a recording chamber (RC-26G, Warner Instruments, Ham-
biotic Zeocin (30 µg/ml) (R25001, Invitrogen). All cells den, CT, USA) and whole-cell voltage-clamp recordings
were cultured at 37 °C in a 5% ­CO2 incubator. were performed at room temperature and under continuous

13
Thallium‑sensitive fluorescent assay reveals loperamide as a new inhibitor of the potassium…

perfusion with a standard bath solution at a flow rate of statistical significance was assessed by one-way ANOVA
2.4 ml/min. The standard bath solution contained (in mM): followed by Games–Howell post hoc test. Differences were
137 NaCl, 5.4 KCl, 2 ­CaCl2, 2 TEA-Cl, 10 HEPES, and 10 considered significant when p < 0.05. In thallium assay plots
d -glucose (300 mOsm/kg, pH 7.4 adjusted with NaOH). (Fig. 1D), data points for DMSO, astemizole, loperamide,
Intracellular patch-pipettes (2–3 MΩ) filling solution con- amitriptyline, and SR 33805 are displayed as mean ± SEM
tained (in mM): 140 KCl, 1 M ­ gCl2, 10 EGTA, 10 HEPES of four independent experiments performed in duplicate;
(300 mOsm/kg, pH 7.2 adjusted with KOH). Ionic cur- the rest of the data points represent the mean ± SEM of two
rents were recorded with a setup composed of Axopatch independent experiments performed in duplicate. In elec-
200B/Digidata 1550/pCamp10 (amplifier/analog–digital trophysiology, data points are displayed as mean ± SEM of
converter/software, all from Molecular Devices), filtered n > 3 independent experiments. All data were analyzed using
at 5 kHz, and digitally sampled at 10 kHz. 60% of the OriginPro9 (OriginLab, USA), IBM SPSP Statistics version
series resistance was electronically compensated. Small 24 and GraphPad Prism 9 (GraphPad Software, USA).
molecules were delivered using a perfusion system VC-6
coupled to a fast-step solution charger (SF-77B, Warner
Instruments, Hamden, CT, USA). Results
Drugs tested on Kv10.1 currents were evaluated from a
holding voltage (Vh) of −70 mV, and currents were gener- Primary screening using a ­Tl+ influx assay
ated by 250 ms duration voltage steps from −70 to +50 mV
applied every 5 s. I–V curves were constructed using a proto- To set up a functional Kv10.1 thallium-sensitive fluorescent
col of 250 ms voltage steps from −100 to  +50 mV in 10 mV assay, HEK-WT and HEK-Kv10.1 were plated in 96-well
increments and a Vh of −70 mV; the amplitude of Kv10.1 plates, loaded with the thallium-sensitive dye, and assayed
currents was measured at the end of voltage steps (aver- for activity with a high thallium/potassium stimulus solu-
age of the last 5 ms) and plotted versus voltage. Concentra- tion (2 mM ­Tl2SO4/30 mM ­K2SO4); this 2 mM ­Tl+ and
tion–effect curves were fitted with the Hill equation using 30 mM ­K+ stimulus was selected from the titration curves
the Levenberg–Marquardt method implemented in Origin (Supplemental Fig. 1C). As shown in Fig. 1A, B, HEK-WT
2016 software (OriginLab Corporation): cells did not respond to the T­ l+/K+ stimulus; this was taken
as proof of the absence or minimal expression of voltage-
cn
Inhibition = Bmax , (1) activated potassium channels. On the other hand, HEK-
IC50n + cn
Kv10.1 cells produced robust responses when the T ­ l+/K+
where Bmax is the maximum block, IC50 is the concentra- stimulus was applied.
tion of half-maximal inhibition, c is the concentration of the Next, we did a primary screening with 21 compounds
molecules, and n is the Hill coefficient. on HEK-Kv10.1 cells to establish the fluorescent assay’s
To determine the voltage dependence of Kv10.1 inactiva- sensitivity (Table 1); these experiments represented two
tion, a double-pulse protocol was used. Beginning from a independent experiments performed in duplicate (n = 2).
holding voltage of −70 mV, conditioning pre-pulses (Vpre) of Then, we obtained the median fluorescent amplitude in the
1.5 s in duration to voltages ranging from −130 to +50 mV presence of each compound and determined that those com-
(in 20 mV increments) were applied every 10 s. After each pounds that showed a decrease in the thallium-fluorescent
prepulse, a 500 ms test pulse was applied to +30 mV. signal greater than 30% (concerning the control) would be
re-evaluated by increasing their repetitions (n). Most of the
Compounds tested compounds (50 µM) did not change the amplitude
of the responses. Only astemizole, loperamide (LP), ami-
All the small molecules (Ek-103) and maurotoxine (STM- triptyline (AMI), and SR 33805 oxalate (SR) decreased the
340) were purchased from Alomone Labs. Only astemizole median fluorescent amplitude (81%, 57%, 65%, and 39%,
(A2861) and DMSO (D2650) were purchased from Sigma- respectively) (Table 1). We also tested maurotoxin (100 nM),
Aldrich. Most stock solutions were made in 100% DMSO a toxin generally used in K­ + channels assays [26], but it did
or water at a concentration of 100 mM. Maurotoxin stock not affect Kv10.1 activity.
solution was made at 5 µM in water. All molecules were As we mentioned above, based on our first screening, we
diluted in a standard bath solution. increased the repetitions for astemizole, LP, AMI, SR and
DMSO (n = 4; four independent experiments performed
Data analysis in duplicate). A statistical analysis using one-way ANOVA
showed astemizole (10  µM), a well-described blocker of
When two groups were compared, statistical significance Kv10.1 channels [27], produce a statistically significant 82%
was assessed by Student’s t test. For multiple comparisons, reduction of the fluorescence intensity (n = 4) (F5,41 = 50.13;

13
A. Loza‑Huerta et al.

Fig. 1  Thallium-sensitive fluorescent assay for Kv10.1 channels. AMI (red). D The summary bar graph of the fluorescence responses
A Thallium-fluorescent recordings in HEK-WT cells (red line) and in HEK-Kv10.1 cells in the presence of different molecules (astemi-
HEK-Kv10.1 cells (black line); ­Tl+ inflow through Kv10.1 channels zole was used at 10  µM, DMSO at 0.1%, Maurotoxin at 100  nM,
was induced with a stimulus solution (2 mM ­Tl2SO4/30 mM ­K2SO4, and the rest of the molecules were tested at a final concentration of
arrow) and measured with a fluorescence FLIPR potassium assay. B 50 µM). Data are normalized to the positive control well. In A and C,
The summary bar graph of the responses in HEK-WT cells and HEK- the solid traces and faded colors show the mean ± SEM, respectively.
Kv10.1 cells. C Thallium-fluorescent responses in HEK-Kv10.1 with Analysis of variance (ANOVA) using the Games–Howell post hoc
control positive response (black), DMSO (orange), LP (blue), and test: **p ≤ 0.01, ***p ≤ 0.001

p < 0.0001, Games–Howell p < 0.001) (Fig.  1D). These The pharmacological effects of the molecules were evalu-
results confirm the functional expression of Kv10.1 chan- ated on Kv10.1-evoked currents by repeated voltages
nels. We evaluated the effect of DMSO (vehicle; n = 4) at pulses from −70 to +50 mV. First, we tested maurotoxin
0.1% on the fluorescence responses, which did not produce (100 nM) on Kv10.1 currents, and similar to the fluores-
any detectable change of the high-potassium-induced thallium cent results, this toxin did not modify the amplitude of the
signal (Fig. 1C, D) (F5,41 = 50.13; p < 0.0001, Games–How- Kv10.1 currents (data not shown). SR has been described
ell p = 0.9). We observed that LP, AMI, and SR significantly as an l -type ­C a 2+ channel antagonist [28]. Various SR
decreased (53%, 55%, and 33%, respectively) (F5,41 = 50.13; concentrations were tested (0.1–300 µM). As shown in
p < 0.0001, Games–Howell p < 0.001 for LP and RS, and Fig. 2A, B, 100 µM SR produced a modest (38.1%; n = 8,
p = 0.002 for AMI) the amplitude of the fluorescence signal p = 0.006; paired t test) and reversible inhibition on Kv10.1
mediated by Kv10.1 channels (Fig. 1C, D; Table 2). currents (data not shown). To quantify SR’s potency as a
Kv10.1 inhibitor, we constructed dose-inhibition curves
SR 33805 oxalate produces a modest inhibition (Fig. 2C) of the Kv10.1 currents evoked at +50 mV. Hill
of Kv10.1 currents fits were performed to these data yielding an IC50 of
130.8 ± 25.3 µM and a Hill coefficient of 1.5 ± 0.1.
To validate the thallium assay’s results, whole-cell patch-
clamp experiments were performed in HEK-Kv10.1 cells.

13
Thallium‑sensitive fluorescent assay reveals loperamide as a new inhibitor of the potassium…

Table 1  Primary screening Compound Cat # Concentration n Median (range)


(µM)

DMSO A2861, Sigma-Aldrich 0.1% 2 0.97 (0.76–1.18)


Astemizole D2650, Sigma-Aldrich 10 2 0.19 (0.15–0.27)
Loperamide L-100, alomone labs 50 2 0.37 (0.28–0.65)
Amitriptyline A-155, alomone labs 50 2 0.34 (0.17–0.53)
Flunarizine F-110, alomone labs 50 2 1.0 (0.88–1.24)
Verapamil V-100, alomone labs 50 2 0.94 (0.87–0.99)
Nifedipine N-120, alomone labs 50 2 0.86 (0.78–0.94)
Lomerizine L-125, alomone labs 50 2 0.85 (0.75–0.97)
Diltiazem D-135, alomone labs 50 2 0.95 (0.88–0.99)
Felodipine F-105, alomone labs 50 2 0.97 (0.92–1.02)
Flavoxate F-140, alomone labs 50 2 0.91 (0.75–1.02)
L-651,582 L-110, alomone labs 50 2 0.94 (0.72–1.21)
Lercanidipine L-115, alomone labs 50 2 1.0 (0.86–1.19)
Nicardipine N-125, alomone labs 50 2 1.1 (0.99–1.27)
Nimodipine N-150, alomone labs 50 2 1.1 (1.05–1.32)
Nisoldipine N-160, alomone labs 50 2 1.0 (0.96–1.10)
Nitrendipine N-155, alomone labs 50 2 1.1 (0.76–1.36)
SR 33805 oxalate S-105, alomone labs 50 2 0.63 (0.53–0.67)
Maurotoxin STM-340, alomone labs 0.1 2 0.98 (0.95–1)
Amlodipine A-110, alomone labs 50 2 1.2 (0.64–1.36)
Isradipine I-100, alomone labs 50 2 0.86 (0.77–1.11)

Normalized fluorescence peak responses in the presence of different tested compounds


All median were normalized relative to the positive control well (molecules-free)

Table 2  Normalized fluorescence peak responses in the presence of to +50  mV (every 5  s) and were measured at the end
different tested compounds of each voltage step and plotted versus time (Fig. 3A).
Compound Concentration n Normalized fluores- 100  µM LP produced almost complete inhibition of
(µM) cence (mean ± SEM) Kv10.1 currents (91.3 ± 1.8%; n = 6, p < 0.0001; paired
t test), this effect was reversible upon washing out the
DMSO 0.1% 4 0.96 ± 0.04
drug (Fig. 3A–C). Figure 4A shows the dose-dependent
Astemizole 10 4 0.18 ± 0.03***
inhibitory effect of LP on Kv10.1 currents. These data
Loperamide 50 4 0.47 ± 0.05***
were fitted with a Hill equation, which yielded an IC50
Amitriptyline 50 4 0.45 ± 0.08**
of 12.8 ± 0.4 µM and a Hill coefficient of 1.2 ± 0.3. The
SR 33805 50 4 0.67 ± 0.03***
electrophysiological results with LP were consistent with
All values were normalized relative to the positive control well. One- the fluorescence thallium assays.
way ANOVA using the Games–Howell post hoc test: **p ≤ 0.01, AMI is a tricyclic antidepressant used to treat differ-
***p ≤ 0.001
ent medical conditions. As well as LP, AMI is a non-spe-
cific modulator of ion channels, including C­ a2+ and N
­ a+
Loperamide and amitriptyline produced a potent voltage-gated channels [33, 34]. As shown in Fig. 3D–F,
inhibitory effect on Kv10.1 currents 100 µM AMI produced a strong inhibitory effect on Kv10.1
currents (92.7 ± 1.1%; n = 10, p < 0.0001; paired t-test);
LP is a non-specific blocker of various ion channels and this effect was reversible upon washing (Fig. 3D–F). Fig-
receptors, including voltage-gated ­Na+, ­Ca2+, and NMDA ure 4B shows the dose-dependent inhibitory effect of AMI
receptors [29–31]. Moreover, LP is a potent agonist of on Kv10.1 currents. These data were fitted with a Hill
µ-opiate receptors (found in the gastrointestinal tract), equation, which yielded an IC50 of 15.2 ± 0.2 µM and
and it is used to treat different diarrheic syndromes [32]. a Hill coefficient of 2.2 ± 0.5. The electrophysiological
Various concentrations of LP (1–300 µM) were tested. results with AMI were also consistent with the fluores-
Kv10.1 currents were evoked by voltage steps from −70 cence thallium assay.

13
A. Loza‑Huerta et al.

Fig. 2  SR 33805 produces a partial inhibition of Kv10.1-evoked cur- on Kv10.1 currents (n = 8). C Dose-inhibition curves of SR 33805 on
rents. A Experimental recordings of Kv10.1 currents (at +50 mV) in Kv10.1 evoked currents (n ≥ 4 for each point). The solid lines repre-
control and in the presence of SR 33805 (100 µM). Dashed lines rep- sent the fit to the Hill equation. **p < 0.01, paired t test
resent 0 pA current. B Summary histogram of the effect 100 µM SR

Fig. 3  Loperamide and amitriptyline produce strong inhibition (100 µM) and AMI (100 µM), respectively. Dashed lines represent 0
of Kv10.1 currents. A, D The temporal course of the effect of LP pA current. C, F Current–voltage relationship of Kv10.1-evoked cur-
(100  µM) and AMI (100  µM) on Kv10.1-evoked currents measured rents and the effect of 100 µM LP (n = 6) and 100 µM AMI (n = 10),
at +50  mV, respectively. B, E Experimental recordings of Kv10.1- respectively. The chemical structure of loperamide and amitriptyline
evoked currents (at +50  mV) in control and in the presence of LP was inserted in Figs. A and D, respectively

13
Thallium‑sensitive fluorescent assay reveals loperamide as a new inhibitor of the potassium…

Fig. 4  Dose-inhibition curves of loperamide and amitriptyline. A, B Dose-inhibition curves for LP (n ≥ 4 for each point) and for AMI (n ≥ 3 for
each point) on Kv10.1-evoked currents. The solid lines represent the fit to the Hill equation

Loperamide accelerates voltage‑dependent Discussion


inactivation of Kv10.1 channels
Patch-clamp electrophysiology is well established as the
Human Kv10.1 channels display a very slow voltage- gold-standard method for ion channel studies. However,
dependent inactivation. However, this biophysical prop- patch clamp could be less effective for many debatable rea-
erty could be pharmacologically accelerated [35]. In our sons for the rapid identification or assessment of active and
initial electrophysiological recordings, we noticed that LP selective compounds on ion channels of therapeutic interest.
accelerates the voltage-dependent inactivation of Kv10.1. In recent years, there has been significant progress in devel-
Therefore, we decided to examine in greater detail the oping new high-throughput technologies and non-invasive
effect of LP and AMI on voltage-dependent inactivation. assays to study ion channels. Indeed, cell-based fluorescent
In the control conditions (Fig. 5A, D), the peak current assays have boosted ion channel drug discovery [21, 22, 24,
(I2) measured during the test pulse decreased slightly with 36].
the most depolarized Vpre (+30 and +50 mV) (Fig. 5C; In this work, we used a thallium-based fluorescent assay
data was normalized I2/I2 max). However, after 100 µM to identify new Kv10.1 channel modulators. We found three
LP was applied, we observed an acceleration in the volt- molecules with variable inhibitory effects on the Kv10.1
age-dependent inactivation of Kv10.1 currents (Fig. 5B). activity. The reliability of the method was proved by: (1)
In the presence of LP, inactivation of Kv10.1 currents no changes in fluorescence observed in HEK-WT cells, (2)
was observed starting from a Vpre of −50  mV (n = 5; the pharmacological inhibition of the Kv10.1 channel using
p < 0.0042; paired t test; Fig.  5C). On the other hand, the well-known blocker astemizole [27], which significantly
AMI (100 µM) did not significantly affect the voltage- decreased the amplitude of the responses, and (3) patch-
dependent inactivation of Kv10.1 currents (Fig. 5D–E). clamp experiments that confirmed the inhibitory effect of
In the presence of AMI, inactivation of Kv10.1 currents SR 33805, LP and AMI on Kv10.1 activity.
was statistically significant with a Vpre of + 30 mV (n = 6; Since its discovery as an oncogenic channel, the Kv10.1
p = 0.0104; paired t test; Fig. 5F). Astemizole (10 µM), a has been proposed as a therapeutic target for drug discov-
proposed pore-blocker of the Kv10.1, produces a similar ery with possible cancer treatment use. Kv10.1 is an ion
effect to LP (Supplemental Fig. 2), which indicates a simi- channel overexpressed in tumor cells but with low expres-
lar mechanism of action. sion in healthy tissue, therefore, the side effects of blocking

13
A. Loza‑Huerta et al.

Fig. 5  Effect of loperamide and amitriptyline on the voltage-depend- of the Vpre, however, in the presence of LP (n = 5), the inactivation
ent inactivation of Kv10.1 currents. A Kv10.1 currents recorded process of the channels was accelerated. D Kv10.1 currents recorded
in control solution using a double-pulse protocol shown below. B in control condition. E Kv10.1 currents recorded in the presence of
Kv10.1 currents recorded in the presence of 100 µM LP. C For inac- AMI (100  µM). F inactivated curves in the control condition (open
tivated curves, maximal Kv10.1 current (I2) measured at + 30  mV, diamond) and in the presence of 100 µM AMI (filled diamond; n = 6).
normalized (I2/I2max) and plotted as a function of Vpre, in the con- Dashed lines represent 0 pA current. *, ** and ***p < 0.05, < 0.01
trol condition (open circles) and in the presence of 100 µM LP (filled and < 0.001 paired t test, respectively
circles). In control, Kv10.1 currents decreased slightly as a function

Kv10.1 could be negligible. In mammals, Kv10.1 expression voltage-gated ion channels, including the cardiac K ­ + chan-
is restricted to the brain. Various studies show that the chan- nel, hERG (Kv11.1). Although none of the molecules tested
nel is found in the olfactory bulb, cerebral cortex, striatum, here are specific for Kv10.1 channels, the potent and revers-
hippocampus, hypothalamus, cerebellum, thalamus, and ible inhibitory effects of LP on Kv10.1 make it a promising
brainstem [7–9, 37], but its physiological role is not clear. tool for further biophysical and molecular studies and for a
In neurons, voltage-gated potassium channels play crucial better understanding of the functional role of Kv10.1 chan-
roles in determining the resting membrane potential and the nels on cancer biology.
shape of action potentials. However, the biophysical acti- In the electrophysiological experiments, tested molecules
vation properties of Kv10.1 channels indicate that they do were ranked by their IC50, with the values in the follow-
not participate in setting the resting membrane potential or ing potency sequence: loperamide > amitriptyline > SR
the shape of single-action potentials [9, 37]. Nevertheless, 33805 oxalate, with the first two compounds exhibiting
in parallel fiber synapses in the cerebellar cortex, Kv10.1 an IC50 below 20 µM. Comparison with previous reports
modulates action potential width during high-frequency revealed that LP can inhibit Nav1.7 channels, heterologously
activity [9]. Notably, Kv10.1 knockout (KO) mice do not expressed in HEK cells [30]. Wu et al. also reported that LP
show apparent cognitive or behavioral differences from their exhibits more potent inhibition on endogenous Nav1.8 cur-
littermates control during development and adulthood [37]. rents expressed in DRG neurons. Nevertheless, this effect
The results suggest that the pharmacological inhibition of is partially mediated by the activation of µ-opioid receptors
Kv10.1 aimed at treating cancer is feasible. This approach in [30]. In our experiments, the inhibitory effect of LP was
conjunction with currently available treatments (e.g., chemo- strictly on Kv10.1 channels since endogenous opioid recep-
therapy) is possible and could improve cancer therapy. tors are not expressed in HEK cells [38, 39]. On the other
To date, there are no small molecules with enough speci- hand, AMI is a structurally related molecule to imipramine
ficity on Kv10.1 activity, with most of them inhibiting other (another non-specific Kv10.1 blocker) and, therefore, it is

13
Thallium‑sensitive fluorescent assay reveals loperamide as a new inhibitor of the potassium…

not surprising that AMI also showed a considerable effect Author contributions  AL–H, EM and EL carried out the experiments;
on the Kv10.1 activity [27]. EL and AH-C designed, analyzed, and interpreted the results; EL wrote
the manuscript. All the authors read and approved the final manuscript.
Human and mouse Kv10.1 channels are characterized by
generating almost non-inactivating macroscopic currents [4, Funding  This work was supported by Grants: 314839, 21887 and
5], but this biophysical property could be pharmacologi- 315803 from Consejo Nacional de Ciencia y Tecnología (CONA-
cally accelerated [27, 35, 40]. In our electrophysiological CYT) and PAPIIT-UNAM IG200119 to AH-C, and SEP-CONACYT
recordings, LP accelerates voltage-dependent inactivation CB2017-2018-A1-S-13646 to E.L. We thank Fis. Cesar Oliver Lara
Figueroa and Dr. Luisa Duran Pasten for technical assistance. The Eng-
of Kv10.1 channels, and this effect was not observed in the lish revision by Dr. Bristol Denlinger is also acknowledged.
presence of AMI. Such differences suggest that both com-
pounds are allosteric inhibitors of the Kv10.1 channel. The Declarations 
electrophysiological behavior in the presence of LP was sim-
ilar to previous reports with astemizole [27], a well-known Conflict of interest  All authors have no conflict of interest.
open-channel blocker of Kv10.1, suggesting LP could be an
open-pore inhibitor. Nevertheless, further biophysical and
pharmacological research (e.g., pharmacological competi-
tion assays or mutagenesis experiments) are needed to test References
this hypothesis. We do not rule out a possible effect of LP on
the voltage-sensor domain as some other molecules acting 1. Alexander SPH, Striessnig J, Kelly E, Marrion NV, Peters JA,
Faccenda E, et al. THE CONCISE GUIDE TO PHARMACOL-
on ­K+ channels have shown [40, 41].
OGY 2017/18: voltage-gated ion channels. Br J Pharmacol.
The potassium channels’ opening and closing regulate 2017;1(174):S160–94.
resting membrane potential in excitable cells, modifying 2. Ashcroft FM. From molecule to malady. Nature. 2006;
the action potential frequency and duration [42]. While in 440:440-47.
3. Yu FH, Catterall WA. The VGL-chanome: a protein superfam-
non-excitable cells, they could participate in other processes,
ily specialized for electrical signaling and ionic homeostasis. Sci
including the cell cycle and proliferation [43, 44]. In can- STKE. 2004;2004:1–17.
cer cells, Kv10.1 channel overexpression is relevant for cell 4. Whicher JR, MacKinnon R. Structure of the voltage-gated ­K+
proliferation [13, 45]. In contrast, their pharmacological channel Eag1 reveals an alternative voltage sensing mechanism.
Science (80-). 2016;353(6300):664–9.
blockage can induce anti-proliferative effects, suggesting
5. Wang X, Chen Y, Zhang Y, Guo S, Mo L, An H, et al. Eag1
a molecular link between Kv10.1 expression and the cell voltage-dependent potassium channels: structure, electrophysi-
cycle mechanisms. The possible connection could be related ological characteristics, and function in cancer. J Membr Biol.
to changes in the voltage membrane potential induced by 2017;250:123−132.
6. Occhiodoro T, Bernheim L, Liu JH, Bijlenga P, Sinnreich M,
the aberrant expression of ion channels in cancer cells that
Bader CR, et al. Cloning of a human ether-a-go-go potassium
impact the nanoscale organization of phospholipids impli- channel expressed in myoblasts at the onset of fusion. FEBS Lett.
cated in mitogen signaling [46]. Thus, small molecules used 1998;434:177-82.
as pharmacological modulators of Kv10.1 channels repre- 7. Bauer CK, Schwarz JR. Ether-à-go-go K + channels: effective
modulators of neuronal excitability. J Physiol. 2018;596:769–83.
sent a reasonable way to induce changes in the voltage mem-
8. Martin S, Lino de Oliveira C, Mello de Queiroz F, Pardo LA,
brane potential that could reduce many hallmarks of cancer. Stühmer W, Del Bel E. Eag1 potassium channel immunohisto-
The idea of a relationship between membrane potential and chemistry in the CNS of adult rat and selected regions of human
cancer has been described thoroughly in recent reviews [47, brain. Neuroscience. 2008;155:833–44.
9. Mortensen LS, Schmidt H, Farsi Z, Barrantes-Freer A, Rubio ME,
48].
Ufartes R et al. Kv10.1 opposes activity-dependent increase in
In summary, given the close correlation with the results ­Ca2+ influx into the presynaptic terminal of the parallel fibre-
obtained by patch-clamp electrophysiology, the thallium- Purkinje cell synapse. J Physiol. 2015;593:181–96.
sensitive fluorescent assay represents an effective and reli- 10. Hemmerlein B, Weseloh RM, de Queiroz FM, Knötgen H,
Sánchez A, Rubio ME, et al. Overexpression of Eag1 potassium
able method for the primary screening of different molecules
channels in clinical tumours. Mol Cancer. 2006;5(41):1–13.
with potential activity on Kv10.1 channels. The thallium- 11. Rodríguez-Rasgado JA, Acuña-Macías I, Camacho J. Eag1
sensitive fluorescent assay opens the possibility to explore channels as potential cancer biomarkers. Sensors (Switzerland).
the effect of peptide toxins isolated from animal venoms or 2012;12(5):5986–95.
12. Ferreira NR, Mitkovski M, Stühmer W, Pardo LA, Del Bel EA.
on Kv10.1 channels. Although peptide toxins exhibit high
Ether-à-go-go 1 (Eag1) potassium channel expression in dopa-
potency and selectivity on ion channels [49], just a few of minergic neurons of basal ganglia is modulated by 6-hydroxydo-
them have been described as showing an inhibitory effect on pamine lesion. Neurotox Res. 2012;21(3):317–33.
Kv10.1 channels [50–52]. 13. Pardo LA, Del Camino D, Sánchez A, Alves F, Brüggemann A,
Beckh S, et al. Oncogenic potential of EAG ­K+ channels. EMBO
J. 1999;18:5540–47.
Supplementary Information  The online version contains supplemen-
14. Gómez-Varela D, Zwick-Wallasch E, Knötgen H, Sánchez A, Hett-
tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 43440-0​ 21-0​ 0304-5.
mann T, Ossipov D, et al. Monoclonal antibody blockade of the

13
A. Loza‑Huerta et al.

human Eag1 potassium channel function exerts antitumor activity. 34. Song JH, Ham SS, Shin YK, Lee CS. Amitriptyline modulation
Cancer Res. 2007;67:7343–49. of ­Na+ channels in rat dorsal root ganglion neurons. Eur J Phar-
15. García-Becerra R, Díaz L, Camacho J, Barrera D, Ordaz-Rosado macol. 2000;401:297-305.
D, Morales A, et al. Calcitriol inhibits Ether-à go-go potassium 35. Garg V, Sachse FB, Sanguinetti MC. Tuning of EAG ­K+ channel
channel expression and cell proliferation in human breast cancer inactivation: Molecular determinants of amplification by muta-
cells. Exp Cell Res. 2010;316:433–42. tions and a small molecule. J Gen Physiol. 2012;140:307–24.
16. Bernal-Ramos G, Hernández-Gallegos E, Vera E, Chávez-López M 36. Vetter I, Carter D, Bassett J, Deuis JR, Tay B, Jami S, et al. High-
de G, Zúñiga-García V, Sánchez-Pérez Y, et al. Astemizole inhibits throughput fluorescence assays for ion channels and GPCRs. In:
cell proliferation in human prostate tumorigenic cells expressing Advances in experimental medicine and biology. 2020;27–72.
ether à-go-go-1 potassium channels. Cell Mol Biol. 2017;63:11–3. 37. Ufartes R, Schneider T, Mortensen LS, Romero C de J, Hen-
17. Wang X, Chen Y, Li J, Guo S, Lin X, Zhang H, et al. Tetrandrine, trich K, Knoetgen H, et al. Behavioural and functional charac-
a novel inhibitor of ether-à-go-go-1 (Eag1), targeted to cervical terization of Kν10.1 (Eag1) knockout mice. Hum Mol Genet.
cancer development. J Cell Physiol. 2019;234:7161–73. 2013;22(11):2247–62.
18. Gavrilova-Ruch O, Schönherr K, Gessner G, Schönherr R, Klap- 38. Gharagozlou P, Demirci H, Clark JD, Lameh J. Activation profiles
perstück T, Wohlrab W, et al. Effects of imipramine on ion chan- of opioid ligands in HEK cells expressing δ opioid receprtors.
nels and proliferation of IGR1 melanoma cells. J Membr Biol. BMC Neurosci. 2002;3:1-8.
2002;188:137–49. 39. Gharagozlou P, Demirci H, Clark JD, Lameh J. Activity of opioid
19. García-Quiroz J, Camacho J. Astemizole: an old anti-histamine as ligands in cells expressing cloned μ opioid receptors. BMC Phar-
a new promising anti-cancer drug. Anticancer Agents Med Chem. macol. 2003;3:1-8.
2012;11(3):307-14. 40. Gómez-Lagunas F, Carrillo E, Pardo LA, Stühmer W. Gating
20. Wu J, Zhong D, Fu X, Liu Q, Kang L, Ding Z. Silencing of Ether modulation of the tumor-related Kv10.1 channel by Mibefradil. J
à go-go 1 by shRNA inhibits osteosarcoma growth and cell cycle Cell Physiol. 2017;232(8):2019–32.
progression. Int J Mol Sci. 2014;15(4) 5570–81. 41. Moreels L, Bhat C, Voráčová M, Peigneur S, Goovaerts H, Mäki-
21. Yu HB, Li M, Wang WP, Wang XL. High throughput screen- Lohiluoma E, et al. Synthesis of novel purpurealidin analogs and
ing technologies for ion channels. Acta Pharmacologica Sinica. evaluation of their effect on the cancer-relevant potassium channel
2016;37:34-43. KV10.1. PLoS One. 2017;12(12):1-18.
22. Philippaert K, Kerselaers S, Voets T, Vennekens R. A thallium- 42. Humphries ESA, Dart C. Neuronal and cardiovascular potassium
based screening procedure to identify molecules that modulate the channels as therapeutic drug targets: promise and pitfalls. J Bio-
activity of C­ a2+-activated monovalent cation-selective channels. mol Screen. 2015;20:1055–73.
SLAS Discov. 2018;23:341–52. 43. Urrego D, Tomczak AP, Zahed F, Stühmer W, Pardo LA. Potas-
23. Huang XP, Mangano T, Hufeisen S, Setola V, Roth BL. Identifi- sium channels in cell cycle and cell proliferation. Philos Trans R
cation of human Ether-à-go-go related gene modulators by three Soc B Biol Sci. 2014;369(1638):1–9.
screening platforms in an academic drug-discovery setting. Assay 44. Rao VR, Perez-Neut M, Kaja S, Gentile S. Voltage-gated ion
Drug Dev Technol. 2010;8:727–42. channels in cancer cell proliferation. Cancers. 2015;7:849–75.
24. Weaver CD, Harden D, Dworetzky SI, Robertson B, Knox RJ. 45. Urrego D, Movsisyan N, Ufartes R, Pardo LA. Periodic expres-
A thallium-sensitive, fluorescence-based assay for detecting and sion of Kv10.1 driven by pRb/E2F1 contributes to G2/M pro-
characterizing potassium channel modulators in mammalian cells. gression of cancer and non-transformed cells. Cell Cycle.
J Biomol Screen. 2004;9:671–77. 2016;15:799-811.
25. Zhang J-H, Chung TDY, Oldenburg KR. A simple statistical 46. Zhou Y, Wong CO, Cho KJ, Van Der Hoeven D, Liang H,
parameter for use in evaluation and validation of high throughput Thakur DP, et al. Membrane potential modulates plasma mem-
screening assays. J Biomol Screen. 1999;4(2):67–73. brane phospholipid dynamics and K-Ras signaling. Science (80-).
26. Visan V, Fajloun Z, Sabatier JM, Grissmer S. Mapping of mau- 2015;349(6250):873–6.
rotoxin binding sites on hKv1.2, hKv1.3, and hlKCa1 channels. 47. Yang M, Brackenbury WJ. Membrane potential and cancer pro-
Mol Pharmacol. 2004;66:1103–12. gression. Front Physiol. 2013;4:1-10.
27. García-Ferreiro RE, Kerschensteiner D, Major F, Monje F, Stüh- 48. Kadir LA, Stacey M, Barrett-Jolley R. Emerging roles of the mem-
mer W, Pardo LA. Mechanism of block of hEag1 ­K+ channels by brane potential: Action beyond the action potential. Front Physiol.
imipramine and astemizole. J Gen Physiol. 2004;124:301-17. 2018;9:1–10.
28. Cazorla O, Lacampagne A, Fauconnier J, Vassort G. SR33805, a 49. King GF. Venoms as a platform for human drugs: translating
­Ca2+ antagonist with length-dependent C ­ a2+-sensitizing properties toxins into therapeutics. Expert Opinion on Biological Therapy.
in cardiac myocytes. Br J Pharmacol. 2003;139:99–108. 2011;11(11):1469–84. Available from: https://doi.org/10.1517/1
29. Church J, Fletcher EJ, Abdel-Hamid K, MacDonald JF. Loperamide 4712598.2011.621940
blocks high-voltage-activated calcium channels and N-methyl- 50. Moreels L, Peigneur S, Galan DT, De Pauw E, Béress L, Waelkens
d-aspartate-evoked responses in rat and mouse cultured hippocam- E, et al. APETx4, a novel sea anemone toxin and a modulator
pal pyramidal neurons. Mol Pharmacol. 1994;45(4):747–57. of the cancer-relevant potassium channel KV10.1. Mar Drugs.
30. Wu Y, Zou B, Liang L, Li M, Tao YX, Yu H, et al. Loperamide 2017;15(9) 1-17.
inhibits sodium channels to alleviate inflammatory hyperalgesia. 51. Twiner MJ, Doucette GJ, Rasky A, Huang XP, Roth BL, Sangui-
Neuropharmacology. 2017;117:282-291. netti MC. Marine algal toxin azaspiracid is an open-state blocker
31. Hagiwara K, Nakagawasai O, Murata A, Yamadera F, Miyoshi I, of hERG potassium channels. Chem Res Toxicol. 2012; 25,
Tan-No K, et al. Analgesic action of loperamide, an opioid ago- 1975-1984.
nist, and its blocking action on voltage-dependent C ­ a2+ channels. 52. Moreels L, Peigneur S, Yamaguchi Y, Vriens K, Waelkens E, Zhu
Neurosci Res. 2003;46(4):493–7. S, et al. Expanding the pharmacological profile of κ-hefutoxin 1
32. Hanauer SB. The role of loperamide in gastrointestinal disorders. and analogues: a focus on the inhibitory effect on the oncogenic
Rev Gastroenterol Disord. 2008;8(1):15–20. channel Kv101. Peptides. 2017;98:43–50.
33. Zahradník I, Minarovič I, Zahradníková A. Inhibition of the car-
diac l-type calcium channel current by antidepressant drugs. J
Pharmacol Exp Ther. 2008;324:977–84.

13

You might also like