Inhibition of Small-Conductance, Ca2+-Activated K+ Current by Ondansetron

Background: Small-conductance Ca2+-activated K+ channels (SK channels) have been proposed as antiarrhythmic targets for the treatment of atrial fibrillation. We previously demonstrated that the 5-HT3 receptor antagonist ondansetron inhibits heterologously expressed, human SK2 (hSK2) currents as well as native cardiac SK currents in a physiological extra-/intracellular [K+] gradient at therapeutic (i.e., sub-micromolar) concentrations. A recent study, using symmetrical [K+] conditions, challenged this result. The goal of the present study was to revisit the inhibitory effect of ondansetron on hSK2-mediated currents in symmetrical [K+] conditions. Experimental Approach: The whole-cell patch clamp technique was used to investigate the effects of ondansetron and apamin on hSK2-mediated currents expressed in HEK 293 cells. Currents were measured in symmetrical [K+] conditions in the presence of 100 nM [Ca2+]o. Results: Expression of hSK2 produced inwardly rectifying whole-cell currents in the presence of 400 nM free cytosolic Ca2+. Ondansetron inhibited whole-cell hSK2 currents with IC 50 values of 154 and 113 nM at −80 and 40 mV, respectively. Macroscopic current inhibited by ondansetron and current inhibited by apamin exhibited inwardly rectifying current-voltage relationships with similar reversal potentials (apamin, ∼5 mV and ondansetron, ∼2 mV). Ondansetron (1 μM) in the continuing presence of apamin (100 nM) had no effect on hSK2-mediated whole-cell currents. Wild-type HEK 293 cells did not express ondansetron- or apamin-sensitive currents. Conclusion: Ondansetron in sub-micromolar concentrations inhibits hSK2 currents even under altered ionic conditions.


INTRODUCTION
Small-conductance Ca 2+ -activated K + channels (SK channels) have been proposed as antiarrhythmic targets for the treatment of atrial fibrillation (Heijman and Dobrev, 2017;Ravens and Odening, 2017). Three SK channel isoforms (SK1, SK2 and SK3, which are encoded by the KCNN1, KCNN2 and KCNN3 genes, respectively) are expressed in the mammalian heart (Tuteja et al., 2005). Under physiological conditions, functional SK channels are predominately present in the atria and significantly contribute to atrial refractoriness and excitability (Xu et al., 2003;Li et al., 2009). In addition to the bee-venom toxin apamin, which selectively blocks SK channels at pico-to nanomolar concentrations (Stocker, 2004;Yu et al., 2014), a number of pharmacological SK channel inhibitors have been developed and shown to reduce susceptibility to atrial fibrillation in animal models (Skibsbye et al., 2011;Skibsbye et al., 2014;El-Haou et al., 2015;Diness et al., 2017;Ravens and Odening, 2017). We previously demonstrated that the 5-HT 3 receptor antagonist ondansetron, a widely prescribed antiemetic, blocks heterologously expressed, human SK2 (hSK2) channels at therapeutic (i.e., sub-micromolar) concentrations (Ko et al., 2018), suggesting its possible repurpose as an atrial-selective antiarrhythmic drug. A recent study by Kirchhoff and coworkers challenged our findings reporting that ondansetron at micromolar concentrations had no significant effect on hSK2mediated currents expressed in human embryonic kidney (HEK) 293 cells (Kirchhoff et al., 2019). Since ionic conditions for hSK2 current measurements that Kirchhoff et al. used (symmetrical [K + ], presence of extracellular Ca 2+ ) were different from those used in our previous experiments (physiological extra-/ intracellular [K + ] gradient, absence of extracellular Ca 2+ ), the goal of the present study was to re-evaluate the efficacy of ondansetron in inhibiting hSK2-mediated currents expressed in HEK 293 cells under ionic conditions replicating those employed by Kirchhoff et al. Our results indicate that ondansetron retains its inhibitory effect on hSK2 channels even under altered ionic conditions.

Electrophysiology
Whole-cell Ca 2+ -activated K + current was recorded from transfected HEK 293 cells at room temperature using the patch-clamp technique in the ruptured patch configuration. The extracellular solution contained (in mmol/L) 0.1 CaCl 2 , 3 MgCl 2 , 154 KCl, 10 HEPES, and 10 glucose . The internal solution contained (in mmol/L) 8.106 CaCl 2 (yielding a free Ca2+ concentration of 400 nM, using the Maxchelator software by C. Patton of Stanford University as we have previously described (Ko, 2018)), 1.167 MgCl 2 , 10 EGTA, 108 KCl, 10 HEPES, 31.25/10 KOH/EGTA, and 15 KOH (pH 7.2). Pipette resistances ranged from 3 to 5 MΩ when filled with the pipette solution. After a gigaohm seal had been achieved, the test pulse current was nulled by adjusting the pipette capacitance compensation. After break in, the whole cell charging transient was nulled by adjusting whole cell capacitance and series resistance. Currents were elicited using a voltage ramp from −80 to 80 mV (0.8 mV/ms) from a holding potential of 0 mV. Voltage ramps were repeated every 2 s. To isolate apamin-or ondansetron-sensitive currents, extracellular solution containing apamin (100 nM) or ondansetron (0.1, 0.5, 1.0 or 5.0 μM) was applied during continuous recordings, and the difference currents between the baseline and the apamin-/ondansetron-containing solution were calculated to be the apamin-or ondansetronsensitive currents (I apamin and I ondan , respectively). To determine the concentration-inhibition relationship for ondansetron of expressed hSK2 currents, cells were first exposed to a given concentration of ondansetron followed by exposure to 100 nM apamin in the continual presence of ondansetron. The amplitude of I ondan was expressed as I ondan / (I ondan + I apamin ) and fit to an equation of the form 1 − I ondan I ondan + I apamin 1 − I ondan,max I ondan,max + I apamin + 1 − 1 − I ondan,max I ondan,max + I apamin where X is the concentration of ondansetron (expressed in logarithmic units), I ondan,max is the maximal I ondan obtainable, and IC 50 is the concentration of ondansetron at which 1-I ondan /(I ondan + I apamin ) equals 0.5. Currents were normalized to cell capacitance to obtain current density. Series resistance was electronically compensated by 70-80%. Voltage-clamp commands were generated with an Axopatch 200B amplifier/Digidata 1440A acquisition system using pCLAMP10 software (Molecular Devices). The electrical signal was filtered with a built-in four-pole low-pass Bessel filter at 1 kHz and digitized at 5 kHz. Whole cell recordings were analyzed using Clampfit 10.2 software (Molecular Devices) and Sigmaplot software (Systat Software, San Jose, CA, United States).

Drugs
Apamin and ondansetron were purchased from Tocris Bioscience. All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO, United States). Apamin and ondansetron were dissolved in double-distilled H 2 O to make stock solutions at concentrations of 10 mM/L and 1 mM/L, respectively. These stock solutions were stored at −20°C and aliquots were solubilized at the desired concentration on the day of the experiment.

Statistics
All numerical values are expressed as mean ± S.E.M. Statistical analysis was performed using Sigmaplot software. Data were analyzed with a t-test where appropriate.

Wild Type Human Embryonic Kidney 293 Cells Do Not Express Apamin-or Ondansetron-Sensitive Currents
In order to rule out the possibility that our results might be confounded by the endogenous expression of apamin-or ondansetron-sensitive currents in HEK 293 cells, we made patch-clamp recordings from wild type HEK 293 cells. Figures  1A,E show representative ramp currents recorded at baseline (before application of apamin or ondansetron, black traces) and after apamin or ondansetron application (red traces). No significant alterations in whole-cell currents by apamin or ondansetron were observed. Plots of peak inward currents at  Figures 1A,E illustrate the absence of significant apamin or ondansetron effects on endogenous currents. Summary data for endogenous currents at test potentials ranging from −80 to +80 mV are shown in Figures 1D,H. Collectively, these data suggest that there is no detectable expression of apamin-or ondansetron-sensitive currents in wild type HEK 293 cells used in the present study.

Block of hSK2-Mediated Currents by Apamin and Ondansetron
Expression of hSK2 produced inwardly rectifying whole-cell currents in the presence of 400 nM free cytosolic Ca 2+ (Figure 2A, black trace). hSK2-mediated currents were blocked by 100 nM apamin ( Figure 2A, red trace, and Figure 2B), but block was incomplete despite addition of a saturating concentration of the bee venom toxin. On average, 100 nM apamin reduced currents measured at −80 and 40 mV by 47.3 ± 4.4% and 41.7 ± 4.8%, respectively (8 hSK2-HEK 293 cells). These values are in close agreement with those previously reported for heterologously expressed whole-cell rat SK2 currents measured in symmetrical [K + ] conditions (Lamy et al., 2010), but significantly smaller than those observed for outside-out macropatch rat SK2 currents (Weatherall et al., 2011). Macroscopic apamin-sensitive currents exhibited  Figures 2E,F), suggesting that the concentration of external Ca 2+ did not significantly affect the potency of apamin block. Next, we sought to examine whether ondansetron at a concentration of 1 μM inhibits hSK2-mediated currents using ionic conditions that are identical to those reported by Kirchhoff et al. Representative whole-cell current traces obtained from ramps from −80 to 80 mV imposed on a voltage-clamped HEK 293 cell expressing hSK2 are demonstrated in Figure 3A.
Ondansetron produced a partial block of hSK2-mediated ramp current (red trace) and addition of apamin on top of ondansetron further reduced its magnitude (green trace). Typically, development of ondansetron block of hSK2-mediated ramp currents took considerably longer time (minutes) to reach steady-state compared to that of apamin-induced block (seconds), as exemplified in Figure 3B. The current-voltage relationships for current inhibited by ondansetron and for current inhibited by cumulatively applied apamin ( Figures  3C,D) suggest that 1 μM ondansetron inhibited a current with properties of apamin-sensitive currents, i.e., a reversal potential (2.2 ± 2.4 mV, n 6 cells.) close to E K (0 mV) and an inwardly rectifying current-voltage relationship. The reversal potential of the ondansetron-sensitive current is not significantly different from that of the apamin-sensitive current (4.6 ± 1.2 mV, n 5 cells; p 0.4 by t-test). To determine the concentrationdependence of ondansetron inhibition, whole-cell ramp currents were measured at a single concentration of ondansetron (0.1, 0.5, 1 or 5 μM) followed by measurements in the presence of cumulatively applied apamin (100 nM). Figure 3D summarizes the I-V relationships for the ondansetron-sensitive component (I ondan ) and the (ondansetron + apamin)-sensitive component (I apamin + I ondan ). Figure 3E shows the fractional block of SK2 currents, plateaued at around 0.7 for micromolar ondansetron concentrations, indicating that not more than ∼30% of the composite, i.e., (ondansetron + apamin)-sensitive, currents were blocked by saturating concentrations of ondansetron under the experimental conditions employed here. Exposure of two hSK2-transfected HEK 293 cells to 1 μM ondansetron in the continuing presence of 100 nM apamin had no effect on ramp currents ( Figures 4A-C), suggesting that apamin occupies the ondansetron binding site or prevents it from binding. The latter finding further indicates that 1 μM ondansetron does not significantly affect native currents expressed in hSK2-HEK 293 cells, confirming our measurements in wild type HEK 293 cells (see Figure 1). Finally, to exclude the possibility that our results are confounded by spontaneous "run-down" of macroscopic currents during extended recordings, ramp currents were elicited every 5 s for 8 min under control conditions in 4 voltage-clamped cells. Representative recordings shown in Figures 4D,E as well as plots of the whole-cell current amplitudes at −80 and 40 mV (normalized to their respective initial values) as a function of time ( Figure 4F) illustrate that the magnitudes of repetitively induced ramp currents remained largely unchanged throughout ∼8-min monitoring periods. Taken together, our data suggest that sub-micromolar concentrations of ondansetron inhibit hSK2-mediated currents expressed in HEK 293 cells in ionic conditions that are identical to those used in the study by Kirchhoff et al.

DISCUSSION
We previously demonstrated, using physiological [K + ] conditions, that ondansetron at sub-micromolar concentrations inhibits hSK2-mediated currents expressed in HEK 293 cells as well as native SK currents in mouse ventricular cardiomyocytes (Ko et al., 2018). The experiments presented here, utilizing symmetrical [K + ], provides additional experimental evidence that ondansetron in sub-micromolar concentrations inhibits hSK2 currents heterologously expressed in HEK 293 cells. We found that ondansetron-sensitive wholecell currents exhibit properties of apamin-sensitive whole-cell currents, i.e., an inwardly rectifying current-voltage relationship and a reversal potential close to the calculated E K .Ondansetron had no significant effect in the continuing presence of apamin, indicating that ondansetron targets hSK2 channels in the experimental conditions employed here. Our findings also exclude the possibility that ondansetron inhibits native currents expressed in wild type HEK 293 cells. These results are in agreement with our previous observations that ondansetron, when applied in the continuing presence of a saturating apamin concentration, had no effect on whole-cell currents of mouse ventricular myocytes (Ko et al., 2018), nor did it prolong ventricular repolarization of isolated perfused rabbit hearts exposed to low [K + ] o or isolated perfused failing rabbit hearts (Chan et al., 2015;Yin et al., 2020). Our results contrast with those recently published by Kirchhoff and co-workers who used identical ionic conditions to determine the effect of ondansetron on hSK2 currents (Kirchhoff et al., 2019). These authors report that ondansetron failed to inhibit macroscopic currents in hSK2-transfected HEK 293 cells at concentrations up to 30 μM. The reasons underlying this apparent discrepancy are unclear. We previously demonstrated that the p. F503L SK2 variant heterologously expressed in HEK 293 cells exhibits reduced sensitivity to inhibition by ondansetron (Ko et al., 2018). Thus, it is possible that the nucleotide sequences of the hSK2 cDNAs that were used for HEK 293 cell transfection differed within segments encoding for the ondansetron-binding motif, altering the affinity of the hSK2 channel to ondansetron. Comparisons of the respective hSK2 cDNA sequences may be useful to obtain information about the SK2 channel motifs that confer ondansetron sensitivity. It is also possible that differences in the experimental protocol are responsible for the discrepant findings. Cells in the study by Kirchhoff et al. were first exposed to bicuculline, a blocker of Ca 2+ -activated K + channels (Khawaled et al., 1999), followed by a brief wash out before ondansetron was applied. Incomplete washout of bicuculline before ondansetron exposure, i.e., recovery from bicuculline block overlapping with ondansetron inhibition, may have masked an inhibitory effect of ondansetron in the study by Kirchhoff et al. Finally, our experiments revealed slow kinetics of the development of ondansetron-induced SK current inhibition (see Figure 3) relative to the kinetics of apamin inhibition, suggesting the possibility that insufficient durations of ondansetron wash-in may have contributed to the reported lack of its effect on hSK2 currents.
Ondansetron at saturating concentrations suppressed maximally 30% of the (ondansetron + apamin)-sensitive current component under the experimental conditions employed here, indicating that ondansetron had lower inhibitory potency than apamin. This finding contrasts with our previous observation that sub-micromolar concentrations of ondansetron and apamin were equipotent in blocking native SK currents in mouse ventricular cardiomyocytes as well as in blocking hSK2 currents expressed in HEK 293 cells in a physiological [K + ] gradient (Ko et al., 2018). We cannot readily explain this apparent discrepancy. Differences in ondansetron binding to and/or differences in ondansetron sensitivity of SK channels may be responsible, resulting from differences in ionic conditions that were used for SK current measurements. Importantly, however, we also demonstrated in the same study that ondansetron at a concentration of 100 nM effectively inhibits apamin-sensitive channels in the intact, isolated perfused mouse heart at physiological temperatures, supporting the drug's utility for in vivo SK current inhibition.
IC 50 values of 154 nM (at −80 mV) and 113 nM (at 40 mV) reported here correspond to ondansetron plasma concentrations of ∼45 and ∼33 ng/ml, respectively. Therapeutic dosages of oral or suppository ondansetron result in peak plasma concentrations ranging from ∼15 to ∼100 ng/ml (de Wit et al., 1996;Jann et al., 1998;Vandenberg et al., 2000), which would suffice to inhibit SK current. The sensitivity of SK channels to ondansetron in clinically relevant concentrations suggests its potential utility as an antiarrhythmic drug. Since under physiological conditions functional SK channels are predominantly present in the atria and control repolarization and excitability (Li et al., 2009;Skibsbye et al., 2014), ondansetron may selectively target atrial SK channels without evoking ventricular proarrhythmia. However, SK channels may importantly contribute to ventricular repolarization in the diseased heart (Chan et al., 2015;Hamilton et al., 2020;Yin et al., 2020), raising the possibility that ondansetron under pathological conditions increases the proarrhythmic risk.

CONCLUSION
Our data indicate that ondansetron at sub-micromolar concentrations blocks hSK2-mediated currents expressed in HEK 293 cells even under altered ionic conditions, largely confirming our previous findings. They further support the notion that ondansetron can be repurposed as an antiarrhythmic agent for treatment of cardiac arrhythmias.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by IACUC, Indiana University School of Medicine.