Bisoprolol, Known to Be a Selective β1-Receptor Antagonist, Differentially but Directly Suppresses IK(M) and IK(erg) in Pituitary Cells and Hippocampal Neurons

Bisoprolol (BIS) is a selective antagonist of β1 adrenergic receptors. We examined the effects of BIS on M-type K+ currents (IK(M)) or erg-mediated K+ currents (IK(erg)) in pituitary GH3, R1220 cells, and hippocampal mHippoE-14 cells. As GH3 cells were exposed to BIS, amplitude of IK(M) was suppressed with an IC50 value of 1.21 μM. The BIS-induced suppression of IK(M) amplitude was not affected by addition of isoproterenol or ractopamine, but attenuated by flupirtine or ivabradine. In cell-attached current, BIS decreased the open probability of M-type K+ (KM) channels, along with decreased mean opening time of the channel. BIS decreased IK(erg) amplitude with an IC50 value of 6.42 μM. Further addition of PD-118057 attenuated BIS-mediated inhibition of IK(erg). Under current-clamp conditions, BIS depolarization increased the firing of spontaneous action potentials in GH3 cells; addition of flupirtine, but not ractopamine, reversed BIS-induced firing rate. In R1220 cells, BIS suppressed IK(M); subsequent application of ML-213(Kv7.2 channel activator) reversed BIS-induced suppression of the current. In hippocampal mHippoE-14 neurons, BIS inhibited IK(M) to a greater extent compared to its depressant effect on IK(erg). This demonstrated that in pituitary cells and hippocampal neurons the presence of BIS is capable of directly and differentially suppressing IK(M) and IK(erg), despite its antagonism of β1-adrenergic receptors.


Introduction
Bisoprolol (zebeta ® ; BIS), a phenoxy-2-propanol derivative, is recognized as an oral synthetic selective blocker of β 1 -adrenoceptor with antioxidant activity which exerts a number of potentially beneficial pharmacological effects such as atrial fibrillation, heart failure and postural tachycardia syndrome [1][2][3][4]. Due to its lipophilic nature, it facilitates entry into brain tissue to produce regulatory actions on central neurons [5,6]. Previous reports have demonstrated that BIS could bind to β 1 -adrenergic receptors inherently existing in brain areas including the pituitary gland and hippocampus [7][8][9][10][11]. An earlier study has also reported the ability of BIS to elevate blood prolactin level [12]. The relationship between the BIS concentration (0.01-30 µM) and the I K(M) amplitude was further constructed and is plotted in Figure 2B. Fitting the concentration-response curve with the modified Hill equation as described in Materials and Methods for details yielded a half-maximal concentration (i.e., IC 50 ) of 1.21 µM and a slope coefficient of 1.2. BIS at a concentration of 30 µM nearly abolished current amplitude. Additionally, in the continued presence of 1-µM BIS, further addition of either 10 µM flupirtine or 10 µM ivabradine was noted to attenuate its suppression of I K(M) ( Figure 2C). Flupirtine was reported to stimulate I K(M) [30], while ivabradine known to suppress hyperpolarization-activated cation current might increase I K(M) amplitude in these cells. Conversely, subsequent application of neither isoproterenol (1 µM) nor ractopamine (10 µM) had any perturbations on BIS-induced block of I K(M) . Isoproterenol or ractopamine (a synthetic phenoethanolamine salt) was previously demonstrated to stimulate β-adrenergic receptors [31][32][33][34][35]. However, addition of neither isoproterenol (1 µM) nor ractopamine (10 µM) alone produced any effects on I K(M) . The results led us to suggest that BIS has a significant depressant action on I K(M) functionally expressed in GH 3 cells, which appears to be unlinked to binding to β-adrenergic receptors.

Effect of Bisoprolol (BIS) on IK(M) in Pituitary GH3 Cells
In the first set of whole-cell current recordings, we bathed the cells in high-K + , Ca 2+ -free Tyrode's solution, the composition of which is described above, and the recording pipette was filled with K +containing solution. Once the whole-cell model was established, the examined cell was maintained at −50 mV and the depolarizing step to −10 mV with a duration of 1 sec was delivered. As shown in Figure 2, upon long-lasting membrane depolarization to −10 mV, K + inward current was readily evoked, which was responsive to low threshold voltage and displayed both the slowly activating and deactivating properties. This type of K + currents in response to depolarizing step was sensitive to inhibition by linopirdine (10 μM) or pioglitazone (10 μM), and it was hence identified as an IK(M) [17][18][19][20]. Of particular interest is how as GH3 cells were exposed to BIS, the IK(M) amplitude in response to such depolarizing step was progressively diminished in a concentration-dependent manner ( Figure  2A and 2B). For example, addition of 1-μM BIS significantly decreased current amplitude from 196 ± 14 to 97 ± 11 pA (n = 12, p < 0.05). After washout of the agent, current amplitude returned to 191 ± 14 pA (n = 11, p < 0.05). Concomitant with this, the value of activation time constant (τact) during the exposure to 1-μM BIS was noted to increase significantly from 102 ± 9 to 182 ± 11 msec (n = 12, p < 0.05) and that of deactivating time constant (τdeact) was decreased from 168 ± 11 to 66 ± 8 msec (n = 12, p < 0.05).
The relationship between the BIS concentration (0.01-30 μM) and the IK(M) amplitude was further constructed and is plotted in Figure 2B. Fitting the concentration-response curve with the modified Hill equation as described in Materials and Methods for details yielded a half-maximal concentration (i.e., IC50) of 1.21 μM and a slope coefficient of 1.2. BIS at a concentration of 30 μM nearly abolished current amplitude. Additionally, in the continued presence of 1-μM BIS, further addition of either 10 μM flupirtine or 10 μM ivabradine was noted to attenuate its suppression of IK(M) ( Figure 2C). Flupirtine was reported to stimulate IK(M) [30], while ivabradine known to suppress hyperpolarization-activated cation current might increase IK(M) amplitude in these cells. Conversely, subsequent application of neither isoproterenol (1 μM) nor ractopamine (10 μM) had any perturbations on BIS-induced block of IK(M). Isoproterenol or ractopamine (a synthetic phenoethanolamine salt) was previously demonstrated to stimulate β-adrenergic receptors [31][32][33][34][35]. However, addition of neither isoproterenol (1 μM) nor ractopamine (10 μM) alone produced any effects on IK(M). The results led us to suggest that BIS has a significant depressant action on IK(M) functionally expressed in GH3 cells, which appears to be unlinked to binding to β-adrenergic receptors.

Inhibitory Effect of BIS on IK(M) Elicited by High Frequency Depolarizing Pulse
Some neurons, neuroendocrine cells, or endocrine cells can display firing behavior with high frequency [25,36,37]. For this reason, in another set of experiments, we sought to evaluate whether BIS affects the magnitude of IK(M) elicited by brief repetitive depolarizations. Under control conditions, a single 1-sec depolarizing step to −10 mV from a holding potential of −50 mV evoked IK(M) amplitude of 156 ± 16 pA (n = 13) with the τact value of 53 ± 7 msec (n = 13). However, the IK(M) amplitude for 1sec repetitive pulses to −10 mV, each of which lasted 6 msec with 4-msec intervals at the level of −50 mV between each abrupt pulse was evidently reduced to 34 ± 5 pA (n = 13), together with a significant prolongation of τact value to 69 ± 8 msec (n = 13). A representative example of IK(M) traces in response to either single pulse or repetitive pulses is illustrated in Figure 3A. As shown in Figure 3B, addition of BIS was able to suppress the IK(M) amplitude elicited by high-frequency depolarizing pulse. For example, as cells were exposed to 3-μM BIS, current amplitude measured at the end of a train of short repetitive stimuli was further reduced by 56 ± 2 % to 19 ± 3 pA (n = 13, p < 0.05). Therefore, IK(M) activation was reduced by repetitive stimuli, and the inhibition by this agent of IK(M) amplitude still remained effective. However, the magnitude of its block appears to become smaller, as compared to those with a single long depolarizing pulse. was compared with the control value (mean ± SEM; n = 12 for each data point). The sigmoidal smooth line represents a best fit to a modified Hill function described in Materials and Methods. The values for IC 50 , maximally inhibited percentage of I K(M) , and the Hill coefficient were 1.21µM, 100 %, and 1.2, respectively. (C) Summary bar graph showing the effects of BIS, BIS plus flupirtine, and BIS plus ivabradine on I K(M) amplitude (mean ± SEM; n = 9 for each bar). BIS: 1-µM BIS; Flu: 10 µM flupirtine; IVA: 10 µM ivabradine. * Significantly different from control (p < 0.05) and ** significantly different from BIS (1 µM) alone group (p < 0.05).

Inhibitory Effect of BIS on I K(M) Elicited by High Frequency Depolarizing Pulse
Some neurons, neuroendocrine cells, or endocrine cells can display firing behavior with high frequency [25,36,37]. For this reason, in another set of experiments, we sought to evaluate whether BIS affects the magnitude of I K(M) elicited by brief repetitive depolarizations. Under control conditions, a single 1-sec depolarizing step to −10 mV from a holding potential of −50 mV evoked I K(M) amplitude of 156 ± 16 pA (n = 13) with the τ act value of 53 ± 7 msec (n = 13). However, the I K(M) amplitude for 1-sec repetitive pulses to −10 mV, each of which lasted 6 msec with 4-msec intervals at the level of −50 mV between each abrupt pulse was evidently reduced to 34 ± 5 pA (n = 13), together with a significant prolongation of τ act value to 69 ± 8 msec (n = 13). A representative example of I K(M) traces in response to either single pulse or repetitive pulses is illustrated in Figure 3A. As shown in Figure 3B, addition of BIS was able to suppress the I K(M) amplitude elicited by high-frequency depolarizing pulse. For example, as cells were exposed to 3-µM BIS, current amplitude measured at the end of a train of short repetitive stimuli was further reduced by 56 ± 2 % to 19 ± 3 pA (n = 13, p < 0.05). Therefore, I K(M) activation was reduced by repetitive stimuli, and the inhibition by this agent of I K(M) amplitude still remained effective. However, the magnitude of its block appears to become smaller, as compared to those with a single long depolarizing pulse.  Current amplitude was measured at the end of single step or repetitive depolarizing pulses (i.e., 1 sec). *or** indicates significantly different from the controls taken with single depolarizing step (p < 0.05) or repetitive stimuli (p < 0.05), respectively.

Effect of BIS on Deactivating IK(M) Elicited Upon Return to Membrane Hyperpolarization with Varying Duration
Previous studies have shown that the magnitude of IK(M) might influence the falling phase of bursting firing or spike after-depolarizations [18,38]. Thus, we wanted to determine how BIS exerted any perturbations on deactivating IK(M) in response to the down sloping ramp pulse from −10 to −50 mV with varying durations. As shown in Figure 4A and 4B, upon return to −50 mV, as the slope of ramp pulse was slowed, the peak amplitude of deactivating IK(M) became exponentially decreased with a time constant of 98 ± 8 msec (n = 11). However, as cells were exposed to 3-μM BIS, the peak amplitude of the current was significantly and exponentially decreased, with a time constant of 65 ± 7 msec (n = 11). For example, as the duration of downslope ramp was set at 40 msec (i.e., slope = −1 V/sec), the addition of 3-μM BIS decreased peak amplitude by 59 ± 2 % from 607 ± 50 to 245 ± 36 pA (n = 11, p < 0.05). The results thus indicated that, as the duration of down sloping ramp pulse increased, the amplitude of deactivating IK(M) would be exponentially decreased, and that the presence of BIS decreased current magnitude in a time-dependent fashion.  In the lower part of (a), current was evoked in response to single membrane depolarization from−50 to −10 mV with a duration of 1 sec (indicated in the upper part of (a)), whereas that in (b) was obtained during rapid repetitive depolarizations to −10 mV (indicated in the lower part of (b)). The individual depolarizing pulse used to elicit inward K + current lasted 6 msec with a total duration of 1 sec. Arrowheads in (a) and (b) indicate the zero current level, while open arrow indicated in (b) depicts the level of I K(M) elicited. The activation time constants for control I K(M) traces in response to single step and repetitive depolarizing pulses were 53 and 69 msec, respectively. (B) Summary bar graph showing the effects of BIS (1 and 3 µM) on the amplitude of I K(M) during repetitive depolarizations (mean ± SEM; n = 13 for each bar). Current amplitude was measured at the end of single step or repetitive depolarizing pulses (i.e., 1 sec). *or** indicates significantly different from the controls taken with single depolarizing step (p < 0.05) or repetitive stimuli (p < 0.05), respectively.

Effect of BIS on Deactivating I K(M) Elicited Upon Return to Membrane Hyperpolarization with Varying Duration
Previous studies have shown that the magnitude of I K(M) might influence the falling phase of bursting firing or spike after-depolarizations [18,38]. Thus, we wanted to determine how BIS exerted any perturbations on deactivating I K(M) in response to the down sloping ramp pulse from −10 to −50 mV with varying durations. As shown in Figure 4A and 4B, upon return to −50 mV, as the slope of ramp pulse was slowed, the peak amplitude of deactivating I K(M) became exponentially decreased with a time constant of 98 ± 8 msec (n = 11). However, as cells were exposed to 3-µM BIS, the peak amplitude of the current was significantly and exponentially decreased, with a time constant of 65 ± 7 msec (n = 11). For example, as the duration of downslope ramp was set at 40 msec (i.e., slope = −1 V/sec), the addition of 3-µM BIS decreased peak amplitude by 59 ± 2 % from 607 ± 50 to 245 ± 36 pA (n = 11, p < 0.05). The results thus indicated that, as the duration of down sloping ramp pulse increased, the amplitude of deactivating I K(M) would be exponentially decreased, and that the presence of BIS decreased current magnitude in a time-dependent fashion.

Inhibitory Effect of BIS on the Activity of Single M-type K + (KM) Channels in GH3 Cells
The cell-attached configuration of the patch-clamp technique was used to investigate the effects of BIS on the activity of single KM channels expressed in these cells. In these experiments, we bathed cells in high-K + , Ca 2+ -free solution and the recording pipette was filled with low-K + (5.4 mM) solution. As shown in Figure 5A, the activity of KM channels, which occurred in rapid open-closed transitions, was clearly detected, as the membrane was maintained at 0 mV relative to the bath. The addition of 1-μM BIS significantly decreased channel open probability from 0.129 ± 0.006 to 0.056 ± 0.003 (n = 12, p < 0.01). After washout of the drug, channel activity was returned to 0.114 ± 0.004 (n = 10, p < 0.01). Based on an amplitude histogram ( Figure 5B), the single-channel amplitude did not differ between the absence and presence of 1-μM BIS. Moreover, in continued presence of 1 μM BIS, further addition of neither diazoxide (10 μM), 9-phenenthrol (10 μM), nor GMQ (10 μM) attenuated its suppression of channel activity. Diazoxide and 9-phenanthrol are the openers of ATP-sensitive K + and intermediate-conductance Ca 2+ -activated K + channels [39], respectively, while GMQ can activate large-conductance Ca 2+ -activated K + channels [40]. These results strongly indicate that the channel activity suppressed by the presence of BIS ascribes primarily from KM channels.

Inhibitory Effect of BIS on the Activity of Single M-type K + (K M ) Channels in GH 3 Cells
The cell-attached configuration of the patch-clamp technique was used to investigate the effects of BIS on the activity of single K M channels expressed in these cells. In these experiments, we bathed cells in high-K + , Ca 2+ -free solution and the recording pipette was filled with low-K + (5.4 mM) solution. As shown in Figure 5A, the activity of K M channels, which occurred in rapid open-closed transitions, was clearly detected, as the membrane was maintained at 0 mV relative to the bath. The addition of 1-µM BIS significantly decreased channel open probability from 0.129 ± 0.006 to 0.056 ± 0.003 (n = 12, p < 0.01). After washout of the drug, channel activity was returned to 0.114 ± 0.004 (n = 10, p < 0.01). Based on an amplitude histogram ( Figure 5B), the single-channel amplitude did not differ between the absence and presence of 1-µM BIS. Moreover, in continued presence of 1 µM BIS, further addition of neither diazoxide (10 µM), 9-phenenthrol (10 µM), nor GMQ (10 µM) attenuated its suppression of channel activity. Diazoxide and 9-phenanthrol are the openers of ATP-sensitive K + and intermediate-conductance Ca 2+ -activated K + channels [39], respectively, while GMQ can activate large-conductance Ca 2+ -activated K + channels [40]. These results strongly indicate that the channel activity suppressed by the presence of BIS ascribes primarily from K M channels.

Effect of BIS on Mean Open Time of KM Channels in GH3 Cells
After observing that during the exposure to BIS, the open-time duration of KM channels in GH3 tended to be shortened with no change in single-channel amplitude ( Figure 5A), we further examined and analyzed the kinetic properties of KM channels obtained with or without addition of BIS. As depicted in Figure 6, the distribution of open durations was least-squares fitted by a monoexponential function. During the exposure to 1-μM BIS, the mean open time of these channels was significantly reduced to 1.67 ± 0.05 msec from the control length of 2.32 ± 0.09 msec (n = 12, p < 0.05). Therefore, BIS-induced changes in KM-channel activity may have been a result of the decrease in the duration of the open state, rather than of any modification in single-channel conductance.

Effect of BIS on Mean Open Time of K M Channels in GH 3 Cells
After observing that during the exposure to BIS, the open-time duration of K M channels in GH 3 tended to be shortened with no change in single-channel amplitude ( Figure 5A), we further examined and analyzed the kinetic properties of K M channels obtained with or without addition of BIS. As depicted in Figure 6, the distribution of open durations was least-squares fitted by a mono-exponential function. During the exposure to 1-µM BIS, the mean open time of these channels was significantly reduced to 1.67 ± 0.05 msec from the control length of 2.32 ± 0.09 msec (n = 12, p < 0.05). Therefore, BIS-induced changes in K M -channel activity may have been a result of the decrease in the duration of the open state, rather than of any modification in single-channel conductance.

Effect of BIS on Mean Open Time of KM Channels in GH3 Cells
After observing that during the exposure to BIS, the open-time duration of KM channels in GH3 tended to be shortened with no change in single-channel amplitude ( Figure 5A), we further examined and analyzed the kinetic properties of KM channels obtained with or without addition of BIS. As depicted in Figure 6, the distribution of open durations was least-squares fitted by a monoexponential function. During the exposure to 1-μM BIS, the mean open time of these channels was significantly reduced to 1.67 ± 0.05 msec from the control length of 2.32 ± 0.09 msec (n = 12, p < 0.05). Therefore, BIS-induced changes in KM-channel activity may have been a result of the decrease in the duration of the open state, rather than of any modification in single-channel conductance.

Lack of BIS Effect on Single-Channel Conductance of K M Channels
We next measured the amplitude and open probability of single K M channels at a series of voltages ranging between −40 and +30 mV. Throughout the voltage range examined, averaged I-V relationships of K M channels with or without BIS (1 µM) addition were analyzed and then compared ( Figure 7A). In the control, fitting these single-channel amplitudes with a linear regression revealed K M channels of 19.8 ± 0.4 pS (n = 9). The value for these channels did not differ significantly from that (19.7 ± 0.4 pS, n = 9, p > 0.05) during exposure to 1-µM BIS. It is clear from the results that addition of this compound is unable to modify single-channel conductance, although it can significantly exert a depressant on the probability of K M channels that would be open.

Lack of BIS Effect on Single-Channel Conductance of KM Channels
We next measured the amplitude and open probability of single KM channels at a series of voltages ranging between -40 and +30 mV. Throughout the voltage range examined, averaged I-V relationships of KM channels with or without BIS (1 μM) addition were analyzed and then compared ( Figure 7A). In the control, fitting these single-channel amplitudes with a linear regression revealed KM channels of 19.8 ± 0.4 pS (n = 9). The value for these channels did not differ significantly from that (19.7 ± 0.4 pS, n = 9, p > 0.05) during exposure to 1-μM BIS. It is clear from the results that addition of this compound is unable to modify single-channel conductance, although it can significantly exert a depressant on the probability of KM channels that would be open.  Figure 7B shows the activation curve of KM channels taken with or without addition of 1-μM BIS. In these experiments, channel open probabilities were measured at different voltages relative to the bath. The plot of channel open probability of KM channels as a function of Δvoltage was derived and then least-squares fitted with the Boltzmann function as described in Materials and Methods for details. In control (i.e., in the absence of BIS), Pmax = 0.99 ± 0.01, V1/2 = −19.2 ± 1.2 mV, q = 2.83 ± 0.05 e (n = 9), whereas in the presence of 1-μM BIS, Pmax = 0.45 ± 0.01, V1/2 = −12.1 ± 0.9 mV, q = 2.85 ± 0.05 e (n = 9). Therefore, it is clear from these results that the presence of BIS not only produced a decrease in the maximal open probability of these channels, but also significantly shifted the activation curve along the voltage axis to depolarized potential by approximately 7 mV. In contrast, minimal change on the gating charge of such activation curve was demonstrated in the presence of BIS. Consequently, as cells were exposed to 1-μM BIS, the difference in free energy (ΔG = -q × F × V1/2, where F = 23,063 cal × V −1 × e −1 ) required for activation of KM channels was significantly reduced from 1.25 ± 0.04 to 0.79 ± 0.04 Kcal (n = 9, p < 0.05).

Effect of BIS on Erg-Mediated K + Current (IK(erg)) in GH3 Cells
Previous studies have reported the importance of IK(erg) functionally expressed in pituitary cells including GH3 cells [41,42]. Some of β-adrenoceptor blockers have been previously demonstrated to suppress HERG K + or other types of voltage-gated K + channels [13,14]. As such, we further studied the possible perturbation by BIS on this type of K + currents (i.e., IK(erg)). Interestingly, as cells were  Figure 7B shows the activation curve of K M channels taken with or without addition of 1-µM BIS. In these experiments, channel open probabilities were measured at different voltages relative to the bath. The plot of channel open probability of K M channels as a function of ∆voltage was derived and then least-squares fitted with the Boltzmann function as described in Materials and Methods for details. In control (i.e., in the absence of BIS), P max = 0.99 ± 0.01, V 1/2 = −19.2 ± 1.2 mV, q = 2.83 ± 0.05 e (n = 9), whereas in the presence of 1-µM BIS, P max = 0.45 ± 0.01, V 1/2 = −12.1 ± 0.9 mV, q = 2.85 ± 0.05 e (n = 9). Therefore, it is clear from these results that the presence of BIS not only produced a decrease in the maximal open probability of these channels, but also significantly shifted the activation curve along the voltage axis to depolarized potential by approximately 7 mV. In contrast, minimal change on the gating charge of such activation curve was demonstrated in the presence of BIS. Consequently, as cells were exposed to 1-µM BIS, the difference in free energy (∆G = -q × F × V 1/2 , where F = 23,063 cal × V −1 × e −1 ) required for activation of K M channels was significantly reduced from 1.25 ± 0.04 to 0.79 ± 0.04 Kcal (n = 9, p < 0.05).

Effect of BIS on Erg-Mediated K + Current (I K(erg) ) in GH 3 Cells
Previous studies have reported the importance of I K(erg) functionally expressed in pituitary cells including GH 3 cells [41,42]. Some of β-adrenoceptor blockers have been previously demonstrated to suppress HERG K + or other types of voltage-gated K + channels [13,14]. As such, we further studied the possible perturbation by BIS on this type of K + currents (i.e., I K(erg) ). Interestingly, as cells were exposed to different concentrations of BIS, the amplitude of I K(erg) elicited by membrane hyperpolarization progressively decreased in a concentration-dependent manner ( Figure 8A,B). From nonlinear least-squares fit to the data points, the IC 50 value needed to exert its inhibitory effect on the amplitude of deactivating I K(erg) was calculated to be 6.42 µM with a Hill coefficient of 1.2, and the agent at a concentration of 100 µM could nearly abolish current amplitude. Moreover, during continued presence of 3-µM BIS, subsequent addition of 10-µM PD-118057 significantly attenuated the decrease of I K(erg) produced by this agent. PD-118057 was previously reported to activate I K(erg) in heart cells [43]. Therefore, BIS could effectively suppress I K(erg) amplitude, despite its inhibitory effect on I K(erg) to a lesser extent as compared with that on I K(M) described above. Similar to previous observations [14] showing the ability of propranolol and its derivative to block HERG channels, BIS-induced inhibition of I K(erg) tends to be independent of its agonistic effect on β-adrenergic receptors.
Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 9 of 20 exposed to different concentrations of BIS, the amplitude of IK(erg) elicited by membrane hyperpolarization progressively decreased in a concentration-dependent manner ( Figure 8A and 8B). From nonlinear least-squares fit to the data points, the IC50 value needed to exert its inhibitory effect on the amplitude of deactivating IK(erg) was calculated to be 6.42 μM with a Hill coefficient of 1.2, and the agent at a concentration of 100 μM could nearly abolish current amplitude. Moreover, during continued presence of 3-μM BIS, subsequent addition of 10-μM PD-118057 significantly attenuated the decrease of IK(erg) produced by this agent. PD-118057 was previously reported to activate IK(erg) in heart cells [43]. Therefore, BIS could effectively suppress IK(erg) amplitude, despite its inhibitory effect on IK(erg) to a lesser extent as compared with that on IK(M) described above. Similar to previous observations [14] showing the ability of propranolol and its derivative to block HERG channels, BISinduced inhibition of IK(erg) tends to be independent of its agonistic effect on β-adrenergic receptors.

The Inhibition of Delayed-Rectifier K + Current (I K(DR) ) in GH 3 cells Caused by the Presence of BIS
An earlier report showed the ability of carvedilol to suppress I K(DR) in NG108-15 neuronal cells [13]. We also explored whether or not the presence of BIS produced any effects on another types of K + currents (e.g., I K(DR) ) in these cells. In these experiments, we bathed cells in Ca 2+ -free Tyrode's solution, and, once whole-cell mode was firmly established, the examined cells were maintained at −50 mV and a series of depolarizing voltage steps ranging between −50 and +60 mV was then delivered. BIS at a concentration of 10 µM was not found to have significant effect on I K(DR) elicited throughout the entire voltage-clamp steps applied (Figure 9). However, upon cell exposure to BIS (30 µM), the I K(DR) amplitude was significantly decreased. For example, at the level of +50 mV, current amplitude decreased from 352 ± 31 to 235 ± 25 pA (n = 11, p < 0.05). Therefore, distinct from I K(M) or I K(erg) , the I K(DR) is relatively resistant to suppression by BIS. An earlier report showed the ability of carvedilol to suppress IK(DR) in NG108-15 neuronal cells [13]. We also explored whether or not the presence of BIS produced any effects on another types of K + currents (e.g., IK(DR)) in these cells. In these experiments, we bathed cells in Ca 2+ -free Tyrode's solution, and, once whole-cell mode was firmly established, the examined cells were maintained at −50 mV and a series of depolarizing voltage steps ranging between −50 and +60 mV was then delivered. BIS at a concentration of 10 μM was not found to have significant effect on IK(DR) elicited throughout the entire voltage-clamp steps applied (Figure 9). However, upon cell exposure to BIS (30 μM), the IK(DR) amplitude was significantly decreased. For example, at the level of +50 mV, current amplitude decreased from 352 ± 31 to 235 ± 25 pA (n = 11, p < 0.05). Therefore, distinct from IK(M) or IK(erg), the IK(DR) is relatively resistant to suppression by BIS.

Effect of BIS on Spontaneous Action Potentials (APs) in GH3 Cells
In a separate set of experiments, we explored whether the presence of BIS perturbs any changes in membrane potential recorded from these cells. As shown in Figure 10A, as the current-clamp voltage recordings were established, spontaneous APs with a firing frequency of 0.85 ± 0.06 Hz (n = 8) were readily observed. As cells were exposed to BIS, membrane became depolarized and the firing frequency was concomitantly raised, as evidenced by a significant raise in the firing frequency to 1.62 ± 0.06 Hz (n = 8, p < 0.05) during the exposure to 3-μM BIS. In continued presence of 3-μM BIS, subsequent addition of 10 μM flupirtine significantly attenuated BIS-induced increase in the firing of spontaneous APs to 1.16 ± .0.06 Hz (n = 8, p < 0.05); however, that of 10 μM ractopamine did not have any effects on it ( Figure 10B). It is therefore conceivable that the observed effect of BIS on the firing patterns of GH3 cells appears to be independent of binding to β-adrenergic receptors and could be partly explained by its suppression of IK(M).

Effect of BIS on Spontaneous Action Potentials (APs) in GH 3 Cells
In a separate set of experiments, we explored whether the presence of BIS perturbs any changes in membrane potential recorded from these cells. As shown in Figure 10A, as the current-clamp voltage recordings were established, spontaneous APs with a firing frequency of 0.85 ± 0.06 Hz (n = 8) were readily observed. As cells were exposed to BIS, membrane became depolarized and the firing frequency was concomitantly raised, as evidenced by a significant raise in the firing frequency to 1.62 ± 0.06 Hz (n = 8, p < 0.05) during the exposure to 3-µM BIS. In continued presence of 3-µM BIS, subsequent addition of 10 µM flupirtine significantly attenuated BIS-induced increase in the firing of spontaneous APs to 1.16 ± 0.06 Hz (n = 8, p < 0.05); however, that of 10 µM ractopamine did not have any effects on it ( Figure 10B). It is therefore conceivable that the observed effect of BIS on the firing patterns of GH 3 cells appears to be independent of binding to β-adrenergic receptors and could be partly explained by its suppression of I K(M) .

Suppressive Effect of BIS on IK(M) in Pituitary R1220 Cells
We next wanted to test if IK(M) in another type of pituitary cells (e.g., pituitary R1220 cells) can be influenced by BIS. This type of rat pituitary cells, originally derived from ScienCell (Carlsbad, CA), was previously reported to express β1-adrenergic receptors [8]. As depicted in Figure 11, as cells were exposed to different BIS concentrations, the amplitude of IK(M) in response to depolarizing step was progressively decreased, in combination with a slowing in activation time course of IK(M). Moreover, with the presence of 1-μM BIS, further addition of ML-213 (10 μM), an activator of IK(M) [44], was able to attenuate BIS-induced inhibition of IK(M), despite inability of further isoproterenol addition to exert any effect on such block. These results thus appear to be indistinguishable from those described above in GH3 cells.

Suppressive Effect of BIS on I K(M) in Pituitary R1220 Cells
We next wanted to test if I K(M) in another type of pituitary cells (e.g., pituitary R1220 cells) can be influenced by BIS. This type of rat pituitary cells, originally derived from ScienCell (Carlsbad, CA, USA), was previously reported to express β 1 -adrenergic receptors [8]. As depicted in Figure 11, as cells were exposed to different BIS concentrations, the amplitude of I K(M) in response to depolarizing step was progressively decreased, in combination with a slowing in activation time course of I K(M) . Moreover, with the presence of 1-µM BIS, further addition of ML-213 (10 µM), an activator of I K(M) [44], was able to attenuate BIS-induced inhibition of I K(M) , despite inability of further isoproterenol addition to exert any effect on such block. These results thus appear to be indistinguishable from those described above in GH 3 cells.

Effect of BIS on I K(M) Recorded from Hippocampal mHippoE-14 Cells
In a final set of experiments, we wanted to determine whether or not I K(M) inherently in central neurons (e.g., hippocampal mHippoE-14 cells) [21] is subject to suppression by this compound. The experimental profiles used were similar to those for pituitary cells. As depicted in Figure 12A,B, the I K(M) elicited by different depolarizing steps was suppressed by the presence of BIS. For example, at the level of −10 mV, the exposure to 3-µM BIS caused a significant reduction of I K(M) amplitude from 109 ± 9 pA to 27 ± 6 pA (n = 9, p < 0.05). After washout of BIS, current amplitude returned to 98 ± 9 pA (n = 9, p < 0.05). The averaged I-V relationships of I K(M) obtained with or without BIS addition are illustrated in Figure 12B. However, isoproterenol (1 µM) had little or no effects on I K(M) amplitude (108 ± 9 pA [control] versus 107 ± 10 pA [in the presence of 1 µM isoproterenol], n = 9, p > 0.05). Therefore, the magnitude of BIS-induced block of I K(M) taken from mHippoE-14 cells is indistinguishable from those described above in pituitary cells.

Effect of BIS on IK(M) Recorded from Hippocampal mHippoE-14 Cells
In a final set of experiments, we wanted to determine whether or not IK(M) inherently in central neurons (e.g., hippocampal mHippoE-14 cells) [21] is subject to suppression by this compound. The experimental profiles used were similar to those for pituitary cells. As depicted in Figure 12A and 12B, the IK(M) elicited by different depolarizing steps was suppressed by the presence of BIS. For example, at the level of −10 mV, the exposure to 3-μM BIS caused a significant reduction of IK(M) amplitude from 109 ± 9 pA to 27 ± 6 pA (n = 9, p < 0.05). After washout of BIS, current amplitude returned to 98 ± 9 pA (n = 9, p < 0.05). The averaged I-V relationships of IK(M) obtained with or without BIS addition are illustrated in Figure 12B. However, isoproterenol (1 μM) had little or no effects on IK(M) amplitude (108 ± 9 pA [control] versus 107 ± 10 pA [in the presence of 1 μM isoproterenol], n = 9, p > 0.05). Therefore, the magnitude of BIS-induced block of IK(M) taken from mHippoE-14 cells is indistinguishable from those described above in pituitary cells.

Discussion
BIS-induced suppression of IK(M) presented herein was found to be attenuated by a further application of flupirtine, but not by isoproterenol or ractopamine, the agonist of β-adrenergic receptor. The present results are of particular importance, as they suggest that the inhibition of both IK(M) and IK(erg) caused by BIS is direct and appears to be unnecessarily linked to its binding to βadrenergic receptor, although binding to β1-adrenergic receptor was previously reported to influence

Discussion
BIS-induced suppression of I K(M) presented herein was found to be attenuated by a further application of flupirtine, but not by isoproterenol or ractopamine, the agonist of β-adrenergic receptor. The present results are of particular importance, as they suggest that the inhibition of both I K(M) and I K(erg) caused by BIS is direct and appears to be unnecessarily linked to its binding to β-adrenergic receptor, although binding to β 1 -adrenergic receptor was previously reported to influence pituitary function [7,8,11,12,45]. However, distinguishable from the inhibitory effect of carvedilol, a non-selective blocker of β-adrenergic receptors, on the amplitude and inactivation kinetics I K(DR) [13], BIS at a concentration greater than 10 µM slightly suppressed I K(DR) amplitude with no discernible change in the inactivation time course of the current.
A previous report has demonstrated the ability of BIS to reverse the down-regulation in mRNA expression of small-conductance Ca 2+ -activated K + channels [15]. In our study, neither continued presence of BIS, further addition of neither diazoxide, GMQ, nor 9-phenanthrol produced attenuating effects on its suppression of I K(M) . Diazoxide and 9-phenanthrol are known to stimulate ATP-sensitive and intermediate-conductance Ca 2+ -activated K + channels, respectively, while 2-guanidine-4-methylquinazoline (GMQ) was reported to enhance the activity of large-conductance Ca 2+ -activated K + channels [39,40]. In this scenario, the observed effect of BIS on whole-cell I K(M) inherently in pituitary GH 3 cells virtually is not involved in the suppression of either ATP-sensitive or Ca 2+ -activated K + channels. The inhibitory effect on I K(M) by BIS could be predominantly responsible for its decrease of macroscopic I K(M) amplitude carried through K M channels.
Block of I K(M) caused by BIS is not instantaneous, but develops with time after the channels are opened upon rapid membrane depolarization, producing an apparent slowing in current activation. Deactivating current of I K(M) in the presence of BIS showed a blunted peak and an increased decay that is consistent with the possibility that the closing (i.e., deactivating process) of channels was raised by unbinding of the BIS molecule. Moreover, as the falling phase upon return to hyperpolarizing potential (i.e., the down sloping ramp) was prolonged, the I K(M) amplitudes were decreased exponentially. The magnitude of BIS-induced block on such deactivating currents was notably increased. The blocking site of this agent thus appears to be located within the channel pore only when the channel is open. KCNQ2/3 mRNA expression was detected in GH 3 and mHippoE-14 cells; however, the present finding showed a lack of effect of BIS on single-channel conductance of K M channels in GH 3 cells, suggesting that the interaction of the channel with it seems to be secondary to alterations that are remote from the pore region of the channel.
Clinically relevant serum concentration of BIS was recently reported to range between 3.8 and 71.1 ng/mL (or 0.011 and 0.22 µM) [46]. These values are similar to the IC 50 required for its suppression of I K(M) presented herein. In this study, the IC 50 value required for BIS-induced block of I K(M) or I K(erg) inherently in GH 3 cells was 1.21 and 6.42 µM, respectively. This agent tends to be selective for I K(M) over I K(erg) or I K(DR) . There was also a depolarizing shift in the activation curve of K M channels during exposure to BIS, strongly suggesting that this compound could alter the voltage dependence of channel activation accompanied by a decrease of free energy for K M -channel activation. During high frequency firing, the amplitude of I K(M) was reduced and the sensitivity to BIS was attenuated. As the duration of the falling phase upon return from membrane potential of −10 to −50 mV increased, the magnitude of BIS-induced block of I K(M) became raised. Taken together, the sensitivity of neurons or pituitary cells to BIS could depend not only on the BIS concentration achieved, but also on the pre-existing level of the resting potential, the discharge pattern or frequency, or their combinations.
Our study showed that in pituitary cells, the presence of BIS does not appear to bind to β-adrenergic receptors exclusively, although the activity of these receptors might be constitutively active [47]. This compound was capable of suppressing I K(M) and I K(erg) directly and effectively. Under current-clamp recordings, BIS was effective at depolarizing the cell and raising the firing rate of spontaneous APs in GH 3 cells. Subsequent addition of flupirtine, yet not ractopamine, could attenuate BIS-enhanced firing. Therefore, apart from its antagonistic effect on β-adrenergic receptors, the inhibition by BIS of these K + currents may also conceivably contribute to its effect on the cellular function (e.g., stimulus-secretion coupling) in these cells. However, it also needs to be noted that the results presented here were derived from cell lines. Whether similar findings occurring in native cells still remains to be clarified.
Neither eugenolol nor eugenodilol was found to suppress I K(M) in GH 3 cells, although these newly designed compounds possess blocking actions on β-adrenergic receptors [48,49]. BIS or its structurally similar compounds may thus be a valuable tool for probing the structure and function of K M or K erg channels, because the pore region of the channel protein to which they bind is of particular relevance for an open-channel blockade. In addition to its antagonistic effect on β 1 -adrenergic receptors, the blockade of K M and K erg channels by BIS, together with an increase in the firing of spontaneous APs, may be responsible for its effects on pituitary or neuronal excitability. To what extent BIS-induced suppression of these K + currents is responsible for its additional effectiveness with favorable metabolic profiles [2,50] has yet to be further delineated. Caution needs to be made in attributing its use to the selective blocking of β 1 -adrenergic receptors. To record whole-cell I K(M) , high K + -bathing solution was composed of the following (in mM): KCl 145, MgCl 2 0.53, and HEPES-KOH 5 (pH 7.4). To measure the activity of single K M channels, the pipette solution was composed of the following (in mM): NaCl 136.5, KCl 5.4, MgCl 2 0.53, and HEPES-NaOH buffer 5 (pH 7.4). All solutions were prepared using deionized water from a Milli-Q water purification system (APS Water Services, Inc., Van Nuys, CA, USA). The pipette solution and culture medium were commonly filtered on the day of use with Acrodisc ® syringe filter with 0.2 µm of Supor ® membrane (Pall Corp., Port Washington, NY, USA). , were maintained in Ham's F-12 medium supplemented with 15% horse serum (v/v), 2.5% fetal calf serum (v/v), and 2 mM L-glutamine in a humidified environment of 5% CO 2 /95% air. Rat pituitary cells (#R1220) were purchased from ScienCell Research Laboratories, Inc. (Carlsbad, CA, USA). These cells reported to express the receptors of gonadotropin releasing hormone (GnRH) were isolated from neonate day-8 CD ® rats and cryopreserved in primary cultures with further purification and expansion (https://www.sciencellonline.com/products-services/primary-cells/animal/cell-types/ pituitarycells/rat-pituitary-cells.html) [51,52]. Cells were routinely grown in Epithelial Cell Medium (Cat #4101; ScienCell).

Cell Preparations
Embryonic mouse hippocampal cell line (mHippoE-14, CLU198) was obtained from Cedarlane CELLutions Biosystems, Inc. (Burlington, ON, Canada) [39,53,54]. Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (v/v) and 2 mM L-glutamine. The culture medium was changed every two to three days, and cells underwent passaged when they reached confluence. The experiments were commonly performed 5 or 6 days after cells had been cultured (60-80% confluence).

RNA Isolation and Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)
To detect the expression of rat KCNQ2 and KCNQ3 mRNAs in different types of cells including GH 3 , R1220 cells, a semiquantitative RT-PCR assay was performed. Total RNA samples were extracted from mHippoE-14 cells with TRIzol reagent (Invitrogen) and reverse-transcribed into complementary DNA using Superscript II reverse-transcriptase (Invitrogen). The sequences of forward and reverse oligonucleotide primers used for KCNQ2 were 5 -CCCTGAAAGTCCAAGAGCAG-3 and 5 -AGGCCCCATAGGTTTGAGTT-3 , respectively, while those for KCNQ3 were 5 -GTGGCTTCA GCATCTCACAA-3 and 5 -CTTGTTGGAAGGGGTCCATA-3 , respectively. Amplication of KCNQ2 or KCNQ3 was made using PCR SuperMix from Invitrogen under the following conditions: 35 cycles composed of 30 sec denaturation at 95 • C. 30 sec primer annealing at 62 • C, 1 min extension at 72 • C, and followed by 72 • C for the final extension for 2 min. PCR products were analyzed on 1.5% (v/v) agarose gel containing ethidium bromide and then visualized under ultraviolet light. Optical densities of DNA bands were scanned and quantified by AlphaImager 2200 (ProteinSimple; Santa Clara, CA, USA).

Electrophysiological Measurements
Immediately prior to the experiment, cells were dissociated and an aliquot of cell suspension was transferred to a custom-made recording chamber mounted on the stage of a DM-IL inverted microscope (Leica, Wetzlar, Germany). There were immersed at room temperature (20-25 • C) in HEPES-buffered normal Tyrode's solution, the composition of which is described above. The patch electrodes were fabricated from borosilicate glass capillaries (No. 34500; Kimble Products, Vineland, NJ, USA) on a Narishige PP-83 puller (Narishige, Tokyo, Japan) or a P-97 Flaming/Brown puller (Sutter, Novato, CA, USA), and electrode tips were fire-polished with MF-83 microforge (Narishige). Their resistances in standard pipette and bathing solutions ranged from 3 to 5 MΩ. Recordings of membrane potential or ion currents were measured in the whole-cell or cell-attached mode of the patch-clamp technique with an RK-400 patch-clamp amplifier (Bio-Logic, Claix, France) [42]. The liquid junction potentials were corrected shortly before seal formation was established.

Data Recordings
The data comprising both potential and current traces were stored online in an Acer SPIN−5 touchscreen laptop computer (SP513-52N-55WE; Taipei, Taiwan) at 10 kHz equipped with Digidata 1440A interface (Molecular Devices, Inc., Sunnyvale, CA, USA), which was used for the analog-to-digital/digital-to-analog conversion. During the recordings, the latter device was controlled by pCLAMP 10.7 software (Molecular Devices) run under Windows 10 (Redmond, WA, USA), and the signals were simultaneously monitored on LCD monitor through a USB type-C connection. Current signals were low-pass filtered at 2 kHz with a FL-4 four-pole Bessel filter (Dagan, Minneapolis, MN, USA) to minimize background noise. Through digital-to-analog conversion, various pCLAMP-generated voltage-clamp profiles with either rectangular or ramp waveforms were applied to determine the current-voltage (I-V) relationship of I K(M) , I K(erg) or I K(DR) . As high-frequency stimuli were necessarily applied to the cell, an Astro-med Grass S88X dual output pulse stimulator (Grass Technologies, West Warwick, RI, USA) was used. After the data were digitally collected, we later analyzed them using different analytical tools that include LabChart 7.0 program (AD Instruments; Gerin, Tainan, Taiwan), OriginPro 2016 (Microcal, Northampton, MA, USA), or custom-made macros built under Microsoft Excel ® 2013 (Redmond).

Data Analyses
To determine percentage inhibition of BIS on I K(M) or I K(erg) , we compared the current amplitudes during cell exposure to different BIS concentrations. To record I K(M) , we bathed the cells in high-K + , Ca 2+ -free solution, and depolarizing step from −50 to −10 mV was applied, while, to measure I K(erg) , the examined cell was hyperpolarized from −10 to −100 mV. The concentration-response data for inhibition of I K(M) or I K(erg) were least-squares fitted by a modified Hill function. That is,  50 and n H are the concentration required for 50% inhibition and the Hill coefficient, respectively; and E max is the maximal inhibition of I K(M) or I K(erg) induced by this drug.

Analyses of Single M-Type K + (K M ) Channels
Single K M -channel currents recorded from GH 3 cells or mHippoE-14 neurons were analyzed using pCLAMP 10.7. We determined single-channel amplitudes taken with or without the BIS addition by fitting Gaussian distributions to the amplitude histograms of the closed and open states. The probabilities of K M channel that would be open were defined as N·P O , which is estimated using the following expression: where N is the number of active channels in the patch, A 0 the area under the curve of an all points histogram corresponding to the closed state, and A 1 . . . A n the histogram area that indicate the level of the distinct open state for 1 to n channels in the patch examined. The single-channel conductance of K M channels was calculated using a linear regression with averaged values of single-channel amplitudes measured at different levels of membrane potentials relative to the bath. Open lifetime distributions of K M channels taken with or without BIS addition were least-squares fitted with logarithmically scaled bin width.
To determine voltage-dependence of the inhibitory effect of BIS on the activity of K M channels, the patch with or without BIS addition was held at different membrane potentials. The activation curve of K M -channel openings taken with or without addition of BIS (3 µM) was fitted by the Boltzmann equation: Relative open probability = P max where P max is the maximal probability of K M -channel openings in the control (i.e., in the absence of BIS) maintained at +30 mV relative to the bath, V 1/2 the voltage at which half-maximal activation of K M channels occurs, q the apparent gating charge, F Faraday's constant, R the universal gas constant, T absolution temperature, and F/RT = 0.04 mV −1 .

Statistical Analyses
The linear or nonlinear (e.g., sigmoidal or exponential curves) least-squares fitting routines were appropriately estimated using either the Solver add-in bundled with Microsoft Excel TM 2013 (Redmond) or OriginPro 2016 (OriginLab, Northampton, MA, USA). The values appearing in this study are provided as the mean values ± standard error of mean (SEM) with sample sizes (n) indicating the number of cells from which the experimental data were taken, and error bars are plotted as SEM.
In this study, we intended to make assertions about the variability of means that could be collected from a random cohort derived from the population concerned. For this reason, the standard error could be more appropriate than the standard deviation. The paired or unpaired Student's t-test, or one-way analysis of variance followed by post-hoc Fisher's least-significance difference test for multiple-group comparisons, were implemented for the statistical evaluation of differences among means. We used non-parametric Kruskal-Wallis test, as the assumption of normality underlying ANOVA was violated. Statistical analyses were performed using IBM SPSS ® version 20.0 (IBM Corp., Armonk, NY, USA). A difference with a p value < 0.05 was considered statistically significant, unless otherwise stated.