Effective Perturbations by Phenobarbital on INa, IK(erg), IK(M) and IK(DR) during Pulse Train Stimulation in Neuroblastoma Neuro-2a Cells

Phenobarbital (PHB, Luminal Sodium®) is a medication of the barbiturate and has long been recognized to be an anticonvulsant and a hypnotic because it can facilitate synaptic inhibition in the central nervous system through acting on the γ-aminobutyric acid (GABA) type A (GABAA) receptors. However, to what extent PHB could directly perturb the magnitude and gating of different plasmalemmal ionic currents is not thoroughly explored. In neuroblastoma Neuro-2a cells, we found that PHB effectively suppressed the magnitude of voltage-gated Na+ current (INa) in a concentration-dependent fashion, with an effective IC50 value of 83 µM. The cumulative inhibition of INa, evoked by pulse train stimulation, was enhanced by PHB. However, tefluthrin, an activator of INa, could attenuate PHB-induced reduction in the decaying time constant of INa inhibition evoked by pulse train stimuli. In addition, the erg (ether-à-go-go-related gene)-mediated K+ current (IK(erg)) was also blocked by PHB. The PHB-mediated inhibition on IK(erg) could not be overcome by flumazenil (GABA antagonist) or chlorotoxin (chloride channel blocker). The PHB reduced the recovery of IK(erg) by a two-step voltage protocol with a geometrics-based progression, but it increased the decaying rate of IK(erg), evoked by the envelope-of-tail method. About the M-type K+ currents (IK(M)), PHB caused a reduction of its amplitude, which could not be counteracted by flumazenil or chlorotoxin, and PHB could enhance its cumulative inhibition during pulse train stimulation. Moreover, the magnitude of delayed-rectifier K+ current (IK(DR)) was inhibited by PHB, while the cumulative inhibition of IK(DR) during 10 s of repetitive stimulation was enhanced. Multiple ionic currents during pulse train stimulation were subject to PHB, and neither GABA antagonist nor chloride channel blocker could counteract these PHB-induced reductions. It suggests that these actions might conceivably participate in different functional activities of excitable cells and be independent of GABAA receptors.


Introduction
Phenobarbital (PHB, phenobarbitone, phenobarb), also known by the trade name Luminal Sodium ® , is a medication belonging to a group known as barbiturates. PHB has long been recognized to be an anticonvulsant and a hypnotic because it can facilitate

Electrophysiological Measurements
In the few hours before the experiments, we dispersed Neuro-2a cells with a 1% trypsin/EDTA solution, and a few drops of cell suspension (10 6 /mL) was quickly transferred to a custom-made chamber firmly mounted on the working stage of a DM-IL inverted microscope (Leica; Major Instruments, Kaohsiung, Taiwan). Cells were then bathed at room temperature (20-25 • C) in normal Tyrode's solution, whose composition is described above. Before each experiment, cells were allowed to settle on the chamber's bottom. The patch pipettes were pulled from Kimax ® -51 borosilicate glass tube (#DWK34500-99; Kimble ® , Merck, Tainan, Taiwan) and were further polished to reach their resistances ranging between 3 and 5 MΩ. During the recordings, the electrodes were mounted in an air-tight holder, which had a suction port on the side, and a silver-chloride wire was used to make contact with the internal electrode solution [39]. We recorded varying types of ionic currents (i.e., I Na , I K(erg) , I K(M) , and I K(DR) ) with the whole-cell mode of a modified patch-clamp technique by using either an Axoclamp-2B (Molecular Devices, Sunnyvale, CA) or an RK-400 amplifier (Bio-Logic, Claix, France), as described elsewhere [19,21,26,40]. The liquid junction potentials that occur when the composition of the pipette internal solution differed from that in the bath were zeroed shortly before giga-Ω formation was made, and the whole-cell data were corrected [21]. As pulse train (PT) stimulation was applied to the tested cell, we used an Astro-Med Grass S88X dual output pulse stimulator (Grass; KYS Technology, Tainan, Taiwan).

Data Recordings
The data were stored online in an ASUSPRO-BU4011LG laptop computer (ASUS, Tainan, Taiwan) at the sampling rate of 10 kHz. The computer was equipped by a Digidata ® Biomedicines 2022, 10, 1968 4 of 19 1440A acquisition device (Molecular Device; Bestgen Biotech, New Taipei City, Taiwan), through which analog-to-digital and digital-to-analog conversions were controlled by pCLAMP ® 10.6 software (Molecular Devices). Current signals were low pass filtered at 3 kHz. We analyzed off-line the signals acquired during the experiments by use of different analytical tools, including LabChart TM 7.0 program (AD Instruments; Gerin, Tainan, Taiwan), OriginPro ® 2021 (OriginLab Corp.; Scientific Formosa, Kaohsiung, Taiwan), and custom-made macros built in Excel ® 2022 (Redmond, DC, USA). We created various voltageclamp protocols used in this work from pCLAMP ® 10.6 and, through digital-to-analog conversion, different waveforms were used to investigate the current versus voltage (I-V) relationship, the steady-state inactivation curve, or the recovery of current inactivation for specific ionic currents (i.e., I Na or I K(erg) ) [12,21].

Whole-Cell Data Analyses
In order to evaluate the percentage inhibition of PHB on the I Na amplitude, each tested cell was 30 ms depolarized from −100 to −10 mV, and current magnitudes during cell exposure to different PHB concentrations were measured and compared at the start of the voltage pulse. The concentration of PHB required to suppress 50% of the peak component of I Na activated in response to short depolarizing pulse was thereafter determined using a Hill function: y = E max/ 1 + IC n H

/[PHB] n H
where y = percentage inhibition (%); [PHB] = the PHB concentration applied; n H = the Hill coefficient; IC 50 = the concentration required for a 50% inhibition of I Na ; E max = PHBinduced maximal inhibition of peak I Na .

Curve-Fitting Approximations and Statistical Analyses
Curve parameter estimation was made by a non-linear or linear fitting routine with a least-squares minimization procedure, in which the "Solver" add-in bundled with Excel ® 2021 (Microsoft, Redmond, WA, USA) was employed [41]. The results are presented as the mean ± standard error of the mean (SEM). The sizes of independent samples (n) indicated the cell number from which the investigation was performed. The statistical significance between the two groups was analyzed by using paired or unpaired Student's t-test, while as the difference among more than two groups was needed, post hoc least-significance tests were further performed. The statistical analyses were performed by using SPSS 20.0 software (IBM Corp., Armonk, NY, USA). A statistical significance was considered when p < 0.05.

Effects of PHB on the Amplitude of Voltage-Gated Na + Current (I Na ) in Neuro-2a Cells
In the beginning, the I Na amplitude activated in response to rapid membrane depolarization was tested in different concentrations of PHB. Cells were placed in Ca 2+ -free, Tyrode's solution, which contained 10 mM tetraethylammonium chloride (TEA) and 0.5 mM CdCl 2 . TEA and CdCl 2 are recognized to block voltage-gated K + and Ca 2+ channels, respectively. Upon the short depolarizing pulse from −100 to −10 mV, the peak amplitude of I Na with a rapidly activating and inactivating time course was decreased at the exposure to 30 and 100 µM PHB ( Figure 1A). For example, in the presence of 100 µM PHB, I Na amplitude was profoundly reduced to 812 ± 25 pA (n = 8, p < 0.05) from a control value of 1867 ± 78 pA (n = 8, Figure 1A). The current amplitude was returned to 833 ± 29 pA (n = 8) after the removal of PHB. The time constants of I Na activation, and fast and slow components of current inactivation in the control period (i.e., no PHB) were respectively 0.61 ± 0.01, 0.91 ± 0.01, and 1.93 ± 0.03 ms (n = 8), while those of activation, and fast and slow components of inactivation in the presence of PHB were respectively 0.61 ± 0.01, 0.92 ± 0.01, and 1.94 ± 0.03 ms (n = 8, p > 0.05). It indicates no obvious change in activa-Biomedicines 2022, 10, 1968 5 of 19 tion or inactivation time course of I Na activated by abrupt membrane depolarization was demonstrated as cells were exposed to PHB. plitude was profoundly reduced to 812 ± 25 pA (n = 8, p < 0.05) from a control value of 1867 ± 78 pA (n = 8, Figure 1A). The current amplitude was returned to 833 ± 29 pA (n = 8) after the removal of PHB. The time constants of INa activation, and fast and slow components of current inactivation in the control period (i.e., no PHB) were respectively 0.61 ± 0.01, 0.91 ± 0.01, and 1.93 ± 0.03 ms (n = 8), while those of activation, and fast and slow components of inactivation in the presence of PHB were respectively 0.61 ± 0.01, 0.92 ± 0.01, and 1.94 ± 0.03 ms (n = 8, p > 0.05). It indicates no obvious change in activation or inactivation time course of INa activated by abrupt membrane depolarization was demonstrated as cells were exposed to PHB. The relationship between the PHB concentrations and the peak INa was further constructed. As demonstrated in Figure 1B, the INa amplitudes obtained at different PHB concentrations were collated and then estimated. The cumulative addition of PHB in the range between 3 µM and 1 mM resulted in a concentration-dependent reduction in the peak amplitude of INa. According to a modified Hill equation stated in Materials and Methods, the IC50 value for the PHB-mediated inhibition of peak INa was optimally yielded to be 83 µM. The data, therefore, reflect that the PHB can exert a depressant action on the depolarization-activated INa in a concentration-dependent fashion in Neuro-2a cells.

PHB-Induced Increase in Cumulative Inhibition of INa Inactivation
High-frequency action potential transmission is important for rapid information processing in the central nervous system. INa inactivation has been previously shown to accumulate prior to being activated during repetitive pulse train (PT) stimulation [14,15,40,42]. Therefore, further experiments were performed to explore whether PHB could modify the inactivation process of INa elicited by the PT stimuli. The PT stimuli were induced by repetitive PT depolarizations to −10 mV (20 ms in each pulse with a rate of 40 Hz for 1 s) when the cell was held at −80 mV. In Figure 2  The relationship between the PHB concentrations and the peak I Na was further constructed. As demonstrated in Figure 1B, the I Na amplitudes obtained at different PHB concentrations were collated and then estimated. The cumulative addition of PHB in the range between 3 µM and 1 mM resulted in a concentration-dependent reduction in the peak amplitude of I Na . According to a modified Hill equation stated in Materials and Methods, the IC 50 value for the PHB-mediated inhibition of peak I Na was optimally yielded to be 83 µM. The data, therefore, reflect that the PHB can exert a depressant action on the depolarization-activated I Na in a concentration-dependent fashion in Neuro-2a cells.

PHB-Induced Increase in Cumulative Inhibition of I Na Inactivation
High-frequency action potential transmission is important for rapid information processing in the central nervous system. I Na inactivation has been previously shown to accumulate prior to being activated during repetitive pulse train (PT) stimulation [14,15,40,42]. Therefore, further experiments were performed to explore whether PHB could modify the inactivation process of I Na elicited by the PT stimuli. The PT stimuli were induced by repetitive PT depolarizations to −10 mV (20 ms in each pulse with a rate of 40 Hz for 1 s) when the cell was held at −80 mV. In Figure 2(Aa),B, the I Na inactivation was evoked in Neuro-2a cells by a 1 s PT stimulus from −80 to −10 mV with the single-exponential time constant (τ) of 104 ± 28 ms (n = 7) in the control period (i.e., no PHB). It indicates a considerable slowing in current inactivation during PT stimulation. When cells exposure to PHB (30 or 100 µM), the exponential time course of I Na evoked by the same PT stimulations was strikingly shortened to 81 ± 22 ms (n = 8, p < 0.05) or 45 ± 14 ms (n = 8, p < 0.05), respectively (Figure 2(Ab),B), in addition to a profound decrease in I Na amplitude in response to rapid membrane depolarization. Furthermore, tefluthrin (Tef, 10 µM) could partially reverse the PHB-mediated reduction of current decay with a τ value of 87 ± 23 ms (n = 8, p < 0.05, Figure 2B). Tef is a synthetic insecticide known to activate I Na [43][44][45].
Neuro-2a cells by a 1 s PT stimulus from −80 to −10 mV with the single-exponential time constant (τ) of 104 ± 28 ms (n = 7) in the control period (i.e., no PHB). It indicates a considerable slowing in current inactivation during PT stimulation. When cells exposure to PHB (30 or 100 µM), the exponential time course of INa evoked by the same PT stimulations was strikingly shortened to 81 ± 22 ms (n = 8, p < 0.05) or 45 ± 14 ms (n = 8, p < 0.05), respectively (Figure 2(Ab),B), in addition to a profound decrease in INa amplitude in response to rapid membrane depolarization. Furthermore, tefluthrin (Tef, 10 µM) could partially reverse the PHB-mediated reduction of current decay with a τ value of 87 ± 23 ms (n = 8, p < 0.05, Figure 2B). Tef is a synthetic insecticide known to activate INa [43][44][45]. To provide a single current trace, the right side of (A) shows the expanded records from the broken boxes in (Aa) and (Ab). The asterisks indicate the occurrence of peak I Na (i.e., transient inward deflection) activated during PT stimulation. (B) Summary graph demonstrating the effect of PHB (30 or 100 µM) and PHB plus tefluthrin (Tef, 10 µM) on the time constant (τ) of I Na decay in response to depolarizing PT stimuli from −80 to −10 mV (mean ± SEM; n = 8 for each point). Of note, the PHB addition results in a decrease in the τ value of I Na activated by PT depolarizing pulses; and the subsequent application of 10 µM Tef can overcome the PHB-induced decrease in τ value. * Significantly different from control (p < 0.05), ** significantly different from PHB (30 µM) alone group (p < 0.05), and + significantly different from PHB (100 µM) alone group (p < 0.05).

Effects of PHB on erg-Mediated K + Current (I K(erg) ) Residing in Neuro-2a Cells
We continued to explore the possible perturbations of PHB on the magnitude of I K(erg) . In order to amplify I K(erg) , cells were placed in a high-K + , Ca 2+ -free solution containing 1 µM TTX. The recording electrode was filled up with a K + -enriched solution [18,19,46]. The different command voltages ranging between −120 and −10 mV for 1 s were imposed to induce hyperpolarization-activated I K(erg) when the tested cell was held at −10 mV ( Figure 3A) [18,19,21,26]. The average current-versus-voltage (I-V) relationship of the peak (filled symbols) or sustained (open symbols) component of I K(erg) in the absence (control) or presence of 300 µM PHB was constructed, respectively ( Figure 3B,C). For example, upon membrane hyperpolarization from −10 mV, PHB decreased the peak I K(erg) from a control value (i.e., no PHB) of 686 ± 156 pA (n = 8) to 343 ± 112 pA (n = 8, p < 0.05) at the level of −60 mV and decreased the sustained I K(erg) from 526 ± 82 pA (n = 8) to 222 ± 88 pA (n = 8, p < 0.05). The presence of PHB (300 µM) could suppress both the peak and sustained component of deactivating I K(erg) throughout the entire measurement ( Figure 3C). depolarizing PT stimuli from −80 to −10 mV (mean ± SEM; n = 8 for each point). Of note, the PHB addition results in a decrease in the τ value of INa activated by PT depolarizing pulses; and the subsequent application of 10 µM Tef can overcome the PHB-induced decrease in τ value. * Significantly different from control (p < 0.05), ** significantly different from PHB (30 µM) alone group (p < 0.05), and + significantly different from PHB (100 µM) alone group (p < 0.05).

Effects of PHB on erg-Mediated K + Current (IK(erg)) Residing in Neuro-2a Cells
We continued to explore the possible perturbations of PHB on the magnitude of IK(erg). In order to amplify IK(erg), cells were placed in a high-K + , Ca 2+ -free solution containing 1 µM TTX. The recording electrode was filled up with a K + -enriched solution [18,19,46]. The different command voltages ranging between −120 and −10 mV for 1 s were imposed to induce hyperpolarization-activated IK(erg) when the tested cell was held at −10 mV ( Figure  3A) [18,19,21,26]. The average current-versus-voltage (I-V) relationship of the peak (filled symbols) or sustained (open symbols) component of IK(erg) in the absence (control) or presence of 300 µM PHB was constructed, respectively ( Figures 3B,C). For example, upon membrane hyperpolarization from −10 mV, PHB decreased the peak IK(erg) from a control value (i.e., no PHB) of 686 ± 156 pA (n = 8) to 343 ± 112 pA (n = 8, p < 0.05) at the level of −60 mV and decreased the sustained IK(erg) from 526 ± 82 pA (n = 8) to 222 ± 88 pA (n = 8, p < 0.05). The presence of PHB (300 µM) could suppress both the peak and sustained component of deactivating IK(erg) throughout the entire measurement ( Figure 3C).   We further investigated whether PHB-induced inhibition of I K(erg) was associated with GABA-mediated inhibition or not. FLM is an antagonist of GABA A receptors, while chlorotoxin is a blocker of Cl-channels. In the presence of PHB (300 µM), neither FLM nor chlorotoxin could overcome the PHB-mediated inhibition of I K(erg) activated by longlasting membrane hyperpolarization ( Figure 3D). The results suggested that PHB-induced inhibition of I K(erg) in these cells is unlinked to its propensity to interact with GABA A receptors, as reported previously in other cell types [7,9,33,[47][48][49][50].

Modification by PHB on the Magnitude of I K(erg) Activated by PT Stimulation
We further explored if PHB could modify the extent of I K(erg) activated in the PT hyperpolarizing stimuli. The stimulus protocol, consisting of repetitive hyperpolarization to −110 mV (40 ms in each pulse with a rate of 20 Hz for 1 s), was imposed over the tested cells held at −10 mV. In the absence of PHB (control), the 1 s repetitive hyperpolarization from −10 to −110 mV evoked the I K(erg) decay with a decaying τ of 147 ± 22 ms (n = 7) (Figure 4(Aa)). Upon exposure to 300 µM PHB, the τ value in the exponential time course of decaying I K(erg) evoked by the same train of hyperpolarizing pulses was shorted to 87 ± 11 ms (n = 7, p < 0.05) (Figure 4(Ab)), and the I K(erg) amplitude was reduced, too ( Figure 4B). E-4031 is an inhibitor of I K(erg) [18,51], and E-4031 could further diminish the I K(erg) amplitude in the presence of 300 µM PHB ( Figure 4B). These results indicate that, apart from the reduction in I K(erg) magnitude, the decrease in the decaying of I K(erg) elicited by a 1 s PT hyperpolarizing pulse (i.e., accumulative inactivation of the current) can be enhanced by PHB in these cells.

Slowing in Recovery from I K(erg) Block Caused by PHB in Neuro-2a Cells
We continued to explore if PHB could lead to any modifications on recovery from the block of I K(erg) . The recovery from the current block was conducted by using a two-step voltage-clamp protocol in which the interval of depolarizing command pulses (i.e., conditioning pulse) varies with a geometrics-based progression. After each conditioning pulse, a hyperpolarizing pulse stepped back to −110 mV for 500 ms was applied to evoke deactivating I K(erg) . As the duration of the conditioning pulse was set at 2048 ms, the amplitude of the tail current (i.e., deactivating I K(erg) ) activated by hyperpolarizing step from −10 to −110 mV with a duration of 500 ms was taken as 1.0. The relative amplitude of I K(erg) with different duration of conditioning pulse was measured and then compared. In the absence of PHB (control), the peak amplitude of deactivating I K(erg) was fully restored from the block when the pulse duration was set at 2048 ms or above ( Figure 5(Aa),B). Interestingly, the time course of recovery from the current block was appropriately fitted by a single exponential, and the τ values could be derived. The τ values were 387 ± 12 ms (n = 7) in the absence of PHB and 823 ± 23 ms (n = 7) in the presence of PHB. As a result, PHB produces a considerable lengthening in the recovery from the block of deactivating I K(erg) in Neuro-2a cells.  with different duration of conditioning pulse was measured and then compared. In the absence of PHB (control), the peak amplitude of deactivating IK(erg) was fully restored from the block when the pulse duration was set at 2048 ms or above ( Figure 5 (Aa),B). Interestingly, the time course of recovery from the current block was appropriately fitted by a single exponential, and the τ values could be derived. The τ values were 387 ± 12 ms (n = 7) in the absence of PHB and 823 ± 23 ms (n = 7) in the presence of PHB. As a result, PHB produces a considerable lengthening in the recovery from the block of deactivating IK(erg) in Neuro-2a cells. The amplitude of deactivating current at the conditioning pulse with a duration of 2048 ms was taken as 1.0. The blue square symbols are control data points, and the orange circle symbols were acquired during cell exposure to 300 µM PHB. (C) PHB-mediated changes in I K(erg) magnitude were evoked by the envelop-of-tail test (mean ± SEM; n = 7 for each point). The relationship of the pulse duration versus the relative amplitude of I K(erg) is illustrated. The relative amplitude appearing at the y axis was measured, when the inwardly gradually increasing current (i.e., activating I K(erg) , I act ) activated by conditioning pulse was divided by peak deactivating I K(erg) (I deact ) obtained following a return of the voltage to −110 mV. The decaying rate of I K(erg) evoked during the envelop-of-tail occurred in a single exponential function. Of notice, the relationships in (B,C) are illustrated with a semi-logarithmic plot. The smooth curves with or without the application of 300 µM PHB indicate the best fits to a single exponential function.

Modification by PHB of the Time Course of I K(erg) Evoked by the Envelope-of-Tail Test
Earlier investigations have demonstrated a time-dependent change in I K(erg) activation during the envelope-of-tail test [21][22][23][24]52]. Here, we then analyzed an inward activating and deactivating I K(erg) evoked by varying durations of conditioning pulse (from 4 to 2048 ms with a geometrics-based progression) to −10 mV from −110 mV, and the relationship of the relative amplitude versus the pulse duration was established in Figure 5C. Like earlier investigations [21,24], the envelope-of-tail test tailored for the activation of I K(erg) seen in Neuro-2a cells exhibits a time-dependent exponential decay in the ratio of the relative amplitude (i.e., I act /I deact ) evoked during the pulse durations between 4 and 2048 ms in a geometrics-based progression. When Neuro-2a-cell exposure to 300 µM PHB, the τ value of I K(erg) activated by the envelope-of-tail method was significantly reduced to 64 ± 3 ms (n = 7, p < 0.05) from a control value of 112 ± 11 ms (n = 7). It is conceivable that PHB has the propensity to reduce the time course of I K(erg) activation evoked during the envelope-of-tail voltage protocol residing in Neuro-2a cells.

Modification by PHB of M-Type K + Current (I K(M) ) in Neuro-2a Cells
Next, we continued examining if PHB could modify the magnitude of I K(M) in Neuro-2a cells. Cells were placed in a high-K + , Ca 2+ -free solution, and the recording electrode was filled up with K + -enriched solution. Upon membrane depolarization from −50 to −10 mV, PHB-treated cells exerted an inhibitory effect on I K(M) amplitude ( Figure 6A,B). This PHBmediated inhibition of I K(M) amplitude could not be reversed by adding FLM (10 µM) or chlorotoxin (1 µM) ( Figure 6B). Thus, unlike the stimulatory action of benzodiazepine derivative or cannabidiol on K V 7-encoded currents, PHB exerts a depressant action of I K(M) in Neuro-2a cells. However, the inactivating properties of I K(M) shown in Figure 6A could be due to the interference by other types of delayed-rectifier K+ currents. The averaged I-V relationships of I K(M) with or without PHB (300 µM) were made, and the results are shown in Figure 6C.

Modification by PHB on IK(M) Inactivation during PT Stimulation
A recent study has demonstrated the capability of IK(M) to maintain the availability of NaV channels during prolonged high-frequency firing [27]. Therefore, we further examined if the PHB could modify the extent of IK(M) during PT stimulation in Neuro-2a cells. The results showed that PHB (100 µM or 300 µM) obviously decreased the amplitudes of activating and deactivating IK(M) during 1 s PT stimulation ( Figure 7A-C). For example, PHB (300 µM) decreased the amplitudes of activating IK(M) to 165 ± 21 pA (n = 7, p < 0.05, Figure 7B) and the amplitudes of activating IK(M) to 459 ± 195 pA (n = 7, p < 0.05, Figure 7C). These disclosed that PHB-mediated inhibition of IK(M) remained efficacious upon high PT

Modification by PHB on I K(M) Inactivation during PT Stimulation
A recent study has demonstrated the capability of I K(M) to maintain the availability of Na V channels during prolonged high-frequency firing [27]. Therefore, we further examined if the PHB could modify the extent of I K(M) during PT stimulation in Neuro-2a cells. The results showed that PHB (100 µM or 300 µM) obviously decreased the amplitudes of activating and deactivating I K(M) during 1 s PT stimulation ( Figure 7A-C). For example, PHB (300 µM) decreased the amplitudes of activating I K(M) to 165 ± 21 pA (n = 7, p < 0.05, Figure 7B) and the amplitudes of activating I K(M) to 459 ± 195 pA (n = 7, p < 0.05, Figure 7C). These disclosed that PHB-mediated inhibition of I K(M) remained efficacious upon high PT stimulation. It is conceivable that the availability of Na V channels during high-frequency firing in unclamped cells became further retarded during the exposure to PHB, despite its ability to suppress the I Na amplitude directly, as described above.

Inhibitory Effect of PHB on Delayed-Rectifier K + Current (IK(DR)) Residing in Neuro-2a Cells
The IK(DR) has been previously revealed to be suppressed by pentobarbital, another barbiturate, as well as to display the resurgent K + tail currents during high PT stimulations [50,53] We additionally explored if the PHB could modify the magnitude of IK(DR) in these cells. In this series of experiments, cells were placed in Ca 2+ -free, Tyrode's solution, which contained 1 µM TTX and 0.5 mM CdCl2, and the recording electrode used was filled up with K + -enriched solution. Under the voltage-clamp current recordings, the tested cell was held at −50 mV, and a series of voltage pulses ranging between −60 and +50 mV was imposed over it. In Figure 8A, upon the presence of 300 µM PHB, the IK(DR) amplitude was clearly suppressed, especially at the voltages above −10 mV. The average I-V relationship of IK(DR) is illustrated in Figure 8B.  in (B,C), respectively, demonstrate the activating and deactivating amplitudes of I K(M) in the absence and presence of 100 or 300 µM PHB (mean ± SEM, n = 7 for each point). Activating or deactivating (or tail) amplitude of I K(M) was measured at the end of each depolarizing pulse from −50 to −10 mV or following return to −50 mV, respectively. * Significantly different from control (p < 0.05) and ** significantly different from PHB (100 µM) alone group (p < 0.05).

Inhibitory Effect of PHB on Delayed-Rectifier K + Current (I K(DR) ) Residing in Neuro-2a Cells
The I K(DR) has been previously revealed to be suppressed by pentobarbital, another barbiturate, as well as to display the resurgent K + tail currents during high PT stimulations [50,53] We additionally explored if the PHB could modify the magnitude of I K(DR) in these cells. In this series of experiments, cells were placed in Ca 2+ -free, Tyrode's solution, which contained 1 µM TTX and 0.5 mM CdCl 2 , and the recording electrode used was filled up with K + -enriched solution. Under the voltage-clamp current recordings, the tested cell was held at −50 mV, and a series of voltage pulses ranging between −60 and +50 mV was imposed over it. In Figure 8A, upon the presence of 300 µM PHB, the I K(DR) amplitude was clearly suppressed, especially at the voltages above −10 mV. The average I-V relationship of I K(DR) is illustrated in Figure 8B. . Of note, current amplitudes measured at the voltages above −10 mV were suppressed by adding PHB (300 µM). Current amplitude was measured at the end of each voltage step. ▇: control; ○: during exposure to 300 µM PLB.

PHB-Induced Increase in Cumulative Inhibition of IK(DR) Inactivation in Neuro-2a Cells
In the last series of measurements, we wanted to investigate if PHB can modify the time course of IK(DR) in response to long-lasting PT stimulation intrinsically in Neuro-2a cells. Under the control condition (i.e., no PHB), a single 10 s depolarizing step from −50 to +50 mV resulted in an exponential decline with a time constant of 2.12 ± 0.08 s (n = 8) ( Figure 9A). In contrast, the time constant for 10 s PT stimulation to +50 mV, each of which lasted 100 ms with 100 ms interval at −50 mV between the depolarizing stimuli, was conceivably reduced to 1.78 ± 0.07 s (n = 8, p < 0.05, Figure 9(Ba)). In the presence of 300 µM PHB, the decaying time constant in response to 10 s repetitive depolarizing pulses was obviously decreased to 0.92 ± 0.05 s (n = 8, p < 0.05, Figure 9(Bb)). This decrease in the decaying time constant mediated by PHB could not be modified by FLM (10 µM). Thus, it is possible that the excessive accumulative inactivation of IK(DR) during the exposure to PHB might be irrelevant to GABAA receptors. . Of note, current amplitudes measured at the voltages above −10 mV were suppressed by adding PHB (300 µM). Current amplitude was measured at the end of each voltage step. : control; : during exposure to 300 µM PLB.

PHB-Induced Increase in Cumulative Inhibition of I K(DR) Inactivation in Neuro-2a Cells
In the last series of measurements, we wanted to investigate if PHB can modify the time course of I K(DR) in response to long-lasting PT stimulation intrinsically in Neuro-2a cells. Under the control condition (i.e., no PHB), a single 10 s depolarizing step from −50 to +50 mV resulted in an exponential decline with a time constant of 2.12 ± 0.08 s (n = 8) ( Figure 9A). In contrast, the time constant for 10 s PT stimulation to +50 mV, each of which lasted 100 ms with 100 ms interval at −50 mV between the depolarizing stimuli, was conceivably reduced to 1.78 ± 0.07 s (n = 8, p < 0.05, Figure 9(Ba)). In the presence of 300 µM PHB, the decaying time constant in response to 10 s repetitive depolarizing pulses was obviously decreased to 0.92 ± 0.05 s (n = 8, p < 0.05, Figure 9(Bb)). This decrease in the decaying time constant mediated by PHB could not be modified by FLM (10 µM). Thus, it is possible that the excessive accumulative inactivation of I K(DR) during the exposure to PHB might be irrelevant to GABA A receptors.

Discussion
The noticeable findings demonstrated in this work are that PHB can regulate the magnitude of multiple types of ionic currents (i.e., INa, IK(erg), IK(M), and IK(DR)) residing in Neuro-2a cells. Consistent with previous observations [54], the presence of PHB itself suppressed peak INa in a concentration-dependent manner with an estimated IC50 of 83 µM ( Figure 1). However, the magnitude of PHB-mediated block in these ionic currents failed to be overcome by the subsequent addition of FLM or chlorotoxin. The cumulative inhibition of ionic currents activated during pulse train stimulation was facilitated by PHB. The PHB-induced rise in INa decay during repetitive stimulation was reversed by Tef. These experimental observations reflected that the reduction of INa, IK(erg), IK(M), and IK(DR) caused by PHB in Neuro-2a cells tends to be acute and may be independent of GABAA receptor-mediated chloride currents. In addition, these actions could potentially contribute to PHB-induced perturbations on the electrical behaviors of excitable cells (e.g., the discharge patterns in the high-frequency firing of action potentials [APs]), presuming that similar in vivo findings appear [55,56].
Previous work has demonstrated that the train of depolarizing pulses could be efficacious in perturbing the magnitude of INa, i.e., current decaying over time in an exponential fashion [14]. This blocking effect is essential in regulating sensory transduction because the driven spike rates can reach hundreds of hertz in response to strong stimuli. As a corollary, the reduced availability of NaV channels during prolonged repetitive activity

Discussion
The noticeable findings demonstrated in this work are that PHB can regulate the magnitude of multiple types of ionic currents (i.e., I Na , I K(erg) , I K(M) , and I K(DR) ) residing in Neuro-2a cells. Consistent with previous observations [54], the presence of PHB itself suppressed peak I Na in a concentration-dependent manner with an estimated IC 50 of 83 µM ( Figure 1). However, the magnitude of PHB-mediated block in these ionic currents failed to be overcome by the subsequent addition of FLM or chlorotoxin. The cumulative inhibition of ionic currents activated during pulse train stimulation was facilitated by PHB. The PHB-induced rise in I Na decay during repetitive stimulation was reversed by Tef. These experimental observations reflected that the reduction of I Na , I K(erg) , I K(M), and I K(DR) caused by PHB in Neuro-2a cells tends to be acute and may be independent of GABA A receptormediated chloride currents. In addition, these actions could potentially contribute to PHB-induced perturbations on the electrical behaviors of excitable cells (e.g., the discharge patterns in the high-frequency firing of action potentials [APs]), presuming that similar in vivo findings appear [55,56].
Previous work has demonstrated that the train of depolarizing pulses could be efficacious in perturbing the magnitude of I Na , i.e., current decaying over time in an exponential fashion [14]. This blocking effect is essential in regulating sensory transduction because the driven spike rates can reach hundreds of hertz in response to strong stimuli. As a corollary, the reduced availability of Na V channels during prolonged repetitive activity (i.e., long-lasting trains of high-frequency APs) would attenuate the spike waveform, and Ca 2+ flux and Ca 2+ -dependent exocytosis would be subsequently impaired at the nerve terminal. In our study, PHB was noted to shorten the τ value of I Na decay responding to PT stimulation ( Figure 2). Thus, it would further diminish Na V -channel availability at the same time.
The magnitude of I Na could be rapidly declined in an exponential manner during highfrequency stimulation, as demonstrated previously [14,40,42] (Figure 2). In other words, during prolonged repetitive activity, the availability of Na V channels was profoundly decreased in a time-dependent fashion. By comparison, the decaying rate of I Na during such high PT stimulation was noticeably faster than that of I K(erg) , I K(M) , or I K(DR) (Figures 4, 7 and 9) seen in Neuro-2a cells. Although the gating kinetics or inactivation-dependent mechanisms residing in these K + currents are overly distinguishable [21,24,31,46,52], the difference in current decaying during PT stimulation could be of particular significance. The main reason for this is because the magnitude of these K + currents (i.e., resurgent K + tail currents) could maintain the membrane in a spike-ready state, thus conferring optimal repriming of I Na during the occurrence of prolonged repetitive activity [27,31]. In this scenario, the augmentation of cumulative I K(erg) , I K(M) , or I K(DR) inactivation caused by PHB during high-frequency activity is relevant because it is able to suppress the rapid firing of neuronal APs and cause an additional loss-of-change in stable waveform and high-fidelity synaptic signaling, particularly in high-frequency firing [27]. Therefore, during high-frequency firing of neurons, the presence of PHB itself would decrease the magnitude of post-spike and steady-state currents leading to diminishing the subthreshold depolarized potential.
In the current study, PHB could slow the recovery of the I K(erg) block and increase the decay of I K(erg) activated by the envelope-of-tail method ( Figure 5). These findings suggested that the molecule of PHB is capable of interacting with the open states (conformations) of the K erg channels present in Neuro-2a cells.
Distinguishable from a recent study [28], the I K(M) present in Neuro-2a cells was directly suppressed by PHB. However, it should be emphasized that the I K(M) amplitude can gradually and progressively arise during repetitive APs because of its slow activation and deactivation kinetics [27]. Upon high-frequency stimulation, the accumulation of I K(M) is allowed to hyperpolarize the after-potential and consequently speed up the recovery of Na V channels from inactivation as well to increase the availability of these channels. The PHB facilitated the cumulative inhibition of I K(M) during PT stimulation (Figure 7). It means that the waveform of neuronal action potentials (APs) in high frequency could be distorted by PHB, and presynaptic signaling occurring across the full dynamic range of the system was thereafter potentially perturbed.
Despite its rapid deactivation kinetics, I K(DR) amplitude (e.g., K V 3.1-encoded current), identified previously as a timely resurgent K + current, can be progressively raised during high-frequency stimulation [31]. During high-frequency stimulation, the accumulation of I K(DR) can be allowed to hyperpolarize the after-potential as well as speed up the recovery of Na V channels from inactivation. In this study, we found that the I K(DR) in Neuro-2a cells can be susceptible to being blocked by PHB, and the cumulative inhibition of I K(DR) during prolonged (i.e., 10 s) repetitive stimulation was enhanced ( Figure 9). Consequently, the magnitude of I Na recovery from current inactivation during PT stimulation would become slowed owing to the enhanced cumulative inhibition of I K(erg) , I K(DR) , and I K(M) [27,31,57]. In this scenario, as cells were exposed to PHB, the high-frequency firing of neuronal APs would lose the emergence of APs' stable waveforms leading to serious perturbations in high-fidelity synaptic signaling (e.g., the presynaptic release of neurotransmitters) [56]. In other words, the negative feedback mechanism on K V channel closure during high PT stimulation would be impaired during the exposure to PHB [27,31]. Therefore, susceptibility to PHB in vivo may depend on the pre-existing resting potential level, the high-frequency firing, and the concentration of PHB.
From previous pharmacokinetic studies, the recommended plasma concentrations of PHB for anti-seizure activities ranged between 10 and 35 µg/mL (or 43 and 151 µM), while for prophylaxis against febrile convulsions was around 15 µg/mM (or 65 µM) [3,5,58]. Because PHB is a lipophilic compound, its membrane concentration would be higher than blood levels. This study showed that the PHB concentration for the half-maximal inhibition of I Na seen in Neuro-2a cells was 83 µM, a value within the clinically applied doses. The observed effects of PHB presented herein can achieve clinical concentration, especially for the treatment of seizure activities lined to high-frequency AP firing.
By synergistic inhibition of multiple ionic currents, including I Na , I K(erg) , I K(M) , or I K(DR) in a frequency-dependent manner, PHB might induce other unpredictable electrophysiological responses, in addition to solely being an activator of GABA A receptors. The clinical use of PHB should take these effects into account. However, the findings in the current study are limited to Neuro-2a cells. Further studies will have to validate these in other excitable cells or in vivo.  Acknowledgments: The authors are grateful to Tzu-Hsien Chuang for his zealous assistance with the experiments.

Conflicts of Interest:
All the authors report no declarations of interest that is directly relevant to this study. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

AP
action potential erg ether-à-go-go related gene FLM flumazenil GABA A receptor γ-aminobutyric acid type A receptor I-V current versus voltage IC 50 the concentration required for 50% inhibition I K(DR) delayed-rectifier K + current I K(erg) erg-mediated K + current I K(M) M-type K + current I Na voltage-gated Na + current K erg channel erg-mediated K + channel K V channel voltage-gated K + channel Na V channel voltage-gated Na + channel PHB phenobarbital (phenobarbitone, phenobarb, Luminal Sodium ® ) PT stimulation pulse train stimulation SEM standard error of mean τ time constant TEA tetraethylammonium chloride Tef tefluthrin TTX tetrodotoxin