Background

Safinamide [SAF; (S)-2-((4-((3-fluorobenzyl)oxy)benzyl)amino)propanamide], an α-aminoamide derivative, is an oral drug used as either an anticonvulsant or an add-on treatment for Parkinson’s disease when a patient is having an “off” episode [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]. However, although this compound was formerly investigated as an anticonvulsant, it has not been approved as a standard antiseizure medication in humans. Several studies also reported the efficacy of SAF as an add-on treatment to subthalamic nucleus deep brain stimulation [19, 20]. SAF exerts its effects through various mechanisms, including inhibition of monoamine oxidase (MAO)-B activity [2, 13,14,15, 21,22,23,24].

Other mechanisms may be involved in SAF-mediated modification of the functional activities [21]. SAF has been demonstrated to protect M17 neuronal cells against amyloid-β-induced oxidative stress and senescence [25]. The evidence supports the notion that SAF might lock NaV channels into the inactivated stage to suppress Na+ current [21, 26,27,28,29]. It also elevates blood pressure; SAF-induced hypertension may be associated with the inhibition of MAO-B activity [15, 23].

Voltage-gated Na+ (NaV) channels, which constitute whole-cell voltage-gated Na+ currents, are essential for the generation, initiation, and propagation of action potentials in electrically excitable membranes. Nine α subunits of NaV channels (NaV1.1 – NaV1.9) have been discovered across excitable mammalian tissues, including the central and peripheral nervous systems, the endocrine system, skeletal muscle, and the heart [29,30,31,32]. Upon brief depolarization, NaV channels undergo a rapid transition from a resting state to an open state and then rapidly return to the inactivated state of the channel. The cumulative inhibition of INa during a train of depolarizing stimuli was demonstrated to affect the electrical behavior of excitable cells [33,34,35,36]. The window INa [INa(W)] has been reported to be responsible for background Na+ conductance and varying firing patterns of action potentials [37,38,39,40,41]. The Hys(V) of persistent INa [INa(P)] induced by a triangular ramp voltage (Vramp) contributes to the electrical behavior [42, 43]. However, the effects of SAF on the magnitude, gating, and Hys(V) behavior of INa remain to be clarified.

In light of the aforementioned observations, in the present study, we investigated the effects of SAF on the magnitude, gating, frequency dependence, and Hys(V) behavior of INa—including transient INa (INa(T)), late INa (INa(L)), INa(W), and INa(P)—in electrically excitable cells.

Methods

Chemicals, drugs, reagents, and solutions

SAF [PNU-151774E, Xadago, Equfina, and Fce-26743; (S)-2-((4-((3-fluorobenzyl)oxy)benzyl)amino)propenamide, (2S)-2-[[4-[(3-fluorophenyl)methoxy]phenyl]methylamino]propanamide;methanesulfonic acid, C17H19FN2O2, https://pubchem.ncbi.nlm.nih.gov/compound/Safinamide], dopamine, serotonin, tetraethylammonium chloride (TEA), and tetrodotoxin (TTX) were purchased from Sigma-Aldrich (Genechain, Kaohsiung, Taiwan). Metofluthrin was obtained from Chung Tai Sing Chemical Industry (Hsinchu, Taiwan), and mirogabalin (MGB) was obtained from Cayman Chemical (Genechain, Kaohsiung, Taiwan). Ham’s F-12 Nutrient Mix, horse serum, fetal calf serum, L-glutamine, and trypsin/ethylenediaminetetraacetic acid (EDTA) were purchased from HyClone (Thermo Fisher, Tainan, Taiwan). All chemicals and reagents were of analytical grade.

The composition of the external or bath solution [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-buffered normal Tyrode’s solution] was as follows: 136.5 mM NaCl, 1.8 mM CaCl2, 5.4 mM KCl, 0.53 mM MgCl2, 5.5 mM glucose, and 5.5 mM HEPES–NaOH (pH 7.4). To measure K+ currents (data not shown), we filled a patch pipette with an internal solution comprising 140 mM KCl, 1 mM MgCl2, 3 mM adenosine 5′-triphosphate disodium salt, 0.1 mM guanosine 5′-triphosphate disodium salt, 0.1 mM ethylene glycol tetraacetic acid, and 5 mM HEPES–KOH buffer (pH 7.2). To record the whole-cell INa, the K+ in the internal solution was substituted with Cs+, and the pH of the solution was adjusted to 7.2 by adding cesium hydroxide. In the single-channel experiments performed to record NaV currents, the pipette was filled with a Na+-rich solution containing 136 mM NaCl, 0.53 mM MgCl2, 5.5 mM glucose, and 5.5 mM HEPES–NaOH (pH 7.2). The bath medium was a K+-rich solution comprising 130 mM KCl, 10 mM NaCl, 3 mM MgCl2, 6 mM glucose, and 10 mM HEPES-KOH (pH 7.4). The solutions and culture media were generally filtered on the day of use by using sterile Acrodisc Syringe Filters containing a 0.2-μm Supor Membrane (Bio-Check; New Taipei City, Taiwan).

Cell preparations

GH3 pituitary tumor cells, acquired from the Bioresources Collection and Research Center (number: 60,015; Hsinchu, Taiwan), were maintained in Ham’s F-12 media containing 15% (v/v) horse serum, 2.5% (v/v) fetal calf serum, and 2 mM L-glutamine. Cells were grown in a monolayer culture at 37 °C in a humidified environment of carbon dioxide/air (1:19) for 5 or 6 days to a confluence of 60–80%. Trypsinization [0.025% trypsin solution (HyClone) containing 0.01 sodium N,N-diethyldithiocarbamate and EDTA] was performed for subculturing. The culture medium was changed every 2 or 3 days; cells were dispersed and passaged every 7–14 days. Experiments were performed after the cells had grown to a confluence of 60–80% (usually 5 or 6 days). The GH3 cell line has been a reliable model for studying the molecular biology, pharmacology, and biophysics of electrically excitable cells, including pituitary endocrine cells.

Electrophysiological measurements

Shortly before experiments, GH3 was carefully suspended in normal Tyrode’s solution at room temperature (20–25 °C). A few drops of the suspension containing cell clumps were immediately added to a custom-built chamber on the stage of an inverted Diaphot-200 microscope (Nikon, Tokyo, Japan). Pipettes were pulled from Kimax-51 soft-glass capillaries (#34500-99; Kimble, Vineland, NJ) by using a Narishige PP-830 Vertical Puller (Tokyo, Japan), and their tips were fire-polished using a microforge (MF-83, Narishige). During the measurements, an electrode with a tip resistance of 2–4 MΩ, which was tightly inserted into a holder, was maneuvered using a WR-98 micromanipulator (Narishige). Patch-clamp experiments were performed in the voltage-clamp mode with either cell-attached or whole-cell configuration (rupturing of the membrane patch after GΩ formation) by using a RK-400 Patch-Clamp Amplifier (Bio-Logic, Claix, France) connected to a laptop [36, 44]. Shortly before GΩ formation, potential correction was performed for a liquid junction potential, which developed at the electrode’s tip because of the difference in the compositions of the internal and bath solutions.

Data collection and recordings

Amplified signals were monitored using an HM-507 oscilloscope (Hameg, East Meadow, NY); the signals were recorded, digitized, and stored online at ≥10 kHz on a laptop (Sony VAIO CS series; Kaohsiung, Taiwan) connected to an Axon Digidata 1440A Digitizer (Molecular Devices) for efficient analog-to-digital and digital-to-analog conversion. Series resistance, always in the range of 6–18 MΩ, was electronically compensated to 80–95%. Voltage-activated currents recorded during whole-cell experiments were stored without leakage correction. The digitizer was operated using pCLAMP (version 10.6; Molecular Devices) on Windows 10 (Microsoft Corporation, Redmond, WA, USA). To ensure digitalization, some recordings were digitally acquired using the PowerLab 2/26 system (AD Instruments; Kuoyang, New Taipei City, Taiwan). During the measurement, the solutions were exchanged through a homemade gravity-driven type of bath perfusion.

Data analyses

To evaluate the concentration-dependent stimulatory effects of SAF on INa(T) and INa(L), INa was induced using a 30-ms depolarizing pulse (−100 to −10 mV). The amplitude of the current in SAF-treated and untreated cells was measured at the beginning [INa(T)] and end [INa(L)] of the voltage pulse. The duration of the voltage-clamp protocol is 30 msec and the INa displaying rapid activation and inactivation can be measured at the beginning and end of depolarizing pulse from −100 to −10 mV. The INa(T) of cells treated with 300 μM SAF was defined as 100% and compared with the current values obtained for different SAF concentrations. The concentration at which SAF increased 50% of the current [INa(T) or INa(L)] amplitude (EC50) was determined using a three-parameter logistic model (modified version of the sigmoidal Hill equation) with goodness-of-fit evaluation:

$Percentage\,increase\,\left( \% \right) = {\raise0.7ex\hbox{${\left\{ {{E_{max}} \times {{\left[ {SAF} \right]}^{{n_H}}}} \right\}}$} \!\mathord{\left/ {\vphantom {{\left\{ {{E_{max}} \times {{\left[ {SAF} \right]}^{{n_H}}}} \right\}} {\left\{ {EC_{50}^{{n_H}} + {{\left[ {SAF} \right]}^{{n_H}}}} \right\}}}}\right.\kern-\nulldelimiterspace}\!\lower0.7ex\hbox{${\left\{ {EC_{50}^{{n_H}} + {{\left[ {SAF} \right]}^{{n_H}}}} \right\}}$}}$

where EC50 is the SAF concentration ([SAF]) required for a 50% increase, nH is the Hill slope, and Emax is the SAF-mediated maximal stimulation of INa(T) or INa(L).

The sigmoidal relationship between Vramp-induced INa(W) and the upsloping Vramp (nonlinear current–voltage relationship) was investigated and fitted with the Boltzmann function as follows:

$${\raise0.7ex\hbox{$I$} \!\mathord{\left/ {\vphantom {I {{I_{max}}}}}\right.\kern-\nulldelimiterspace}\!\lower0.7ex\hbox{${{I_{max}}}$}} = {\raise0.7ex\hbox{$G$} \!\mathord{\left/ {\vphantom {G {\left\{ {1 + exp\left[ { - \frac{{\left( {V - {V_h}} \right)qF}}{{RT}}} \right]} \right\}}}}\right.\kern-\nulldelimiterspace}\!\lower0.7ex\hbox{${\left\{ {1 + exp\left[ { - \frac{{\left( {V - {V_h}} \right)qF}}{{RT}}} \right]} \right\}}$}} \times \left( {V - {E_{Rev}}} \right)$$

where V is the membrane potential in millivolts, Erev is the reversal potential of INa, G is the INa conductance in nanosiemens, I is the current, Vh is the voltage at which half-maximal activation or inactivation of the current occurs, q is the apparent gating charge, F is Faraday’s constant, R is the universal gas constant, and T is the absolute temperature.

The free energy ∆G0 for the gating of INa(W) was determined by assuming a two-state gating model [equilibrium between closed (resting) and open states] of the NaV channel. The ∆G0 for the activation of INa(W) at 0 mV could be calculated as follows: q × F × V1/2 [45, 46]. The standard errors in ∆G0 (\({\sigma _{q{V_{1/2}}}}\)) could be calculated as follows:

\({\sigma _{qF{V_{\frac{1}{2}}}}} = F \times \sqrt {V_{\frac{1}{2}}^2\sigma _q^2 + {q^2}\sigma _{{V_{\frac{1}{2}}}}^2}\)

where σq and σV1/2 represent the standard error in q and V1/2, respectively.

Recordings and analyses of single-channel NaV currents

Single-channel NaV currents induced by depolarizing pulses ranging from −100 to −10 mV were measured and subsequently analyzed using pCLAMP 10.7. The opening events of the channels were generally evaluated through multi-Gaussian adjustments of the distribution of amplitude across channels. Functional independence between channels was determined by comparing the observed stationary probabilities with the values calculated based on the binomial law. For dwell-time analyses, only a single channel was used in the patch-clamp experiment.

Curve-fitting approximations and statistical analyses

Linear or nonlinear curve fitting to different data sets was implemented using the least-squares minimization method through various maneuvers, including the Excel-embedded Solver (Microsoft Corporation) and 64-bit OriginPro (OriginLab; Scientific Formosa, Kaohsiung, Taiwan). The averaged results (whole-cell or single-channel data) are presented in terms of the mean ± standard error of the mean; the number of independent samples (n) indicates the number of cells used for experimental data collection. Between-group differences were analyzed using paired or unpaired Student’s t-test. The differences between more than two groups were evaluated through multiple comparisons performed using analysis of variance (ANOVA)-1 or ANOVA-2 with or without repeated-measures analysis, which was followed by a post-hoc Fisher’s least-significant difference test. Statistical significance was set at P < 0.05 (indicated using *, **, or + in the figures).

Results

Effects of SAF on INa magnitude

We investigated the effects of SAF on the magnitude of INa induced by rapid membrane depolarization. The cells were bathed in Ca2+-free Tyrode’s solution containing 10 mM TEA and 0.5 mM CdCl2. TEA and CdCl2 were used to block K+ and Ca2+ currents, respectively. The recording pipettes were filled with a solution containing Cs+. As shown in Fig. 1, the tested cell was maintained at –80 mV. Subsequently, a hyperpolarizing step was applied, bringing the voltage down to –100 mV for a duration of 30 ms. This was followed by a brief depolarization step to –10 mV for 30 ms was applied to evoke INa. The voltage was then returned to –50 mV for 30 ms to observe the tail current, and finally, the voltage was returned to the holding potential. Depolarizing the voltage from –100 to –10 mV over 30 ms (from a holding potential of –100 mV) robustly induced an inward current with the properties of being rapidly activated and inactivated. The rapid inward current induced by the short depolarizing pulse was identified as INa [36, 47,48,49] because it could be blocked by TTX (1 μM) and stimulated by either tefluthrin (Tef; 10 μM) or metofluthrin (10 μM). TTX is a potent inhibitor of INa; Tef and metofluthrin effectively stimulate INa [39, 50, 51]. The results are summarized in Fig. 2.

Fig. 1
figure 1

Effect of safinamide (SAF) on the voltage-gated Na+ current (INa) in pituitary tumor (GH3) cells. To record macroscopic currents, calcium ion–free Tyrode’s solution containing 10 mM tetraethylammonium chloride and 0.5 mM cadmium chloride was added to the cells; the recording electrode was filled with solution containing cesium ions. (A) Current traces during the control period (a, black; untreated cells) and during exposure to 10 (b, blue) and 30 (c, red) μM SAF. The voltage-clamp protocol used is indicated atop the current traces. In panel A, the third graph from the top is an expanded version the second graph (purple dashed box). (B) Time course showing effect of 10 and 30 μM SAF on the amplitude of peak INa. Each Current amplitude (indicated with black circles) was measured at the beginning of depolarizing pulse at a rate of 2 Hz. Horizontal bar shown above indicates the SAF application. (C) Concentration–response curves corresponding to SAF-mediated stimulation of transient INa [(INa(T)): blue filled squares] and late INa [INa(L); sustained: red open circles] in GH3 cells (mean ± standard error of the mean; n = 8 for each point). The current amplitude was measured at the beginning and end of a 30-ms depolarizing pulse (−100 to −10 mV). The gray smooth line indicates the goodness of fit of our model to the modified Hill equation. The EC50 values corresponding to the SAF-induced stimulation of INa(T) and INa(L) were 27.1 and 4.8 μM, respectively (least-squares minimization)

Full size image
Fig. 2
figure 2

Graph showing effects of tetrodotoxin (TTX), tefluthrin (Tef) and metofluthrin on the peak amplitude of INa in GH3 cells. Current amplitude was measured at the beginning of each depolarizing pulse from −100 to −10 mV for a duration of 30 ms. Each point represents the mean ± standard error of the mean (n = 7). The statistical analyses were done by ANOVA-1, P < 0.05, followed by post-hoc Fisher’s least-significant different test, P < 0.05. *Significantly different from control (P < 0.05)

Full size image

SAF exposure resulted in gradual increases in the magnitudes of INa(T) and INa(L) induced by abrupt membrane depolarization pulses (Fig. 1A). For example, 1 min after the addition of 10 and 30 μM SAF, the amplitude of INa(T) had increased from a control value of 512 ± 23 pA (n = 8) to 1387 ± 131 pA (n = 8, P < 0.05) and 1991 ± 153 pA (n = 8, P < 0.05), respectively; the corresponding values for INa(L) had increased from 18 ± 3 pA (n = 8) to 54 ± 9 pA (n = 8, P < 0.05) and 106 ± 17 pA (n = 8, P < 0.05), respectively. The amplitude of INa(L) was measured at the end of the depolarizing pulse from −100 to −10 mV for a duration of 30 ms. After the removal of SAF, the amplitudes of INa(T) and INa(L) returned to 523 ± 24 and 20 ± 3 pA, respectively (n = 8). SAF (30 μM) also increased the time constant τinact(S) corresponding to the slow component of INa(T) inactivation, with no evident change in that corresponding to the fast component. As summarized in Fig. 3, SAF (30 μM) markedly increased the τinact(S) of INa(T) inactivation from 5.1 ± 0.7 to 10.2 ± 1.1 ms (n = 8, P < 0.05). Fig. 1B illustrates the time course of stimulatory effect of SAF (10 and 30 μM) on INa. The presence of TTX (1 μM) alone decreased INa(T) and INa(L) to 18 ± 2 pA (n = 8, P < 0.05) and 2 ± 1 pA (n = 8, P < 0.05), respectively, from a control value of 511 ± 17 pA and 32 ± 5 pA (n = 8).

Fig. 3
figure 3

Graph showing effects of SAF on the slow component (τinact(S)) in inactivation time constant of INa in GH3 cells. Each point represents the mean ± standard error of the mean (n = 8). The statistical analyses were done by ANOVA-1, P < 0.05, followed by post-hoc Fisher’s least-significant different test, P < 0.05. *Significantly different from control (P < 0.05) and **significantly different from SAF (30 μM) alone group (P < 0.05)

Full size image

SAF increased the amplitudes of INa(T) and INa(L) in a concentration-dependent manner (Fig. 1C). Using the Hill equation described in Materials and Methods, the EC50 values corresponding to SAF-mediated stimulation of INa(T) and INa(L) were estimated to be 27.1 ± 2.1 and 4.8 ± 0.7 μM, respectively. INa(T) and INa(L) induced by rapid depolarization pulses differentially increased in a concentration-dependent manner in GH3 cells.

Effects of SAF on the steady-state I–V relationship and inactivation curve of INa(T)

To further characterize the stimulatory effects of SAF on INa(T), we investigated whether this drug perturbs the steady-state I–V relationship of INa(T) in GH3 cells. Fig. 4a illustrates INa(T) traces induced by different voltage steps in the presence and absence of SAF. Fig. 4B depicts the mean I–V relationship of INa(T) (i.e., V-shaped) in the absence and presence of 3 or 10 μM SAF. Fig. 4C also illustrates mean conductance versus voltage (G–V) relationship of INa(T) obtained in the control period and with the addition of 3 or 10 μM SAF. The value required for half-maximal activation voltage was found to be shifted to more negative potentials in the presence of SAF. Additionally, the steady-state inactivation curve of INa was further characterized (Fig. 4D). In these experiments, a two-step voltage-clamp protocol was applied (indicated in the legend of Fig. 4D). The results showed that cell exposure to 10 mM SAF not only increased the maximal conductance of INa, but also shifted the inactivation curve to the rightward direction by approximately 14 mV with no change in the sloping factor of the curve.

Fig. 4
figure 4

Mean current–voltage (I-V) or conductance-voltage relationship of INa(T) in GH3 cells. The measurements were performed as described in the legend of Fig. 1. Because we used the whole-cell mode, the test cells were maintained at −80 mV and subjected to a series of command voltages ranging from −100 to +30 mV in 10-mV increments. (A) Current traces during the control period (upper) and during the exposure to 3 μM SAF. The voltage-clamp protocol used is indicated atop the current traces. (B) Mean I–V relationship of INa(T) in the absence (blue filled squares) and with cell exposure to 3 μM SAF (red open circles) or 10 μM SAF (brown open triangles) (mean ± standard error of the mean; n = 8 for each point). INa(T) amplitude was measured at the beginning of each voltage pulse. Notably, the I–V relationship of INa(T) (or peak INa) induced by 30-ms voltage pulses was shifted to more negative potentials upon SAF (3 or 10 μM) exposure. (C) Mean conductance versus voltage (G-V) of INa(T) in the absence (blue filled squares) and with cell exposure to 3 μM SAF (red open circles) or 10 μM (brown open triangles) (mean ± standard error of the mean; n = 8 for each point). The conductance-voltage relationship of INa(T) was shifted to more negative potentials during exposure to SAF. (D) Effect of SAF (10 μM) on the steady-state inactivation curve of INa(T) (mean ± standard error of the mean; n = 7 for each point). In these experiments, the conditioning voltage pulse with a duration of 30 ms to various membrane potentials between −100 and +10 mV was applied from a holding potential of −80 mV. Following each conditioning potential, a test pulse to −10 mV for a duration of 30 ms was given to activate INa. The normalized amplitude of INa (I/Imax) was constructed against the conditioning potential

Full size image

Effects of dopamine, serotonin, SAF, SAF plus dopamine, and SAF plus serotonin on INa(T) amplitude

Studies have demonstrated the existence of MAO activity in pituitary cells [52, 53]. In the inhibition of MAO-B, the stimulatory effects of SAF on INa may result primarily from an increase in the extracellular concentration of dopamine or serotonin. We investigated whether dopamine or serotonin affects INa(T) in these cells and whether the addition of dopamine and serotonin during SAF exposure reverses the SAF-mediated increase in INa. As shown in Fig. 5, the addition of neither dopamine nor serotonin altered the magnitude of INa(T); similarly, during SAF exposure, the addition of neither dopamine nor serotonin reversed the SAF-mediated increase in INa(T). Thus, under the experimental conditions employed in the present study, the SAF-mediated stimulation of INa(T) in pituitary cells may not have involved the inhibition of MAO-B activity.

Fig. 5
figure 5

Graph illustrating the effects of dopamine, serotonin, tetrodotoxin, SAF, SAF plus dopamine, and SAF plus serotonin on the amplitude of INa(T) in GH3 cells. Each current amplitude was measured at the beginning of a short depolarizing pulse (−100 to −10 mV). In the experiments of SAF plus dopamine and SAF plus serotonin, dopamine (10 μM) or serotonin (10 μM) was added when the cells were exposed to SAF (30 μM). Each point represents the mean ± standard error of the mean (n = 7). The statistical analyses were done by ANOVA-1, P < 0.05, followed by post-hoc Fisher’s least-significant different test, P < 0.05. *Significantly different from control (P < 0.05)

Full size image

SAF-induced increase in the cumulative inhibition of INa(T) during a train of depolarizing stimuli

The inactivation of INa(T) has been demonstrated to accumulate before being elicited during repetitive short pulses [33, 34, 54]. SAF is efficacious as an add-on therapy following subthalamic nucleus deep brain stimulation in patients with Parkinson’s disease [19, 20]. Therefore, we investigated whether SAF could modify the inactivation of currents induced by a train of depolarizing stimuli. The test cells were maintained at −80 mV and subjected to repetitive depolarization to −10 mV (40 ms per pulse; rate of 20 Hz; duration of 1 s). Similar to the findings of relevant studies [34, 36], during the control period (the absence of SAF), INa(T) inactivation was noticed to be induced by 1 s of repetitive depolarization stimuli (−80 to −10 mV) with a decaying time constant of 65 ± 4 ms (n = 7; Fig. 6A, 6B and 6C). This indicated that the single-exponent process resulted in a sudden decay in the current. In the presence of 3 and 10 μM SAF, the exponential time course of INa(T) induced by the same train of depolarizing pulses was longer at 107 ± 5 ms (n = 7, P < 0.05) and 124 ± 6 ms (n = 7, P < 0.05), respectively. Furthermore, as MGB and ranolazine have been reported to suppress the amplitude of INa effectively [36, 47, 55, 56], we added ranolazine (Ran; 10 μM) and mirogabalin (MGB; 10 μM) separately in the presence of 10 μM SAF. We found they effectively attenuated the SAF-induced increase in the decaying time constant of INa(T) induced by a rapid train of pulses (Fig. 6B). The application of Ran (10 μM) or MGB (10 μM) alone decreased the decaying time constant of INa(T) during the same train of depolarizing pulses to to 39 ± 4 ms (n = 7, P < 0.05) or 42 ± 4 ms (n = 7, P < 0.05), respectively, from a control value of 66 ± 5 ms (n = 7). Thus, in addition to increasing the magnitude of INa(T), SAF prominently affects the decaying of INa(T) subjected to a 1-s train of depolarizing pulses.

Fig. 6
figure 6

Effect of SAF on INa(T) decay induced by a train of depolarizing pulses in GH3 cells. The train of pulses comprised twenty 40-ms pulses (voltage increased to −10 mV) with 10-ms intervals at −80 mV for a total duration of 1 s. (A) Current traces during the control period (a, blue) and during exposure to 10 μM SAF (b, red). The voltage-clamp protocol used is indicated atop the current traces. In panel A, the third graphs (blue and red) from the top are the expanded forms of the second graphs (brown dashed boxes). (B) The relative amplitude of INa(T) versus pulse train duration in the absence (blue open circles) and presence (red open squares) of 10 μM SAF (mean ± standard error of the mean; n = 7 for each point). The INa(T) amplitudes were normalized by dividing the current amplitudes at the end of each pulse-train stimulation by those obtained at the beginning of the pulse train stimulation. The gray continuous lines on which the data points are overlaid are reliably fitted with a single exponential. (C) Summary graph depicting the effects of SAF (3 and 10 μM), SAF plus ranolazine (Ran), and SAF plus MGB on the decaying time constant of the current induced by a train of depolarizing command voltages ranging from –80 to –10 mV (mean ± standard error of the mean; n = 7 for each point). *Significantly different from the control (P < 0.05), **significantly different from the SAF (3 μM) alone group (P < 0.05), and +significantly different from the SAF (10 μM) alone group (P < 0.05)

Full size image

Stimulatory effects of SAF on INa(W)

The induction of instantaneous INa(W) by ascending (or upsloping) Vramp has been demonstrated in various excitable cells [35, 36, 38, 40, 41, 48]. In the present study, we investigated whether the addition of SAF to GH3 cells modulates the magnitude of INa(W) induced by ascending Vramp. Test cells were maintained at −80 mV and subjected to Vramp ascending from −100 to +40 mV over 200 ms (i.e., ramp speed of 0.7 mV/ms) to induce INa(W) [40, 48]. The amplitude and strength (∆area) of INa(W) induced by the ascending Vramp sharply increased within 1 min of SAF exposure (Fig. 7A and 7B). The ∆area values of INa(W) in the absence and presence of SAF and SAF plus Ran were calculated (Fig. 7B).

Fig. 7
figure 7

Stimulatory effects of SAF on window INa [INa(W)] induced by an ascending ramp voltage Vramp in GH3 cells. For the experiments, the test cells were maintained at −80 mV and subjected to Vramp ranging from −100 to +40 mV over 200 ms (ramp speed of 0.7 mV/ms). (A) INa(W) trace in the absence (a, blue) and presence (b, red) of 10 μM SAF. The inset indicates the Vramp protocol, and the downward deflection indicates inward-directed current. The continuous gray lines corresponding to the untreated or SAF-treated (10 μM) cells were fitted (least-squares minimization) with the Boltzmann equation. The values of V1/1 and q (apparent gating charge) induced by ascending Vramp over 200 ms in the absence (blue) of SAF were, respectively, −34 mV and 2.1 e; the corresponding values in the presence (red) of 10 μM SAF were −38 mV and 2.8 e, respectively. (B) Summary graph illustrating the effects of SAF (3 and 10 μM) and SAF (10 μM) plus Ran (10 μM) on the Δarea of INa(W) in GH3 cells (mean ± standard error of the mean; n = 7 for each point). The value of Δarea was measured at a voltage ranging from −80 and +40 mV, which corresponded to INa(W) induced by the ascending Vramp. The statistical analyses were done by ANOVA-1, P < 0.05, followed by post-hoc Fisher’s least-significant different test, P < 0.05. *Significantly different from the control (P < 0.05), **significantly different from the SAF (3 μM) alone group (P < 0.05), and +significantly different from the SAF (10 μM) alone group (P < 0.05). (C) Summary graph illustrating the effects of SAF (3 and 10 μM) and SAF plus Ran (10 μM) on ΔG0 (mean ± standard error of the mean; n = 7 for each point). The estimation of ΔG0 for the induction of instantaneous INa(W) is described in the Materials and Methods section. Notably, increasing the concentration of SAF increased ΔG0 in GH3 cells; further addition of Ran effectively reversed the SAF-induced increase in ΔG0. The statistical analyses were done by ANOVA-1, P < 0.05, followed by post-hoc Fisher’s least-significant different test, P < 0.05.*Significantly different from the control (P < 0.05), **significantly different from the SAF (3 μM) alone group (P < 0.05), and +significantly different from the SAF (10 μM) alone group (P < 0.05)

Full size image

Effect of SAF on the activation energy required for the induction of INa(W) by Vramp

Experimental data points corresponding to INa(W) were optimally fitted with the Boltzmann isotherm to estimate the values of q and V1/2 for the instantaneous INa(W) induced by Vramp. Using these values, the ∆G0 for the gating of INa(W) activation at 0 mV in the absence and presence of SAF was calculated (∆G0 = q × F × V1/2; Fig. 7C).As the SAF concentration was increased, ∆G0 for the induction of INa(W) by a 200-ms-long Vramp also increased; the subsequent addition of Ran effectively attenuated the SAF-mediated increase in ∆G0. In GH3 cells exposed to 3 and 10 μM SAF, the ∆Go values increased from a control value of 6.78 ± 1.1 to 8.81 ± 1.6 kJ/mol (n = 7, P < 0.05) and 10.36 ± 1.8 kJ/mol (n = 7, P < 0.05), respectively.

Attenuation of the SAF-induced increase in the amplitude and Hys(V) of INa(P) by MGB and Ran

We further investigated whether SAF exposure modulated the magnitude and Hys(V) behavior of INa(P) induced by an isosceles-triangular Vramp in GH3 cells. To record whole-cell currents, the test cells were maintained at −80 mV and subjected to an upright isosceles-triangular Vramp ascending from –110 and +50 mV over 3.2 s (digital-to-analog conversion; Fig. 8A). Consistent with the findings of relevant studies [43, 49, 57], we found that SAF exposure markedly increased the high and low amplitudes of INa(P) induced by the upsloping (ascending) and downsloping (descending) ends of the upright triangular Vramp, respectively; consequently, we observed a figure-of-eight (∞-shaped) configuration of the instantaneous I–V relationship for INa(P) and found that the configuration was enhanced by SAF. For example, when test cells were subjected to an isosceles-triangular Vramp over 3.2 s (ramp speed of 0.1 mV/ms), the INa(P) amplitudes measured at −10 mV (high threshold) and −80 mV (low threshold) during the control period were 175 ± 14 pA (n = 7) and 288 ± 25 pA (n = 7), respectively. After the addition of 3 and 10 μM SAF, the INa(P) amplitude at −10 mV [high-threshold Hys(V) loop] was 194 ± 17 pA (n = 7, P < 0.05) and 219 ± 18 pA (n = 7, P < 0.05), respectively; the corresponding amplitudes at −80 mV [low-threshold Hys(V) loop] were 348 ± 29 pA (n = 7, P < 0.05) and 389 ± 31 pA (n = 7, P < 0.05), respectively. Adding MGB and Ran separately during SAF exposure reversed the SAF-mediated increase in the high- and low-threshold INa(P) induced by the triangular Vramp (Fig. 8B). These findings indicate the unique Hys(V) behavior of INa(P) induced by an isosceles-triangular Vramp in GH3 cells; SAF exposure may increase the strength of Hys(V).

Fig. 8
figure 8

Stimulatory effects of SAF on the voltage-dependent hysteresis [Hys(V)] behavior of persistent INa [INa(P)] induced by upright isosceles-triangular Vramp. Vramp was supplied for 3.2 s (ramp speed of 0.1 mV/ms; digital-to-analog conversion) to induce Hys(V) behavior in GH3 cells. (A) Current traces during the control period (left) and during the exposure to 10 μM SAF (right). The blue and red traces shown in each panel represent currents induced by the upsloping (ascending) and downsloping (descending) limbs of the upright isosceles-triangular Vramp, respectively. The inset indicates the voltage protocol. The dashed arrows indicate the direction of the current trajectory over time. (B) Summary graphs illustrating the effects of SAF (3 and 10 μM), SAF plus MGB, and SAF plus Ran on the amplitude of Vramp-induced INa(P) measured at −10 mV (ascending limb; left side) and −80 mV (descending limb; right side). Each point represents the mean ± standard error of the mean (n = 7). *Significantly different from the control (P < 0.05), **significantly different from the SAF (3 μM) alone group (P < 0.05), and +significantly different from the SAF (10 μM) alone group (P < 0.05)

Full size image

Effect of SAF on single-channel NaV currents

To elucidate the mechanisms underlying the effects of SAF on the magnitude of INa, we investigated the actions of SAF and SAF plus MGB on single-channel NaV currents. This experiment was performed using the cell-attached configuration of the voltage-clamp test. Test cells were placed in K+-rich solution, and the recording pipette was filled with Na+-rich solution. SAF (10 μM) increased channel activity and decelerated current inactivation when the cells were exposed to depolarization stimuli ascending from –100 to –10 mV (rate of 0.1 Hz; Fig. 9A). Furthermore, the addition of MGB (10 μM) in the presence of 10 μM SAF reduced the probability of channel opening. The addition of 10 μM SAF markedly increased channel activity from 0.027 ± 0.007 to 0.091 ± 0.006 (n = 7, P < 0.05); the addition of 10 μM MGB in the presence of SAF reduced the open-state probability of the channel to 0.041 ± 0.007 (n = 7, P < 0.05). Moreover, with the presence of 10 μM SAF, the mean open time of NaV channels was prolonged to 6.5 ± 0.7 (n = 7, P < 0.05) msec from a control value of 2.3 ± 0.3 msec (n = 7). However, no considerable modifications were noted in the amplitude of the single-channel current in the presence of SAF or SAF plus MGB (control, 2.01 ± 0.35 pA; SAF, 2.03 ± 0.37 pA; SAF plus MGB, 2.00 ± 0.44 pA; n = 7; P > 0.05). Consistent with these findings, the mean open time of the NaV channel in the presence of 10 μM SAF (5.9 ± 1.1 ms; n = 7; P < 0.05) was longer than that in the control period (2.3 ± 0.3 ms, n = 7); the subsequent further addition of 10 μM MGB decreased the mean open time to 3.7 ± 0.7 ms (n = 7, P < 0.05) (Fig. 9B). Although SAF did not change the amplitude of single-channel currents, it enhanced channel activity and decelerated inactivation of NaV-channel opening in GH3 cells. MGB added during SAF exposure reversed the SAF-induced increase in NaV channel activity.

Fig. 9
figure 9

Effects of SAF on single-channel NaV currents in GH3 cells. In (A), single-channel currents were induced through successive depolarizations (rate of 0.1 Hz) from a holding potential of −100 to −10 mV. Black graphs on the left indicate current traces during the control period (in the absence of SAF or SAF plus MGB); the red and blue graphs indicate the exposure to 10 μM SAF and 10 μM SAF plus 10 μM MGB, respectively. In the SAF plus MGB experiment, MGB was added 2 min after the addition of SAF. Channel opening in each record is shown as a downward deflection; the lowest traces on each side represent the average of 50 sweeps. Notably, SAF effectively increased the open-state probability of NaV channels, but the subsequent addition of MGB attenuated the SAF-induced increase in the likelihood of channel opening. However, the presence of neither SAF nor MGB in addition to SAF changed the single-channel amplitude of NaV channels in GH3 cells. (B) Graph showing effects of SAF, SAF plus MGB on the mean open time of NaV channel (mean ± standard error of the mean; n = 7). *Significantly different from control (P < 0.05) and **significantly different from SAF (10 μM) alone group (P < 0.05)

Full size image

Docking prediction of SAF on human MAO_B and NaV channel

Using PyRx, we further explored the molecular docking between human MAO-B (structure: https://www.rcsb.org/structure/1GOS) and SAF. Figure 10 illustrates the predicted binding sites of SAF. SAF engages in hydrophobic interactions with certain amino acid residues, such as Phe 103, Val 106, Arg 120, Asp 123, Arg 127, Thr 479, and Glu 483. In addition to interacting with intramolecular hydrogen bonds [58], SAF forms three hydrogen bonds with the NaV-channel residues Pro 104, His 115, and Trp 119, with the bond lengths being 2.81, 3.16, and 2.90 Å, respectively. The binding affinity for the interaction between SAF and MAO-B is –7.8 kcal/mol, and the upper and lower root-mean-square deviations (RMSD) in atomic positions were 49.76 and 60.68, respectively. In line with the findings of relevant studies [21, 23, 59], we observed that the interaction between MAO-B and SAF resulted in a substantial decrease in MAO-B activity.

Fig. 10
figure 10

Predicted docking interactions between SAF and monoamine oxidase B (MAO-B). The protein structure was obtained from the Protein Data Bank (ID: 1GOS); the chemical structure of SAF was obtained from PubChem [compound CID: 131682 (3D conformer)]. MAO-B was docked with SAF (yellow dashed box on the left) through PyRx, and the corresponding interaction diagram was generated using LigPlot+. The red arcs with spokes radiating toward the ligand (SAF) indicate hydrophobic interactions between SAF and MAO-B. The green dotted line indicates the hydrogen bond between SAF and Pro 104, His 115, or Trp 119, with the corresponding bond lengths being 2.81, 3.16, or 2.90 Å

Full size image

We further explored the molecular docking between NaV channels and SAF. Figure 11 and Supplementary Fig. 1 depicts the predicted binding sites of SAF. After docking, SAF forms a hydrogen bond with Lys 63, with the bond length being 2.97 Å. SAF further engages in hydrophobic interactions with several residues, including Ile 9, Gln 15, Tyr 67, Asn 78, Ser 112, and Val 113. The binding affinity for the interaction between SAF and a NaV channel was found to be −6.8 kcal/mol, and the upper and lower RMSD values were 23.00 and 25.59, respectively. The affinity energy was close to the estimated ∆G0 for the induction of INa(W) by Vramp in the presence of SAF. Thus, SAF can dock with both MAO-B and NaV channels, thereby presumably reducing structural constraints and increasing channel activity. Collectively, the dual effects of SAF on MAO-B and NaV channel activities [23] may considerably affect the functional activities and thus may be beneficial in the treatment of various neurological disorders [11, 15, 25, 29]. However, since a prokaryotic NaV channel (i.e., NaVM) was used in this prediction, whether SAF can modulate the function of NaVM as observed in GH3-cells’ NaV channel needs to be further examined.

Fig. 11
figure 11

Predicted docking interactions between NaV channels and SAF. The protein structure of a NaV channel was obtained from the Protein Data Bank (ID: 6Z8C), and the chemical structure of SAF was obtained from PubChem [compound CID: 131682 (3D conformer)]. A NaV channel was docked with SAF (yellow dashed box on the left) using PyRx; the corresponding interaction diagram was generated using LigPlot+. In the image on the right, the red arcs with spokes radiating toward the ligand (SAF) indicate hydrophobic interactions between SAF and several amino acid residues, whereas the green dashed line indicates the hydrogen bond between SAF and Lys 63, the length of which was 2.97 Å. The docking regions appear to be adjacent to the transmembrane region (position: residues 82–102) and the membrane segment (position: residues 46–67). The interactions probably alter structural constraints, thereby increasing the open-state probability of NaV channels

Full size image

Discussion

Our key findings are as follows. SAF stimulated INa in a concentration-, time-, and frequency-dependent manner. It differentially stimulated INa(T) and INa(L) induced by short depolarizing pulses. SAF increased the time constant of the decay of INa(T) induced by a train of depolarizing pulses but increased the strength and ∆G0 of Vramp-induced INa(W). The Hys(V) strength of INa(P) (in both low- and high-threshold loops) was greater when the cells were exposed to an upright isosceles-triangular Vramp. Cell-attached single-channel current recordings revealed a SAF-induced increase in the open-state probability of the channel without any change in the single-channel amplitude. Molecular docking between SAF and both MAO-B and NaV channels indicated the existence of similar structural motifs, which facilitate SAF binding to MAO-B and NaV channels. SAF may reach the binding site once the NaV channel protein is highly activated and is in its open state or conformation. Overall, these findings suggest that SAF-mediated modulation of the magnitude, gating, and Hys(V) behavior of INa may be independent and upstream of its inhibitory action on MAO-B activity.

As mentioned, SAF inhibits the activity of MAO-B [2, 14, 23, 24, 52, 53, 60]. Thus, in the present study, the stimulation of INa by SAF was expected to be associated with the inhibition of MAO-B by SAF and with subsequent increases in the concentrations of dopamine and serotonin. However, the exposure of cells to dopamine and serotonin did not lead to any changes in the magnitude of INa(T) induced by rapid membrane depolarization pulses. Adding dopamine and serotonin separately to the bath solution in the presence of SAF exerted no effects on the SAF-stimulated INa(T). Therefore, the stimulatory effects of SAF on INa(T) and INa(L) may be mediated by a mechanism other than that involving the inhibition of MAO-B activity.

The time-dependent decrease in INa(T) induced by a 20-Hz train of depolarizing pulses (40-ms pulses of voltage ascending from –80 to –10 mV; rate of 20 Hz; duration of 1 s) was decelerated by SAF. The results indicate that there is use dependence of INa(T) during repetitive depolarization, as demonstrated previously [34,35,36, 60]. SAF may lead to progressive gain-of-function changes by altering and decelerating the inactivation of currents. Thus, the SAF-mediated increase in INa(T) may be closely associated with use-dependent attenuation of the magnitude of INa(T) induced by a train of depolarizing pulse stimuli.

We further estimated the ∆area and ∆G0 values for the instantaneous INa(W) induced by an ascending Vramp, both of which were found to be markedly higher when SAF was present. Adding Ran in the presence of SAF reversed the SAF-mediated increase in the ∆area and ∆G0 of the current. Because the magnitude of INa(W) is primarily responsible for the background (steady state) conductance of Na+ and the electrical firing of excitable cells [37, 38, 41, 61,62,63], SAF may increase the firing frequency of action potentials by enhancing the strength of INa(W). It is important to note that the currents elicited by Vramp may also result from late/persistent INa and slow closed-state inactivation. Whether the SAF-mediated augmentation of ramp currents was caused by increasing INa(W) still needs to be investigated. Moreover, how the activation energy of Vramp-induced INa(W) can be increased in the SAF presence remains to be further studied.

We observed the nonlinear voltage-dependent Hys(V) behavior of INa(P) during the control period and the exposure of the cells to SAF, SAF plus MGB, or SAF plus Ran [36, 42]. The Hys(S) behavior was induced by exposing the cells to an upright isosceles-triangular Vramp for 3.2 s. SAF increased the peak of INa(P) induced by the ascending (upsloping) limb of the triangular Vramp, particularly at –10 mV, and the amplitude of INa(P) induced by the descending (downsloping) end of Vramp, particularly at –80 mV. Thus, we noted a figure-of-eight (∞-shaped) configuration of the Hys(V) loop of current induced by the triangular Vramp; the strength of this behavior was discovered to be considerably enhanced in the presence of SAF. Thus, Vramp induced two distinct types of INa(P): high-threshold current and low-threshold current. The high-threshold INa(P) was excessively induced [at a voltage range where peak INa(T) was induced maximally] by the upsloping limb of the triangular Vramp; by contrast, the low-threshold INa(P) was induced by the downsloping end of the triangular Vramp. Furthermore, the trajectories of currents induced by the ascending and descending limbs of the triangular Vramp followed the counterclockwise and clockwise directions, respectively. Adding MGB and Ran separately in the presence of SAF reduced the SAF-mediated increases in high- and low-threshold INa(P) induced by Vramp. It also needs to be noted that the increase of the current at –80 mV in the descending limb of the voltage protocol might be caused by the ability of SAF to activate the persistent INa. Thus, whether SAF affect the NaV-channel deactivation still warrants further investigations.

It needs to be noted that as the concentration of extracellular Ca2+ was decreased, the gating of INa might be altered [64]. However, the presence of extracellular Ca2+ was also allowed to activate voltage-gated Ca2+ current and then to interfere with the examination of voltage-gated Na+ current. Therefore, whether safinamide affects the amplitude and gating of INa when the concentration of external Ca2+ is within the physiological range remains to be further studied.

The cell-attached current recordings revealed that SAF increased the open-state probability of NaV; this increase was reversed by the subsequent addition of MGB. Furthermore, SAF extended the mean open time of single NaV channels; however, this effect of SAF was reversed by MGB. SAF did not modify the amplitude of single NaV channels; this indicates that the interaction between SAF and NaV channels is secondary to alterations, which possibly occur at a location remote from the pore region of the channels.

Pharmacokinetic studies have reported that the maximal plasma concentrations of SAF after single 2.5, 5.0, and 10.0 mg/kg doses can be approximately 1000, 2500, and 5500 ng/mL (or 3.3, 8.3 and 18.2 μM), respectively [65]. The EC50 values for the SAF-mediated stimulation of INa(T) and INa(L) were 27.1 ± 2.1 and 4.8 ± 0.7 μM, respectively; this indicates that the stimulatory effects of SAF on INa are within the clinical therapeutic range, although the SAF concentration in cerebrospinal fluid is still unclear [60]. The IC50 for the SAF-mediated inhibition of MAO-B activity has been reported to be approximately 30–40 nM [23]. Notably, the magnitude of SAF-stimulated INa depends strongly on various factors, including the pre-existing resting potential, the firing pattern of the action potential, the concentration of SAF in cells, and a combination of the aforementioned factors. Regarding the inhibitory effect of SAF on MAO-B, SAF may lyse and remove the components of the surface membrane of various host cells (Edmondson and Binda, 2018), thus accessing cytosolic or mitochondrial enzymes. Thus, the stimulatory perturbation of INa [INa(T), INa(L), INa(W), and INa(P)] is clinically achievable by using SAF and may have pharmacological, therapeutic, and toxicological relevance in humans [12, 16, 17].

SAF may increase blood pressure. NaV channels are functionally distributed across vascular smooth muscle cells [66,67,68]. The mRNA transcripts of the α subunits of NaV1.1, NaV1.2, NaV1.3, and NaV1.6, together with those of the β1 and β3 subunits, have been detected in GH3 cells [69]. The extent to which SAF-induced hypertensive events [4, 10] are associated with the stimulatory effects of SAF on INa in vascular smooth muscle cells (i.e., NaV1.7), heart cells (i.e., NaV1.5 or NaV1.6), and skeletal muscle cells (i.e., NaV1.4) is worth investigation.

In contrast to our findings, earlier studies showed SAF might suppress INa magnitude in different preparations [21, 26,27,28,29]. It will be important to determine whether the inhibitory effect of SAF on INa is associated with either the decrease of the activity of monoamine oxidase or the production of reactive oxygen species [13, 14, 22,23,24,25]. Of note, although SAF was considered a potential anticonvulsant based on prior reports of INa attenuating property, clinically, the anticonvulsive activity of SAF was not proven, as only open-label studies comparing with baseline were provided [70, 71]. The inhibitor of NaV channels could generally be considered as an anticonvulsant, however, despite the unclear underlying ionic mechanism, the stimulator of NaV channels virtually might not become a pro-epileptic drug, because of a wide range of epileptic disorders through which the initiation or epileptogenesis is largely unclear. Furthermore, the difference on the effect of INa may be the result of dissimilar channel isoforms, expression levels of isoforms, the species, the auxiliary proteins in the cell types, and the different concentrations of compounds used for each cell type. Direct comparisons of sodium channel kinetic properties were thus restricted to data within the same cell type. Nevertheless, it is likely that the stimulatory effect of SAF on INa is preferentially linked to its bindings to NaV1.5 and/or NaV1.6 isoforms of the channel. Further characterization and interpretation of the modulatory effect of SAF on INa and overall cellular excitability in different cell types or network should be implemented.