- Email: [email protected]
Inhibition of the rat brain sodium channel Nav1.2 after prolonged exposure to gabapentin
Epilepsy Research 70 (2006) 263–268
Short communication
Inhibition of the rat brain sodium channel Nav1.2 after prolonged exposure to gabapentin Yi Liu ∗ , Ning Qin, Tasha Reitz, Yan Wang, Christopher M. Flores Analgesics Drug Discovery, Johnson & Johnson Pharmaceutical Research & Development, Welsh & McKean Roads, P.O. Box 776, Spring House, PA 19477-0776, United States Received 20 September 2005; received in revised form 2 March 2006; accepted 15 March 2006 Available online 18 April 2006
Abstract Prolonged exposure of neurons to gabapentin inhibits repetitive firing of Na+ -dependent action potentials. Here, we studied the effect of such prolonged exposure to gabapentin on a rat sodium channel, Nav1.2. After 3 days of continuous incubation with gabapentin (10–1000 M), Nav1.2 current density was decreased dose-dependently relative to untreated cells. The reduction was 57% at 30 M gabapentin, while higher concentrations (100–1000 M) did not result in greater effects. Prolonged treatment with gabapentin also caused the channel to inactivate at more hyperpolarized potentials. These effects provide a mechanistic basis for the inhibition of Na+ -dependent repetitive firing upon prolonged exposure to gabapentin and may contribute to its anticonvulsant activity. © 2006 Elsevier B.V. All rights reserved. Keywords: Gabapentin; Anticonvulsant; Patch clamp; Nav1.2; Prolonged exposure
1. Introduction Gabapentin (GBP) is a second-generation antiepileptic drug (AED) that is effective in treating various types of seizures in animals (Taylor, 1993a) and humans (McLean, 1995). Recently, clinical efficacy of GBP has also been demonstrated for the treatment of neuropathic pain (Pappagallo, 2003). The mechanism by which GBP exerts these anticonvulsant ∗
Corresponding author. Tel.: +1 215 628 5427; fax: +1 215 628 3297. E-mail address: [email protected] (Y. Liu). 0920-1211/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2006.03.007
and antinociceptive actions is not clearly understood. However, it appears unique among known AEDs and likely involves multiple targets. For example, GBP increases brain ␥-aminobutyric acid (GABA) concentrations by enhancing GABA synthesis/release (Taylor et al., 1998). It binds with high affinity to two auxiliary subunits of high voltage-activated (HVA) calcium channels, ␣2 ␦-1 and ␣2 ␦-2 (Gee et al., 1996; Marais et al., 2001), and at therapeutic concentrations inhibits multiple subtypes of these channels in cultured human and rat neurons (Fink et al., 2001; Stefani et al., 2001; Sutton et al., 2002; Vega-Hernandez and Felix, 2002). However, GBP inhibition of HVA calcium channels
264
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
appears incomplete (Sutton et al., 2002) or occurs only after long periods (≥3 days) of incubation with GBP (Vega-Hernandez and Felix, 2002; Kang et al., 2002), suggesting that these effects may be subtle and only exist under certain experimental conditions. Similarly, sodium channel currents are not inhibited by transient application of GBP (Taylor, 1993b; Stefani et al., 2001), and neither is repetitive firing of Na+ -dependent action potentials in cultured rodent neurons (Rock et al., 1993). However, whereas a preliminary report (Taylor, 1993b) showed that GBP had no effect on rat Nav1.2 (rNav1.2) currents after a 24 h exposure, it has been shown that, in mouse spinal and neocortical neurons, therapeutic concentrations of GBP inhibited high frequency action potential firing in a voltage- and use-dependent manner after prolonged periods (12–48 h) of incubation (Wamil and McLean, 1994). Since GBP effects on calcium channels and Na+ dependent action potentials appear to be dependent on the duration of exposure, the present study was undertaken to better characterize the effect of prolonged GBP treatment on sodium channels. Our results reveal that rNav1.2 was inhibited by GBP only after long incubation periods and thus provide a plausible mechanistic explanation for the inhibition of Na+ -dependent repetitive firing after prolonged exposure to GBP.
2. Methods CHL1610 cells (derived from a Chinese Hamster lung fibroblast cell line) stably expressing rNav1.2 were plated on glass coverslips and incubated with GBP (0–1000 M) for 3 days at 37 ◦ C and in 5% CO2 . The culture media contained DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin and 200 g/ml G418. Recordings were performed using whole-cell patch clamp. In one protocol, a 45-s preconditioning pulse (from −107 to −47 mV with 10 mV increments) was followed by 3 s of brief (5 ms) depolarizations to −7 mV at 10 Hz. The membrane potential during the period between brief depolarizations was the same as the preconditioning voltage. The interval between preconditioning pulses was 15 s, during which time the cell was held at −107 mV. Both fast and slow inactivation was thus allowed to develop at depolarized preconditioning voltages in this protocol. A second protocol
was used to specifically examine fast inactivation. In this protocol, a 20 ms preconditioning pulse (from −87 to +3 mV with 10 mV increments) was followed by a brief (10 ms) depolarization to −7 mV. The interval between preconditioning pulses was 2 s, during which time the cell was held at −107 mV. The extracellular, cell-perfusing solution was applied at 0.5 ml/min and contained (mM): NaCl (132), KCl (5.4), CaCl2 (1.8), MgCl2 (0.8), HEPES (10) and glucose (10) at pH 7.4. The pipette solution contained (mM): CsCl (45), CsF (100), EGTA (5), HEPES (10) and glucose (5) at pH 7.4. The junction potential was corrected using pClamp9.0 (Molecular Devices, Sunnyvale, CA). Pipette resistance ranged from 1 to 2 M. Capacitance transients were cancelled and series resistance was 85% compensated. Whole cell currents were amplified, filtered (2 kHz; Axopatch 200B), sampled (20 kHz; Digidata 1322A) and acquired using pClamp 9 (Molecular Devices). All experiments were conducted at 22 ◦ C. For cells that had been incubated with GBP at a particular concentration, the same concentration was also present in the extracellular solution unless otherwise specified. GBP and lamotrigine (LTG) were from Sigma (St. Louis, MO) and stored either as a 300 mM stock in distilled water (GBP) or as a 100 mM stock in 100% dimethyl sulfoxide (LTG) at −20 ◦ C. The stock solution was diluted freshly before each use. The peak current amplitudes during brief depolarizations (5–10 ms) to −7 mV were used for calculations. Concentration-response relationships were fitted to a logistic function to determine the IC50 value. The voltage dependence of inactivation was fitted to a Boltzmann function to determine V1/2 , the voltage at which 50% of the channels are inactivated. Data fitting was performed using Origin 7.0 (OriginLab, Northampton, MA). A two-tailed t-test or two-way ANOVA was performed as appropriate to determine statistical significance at p ≤ 0.05. Data are expressed as the mean ± S.E.M. The surface proteins on CHL1610 cells stably expressing rNav1.2 were first biotinylated and subsequently isolated with Avidin gel by using PinpointTM Cell Surface Protein Isolation Kit following the manufacturer’s protocol (Pierece Biotechnology, Rockland, IL). Briefly, 1 × 107 cells were plated on each 150 mm plate. Four plates were cultured in the presence of 30 M GBP and two plates in the absence of GBP for 3 days at 37 ◦ C and in 5% CO2 . Cells
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
cultured without GBP were also biotinylated in the absence of GBP. Cells cultured in the presence of GBP were further divided into two groups (2 plates/group). One group was biotinylated in the presence and the other group in the absence of 30 M GBP. Before labeling, the cells were washed twice with ice-cold PBS. The cell surface proteins were then labeled with Sulfo-NHS-SS-Biotin for 30 min at 4 ◦ C. The labeled cells were harvested and lysed with 500 l lysis buffer. The biotinylated proteins in the lysates were isolated with immobilized NeutraAvidinTM gel and the bound proteins were then released by 450 l of SDS-PAGE sample buffer containing 50 mM DTT. Finally, the eluted proteins were concentrated to 200 l by speed-vacuum and used for Western blot analysis. The labeled surface protein samples (30 l each) were subject to 1.5 mm/4–20% SDS-PAGE. After elec-
265
trophoresis at 30 mA for 80 min, the proteins in the gel were transferred to nitrocellulose membrane at 25 V for 90 min. After being blocked with 2% dry milk in TTBS (0.5% Tween 20/100 mM NaCl/10 mM Tris–HCl pH 7.4) for 1 h at room temperature, the blot was incubated overnight with 1:200 dilution of an anti-rNav1.2 polyclonal antibody (Chemicon, Temecula, CA) in 5% dry milk/TTBS at 4 ◦ C. A 1:15,000 dilution of a secondary goat anti-rabbit IgG conjugated with HRP (Pierce Biotechnology) was applied to the blot for 1 h at room temperature. Finally, the signal was visualized on X-ray film using an ECL-plus kit (Amersham Biosciences). All the X-ray films were photographed and the band densities analyzed using FlurochemTM 8000 (Alpha Innotech). The data were normalized to the control sample for each individual experiment (no GBP treatment during cell culture and biotinylation) before averaging.
Fig. 1. Effect of transient (2–3 min) application of GBP and LTG on rNav1.2 channels. (A) Voltage protocol used in (B)–(D). (B) Currents with and without 300 M GBP (normalized to the pre-GBP value at −107 mV; n = 4) have the same amplitude and use- and voltage-dependence. Inset: current traces at Vhold = −107 and −67 mV. (C) LTG (30 M; n = 4) caused voltage- and use-dependent block of rNav1.2. Inset: current traces at Vhold = −107 and −67 mV. (D) Concentration- and voltage-dependent block of rNav1.2 by LTG (n = 4). Denotation of statistical significance (* at p = 0.05, two-tailed t-test) applies only to data at −67 mV, where IC50 = 55.2 and 35.4 M for the 1st and 30th pulses, respectively.
266
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
3. Results Using the voltage protocol depicted in Fig. 1A, application of a relatively high concentration (300 M) of GBP for 2–3 min resulted in no change in rNav1.2 current amplitude from control values (Fig. 1B). This lack of GBP effect was not dependent on the holding potential (−107 or −67 mV) or a high-frequency train of brief (5 ms) depolarizations (30 pulses to −7 mV at 10 Hz). In contrast, Fig. 1C and D demonstrate that transient application of LTG resulted in significant voltage- and use-dependent block of rNav1.2 (IC50 = 55.2 and 35.4 M for the 1st and 30th pulses, respectively, at −67 mV). Next, we studied the functions of rNav1.2 treated with GBP for 3 days.
This prolonged treatment had two appreciable effects on rNav1.2 function. First, the current density was 18–57% smaller in GBP-treated cells (10–1000 M) than that in untreated cells (Fig. 2A). This current reduction by GBP was partial and plateaued at concentrations above 30 M GBP. The reduction was also not acutely reversible, since subsequent perfusion of the same cells with GBP-free solution for 5 min did not result in larger currents (Iwash /I30 M GBP = 1.00 ± 0.06, n = 3). Apparently, it was not due to a decrease in the surface expression of rNav1.2, since GBP treatment (for 3 days) had no effect on the level of rNav1.2 expression in the membrane as determined by Western blot analysis of biotinylated/isolated cell surface proteins with an anti-rNav1.2 antibody (Fig. 2B and C). Second,
Fig. 2. Effect of prolonged (3-day) application of GBP on current density and surface expression of rNav1.2 channels. (A) Average rNav1.2 current density in the absence and presence of GBP (0–1000 M). Vhold = −107 mV and Vtest = −7 mV. Number of cells is indicated in the parentheses. Where applicable, statistical significance (* at p = 0.05, two-tailed t-test) is indicated. (B) Western blot analysis of cell surface expression of rat Nav1.2 in the presence and absence of GBP. Cell surface proteins were biotinylated/isolated and subject to Western blot analysis 3 days after the cells were in culture in the presence or absence of 30 M GBP. Lane 1: no GBP during cell culture and biotinylation; lane 2: cell culture in the presence and biotinylation in the absence of 30 M GBP; lane 3: cell culture and biotinylation in the presence of 30 M GBP. (C) Average of the band density (normalized to the control from each individual experiment) of three experiments as that shown in (B).
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
267
Fig. 3. Effect of prolonged (3-day) application of GBP on inactivation of rNav1.2 channels. (A) Inhibition of rNav1.2 currents by GBP (30 M) at two voltages. Control (n = 12) and GBP-treated (n = 6) current amplitudes were normalized to the respective values at −97 mV. The voltage protocol is similar to that in Fig. 1A. (B) Voltage dependence of rNav1.2 inactivation obtained using the protocol in (A). V1/2 = −60.9, −64.2, −65.3 mV and n = 12, 6, 8 for 0, 30 and 300 M GBP, respectively. The hyperpolarizing shifts by 30 and 300 M GBP are statistically significant at p = 0.05 (two-way ANOVA Bonferroni post-hoc test). The peak value of the first 5-ms depolarization for each condition is used. (C) Voltage dependence of fast inactivation of rNav1.2 (inset: voltage protocol). V1/2 = −38.1, −40.0, −40.6 mV and n = 12, 7, 8 for 0, 30 and 300 M GBP, respectively. The hyperpolarizing shift by 300 M GBP is statistically significant at p = 0.05 (two-way ANOVA Bonferroni post-hoc test). Best fits with a Boltzmann function are superimposed to the data in both (B) and (C) for 0 (solid line), 30 (dotted line) and 300 M GBP (dashed line).
the relative current amplitude at depolarized potentials was significantly smaller in GBP-treated cells than in untreated cells (Fig. 3A). For example, it is 20% and 70% smaller at −67 and −57 mV, respectively, in 30 M GBP-treated than untreated cells for the 30th pulse. This resulted in a significant hyperpolarizing shift of the voltage dependence of (fast plus slow) inactivation for GBP-treated versus untreated channels, by 3.3 and 4.4 mV for 30 and 300 M GBP, respectively (Fig. 3B). We further examined the contribution of fast inactivation to this shift. Both 30 and 300 M GBP caused a small (1.9 and 2.5 mV, respectively) left shift, although it was statistically significant only at 300 M GBP (Fig. 3C).
4. Discussion GBP is a unique AED in its mechanism(s) of action. In addition to blocking voltage-activated calcium channels and increasing brain GABA concentrations, GBP also decreases the conductance of voltage-gated sodium channels. Unlike other known sodium-channel blocking AEDs (e.g., LTG), however, transiently applied GBP does not block these channels. Instead, we found that after 3 days of treatment, GBP significantly decreased the rNav1.2 current density at therapeutically relevant concentrations (e.g., by 57% at 30 M). In addition, this prolonged treatment also caused the channel to inactivate at more hyperpolarized
268
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
potentials, further decreasing the effective sodium conductance during action potentials. Modeling studies indicate that decreases in sodium conductance significantly reduce the likelihood and frequency of repetitive firing (Matzner and Devor, 1992). It is not yet clear what underlies the partial inhibition and slow kinetics of the GBP actions on rNav1.2. Similarly slow effects of GBP have also been observed for calcium channels (Vega-Hernandez and Felix, 2002; Kang et al., 2002). It is conceivable that these effects may result indirectly from GBP actions on other cellular processes, such as interaction of modulatory protein(s) with sodium channels (which could also cause the observed hyperpolarizing shift of channel inactivation). We examined the level of sodium channel expression in the plasma membrane using Western blot analysis of biotinylated cell surface proteins with an anti-rNav1.2 antibody. We found no significant differences between expression levels in untreated and 3-day GBP-treated cells. Therefore, it does not appear that trafficking of sodium channels to the membrane surface was affected by GBP treatment. In addition, the decrease in current density was not a result of the GBP-induced left shift of inactivation because the shift occurred at much more depolarized potentials than the holding potential (−107 mV) used in the experiments to measure current density. The effect of prolonged GBP treatment on rNav1.2 currents was not acutely reversible, suggesting that prolonged GBP exposure likely induced a persistent state in these channels, such that the continued presence of GBP was not necessary to keep the channel in this state (at least on the time scale of our experiments). Alternatively, the effect of GBP might require longer wash periods to reverse if this effect had resulted from highaffinity GBP binding to sodium channels. However, this seems unlikely because the concentration dependence of GBP-induced inhibition does not suggest the existence of such a high-affinity binding site. In conclusion, our results are consistent with the inhibitory effects of prolonged GBP exposure on Na+ dependent repetitive firing in cultured neurons and provide a plausible mechanistic basis for these effects. The inhibition of sodium channels occurs at therapeutic concentrations and thus may contribute to the anticonvulsant and analgesic effects of the drug.
References Fink, K., Dooley, D.J., Meder, W.P., Suman-Chauhan, N., Duffy, S., Clussman, H., Gothert, M., 2001. Inhibition of neuronal Ca2+ influx by gabapentin and pregabalin in the human neocortex. Neuropharmacology 42, 229–236. Gee, N.S., Brown, J.P., Dissanayake, V.U.K., Offord, J., Thurlow, R., Woodruff, G.N., 1996. The novel anticonvulsant drug, gabapentin (Neurotin), binds to the ␣2␦ subunit of a calcium channel. J. Biol. Chem. 271, 5768–5776. Kang, M.G., Felix, R., Campbell, K.P., 2002. Long-term regulation of voltage-gated Ca2+ channels by gabapentin. FEBS Lett. 528, 177–182. Marais, E., Klugbauer, N., Hofmann, F., 2001. Calcium channel ␣2␦ subunits – structure and gabapentin binding. Mol. Pharm. 59, 1243–1248. Matzner, O., Devor, M., 1992. Na+ conductance and the threshold for repetitive neuronal firing. Brain Res. 597, 92– 98. McLean, M.J., 1995. Gabapentin. Epilepsia 36 (Suppl. 2), S73– S86. Pappagallo, M., 2003. Newer antiepileptic drugs: possible uses in the treatment of neuropathic pain and migraine. Clin. Ther. 25, 2506–2538. Rock, D.M., Kelly, K.M., Macdonald, R.L., 1993. Gabapentin actions on ligand- and voltage-gated responses in cultured rodent neurons. Epilepsy Res. 16, 89–98. Stefani, A., Spadoni, F., Giacomini, P., Lavaroni, F., Bernardi, G., 2001. The effects of gabapentin on different ligand- and voltagegated currents in isolated cortical neurons. Epilepsy Res. 43, 239–248. Sutton, K.G., Martin, D.J., Pinnock, R.D., Lee, K., Scott, R.H., 2002. Gabapentin inhibits high-threshold calcium channel currents in cultured rat dorsal root ganglion neurons. Br. J. Pharmacol. 135, 257–265. Taylor, C.P., 1993a. Mechanisms of action of new anti-epileptic drugs. In: Chadwick, D. (Ed.), New Trends in Epilepsy Management: The Role of Gabapentin. Royal Society of Medicine Services, London, pp. 13–40. Taylor, C.P., 1993b. The anticonvulsant lamotrigine blocks sodium currents from cloned ␣-subunits of rat brain Na+ channels in a voltage-dependent manner but gabapentin does not. Soc. Neurosci. Abstr. 23, 1631. Taylor, C.P., Gee, N.S., Su, T., Kocsis, J.D., Welty, D.F., Brown, J.P., Dooley, D.J., Boden, P., Singh, L., 1998. A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Res. 29, 233–249. Vega-Hernandez, A., Felix, R., 2002. Down-regulation of N-type voltage-activated Ca2+ channels by gabapentin. Cell. Mol. Neurobiol. 22, 185–190. Wamil, A.W., McLean, M.J., 1994. Limitation by gabapentin of high frequency action potential firing by mouse central neurons in cell culture. Epilepsy Res. 17, 1–11.
Short communication
Inhibition of the rat brain sodium channel Nav1.2 after prolonged exposure to gabapentin Yi Liu ∗ , Ning Qin, Tasha Reitz, Yan Wang, Christopher M. Flores Analgesics Drug Discovery, Johnson & Johnson Pharmaceutical Research & Development, Welsh & McKean Roads, P.O. Box 776, Spring House, PA 19477-0776, United States Received 20 September 2005; received in revised form 2 March 2006; accepted 15 March 2006 Available online 18 April 2006
Abstract Prolonged exposure of neurons to gabapentin inhibits repetitive firing of Na+ -dependent action potentials. Here, we studied the effect of such prolonged exposure to gabapentin on a rat sodium channel, Nav1.2. After 3 days of continuous incubation with gabapentin (10–1000 M), Nav1.2 current density was decreased dose-dependently relative to untreated cells. The reduction was 57% at 30 M gabapentin, while higher concentrations (100–1000 M) did not result in greater effects. Prolonged treatment with gabapentin also caused the channel to inactivate at more hyperpolarized potentials. These effects provide a mechanistic basis for the inhibition of Na+ -dependent repetitive firing upon prolonged exposure to gabapentin and may contribute to its anticonvulsant activity. © 2006 Elsevier B.V. All rights reserved. Keywords: Gabapentin; Anticonvulsant; Patch clamp; Nav1.2; Prolonged exposure
1. Introduction Gabapentin (GBP) is a second-generation antiepileptic drug (AED) that is effective in treating various types of seizures in animals (Taylor, 1993a) and humans (McLean, 1995). Recently, clinical efficacy of GBP has also been demonstrated for the treatment of neuropathic pain (Pappagallo, 2003). The mechanism by which GBP exerts these anticonvulsant ∗
Corresponding author. Tel.: +1 215 628 5427; fax: +1 215 628 3297. E-mail address: [email protected] (Y. Liu). 0920-1211/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2006.03.007
and antinociceptive actions is not clearly understood. However, it appears unique among known AEDs and likely involves multiple targets. For example, GBP increases brain ␥-aminobutyric acid (GABA) concentrations by enhancing GABA synthesis/release (Taylor et al., 1998). It binds with high affinity to two auxiliary subunits of high voltage-activated (HVA) calcium channels, ␣2 ␦-1 and ␣2 ␦-2 (Gee et al., 1996; Marais et al., 2001), and at therapeutic concentrations inhibits multiple subtypes of these channels in cultured human and rat neurons (Fink et al., 2001; Stefani et al., 2001; Sutton et al., 2002; Vega-Hernandez and Felix, 2002). However, GBP inhibition of HVA calcium channels
264
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
appears incomplete (Sutton et al., 2002) or occurs only after long periods (≥3 days) of incubation with GBP (Vega-Hernandez and Felix, 2002; Kang et al., 2002), suggesting that these effects may be subtle and only exist under certain experimental conditions. Similarly, sodium channel currents are not inhibited by transient application of GBP (Taylor, 1993b; Stefani et al., 2001), and neither is repetitive firing of Na+ -dependent action potentials in cultured rodent neurons (Rock et al., 1993). However, whereas a preliminary report (Taylor, 1993b) showed that GBP had no effect on rat Nav1.2 (rNav1.2) currents after a 24 h exposure, it has been shown that, in mouse spinal and neocortical neurons, therapeutic concentrations of GBP inhibited high frequency action potential firing in a voltage- and use-dependent manner after prolonged periods (12–48 h) of incubation (Wamil and McLean, 1994). Since GBP effects on calcium channels and Na+ dependent action potentials appear to be dependent on the duration of exposure, the present study was undertaken to better characterize the effect of prolonged GBP treatment on sodium channels. Our results reveal that rNav1.2 was inhibited by GBP only after long incubation periods and thus provide a plausible mechanistic explanation for the inhibition of Na+ -dependent repetitive firing after prolonged exposure to GBP.
2. Methods CHL1610 cells (derived from a Chinese Hamster lung fibroblast cell line) stably expressing rNav1.2 were plated on glass coverslips and incubated with GBP (0–1000 M) for 3 days at 37 ◦ C and in 5% CO2 . The culture media contained DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin and 200 g/ml G418. Recordings were performed using whole-cell patch clamp. In one protocol, a 45-s preconditioning pulse (from −107 to −47 mV with 10 mV increments) was followed by 3 s of brief (5 ms) depolarizations to −7 mV at 10 Hz. The membrane potential during the period between brief depolarizations was the same as the preconditioning voltage. The interval between preconditioning pulses was 15 s, during which time the cell was held at −107 mV. Both fast and slow inactivation was thus allowed to develop at depolarized preconditioning voltages in this protocol. A second protocol
was used to specifically examine fast inactivation. In this protocol, a 20 ms preconditioning pulse (from −87 to +3 mV with 10 mV increments) was followed by a brief (10 ms) depolarization to −7 mV. The interval between preconditioning pulses was 2 s, during which time the cell was held at −107 mV. The extracellular, cell-perfusing solution was applied at 0.5 ml/min and contained (mM): NaCl (132), KCl (5.4), CaCl2 (1.8), MgCl2 (0.8), HEPES (10) and glucose (10) at pH 7.4. The pipette solution contained (mM): CsCl (45), CsF (100), EGTA (5), HEPES (10) and glucose (5) at pH 7.4. The junction potential was corrected using pClamp9.0 (Molecular Devices, Sunnyvale, CA). Pipette resistance ranged from 1 to 2 M. Capacitance transients were cancelled and series resistance was 85% compensated. Whole cell currents were amplified, filtered (2 kHz; Axopatch 200B), sampled (20 kHz; Digidata 1322A) and acquired using pClamp 9 (Molecular Devices). All experiments were conducted at 22 ◦ C. For cells that had been incubated with GBP at a particular concentration, the same concentration was also present in the extracellular solution unless otherwise specified. GBP and lamotrigine (LTG) were from Sigma (St. Louis, MO) and stored either as a 300 mM stock in distilled water (GBP) or as a 100 mM stock in 100% dimethyl sulfoxide (LTG) at −20 ◦ C. The stock solution was diluted freshly before each use. The peak current amplitudes during brief depolarizations (5–10 ms) to −7 mV were used for calculations. Concentration-response relationships were fitted to a logistic function to determine the IC50 value. The voltage dependence of inactivation was fitted to a Boltzmann function to determine V1/2 , the voltage at which 50% of the channels are inactivated. Data fitting was performed using Origin 7.0 (OriginLab, Northampton, MA). A two-tailed t-test or two-way ANOVA was performed as appropriate to determine statistical significance at p ≤ 0.05. Data are expressed as the mean ± S.E.M. The surface proteins on CHL1610 cells stably expressing rNav1.2 were first biotinylated and subsequently isolated with Avidin gel by using PinpointTM Cell Surface Protein Isolation Kit following the manufacturer’s protocol (Pierece Biotechnology, Rockland, IL). Briefly, 1 × 107 cells were plated on each 150 mm plate. Four plates were cultured in the presence of 30 M GBP and two plates in the absence of GBP for 3 days at 37 ◦ C and in 5% CO2 . Cells
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
cultured without GBP were also biotinylated in the absence of GBP. Cells cultured in the presence of GBP were further divided into two groups (2 plates/group). One group was biotinylated in the presence and the other group in the absence of 30 M GBP. Before labeling, the cells were washed twice with ice-cold PBS. The cell surface proteins were then labeled with Sulfo-NHS-SS-Biotin for 30 min at 4 ◦ C. The labeled cells were harvested and lysed with 500 l lysis buffer. The biotinylated proteins in the lysates were isolated with immobilized NeutraAvidinTM gel and the bound proteins were then released by 450 l of SDS-PAGE sample buffer containing 50 mM DTT. Finally, the eluted proteins were concentrated to 200 l by speed-vacuum and used for Western blot analysis. The labeled surface protein samples (30 l each) were subject to 1.5 mm/4–20% SDS-PAGE. After elec-
265
trophoresis at 30 mA for 80 min, the proteins in the gel were transferred to nitrocellulose membrane at 25 V for 90 min. After being blocked with 2% dry milk in TTBS (0.5% Tween 20/100 mM NaCl/10 mM Tris–HCl pH 7.4) for 1 h at room temperature, the blot was incubated overnight with 1:200 dilution of an anti-rNav1.2 polyclonal antibody (Chemicon, Temecula, CA) in 5% dry milk/TTBS at 4 ◦ C. A 1:15,000 dilution of a secondary goat anti-rabbit IgG conjugated with HRP (Pierce Biotechnology) was applied to the blot for 1 h at room temperature. Finally, the signal was visualized on X-ray film using an ECL-plus kit (Amersham Biosciences). All the X-ray films were photographed and the band densities analyzed using FlurochemTM 8000 (Alpha Innotech). The data were normalized to the control sample for each individual experiment (no GBP treatment during cell culture and biotinylation) before averaging.
Fig. 1. Effect of transient (2–3 min) application of GBP and LTG on rNav1.2 channels. (A) Voltage protocol used in (B)–(D). (B) Currents with and without 300 M GBP (normalized to the pre-GBP value at −107 mV; n = 4) have the same amplitude and use- and voltage-dependence. Inset: current traces at Vhold = −107 and −67 mV. (C) LTG (30 M; n = 4) caused voltage- and use-dependent block of rNav1.2. Inset: current traces at Vhold = −107 and −67 mV. (D) Concentration- and voltage-dependent block of rNav1.2 by LTG (n = 4). Denotation of statistical significance (* at p = 0.05, two-tailed t-test) applies only to data at −67 mV, where IC50 = 55.2 and 35.4 M for the 1st and 30th pulses, respectively.
266
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
3. Results Using the voltage protocol depicted in Fig. 1A, application of a relatively high concentration (300 M) of GBP for 2–3 min resulted in no change in rNav1.2 current amplitude from control values (Fig. 1B). This lack of GBP effect was not dependent on the holding potential (−107 or −67 mV) or a high-frequency train of brief (5 ms) depolarizations (30 pulses to −7 mV at 10 Hz). In contrast, Fig. 1C and D demonstrate that transient application of LTG resulted in significant voltage- and use-dependent block of rNav1.2 (IC50 = 55.2 and 35.4 M for the 1st and 30th pulses, respectively, at −67 mV). Next, we studied the functions of rNav1.2 treated with GBP for 3 days.
This prolonged treatment had two appreciable effects on rNav1.2 function. First, the current density was 18–57% smaller in GBP-treated cells (10–1000 M) than that in untreated cells (Fig. 2A). This current reduction by GBP was partial and plateaued at concentrations above 30 M GBP. The reduction was also not acutely reversible, since subsequent perfusion of the same cells with GBP-free solution for 5 min did not result in larger currents (Iwash /I30 M GBP = 1.00 ± 0.06, n = 3). Apparently, it was not due to a decrease in the surface expression of rNav1.2, since GBP treatment (for 3 days) had no effect on the level of rNav1.2 expression in the membrane as determined by Western blot analysis of biotinylated/isolated cell surface proteins with an anti-rNav1.2 antibody (Fig. 2B and C). Second,
Fig. 2. Effect of prolonged (3-day) application of GBP on current density and surface expression of rNav1.2 channels. (A) Average rNav1.2 current density in the absence and presence of GBP (0–1000 M). Vhold = −107 mV and Vtest = −7 mV. Number of cells is indicated in the parentheses. Where applicable, statistical significance (* at p = 0.05, two-tailed t-test) is indicated. (B) Western blot analysis of cell surface expression of rat Nav1.2 in the presence and absence of GBP. Cell surface proteins were biotinylated/isolated and subject to Western blot analysis 3 days after the cells were in culture in the presence or absence of 30 M GBP. Lane 1: no GBP during cell culture and biotinylation; lane 2: cell culture in the presence and biotinylation in the absence of 30 M GBP; lane 3: cell culture and biotinylation in the presence of 30 M GBP. (C) Average of the band density (normalized to the control from each individual experiment) of three experiments as that shown in (B).
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
267
Fig. 3. Effect of prolonged (3-day) application of GBP on inactivation of rNav1.2 channels. (A) Inhibition of rNav1.2 currents by GBP (30 M) at two voltages. Control (n = 12) and GBP-treated (n = 6) current amplitudes were normalized to the respective values at −97 mV. The voltage protocol is similar to that in Fig. 1A. (B) Voltage dependence of rNav1.2 inactivation obtained using the protocol in (A). V1/2 = −60.9, −64.2, −65.3 mV and n = 12, 6, 8 for 0, 30 and 300 M GBP, respectively. The hyperpolarizing shifts by 30 and 300 M GBP are statistically significant at p = 0.05 (two-way ANOVA Bonferroni post-hoc test). The peak value of the first 5-ms depolarization for each condition is used. (C) Voltage dependence of fast inactivation of rNav1.2 (inset: voltage protocol). V1/2 = −38.1, −40.0, −40.6 mV and n = 12, 7, 8 for 0, 30 and 300 M GBP, respectively. The hyperpolarizing shift by 300 M GBP is statistically significant at p = 0.05 (two-way ANOVA Bonferroni post-hoc test). Best fits with a Boltzmann function are superimposed to the data in both (B) and (C) for 0 (solid line), 30 (dotted line) and 300 M GBP (dashed line).
the relative current amplitude at depolarized potentials was significantly smaller in GBP-treated cells than in untreated cells (Fig. 3A). For example, it is 20% and 70% smaller at −67 and −57 mV, respectively, in 30 M GBP-treated than untreated cells for the 30th pulse. This resulted in a significant hyperpolarizing shift of the voltage dependence of (fast plus slow) inactivation for GBP-treated versus untreated channels, by 3.3 and 4.4 mV for 30 and 300 M GBP, respectively (Fig. 3B). We further examined the contribution of fast inactivation to this shift. Both 30 and 300 M GBP caused a small (1.9 and 2.5 mV, respectively) left shift, although it was statistically significant only at 300 M GBP (Fig. 3C).
4. Discussion GBP is a unique AED in its mechanism(s) of action. In addition to blocking voltage-activated calcium channels and increasing brain GABA concentrations, GBP also decreases the conductance of voltage-gated sodium channels. Unlike other known sodium-channel blocking AEDs (e.g., LTG), however, transiently applied GBP does not block these channels. Instead, we found that after 3 days of treatment, GBP significantly decreased the rNav1.2 current density at therapeutically relevant concentrations (e.g., by 57% at 30 M). In addition, this prolonged treatment also caused the channel to inactivate at more hyperpolarized
268
Y. Liu et al. / Epilepsy Research 70 (2006) 263–268
potentials, further decreasing the effective sodium conductance during action potentials. Modeling studies indicate that decreases in sodium conductance significantly reduce the likelihood and frequency of repetitive firing (Matzner and Devor, 1992). It is not yet clear what underlies the partial inhibition and slow kinetics of the GBP actions on rNav1.2. Similarly slow effects of GBP have also been observed for calcium channels (Vega-Hernandez and Felix, 2002; Kang et al., 2002). It is conceivable that these effects may result indirectly from GBP actions on other cellular processes, such as interaction of modulatory protein(s) with sodium channels (which could also cause the observed hyperpolarizing shift of channel inactivation). We examined the level of sodium channel expression in the plasma membrane using Western blot analysis of biotinylated cell surface proteins with an anti-rNav1.2 antibody. We found no significant differences between expression levels in untreated and 3-day GBP-treated cells. Therefore, it does not appear that trafficking of sodium channels to the membrane surface was affected by GBP treatment. In addition, the decrease in current density was not a result of the GBP-induced left shift of inactivation because the shift occurred at much more depolarized potentials than the holding potential (−107 mV) used in the experiments to measure current density. The effect of prolonged GBP treatment on rNav1.2 currents was not acutely reversible, suggesting that prolonged GBP exposure likely induced a persistent state in these channels, such that the continued presence of GBP was not necessary to keep the channel in this state (at least on the time scale of our experiments). Alternatively, the effect of GBP might require longer wash periods to reverse if this effect had resulted from highaffinity GBP binding to sodium channels. However, this seems unlikely because the concentration dependence of GBP-induced inhibition does not suggest the existence of such a high-affinity binding site. In conclusion, our results are consistent with the inhibitory effects of prolonged GBP exposure on Na+ dependent repetitive firing in cultured neurons and provide a plausible mechanistic basis for these effects. The inhibition of sodium channels occurs at therapeutic concentrations and thus may contribute to the anticonvulsant and analgesic effects of the drug.
References Fink, K., Dooley, D.J., Meder, W.P., Suman-Chauhan, N., Duffy, S., Clussman, H., Gothert, M., 2001. Inhibition of neuronal Ca2+ influx by gabapentin and pregabalin in the human neocortex. Neuropharmacology 42, 229–236. Gee, N.S., Brown, J.P., Dissanayake, V.U.K., Offord, J., Thurlow, R., Woodruff, G.N., 1996. The novel anticonvulsant drug, gabapentin (Neurotin), binds to the ␣2␦ subunit of a calcium channel. J. Biol. Chem. 271, 5768–5776. Kang, M.G., Felix, R., Campbell, K.P., 2002. Long-term regulation of voltage-gated Ca2+ channels by gabapentin. FEBS Lett. 528, 177–182. Marais, E., Klugbauer, N., Hofmann, F., 2001. Calcium channel ␣2␦ subunits – structure and gabapentin binding. Mol. Pharm. 59, 1243–1248. Matzner, O., Devor, M., 1992. Na+ conductance and the threshold for repetitive neuronal firing. Brain Res. 597, 92– 98. McLean, M.J., 1995. Gabapentin. Epilepsia 36 (Suppl. 2), S73– S86. Pappagallo, M., 2003. Newer antiepileptic drugs: possible uses in the treatment of neuropathic pain and migraine. Clin. Ther. 25, 2506–2538. Rock, D.M., Kelly, K.M., Macdonald, R.L., 1993. Gabapentin actions on ligand- and voltage-gated responses in cultured rodent neurons. Epilepsy Res. 16, 89–98. Stefani, A., Spadoni, F., Giacomini, P., Lavaroni, F., Bernardi, G., 2001. The effects of gabapentin on different ligand- and voltagegated currents in isolated cortical neurons. Epilepsy Res. 43, 239–248. Sutton, K.G., Martin, D.J., Pinnock, R.D., Lee, K., Scott, R.H., 2002. Gabapentin inhibits high-threshold calcium channel currents in cultured rat dorsal root ganglion neurons. Br. J. Pharmacol. 135, 257–265. Taylor, C.P., 1993a. Mechanisms of action of new anti-epileptic drugs. In: Chadwick, D. (Ed.), New Trends in Epilepsy Management: The Role of Gabapentin. Royal Society of Medicine Services, London, pp. 13–40. Taylor, C.P., 1993b. The anticonvulsant lamotrigine blocks sodium currents from cloned ␣-subunits of rat brain Na+ channels in a voltage-dependent manner but gabapentin does not. Soc. Neurosci. Abstr. 23, 1631. Taylor, C.P., Gee, N.S., Su, T., Kocsis, J.D., Welty, D.F., Brown, J.P., Dooley, D.J., Boden, P., Singh, L., 1998. A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Res. 29, 233–249. Vega-Hernandez, A., Felix, R., 2002. Down-regulation of N-type voltage-activated Ca2+ channels by gabapentin. Cell. Mol. Neurobiol. 22, 185–190. Wamil, A.W., McLean, M.J., 1994. Limitation by gabapentin of high frequency action potential firing by mouse central neurons in cell culture. Epilepsy Res. 17, 1–11.