Effects of Carbamazepine, Lacosamide and Zonisamide on Gliotransmitter Release Associated with Activated Astroglial Hemichannels

Abstract

Recent studies using the genetic partial epilepsy model have demonstrated that hyperfunction of astroglial hemichannels contributes to pathomechanism of epileptic seizure. Therefore, to explore the novel anticonvulsive mechanisms, the present study determined the effects of voltage-dependent Na⁺ channel (VDSC)-inhibiting anticonvulsants, carbamazepine (CBZ), lacosamide (LCM), and zonisamide (ZNS) on the astroglial release of l-glutamate and adenosine triphosphate (ATP). The effects of subchronic administration of therapeutic-relevant dose of three anticonvulsants on the release of l-glutamate and ATP in the orbitofrontal cortex (OFC) were determined using microdialysis. The concentration-dependent effects of acute and subchronic administrations of anticonvulsants on astroglial gliotransmitter release were determined using primary cultured astrocytes. The concentration-dependent effects of subchronic administrations of anticonvulsants on connexin43 (Cx43) expression in the plasma membrane of primary cultured astrocytes were determined using the Simple Western system. An increase in the levels of extracellular K⁺ resulted in a concentration-dependent increase in the astroglial release of l-glutamate and ATP. The depleted levels of extracellular Ca²⁺ alone did not affect astroglial gliotransmitter release but did accelerate K⁺-evoked gliotransmitter release via activation of astroglial hemichannels. Both non-selective hemichannel inhibitor carbenoxolone (CBX) and selective Cx43 inhibitor GAP19 prevented both gliotransmitter release through activated astroglial hemichannels and the hemichannel-activating process induced by elevation of the levels of extracellular K⁺ with depletion of the levels of extracellular Ca²⁺. ZNS subchronically decreased Cx43 expression and acutely/subchronically inhibited Cx43 hemichannel activity. LCM acutely inhibited hemichannel activity but did not subchronically affect Cx43 expression. Therapeutic-relevant concentration of CBZ did not affect hemichannel activity or Cx43 expression, but supratherapeutic concentration of CBZ decreased Cx43 expression and hemichannel activity. Therefore, the present study demonstrated the distinct effects of CBZ, LCM, and ZNS on gliotransmitter release via modulation of astroglial hemichannel function. The different features of the effects of three VDSC-inhibiting anticonvulsants on astroglial transmission associated with hemichannels, at least partially, possibly contributing to the formation of the properties of these three anticonvulsants, including the antiepileptic spectrum and adverse effects regarding mood and cognitive disturbance.

Keywords: astrocyte; carbamazepine; hemichannel; lacosamide; microdialysis; zonisamide.

Figure 1

Effects of the extracellular Ca²⁺ and K⁺ and hemichannel inhibitors carbenoxolone (CBX; a non-selective inhibitor, 100 μM) and GAP19 (a selective Cx43 inhibitor, 20 μM) on the astroglial release of (A) l-glutamate and (B) adenosine triphosphate (ATP). Primary cultured astrocytes were incubated in artificial cerebrospinal fluid (ACSF), Ca²⁺-free ACSF (FC-ACSF), K⁺-containing ACSF (HK-ACSF; 100 mM), and Ca²⁺-free with 100 mM K⁺-containing ACSF (FCHK-ACSF) for 20 min. Ordinate: mean ± standard deviation (SD) (n = 6) of the extracellular levels of l-glutamate (μM) and ATP (nM). ** p < 0.01 relative to ACSF, and ^@ p < 0.05 and ^@@ p < 0.01 relative to the control (without hemichannel inhibitors) by MANOVA with Tukey’s post-hoc test.

Figure 2

Effects of the hemichannel inhibitors (20 μM GAP19 and 100 μM CBX) on the astroglial release of (A) l-glutamate, (B) ATP induced by FCHK-ACSF associated with activated astroglial hemichannels, and (C) the experimental design. Ordinate: mean ± SD (n = 6) of the levels of l-glutamate (μM) and ATP (nM). ** p < 0.01 relative to the control by two-way ANOVA with Tukey’s post-hoc test.

Figure 3

Effects of the hemichannel inhibitors (20 μM GAP19 and 100 μM CBX) on the repetitive K⁺-evoked (FCHK-ACSF) astroglial release of (A) l-glutamate, (B) ATP, and (C) the experimental design. Ordinate: mean ± SD (n = 6) of the levels of l-glutamate (μM) and ATP (nM). ** p < 0.01 relative to the control by two-way ANOVA with Tukey’s post-hoc test.

Figure 4

Concentration-dependent effects of the acute (open circles) and subchronic (closed circles) administration of (A,D) zonisamide (ZNS) (30, 100, or 300 μM), (B,E) carbamazepine (CBZ) (10, 30, or 100 μM), and (C,F) lacosamide (LCM) (10, 30, or 100 μM) on the basal astroglial release of (A–C) l-glutamate and (D–F)ATP. Ordinate: mean ± SD (n = 6) of the basal astroglial release of l-glutamate (μM) and ATP (nM). Abscissa: concentration of ZNS, CBZ, and LCM (μM).

Figure 5

Concentration-dependent effects of the acute (squares) and subchronic (circles) administration of (A,D) ZNS (30, 100, or 300 μM), (B,E) CBZ (10, 30, or 100 μM), and (C,F) LCM (10, 30, or 100 μM) on the repetitive K⁺-evoked (first: closed marks; second: opened marks) astroglial release of (A–C) l-glutamate and (D–F) ATP. Ordinate: mean ± SD (n = 6) of the astroglial release of l-glutamate (μM) and ATP (nM). Abscissa: concentration of ZNS, CBZ, and LCM (μM). * p < 0.05 and ** p < 0.01 relative to the anticonvulsant-free condition, and ^@ p < 0.05 and ^@@ p < 0.01 relative to acute administration by MANOVA with Tukey’s multiple comparison. The F-values of the effects of ZNS on l-glutamate and ATP were F_ZNS (3,30) = 77.4 (p < 0.01), F_event (1,10) = 147.6 (p < 0.01), F_treatment (1,10) = 0.4 (p > 0.05), F_{ZNS*treatment} (3,30) = 3.2 (p < 0.05), F_{event*treatment} (1,10) = 0.2 (p > 0.05), F_ZNS*event (3,30) = 0.5 (p > 0.05), and F_{ZNS*event*treatment} (3,30) = 0.2 (p > 0.05), and on ATP were F_ZNS (3,30) = 167.9 (p < 0.01), F_event (1,10) = 422.5 (p < 0.01), F_treatment (1,10) = 0.8 (p > 0.05), F_{ZNS*treatment} (3,30) = 10.2 (p < 0.01), F_{event*treatment} (1,10) = 0.7 (p > 0.05), F_ZNS*event (3,30) = 19.1 (p < 0.01), and F_{ZNS*event*treatment} (3,30) = 1.3 (p > 0.05), respectively. The F-values of the effects of CBZ on l-glutamate and ATP were F_CBZ (3,30) = 64.3 (p < 0.01), F_event (1,10) = 402.2 (p < 0.01), F_treatment (1,10) = 0.4 (p > 0.05), F_{CBZ*treatment} (3,30) = 12.4 (p < 0.01), F_{event*treatment} (1,10) = 1.1 (p > 0.05), F_CBZ*event (3,30) = 0.5 (p > 0.05), and F_{CBZ*event*treatment} (3,30) = 0.2 (p > 0.05), and on ATP were F_CBZ (3,30) = 26.9 (p < 0.01), F_event (1,10) = 388.0 (p < 0.01), F_treatment (1,10) = 0.4 (p > 0.05), F_{CBZ*treatment} (3,30) = 8.7 (p > 0.05), F_{event*treatment} (1,10) = 0.1 (p > 0.05), F_CBZ*event (3,30) = 1.0 (p > 0.05), F_{CBZ*event*treatment} (3,30) = 1.8 (p > 0.05), respectively. The F-values of the effects of LCM on l-glutamate and ATP were F_LCM (3,30) = 9.4 (p < 0.01), F_event (1,10) = 286.3 (p < 0.01), F_treatment (1,10) = 0.3 (p > 0.05), F_{LCM*treatment} (3,30) = 2.4 (p > 0.05), F_{event*treatment} (1,10) = 6.4 (p < 0.01), F_LCM*event (3,30) = 1.3 (p > 0.05), and F_{LCM*event*treatment} (3,30) = 2.3 (p > 0.05), and on ATP were F_LCM (3,30) = 11.3 (p < 0.01), F_event (1,10) = 441.7 (p < 0.01), F_treatment (1,10) = 0.8 (p > 0.05), F_{LCM*treatment} (3,30) = 3.9 (p < 0.05), F_{event*treatment}(1,10) = 6.6 (p < 0.05), F_LCM*event (3,30) = 5.5 (p < 0.01), and F_{LCM*event*treatment} (3,30) = 8.8 (p < 0.01), respectively.

Figure 6

Concentration-dependent effects of the subchronic administration of LCM (30 or 100 μM), CBZ (30 or 100 μM), and ZNS (30, 100, or 300 μM) on the Cx43 expression in the plasma membrane fraction of primary cultured astrocytes (A) and pseudo-gel images using the Simple Western results with anti-Cx43 (B) and anti-GAPDH (C) antibodies for blotting of the plasma membrane fractions. In (A), ordinate: mean ± SD (n = 6) of the relative protein level of Cx43. * p < 0.05 relative to the control by one-way ANOVA with Tukey’s multiple comparison.

Figure 7

Effects of perfusion with 50 mM K⁺ containing modified Ringer’s solution (MRS) (MK-MRS), 100 mM K⁺ containing MRS (HK-MRS), and Ca²⁺-free with 100 mM K⁺ containing MRS (FCHK-MRS) on the extracellular levels of (A,B) l-glutamate and (C,D) ATP in the orbitofrontal cortex (OFC). Ordinates: the mean extracellular levels of l-glutamate (μM) and ATP (nM) (n = 6); abscissas: time after perfusion with MK-MRS, HK-MRS, or FCHK-MRS for 20 min (black bars). The area under the curve (AUC) values of the extracellular levels of l-glutamate (nmol) and ATP (pmol) after perfusion with MK-MRS, HK-MRS, or FCHK-MRS during 20–180 min of (A,B) are represented in (C,D), respectively. * p < 0.05 and ** p < 0.01 relative to MK-MRS, and ^@ p < 0.05 relative to HK-MRS by MANOVA with Tukey’s post-hoc test.

Figure 8

Effects of the subchronic administration of a therapeutic-relevant dose of (B,F) ZNS, (C,G) CBZ, and (D,H) and LCM (25 mg/kg/day for 7 days) on the repetitive K⁺-evoked (first stimulation: opened circles; second stimulation: closed circles) release of (A–D) l-glutamate and (E–H) ATP in the OFC. Rats were subchronically administrated with a therapeutic-relevant dose of anticonvulsants. Perfusion medium in the OFC was commenced with MRS. After the stabilization of the levels of l-glutamate and ATP in the OFC, the perfusion medium was switched from MRS to FCHK-MRS for 20 min (first stimulation). After the first K⁺-evoked stimulation, the perfusate was returned to MRS for 240 min (recovery). Following recovery, the perfusate was switched to FCHK-MRS for 20 min again (second stimulation). After the second FCHK-MRS stimulation, the perfusate was returned to MRS again. Ordinates: mean extracellular levels of l-glutamate (μM) and ATP (nM) (n = 6); abscissas: time after the first or second FCHK-MRS stimulations (min). Black bars indicate the perfusion with FCHK-MRS (first and second K⁺-evoked stimulation).

Figure 9

Effects of the subchronic administration of a therapeutic-relevant dose of ZNS, CBZ, and LCM (25 mg/kg/day for 7 days) on the repetitive FCHK-MRS-evoked (first stimulation: HKMRS1; second stimulation: HKMRS2) release of (A) l-glutamate and (B) ATP in the OFC. Ordinates: mean AUC values of the extracellular levels of l-glutamate (nmol) and ATP (pmol) before (basal extracellular l-glutamate levels) and after K⁺-evoked stimulation (from 20 to 180 min) of Figure 1. Gray and colored columns indicate the AUC values of the basal and K⁺-evoked releases, respectively. * p < 0.05 and ** p < 0.01; relative to the control, ^@ p < 0.05 and ^@@ p < 0.01 relative to HKMRS (first simulation) by MANOVA with Tukey’s post-hoc test. The F-values of the effects of ZNS on l-glutamate and ATP were F_ZNS (1,200) = 185.0 (p < 0.01), F_time (9,200) = 53.4 (p < 0.01), F_event (1,10) = 47.0 (p < 0.01), F_ZNS*time (9,200) = 3.7 (p < 0.01), F_ZNS*event(1,200) = 16.9 (p < 0.01), F_event*time (9,200) = 2.3 (p < 0.05), and F_{ZNS*event*time} (9,200) = 1.1 (p > 0.05), and on ATP were F_ZNS (1,200) = 216.2 (p < 0.01), F_time (9,200) = 11.5 (p < 0.01), F_event (1,10) = 70.5 (p < 0.01), F_ZNS*time (9,200) = 4.8 (p < 0.01), F_ZNS*event (1,200) = 45.7 (p < 0.01), F_event*time (9,200) = 0.9 (p > 0.05), and F_{ZNS*event*time} (9,200) = 1.7 (p > 0.05), respectively. The F-values of the effects of CBZ on l-glutamate and ATP were F_CBZ (1,200) = 76.9 (p < 0.01), F_time (9,200) = 50.4 (p < 0.01), F_event (1,10) = 63.2 (p < 0.01), F_CBZ*time (9,200) = 2.5 (p < 0.05), F_CBZ*event (1,200) = 5.7 (p < 0.05), F_event*time (9,200) = 1.8 (p > 0.05), and F_{CBZ*event*time} (9,200) = 1.0 (p > 0.05), and on ATP were F_CBZ (1,200) = 57.6 (p < 0.01), F_time (9,200) = 11.4 (p < 0.01), F_event (1,10) = 138.4 (p < 0.01), F_CBZ*time (9,200) = 0.6 (p > 0.05), F_CBZ*event (1,200) = 9.6 (p < 0.05), F_event*time (9,200) = 1.7 (p > 0.05), F_{CBZ*event*time} (9,200) = 1.1 (p > 0.05), respectively. The F-values of the effects of LCM on l-glutamate and ATP were F_LCM (1,200) = 147.8 (p < 0.01), F_time (9,200) = 58.0 (p < 0.01), F_event (1,10) = 44.2 (p < 0.01), F_LCM*time (9,200) = 2.2 (p < 0.05), F_LCM*event (1,200) = 16.5 (p < 0.01), F_event*time (9,200) = 2.7 (p < 0.01), and F_{CBZ*event*time} (9,200) = 0.8 (p > 0.05), and on ATP were F_LCM (1,200) = 148.3 (p < 0.01), F_time (9,200) = 13.8 (p < 0.01), F_event (1,10) = 103.5 (p < 0.01), F_LCM*time (9,200) = 3.8 (p < 0.05), F_LCM*event (1,200) = 35.6 (p < 0.01), F_event*time (9,200) = 3.2 (p < 0.01), and F_{LCM*event*time} (9,200) = 0.7 (p > 0.05), respectively.