Pharmacological correction of gating defects in the voltage- gated Ca_v2.1 Ca²⁺ channel due to a familial hemiplegic migraine mutation

SUMMARY

Voltage-gated ion channels exhibit complex properties, which can be targeted in pharmacological therapies for disease. Here, we report that the pro-oxidant, tert-butyl dihydroquinone (BHQ), modulates Ca_v2.1 Ca²⁺ channels in ways that oppose defects in channel gating and synaptic transmission resulting from a familial hemiplegic migraine mutation (S218L). BHQ slows deactivation, inhibits voltage-dependent activation, and potentiates Ca²⁺- dependent facilitation of Ca_v2.1 channels in transfected HEK293T cells. These actions of BHQ help offset the gain-of-function and reduced Ca²⁺-dependent facilitation of Ca_v2.1 channels with the S218L mutation. Transgenic expression of the mutant channels at the Drosophila neuromuscular junction causes abnormally elevated evoked postsynaptic potentials and impaired synaptic plasticity, which are largely restored to the wild-type phenotypes by BHQ. Our results reveal a new mechanism by which a Ca_v2.1 gating modifier can ameliorate defects associated with a disease-causing mutation in Ca_v2.1.

INTRODUCTION

Voltage-gated Ca_v2.1 Ca²⁺ channels mediate P/Q-type Ca²⁺ currents that control neuronal excitability and neurotransmitter release throughout the nervous system (Catterall, 2000). The neurophysiological requirement for Ca_v2.1 is illustrated by the severe ataxia, dyskinesia, and absence epilepsy in mice lacking Ca_v2.1 (Jun et al., 1999; Maejima et al., 2013; Mark et al., 2011). In addition, human mutations in the CACNA1A gene encoding the pore-forming α₁ subunit of Ca_v2.1 (α₁2.1) cause multiple neurological disorders including episodic ataxia (Ophoff et al., 1996), spinocerebellar ataxia (Zhuchenko et al., 1997), and familial hemiplegic migraine (FHM1) (Ophoff et al., 1996).

Among Ca_v channels, Ca_v2.1 exhibits particularly complex behavior and modulation. Single Ca_v2.1 channels can alternate between gating modes characterized by distinct kinetics and voltage-dependent activation (Fellin et al., 2004; Luvisetto et al., 2004). In addition, Ca_v2.1 channels undergo both positive and negative feedback regulation by Ca²⁺, Ca²⁺-dependent facilitation (CDF) and inactivation (CDI), respectively, which contributes to short-term synaptic plasticity (Catterall and Few, 2008; Christel and Lee, 2012). While dihydropyridine analogs have greatly advanced understanding of Ca_v1 channel gating in normal and diseased states (Erxleben et al., 2006; Nowycky et al., 1985), the scarcity of pharmacological modulators for Ca_v2.1 limits similar mechanistic insights for these channels.

2,5′-di(tertbutyl)-1,4,-benzohydroquinone (BHQ) is a synthetic phenolic compound that inhibits sarco-endoplasmic Ca²⁺ ATP-ases (SERCAs) (Moore et al., 1987) and has pro-oxidant properties (Fusi et al., 1999). Here, we report that independent of these actions, BHQ inhibits Ca_v2.1 voltage-dependent activation, and enhances CDF during high-frequency stimuli. These actions of BHQ ameliorate gating defects in Ca_v2.1 and synaptic transmission due to a S218L mutation in α₁2.1 that causes a severe form of FHM1 in humans (Kors et al., 2001). Our results reveal unique features of Ca_v2.1 gating that may be exploited for therapeutic benefit.

RESULTS

BHQ inhibits Ca_v2.1 activation and slows deactivation

During studies to evaluate the effect of intracellular Ca²⁺ on Ca_v2.1 in transfected HEK293T cells, we discovered novel modulatory actions of BHQ on Ca_v2.1 properties. BHQ reversibly inhibited Ca_v2.1 Ba²⁺ currents (I_Ba) in a dose-dependent manner (IC₅₀ of 7.9 ± 1.0 μM; Fig. 1A–C) and across a range of voltages (Fig. 1D,E). To determine the underlying mechanism, we analyzed tail currents, which are evoked by the repolarization-induced change in driving force and reflect the population of open channels at the end of the test depolarization. BHQ suppressed tail currents most strongly at moderate voltages (Fig. S1), which indicated a voltage-dependent mechanism of inhibition rather than a pore-blocking effect. Two-component Boltzmann analysis of the normalized tail current-voltage curves with BHQ were positively shifted with more shallow slope than controls, with BHQ enhancing the fraction (F₂) of I_Ba activating at positive voltages (Table 1, Fig. 1F). BHQ reduced tail currents to a similar extent and with similar voltage-dependence as the steady-state current (Fig. 1G), which indicated that BHQ inhibits Ca_v2.1 currents primarily by modifying voltage-dependent activation. These effects of BHQ were generally similar for Ca²⁺ currents (I_Ca; Fig. 1H–K). However, unlike I_Ba, tail currents for I_Ca were best fit by a single Boltzmann function under control conditions and exhibited maximal amplitudes at intermediate voltages. This latter feature has been attributed to Ca²⁺-dependent facilitation (CDF), which is particularly prominent for Ca_v2.1 (Chaudhuri et al., 2007). BHQ introduced a second high-threshold component to the activation curve (Fig. 1J, Table 1). These results demonstrate that BHQ inhibits voltage-dependent activation of Ca_v2.1 and has distinct actions on I_Ba and I_Ca.

Figure 1. BHQ inhibits voltage-dependent activation of Ca_v2.1.

A–C: Dose-response properties. (A) Representative I_Ba evoked by a 30-ms test pulse to +10 mV from −80 mV before (Control), during (+BHQ), and after (Washout) bath perfusion with BHQ (10 μM). (B) Same as in (A) except that cell was exposed to the indicated concentrations of BHQ (in μM, grey traces). (C) Dose-response curve for BHQ inhibition of I_Ba evoked as in (A). Smooth line represents fit by the Hill equation, n=8 cells. D–G: Effects of BHQ on I_Ba. (D) Representative I_Ba evoked by a 15-ms test pulse to +10 mV from −80 mV before (Control) and during perfusion with BHQ (+BHQ, 10 μM). (E) I–V relationship. I_Ba was evoked by 15-ms steps from −80 mV to various voltages with repolarization to −40 mV to facilitate measurement of tail currents in F. I_Ba amplitudes were normalized to cell capacitance and plotted against voltage, n=9 cells. (F) Normalized tail current-voltage plots obtained from data in E. Peak tail current amplitudes were normalized to that obtained with a +85-mV test pulse. Smooth lines represent fits by a double- Boltzmann function. (G) Voltage-dependence of BHQ inhibition of the steady-state current measured at the end of the test depolarization (Test I_Ba) and tail current measured upon repolarization to −40 mV (Tail I_Ba). Test and tail I_Ba recorded in the presence of BHQ were expressed as the percent of the corresponding values under control conditions, and plotted against test voltage. H–K: Effects of BHQ on I_Ca. Same as in D–G, except for I_Ca. In H, test pulse (to +20 mV) was 50 ms to achieve steady-state I_Ca. In J, smooth lines represent fits by a single- (for Control) or double-Boltzmann (+BHQ) function, n=7 cells. All averaged data represent mean ± SEM. See also Figure S1.

Table 1.

Parameters for voltage-dependent activation.

WT IBa n = 10	Control**	+BHQ**	p-value
Vh1 (mV)	−7.1 ± 1.1	−0.2 ± 0.8	<0.001
k1	3.9 ± 0.3	6.6 ± 0.2	<0.001
F1	0.65 ± 0.02	0.49 ± 0.01	<0.001
Vh2 (mV)	10.4 ± 1.7	26.7 ± 1.2	<0.001
k2	9.3 ± 0.5	13.3 ± 0.4	<0.001
F2	0.35 ± 0.02	0.51 ± 0.01	<0.001
WT ICa n = 7	Control*	+BHQ**	p-value
Vh1 (mV)	0.0 ± 1.6	−3.3 ± 3.1	0.21
k1	6.3 ± 0.5	5.6 ±0.1	0.17
F1	1.0	0.69 ± 0.03	0.02
Vh2 (mV)	--	28.0 ± 2.5	--
k2	--	5.8 ± 1.2	--
F2	--	0.29 ± 0.04	--
S218L IBa n = 12	Control**	+BHQ**	p-value
Vh1 (mV)	−14.5 ± 0.9	−12.5 ± 1.2	0.08
k1	5.8 ± 0.1	6.9 ± 0.2	< 0.001
F1	0.60 ± 0.01	0.42 ± 0.02	< 0.001
Vh2 (mV)	11.7 ± 1.6	23.6 ± 1.4	< 0.001
k2	12.5 ± 0.3	16.6 ± 0.5	< 0.001
F2	0.38 ± 0.01	0.56 ± 0.02	< 0.001
S218L ICa n = 7	Control*	+BHQ**	p-value
Vh1 (mV)	−6.8 ± 1.7	−8.7 ± 1.9	0.19
k1	7.2 ± 0.7	6.2 ± 0.4	0.27
F1	1.0	0.50 ± 0.02	< 0.001
Vh2 (mV)	--	17.0 ± 4.1	--
k2	--	13.8 ± 0.8	--
F2	--	0.49 ± 0.03	--

Normalized tail current- voltage were fit with either a single (*) or double (**) Boltzmann function. P-values were determined by paired t-test. V_h1, V_h2: half-maximal activation; k₁, k₂: slope factors; F₁, F₂: fractional current.

Although BHQ can raise intracellular Ca²⁺ due to SERCA blockade (Moore et al., 1987), a structurally distinct SERCA inhibitor, thapsigargin, had no effect on Ca_v2.1 properties (Fig. 2A). BHQ was capable of modestly increasing intracellular Ca²⁺ in intact HEK293T cells transfected with Ca_v2.1 as indicated by the ratiometric Ca²⁺ indicator, Fura-2 (Fig. 2B). However, under our whole-cell recording conditions that included 5 mM EGTA in the intracellular recording solution, BHQ did not produce any change in the level of intracellular Ca²⁺ (Fig. 2C). Therefore, the modulatory effects of BHQ on Ca_v2.1 are independent of SERCA antagonism.

Figure 2. Effects of BHQ on Ca_v2.1 are independent of SERCA inhibition.

(A) Same as in Fig. 1H–J except for I_Ca evoked by a 30-ms test pulse before (Control) and after (+THP) bath application of thapsigargin (10 μM). Points represent mean ± SEM, n = 10 cells. (B,C) Representative traces showing the effects of BHQ on [Ca²⁺]_iin HEK293T cells transfected with Ca_v2.1 cDNAs (n=27 cells in B, n=6 cells in C). Each trace represents an individual cell. In B, cells were loaded with Fura-2/AM and subject to Ca²⁺ imaging. In C, Ca²⁺ imaging was done on cells that were subject to whole-cell patch clamp recording conditions as in A, except that the intracellular recording solution contained 100 μM Fura-2.

In addition to inhibiting Ca_v2.1 activation, BHQ caused a prominent slowing of deactivation which was evident as a prolongation of the tail currents. For I_Ca, the dominant effects of BHQ were a significant increase in the fraction and time constant of a slow component of deactivation (Fig. 3A). To better understand the modulatory effects of BHQ on inhibiting activation and slowing deactivation, we developed a 10-state model (Fig. S2). According to the model, despite slower deactivation, decreased open probability of channels modulated by BHQ at moderate test voltage would cause the decline in steady-state current amplitude with BHQ (Fig. S2). Consistent with the model, the proportion of whole cell I_Ca that was inhibited by BHQ was identical to the fraction of the slow tail I_Ca across a range of test voltages (Fig. S2; p=1.0, by two-way ANOVA). We conclude that BHQ stabilizes the opening and weakens voltage-dependent activation of Ca_v2.1, which causes slower tail currents and a reduction in the steady- state current, respectively.

Figure 3. BHQ-modulated Ca_v2.1 channels exhibit slow deactivation that is greater for I_Ca than for I_Ba.

(A) I_Ca was evoked by 15-ms test pulses from −80 mV to various voltages with repolarization to −40 mV under control conditions or in the presence of BHQ (10 μM). Representative current traces are shown (left). For +20-mV to −40 mV step, the decay of tail currents was fit with a double exponential function. Averaged data (mean ± SEM, right) represent the time constant (τ) and fractional contribution (Fraction) of the fast and slow component. *, p < 0.01 by paired t-test. (B) Representative traces for I_Ba (left) evoked as in A. Averaged data (mean ± SEM, right) show fast and slow time constants from double exponential fits of the decay of tail I_Ba (for +10 mV to −40 mV step, n=10) and I_Ca (n=7) in the presence of BHQ. *, p = 0.03 by t-test. (C) For I_Ca and I_Ba evoked as in A,B, tail current amplitudes were measured 5 ms following the repolarization to −40 mV and plotted against test voltage. The percent increase caused by BHQ is indicated for I_Ca (at +20 mV) and I_Ba (at +10 mV). Points represent mean ± SEM. See also Figure S2.

BHQ enhances Ca²⁺-dependent facilitation induced by action potential waveforms

Further analyses revealed a significantly larger effect of BHQ on the decay of tail currents carried by Ca²⁺ than by Ba²⁺ (Fig. 3A–C). While there was no difference in the relative amounts of A_slow and A_fast for I_Ca and I_Ba, the slow time constant for tail I_Ca decay (6.7 ± 0.8 ms) was significantly greater than that for I_Ba (4.8 ± 0.4 ms, p = 0.03; Fig. 3B). The more prominent effect of BHQ on deactivation of I_Ca was particularly apparent in plots of the slow tail current amplitude measured 5 ms after repolarization to the holding voltage. At this time point, the current amplitude should be practically devoid of the contribution of the rapidly decaying tail current resulting largely from channels that are not modulated by BHQ. By this metric, BHQ had nearly double the effect in prolonging tail I_Ca compared to tail I_Ba (Fig. 3C).

The stronger effects of BHQ on deactivation of I_Ca compared to I_Ba suggested the potential for BHQ to enhance CDF of Ca_v2.1 currents during trains of action potential waveforms (APWs). Due to the short duration (half-width< 1 ms) of APWs and relatively slow activation kinetics of Ca_v2.1, APWs do not evoke significant I_Ca during the depolarizing phase but produce large tail currents upon repolarization. During trains of APWs, I_Ca undergoes both voltage-dependent facilitation (VDF) (Brody and Yue, 2000; Currie and Fox, 2002) as well as CDF due to Ca²⁺/calmodulin (Benton and Raman, 2009; Chaudhuri et al., 2007; Kreiner et al., 2010). In contrast, I_Ba exhibits primarily VDF. Because CDF and VDF occur on similar timescales, net CDF is measured as the fractional increase in the amplitude of I_Ca over I_Ba at the end of a train of APWs (F_CDF; Fig. 4A–D). With this protocol, CDF results in an increase in I_Ca (Chaudhuri et al., 2007), which may be further amplified by BHQ.

Figure 4. BHQ enhances CDF.

(A) Voltage protocol for APWs (200 Hz) and representative I_Ca and I_Ba recorded before (control) and after (+BHQ) bath perfusion with BHQ (10 μM). (B,C) Peak current amplitudes of I_Ca or I_Ba were normalized to the first in the series (I_norm.) and plotted against time. Points represent mean ± SEM. Dashed lines indicate initial I_norm. (D) CDF was measured as the difference in I_norm. for I_Ca and the mean I_norm. for I_Ba at 0.3 s (F_CDF, mean ± SEM). *, p = 0.001 by t-test. (E) The effect of BHQ was measured as the ratio of I_norm. in the presence of BHQ to the mean I_norm. under control conditions at 0.3 s (F_BHQ, mean ± SEM). For control, n=17 for I_Ca, n=8 for I_Ba; for +BHQ, n=17 for I_Ca, n=8 for I_Ba. See also Figure S3.

To test this, we analyzed the effect of BHQ on I_Ca and I_Ba during trains of APWs. To restrict analysis to CDF in the absence of the competing effects of Ca²⁺-dependent inactivation (CDI), the whole-cell intracellular recording solution contained 5 mM EGTA, which blunts CDI (Fig. S3), but spares CDF (Lee et al., 2000). With this protocol, CDF was significantly greater with BHQ (F_CDF = 0.25 ± 0.02) than without BHQ (F_CDF = 0.17 ± 0.02, p = 0.04; Fig. 4B–D). To more clearly discern the effect of BHQ on I_Ca and I_Ba, currents recorded after exposure to BHQ were normalized to the maximal amplitude of the current before bath perfusion with BHQ (i.e., at 0.3 s; F_BHQ in Fig. 4E). By this measure, BHQ inhibited I_Ba but clearly increased I_Ca (Fig. 4E). The inhibition of I_Ba was not due to an effect of BHQ on voltage-dependent inactivation since BHQ did not enhance the decay of I_Ba evoked by a sustained depolarization (Fig. S3). When CDF was prevented by co-expression of a CaM mutant (DeMaria et al., 2001; Lee et al., 2003), I_Ca was not maintained to control levels in the presence of BHQ (Fig. S3). Taken together, these results show that BHQ inhibits VDF but also increases CaM-dependent CDF during physiological stimuli.

BHQ offsets gain-of function of Ca_v2.1 due to FHM1 mutation

The S218L mutation enhances voltage-dependent activation of Ca_v2.1, which augments excitatory transmission in S218L knock-in mice (Tottene et al., 2005; van den Maagdenberg et al., 2004; van den Maagdenberg et al., 2010). However, S218L also inhibits CDF, which is associated with impaired synaptic plasticity (Adams et al., 2010). The dual effects of BHQ on inhibiting voltage-dependent activation and enhancing CDF suggested it could oppose dysfunction associated with the S218L FHM1 mutation in α₁2.1. To test this, we analyzed the effect of BHQ on I_Ca and I_Ba in HEK293T cells transfected with S218L-containing channels. As shown previously (Tottene et al., 2005), I_Ba for S218L cells exhibits more negative voltage- dependence of activation than wild-type (WT) Ca_v2.1 (Table 1; Fig. 5A,B). Reductions in current density seen with S218L compared to WT (Fig. 5A) may be an artifact of heterologous overexpression, as this is not observed for Ca_v2.1 currents in neurons from knock-in mice expressing S218L channels (Pietrobon, 2010). BHQ had similar effects on S218L as on WT channels in inhibiting current density and voltage-dependent activation of I_Ba and I_Ca (Fig. 5C–H, Table 1). Across a physiological voltage range (−30 mV to −10 mV), BHQ offset by 25–44% the enhancement of I_Ba caused by the S218L mutation (Fig. S4). Therefore, BHQ reduces the gain-of function in Ca_v2.1 caused by S218L.

Figure 5. BHQ opposes enhanced activation of S218L.

(A,B) I–V (A) and normalized tail I_Ba-voltage (B) relationships for cells transfected with S218L. Same voltage protocols as in Fig. 1E,F. (C–H) Same as in Fig. 1D–F, H–J except in cells transfected with S218L. See also Figure S4. Points represent mean ± SEM; n = 12 for control I_Ba, n= 10 for I_Ca.

BHQ restores CDF loss-of function due to FHM1 mutation

Similar to its effects on WT channels, BHQ slowed deactivation of S218L (Fig. 6A,B). However, this effect of BHQ was greater for S218L than WT (Fig. 6C,D). Compared to WT channels, BHQ caused a more dramatic shift in the balance of fast and slow fractions of S218L tail I_Ca decay, such that total decay was dominated by the slow fraction (0.58 ± 0.02 for S218L vs. 0.42 ± 0.05 for WT; p=0.01 by t-test; Fig. 6D). The slow fraction of I_Ca decay under control conditions was twice as large for S218L compared to WT (0.16 ± 0.03 for S218L vs. 0.08±0.01 for WT, p = 0.03; Fig. 6D). BHQ also increased the slow time constant of I_Ca decay for S218L (τ_slow = 3.9 ± 0.8 ms for control vs. 10.2 ± 0.7 ms for BHQ) as it did for WT channels (Fig. 6B,D). These results suggested that BHQ stabilizes a fraction of slowly deactivating channels, which are more prevalent in S218L than WT.

Figure 6. BHQ slows deactivation of S218L more than WT Ca_v2.1.

(A) Representative I_Ca and I_Ba obtained from cells transfected with S218L. Currents were evoked by 15-ms steps to various voltages from −80 mV before (control) and after exposure to BHQ (10 μM). (B) For +10- mV(I _Ca) and +0-mV (I_Ba) test pulses, the decay of tail currents evoked upon repolarization to −40 mV was fit with a double exponential function. Averaged data (mean ± SEM) represent the time constant (τ) and fractional contribution (Fraction) of the fast and slow component. *, p < 0.01 by paired t-test, n = 8 for I_Ca, n = 12 for I_Ba. (C,D) Larger effects of BHQ on deactivation of S218L compared to WT channels. Representative traces (C) show I_Ca evoked by +20-mV (WT) and +10-mV (S218L) test pulses with repolarization to −40 mV before (control) and after exposure to BHQ (10 μM). In D, τ_slow and F_slow were obtained by double exponential fits of tail I_Ca obtained as in C. Data represent mean ± SEM. *, p < 0.05 by t-test, n = 7 for WT, n=8 for S218L.

To determine if BHQ could restore CDF to S218L, we compared I_Ca and I_Ba evoked by trains of AP waveforms (Fig. 7A–C). Because the S218L mutation accelerates fast voltage- dependent inactivation of I_Ca and I_Ba (Tottene et al., 2005; Weiss et al., 2008), I_Ba progressively decreased during the train (Fig. 7B). For this reason, we evaluated I_Ca and I_Ba during 100-ms trains as opposed to the 300-ms trains used for WT channels (Fig. 4). As shown previously (Adams et al., 2010), S218L did not undergo as much CDF as WT (F_CDF = 0.09 ± 0.02 for S218L vs. 0.17 ± 0.01 for WT, p=0.005 by t-test; Fig. 4D, 7D). BHQ had a larger effect in enhancing I_Ca at the end of the train in S218L compared to WT channels (F_BHQ = 1.06 ± 0.01 for S218L at 0.1 s vs. 1.02 ± 0.01 for WT at 0.3 s, p = 0.03 by t-test; Fig. 4E,7E,) and increased CDF for S218L (F_CDF= 0.15 ± 0.02) such that it was not different from WT (F_CDF = 0.17 ± 0.01; p = 0.63; Fig. 4D,7D). Taken together, our findings show that BHQ opposes both the enhanced activation and limited CDF of S218L in transfected HEK293T cells.

Figure 7. BHQ enhances CDF in S218L.

(A–E) Representative I_Ca and I_Ba obtained by APWs and analysis as in Fig. 4A–E but in cells transfected with S218L. n=10 for I_Ca, n=8 for I_Ba; for +BHQ. *, p = 0.003 by paired t-test. In B,C, dashed lines indicate level of I_norm. for I_Ca of WT channels (Control and +BHQ) in Fig. 4B,C. In D, *, p = 0.005 by t-test. Data in B–E represent mean ± SEM.

BHQ offsets gain-of function in synaptic transmission due to S218L

To determine if BHQ could offset the neurophysiological deficits associated with S218L, we turned to the Drosophila neuromuscular junction as an in vivo model. Here, as in mammals, Ca_v2.1 channels mediate presynaptic Ca²⁺ influx that triggers neurotransmitter release (Kawasaki et al., 2000; Macleod et al., 2006; Uchitel et al., 1992). Subsequent activation of postsynaptic ionotropic glutamate receptors depolarizes the innervated muscle membrane. The resulting excitatory postsynaptic potentials (EPSPs) scale with the amount of neurotransmitter release and promote muscle contraction. Based on previous reports of enhanced transmission at neuromuscular synapses of S218L knock-in mice (Kaja et al., 2010), we predicted a similar impact of S218L in Drosophila.

To test this, the mutation analogous to S218L was introduced into the Drosophila orthologue of CACNA1A, cacophony, and transgenically expressed in the Drosophila nervous system (“S218L”). The C155-GAL4 (elaV) neuronal driver allowed expression of the transgene in the presynaptic motor neuron but not muscle (Lin and Goodman, 1994). Compared to controls expressing the wild-type transgene (“WT”), the amplitude of EPSPs was significantly greater in NMJs expressing the mutant channel (“S218L”; Fig. 8A). This result was not due to increased numbers of synaptic boutons in “S218L” vs. “WT”, which were not significantly different between genotypes (Table 2). Analyses of quantal content corrected for non-linear summation and plotted against different extracellular Ca²⁺ concentrations revealed nearly identical Ca²⁺-cooperativity slope values for “WT” and “S218L” (2.65 and 2.61, respectively). Thus, increased EPSP amplitudes in “S218L” are likely due to greater presynaptic Ca²⁺ currents mediated by “S218L” rather than increases in the Ca²⁺ cooperativity of neurotransmitter release.

Figure 8. BHQ reverses gain-of function in synaptic transmission due to S218L.

EPSPs were recorded from muscles of transgenic flies expressing the wild-type cacophony(“WT”) or the channel with mutation corresponding to S218L (“S218L”). Extracellular solution contained 0.4 mM Ca²⁺. (A) Representative traces for averaged evoked EPSPs (left) and the mean ± SEM for EPSP amplitudes. *, p = 0.013 by t-test. (B) Left, Representative averaged EPSPs overlaid for comparison in the presence (5 or 10 μM) or absence of BHQ (control). Right, mean ± SEM for EPSP amplitudes. (C) mEPSP amplitudes measured in “WT” and “S218L” with the indicated concentrations of BHQ. Data represent mean ± SEM. In A- C, parentheses indicate numbers of cells. In B,C *, p < 0.05 compared to control (no BHQ) by one-way ANOVA and post-hoc Bonferroni t-test.

Table 2.

Synaptic bouton counts in “WT” and “S218L” flies.

Genotype	Muscle A2 6/7	Muscle A3 6/7
“WT”	97.6 ± 3.2 (n=19)	79.7 ± 4.3 (n=19)
“S218L”	93.0 ± 3.2 (n=27) p = 0.31	74.5 ± 2.0 (n=26) p = 0.28

Drosophila larval NMJ filet preparations were double-labeled with anti-synapsin and anti-Dlg primary antibodies as described in Experimental Procedures. Synapses for muscle group 6/7 in segments A2 and A3 were examined by immunofluorescence, and synaptic boutons per NMJ 6/7 were counted. For both muscle segments examined, presynaptic expression of “S218L” has no significant effect on synapse growth by Student’s t-test (compared to “WT”, p-values indicated). Numbers represent synapse counts (mean ± S.E.M.; n = number of synapses counted per genotype).

At 5 μM, BHQ caused nearly ~50% inhibition of EPSP amplitudes in “WT” NMJs but did not affect miniature EPSP (mEPSP) amplitudes (Fig. 8B,C). These results were consistent with a presynaptic action of BHQ, most likely due to inhibition of Ca_v2.1 activation (Fig. 1). However, this concentration of BHQ had no effect on EPSP amplitudes in “S218L” NMJs. At 10 μM, BHQ strongly inhibited EPSP amplitudes, but also affected mEPSPs in both genotypes (Fig. 8B,C). This result suggested an additional effect of BHQ at higher concentrations on spontaneous release properties. Therefore, BHQ at 10 μM reversed the gain-of function in synaptic transmission in “S2l8L” NMJs, although this effect could not be attributed to actions on Ca_v2.1 alone.

BHQ restores synaptic plasticity defect due to S218L

Short-term facilitation (STF) of synaptic transmission during bursts of impulses is a form of synaptic plasticity that can be driven by the accumulation of presynaptic Ca²⁺ (Zucker and Regehr, 2002). At cerebellar synapses of S218L knock-in mice, STF is impaired, which has been attributed to weak CDF of S218L channels (Adams et al., 2009). However, a caveat is that basal release probability is inversely related to the magnitude of STF (Zucker and Regehr, 2002). Therefore, it is also possible that the gain-of function of S218L channels would increase basal synaptic transmission, which would subsequently suppress STF in S218L knock-in mice, independent of any changes in CDF (Adams et al., 2009). The fact that BHQ at 5 μM increased CDF of S218L channels in HEK293T cells (Fig. 7), but did not affect evoked release in “S218L” NMJs (Fig. 8B,C), provided an opportunity to determine if weak CDF contributes to the synaptic plasticity defect due to the S218L mutation. If so, this concentration of BHQ should increase STF of “S218L” NMJs.

To test this prediction, we analyzed trains of EPSPs evoked by repetitive (10 Hz) stimuli of presynaptic motor neurons. To improve resolution of STF, we initially maintained extracellular Ca²⁺ concentrations at 0.4 mM so as not to saturate the release machinery and create problems with non-linear summation. STF was measured as the amplitude of the evoked EPSPs normalized to that for the initial EPSP of the train (F_ratio). Compared to in”WT” NMJs (maximal F_ratio at ~1.1 s (F_ratio,max) = 1.22 ± 0.09), EPSPs showed significantly less STF in “S218L” – expressing NMJs (F_ratio,max = 0.99 ± 0.04; p = 0.01 by t-test; Fig. 9A,B). These results confirm that the synaptic defects that are associated with the S218L mutation (Kaja et al., 2010; van den Maagdenberg et al., 2010) are reproduced by the “S218L” at the DrosophilaNMJ. BHQ (5 μM) essentially restored STF in “S218L” NMJs: there was no significant difference in STF of “WT” controls (F_ratio,max =1.22 ± 0.09) and “S218L” with BHQ (F_ratio,max = 1.20 ± 0.10, p = 0.70; Fig. 9A–C). This result is consistent with the effects of BHQ in enhancing CDF of S218L in HEK293T cells (Fig. 7).

Figure 9. BHQ (5.

μM) restores STF to S218L.

(A) Traces represent averaged EPSPs evoked by 10 Hz stimulation. Vertical scales bars represent 20 mV each. (B) STF measured as averaged EPSP amplitudes normalized to the first in the series (F_ratio) and plotted against time with extracellular [Ca²⁺] at 0.4 mM (left) or 0.2 mM (right). Points represent mean ± SEM. (C) For data in B, STF at 1 s (F_{ratio,1 s}, mean ± SEM) is compared. p-values shown were determined by Mann-Whitney rank sum test. Parentheses indicate numbers of cells. (D) At the indicated concentrations of extracellular Ca²⁺, basal evoked EPSP amplitudes (mean ± SEM) were compared for control and +BHQ in “WT” (left) and “S218L” (right). *, p < 0.05 by t-test. (E) Same as in B except that extracellular Ca²⁺ was 1.5 mM. *, p < 0.05 by t-test. See also Figure S5.

To rule out that these effects were due to effects of BHQ on decreasing release probability, we performed similar experiments but with a reduced extracellular concentration of Ca²⁺ (0.2 mM, Fig. 9B,C). Similar to our results obtained with 0.4 mM Ca²⁺, BHQ did not affect the amplitude of the evoked EPSP in “S218L” NMJs (5.5 ± 0.8 mV for control vs. 5.4 ± 1.2 mV for +BHQ; p = 0.97; Fig. 9D), but still restored STF to the level of the “WT” control in 0.2 mM Ca²⁺ (Fig. 9B,C). This result argues against the likelihood that BHQ restored STF to “S218L” simply by inhibiting presynaptic Ca²⁺ channels and/or release properties. BHQ did seem to enhance the Ca²⁺cooperativity of release for “S218L”, but after normalizing for this effect, BHQ still increased STF to “WT” control levels (Fig. S5). At a physiological extracellular Ca²⁺ concentration (1.5 mM), we observed short-term depression (STD), as opposed to STF for both “WT” and “S218L”. Despite this difference, BHQ significantly abated STD at early time points without affecting evoked EPSP amplitudes (Fig. 9D,E), consistent with the STF data at lower levels of external Ca²⁺. Collectively, our findings support the argument that by enhancing CDF, BHQ can promote normal synaptic plasticity in the context of the S218L mutation.

DISCUSSION

BHQ is a Ca_v2.1 gating modifier

Divalent cations, toxins, and small-molecules have yielded rich insights into the structure, function, and physiological roles of Ca_v channels in excitable cells. Like gating modifiers such as agatoxin IVA (McDonough et al., 1997), BHQ inhibits the voltage-dependence of Ca_v2.1 activation in a manner that can account for the decrease in the steady-state current (Fig. 1D–K). While agatoxin IVA speeds deactivation of neuronal Ca_v2.1 channels (McDonough et al., 1997), BHQ has the opposite effect (Fig. 3). In this regard, BHQ is similar to roscovitine – an inhibitor of cyclin-dependent kinases that also slows deactivation of Ca_v2.1 and Ca_v2.2 channels (Buraei et al., 2005; Yan et al., 2002). At higher concentrations that involve a kinase-dependent mechanism, roscovitine also inhibits Ca_v2.2 (Buraei et al., 2005). In contrast, our results indicate that the opposing effects of BHQ on activation and deactivation are mechanistically linked. These modulatory effects may have been missed in previous studies due to the use of high concentrations of BHQ (> 100 μM; (Fusi et al., 2001; Scamps et al., 2000)) that strongly depress current amplitudes (Fig. 1B). The dual agonist and antagonist actions of BHQ are unique among substances that modulate Ca_v2.1, and should prove useful in mechanistic inquiries into the complex gating properties of these channels.

Modulation of CDF by BHQ: insights into a novel aspect of S218L gain-of function

During trains of APWs, I_Ca in HEK293T cells transfected with Ca_v2.1 may undergo VDF, perhaps due to relief from G-protein inhibition (Brody et al., 1997), and CDF (Chaudhuri et al., 2007) due to Ca²⁺/CaM binding to the C-terminal domain of the Ca_v2.1 α₁ subunit (DeMaria et al., 2001; Lee et al., 2000; Lee et al., 1999). Elementary analyses indicate that both Ca²⁺/CaM and G-proteins may stimulate transitions to a gating mode of higher open probability (P_o), but via distinct biophysical pathways (Chaudhuri et al., 2007; Colecraft et al., 2001). Assuming that VDF and CDF are mechanistically separable gating modes, the inhibition of VDF by BHQ may increase the availability of WT Ca_v2.1 channels to undergo CDF (Fig. 4E). An alternate, although not mutually exclusive, explanation is suggested by the effects of BHQ on the S218L mutant. It has been proposed that the hyperpolarized shift in the voltage-dependence of activation of S218L may cause these channels to be “pre-facilitated”, such that they are not sensitive to Ca²⁺/CaM-stimulated transitions to the high P_o gating mode (Adams et al., 2010). By inhibiting voltage-dependent activation (Fig. 5H; Table 1), BHQ may return a fraction of the mutant channels to the normal low P_o gating mode so they can undergo CDF. An understanding of the mechanistic basis for how BHQ modulates CDF awaits detailed single-channel analyses of I_Ca and I_Ba in WT and S218L channels.

In the absence of BHQ, we identified a slow component of deactivation that was significantly greater for I_Ca mediated by S218L compared to WT channels (Fig. 6D). Because CDF is associated with a broadening of tail I_Ca during APW trains (Chaudhuri et al., 2007), this difference may arise from the “pre-facilitated” S218L channels. Although not noted previously, the ~50% increase in the slow tail I_Ca in S218L compared to WT channels (Fig. 6D) would augment AP-driven neurotransmitter release and subsequently accelerate the development of synaptic depression. Indeed, compared to WT, EPSPs at the NMJ in S218L KI mice are of greater amplitude and show faster run-down at physiological firing frequency (Kaja et al., 2010). Moreover, EPSPs at S218L KI NMJ are broader than in WT and often exhibit an irregular plateau during the falling phase (Kaja et al., 2010). These results could be explained by the increased slow deactivation of S218L channels, which may cause abnormal prolongation of neurotransmitter release. Similar alterations in S218L deactivation at central synapses may contribute to the increased excitatory transmission and lower thresholds for cortical spreading depression that is thought to underlie the headache symptoms associated with migraine (van den Maagdenberg et al., 2010).

Dysregulation of CDF and impaired synaptic plasticity caused by S218L

CDF of presynaptic Ca_v2.1 channels has emerged as an important determinant of STF (Mochida et al., 2008). At the parallel fiber – Purkinje cell synapse, STF is robust in WT but not in S218L knock-in mice (Adams et al., 2010), which could be explained by reduced net CDF caused by the S218L mutation. However, a complicating factor is that increased Ca²⁺ influx in parallel fiber terminals due to “pre-facilitated” channels (Adams et al., 2010) would likely saturate the Ca²⁺ sensor for exocytosis in S218L knock-in mice. This would decrease release probability, which itself would reduce STF. Our findings that BHQ not only restores CDF of S218L channels in HEK293T cells (Fig. 7) but also STF at “S218L” NMJs (Fig. 9) indicate that dysregulation of CDF contributes to the impaired synaptic plasticity caused by the S218L mutation. Given the localization of Ca_v2.1 channels at most central synapses (Westenbroek et al., 1995), our results predict widespread deficits in short-term plasticity caused by the S218L mutation. Whether such alterations contribute to the neurological phenotypes associated with FHM-1 remains an important challenge for future studies.

Therapeutic implications for Ca_v2.1 channelopathies

The CACNA1A gene encoding Ca_v2.1 channels is a major target of mutations that cause severe neurological disorders in humans (Pietrobon, 2010). Significant advances have been made in understanding the pathogenic mechanisms associated with some of the mutations, thanks to the generation of appropriate knock-in mice (van den Maagdenberg et al., 2004; van den Maagdenberg et al., 2010; Watase et al., 2008). Our study highlights the use of Drosophila as a novel alternate strategy to investigate the synaptic defects caused by CACNA1A mutations. The gain-of function in EPSPs and impaired STF at the NMJs of “S218L”- expressing flies (Figs. 8,9) are similar to that seen at the NMJs and parallel fiber-Purkinje cell synapses, respectively, of S218L knock-in mice (Adams et al., 2010; Kaja et al., 2010). Moreover, BHQ regulates synaptic transmission at “WT” and “S218L” NMJs in a manner consistent with its modulation of the corresponding channels in HEK293T cells. These results indicate that the incorporation of human mutations into the cacophony gene, and expression of the mutant construct in Drosophila motor neurons, can simulate the functional abnormalities associated with Ca_v2.1 channelopathies. This approach should prove useful in the search for pharmacological and genetic modifiers that can ameliorate the often severe neurophysiological consequences of CACNA1A mutations.

Despite the known activity of BHQ on SERCAs and Ca_v channels (Fusi et al., 2001; Moore et al., 1987; Scamps et al., 2000), the effects of BHQ on synaptic transmission at the Drosophila NMJ could be explained largely by its modulation of Ca_v2.1. Due to its pro-oxidant effects and actions on targets other than Ca_v2.1, BHQ itself may not be therapeutically suitable as a specific gating modifier for neuronal Ca_v2.1 channels. However, the ability of BHQ to oppose the defects in Ca_v2.1 gating and synaptic transmission associated with the S218L mutation suggests that modification of the chemical structure of BHQ may yield small-molecule modulators that could selectively inhibit or enhance activation/ deactivation gating of Ca_v2.1. Such compounds would greatly extend current capabilities to dissect the functions of Ca_v2.1 channels in normal and diseased states of the nervous system.

EXPERIMENTAL PROCEDURES

Cell culture and transfection

The following Ca_v subunit cDNAs were used: human WT α₁2.1 (Genbank no. NM_023035.1) or S218L mutant in pcDNA 3.1 Zeo (+) (generously provided by Dr. T. Snutch, described in (Adams et al., 2010)), rat β_2a(Genbank no. NM053581), and rabbit α₂δ₁ (Genbank no. M21948). Human embryonic kidney cells transformed with SV40 T-antigen (HEK293T) were maintained in DMEM with 10% fetal bovine serum (Life Technologies), penicillin (50 U/ml), and streptomycin (50 μg/ml ) at 37º C in a humidified atmosphere with 5% CO₂. Cells were grown to ~80% confluency and transfected with Fugene 6 (Roche) according to the manufacturer’s protocol. Cells were transfected with cDNAs encoding WT-α₁2.1 or S218L- α₁2.1 (1 μg), β_2a (0.5 μg), and α₂δ₁(0.5 μg) with EGFP -N1 (Life Technologies, 0.1 μg).

Electrophysiological recordings of transfected HEK293T cells

Ca²⁺ or Ba²⁺ currents were recorded in transfected HEK293T cells in whole cell patch-clamp configuration at room temperature. Recordings were performed between 48 to 72 hours after transfection. Extracellular recording solutions contained (in mM): 150 Tris, 1 MgCl₂, and 10 CaCl₂ or 10 BaCl₂. Intracellular recording solutions contained (in mM): 140 N-methyl-D- glucamine, 10 HEPES, 2 MgCl₂, 2 Mg-ATP, and 5 EGTA. The pH of extracellular and intracellular recording solutions were adjusted to 7.3 with methanesulfonic acid. Reagents used for electrophysiological recordings were obtained from Sigma – Aldrich (St. Louis, MO). 1,4-Dihydroxy-2,5-di-tert-butylbenzene (BHQ, Tocris bioscience) was dissolved in DMSO at a concentration of 100 mM to make a stock solution. BHQ was diluted with the extracellular solution immediately before recording. For each cell, voltage protocols were run first under control conditions (0.01% DMSO) followed by extracellular perfusion with BHQ and finally upon washout of BHQ with the control solution. Data from cells in which the effect of BHQ was reversible were included in the analysis.

Recording electrodes were pulled from borosilicate glass (World Precision Instruments, Sarasota, FL) on a Sutter P97 puller (Sutter Instruments, Novato, CA). Electrode resistances in the recording solutions were typically 2–4 MΩ. Series resistance ranged from 4–6 MΩ and compensated at 50–80%. Currents were recorded with an EPC-10 patch-clamp amplifier driven by Patch-Master software (HEKA Electronics, Lambrecht/Pfalz, Germany). Leak current was subtracted using a P/8 protocol. Whole-cell currents were digitized at 50 kHz after filtering at 10 kHz. Action potential (AP) waveforms were constructed by averaging traces corresponding to spontaneous APs recorded in current clamp from dissociated mouse Purkinje neurons (Kreiner et al., 2010). Data were analyzed using routines written in IGOR Pro software (WaveMetrics, Lake Oswego, OR). Averaged data represent the mean ± SEM. Statistical differences between groups were determined by Student’s t-test unless otherwise indicated. Tail current-voltage relationships were fit with a single (I /{1 + exp[ − (V − V_h)/k] + b}) or double Boltzmann function (I₁/{1 + exp[− (V − V_h1)/k₁] + I₂/{1 + exp[− (V − V_h2)/k₂] + b}), where I is the maximal current, V is the test voltage, V_h is the voltage of half-activation, k is the slope factor, and b is the baseline. Due to the complicating effects of CDF, tail I_Ca-voltage curves under control conditions were fit through the maximum I_Ca (usually obtained at +30–40 mV). Tail current decay kinetics were with a single (y₀+[A₁ * exp(− (t–t₀)/τ₁)]) or a double exponential function (y₀+[A₁ * exp(− (t–t₀)/τ₁)] + [A₂*exp(− (t–t₀)/τ₂)]), where y₀ is the baseline, t is time, t₀ is initial time, A is amplitude. The decay was fit over an interval long enough for the current to reach steady state in control and in the presence of BHQ. In comparisons of I_Ca and I_Ba evoked by single test pulses, results are compared at different voltages to account for the differences in voltage-dependent activation due to distinct surface charge screening effects of Ca²⁺ and Ba²⁺ (see V_h, Table 1).

Measurements of intracellular Ca²⁺ concentration ([Ca²⁺]_i)

Transfected HEK293T cells were either incubated with Fura-2 AM or loaded with Fura-2 (Invitrogen, Carlsbad, CA) via the patch pipette (100 μM), which also contained the intracellular recording solution described for electrophysiological recordings. Cells were placed in a flow-through chamber mounted on the stage of an inverted IX-71 microscope (Olympus, Japan). Fluorescence was alternately excited at 340 (12 nm bandpass) and 380 (12 nm bandpass) nm using the Polychrome IV monochromator (TILL Photonics, Germany) via a 40x oil-immersion objective (NA=1.35, Olympus, Japan). Emitted fluorescence was collected at 510 (80) nm using an IMAGO CCD camera (TILL Photonics, Germany). Pairs of 340/380 nm images were sampled at 10 Hz. Fluorescence was corrected for background, as determined in an area that did not contain a cell. Data were processed using TILLvisION 4.0.1.2 (TILL Photonics, Martinsried, Germany) and presented as a fluorescence ratio of F₃₄₀/F₃₈₀, where F₃₄₀ and F₃₈₀ are fluorescence intensities at the excitation wavelengths 340 and 380 nm, respectively.

Electrophysiological recordings of Drosophila NMJ

The cacophony transgenic line UAS-cac-EGFP^786c (Kawasaki et al., 2004) was expressed under control of the C155-GAL4 (elaV) driver (Lin and Goodman, 1994) as the “WT” condition. For the “S218L” condition, we used PCR to mutate the serine161 codon in the UAS-cac-EGFP cDNAanalogous to serine218 of CACNA1A (Kawasaki et al., 2004). The resulting cDNA was used to generate a UAS-cac-EGFP^S/L transgene and expressed using the same C155-GAL4 driver.

For electrophysiological recordings of NMJs, wandering third instar larvae were selected for analysis. Sharp electrode recordings were taken from muscle 6 of abdominal segments 2 and 3, as previously described (Davis et al., 1998; Frank et al., 2006; Frank et al., 2009). Larvae were dissected in a modified HL3 saline with the following components (and concentrations): NaCl (70 mM), KCl (5 mM), MgCl₂ (10 mM), NaHCO₃ (10 mM), sucrose (115 mM = 3.9%), trehalose (4.2 mM = 0.16%), HEPES (5.0 mM = 0.12%), and CaCl₂ (0.4 mM, unless otherwise indicated). Data were collected using an Axopatch 200B electrode amplifier (Molecular Devices, Sunnyvale, CA), digitized using a Digidata 1440A data acquisition system (Molecular Devices), and recorded with pCLAMP 10 acquisition software (Molecular Devices). For presynaptic nerve stimulation, a Master-8 pulse stimulator (A.M.P. Instruments, Jerusalem, Israel) and an ISO-Flex isolation unit (A.M.P. Instruments) were utilized to deliver stimuli to the nerve (1-ms stimulus duration).

A custom function was written in Igor Pro for amplitude analysis of trains of EPSPs evoked at the NMJ. For each EPSP, a single exponential was used to fit the decay of the EPSP peak to the start of the next EPSP. The exponential decay was then extrapolated to the time point of the subsequent peak response. This point was used as the baseline for peak amplitude measurements of EPSPs 2 – 50. The first EPSP peak amplitude was simply the baseline measured before the train subtracted from the first peak. Two to four trains were evoked per muscle. The EPSP amplitudes measured by the above procedure were then averaged. These averaged amplitudes were then normalized to the first averaged EPSP amplitude to yield F_ratio values for each muscle such that the standard error for F_ratio plotted in Fig. 9 represents the variability between muscles. Averaged data represent mean ± SEM from 4–6 animals per genotype.

Neuromuscular junction analyses by immunocytochemistry

Dissected third instar larvae were fixed for 2–4 min in Bouin’s fixative (Ricca Chemical Company or Sigma-Aldrich), washed with PBS and PBT, and incubated overnight at 4 C with primary antibodies. After three washes in PBT, secondary antibodies were applied for 2 hours at room temperature. The following primary antibodies were used: mouse anti-Synapsin, 1:50 (Developmental Studies Hybridoma Bank, University of Iowa; (Klagges et al., 1996)), mouse anti-Bruchpilot, 1:250 (DSHB; Wagh et al., 2006), rabbit anti-GluRIIC, 1:3000 (Marrus and DiAntonio, 2004), and rabbit anti-Dlg, 1:3000 (Budnik et al., 1996). Fluorophore-conjugated goat anti-mouse and goat anti-rabbit secondary antibodies were used (Jackson Immunoresearch Laboratories, Inc.).

HIGHLIGHTS.

• BHQ inhibits voltage-dependent activation and slows deactivation of Ca_v2.1.

• BHQ enhances Ca²⁺-dependent facilitation of Ca_v2.1.

• BHQ ameliorates defects in Ca_v2.1 gating and synaptic transmission associated with S218L.