The Headache of a Hyperactive Calcium Channel

Abstract

Migraine is thought to be triggered by excessive neocortical neuronal excitability that leads to cortical spreading depression. In this issue of Neuron, Tottene et al. study a mouse model of familial hemiplegic migraine type 1, and provide evidence for the hyperactivity of P/Q-type calcium channel-mediated cortical glutamatergic synaptic transmission as an underlying mechanism for the susceptibility of cortical spreading depression initiation in migraine disorders.

Migraine is a common neurological syndrome characterized by episodic disabling headache pain that is often associated with a preceding aura (migraine with aura). Functional neuroimaging studies observe cortical hemodynamic changes in migraine with aura patients that resemble those observed during cortical spreading depression (CSD), indicating that CSD may cause migraine aura and trigger migraine headache (Sanchez del Rio and Alvarez Linera, 2004). CSD is a wave of neuronal excitation that slowly propagates across the cortex, causing brief intense spike trains followed by long-lasting suppression of neuronal activities (Smith et al., 2006). Are these waves causal for migraine and how are they generated in the first place? Tottene et al. (2009) in this issue of Neuron provide evidence that a potentiation of glutamate release may alter the excitation and inhibition balance in neuronal networks that could contribute to the vulnerability of CSD initiation in migraine disorders.

Familial hemiplegic migraine type 1 (FHM1), a rare autosomal dominant subtype of migraine with aura, is caused by a series of missense mutations in CACNA1A gene that encodes the poreforming subunit (α₁ subunit) of voltage-gated P/Q-type (Ca_v2.1) calcium channel (Ophoff et al., 1996). It seems very likely that FHM1 mutations may alter synaptic transmission because of the prominent location of these channels at the presynaptic termini and their crucial role in controlling neurotransmitter release. Indeed, when one of these mutations (R192Q) was introduced into the mouse-endogenous cacna1a gene to create a knockin mouse model of FHM1, the mutant mice show increased P/Q-type calcium current density in cerebellar granule cells and enhanced neurotransmitter release at neuromuscular junctions (Kaja et al., 2005; van den Maagdenberg et al., 2004). More importantly, the R192Q mutant mice show reduced induction threshold and increased velocity of CSD (van den Maagdenberg et al., 2004), but how this gain-of-function mutation contributes to the vulnerability of CSD initiation in migraine remained unclear.

Tottene et al. (2009) now take advantage of autaptic cortical cultures to clearly show an enhanced excitatory synaptic transmission in R192Q mutant mice, and demonstrate that this enhancement is due to a specific enhancement of P/Q-type calcium current and an increase in release probability without alteration of other synaptic parameters. This conclusion is further strengthened by paired recordings of layer 2/3 pyramidal cells and fast-spiking interneurons in acute cortical slices, in which Tottene et al. show that glutamatergic connections, but surprisingly not GABAergic connections, are enhanced in mutant mice. These results allow the authors to develop a plausible model in which specific enhancement of excitatory drive by R192Q mutation in Ca_v2.1 alters the balance between excitation and inhibition during cortical activity, which in turn facilitates the induction and propagation of CSD. Although Tottene et al. (2009) do not directly show that the excitatory inputs onto pyramidal cells are increased and pyramidal cells are therefore hyperactive, they do present a convincing and elegant experiment linking the enhancement of P/Q-type calcium channel function to the facilitation of CSD. They show that application of a subsaturating concentration of ω-AgalVA (a specific P/Q-type calcium channel blocker) that reduces glutamate release at mutant synapses to the levels of wild-type synapses completely restores the facilitation of CSD induction and propagation in cortical slices from the mutant mice.

In addition to the frank upregulation of excitatory transmission by the R192Q mutation, other mechanisms may also contribute to the hyperactivity of cortical circuits. The increased P/Q-type calcium current results in a relative shift in the dependence of neurotransmission on P/Q-, N- and R-type calcium channels toward P/Q-type channels. This shift in the presynaptic calcium entry pathway could decrease G protein inhibition of neurotransmitter release because P/Q-type channels undergo less modulation than N-type channels (Zamponi, 2001). For instance, during cortical activity activation of presynaptic GABA_B receptors at excitatory synapses may cause less inhibition of glutamate release in R192Q mutant mice, leading to more excitation of neural network.

The findings of Tottene and colleagues contribute novel insights into our understanding of the cellular and synaptic mechanisms of CSD and migraine, and highlight how good basic science research may point out the direction in which to search for novel therapies. Meanwhile, these exciting results also raise several interesting and even puzzling questions for future studies. The first seemly naive but crucial question is whether the R192Q mutant mice actually exhibit “migraine,” or whether the trigeminal nociception is altered in these mice. So far no behavioral abnormalities in these mice have been reported, but it will be much more useful for therapeutic intervention to identify behavioral migraine phenotypes in this FHM1 model. It will be also interesting to examine the cellular and behavioral phenotypes in the R192Q heterozygous mice since FHM1 is a dominant inherited disorder.

Both loss-of-function and gain-of-function phenotypes in P/Q-type calcium channel function and synaptic signaling have been reported for the FHM1 mutations in various experimental conditions (Barrett et al., 2005; Cao et al., 2004; Cao and Tsien, 2005; Hans et al., 1999; Kraus et al., 1998; Tottene et al., 2002, 2009; van den Maagdenberg et al., 2004). Particularly, rescue experiments using a human Ca_v2.1 splice variant in hippocampal neurons from cacna1a knockout mice show that several FHM1 mutations including R192Q decrease P/Q-type calcium current and its contribution to neurotransmitter release. However, the overall synaptic strength is well maintained because N-type calcium channels compensate for the defective P/Q-type calcium channels (Cao et al., 2004; Cao and Tsien, 2005). One should not take this controversy lightly because the exact functional consequences of these mutations will determine the directions in which we should look for the potential therapies. One possibility suggested by Tottene et al. (2009) is that although R192Q mutation induces gain-of-function changes in Ca_v2.1 function, insufficient surface expression of mutant Ca_v2.1 in transfected neurons leads to apparent loss-of-function phenotypes. However, this explanation cannot readily account for the dominant-negative effect of R192Q when it is overexpressed in wild-type neurons (Cao et al., 2004; Cao and Tsien, 2005). Another possible explanation suggested by Cao and colleagues is the difference between species of Ca_v2.1 used in different studies because, for example, Ca_v2.1 of rabbit and human are differentially affected by FHM1 mutations and trinucleotide expansions that cause spinocerebellar ataxia type 6 (Cao et al., 2004). It should be straightforward to test this idea by comparing wild-type and mutant Ca_v2.1 from different species in cacna1a knockout neurons. Regardless of the outcome of this comparison, the ultimate question is what happens in FHM1 patients and what is a good model for FHM1. A subgroup of migraineurs with aura exhibits abnormal neuromuscular transmission, supporting the loss-of-function in P/Q-type calcium channel model (Ambrosini et al., 2001), but it is not clear whether these patients carry FHM1 mutations. In fact, abnormal neuromuscular transmission is not detected in some FHM1 patients (Terwindt et al., 2004). In this context, whether the R192Q mutant mice display migraine is vital for future research. Additional transgenic mice carrying other FHM1 mutations will also be helpful and valuable for reconciling the controversy, understanding the mechanism of migraine, and testing potential therapeutic strategies.

Another intriguing observation that Tottene and colleagues made in this study is that the R192Q mutation does not affect synaptic transmission at the GABAergic synapses from fast-spiking interneurons to pyramidal cells, even though P/Q-type calcium currents exclusively mediate neurotransmitter release at these synapses (Tottene et al., 2009). Is it possible that at these synapses, the P/Q-type calcium currents evoked by action potentials are barely affected by the R192Q mutation? Could it be that a different set of calcium channel auxiliary subunits, channel-associated proteins, or both is expressed in fast-spiking interneurons that somehow modulate channel function so that the effect of R192Q mutation is minimal? This finding also poses a challenge to future drug discovery of how to selectively target affected glutamatergic synapses without compromising normal synapses. What about the effects of R192Q mutation on other inhibitory and modulatory synapses and their impacts on the excitation and inhibition balance in cortical circuits? It will not be surprising to see differential effects of R192Q mutation at different synapses given the great diversity of cell types in the brain, specially the GABAergic interneurons. In fact, Tottene et al. (2009) already observe a difference in the activation properties of Ca_v2.1 between cerebellar granule cells and cortical pyramidal cells, indicating a diversity of P/Q-type calcium channels. A better appreciation of this diversity at the molecular level will not only further our understanding of calcium channels in general, but also certainly aid the discovery of therapeutic approaches for calcium channelopathies.