CaV3.2 calcium channels control NMDA receptor-mediated transmission: a new mechanism for absence epilepsy

Low-threshold voltage-gated T-type CaV3.1, CaV3.2, and CaV3.3 calcium channels, encoded by genes CACNA1G, CACNA1H, and CACNA1I, respectively, are widely expressed in various types of neurons throughout the brain (Catterall 2011; Cheong and Shin 2013; Simms and Zamponi 2014). T-type calcium channels activate near the resting membrane potential (RMP). Therefore, it is generally believed that these channels can promote the synchronization of sleep rhythms and absence seizures by modulating membrane properties with low-threshold calcium spikes and calcium-dependent potassium currents. This notion is supported by the recent genetic knockout experiments that demonstrated CaV3.1 and CaV3.3 to be indispensable for the generation of low-threshold calcium spikes (Kim et al. 2001; Astori et al. 2011; Lee et al. 2014a). However, the genetic deletion of CaV3.1 and CaV3.3 failed to prevent absence epilepsy in the knockout mice, raising the question of whether these calcium channels are essential for absence seizures. In contrast, the primary function of CaV3.2 channels remains elusive. Interestingly, genetic studies have linked a large number of gain-of-function mutations on the CACNA1H gene, but no confirmed mutation on the CACNA1G and CACNA1I genes, with human childhood absence epilepsy (CAE) (Chen et al. 2003; Liang et al. 2006; Heron et al. 2007), which is the most common form of pediatric epilepsy (Snead 1995; Tenney and Glauser 2013). More perplexingly, ∼50% of CAE patients did not respond to ethosuximide, a T-type calcium channel antagonist and first-line drug used to treat CAE (Coulter et al. 1989; Huguenard and Prince 1994), and the majority of the nonresponsive patients carry CaV3.2 channel mutations (Glauser et al. 2010). These findings establish the functional significance of CaV3.2 channels but raise the fundamental questions of what the primary function of CaV3.2 channels is and how CAE-linked CaV3.2 mutations may be epileptogenic. At central synapses, NMDA transmission plays a decisive role in controlling AMPA transmission strength (Kessels and Malinow 2009; Stornetta and Zhu 2011; Huganir and Nicoll 2013). It is also clear that NMDA transmission strength itself is regulated at synapses, albeit the detailed mechanisms and dynamics of this regulation are much less explored (Lau and Zukin 2007; Hunt and Castillo 2012; Paoletti et al. 2013). Since synaptic NMDA transmission is central to fundamental cognitive functions, including sensory perception and behavior adaption, it is not surprising to see that dysregulation of NMDA transmission can lead to various neurological, mental, and psychiatric disorders, including addiction, Alzheimer's disease, autism, depression, pathological pain, and schizophrenia (Lau and Zukin 2007; Russo et al. 2010; Hunt and Castillo 2012; Paoletti et al. 2013; Monteggia and Zarate 2015). However, what controls NMDA transmission strength remains unknown. We report here an investigation of regulation and function of CaV3.2 channels in multiple distinct types of rat central neurons in intact circuits and intact brains. Using simultaneous multiple patch clamp recordings and/or multiple two-photon imaging techniques (Wang et al. 2015), we found that, in sharp contrast to the other T-type calcium channels (i.e., CaV3.1 and CaV3.3 channels), CaV3.2 channels did not contribute to either modulation of membrane properties or production of low-threshold calcium spikes in central neurons. Instead, functional CaV3.2 channels primarily incorporated into synapses by replacing existing synaptic CaV3.2 channels and served to control the strength of NMDA transmission. This CaV3.2 channel-dependent regulation of NMDA transmission required synaptic activity, activation of CaV3.2 channels, and calcium influx. Consistent with these findings, expression of human CAE-linked mutant hCaV3.2(C456S) channels in rats resulted in replacement of endogenous CaV3.2 channels with higher open probability mutant channels that led to 2- to 4-Hz spike and wave discharges (SWDs) and absence-like epilepsy characteristic of CAE patients. The SWDs and absence-like epilepsy were suppressed by AMPA receptor (AMPA-R) and NMDA-R antagonists but not T-type calcium channel antagonists. These results reveal a surprising role of CaV3.2 channels in regulation of synaptic NMDA transmission strength and establish the first genetic model for CAE patients carrying CaV3.2 channel mutations.