Here we demonstrate that the N-type voltage-gated calcium channel, a major presynaptic calcium channel, is a Cdk5 substrate. Phosphorylation of the CaV2.2 pore-forming

α1 subunit by Cdk5 increases calcium influx by enhancing channel open probability and also facilitates neurotransmitter release. These events are mediated by an interaction between CaV2.2 and RIM1, which impacts vesicle docking at the active zone. Our results outline a mechanism by which Cdk5 regulates N-type calcium channels and affects presynaptic function. To investigate whether the N-type calcium channel is a Cdk5 substrate, click here we cloned the intracellular domains of the CaV2.2 α1 subunit into glutathione S-transferase (GST) fusion protein constructs for in vitro kinase assays (Figure 1A). Each UMI-77 cell line purified GST-CaV2.2 protein fragment was incubated with an activated Cdk5/p25 protein complex along

with radioactive [γ32-P]ATP to assay the level of Cdk5 kinase activity (Figure 1B). Two GST-CaV2.2 fusion protein fragments, the C-terminal 3 (CT 3, amino acids 1981–2120) and C-terminal 4 fragments (CT 4, amino acids 2121–2240) were consistently phosphorylated by Cdk5 (Figure 1C). Mutagenesis of serine 2013 (S2013), a consensus Cdk5 site on the CT 3 fragment, to alanine abolished Cdk5 phosphorylation. However, several combinations of point mutations on the CT 4 fragment were insufficient to reduce Cdk5/p25 phosphorylation (Figure S1 available online). Only mutagenesis of all seven putative Cdk5 phosphorylation sites on the CT 4 fragment resulted in undetectable phosphorylation levels (Figure 1D). These kinase assays identify the N-type calcium channel as a Cdk5 substrate. To confirm phosphorylation of the Sitaxentan N-type calcium channel, we generated and purified a phosphorylation-state-specific

antibody to S2013, a well-conserved residue (Figure 2A). The phospho-CaV2.2 antibody (pCaV2.2) signal was robust when the CT 3 fragment, but not the CT 3 (S2013A) fragment, was coincubated with Cdk5/p25, indicating that the antibody was specific to S2013-phosphorylated CaV2.2 in vitro (Figure S2). Furthermore, the pCaV2.2 antibody signal was observed only in the presence of Cdk5/p35 in a cell line stably expressing the rat isoform of CaV2.2 (Lin et al., 2004), and alkaline phosphatase (CIP) treatment abolished the signal (Figure 2B). Since S2013 is also conserved in P/Q-type calcium channels, we tested the specificity of the Cdk5-dependent S2013 phosphorylation by immunoprecipitation of brain lysates with an anti-CaV2.2 antibody, followed by immunoblotting for pCaV2.2 in lysates of control and Cdk5 conditional knockout (cKO) mice (Guan et al., 2011). We noted that pCaV2.