The CACNA1C gene encoding the Cav1.2 subunit of the L-type calcium channel has emerged as a new candidate gene for neuropsychiatric disease, including bipolar disorder, major depression, schizophrenia and autism.1, 2, 3 We report that global haploinsufficiency, forebrain-specific elimination and prefrontal cortex (PFC)-specific knockdown of cacna1c all increase anxiety-related behavior in mice, a prominent component of the forms of neuropsychiatric disease in which aberrations in CACNA1C have been implicated, without affecting compulsive behavior. Constitutive cacna1c heterozygous mice (HET) were evaluated in three behavioral assays related to anxiety: open field test, light–dark conflict test and elevated plus maze (EPM). HETs displayed anxiety-like behavior in the EPM (Figure 1a), spending significantly less time exploring the open arms compared with wild-type littermate controls (WT; F1,19=6.437; P<0.05). However, no differences were observed between HETs and WTs in the open field and light–dark conflict test, (Figures 1a and b, Supplementary Material). We also observed a similar statistically significant effect of increased anxiety-like behavior compared with WTs in EPM in adult female HETs (Figure 1d, Supplementary Material) and adolescent male HETs (Figure 1e, Supplementary Material). To more specifically investigate the function of cacna1c in the brain, we generated forebrain-specific conditional cacna1c-deficient mice (forebrain-cacna1c cKO) by crossing cacna1c-floxed mice with mice harboring alphaCaM Kinase II promoter-driven expression of Cre recombinase.4 Relative to WTs, this strategy achieved ∼70% elimination of cacna1c mRNA in the hippocampus, PFC, basolateral amygdala, striatum and nucleus accumbens, as assessed by quantitative PCR (Table 1, Supplementary Material). Cacna1c mRNA levels were unaffected in the ventral tegmental area and cerebellum. With this greater reduction in cacna1c in forebrain than could be achieved in HETs, significantly increased anxiety-like behavior was observed across all three behavioral assays. In EPM, forebrain-cacna1c cKO mice spent significantly less time exploring the open arms compared with WTs (Figure 1b, F1,16=68.587; P<0.0001 and Figure 2c, Supplementary Material). In the open field test, forebrain-cacna1c cKO mice spent less time exploring the center of the chamber compared with WTs (Figures 2a and 3a, Supplementary Material). In the light-dark conflict test, forebrain-cacna1c cKO mice spent significantly less time in the brightly lit side compared with WTs (Figures 2b and 3b, Supplementary Material). Figure 1 Anxiety-like behavior as measured in the elevated plus maze (EPM) assay is shown for (a) cacna1c haplosufficient (cacna1c HET; n=10) and wild-type (WT; n=11) littermates, (b) forebrain-specific cacna1c knockout (forebrain-cacna1c cKO; n=8) and WT controls ... Clinically, anxiety is often accompanied by compulsive behavior, such as in obsessive-compulsive disorder (OCD), in which patients seek alleviation from recurrent bouts of anxiety-inducing intrusive thoughts by engaging in compulsively repetitive behaviors. Experimental models for OCD, such as SAPAP3-5 or SLITRK5-deficient6 mice, display pathologically high compulsive grooming that is readily quantified by the spray-induced grooming test. Compared with respective WTs, we did not observe elevated grooming in either HETs or forebrain-cacna1c cKO mice (Figures 1c and 3c, Supplementary Material). Thus, the form of anxiety associated with cacna1c function is distinct from that associated with OCD spectrum illnesses. Some genetic variations in CACNA1C have been associated with altered PFC function7, 8, 9 in neuropsychiatric disease, so we next generated focal elimination of cacna1c in the PFC with adeno-associated viral (AAV) vector-expressing Cre recombinase (AAV-Cre).10 AAV-Cre was stereotaxically delivered bilaterally into the PFC of floxed cacna1c mice, and regional elimination of Cav1.2 was immunohistochemically confirmed (Figures 1d and e). Following elimination of cacna1c in the PFC, mice showed no differences in basal locomotor activity compared with AAV-GFP control injected mice (Figure 4, Supplementary Material). However, selective elimination of cacna1c in the PFC was associated with less time spent exploring open arms of the EPM, compared with control AAV-GFP injected mice (Figure 1c, F1,16=5.477; P<0.05). To evaluate the specificity of PFC cacna1c knockdown in mediating anxiety, we used AAV-expressing cacna1d siRNA10 to selectively eliminate cacna1d in the PFC, the other L-type Ca2+ channel isoform expressed in brain. Selective knockdown of cacna1d in the PFC had no effect on locomotor behavior (Figure 5a, Supplementary Material) or time spent in open arms in the EPM (Figure 5b, Supplementary Material). In summary, we report here the first direct evidence for a role of forebrain cacna1c in regulating anxiety. Mice harboring forebrain-specific elimination of cacna1c may thus provide a useful tool for studying the pathophysiology of anxiety in forms of neuropsychiatric diseases in which CACNA1C is implicated.