CACNA1C: Association With Psychiatric Disorders, Behavior, and Neurogenesis

Abstract

Large-scale genome-wide association studies have consistently shown that genetic variation in CACNA1C, a gene that encodes calcium voltage-gated channel subunit alpha1C, increases risk for psychiatric disorders. CACNA1C encodes the Cav1.2 subunit of voltage-gated calcium channels, which themselves have been functionally implicated in a broad spectrum of neuropsychiatric syndromes. Research has concentrated on uncovering the underlying biological mechanisms that could be responsible for this increased risk. This review presents an overview of recent findings regarding Cacna1c variation in animal models, particularly focusing on behavioral phenotypes associated with neurodevelopmental disorders such as cognition, anxiety and depressive phenotypes, and fear conditioning. The impact of reduced gene dosage of Cacna1c on adult hippocampal neurogenesis is also assessed, including new data from a novel Cacna1c+/− rat model.

Introduction

Genetic Association

The growth of psychiatric genetics has heralded in a new era of knowledge about psychiatric and neurodevelopmental disorders. Genome-wide association studies (GWAS) have been highly influential in identifying common variation in genes that are over or underrepresented in individuals with a certain disorder. These studies identify single nucleotide polymorphisms (SNPs) that occur throughout the genome that increase risk for neuropsychiatric disorders. One of the first, and now well replicated, GWAS finding in psychiatry was the association of SNP rs1006737 within the calcium voltage-gated channel subunit alpha1c (CACNA1C) gene with bipolar disorder.¹ This association was confirmed in a larger data set,² and subsequent studies showed a further association of this SNP with schizophrenia, major depressive disorder (MDD) and autism (table 1). Further SNPs within CACNA1C have since been associated with these disorders in multiple studies (table 1).

Table 1.

Summary of Published Association Studies of SNPs Within CACNA1C With Psychiatric/Neurodevelopmental Disorders

SNP	Disorder	Risk Allele	Main References
rs1006737	BPD	A	Ferreira et al2
			Sklar et al1
			Green et al3
			Gonzalez et al4
			Liu et al5
			Ruderfer et al6
			Lett et al7
	SCZ	A	Green et al8
			Nyegaard et al9
			He et al10
			Ivorra et al11
			Guan et al12
			Zheng et al13
			Hori et al14
			Ruderfer et al6
	Autism	G	Li et al15
	MDD	A	Liu et al5
			Green et al8
			Wray et al16
			Casamassima et al17
rs4765905	SCZ	A	Hamshere et al18
			Takahashi et al19
	Autism	G	Li et al15
rs4765913	BPD	A	Ripke et al20
			Mühleisen et al21
	SCZ	A	Ripke et al20
	MDD	A	Ripke et al20
rs1024582	BPD, SCZ, ADHD, MDD, autism	A	Cross-Disorder Group of the Psychiatric Genomics Consortium22
rs2007044	SCZ	A	Ripke et al20
			Pardiñas et al23
rs7297582	BPD	T	Liu et al5
	MDD	T	Liu et al5
rs12898315	SCZ	A	Pardiñas et al23
rs10744560	BPD	T	Stahl et al24

Note: BPD, bipolar disorder; SCZ, schizophrenia; MDD, major depressive disorder; ADHD, attention deficit hyperactivity disorder.

The majority of these SNPs are in known linkage disequilibrium with each other, except rs7297582 and rs12898315, potentially due to the fact they are less studied. The SNPs lie within introns, within predicted enhancers which can interact with the CACNA1C transcription start site²⁵ and therefore may determine gene expression.^26,27 rs1006737 has been shown to be an expression quantitive trait loci (eQTL) for CACNA1C expression: associated with decreased expression.²⁷

CACNA1C SNPs were found to have shared effects across attention deficit hyperactivity disorder (ADHD), autism, BPD, SCZ, and MDD,²² implying that common variation in CACNA1C may be associated with particular symptom clusters instead of one particular disorder.

In addition to GWAS findings, large exome sequencing studies have shown that rare disruptive mutations within calcium ion channels are enriched in patients with schizophrenia²⁸ and autism.^29,30 Furthermore, missense mutations in exon 8, or the alternatively spliced exon 8a, of CACNA1C can cause an autosomal dominant genetic disorder named Timothy syndrome (TS).³¹ TS is a multisystem channelopathy characterized by cardiac defects, craniofacial abnormalities, autism, and cognitive impairments. There are 2 common types of TS characterized by mutation; TS1 (G406R in exon 8a) and the more severe form TS2 (G406R or G402S in exon 8). Both TS1 and TS2 are characterized by gain-of-function mutations in CACNA1C.³²

Cacna1c, Gene Transcription, and Synaptic Plasticity

CACNA1C encodes for the alpha1c subunit of the Cav1.2 L-type voltage-gated calcium channel (LTCC). This subunit forms the pore through which calcium influxes into a cell and initiates downstream signaling cascades.³³ LTCCs have a prominent role in controlling gene expression through coupling membrane depolarization with cAMP response element-binding protein (CREB) phosphorylation via local Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) signaling.³⁴ CREB can bind to a critical Ca²⁺ response element within brain-derived neurotrophic factor (BDNF) to trigger its transcription.^35,36 This pathway, and particularly CREB and BDNF, are thought to be essential for learning and memory processes. Synaptic plasticity, which is thought to underlie learning and memory, can be modulated by LTCCs^37,38; LTCC antagonists reduce induction of long-term potentiation (LTP) in the CA1 of the rat hippocampus.³⁹ Cav1.2 knockdown models have shown reduced CREB transcription and hippocampal LTP,^40,41 implicating the important of these channels in gene expression and plasticity.

This review aims to give a brief overview of the current phenotypes relevant to psychiatric and neurodevelopmental disorders studied so far in animal models of Cacna1c/Cav1.2 dysfunction, including new findings on impacts on neurogenesis in a rat model of reduced gene dosage of Cacna1c.

Models

Genetic Cacna1c/Cav1.2 rodent models have mostly concentrated on reduced gene dosage. Some studies utilize a constitutive heterozygote model (Cacna1c^+/−) to study gene dosage effects as the homozygote model is embryonically lethal. However, other studies have utilized region-specific complete knockouts of Cacna1c (Cacna1c^−/−) driven by specific promotors to disentangle the neuronal contribution of this gene compared with the cardiac properties. Bader and colleagues⁴² developed a genetic mouse model based on TS2. While both homozygote and heterozygote knockout of exon 8a were lethal, a heterozygote model that included an inverted neomycin cassette was viable (TS2_neo).⁴² An overview of the genetic Cacna1c/Cav1.2 mouse models and their associated phenotypes are presented in table 2.

Table 2.

Overview of the Currently Studied Mouse Models of Cacna1c Dysfunction and Their Associated Psychiatric and Mood Phenotypes

Study	Model	Phenotype
Bader et al42	TS2_neo+/−	↓ novelty induced locomotion
		↓ sociability
		↑ cued and contextual fear memory
		↓ extinction of fear memory
		↑ preservation in Y maze
Kabitzke et al43	TS2_neo+/−	↓ social-induced locomotion
		↓ sociability
Dao et al44	Cacna1c +/−	↓ exploratory activity in females
		↓ locomotion in females
		↓ depressive phenotype
		↑ anxiety in females
Bader et al42	Cacna1c +/−	↓ basal and novelty induced locomotion
		↑ anxiety
Bavley et al45	Cacna1c +/−	↓ depressive phenotype
Moosmang et al40	Cacna1c −/− (forebrain only)	↓ spatial discrimination
McKinney et al46	Cacna1c −/− (forebrain excitatory neurons only)	No effect on contextual fear memory
White et al47	Cacna1c −/− (forebrain only)	↓ long-term spatial memory
Langwieser et al48	Cacna1c −/− (CNS only)	No effect on cued fear memory
Lee et al49	Cacna1c −/− (prefrontal cortex only)	↑ anxiety
Temme et al50	Cacna1c −/− (neurons only)	↓ context discrimination
		↓ spatial memory (complex task)
		↓ neurogenesis
Lee et al51	Cacna1c −/− (forebrain only)	↓ neurogenesis
Kabir et al52	Cacna1c −/− (prefrontal cortex only)	↓ depressive phenotype
Dedic et al53	Cacna1c +/− (excitatory neurons only)	↓ sociability
		↓ depressive phenotypes
		↑ susceptibility to chronic social defeat stress
		↑ anxiety

Motor Function

Neurodevelopmental disorders, particularly autism, can present with neurological disturbance of the motor system resulting in abnormal gait^54,55 and dysfunctions in movement planning and execution.⁵⁶ Bader et al⁴² reported that TS2_neo mice had similar motor abilities and reflexes in their home cage, however had decreased locomotion when placed in a novel environment. Consistently, another study reported that while TS2_neo mice had no deformities in gait, they had reduced locomotion in social tests such as reciprocal social interaction, urine open field test and increased freezing in the Smartcube platform challenge.⁴³Cacna1c^+/− mice were reported to be markedly hypoactive in both a home cage and novel environment,⁴² however studies on Cacna1c^+/− mice using a rotarod paradigm^40,44,48 did not report any differences in motor ability or co-ordination. A prefrontal cortex specific elimination of Cacna1c also did not result in any different basal locomotor behavior.⁴⁹ Dao et al⁴⁴ also reported no genotype differences in motor activity in the home cage however did report a slight hypoactivity in females in the open field test, as well as reduced exploratory activity in the holeboard test.

The role of Cav1.2 in motor activity thus requires further clarification, models suggest that dysfunction in Cacna1c may lead to elements of hypoactivity. It is important to consider that this reduced locomotion could in part reflect an indication of anxiety in contrast to a motor deficit per se.

Sociability

Social interactions, and the perceptions of them, are often altered in psychiatric patients. TS2_neo mice show no sociability defects in the 3 chamber test⁴³ and maintained social memory,⁴² however present decreased activity in social interactions. They also initiate less social events, but maintain them longer.^42,43 The Cacna1c^+/− knock out mouse did not show any differences in social beahvior,⁴² however a Cacna1c^+/− excitatory neuron knockout showed decreased sociality.⁵³ This suggests that some subtle elements of social interactions may be affected in Cav1.2 dysfunction, but no global social deficits are present.

Fear Conditioning

Aversive associative learning processes such as fear conditioning can be used to investigate learning, memory, and cognitive processes in animal models. They can give us an understanding on the neural circuity, ie, affected in a wide range of psychiatric disorders. Interestingly, it has been shown that Cav1.2 levels are increased in the amygdala following fear conditioning.⁵⁷ In genetic models, deletion of Cav1.2 in the anterior cingulate cortex results in decreased observational fear learning, where unconditioned mice develop freezing behavior by observing conditioned mice receiving foot shocks.⁵⁸ TS2-neo mice can acquire cued fear conditioning correctly, however demonstrate increased freezing in context and cue recalls, as well as reduced extinction.⁴² The authors suggest that this is due to an enhanced perseverance of both tone and context memory. However, other models of Cacna1c knockdown do not show alterations of fear memory. Animals with neuron specific knockout of Cacan1c^−/− show no impairments in acquisition, consolidation or recall of auditory,⁴⁸ or contextual⁵⁰ fear conditioning paradigms. However, Temme et al⁵⁰ did show significant context discrimination deficits in their neuronal knockout model. The Cacna1c^−/− (forebrain excitatory neurons only) model also maintained successful consolidation and extinction of conditioned fear.⁴⁶ This disparity between the Cacna1c knockdown models and TS models is interesting and may suggest that there are some compensatory adaptations.⁴⁸ Future studies on Cacna1c^+/− models would be beneficial to further investigate Cav1.2’s influence over fear memory.

Anxiety and Depressive Phenotypes

Cacna1c ^+/− mice have shown decreased depressive-related phenotypes as assessed by the tail suspension test⁴⁵ at 5–7 days following a chronic stress. Cacna1c heterozygosity has also been associated with protection against depressive-like phenotypes in the forced swim, sucrose preference, and tail suspension tests.^44,52,53 However, Cacna1c^+/− deletion during development increases susceptibility to chronic social defeat stress.⁵³ In addition, a gene × environment human study revealed that SNPs in CACNA1C interact with trauma to predict depressive symptoms,⁵³ suggesting that depressive phenotypes may be subject to environment factors interacting with CACNA1C.

Dao et al⁴⁴ reported increased anxiety-related phenotypes in female Cacna1c^+/− mice only.⁴⁴ Increased anxiety-like phenotypes in males has been reported in Cacna1c^+/− mice in an annex test,⁴² dark-light box,⁵³ and in the open field,⁴⁹ however, these findings are not consistent across all models.^40,45 The TS2-neo model has not been associated with alterations in anxiety.^42,43

The association between Cav1.2 and anxiety is still not fully understood. However, the current literature seems to suggest that Cacan1c heterozygosity may result in increased anxiety and this effect may be stronger in females.

Cognition

Elements of cognitive dysfunction, such as working memory, are common in psychiatric disorders and may represent core features of these conditions.⁵⁹ The SNP rs1006737 was associated with increased prefrontal activity during executive cognition in healthy humans⁶⁰ and impaired logical memory performance¹⁴ in those with schizophrenia. SNP rs2007044 was also associated with poor working memory in schizophrenia patients, potentially through decreased prefrontal cortex connectivity to other cortical regions.⁶¹

No significant differences were seen between TS2-neo mice and wild-types in the procedural T-maze,⁴³ however, increased preservative behavior was observed in the Y maze.⁴² Elements of spatial memory have been shown to be affected in Cacna1c^−/− conditional forebrain knockout mice.^40,47,50 In the Morris water maze, knockout mice could learn the spatial task correctly,^46,47 but they display spatial memory impairments when tested 30 days later.⁴⁷ In a neuronal specific Cacna1c knockdown, mice could successful learn a simple Morris water maze but had profound deficits in the acquisition of spatial learning within a more complex maze when visual cues around the room were limited.⁵⁰ Impairments in a water maze spatial discrimination task have also been reported.⁴⁰

These findings have implications for understanding how genetic variants can have an impact on underlying cognitive impairments in psychiatric disorders, however more research is needed to clarify the primary domains affected. It will also be important to test animal models on tasks with a high degree of translational potential, such as rodent analogues of human touchscreen tasks, to facilitate future integrative research and drug development.

Neurogenesis

Cav1.2 may be required for more complex cognitive behaviors such as limited cued Morris water mazes where allocentric spatial representations are required. Data have shown that adult hippocampal neurogenesis is required for formation of complex forms of spatial representations but not simple,⁶² mirroring results seen in cognitive tasks following Cav1.2 knockdown. This suggests a possible deficit in adult hippocampus may be responsible for elements of behavior dysfunction in these models.

Psychiatric disorders, and in particular mood disorders, have been linked to alterations in adult neurogenesis.⁶³ In rodents and humans, neurogenic niches have been found in the ventricular-subventricular zone in the lateral ventricles and the subgranular zone (SGZ) of the dentate gyrus (DG) in the hippocampus.⁶⁴ These neurogenic cells of the hippocampus have been commonly associated with psychiatric and affective disorders⁶³ although this is still controversial in the current literature. Adult hippocampal neurogenesis is a complex multistep process, ie, necessary for the generation of new neurons from neural stem cells (NSCs). A range of psychotropic medications have been associated with increasing neurogenesis in rodent models (including SSRIs, selective SNRIs, and tricyclic antidepressants).⁶⁵ There is also increasingly evidence that hippocampal neurogenesis contributes to some forms of hippocampus-dependent learning and memory.⁶⁶ There are many factors regulating this process: environmental cues, growth factors such as BDNF, glucocorticoids, and neurotransmitters.⁶³ As the literature suggests that stress may interact with CACNA1C to cause depressive symptoms, and CACNA1C is known to mediate BDNF production, it may be hypothesized that Cav1.2 has a role in neurogenesis, through interacting with stress or BDNF.

LTCCs have been shown to regulate the conversion of adult hippocampal neural precursors to immature neurons in a bidirectional manner.⁶⁷ This agrees with findings in genetic models; Cacna1c^−/− deletions in the forebrain and neurons both show decreases in immature neurons (table 3). In the forebrain-Cav1.2 knockout, this was attributed to increased cell death of young neurons, correlated with decreased BDNF levels.⁵¹ However, in a pan-neuronal Cacna1c deletion marked decreases in cell proliferation were seen which is likely to be the cause of decreased numbers of immature neurons.⁵⁰ Völkening et al⁶⁸ deleted Cacna1c on type 1 cells and reported decreased proliferation and immature neuron production (table 3). These mice also showed deficits in a pattern separation paradigm—a type of learning thought to require intact hippocampal neurogenesis.⁶⁸

Table 3.

Summary of the Findings Involving Hippocampal Neurogenesis in Rodent Models of Cacna1c/Cav1.2 Dysfunction

	Model	Proliferation Rate	Immature Neurons	Cell Survival	Dentate Gyrus Size
Lee et al51	Forebrain-Cav1.2 CkO (−/−)	n.d.	↓	↓	n.d.
Temme et al50	Neuron-specific deletion of Cav1.2 CkO (−/−)	↓	↓		n.d.
Völkening et al68	TgGLAST-CreERT2 /Cacna1cfl/fl /RCE:loxP mice (Mouse model with Cacna1c-deficient Type-1 cells)	↓	↓
Moon et al, this study	Cacna1c+/− rat model	↓	n.d.		n.d.

Note: n.d., no significant difference seen.

We have used a novel Cacna1c heterozygote (Cacna1c^+/−) rat model to investigate if these findings could be replicated in another rodent species.⁶⁹ We show a marked decrease in cells incorporating 5-bromo-2-deoxyuridine (BrdU)—a nucleotide analogue that marks dividing cells—suggesting that proliferation is significantly decreased in the SGZ in this model, confirming a key role for Cacna1c in neurogenesis across species. However, we do not see any difference in the number of immature neurons, contrasting with the findings in the mouse models (table 3, figures 1 and 2). This may be due to compensatory mechanisms such as decreased apoptosis resulting in increased cell survival. Further studies assessing long-term survival, over following the proliferation, survival, differentiation, and integration of newly born neurons following BrdU incorporation, would give valuable insight into the functional consequences of Cacna1c knockdown on psychology and behavior related to this process.

Fig. 1.

Cacna1c ^+/− rats show decreased BrdU, a marker of cell proliferation, in both suprapyramidal and infrapyramidal blades of the dentate gyrus (F = 11.9133, P = .0043, one-way ANOVA). There are no differences in doublecortin positive cells between Cacna1c^+/− rats and wild-type littermates. Bars represent normalized mean per mm² ±SEM, n = 8/genotype, all males.

Fig. 2.

Representative immunofluorescent image of BrdU+ cells (green) and DCX+ cells (red) in the dentate gyrus of the hippocampus. For color, please see the figure online.

Conclusions

Associations of psychiatric disorders with the CACNA1C locus has been one of the most robust findings from genetic studies in mental health. This has led to the investigation of a number of animal models of genetic variation in Cacna1c to study potential risk pathways. These models have yielded some clues as to functional impacts—including potential alterations in motor behaviors, social interactions as well as increased anxiety, and preservative behavior. Interestingly, there may also be a subtle anti-depressive effect of a reduced gene dosage of Cacna1c, although interactions with stress may alter this phenotype.

Cav1.2 also appears to play an essential role in elements of hippocampal neuron production, suggesting that the alterations seen in neurogenesis in rodent models have also play a part in other phenotypes seen. This is of interest as disruptions in SGZ neurogenesis have been associated with both psychiatric disorders and treatment response. However, more work is needed to determine if this, in fact, a causative effect.

There are, of course, limitations to the work so far. The majority of the Cacna1c^+/− models have focused on reduced gene dosage, whereas some of the genetic literature suggests that both loss and gain-of-function phenotype may be relevant to disease. Additionally, it is important to note that there is an imprecise relationship between rodent behavioral tests and human psychiatric disorders. Further studies using translational tasks and assessments in both animal models and human subjects with specific genetic variants in CACNA1C will be needed to build up the knowledge required for potential therapeutic targeting of LTCCs and associated pathways in psychiatric disorders.

Supplementary Material

Supplementary data are available at Schizophrenia Bulletin online.

Laboratory Animals

All procedures were carried out in accordance with local ethics guidelines, the UK Home Office Animals Act 1986 and the European Communities Council Directive of 24 November 1986 (86/609/EEC). For further details on methods, please see supplementary material.

Funding

This work was supported by the Medical Research Council (PhD scholarship awarded to A.M.) and a Wellcome Trust Strategic Award (503147).