Calcium Channel Ca_Vα₁ Splice Isoforms - Tissue Specificity and Drug Action

Abstract

Voltage-gated calcium ion channels are essential for numerous biological functions of excitable cells and there is wide spread appreciation of their importance as drug targets in the treatment of many disorders including those of cardiovascular and nervous systems. Each Cacna1 gene has the potential to generate a number of structurally, functionally, and in some cases pharmacologically unique Ca_Vα1 subunits through alternative pre-mRNA splicing and the use of alternate promoters. Analyses of rapidly emerging deep sequencing data for a range of human tissue transcriptomes contain information to quantify tissue-specific and alternative exon usage patterns for Cacna1 genes. Cell-specific actions of nuclear DNA and RNA binding proteins control the use of alternate promoters and the selection of alternate exons during pre-mRNA splicing, and they determine the spectrum of protein isoforms expressed within different types of cells. Amino acid compositions within discrete protein domains can differ substantially among Ca_V isoforms expressed in different tissues, and such differences may be greater than those that exist across Ca_V channel homologs of closely related species. Here we highlight examples of Ca_V isoforms that have unique expression patterns and that exhibit different pharmacological sensitivities. Knowledge of expression patterns of Ca_V isoforms in different human tissues, cell populations, ages, and disease states should inform strategies aimed at developing the next generation of Ca_V channel inhibitors and agonists with improved tissue-specificity.

1. INTRODUCTION

The functional core of the voltage-gated calcium ion channel is the Ca_Vα₁ subunit and each may be associated with a cytoplasmic calcium channel Ca_Vβ subunit, an extracellular Ca_Vα₂δ subunit, and other proteins such as gamma subunits, calmodulin, scaffold proteins, kinase anchoring proteins, and downstream target proteins [1]. In mammals, 10 different genes (Cacna1A-1I and 1S) encode 10 Ca_Vα₁ subunits, and in humans these reside on nine different chromosomes. Cacna1 genes encode three main classes of Ca_Vα₁ proteins grouped according to their degree of sequence homology: Ca_V1 (1S, 1F, 1C, 1D), Ca_V2 (1A, 1B, 1E) and Ca_V3 (1G, 1H, 1I) [2–4]. Expansion within each of the 3 main Cacna1 gene families likely occurred some 450 million years ago probably as a result of genome doublings from single ancestral vertebrate Cacna1 genes [5]. For example, phylogenic analyses suggest that Cacna1s, Cacna1c, Cacna1d, and Cacna1f genes are located within paralogous chromosome regions (spread across four different chromosomes) that contain several gene families including visual opsins [5].

As individual calcium channel genes evolved, Ca_V proteins acquired new functions and phenotypes that account for their unique biological activities in specific tissues and cell types. For example, of the four mammalian Cacna1 genes that encode Ca_V1 subunits: Cacna1s expression is limited to skeletal muscle and Ca_V1.1 proteins contain a unique sequence in the II–III linker region that couples the channel to the ryanodine receptor. This is the critical first step in excitation-contraction coupling in skeletal muscle [6–8]; Cacna1f is highly expressed in retina but also in other human tissues (CACNA1F, the GTEx Portal on 12/06/2014) [9], and Ca_V1.4 channels are important for transmission at glutamatergic ribbon synapses between photoreceptors and second-order neurons [10–13]; and Cacna1c and Cacna1d genes have relatively broad expression patterns including in brain, heart, smooth muscle, endocrine cells, and auditory hair cells [1, 14] (CACNA1C and CACNA1D, the GTEx Portal on 12/06/2014) [9]. The expression of functionally different splice isoforms of Ca_V1.2 and Ca_V1.3 in different tissues enables Cacna1c and Cacna1d genes to support a wider range of calcium-dependent cellular functions [2, 3, 15–17].

2. PHARMACOLOGICAL TOOLS THAT DISCRIMINATE AMONG MAJOR SUBTYPES OF Ca_V CHANNELS

Pharmacological approaches are invaluable to delineate the cellular functions of different Cacna1 genes. For example, dihydropyridine (DHP) antagonists and agonists are the molecules of choice to establish the contributions that Ca_V1 channels make to various cellular responses, whereas the spider toxin ω-agatoxin IVA and the cone snail toxin ω-conotoxin GVIA are highly selective inhibitors of Ca_V2.1 and Ca_V2.2 channels respectively (Table 1). As is well known, and discussed throughout this special issue, DHP antagonists are used in the clinic to lower blood pressure. All four Ca_V1 channels are sensitive to DHPs, but their anti-hypertensive action in vivo is dominated by the inhibitory actions of several DHPs on vascular smooth muscle, DHPs have few side effects, a surprising fact considering the relatively broad expression patterns of Ca_V1.2 and Ca_V1.3 channels in many excitable cells.

Table 1.

CACNA1 Genes, Proteins, Currents.

Gene name: human chromosome	Protein name	Exon, splicing factor	Current	Pharmacology (influenced by exon)
CACNA1S: chr 1	CaV1.1	e29, rbFox [121]	L	DHP– Isradipine and nisoldipine
CACNA1C: chr 12	CaV1.2	e8a, PTB [47] e9*, rbFox [48] e21/e22, ND e33, rbFox [48]	L	(e8/e8a; e21/22) [77,78,81,84,85] Diltiazem (e9*) [79]
CACNA1D: chr 3	CaV1.3	e8b, rbFox [121] e42a, ND	L	DHP – Nimodipine (e42a) [80]
CACNA1A: chr 19	CaV2.1	e24a, Nova(+) [49] e31a, Nova (−) [49]	P/Q	ω-Agatoxin IVA (e31a) [61]
CACNA1B: chr 9	CaV2.2	e24a, Nova(+) [122] e31a, Nova(−) [49] e18a, rbFox e37a/e37b, ND	N	μ-opioid and GABAB receptors - DAMGO, Morphine, Baclofen [107–109]

(+) indicates enhancer of exon inclusion and (−) repressor; ND = Not determined; All genes contain numerous sites of alternative splicing but only those for which the splicing factor has been determined, or that are known to influence drug sensitivity are listed here. For a more comprehensive review of alternative splicing in Cacna1 genes see [2, 3, 15, 17, 31, 47–49, 61, 77–80, 91, 92, 108, 109, 121–124].

As shown by Bruce Bean, and by Reuter and colleagues, the efficacy of DHP antagonists is strongly voltage-dependent and these drugs are more potent inhibitors of Ca_V1 channels that are activated from relatively depolarized membrane potentials typical of smooth muscle cells in vivo [18, 19]. Preferential inhibition of inactivated Ca_V1 channels also results in reduced inhibition by DHPs of Ca_V1 channels activated by brief, non-inactivating stimuli applied from negative membrane potentials – such as action potential-like waveforms [20]. Ca_V1.3 channels have different biophysical properties and have somewhat lower intrinsic sensitivities to certain DHPs as compared to Ca_V1.2 channels [21, 22]; although these differences are generally insufficient to permit unambiguous functional separation of Ca_V1.2 and Ca_V1.3 currents at cellular and behavioral levels. Ca_V1.3 channels support low threshold calcium entry during pacemaking in some cells [23–25] and data link de novo mutations in CACNA1D to increased risk of sporadic autism spectrum disorder [26]. Risk alleles for bipolar disorder, schizophrenia and other psychiatric illnesses have also been linked to CACNA1C [27–30]. Such findings have increased efforts to develop compounds to discriminate between Ca_V1.2 and Ca_V1.3 channels, as well as among their various splice isoforms. There is also evidence discussed below, that splice isoforms of Ca_V1.2 have different sensitivities to DHPs.

3. CELL-SPECIFIC ALTERNATIVE PRE-mRNA PROCESSING

All mammalian Cacna1 genes studied to date contain alternatively spliced exons and many have alternative promoters; these exons combine in different patterns according to cell-specific control, with the potential to generate thousands of Ca_V channels [2]. We know that alternative splicing of Cacna1 pre-mRNAs is extensive in mammals, but we still have only a cursory understanding of their cell-specific functions. Recent reviews have summarized what was known of the tissue-specific expression pattern of exon usage for several Cacna1 genes and, where tested, their functional impact on Ca_V channel properties [2, 15, 31, 32]. More recently, the Genotype-Tissue Expression (GTEx) project provides the most comprehensive publicly available dataset on the expression pattern of human genes, including CACNA1, and the quantitative comparison of individual exon usage, genome-wide, across different tissues [9].

Multi-exon genes have the capacity to generate several – sometimes thousands – of proteins with unique structures through cell-specific selection (or elimination) of certain exons. Alternatively spliced exons often encode autonomous protein modules that introduce or eliminate specific functions or that promote or inhibit coupling to downstream signaling cascades [33–35]. Cell-specific exon choice can involve the inclusion or exclusion of a cassette exon, the choice of one of a pair or more of mutually exclusive exons, or the use of alternate splice donor or acceptor sites (Fig. 1). Cell-specific inclusion or exclusion of certain exons can also introduce premature termination codons that promote nonsense-mediated decay (NMD) thereby regulating the overall protein expression level (Fig. 1). For example, e20a in Cacna1b introduces a premature stop codon if inserted during pre-mRNA splicing and Ca_V2.2 mRNAs containing this exon are present in brain of mice and human (Fig. 1) (GenBank FJ609386.1; see also CACNA1B in GTEx portal [9]). Alternative splicing-linked NMD is common during early differentiation of cell-types and often involves the family of polypyrimidne tract binding protein (PTB) splicing factors, and NMD has recently been shown to involve the action of neuronal splicing factors of the neuro-oncological ventral antigen (NOVA) family, that regulate splicing of cryptic exons within introns [33, 36–38].

Fig. (1).

Examples of alternative pre-mRNA splicing that have been observed for Cacna1 genes. Alternative splicing can involve the selection of one of a pair of mutually exclusive exons that are the same (or similar) lengths but that encode slightly different amino acid sequences. A pair of mutually exclusive exons, e21 and e22, are present in Cacna1c. E21 is selected more frequently during pre-mRNA splicing in brain, whereas in heart and smooth muscle, e22 is selected more often (see Table 1) [85]. E21 and e22 influence the sensitivity of Ca_V1.2 currents to DHPs [9, 77, 78, 85]. The inclusion of one mutually exclusive exon is often required to maintain the correct reading frame, and if neither or both are included during splicing a premature stop codon is often introduced resulting in NMD. Cassette exons are included or excluded during alternative pre-mRNA splicing. E31a of Cacna1a is excluded during pre-mRNA splicing in Purkinje neurons of the cerebellum (P-type currents), whereas in other neurons including cerebellar granule cells, e31a is included (Q-type currents). The presence of e31a influences inhibition by ω-agatoxin IVA [61]. The cassette e20a in Cacna1b is repressed in most neurons but if included it introduces a premature stop codon (see GenBank FJ609386.1). Certain cells may use e20a to down regulate Ca_V2.2 protein expression levels.

Recent large-scale transcriptome sequencing projects suggest that alternative exon usage (pre-mRNA splicing and alternate promoters) occurs in >97% of multi-exon genes and it is extensive in mammalian brains [39–43]. Cell-specific splicing (inclusion or exclusion) of alternative exons in the nervous system is essential for development, axon targeting, neuronal excitability, circuit formation, and drug action [2, 3, 42–46]. Cell-specific splicing factors determine when and which Ca_V channel isoforms are expressed in a given cell type [38, 41, 47–49]. Furthermore, upstream factors regulate the expression and activity of cell-specific splicing factors including other splicing factors, miRNAs, kinases, as well as epigenetic modifiers [38, 50–52].

Cell-specific splicing factors work in concert with the ubiquitous splicing machinery to select (or skip) optional exons during pre-mRNA processing (Fig. 2) (Table 1) [33, 38, 40, 53, 54]. The expression patterns and activities of cell-specific splicing factors differ according to cell-type, age, activity, and potentially metabolic state and disease states [33]. Certain master splicing factors control the selection of exons that are essential for cell-fate determination, whereas others regulate exon choice that exert control over specific cell-functions [38]. Tremendous progress has been made in mapping splicing factor binding on a genome-wide scale, leading to the identification of splicing networks for NOVA, PTB, and rbFOX and inevitable insights into genome-wide biological significance [38, 40, 55, 56]. Splicing factors known to regulate cell-specific splicing of Ca_V pre-mRNAs are summarized in Table 1.

Fig. (2).

Cell-specific RNA binding proteins promote or repress exon inclusion during alternative splicing of pre-mRNAs. The example illustrates the difference in splicing patterns of the same region of pre-mRNA in Cell 1 that expresses a RNA binding splicing factor and in Cell 2 that does not. The splicing factor represses the inclusion of a cassette exon during alternative splicing of pre-mRNA. The splicing factor binds to specific nucleotide sequences upstream of the target exon in pre-mRNA. In cell 2, the absence of the splicing repressor results in exon inclusion during pre-mRNA splicing. The fully processed mRNAs are transported into the cytoplasm and are translated into different Ca_V splice isoforms. Certain RNA binding proteins including NOVA, rbFOX and PTB families can enhance or repress exon inclusion depending on where they bind (upstream or downstream) relative to the target exon [33, 40].

Here we focus on a few examples from three Cacna1 genes to illustrate approaches that are being employed to assess the pharmacological consequences of alternative splicing on Ca_V channels, in expression systems and within their native environment in vivo. Other reviews illustrate the capacity of alternative splicing to stratify Ca_V channel phenotypes modifying biophysical characteristics, membrane trafficking, subcellular targeting, association with downstream target proteins, and pharmacology [2, 3, 15].

4. SPLICE ISOFORMS EXPRESSED IN DIFFERENT TISSUES ARE DIFFERENT

Most, although not all, alternatively spliced cassette exons in Cacna1 genes are located in hyper-variable domains of Ca_V proteins that can accommodate as few as one amino acid or larger insertions comprised of several amino acids, without disrupting basic protein function. Cassette exons are either inserted or skipped and they are located in the intracellular C-termini and N-termini, intracellular I–II, II–III, and III–IV linkers, as well as S3-S4 extracellular linkers in domains III and IV of Ca_V channels [2, 3, 15]. Pairs of mutually exclusive exons often encode regions that do not tolerate large change in protein structure, and these have been shown to modify amino acid composition in transmembrane spanning domains including IS6, IIIS2, IVS3 and the proximal region of the C-terminus of Ca_V channels (Fig. 3) [2, 3, 15].

Fig. (3).

Schematic of major transmembrane spanning, intracellular, and extracellular domains of Ca_V channels as well as variable regions encoded by alternatively spliced exons that influence Ca_V channel pharmacology. Mutually exclusive exons e8a/e8, e21/e22 and e31/e32 of Cacna1c encode transmembrane spanning segments of Ca_V1.2 and they influence DHP inhibition by various mechanisms, including modifying voltage-dependent inhibition by DHPs [77, 78, 81, 84, 85]. Putative high-affinity DHP binding sites in Ca_V1.2 (shaded). Cassette e31a of Cacna1a encodes two amino acids asparagine proline (NP) in the extracellular linker IVS3–IVS4 of Ca_V2.1 and modifies inhibition by ω-agatoxin IVA [61]. Mutually exclusive e37a and e37b of Cacna1b encode 33 amino acids in the proximal C-terminus of Ca_V2.2 and they influence G protein coupled receptor mediated inhibition [107, 108].

Certain alternatively spliced exons are conserved among Cacna1 genes. For example, all but one Cacna1 gene contains an alternatively spliced cassette exon that encodes a region in the putative extracellular linker between transmembrane spanning helices S3 and S4 in domain IV of Ca_V channels (e.g. e29 in Cacna1s; e31a in Cacna1a; e31a in Cacna1b and e33 in Cacna1c [3] and GTEx portal [9]). The composition and length of these exons differ across genes, but all affect voltage-dependent gating of Ca_V channels [3]. Many other alternatively spliced exons influence channel gating including the voltage dependence of channel activation and inactivation, gating kinetics, and calcium-dependent inactivation [57–67].

Analyses of tissue-specific transcriptomes across mammals in a variety of tissues have demonstrated that alternative splicing is also frequently lineage-specific such that there is variation in the splicing of alternative exons between species [55]. This finding emphasizes the importance of study the pattern of alternative splicing in human tissue, if the ultimately goal is drug development for human use. Interestingly, transcriptome-wide analyses across several species of mammals also show that alternative splicing frequently alters the degree of protein phosphorylation implicating an important role in defining second messenger signaling cascades [55]. It is also important to point out, however, that difference in amino acid composition and protein function among Ca_V splice isoforms, generated from the same gene but expressed in different tissues, can also be substantial in highly variable domains including C-termini. Indeed, tissue-specific differences in amino acid compositions among Ca_V channel isoforms generated from one gene may be greater than those that exist across Ca_V channels of the same family across closely related species. This is especially dramatic for large alternative exons. For example, several Cacna1 genes have different tissue-specific promoters that generate Ca_V1α subunits with substantially different N-termini. Exon 1a of Cacna1c is expressed in heart and encodes 46 amino acids defining the start of the N-terminus, whereas in other tissue including brain and smooth muscle, exon 1b is expressed and it encodes a completely different 16 amino acid sequence [68–70]. By contrast, exons 1a and 1b are highly conserved across mammals and vertebrates (see multi-alignment of vertebrates in UCSC genome browser [71] and GTEx portal [9]). Similarly, exon 18a of Cacna1b, and an equivalent exon in Cacna1e, encode 21 amino acids in the II–III linkers of Ca_V2.2 and Ca_V2.3 respectively. The amino acid sequences of these alternative exons and of constitutive flanking exons, are highly conserved across species, but expression patterns of the alternative exons differ according to tissue and development [45, 71].

5. SPLICE ISOFORMS ARE TISSUE-SPECIFIC AND CELL-SPECIFIC

By definition, cell-specific processing of pre-mRNAs results in the expression of unique patterns of Ca_V isoforms in different cells and different tissues. This offers the opportunity to achieve a higher level of tissue targeting that might be achieved otherwise, if drugs can be identified that differentially act on splice isoforms. Indeed, the therapeutic efficacy of certain drugs can depend on their preferential action on specific splice isoforms, such as the action of non steroidal anti-inflammatory drugs that have different therapeutic profiles that depend on their differential action on isoforms of cyclooxygenase (COX-1, COX-2, and COX-3) [72–75].

Within Ca_V1 channels, there is interest in identifying drugs that target brain-specific Ca_V1.2 channel isoforms over those that dominate in cardiac and smooth muscle. The precise composition and relative abundance of Ca_V1.2 isoforms across different human tissue can now be analyzed from transcriptome sequencing projects (GTEx portal [9]). Certain CACNA1C exons appear to be preferentially expressed in heart including e1a (which is controlled by a cardiac tissue-specific promoter; [68, 76]), while other exons might be preferentially expressed in brain or smooth muscle [15].

6. ALTERNATIVE SPLICING OF Ca_V PRE-mRNAs INFLUENCE DRUG ACTION

6.1. Splice Isoforms of Ca_V1.2 have Different Sensitivities to DHPs

In 1995 Soldatov, Bouron, and Reuter published an important paper showing - for the first time - that different splice isoforms of Ca_V1.2 channels exhibit measurable differences in their biophysical as well as pharmacological properties. They suggested that the different actions of DHP antagonists on splice isoforms of Ca_V1.2 might account for some of their higher selectively on smooth muscle channels [77, 78] (Fig. 3). Three pairs of mutually exclusive exons in Cacna1c have been reported to modify the sensitivity of Ca_V1.2 calcium currents to DHPs: e8a/e8, e21/e22 and e31/e32 (>10-fold changes). While, more subtle effects of DHPs and other calcium channel blockers on inhibition of Ca_V1 channels (less than 10-fold changes in potency), have been reported for e9* of Cacna1c (diltiazem) [79] and for e42a of Cacna1d (nimodipine) [80].

E8a and e8 encode sequences of IS6 and part of the S5–S6 pore forming loop; photoaffinity studies indicate that the extracellular portion of this region can interact with DHPs [81–83]. Inclusion of e8 over e8a increases the sensitivity of Ca_V1.2 currents to DHPs [81, 84]. Interestingly, e8 and e8a have different patterns of tissue expression in mammals. E8 expression is higher in lungs and aorta than in cardiac muscle, while e8a usage is more prevalent in heart than smooth muscle of rodents and human [9, 81, 85]. This observation may also contribute to the dominant biological actions of certain DHPs as anti-hypertensive [86].

Cacna1c e21 and e22 are mutually excusive and each encodes part of the extracellular linker IIIS1–IIIS2 as well as most of the IIIS2 transmembrane helix of Ca_V1.2. As discussed above, e22 increases the sensitivity of Ca_V1.2 channels to DHP antagonists compared to e21 [77, 78, 85]. E21 and e22 are expressed throughout heart, aortic tissue and brain [9, 85, 87], although e22 is relatively more abundant in rat cardiac ventricular muscle than in aortic tissue [85]. Mutually exclusive Cacna1c exons e31 and e32 each encode most of the intracellular IVS2–IVS3 linker and the entire IVS3 transmembrane helix of Ca_V1.2. In heart, the selection of e31 over e32 (and vice versa) varies with development and disease state [88]. E31 dominates early in development and it reemerges in heart failure [89, 90], while e32-containing Ca_V1.2 mRNAs dominate in normal adult heart [89].

6.2. Splice Isoforms of Ca_V2.1 have Different Sensitivities to ω-Agatoxin IVA

Zamponi and colleagues showed that the actions of ω-agatoxin IVA are influenced by the presence of a two amino acid sequence, NP, encoded by e31a of the Cacna1a gene. This sequence comprises part of the putative extracellular linker between domains IVS3 and IVS4 of Ca_V2.1 [61]. Ca_V2.1 channels including NP are up to 11-fold less sensitive to ω-agatoxin IVA inhibition compared to Ca_V2.1 channels that lack NP. Ca_V2.1 mRNAs containing e31a are more abundant in Purkinje cells than in cerebellar granule cells; a result that nicely parallels the greater sensitivity of so called P-type currents, as compared to Q-type currents, to ω-agatoxin IVA [61]. The brain-specific splicing factor NOVA-2 represses the inclusion of e31a during alternative splicing of Ca_V2.1 pre-mRNA, and notably has the opposite action on the homologous e31a of Cacna1b [49]. NOVA-2 is known to be essential for coordinating cell-specific alternative splicing among a network of genes critical for synaptic function [91, 92].

6.3. Splice Isoforms of Ca_V2.2 have Different Sensitivities to G Protein Coupled Receptors

Ca_V2.2 channels underlie N-type currents that carry calcium entry into presynaptic terminals to control transmitter release from many synapses including primary nociceptive afferent synapses to dorsal horn neurons in spinal cord [93–98]. Ca_V2.2 channels are important drug targets that act directly to inhibit channel activity, as well as indirectly via G protein coupled receptors (GPCRs). Ziconotide (also known as ω-conotoxin MVIIC) is analgesic in humans and rodents when administered by intrathecal injection and it can prevent, as well as reverse symptoms of chronic pain induced by inflammation or nerve injury [99, 100]. In spinal cord, μ-opioid GPCRs inhibit presynaptic Ca_V2.2 channels and mediate intrathecal opioid analgesia [101–105]. We discovered that e37a, one of a pair of mutually exclusive exons the other being e37b, was expressed at higher levels in nociceptors compared to other neurons that primarily express e37b-containing Ca_V2.2 mRNAs [106]. We went on to show that Ca_V2.2 channels which contain e37a sequence exhibited different sensitivity to inhibition by GPCRs, including μ-opioid receptors, as compared to e37b-containing Ca_V2.2 isoforms [107]. By eliminating the contribution of e37a using either isoform-specific siRNAs or by removing e37a sequence from the Cacna1b mouse gene (replacing it with e37b; Fig. 4), we showed that e37a contributed to nociception and that it enhanced intrathecal morphine analgesia to thermal stimuli [108–110]. By contrast, ziconotide was equally effective at inhibiting e37a and e37b Ca_V2.2 splice isoforms and was an equally effective analgesic in vivo independent of e37a [109].

Fig. (4).

Exon substitution strategy employed to elucidate the functional and behavioral contributions of e37a and e37b of Cacna1b. Mutually exclusive e37a and e37b are the same length but about 30% of amino acids they encode differ. E37b was replaced by a second copy of e37a, in the e37a-only mouse strain, and e37a was replaced by a second copy of e37b, in the e37b-only mouse strain. E37a is repressed in most region of the nervous system but is found at higher levels in dorsal root ganglia, e37b is expressed throughout the nervous system. E37a does not substitute for e37b in the e37a-only mouse, there is increased e37 lacking Ca_V2.2 mRNA in neurons, and significantly reduced Ca_V2.2 protein levels. Whereas, in the e37b-only mouse, e37b is inserted in place of e37a and wild-type levels of Ca_V2.2 protein is preserved [108, 120].

7. METHODS FOR STUDYING THE BEHAVIORAL IMPORTANCE OF ALTERNATIVELY SPLICED EXONS

There has been relatively slow progress in discerning the biological and pharmacological significance of cell-specific alternative splicing of Cacna1 pre-mRNAs. This is more related to the technical challenges of separating the relative contributions of each Ca_V splice isoforms especially in vivo, and not the potential significance of their biological functions [2, 3]. Gene deletion and mutation studies in animal models, in conjunction with analyses of disease-causing mutations in human CACNA1 genes, have been essential in determining the biological functions of different Cacna1 genes [10, 111–117]. Interestingly, rare mutations in either one of a pair of alternatively spliced exons in CACNA1C (e8 or e8a) underlie the devastating Timothy Syndromes characterized by severe multi-organ disorders including lethal arrhythmias, webbing of fingers and toes, congenital heart disease, immune deficiency, intermittent hypoglycemia, cognitive abnormalities, and autism [86, 118, 119]. The different clinical manifestations of Type I and Type II Timothy Syndrome are correlated with the exon of CACNA1C, e8 or e8a, which carries the gain-of-function mutation [118]. Two strategies have been used to delineate the functional and behavioral consequences of different Ca_V2.2 splice isoforms in vivo: i) isoform-specific siRNA to achieve acute knockdown of a specific Ca_V protein [110]; and ii) eliminating one of a pair of alternative exons in the Cacna1b mouse gene generating novel mouse strains (Fig. 4) [108, 109]. The latter approach has the advantage of forcing inclusion of a particular exon (e37a or e37b) during splicing without disrupting intron sequence, and in the case of the e37b-only mouse, without affecting total Ca_V2.2 protein levels. We utilized e37b-only and wild-type mouse strains to extract the contribution of Ca_V2.2-e37a splice isoform in an otherwise normal background (wild-type levels of Ca_V2.2 current and protein expression, and no compensatory changes in non-Ca_V2.2 current or Ca_V2.1 protein) [108, 109, 120]. By comparing behavioral phenotypes we found evidence that intrathecal morphine analgesia to noxious thermal stimuli was reduced in mice lacking e37a (e37b-only) as compared to wild-type mouse strains [108, 109]. Similar analyses of other alternatively spliced exons should help establish their biological functions.

CONCLUSION

Voltage-gated calcium ion channels are essential for mediating depolarization-induced calcium entry into excitable cells, and they are important drug targets for treating several disorders including those of cardiovascular and nervous systems. Human Ca_V channel clones are routinely used to screening for new drugs to inhibit, or activate, calcium channels for use in humans. Even a single amino acid difference between human and non-human Ca_V proteins could affect drug efficacy. The amino acid composition of discrete regions of Ca_V channels, originating from cell-specific alternative splicing, can vary substantially across isoforms in different tissues. Intensive transcriptome sequencing projects for a range of human tissues are allowing quantification of expression patterns for each gene, including CACNA1 genes, as well as alternative exon usage according to tissue-type, age and disease state. Cell-specific alternative splicing can modify the pharmacological sensitivities of Ca_V channels to drugs, toxins and G protein coupled receptors. Knowledge of expression patterns of Ca_V isoforms in different human tissues and disease states should inform strategies aimed at developing the next generation of Ca_V channel inhibitors and agonists with improved tissue-specificity.

Biography

Diane Lipscombe