CACNA1I

CACNA1I encodes CaV3.3, the pore-forming α-subunit of a T-type (low-voltage-activated) calcium channel whose distinctively slow gating kinetics — determined by multiple transmembrane domains with domain IV as the principal contributor — generate low-threshold calcium spikes, rebound burst firing, and window-current-driven membrane potential oscillations that are essential for sleep spindle rhythmogenesis in thalamic reticular nucleus (TRN) neurons (PMID:26612388, PMID:32066662, PMID:16996222, PMID:16706840). CaV3.3 surface expression and current amplitude are regulated by Gαq/11-coupled muscarinic receptors acting through two channel-intrinsic regions (PMID:17535809), by endogenous polyunsaturated lipids competing at a shared site with synthetic T-channel antagonists (PMID:24214826), by neuritin signaling via the insulin receptor/MEK-ERK pathway (PMID:28475719), and by TET1-mediated DNA hydroxymethylation of the CACNA1I locus (PMID:36370755). Gain-of-function missense variants at residues I860, I1306, M1425, and A398 slow channel gating, increase window current, and enhance neuronal excitability — correlating with seizure phenotypes — while the schizophrenia-associated R1346H variant reduces glycosylation and surface expression, diminishing CaV3.3 current density sufficiently to abolish rebound bursting and disrupt sleep spindles in vivo (PMID:33704440, PMID:27756899, PMID:32066662, PMID:40825030).

1. 2003 Medium

   Establishing that CaV3.3 protein exists as developmentally regulated, differentially glycosylated isoforms with region-specific brain expression provided the first protein-level characterization beyond the cloned cDNA.

   Evidence Anti-peptide antibody Western blotting and immunohistochemistry across rodent and human brain regions and developmental stages

   PMID:12614673

   Open questions at the time

   • Antibody-based detection without independent mass-spectrometry validation
   • Functional significance of neonatal vs. adult isoforms unknown
   • No link to channel electrophysiology
2. 2004 High

   Systematic chimeric analysis between CaV3.1 and CaV3.3 revealed that CaV3.3's hallmark slow gating kinetics arise from distributed structural determinants across multiple domains rather than a single transferable module, while alternative splicing of exons 9 and 33 modulates gating and burst firing properties.

   Evidence Chimeric channel electrophysiology in Xenopus oocytes and tsA-201 cells; RT-PCR cloning of splice variants from human brain with patch-clamp characterization

   PMID:15016809 PMID:15254077

   Open questions at the time

   • Structural basis at atomic resolution unknown
   • Which splice variants predominate in TRN neurons not determined
3. 2006 High

   Domain IV was identified as the principal determinant of CaV3.3 half-activation potential, activation kinetics, and recovery from inactivation, resolving the earlier finding that no single domain was sufficient by showing graded contributions with domain IV dominant; separately, CaV3.3 window current was shown to drive intrinsic membrane potential and Ca²⁺ oscillations.

   Evidence Systematic domain-swap chimeras with patch-clamp in tsA-201 cells; CaV3.3 overexpression in NG108-15 cells with Ca²⁺ imaging

   PMID:16706840 PMID:16996222

   Open questions at the time

   • Molecular identity of domain IV residues conferring slow kinetics not pinpointed
   • Window current contribution not tested in native TRN neurons
4. 2007 High

   Two regulatory mechanisms were defined: Gαq/11-coupled muscarinic receptors selectively inhibit CaV3.3 (not CaV3.1/3.2) through two channel-intrinsic regions, and neonatal CaV3.3 carries polysialic acid modifications that fully account for developmental molecular weight differences.

   Evidence Chimeric channel analysis with genetically encoded Gα/Gβγ antagonists; PNGase F and Endo-N enzymatic deglycosylation and Western blotting

   PMID:17317015 PMID:17535809

   Open questions at the time

   • Identity of channel residues mediating muscarinic inhibition unknown
   • Functional impact of polysialylation on channel gating or trafficking not tested
5. 2013 High

   Identification of a shared binding site on CaV3.3 for endogenous polyunsaturated lipids and synthetic T-channel antagonists revealed that the channel is a direct lipid sensor, establishing a molecular mechanism for endogenous modulation.

   Evidence Radioligand displacement assay with [³H]TTA-A1 on CaV3.3-expressing membranes combined with patch-clamp electrophysiology across multiple lipid species

   PMID:24214826

   Open questions at the time

   • Binding site not mapped to specific channel residues
   • Physiological relevance of lipid modulation in TRN neurons not demonstrated in vivo
6. 2016 High

   CaV3.3 was established as the dominant T-type channel for TRN burst firing and sleep spindle generation: the schizophrenia-associated R1346H variant reduced glycosylation and surface expression to suppress current density by ~50%, and CaV3.2/CaV3.3 double knockout abolished low-threshold bursting and fragmented NREM sleep.

   Evidence Biochemistry and patch-clamp in human cell lines with NEURON modeling (R1346H); patch-clamp in brain slices and polysomnography from CaV3.2/CaV3.3 double-KO mice

   PMID:26612388 PMID:27756899

   Open questions at the time

   • Whether R1346H acts purely via reduced surface expression or also affects trafficking pathway
   • Relative contribution of CaV3.3 vs. CaV3.2 to spindle generation not fully separated
7. 2017 Medium

   Neuritin was identified as an upstream regulator that increases CaV3.3 surface expression via insulin receptor/MEK-ERK signaling, linking extracellular trophic signals to T-type channel-dependent glutamate release in medial prefrontal cortex.

   Evidence Western blotting of membrane fractions, patch-clamp, HPLC glutamate measurement in mPFC slices with pharmacological pathway dissection

   PMID:28475719

   Open questions at the time

   • No direct CaV3.3 mutagenesis to confirm subtype specificity
   • Mechanism of ERK-dependent surface trafficking not defined
   • Single-lab finding
8. 2020 High

   In vivo validation that the R1346H variant specifically disrupts sleep spindles came from knock-in mice showing altered TRN excitability and marked spindle deficits during NREM sleep, while CaV3.3 haploinsufficiency alone was insufficient — demonstrating that the variant acts as more than simple loss-of-function.

   Evidence CaV3.3 R1346H knock-in and heterozygous KO mice with in vivo polysomnography and ex vivo TRN patch-clamp

   PMID:32066662

   Open questions at the time

   • Molecular mechanism distinguishing R1346H from haploinsufficiency not resolved
   • Whether spindle deficits cause cognitive/psychiatric phenotypes not tested
9. 2021 High

   Gain-of-function neurodevelopmental variants at I860, I1306, and M1425 were shown to stabilize hydrogen bonds in the channel gate, slow gating transitions, increase window current, and shift TRN model neurons from rebound bursts to slow oscillations — establishing a gain-of-function disease mechanism distinct from R1346H loss-of-function.

   Evidence Patch-clamp in HEK293T cells, structural homology modeling, site-directed mutagenesis, computational TRN modeling, chromaffin cell recordings

   PMID:33704440

   Open questions at the time

   • No in vivo knock-in model for gain-of-function variants
   • Downstream consequences for thalamocortical circuit and sleep architecture not tested
10. 2022 Medium

    Epigenetic regulation of CACNA1I by TET1-mediated DNA hydroxymethylation was demonstrated, and rare hemiplegic migraine-associated variants were functionally characterized, broadening the regulatory and disease landscape of CaV3.3.

    Evidence MeDIP/hMeDIP with TET1 overexpression/knockdown in Leydig cells; patch-clamp of five migraine-associated variants in HEK293T cells

    PMID:35928792 PMID:36370755

    Open questions at the time

    • TET1 regulation shown only in non-neuronal cells
    • Migraine variants not validated in neuronal systems or in vivo
    • Causal link between CaV3.3 dysfunction and migraine pathophysiology not established
11. 2025 High

    Position A398 was revealed as a hotspot where different substitutions cause opposite functional effects (A398E gain-of-function vs. A398V loss-of-function), and genotype–phenotype correlation across multiple variants established that presence or absence of increased neuronal excitability in silico predicts seizure occurrence in patients.

    Evidence Site-directed mutagenesis, voltage-clamp electrophysiology, structural homology modeling, computational neuronal excitability modeling with clinical genotype-phenotype correlation

    PMID:40825030

    Open questions at the time

    • No in vivo animal model for A398 variants
    • Structural basis of opposite effects at same residue not resolved at atomic level

• A high-resolution cryo-EM structure of CaV3.3, the precise residues mediating muscarinic receptor inhibition and lipid binding, the trafficking pathway by which R1346H differs from haploinsufficiency, and in vivo validation of gain-of-function variants in thalamocortical circuits remain unresolved.
• No experimental 3D structure of CaV3.3
• Muscarinic inhibition and lipid binding sites not mapped to specific residues
• In vivo models for gain-of-function variants lacking
• How CaV3.3 dysfunction translates to psychiatric and cognitive phenotypes is mechanistically unclear