CALM3

CALM3 is one of three human genes encoding the identical 148-amino-acid calmodulin protein, a dumbbell-shaped Ca²⁺ sensor with four EF-hand domains that transduces calcium signals to regulate cardiac ion channels, neuronal mRNA trafficking, and cell differentiation (PMID:1474585, PMID:8314583). Despite being the most transcriptionally active CALM gene in some cell types, CALM3 contributes only ~11% of total calmodulin protein in the human left ventricle owing to lower translational efficiency, which correlates with milder clinical phenotypes in CALM3 variant carriers compared with CALM1 or CALM2 (PMID:9681195, PMID:41846582). Pathogenic CALM3 missense variants (E141G, A103V, N138K, D130G) reduce C-lobe Ca²⁺-binding affinity and dominantly dysregulate specific cardiac ion channels—impairing CaV1.2 inactivation, augmenting NaV1.5 late current, or activating RyR2-mediated Ca²⁺ release—causing long QT syndrome and catecholaminergic polymorphic ventricular tachycardia (calmodulinopathy), phenotypes rescuable by suppression-and-replacement gene therapy in patient iPSC-cardiomyocytes (PMID:26969752, PMID:27516456, PMID:35225649, PMID:39069900). In neurons, a retained 3′-UTR intron in CALM3 mRNA recruits Staufen2 for NMDA receptor activity-dependent dendritic localization, providing a mechanism for local calmodulin synthesis at synapses (PMID:28765142).

1. 1992 High

   Determination of the high-resolution crystal structure of calmodulin established the bilobed, four-EF-hand architecture that underpins all subsequent structure–function studies of CALM3-encoded protein.

   Evidence X-ray crystallography at 1.7 Å resolution of recombinant calmodulin

   PMID:1474585

   Open questions at the time

   • Structure solved without a bound target peptide, leaving the activation mechanism unclear
   • Conformational flexibility of the central linker helix in solution was not captured
2. 1993 High

   Mapping CALM3 to chromosome 19q13.2-q13.3 demonstrated that the three human calmodulin genes are dispersed across different chromosomes, raising the question of whether they are differentially regulated despite encoding an identical protein.

   Evidence PCR on human–hamster somatic cell hybrids and in situ hybridization on metaphase chromosomes

   PMID:8314583

   Open questions at the time

   • No information on differential tissue-level expression or translational output
   • Evolutionary forces maintaining three identical coding sequences were uncharacterized
3. 1998 High

   Quantification of CALM3 transcription revealed it is at least 5-fold more active than CALM1 or CALM2 in proliferating cells, establishing that the three genes contribute unequally to the calmodulin mRNA pool and that 5′-UTR elements drive this difference.

   Evidence Nuclear run-on, mRNA abundance measurement, and luciferase reporter assays in human teratoma cells

   PMID:9681195

   Open questions at the time

   • Transcriptional dominance measured in one cell type; generalizability to heart or brain unknown
   • Post-transcriptional and translational regulation were not assessed
4. 2016 High

   Functional characterization of CALM3 pathogenic variants (E141G, A103V) showed that different mutations target distinct cardiac ion channels—E141G dominantly impairs CaV1.2 inactivation and augments NaV1.5 late current (LQTS mechanism), while A103V activates RyR2-mediated spontaneous Ca²⁺ release (CPVT mechanism)—establishing a variant-specific functional hierarchy among calmodulin's channel targets.

   Evidence Ca²⁺-binding assays, whole-cell patch-clamp of CaV1.2 and NaV1.5, confocal Ca²⁺ spark/wave imaging and electrophysiology in cardiomyocytes

   PMID:26969752 PMID:27516456

   Open questions at the time

   • Structural basis of variant-specific channel selectivity (e.g., why E141G spares RyR2) was not resolved
   • In vivo cardiac phenotypes were not modeled
5. 2017 High

   Discovery that a retained 3′-UTR intron in CALM3 mRNA recruits Staufen2 for NMDA receptor activity-dependent dendritic transport revealed a neuron-specific post-transcriptional mechanism that could enable local calmodulin synthesis at activated synapses.

   Evidence iCLIP, FISH, Staufen2 knockdown, and NMDA stimulation/inhibition in neurons

   PMID:28765142

   Open questions at the time

   • Whether dendritic CALM3 mRNA is locally translated and whether this pool is functionally distinct from somatic calmodulin was not tested
   • Relevance to in vivo synaptic plasticity or neurological disease was not assessed
6. 2022 High

   The CALM3-N138K variant was shown to impair CaV1.2 inactivation and uniquely potentiate IKs current, providing a mechanistic basis for the milder and variable LQTS phenotype in a multigenerational family carrying this mutation.

   Evidence Stoichiometric and equilibrium Ca²⁺ titration; patch-clamp of ICaL and IKs; optical membrane expression assays

   PMID:35225649

   Open questions at the time

   • Structural mechanism by which N138K potentiates IKs while D130G does not was unexplained
   • Effect on other calmodulin targets (CaMKII, calcineurin) was not examined
7. 2024 Medium

   A triple-CALM suppression-and-replacement gene therapy construct rescued pathologically prolonged APD in patient iPSC-cardiomyocytes carrying CALM3-D130G, demonstrating that a single WT CALM1 cDNA can functionally compensate for all three CALM genes and validating gene therapy as a potential calmodulinopathy treatment strategy.

   Evidence shRNA knockdown (RT-qPCR) and voltage-sensing dye APD90 measurement in patient-derived iPSC-cardiomyocytes

   PMID:39069900

   Open questions at the time

   • Efficacy demonstrated only in iPSC-CMs; in vivo delivery and long-term safety are untested
   • Whether chronic knockdown of endogenous CALM3 transcript affects neuron-specific functions (e.g., dendritic localization) was not addressed
8. 2024 Medium

   CALM3 knockdown partially reverses neuronal differentiation of mesenchymal stem cells and is negatively regulated by miR-543, placing CALM3 in a regulatory axis controlling neurogenesis.

   Evidence siRNA knockdown of CALM3 and miR-543 mimic transfection with RT-qPCR and western blot readout in hUC-MSCs

   PMID:39444227

   Open questions at the time

   • Single-study finding in a non-neuronal progenitor system; relevance to physiological neurogenesis is unconfirmed
   • Whether the effect is CALM3-specific or reflects total calmodulin levels was not distinguished
9. 2026 Medium

   Ribosome profiling of human left ventricle revealed CALM3 contributes only ~11% of total calmodulin protein despite substantial mRNA levels, explaining both its lower pathogenic penetrance and weaker negative selection compared to CALM1 and CALM2.

   Evidence Ribosome profiling and RNA-seq from GTEx left ventricular tissue; gnomAD constraint analysis; International Calmodulinopathy Registry clinical correlation

   PMID:41846582

   Open questions at the time

   • Translational efficiency measured in bulk left ventricle; cell-type-resolved (cardiomyocyte vs. fibroblast) data are lacking
   • Cis-regulatory elements responsible for lower translational output have not been identified

• Key open questions include the structural basis for variant-specific ion channel selectivity, whether dendritically localized CALM3 mRNA is locally translated to support synaptic plasticity, and the in vivo efficacy of suppression-and-replacement gene therapy for calmodulinopathy.
• No co-crystal structures of calmodulinopathy mutant CaM with CaV1.2, RyR2, or KCNQ1
• Local translation from dendritic CALM3 mRNA has not been demonstrated
• In vivo gene therapy efficacy and safety data are absent