Brain voltage-gated Ca²⁺ channels and calmodulin: Teaming up in calmodulinopathies?

The International Calmodulinopathy Registry provides valuable information on clinical and genetic data from subjects with CALM variants. To date, 140 subjects have been enrolled, including 97 index cases and 43 family members (Crotti et al., 2023) with 59 distinct amino acid substitutions in the three CALM genes. The major phenotypes are long-QT syndrome (LQTS) (53%), catecholaminergic polymorphic ventricular tachycardia (CPVT) (26%), and LQTS/CPVT overlap (7%). Adrenergic stimulation was the major trigger of cardiac events in the entire population, irrespective of the phenotype. Interestingly, 20 subjects also had a primary neurological/neurodevelopmental disorder. This raises the important question of additional extracardiac manifestations of calmodulinopathies.

Pathogenic mutations are primarily located in the Ca²⁺-binding EF-hand motifs III and IV, comprising the C-terminal lobe of calmodulin (CaM). The C-lobe binds Ca²⁺ with a higher affinity than the N-lobe. Both lobes can mediate different effects on a target protein, as also found for Ca²⁺-channel modulation (Hussey et al., 2023, see below).

Over the last ∼15 years (Nyegaard et al., 2012; Crotti et al., 2013), a mechanistic genotype–phenotype correlation has emerged. CALM variants associated with LQTS prolong the cardiac action potential in transfected or human iPSC-derived cardiomyocytes, which prolongs the QT interval (Limpitikul et al., 2017; Yamamoto et al., 2017). This is explained by the impaired inactivation of Cav1.2. Cav1.2 is the predominant L-type Ca²⁺ channel in the working myocardium (Zamponi et al., 2015) and triggers Ca²⁺-induced Ca²⁺-release of adjacent RYR2 ryanodine receptors within the same nanodomain in the junctional sarcoplasmic reticulum (reviewed by Dixon, 2021). The C-lobe of CaM enables an essential Ca²⁺-dependent autoinhibitory feedback mechanism of Cav1.2, Ca²⁺-dependent inactivation (CDI): When action potential-triggered Cav1.2 Ca²⁺ entry and subsequent SR Ca²⁺ release increase nanodomain Ca²⁺ (reviewed by Bers and Perez-Reyes, 1999), Ca²⁺ occupies apo-CaM, which is pre-associated with the channel at resting free Ca²⁺ concentrations by binding to an IQ motif within the channel’s C terminus (Limpitikul and Dick, 2025). Ca²⁺/CaM accelerates channel inactivation and reduces Ca²⁺ flux during an action potential (Bers and Perez-Reyes, 1999). Because LQTS CaM mutations reduce Ca²⁺-affinity for the C-lobe, they are unable to support CDI, permitting excess Ca²⁺ entry.

Notably, these mutations do not prevent the pre-association of the mutated apo-CaM to Cav1.2 (for reviews see Jensen et al., 2018; Tsai et al., 2021; Crotti et al., 2023; Hussey et al., 2023; Schwartz et al., 2024; Gao et al., 2026). This pre-association answers two key questions:

• (1)

  Heterozygous mutations in one of the three CALM genes affect only one out of their six alleles. How can this cause an apparently dominant effect (i.e., inhibition of CDI), despite an apparent loss-of-function (i.e., decreased CaM Ca²⁺ binding) in the presence of a (predicted) excess of unmutated CaM protein?

• (2)

  Why is the clinical phenotype limited to cardiac arrhythmia in most patients, despite the key involvement of CaM in so many physiological processes?

The key answer to both questions is that the mutant CaM, like WT CaM, can still pre-associate as apo-CaM with Cav1.2. This allows a small fraction of mutant CaM proteins to prevent CDI in a subset of channels. This is sufficient to significantly affect Cav1.2 function (Hussey et al., 2023; Schwartz et al., 2024; Limpitikul et al., 2017). In contrast, other CaM-modulated proteins that bind CaM primarily in its Ca²⁺-bound form would preferentially bind free WT Ca²⁺/CaM because the activation of mutant CaM is compromised due to their reduced Ca²⁺ affinity. CaM pre-association is observed for many ion channels, including several voltage-gated Ca²⁺, Na⁺, and K⁺ channels and various TRP channels (Alemayhu and Paulsen, 2026). In contrast, no pre-association is observed for many other CaM targets, including many GPCRs, connexins, and CaM-dependent enzymes (Andrews et al., 2020).

In addition to CaM interactions with voltage-gated channels, CaM can also modulate RyR2 by direct interaction. Both apo-CaM and Ca²⁺/CaM bind to RyR2 (Gao et al., 2026). The CaM-mediated decrease of RyR2 open probability appears stronger at low intracellular Ca²⁺ and helps with the termination of SR Ca²⁺-release. Therefore, CaM serves as a constitutive Ca²⁺ sensor that inhibits RyR2 Ca²⁺ release in a Ca²⁺-dependent manner to support the termination of Ca²⁺ release and to maintain a low RyR2 activity during diastole (Gao et al., 2026).

There is substantial experimental evidence that mutant CaM proteins causing CPVT induce a disinhibition of RyR2 channels, which facilitates especially spontaneous SR Ca²⁺ release at high ER Ca²⁺ levels (Sondergaard et al., 2019), driving arrhythmias. CPVT is often observed in subjects with LQTS mutations (Crotti et al., 2023). Only four pathogenic mutations were reported to cause only CPVT without LQTS. In these mutants C-lobe Ca²⁺-binding affinity is unaltered or decreased only weakly (Gomez-Hurtado et al., 2016; Sondergaard et al., 2019; Gao et al., 2026).

Taken together, this provides strong evidence that CaM missense mutations in one of six CALM alleles induce disease-relevant changes in Ca²⁺-homeostasis by enhancing Ca²⁺ signals of these two functionally coupled Ca²⁺ channels, especially under conditions in which β-adrenergic receptor stimulation further enhances their activity.

Neurological/neurodevelopmental symptoms associated with (mostly LQTS associated) CaM mutations in the International Calmodulinopathy Registry include autism, ADHD, developmental delay, intellectual disability, hyperactivity, and seizures. This raises the important question about molecular targets dysregulated by mutant CaMs in the brain. Given the preferential modulation by mutant CaMs in the heart, voltage-gated L-type Ca²⁺ channels and RYR2 are again obvious targets to consider. Both regulate neuronal functions, including synaptic plasticity, learning, memory, and dendritic spine remodeling (Zamponi et al., 2015; Bertan et al., 2020; Striessnig et al., 2022).

Although CPVT mutations may affect neuronal function through RyR2 dysfunction (Gao et al., 2023), LQTS mutations are more likely candidates because they not only affect the CDI of L-type Ca²⁺ channels but, through their reduced C-lobe Ca²⁺ affinity, should also affect Cav2.1 channels, a major presynaptic Ca²⁺ channel controlling fast neurotransmitter release (Zamponi et al., 2015). Although their CDI is modulated by the N-lobe of CaM, upregulation of the activity of Cav2.1 following strong depolarizations, also known as Ca^2+-dependent facilitation (CDF), is controlled by the C-lobe (Hussey et al., 2023). CDF allows presynaptic Cav2.1 channels to contribute to short-term synaptic plasticity and the CDF of synaptic transmission in response to repeated trains of depolarizing stimuli (Lee et al., 1999).

In the brain Cav1.3 L-type channels, expressed at lower levels than neuronal Cav1.2 channels (Zamponi et al., 2015), are also obvious candidates. Like Cav1.2, Cav1.3 is also pre-associated with apo-CaM and primarily located in somatodendritic postsynaptic compartments. Both support dendritic Ca²⁺-spike propagation, activity-dependent gene transcription, and shape neuronal firing. Through their different gating properties they contribute in different ways to normal neuronal development, synaptic pruning, and various forms of learning and memory (Zamponi et al., 2015; Striessnig et al., 2022). Moreover, missense mutations facilitating voltage-dependent channel opening and/or inhibiting inactivation cause abnormal brain function, both in mice (Bader et al., 2011; Ortner et al., 2023) and humans (Striessnig, 2021). Such Cav1.2 mutations (CACNA1C gene) not only cause LQTS but also neurological/neurodevelopmental symptoms, including neurodevelopmental delay and autism (Striessnig, 2021; Herold et al., 2023). In Cav1.3 (CACNA1D, Fig. 1) they lead to a neurodevelopmental disorder, including autism, autoaggressive behaviors, developmental delay, and sometimes also endocrine symptoms (Ortner et al, 2023). Cav1.3 does not contribute to QT prolongation because it is not expressed in adult ventricular myocytes. However, in both mice and humans, Cav1.3 expression contributes to normal sinoatrial node pacemaking and AV-node conduction (Zamponi et al., 2015).

The importance for CDI fine-tuning of Cav1.3 is evident from the fact that alternative splicing generates full-length (Cav1.3_L) and C-terminally short variants (Cav1.3_S), which differ with respect to CaM binding and the strength of CDI (Zamponi et al., 2015). Moreover, depending on cellular function, Cav1.3 CDI can be turned off completely by alternative splicing or CaM-like Ca²⁺-binding proteins (Zamponi et al., 2015; Limpitikul and Dick, 2025).

In their comprehensive analysis in this issue of the JGP (Hussey et al., 2026) investigated the modulation of CDI in Cav1.3 and CDF in Cav2.1 by heterologous expression of channels and mutant CaM in HEK-293 cells, followed by whole-cell patch-clamp recordings. To interpret these findings, they determined changes in the affinity of mutant CaM for binding to a peptide containing the channels’ IQ domains using live-cell flow cytometric FRET. Finally, they investigated the Ca²⁺ dependence of CaM interaction with the IQ domains under low and high Ca²⁺ conditions with a fluorescence anisotropy-binding assay.

They investigated C-lobe mutations strongly associated with LQTS (D96V, F142L, D130G, and D132G) or with sudden death and cardiac arrest (G114R) previously shown to blunt CDI in Cav1.2. N-lobe mutation N54I, which is associated with CPVT without LQTS and does not affect CDI, was also included.

CDI is measured by comparing the inactivation of the channel with Ba²⁺ (no CDI) or Ca²⁺ (CDI) as the conducting ion. All mutants except for N54I reduced CDI in both Cav1.3_L and Cav1.3_S by about 60%. Incomplete inhibition of CDI in Cav1.3 is consistent with the previous finding that, in contrast to Cav1.2, a small fraction of CDI is only inhibited by the CaM N-lobe (Hussey et al., 2026). G114R was unable to affect CDI in Cav1.3_S, which suggested that it loses binding affinity and, in contrast to the other mutants, cannot efficiently compete with endogenous CaM. This was confirmed in their live-cell FRET-measurements; impaired G114R pre-association was only partly rescued by high intracellular Ca²⁺.

Fluorescence anisotropy confirmed a 10-fold reduced affinity of apo-G114R for the Cav1.3 IQ domain, an increased affinity for all mutants in Ca²⁺-bound conditions and reduced Ca²⁺ sensitivity for binding to the channel except, as expected, for N54I.

The same approach was used to monitor changes of CDF modulation of Cav2.1. While mutants D130G, D132G, and F142L strongly impaired CDF, inhibition was absent for D96V and minimal for G114R. This could not be readily explained by the small reductions in apo-CaM affinity or the reduced Ca²⁺ dependency of the CaM/Cav2.1 IQ interaction. The absence of an effect of D96V shows that dysregulation by CaM mutations can be channel specific. The previously shown major distortion of the C-terminal lobe architecture in D96V that alters its C-lobe interaction with the IQ domain (Wang et al., 2018) may contribute to this difference.

In this series of elegant experiments, the authors convincingly demonstrated that CaM mutants can affect the autoregulatory fine-tuning of neuronal Ca²⁺ channels. However, the potential implications for neurodevelopmental symptoms in calmodulinopathy subjects remain speculative. Limitations include the broad repertoire of brain CaM targets and the artificial nature of heterologous expression systems, which cannot accurately model a native environment, and the predicted lower abundance of mutant CaM compared with WT. Although it has been convincingly shown that effects on Cav1.2 and RYR2 function can also be detected when mutant proteins represent only half (Hussey et al., 2026) or even much less of the total CaM protein (Limpitikul et al., 2017; Schwartz et al., 2024), actual mutant expression levels across different cells remain largely unknown. Surprisingly, a recent study in Calm1^N98S knockin mice found that high levels of mutant protein accumulated in the hearts of heterozygous animals, even exceeding WT levels, despite very low mutated transcript levels (Tsai et al., 2025).

While iPSC-derived human cardiomyocytes have been successfully used to characterize CaM-mutant pathology, this has not yet been extended to neurons. The same is true for CaM-mutation knockin mice, which focused on the characterization of the severe cardiac phenotype (Tsai et al., 2025). Studies with iPSC neurons and mouse knockin models should focus on CaM mutations for which a primary, not hypoxia-associated, neurological/neurodevelopmental phenotype can be unambiguously validated and which exert strong effects on CaM modulation of Cav1.2, Cav1.3, and Cav2.1 channels in heterologous systems.

The study by Hussey et al. represents an important first step toward a better understanding of how a calmodulinopathy can become a neuronal “channelopathy” and cause human central nervous system diseases.

Christopher J. Lingle served as editor.

This work was funded in whole or in part by the Austrian Science Fund (FWF, DOI 10.55776/P35722).

For open access purposes, the author has applied a CC-BY public copyright license to the author-accepted manuscript version arising from this submission.

Author contributions: Jörg Striessnig: conceptualization, funding acquisition, project administration, supervision, validation, and writing—original draft, review, and editing.