Contribution of CACNA1H Variants in Autism Spectrum Disorder Susceptibility

Abstract

Autism Spectrum Disorder (ASD) is a highly heterogeneous neuropsychiatric disorder with a strong genetic component. The genetic architecture is complex, consisting of a combination of common low-risk and more penetrant rare variants. Voltage-gated calcium channels (VGCCs or Ca_v) genes have been implicated as high-confidence susceptibility genes for ASD, in accordance with the relevant role of calcium signaling in neuronal function. In order to further investigate the involvement of VGCCs rare variants in ASD susceptibility, we performed whole genome sequencing analysis in a cohort of 105 families, composed of 124 ASD individuals, 210 parents and 58 unaffected siblings. We identified 53 rare inherited damaging variants in Ca_v genes, including genes coding for the principal subunit and genes coding for the auxiliary subunits, in 40 ASD families. Interestingly, biallelic rare damaging missense variants were detected in the CACNA1H gene, coding for the T-type Ca_v3.2 channel, in ASD probands from two different families. Thus, to clarify the role of these CACNA1H variants on calcium channel activity we performed electrophysiological analysis using whole-cell patch clamp technology. Three out of four tested variants were shown to mildly affect Ca_v3.2 channel current density and activation properties, possibly leading to a dysregulation of intracellular Ca²⁺ ions homeostasis, thus altering calcium-dependent neuronal processes and contributing to ASD etiology in these families. Our results provide further support for the role of CACNA1H in neurodevelopmental disorders and suggest that rare CACNA1H variants may be involved in ASD development, providing a high-risk genetic background.

Keywords: ASD; CACNA1H; Cav3.2; VGCCs; calcium channel; rare variants.

Figure 1

Family 22 and 105 CACNA1H variants. (A) Family segregations and gene location of CACNA1H biallelic variants. Filled shapes indicate ASD individuals. *DNA was not available for individual 105.4. UCSC hg38 Genome Browser screenshot shows the location of biallelic variants within the CACNA1H gene. (B) Schematic of the CACNA1H protein channel (Ca_v3.2). Ca_v3.2 channel consists of the single α1 pore-forming subunit of about 260 kDa, organized in four homologous domain each composed of six transmembrane segments (S1–S6). Within each domain, the arginine/lysine-rich S4 segment represents the voltage-sensing region of the channel, while the extracellular loop linking S5 and S6 segments (P loop) ensures the ion conductivity and selectivity of the channel (14, 18). Protein visualization was generated using Protter–visualize proteoforms (19).

Figure 2

Immunofluorescence (IF) assay of 3xFLAG-Ca_v3.2. IF assay was performed in HEK-293T cells, by transiently transfecting cells with p3xFLAG-CACNA1H plasmid constructs. An empty vector (EV) was used as negative control, while 3xFLAG-ABCC3 transmembrane fusion protein was used as positive control. Mouse anti-FLAG M2 antibody was used for detection of the recombinant proteins. Goat anti-mouse Cy3 antibody (red signal) and Hoechst dye (blue signal) were employed to detect anti-FLAG antibody and nuclei respectively.

Figure 3

Lys785Met is a gain of function Ca_v3.2 subtype. (A) Representative current traces recorded from two HEK-293 cells transiently transfected with the WT (left) or Lys785Met (right) CACNA1H plasmid construct. The currents were elicited by step depolarizations from a holding potential of −120 mV to various test potentials. (B) Representative current traces recorded from three HEK-293 cells transiently transfected with the Pro2124Leu (left) or Ser2338Phe (middle) or Pro849Ser (right) plasmid construct, same protocol as (A). (C) Mean activation curves for WT, Pro2124Leu and Ser2338Phe plasmid constructs transfected cells, as indicated. Solid lines represent data fit to the Boltzmann equation (Vh values are −45.5, −49.3 and −43.4 mV for WT, Pro2124Leu and Ser2338Phe, respectively). Data were averaged from 16, 19 and 17 cells, for WT, Pro2124Leu and Ser2338Phe, respectively. (D) Mean activation curves for WT (same data as C), Pro849Ser and Lys785Met plasmid constructs transfected cells, as indicated. Solid lines represent data fit to the Boltzmann equation (Vh values are −45.5, −49.5 and −50.1 mV for WT, Pro849Ser and Lys785Met, respectively). Data were averaged from 16 (same cells as C), 16 and 14 cells, for WT, Pro849Ser and Lys785Met, respectively. (E) Normalized mean activation curves for WT, Pro2124Leu and Ser2338Phe plasmid constructs transfected cells, same data as (C). Solid lines represent data fit to the activation Boltzmann equation. (F) Normalized mean activation curves for WT, Pro849Ser and Lys785Met plasmid constructs transfected cells, same data as (D). Solid lines represent data fit to the activation Boltzmann equation. (G) Inactivation open probability-voltage relationships for WT, Pro2124Leu and Ser2338Phe plasmid constructs transfected cells. Solid lines represent data fit to the inactivation Boltzmann equation (Vi values are −78.9, −79.5 and −78.2 for WT, Pro2124Leu and Ser2338Phe, respectively). No significant differences were detected. (H) Inactivation open probability-voltage relationships for WT, Pro849Ser and Lys785Met plasmid constructs transfected cells. Solid lines represent data fit to the inactivation Boltzmann equation (Vi values are −78.9, −75.3 and −77.1 mV for WT, Pro849Ser and Lys785Met, respectively). No significant differences were detected.

Figure 4

Activation and kinetic parameters of Lys785Met, Ser2338Phe and Pro849Ser Ca_v3.2 subtypes are different from WT. (A) Histogram representing the mean conductance density of WT and mutant channels expressed in HEK-293 cells, measured at −40 mV test potential for each Ca_v3.2 subtype, as indicated. Mean values were averaged from 16, 15, 13, 12 and 10 cells, from left to right. The conductance density was significantly higher for Lys785Met mutant, as compared to WT (a, p < 0.001). (B) Histogram representing the mean Vh value measured at −40 mV test potential for each Ca_v3.2 subtype, as indicated. Same cells as (A). The Vh value was significantly higher for Lys785Met mutant, as compared to WT (b, p = 0.037). (C) Left, histogram representing the mean values of time to peak measured at −40 mV test potential for each Ca_v3.2 subtype, as indicated. Mean values were averaged from 10, 13, 10, 18 and 15 cells, from left to right. The time to peak was significantly higher for Ser2338Phe, Pro849Ser and Lys785Met mutants, as compared to WT (c, p = 0.002; d, p = 0.002; e, p = 0.003). (D) Histogram representing the mean values of exponential τ decay measured at −40 mV test potential for each Ca_v3.2 subtype, as indicated. Same cells as (C). The time to peak was significantly higher for Pro849Ser mutant, as compared to WT (f, p = 0.039). (E) Histograms representing the mean values of charge transfer measured at −40 mV test potential for each Ca_v3.2 subtype, as indicated, at different current times (left, 2 ms; center, 10 ms; right, 100 ms). Same cells as (C). The charge transfer value was significantly higher only for Lys785Met mutant, at each time point, as compared to WT (g, p = 0.004; h, p = 0.010; i, p < 0.001).