CALM1, CALM2, and CALM3 expression and translation efficiency provide insight into the severity of calmodulinopathy

Background Missense variants in the CALM1, CALM2, and CALM3 genes cause calmodulinopathy, an ultra-rare spectrum of clinical manifestations, including ventricular arrhythmias and sudden cardiac death. The three genes encode an identical calmodulin protein. However, little is known about the potential different clinical consequences of protein-altering variants in each of the three genes. Methods We compared variants from the International Calmodulinopathy Registry with population data from the Genome Aggregation Database (gnomAD). RNA sequencing data from the Genotype-Tissue Expression (GTEx) project was used to investigate the relative expression of CALM1, CALM2, and CALM3 among 49 different human tissues. We used ribosome profiling data of left ventricle tissue to estimate the percentage of calmodulin that could be attributed to each of CALM1 , CALM2 , and CALM3 . Results The percentage of carriers with cardiac events was different among carriers with missense variants in CALM1, CALM2, and CALM3 ( p = 0.004). Observed-to-expected ratios in population data were higher in CALM3 (0.29) in comparison to CALM2 (0.20) and especially, CALM1 (0.11) suggesting differential selection against missense variants in the three genes. RNA sequencing data revealed that CALM3 was less expressed than CALM1 and CALM2 in 39 of the 49 investigated tissues. Finally, the contribution to the total calmodulin amount in the left ventricle differed significantly among CALM1 , CALM2 , and CALM3 ( p < 2×10 -16 ), with estimated contributions of 45%, 44%, and 11%, respectively. Conclusions Our findings reveal a four-fold greater contribution of CALM1 and CALM2 to the total calmodulin levels in the human heart compared to CALM3 , suggesting why missense variants in CALM3 are associated with a less severe cardiac phenotype than those in CALM1 and CALM2 .