Burden of Rare Sarcomere Gene Variants in the Fra-mingham and Jackson Heart Study Cohorts

Summary

The aim of the study was to evaluate whether allelic variants in eight sarcomeric genes are asso-ciated with cardiac morphology and function in the community. Alexander Bick and coworkers sequenced 3,600 unrelated individuals from the Framingham Heart Study (FHS) and Jackson Heart Study (JHS) cohorts. FHS is a three-generation prospective, community –based family study to identify cardiovascular risk factors. The individuals originally recruited in 1948 were primarily of European ancestry and had not yet developed cardiovascular disease. The FHS Offspring cohort investigated in this study comprised the original participants’ adult children and spouses. JHS is a community-based observational study. The participants were recruited from Jackson, Mississippi, metropolitan statistical area and the study population included also individuals of African American origin.

Sarcomere genes ACTC1 (MIM 102540), MYBPC3 (MIM 600958), MYH7 (MIM 160760), MYL2 (MIM 160781), MYL3 (MIM 160790), TNNI3 (MIM 191044), TNNT2 (MIM 191045), and TPM1 (MIM 191010) were sequenced. Targeted genes were selected with a custom designed hybrid capture array and sequenced with an Illumina HiSeq. Sequences were aligned to human genome assembly hg19 with Burrow-Wheeler Alligner and the recalibrated with the Genome Analysis Toolkit (GATK). The pathogenicity of the variants was evaluated using Clinical Laboratory Improvement Amendments (CLIA)-approved laboratory variant calling algorithm and the PolyPhen HCM algorithm. The variants were classified as pathogenic, benign or having unknown significance.

In the two cohorts studied, 11,2% of individuals turned out to have one or more rare (MAF <1%) nonsynonymous variants (missense, nonsense, indels, and splice variants). In the FHS cohort three individuals and in the JHS study population 19 subjects had multiple rare nonsynonymous sarcomere variants. The frequency of variants for proportional to protein length, except for MYBPC3 and MYL3, which were overrepresented (p<0.0001). The prevalence of individuals rare known or likely pathogenic sarcomeric variants was 14 of 1,637 FHS study participants and 8 of 1,963 sequenced JHS participants. The estimated prevalence of known likely pathogenic sarcomeric variants was 0.85% among European Americans and 0.4% among African Americans.

The investigators also selected a set of 30 comparably sized genes for negative controls to the permutation analyses. The chosen genes are known or presumed to be pathogenic either for cardiac electrophysiology or anthropomorphic features but they are not known to affect cardiac dimensions.
The reported p-values were robust to this analysis.

The possible effect of sarcomeric mutations was assessed by using data from previous echocardio-graphies (left ventricular diastolic thickness LVWT, left ventricular diastolic diameter LVDD, left atrial diameter LAD, and fractional shortening FS). The authors report significant increases in LVWT and LAD among subjects carrying one or more rare sarcomeric variant. LVWT was increased by 5% in the FHS cohort (p<0.01) and by 2% in the JHS cohort (p< 0.05). Individuals with sarcomere variants of know or likely pathogenicity had an 8.6% increased wall thickness in the FHS population (p<0.05) but not in the JHS cohort. LAD was increased by 3% in FHS and JHS participants (p<0.05, FHS; p<0.01 JHS).

22 individuals carried variants that in the literature were considered likely pathogenic. Only four of these subjects fulfilled clinical criteria of HCM, two had LVWT > 12mm and two ECG criteria for LVH.

Bick and coworkers report that individuals with one or more nonsynonymous sarcomeric variants either with or without physiologic risk factors had significantly earlier onset of first cardio-vascular events (hazard ratio: 2.3) in the FHS cohort. The authors suggest that cardiovascular risk assessment in the general population could benefit from rare variant analysis.

Comments

Until recently, due to the large size of many of the genes associated with cardiomyopathies, much of the data concerning the genotype-phenotype relations has been obtained from family studies. The developments of DNA sequencing techniques and data analysis have made it possible to search for and evaluate rare variants in large genes in large study populations. The observed prevalence of the rare sarcomeric variants was 0.6% in the general population, which is about twice the estimated prevalence of HCM. Previously, it has been reported that also desmosomal mutations that may predispose to ARVC and its clinical sequelae were more prevalent in the Finnish population ( approximately 1:200) than expected based on the published ARVC prevalence  (1:1,000-1:5,000) data. 

 

Conclusion:

In this study carried out by Alexander Bick and coworkers, the sequence data could be combined with long follow-up clinical data collected from the study participants. The present study adds to our knowledge on the prevalence and  the significance of rare sarcomeric variants in the general population. It seems that the clinical cardiomyopathy that we see in our patients may be like the tip of the iceberg reflecting only the easily visible part of genetically determined cardiac pathogenicity.