9 CHAPTER 9 200 Multiple-variant burden testing of CACNA1x genes Considering the results from the individual variant analysis, a multiple-variant burden analysis was performed to test whether there is an over-representation of CACNA1x missense variants in HM compared to controls. As shown in Table 2, in the Australian HM cohort this analysis revealed a significantly increased burden of missense variants for those with HM in CACNA1E (p = 1.3 x 10-4), CACNA1H (p < 2.2 x 10-16) and CACNA1I (p < 2.2 x 10-16). For the Dutch cohort this replicated for CACNA1H (p = 3.5 x 10-8, p cor = 1.04 x 10 -7) and CACNA1I (p = 0.019, p cor = 0.056), but not CACNA1E (p = 0.85), albeit that CACNA1I did not remain significant after correction for multiple testing. In addition, the number of subjects carrying a variant was also higher cases in CACNA1E (p = 6.2 x 10-3), CACNA1H (p < 2.2 x 10-16) and CACNA1I (p < 2.2 x 10-16) (Table 2). The results showed evidence of replication in the Dutch cohort for CACNA1H (p = 1.2 x 10-2; p cor = 3.6 x 10 -2) and CACNA1I (p = 4.4 x 10-2; p cor =0.13), but not for CACNA1E (p = 0.88), albeit CACNA1I did not remain significant after correction for multiple testing. All but four variants were outside the transmembrane domains that are typically affected in case of a pathogenic mutation in either CaV channel (Figure 2) Table 2 Missense variant burden in the Discovery cohort consisting of Australian cases and GnomAD controls. Gene Gene length (bp) Case count HET Case count HOM Control count Case allele count Control allele count P-value Subjectsa P-value Allele Countsa CACNA1A 8392 54 1 23977 60 26841 1 1 CACNA1B 9790 6 0 1815 7 1819 0.54 0.39 CACNA1C 8425 4 0 1221 4 1225 0.56 0.57 CACNA1D 7636 4 0 2014 4 2024 0.89 0.90 CACNA1E 7067 109 41 29644 228 53792 6.21 x 10-3* 1.30 x 10-4* CACNA1G 7648 12 0 1768 12 1770 1.33 x 10-2 1.65 x 10-2 CACNA1H 8084 155 88 20456 637 22942 <2.2 x 10-16* < 2.2 x 10-16* CACNA1I 10004 126 44 4652 277 4694 <2.2 x 10-16* < 2.2 x 10-16* Case count = number of cases with at least one variant (Het = heterozygous variant Hom = homozygous variant), Control count = number of controls derived from gnomAD with a variant, Case allele count = Total allele count in cases, Control allele count = Total allele count in controls derived from gnomAD. * Significant results; a p-values < 6.25 x 10-3 were considered significant (Bonferroni corrected for testing 8 genes) Discussion Here we used WES data from 184 suspected HM patients from an Australian clinically-referred cohort, and compared these to the publicly available gnomAD control dataset using TRAPD, finding that CACNA1E, CACNA1I and CACNA1H missense variants were more prevalent in cases. Furthermore, we show evidence for replication of these findings for CACNA1H and CACNA1I in a Dutch clinical HM cohort. This finding emphasizes that although the cohorts differ in terms of inclusion criteria, the results are transferable to both groups.
RkJQdWJsaXNoZXIy MTk4NDMw