Reevaluation of the South Asian MYBPC3Δ25bp Intronic Deletion in Hypertrophic Cardiomyopathy

Supplemental Digital Content is available in the text.

H ypertrophic cardiomyopathy (HCM) is the most common inherited cardiac condition, affecting at least ≈1:500 individuals. 1 It is a genetically heterogeneous disorder, typically attributable to pathogenic variants in genes encoding cardiac sarcomere proteins, predominantly MYBPC3 and MYH7. 2 Truncating variants in MYBPC3 are a well-recognized cause of HCM, and the majority are considered to cause autosomal dominant disease with high age-related penetrance; consequently, such variants are extremely rare in the wider nondisease population. 2 Circ Genom Precis Med. 2020;13:e002783. DOI: 10.1161/CIRCGEN.119. 002783 June 2020 102 A 25 base pair deletion located within intron 32 of MYBPC3 (MYBPC3 Δ25 ), the c.3628-41_3628-17del variant, is a notable exception. Detected in 4% to 8% of individuals of South Asian ancestry, 3,4 and with an estimated 100 million carriers worldwide, this common variant is considered to be associated with cardiomyopathy, with an almost 7-fold increased risk of cardiomyopathy in heterozygous carriers. 3 Although previous studies have considered the possibility that MYBPC3 Δ25 lies in linkage disequilibrium with another MYBPC3 variant that causes or contributes to disease risk, 3,4 comprehensive analyses in large patient cohorts have not been performed.
Here, using genetic data from 2 large HCM cohorts, we present data suggesting that MYBPC3 Δ25 is not a pathogenic risk factor in HCM. Rather, the increased frequency of this variant in South Asian cardiomyopathy cohorts reflects the enrichment of a derived haplotype, which bears both the common MYBPC3 Δ25 variant and a rare pathogenic variant, MYBPC3 c.1224-52G>A. Additionally, we find that MYBPC3 c.1224-52G>A-an intronic variant that is not routinely detected on gene panel or exome sequencing-is the single most common pathogenic variant in individuals of South Asian ancestry in our cohort and the second most common in individuals of European ancestry.

METHODS
The complete methods are available in Materials in the Data Supplement. Due to the confidential nature of some of the research materials supporting this publication, not all of the data can be made accessible to other researchers. Please contact the corresponding author for more information. The study was approved by the local ethics committees, and all patients signed an informed consent.

Oxford Medical Genetics Laboratory Demographic and Clinical Details
Within the Oxford Medical Genetics Laboratory (OMGL) cohort, demographic information was available for 98.0% of individuals (2703/2757). The majority of referrals were provided by inherited cardiac condition centers within the United Kingdom (80.1%; 2166/2757). The average age was 54.5 years (±16.2), and 68.4% were men (n=1845; Table 1). No self-identified, or genetically derived, ancestry information was available.

HCMR Demographic and Clinical Details
Within the HCMR cohort, the average age was 49.5 years (±11.3), and 71.4% were men. Genetically derived ancestry predictions, determined through principal components analysis, demonstrated European ancestry in 78.3%, African ancestry in 9.0%, and South Asian ancestry in 5.1% of individuals (Table 1).

Population Frequency of MYBPC3 Δ25
In the Genome Aggregation Database (gnomAD; v2.

Oxford Clinical Laboratory Cohort
In the OMGL HCM cohort, pathogenic variants were detected in 17.1% (471/2757), likely pathogenic variants in 6.9% (191/2757), and variants of uncertain significance in an additional 14.2% (392/2757) of individuals. A summary of the most frequently detected variants is presented in Table I   In individuals of South Asian ancestry in the HCMR cohort, the MYBPC3 c.1224-52G>A variant was found    individuals with the c.1224-52G>A variant generated an aberrant product. Sequencing of this product confirmed in silico predictions and showed inclusion of 50 intronic nucleotides in the transcript (Figure 2). Inclusion of these nucleotides is predicted to lead to a frameshift in the amino acid sequence and insertion of a premature termination codon at position 438 (p.Ser408fs*31).
Pathogenicity Classification for MYBPC3 c.1224-52G>A Using the American College of Medical Genetics framework, 6 the MYBPC3 c.1224-52G>A variant was classified as pathogenic based on the following criteria: PS3: RNA studies have provided evidence of an aberrant effect on splicing (our analyses and published data 7 ); PS4: the variant is significantly more frequent in probands with HCM than in population controls; PM2: the variant is very rare in the wider population; and PP1: there is evidence of cosegregation with HCM in multiple families (4 in our cohort and published data 7 ).

DISCUSSION
When the MYBPC3 Δ25 variant was first reported to be associated with cardiomyopathy in the South Asian population, it was thought likely to have a direct role in disease pathogenesis; since the initial report, it has come to be considered as one of the most compelling examples of a common, low-penetrance variant contributing to the genetic architecture of HCM. 3,[8][9][10][11][12] Genetic analyses undertaken in this study challenge these previous assertions and show that the MYBPC3 Δ25 variant does not directly confer an increased risk of cardiomyopathy but instead acts as a proxy marker for a rare, large effect size, intronic pathogenic variant, MYBPC3 c.1224-52G>A ( Figure 3). Consequently, we conclude that heterozygosity for the MYBPC3 Δ25 common variant is not pathogenic for HCM. Through RNA studies and segregation analyses, we provide robust evidence to support the pathogenicity of the MYBPC3 c.1224-52G>A variant. This variant has previously been described in the literature as a pathogenic variant 7 ; however, neither its high prevalence nor its relationship with MYBPC3 Δ25 has been reported. Our analyses reveal MYBPC3 c.1224-52G>A to be a recurrent variant, and one of the most frequent pathogenic variants across all known HCM genes in both European and South Asian populations, comparable to other wellestablished recurrent and founder pathogenic variants (eg, MYBPC3 c.2373dup 13 and MYBPC3 p.Glu258Lys 2 ), and exceeded only by the MYBPC3 p.Arg502Trp variant, the most common pathogenic variant in HCM. 2,5,14 Further, the MYBPC3 c.1224-52G>A variant has a strikingly high OR for disease (≈700), suggesting that it is a high penetrance allele.
Haplotype analyses indicate that an ancestral MYBPC3 c.1224-52G>A variant arose on a haplotype bearing the common MYBPC3 Δ25 variant and that the reported association between MYBPC3 Δ25 and HCM in the South Asian population was due to the increased frequency of the derived MYBPC3 Δ25/−52 haplotype, which had not previously been differentiated from the common MYBPC3 Δ25 haplotype. In our cohort, after accounting for the MYBPC3 Δ25/−52 haplotype, the frequency of the MYBPC3 Δ25 allele appears equivalent between HCM cases and reference controls, which casts doubt upon previous pathogenic inferences from risk associations and suggests that it is not clinically appropriate to type the MYBPC3 Δ25 in isolation. Indeed, the ability to detect the MYBPC3 Δ25/−52 haplotype is critical not only for individuals with a clinical diagnosis of HCM but for the vast majority of the 100 million individuals of South Asian ancestry heterozygous for the MYBPC3 Δ25 alone, who would previously have been declared at increased risk of HCM.

Limitations
Our conclusions rely on the observed MYBPC3 Δ25 and MYBPC3 Δ25/−52 haplotype frequencies being representative of the wider South Asian population. Here, direct evaluation of MYBPC3 Δ25 and MYBPC3 Δ25/−52 and HCM disease risk has relied on analysis performed using individuals ascribed South Asian ancestry based on genetic principal components analysis from 2 independent, but relatively small, cohorts. Large reference cohorts, specifically gnomAD and Trans-Omics for Precision Medicine, were useful in quantifying the allele frequencies of both MYBPC3 Δ25 and MYBPC3 c.1224-52G>A but were not suitable for the direct evaluation of the MYBPC3 Δ25/−52 haplotype, given the lack of individuallevel data.
Our case series comprised 2 large HCM cohorts with a combined total of 5394 HCM probands (OMGL, n=2757; HCMR, n=2636), representing the largest published HCM cohort to date. MYBPC3 Δ25 and MYBPC3 Δ25/−52 haplotype frequencies were equivalent within these mixed ancestry HCM cohorts. Ancestry data were only available from the HCMR cohort, in which 134 cases were defined as South Asian; additional analyses in other South Asian cohorts will refine MYBPC3 Δ25/−52 haplotype frequency estimates and allow more accurate quantification of the strength of the association of this haplotype to HCM in this population.
The findings in this study relate specifically to HCM. In the original case-control study by Dhandapany et al, 3 2 composite case groups were assembled that included individuals diagnosed with HCM (n=357), dilated cardiomyopathy (n=395), and restrictive cardiomyopathy (n=15). While our findings refute a pathogenic role for the MYBPC3 Δ25 variant in HCM, at present, our conclusions do not extend to these other cardiomyopathies or to homozygosity for this variant. However, given current understanding of the diametrically opposing molecular mechanisms that underpin sarcomeric HCM and The MYBPC3 Δ25 is a common variant present in 4% to 8% of the South Asian population (estimated to be carried by ≈100 million people). In a cohort of South Asian hypertrophic cardiomyopathy (HCM) cases, we detected a rare derived haplotype, bearing both MYBPC3 Δ25 and a pathogenic variant, MYBPC3 c.1224-52G>A. The rare MYBPC3 Δ25/−52 haplotype is strongly associated with HCM with high penetrance. Haplotypes bearing MYBPC3 Δ25 without the MYBPC3 c.1224-52G>A variant, which account for the vast majority of South Asian individuals carrying the MYBPC3 Δ25 variant, are not associated with HCM.

Conclusions
The results of this study provide strong evidence to refute a direct pathogenic link between the MYBPC3 Δ 25 variant and HCM risk; this is important for the very large number of South Asian individuals who will be found to have this variant when undergoing either targeted or genome-wide genetic analysis. Additionally, they highlight MYBPC3 c.1224-52G>A as an important HCM variant. They also reiterate the importance of sequencing deeper intronic regions in the MYBPC3 gene, and, indeed, other cardiomyopathy genes where truncating variants are believed to cause the disease. Collectively, these findings have significant implications for our understanding of the genetic architecture of HCM and for the clinical management of patients with HCM.