Significance of α-Myosin Heavy Chain (MYH6) Variants in Hypoplastic Left Heart Syndrome and Related Cardiovascular Diseases

Hypoplastic left heart syndrome (HLHS) is a severe congenital heart disease (CHD) with complex genetic inheritance. HLHS segregates with other left ventricular outflow tract (LVOT) malformations in families, and can present as either an isolated phenotype or as a feature of a larger genetic disorder. The multifactorial etiology of HLHS makes it difficult to interpret the clinical significance of genetic variants. Specific genes have been implicated in HLHS, including rare, predicted damaging MYH6 variants that are present in >10% of HLHS patients, and which have been shown to be associated with decreased transplant-free survival in our previous studies. MYH6 (α-myosin heavy chain, α-MHC) variants have been reported in HLHS and numerous other CHDs, including LVOT malformations, and may provide a genetic link to these disorders. In this paper, we outline the MYH6 variants that have been identified, discuss how bioinformatic and functional studies can inform clinical decision making, and highlight the importance of genetic testing in HLHS.

We previously identified 19 distinct, rare, predicted damaging MYH6 variants in a cohort of 190 unrelated HLHS subjects, comprising >10% of the cohort [8]. These findings are consistent with previous studies of mutations in the zebrafish MYH6 homologue, amhc/myh6, wherein loss-of-function mutations, along with amhc morpholino knockdown, disrupted atrial sarcomere assembly, impaired atrial contractility, and resulted in atrial dilation in zebrafish embryos. Mutant embryos also exhibited ventricular wall thickening and a narrowed ventricular lumen, mimicking the HLHS phenotype [15]. The additional characterization of knockdown embryos revealed that abnormal ventricular morphology was not due to differences in cardiomyocyte number, but rather due to differences in the size and shape of ventricular cardiomyocytes in myh6 mutants, compared to wild-type [16]. Similarly, developing myh6−/− Xenopus tropicalis hearts lacked cardiac contractility, which was accompanied by atrial and ventricular dilation, and impaired outflow tract development [17]. Although murine models are widely used to study CHD, their cardiac chamber-specific expression of MHC is opposite that of humans, making them unsuitable for modeling MYH6-associated disease.
MYH6 encodes for the alpha isoform of the cardiac myosin heavy chain (α-MHC), which is expressed throughout the myocardium during early cardiac development. As development proceeds, MYH6 expression decreases in the ventricles and is replaced with MYH7 (β-MHC) throughout gestation; α-MHC is the dominant atrial isoform postnatally [18][19][20]. Genetic variants in both MYH6 and MYH7 have been linked to numerous human cardiac pathologies, including hereditary cardiomyopathies, arrhythmias, as well as CHD. While MYH7 variants have been characterized more extensively, the specific mechanisms underlying MYH6 variants are less understood. In this paper, we outline the MYH6 variants that have been reported in HLHS and other CHDs, discuss the benefits and limitations of biostatistical methods for interpreting variants, and emphasize the importance of mechanistic studies designed to improve personalized treatment strategies.

Clinical Interpretation of Genetic Studies
These genetic studies have been highly informative, but limitations remain in connecting the knowledge of MYH6 variants to clinical phenotypes. In cohort studies, the designation of a variant as disease-causing relies on a combination of allele frequency and bioinformatic tools of pathogenicity, which are subject to change as new information is learned. There is not a single consensus on what constitutes a "rare" variant, or what is considered damaging when using a continuous variant scoring system, such as Combined Annotation Dependent Depletion (CADD). It is even more challenging to interpret the significance of variants when there are conflicting assessments between computational predictions and variant frequency, or between multiple predictive methods, for example, the MYH6 variants Q277H, M436V, I512T, V606I, D629N, R860H, A936V, R1151Q, A1298V, D1316E, and R1398Q (Table 1), which have predicted opposite effects when compared using the popular tools SIFT and Polyphen2.
Family-based studies remain a useful way of identifying pathogenic variants, as the segregation of a gene variant within multiple affected family members provides strong evidence that a variant is disease-causing. Familial studies are also directly informative for clinical practice when determining whether family members of an affected individual should be screened for the variant, and if carriers should be surveilled for future disease development. These considerations are particularly important in HLHS due to its high heritability and segregation with other LVOT malformations in family members [66,67], many of which are also associated with MYH6 variants.

Impact of MYH6 Variants on Outcomes in HLHS
Our group has examined outcomes of patients with HLHS stratified by presence of an MYH6 variant. Specifically, we compared a composite endpoint of cardiac arrest, need for mechanical circulatory support, and heart transplant or death between 12 HLHS patients with MYH6 and 24 HLHS patients without MYH6 variants. In this cohort, each MYH6 variant carrier was matched to two controls based on anatomical subtype (i.e., aortic and mitral valve anatomy), stage I surgical shunt type, age/era, and sex when possible. Patients with chromosomal abnormalities and those carrying MYH7 variants were excluded from this analysis. The difference in reaching the composite endpoint at 15 years between MYH6 variant and control groups did not reach statistical significance in this small study ( Figure 1). However, there is certainly a trend towards improved short-term event-free survival in the control group. Control group outcomes appear better than previously reported transplantfree survival of HLHS patients during follow-up of the single ventricle reconstruction (SVR) randomized trial cohort, which examined differences in transplant-free survival and interventions based on stage I surgical shunt type [68]. These findings warrant further investigation with a larger sample size and emphasize the importance of genetic testing for all HLHS patients to identify variants that may impact survival even more so than surgical shunt type.
transplant-free survival and interventions based on stage I surgical shunt type [68]. These findings warrant further investigation with a larger sample size and emphasize the importance of genetic testing for all HLHS patients to identify variants that may impact survival even more so than surgical shunt type.

Importance of Mechanistic Studies
Understanding the specific mechanism of MYH6 variant pathogenicity would be especially relevant to clinical decision making, considering the availability of the drugs omecamtiv mecarbil and mavacamten, which act specifically on the cardiac myosin heavy chains (MHC) to improve systolic and diastolic function, respectively. In phase III clinical trials, both drugs showed efficacy in the treatment of heart failure in adults [69−72], irrespective of genetic background; omecamtiv mecarbil was FDA-approved for use earlier this year, and FDA approval of mavacamten is pending. In HLHS patients with a known pathogenic MYH6 variant, the cardiac specificity of omecamtiv mecarbil and mavacamten may offer a way to prevent disease progression. This treatment may be particularly important in variant carriers, given our previous report that HLHS patients with MYH6 variants have decreased cardiac transplant-free survival compared to HLHS patients without MYH6 variants [8]. However, choosing to use a cardiac MHC-specific activator vs. inhibitor requires the understanding of whether a specific variant will cause systolic or diastolic dysfunction. This highlights the importance of mechanistic studies designed to understand phenotypes at the cellular and tissue levels. Relative to the large body of literature assessing MYH7 variants, few studies have sought to understand MYH6 variant pathology at the molecular level; the findings from these studies are summarized in Table 2.

Importance of Mechanistic Studies
Understanding the specific mechanism of MYH6 variant pathogenicity would be especially relevant to clinical decision making, considering the availability of the drugs omecamtiv mecarbil and mavacamten, which act specifically on the cardiac myosin heavy chains (MHC) to improve systolic and diastolic function, respectively. In phase III clinical trials, both drugs showed efficacy in the treatment of heart failure in adults [69][70][71][72], irrespective of genetic background; omecamtiv mecarbil was FDA-approved for use earlier this year, and FDA approval of mavacamten is pending. In HLHS patients with a known pathogenic MYH6 variant, the cardiac specificity of omecamtiv mecarbil and mavacamten may offer a way to prevent disease progression. This treatment may be particularly important in variant carriers, given our previous report that HLHS patients with MYH6 variants have decreased cardiac transplant-free survival compared to HLHS patients without MYH6 variants [8]. However, choosing to use a cardiac MHC-specific activator vs. inhibitor requires the understanding of whether a specific variant will cause systolic or diastolic dysfunction. This highlights the importance of mechanistic studies designed to understand phenotypes at the cellular and tissue levels. Relative to the large body of literature assessing MYH7 variants, few studies have sought to understand MYH6 variant pathology at the molecular level; the findings from these studies are summarized in Table 2.  A822T SCD • LV and conduction system fibrosis [36] K849del HLHS • Atrial sarcomere disarray (no effect on ventricular sarcomere organization) [8,75] E1503V HLHS • Atrial sarcomere disarray (no effect on ventricular sarcomere organization) [75] S385L & M436V HLHS • No effect on atrial or ventricular sarcomeres [75]

In Vitro Mechanistic Studies
Most of the mechanistic studies of MYH6 variants have utilized in vitro methods. The first variant to be functionally assessed was MYH6-I820N; this ASD-associated variant is located within the regulatory light chain (RLC) binding region of α-MHC and was found in cellular studies to decrease the binding affinity of α-MHC for RLC [31]. However, RLC binding is thought to modulate MHC activation [76] and it is unclear how a disruption in this process could lead to an ASD. Similarly, the SSS-associated MYH6-E933del variant enhanced binding to myosin-binding protein C (MyBP-C), an effect consistent with the location of E933del within the MyBP-C binding region of the α-MHC protein [39]. HL-1 mouse atrial cardiomyocytes transfected with human MYH6-E933del also exhibited a slower electric propagation velocity compared to human MYH6-WT [39], and neonatal rat ventricular cardiomyocytes (NRVCMs) transfected with either the E933del or the R721W [35] variant exhibited disrupted sarcomere structure and the perinuclear aggregation of α-MHC [39]. Together, these findings led the authors to suggest that structural changes within the atrial cardiomyocytes surrounding the sinus node leads to node dysfunction and conduction defects. However, given the role of MyBP-C in modulating contractile strength [77], it is unclear how changes in the α-MHC/MyBP-C interaction are linked to the identified conduction deficits.
Many other MYH6 variants have shown similar changes in sarcomere structure. Cultured cardiomyocytes expressing the A230P [26], A1366D [26], E526K [64], R1822_E1823dup [73], and HLHS-associated R443P [8] variants exhibited decreased sarcomere organization in variant-carrying cells, while H252Q actually increased myofibril striations [26]. Interestingly, the MYH6-V700M variant did not appear to impact sarcomere organization, despite its rare frequency (<0.001%) and being predicted as "likely damaging" by CADD, SIFT, and PolyPhen2. Our lab also assessed sarcomere organization in cardiac tissue from HLHS patients and found that atrial sarcomeres were disrupted with the R443P, K849del, and E1503V variants, while the ventricular sarcomere structure remained intact [75], consistent with α-MHC being the predominant atrial MHC isoform postnatally. Other groups have also evaluated cardiac tissue from MYH6 variant carriers and reported fibrosis in the conduction system [36], ventricular walls [36,44,54], and ventricular septum [44,54]. Given the predominance of β-MHC in the postnatal ventricles, it is possible that ventricular fibrosis is, at the cellular level, a downstream response to atrial cardiomyocyte dysfunction caused by MYH6 variants in the patients studied.
Some of the most informative in vitro studies are those that examine contractility at the cellular level. NRVCMs expressing human MYH6-A1004S shortened at a slower rate, and consequently shortened less overall, when compared to NRVCMs expressing human MYH6-WT. In the same study, NRVCMs expressing human MYH6-P830L showed no difference in shortening rate, compared to MYH6-WT [74]. This finding is somewhat unexpected, given that A1004S is located on the MHC backbone and is predicted by both SIFT and PolyPhen2 to be non-damaging (Table 1). The MYH6-A1004S variant is also found at a frequency of 1.1% in the general population, which is more common than CHD.
Meanwhile, MYH6-P830L is a novel variant and located near the RLC binding region and thus one would predict greater changes in function with P830L than A1004S. Our lab also found that the MYH6-R443P variant decreased the shortening rate, relaxation rate, extent of shortening, percent shortening, and calcium transient amplitude at the single CM level in patient-specific induced pluripotent stem-cell-derived cardiomyocytes (iPSC-CMs), without affecting action potentials. These MYH6-R443P iPSC-CMs also demonstrated sarcomere disorganization and the upregulation of MYH7, recapitulating the phenotype found in atrial tissue from an HLHS patient carrying the R443P variant [8,75].

In Vivo Mechanistic Studies
To date, zebrafish embryos are the only animal model that has been used to study human MYH6 variants. Specifically, researchers evaluated the ability of the human MYH6-E933del and MYH6-R1252Q variants, which are associated with cardiac conduction disease, to rescue cardiac impairments resulting from myh6 knockdown. In both sets of experiments, the authors reported bradycardia in knockdown embryos at 48 hpf −137.7 ± 2.2 bpm in myh6−/− vs. 150.2 ± 1.6 in uninjected [39], and 144 ± 16 bpm in myh6−/− vs. 153 ± 13 in uninjected [67]. While human MYH6-WT and MYH6-R1252Q increased heart rate in knockdowns, human MYH6-E933del failed to rescue this phenotype.

Structural Considerations
The structure of human α-MHC has not been solved, thus most hypotheses regarding the effect of MYH6 variants on the α-MHC structure are based on comparison to the solved structures of β-MHC. Mutational clustering analysis using population-level data found that pathologic variants in MYH7 cluster in certain regions [78], which has been used to inform ACMG/AMP variant classification framework [79]. At present, no such "mutational hotspots" have been identified in MYH6 ( Figure 2); however, similar patterns could emerge as new pathological variants are discovered. Investigating structure-function relationships in MYH7 has been successful in elucidating a mechanism for HCM that results from variants in a surface region of β-MHC referred to as the "myosin mesa" [80,81]. However, the interpretation of such studies may be complicated by the finding that the same MYH7 variant can cause clinically opposite phenotypes (i.e., both HCM and DCM), depending on the person [82].
Some research groups have begun employing advanced in silico methods to model the effect of specific MYH6 variants. Molecular dynamics simulations predicted that the MYH6 variants E1207K and T1379M alter the helicity and flexibility of the tail domain, which would likely impact the rigidity and movement of the thick filament as a whole [9]. Similarly, simulations found that MYH6-R1822_E1823dup likely increases the strength of the dimerized α-MHC tail domain, decreasing its flexibility [73]. However, this information does not explain either the clinical or cellular phenotypes associated with these variants. the effect of specific MYH6 variants. Molecular dynamics simulations predicted that the MYH6 variants E1207K and T1379M alter the helicity and flexibility of the tail domain, which would likely impact the rigidity and movement of the thick filament as a whole [9]. Similarly, simulations found that MYH6-R1822_E1823dup likely increases the strength of the dimerized α-MHC tail domain, decreasing its flexibility [73]. However, this information does not explain either the clinical or cellular phenotypes associated with these variants. Variants were considered rare if allele frequency was <1 × 10 −3 in both the Genome Aggregation Database (gnomAD) Genomes dataset v3.1.2 and the Allele Frequency Aggregator (ALFA) dataset (release version 20201027095038). Single nucleotide variants were considered predicted damaging if the scaled Combined Annotation Dependent Depletion (CADD) score was >22.0 (GRCh37, v1.6) [60], or if the variant was predicted "damaging" or "probably damaging" by SIFT [61] and Poly-Phen2 [62]. Deletions are not shown as CADD scores cannot be calculated. (B) Schematic of α-MHC domains. ELC, essential light chain; RLC, regulatory light chain; MyBP-C, myosin binding protein C; ACD, assembly competent domain. Variants were considered rare if allele frequency was <1 × 10 −3 in both the Genome Aggregation Database (gnomAD) Genomes dataset v3.1.2 and the Allele Frequency Aggregator (ALFA) dataset (release version 20201027095038). Single nucleotide variants were considered predicted damaging if the scaled Combined Annotation Dependent Depletion (CADD) score was >22.0 (GRCh37, v1.6) [60], or if the variant was predicted "damaging" or "probably damaging" by SIFT [61] and PolyPhen2 [62]. Deletions are not shown as CADD scores cannot be calculated. (B) Schematic of α-MHC domains. ELC, essential light chain; RLC, regulatory light chain; MyBP-C, myosin binding protein C; ACD, assembly competent domain.

Conclusions
HLHS is a complex and genetically heterogenous disease, and the origins of HLHS are likely multigenic. Evidence suggests MYH6 variants are etiologic in a significant percentage of HLHS. Many of the studies cited in this paper identified additional candidate variants that may be contributing to cardiac disease development, including some patients carrying compound heterozygous MYH6 variants. New bioinformatic tools, such as Oligogenic Resource for Variant AnaLysis (ORVAL) [83], are designed to identify candidate pathogenic combinations of variants and are likely to be useful in elucidating the multigenic origins of HLHS and related disorders. The presence of additional genetic variants and environmental influences may explain some of the discrepancies between bioinformatic predictions of MYH6 variant pathogenicity and the reported cellular and clinical phenotypes. In any event, MYH6 variants will remain an important genetic risk factor for HLHS, having prognostic significance irrespective of other factors.
Our published work, [8,75] and many of the studies discussed in this paper, supports our hypothesis that atrial dysfunction due to sarcomere disorganization impairs atrial contractility during cardiac development leading to HLHS. These changes in atrial cardiomyocytes would likely impair atrial contractility in single ventricle patients postnatally, leading to heart failure over time. We anticipate that future longitudinal analyses will allow us to better understand the impact of MYH6 variants on long-term cardiac function in HLHS. Institutional Review Board Statement: This study was conducted in accordance with the principles outlined in the Declaration of Helsinki and institutionally approved research (IRB) protocols by Children's Wisconsin (Milwaukee, WI, United States). Subjects were consented through the CHD Tissue Bank (IRB #CHW 06/229, GC 300) and the Wisconsin Pediatric Cardiac Registry (IRB #CHW 09/91, GC889), IRB-approved research databases housed at Children's prior to inclusion in the study. All associated clinical outcome variables were obtained through these protocols.

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
Restrictions apply to the availability of these data. Data is restricted as it contains protected patient information.