Rare Variants in Genes Encoding MuRF1 and MuRF2 Are Modifiers of Hypertrophic Cardiomyopathy

Modifier genes contribute to the diverse clinical manifestations of hypertrophic cardiomyopathy (HCM), but are still largely unknown. Muscle ring finger (MuRF) proteins are a class of muscle-specific ubiquitin E3-ligases that appear to modulate cardiac mass and function by regulating the ubiquitin-proteasome system. In this study we screened all the three members of the MuRF family, MuRF1, MuRF2 and MuRF3, in 594 unrelated HCM patients and 307 healthy controls by targeted resequencing. Identified rare variants were confirmed by capillary Sanger sequencing. The prevalence of rare variants in both MuRF1 and MuRF2 in HCM patients was higher than that in control subjects (MuRF1 13/594 (2.2%) vs. 1/307 (0.3%), p = 0.04; MuRF2 22/594 (3.7%) vs. 2/307 (0.7%); p = 0.007). Patients with rare variants in MuRF1 or MuRF2 were younger (p = 0.04) and had greater maximum left ventricular wall thickness (p = 0.006) than those without such variants. Mutations in genes encoding sarcomere proteins were present in 19 (55.9%) of the 34 HCM patients with rare variants in MuRF1 and MuRF2. These data strongly supported that rare variants in MuRF1 and MuRF2 are associated with higher penetrance and more severe clinical manifestations of HCM. The findings suggest that dysregulation of the ubiquitin-proteasome system contributes to the pathogenesis of HCM.


Introduction
Hypertrophic cardiomyopathy (HCM) is defined by the presence of left ventricular hypertrophy in the absence of loading conditions sufficient to cause the observed abnormality [1]. It is the most common monogenic cardiovascular disease, caused mainly by mutations in genes encoding sarcomere proteins [2][3][4][5]. Mutations in other genes, such as those encoding Z-disc and calcium-handling proteins, are also reported to be associated with the disease, but with low prevalence and less certainty [6].
HCM is characterized by marked clinical heterogeneity [1,7]. There is variable penetrance and variable age of onset, ranging from infancy to old age. The severity of cardiac hypertrophy and left ventricular outflow obstruction vary considerably, and clinical outcomes are heterogeneous. This phenotypic diversity is present even among patients with same genetic mutation [8], suggesting a role for modifier genetic or environmental factors.
A delicate balance between protein synthesis and degradation is necessary for cardiac homeostasis. The ubiquitin-proteasome system (UPS) is responsible for the degradation of 70%-90% intracellular proteins. Impairment of the UPS has been implicated in various cardiac diseases, including HCM [9][10][11][12][13]. Muscle ring finger (MuRF) proteins MuRF1, 2 and 3 (encoded by TRIM63, TRIM55 and TRIM54, respectively) comprise a subfamily of the RING-finger E3 ubiquitin ligases that are specifically expressed in striated muscles [14,15]. They can create complexes as homodimers and heterodimers, and regulate myocyte size and contractility through proteasome-dependent degradation of sarcomere proteins and transduction factors of hypertrophic signaling [16][17][18][19][20][21][22][23]. Recently, mutations in the gene encoding MuRF1 were reported to cause HCM by impairing protein degradation in cardiomyocytes [24], suggesting genetic variants in genes encoding MuRF proteins might be involved in the pathogenesis of HCM. In the present study, we evaluated the association between genetic variants of MuRF genes and HCM phenotype by screening all the three members of MuRF subfamily in a large HCM cohort and matched healthy controls.

Rare Variants in MuRF3 Gene
In the MuRF3 gene, 19 rare variants were identified in 14 (2.4%) of the HCM patients and 8 (2.6%) of the healthy controls ( Table 2). All detected variants were missense and heterozygous (Table S2). The frequency of rare variants in MuRF3 was equal between HCM patients and healthy controls (χ 2 = 0.053, p = 0.82). No common variants in the MuRF3 gene were detected in the study population.

Coexistence of Sarcomere Mutations with MuRF1 and MuRF2 Rare Variants
In order to determine whether the rare variants in genes encoding MuRF1 and MuRF2 were independent causes of HCM, the coexistence of mutations in 8 sarcomere genes was analyzed (Table 1). Among 34 HCM patients with rare variants in the MuRF1 or the MuRF2 genes, 19 (55.9%) carried mutations in sarcomere genes. Similarly, among the 22 HCM patients with potentially pathogenic variants in the MuRF1 or the MuRF2 genes, sarcomere mutations were present in 12 (54.5%).

Association of MuRF1 and MuRF2 Rare Variants with Phenotypes of Hypertrophic Cardiomyopathy (HCM) Patients
The phenotypes of HCM patients were ascertained to determine whether rare variants of MuRF1 and MuRF2 modify the clinical expression of the disease (Tables 3 and S3). We found that patients with rare variants of MuRF1 and MuRF2 were younger (44.5 ± 14.2 vs. 49.6 ± 14.0 years, p = 0.04), and had greater maximum left ventricular wall thickness (23.8 ± 5.2 vs. 21.5 ± 4.7 mm, p = 0.006) than those without such variants. Likewise, patients with potentially pathogenic rare variants were also younger (p = 0.02), with thicker left ventricular wall (p = 0.02).

Discussion
Although polymorphisms in genes encoding members of the renin-angiotensin-aldosterone system and other genes implicated in cardiac growth have been reported to be associated with the penetrance and severity of cardiac hypertrophy in HCM [25][26][27], modifier gene effects in HCM are still largely unknown. In the present study, we sequenced genes encoding all three members of MuRF E3-liagse subfamily, including MuRF1, MuRF2 and MuRF3, in 594 HCM patients and 307 healthy controls, and found that rare nonsynonymous variants in genes encoding MuRF1 and MuRF2, but not MuRF3, were more prevalent in HCM patients (respectively, approximately seven times and five times). Patients with these variants were younger and had a greater left ventricular wall thickness. Thus, our findings indicate that rare variants in MuRF1 and MuRF2 can modify the phenotypic expression of HCM.
In most mammalian cells, the UPS controls many fundamental biological processes through targeting the degradation of most cytosolic, nuclear and myofibrillar proteins in a highly regulated manner [9]. Impairment of this process has been observed in various cardiac diseases. In human HCM, proteasome activity is decreased, accompanied by increased mediators of cardiac hypertrophy and apoptosis [28]. Expression of mutant MYPBC3 proteins, one of the most common causes of HCM, results in UPS impairment in both cultured neonatal cardiomyocytes and animal models [11][12][13]29,30]. Furthermore, inhibition of proteasome activity produces cardiac hypertrophy in vivo. Thus, the UPS may play an important role in HCM pathogenesis.
The rate-limiting enzyme in UPS-dependent protein degradation is the ubiquitin E3-ligase, which recognizes specific substrates and catalyzes the transfer of activated ubiquitin to them. MuRF1, MuRF2 and MuRF3 belong to the same E3-ligase subfamily, and are restrictedly expressed in cardiac and skeletal muscles [14,15]. MuRF1 and MuRF2 interact with a number of the same sarcomere and sarcomere-related proteins, and are implicated in contractile regulation, myogenic responses and regulation of cardiac hypertrophy [31,32]. Whereas increased expression of MuRF1 in cardiomyocytes inhibits the development of cardiac hypertrophy [16], deletion of MuRF1 leads to exaggerated cardiac hypertrophy in response to pressure overload [19]. Although mice with deletion of either MuRF-1 or MuRF-2 alone exhibit a normal cardiac phenotype at baseline, MuRF-1/MuRF-2 double knockout mice develop massive spontaneous hypertrophic cardiomyopathy [31]. Impairment of the UPS may underlie the association of rare variants in the genes encoding MuRF1 and MuRF2 with increased risk for HCM and greater severity of phenotypic expressions observed in our study. Variants in these two genes may reduce protein degradation through the UPS, and thereby lead to accumulation of mutant proteins and cardiac hypertrophy regulators. Consistent with our findings, previous studies have shown that expression of mutant MYBPC3 protein in an HCM model is markedly increased by proteasome inhibition [30]. In a recent study, cardiac-specific over-expression of MuRF1 mutations led to increased sarcomere protein expressions and activation of the MTOR-S6K and calcineurin pathways [24].   MuRF3 appears to have a different function from MuRF1 and MuRF2. MuRF3 mainly interacts with microtubule proteins, whereas MuRF1 and MuRF2 interact with sarcomeric contractile proteins [18]. This functional difference may explain why variants in the MuRF3 gene were not associated with HCM in our study.
Recently, Chen et al. reported mutations in the gene encoding MuRF1 as a likely cause of HCM [24]. They identified three mutations in the MuRF1 gene in 5 of 302 HCM probands. These mutations were absent in 1090 controls and coexistence of mutations in HCM-causing sarcomere genes were excluded. Although the size of families in was not sufficient for genetic linkage analysis, transgenic over-expression of any of the three mutations led to cardiac hypertrophy in mice, suggesting MuRF1 mutations are independent causes of HCM. Nevertheless, controversy remains since other study found that individuals with MuRF1 p.Q247X, one of the disease-causing mutations, reported by Chen et al. [24], had no evidence of HCM [33]. In our study, rare variants in the MuRF1 and MuRF2 genes were present in healthy control subjects, albeit at a much lower prevalence than in HCM patients. Bioinformatics analysis predicted that the rare variants identified in controls were damaging. More importantly, sarcomere mutations were identified in more than half of the HCM patients with rare variants in the MuRF1 and MuRF2 genes, comparable to its prevalence in the general HCM population. Such a high frequency of coexisting sarcomere mutations and MuRF variants suggests that most rare MuRF variants were modifiers and not independent causes of HCM.
Limitation: First, the frequency of MuRF rare variants in healthy sarcomere-mutation carriers was not analyzed in the present study. A comparison of the prevalence of MuRF rare variants between HCM patients with sarcomere mutations and healthy sarcomere-mutation carriers would help to confirm the modifier effect of rare variants in MuRF1 and MuRF2 genes on the development and phenotypic expression of HCM; Second, in the present study, although more than 50% of the patients either with rare variants or with pathogenic rare variants of MuRF1 and MuRF2 genes carried mutations of sarcomere genes, the HCM-causing roles of these variants, especially those identified in patients without sarcomere mutation detected, cannot be excluded completely. Finally, the number of patients with rare variants of MuRF1 and MuRF2 is relatively small. The findings in our study require validation in other large cohorts, and should be interpreted with caution.

Ethics Statement
This study was performed in accordance with the principle of the Declaration of Helsinki and approved by the Ethics Committees of Fuwai Hospital. All participants provided written informed consent.

Study Subjects
A total of 594 unrelated HCM patients and 307 age-and sex-matched healthy controls were included in this study. HCM was defined by an unexplained maximal left ventricular wall thickness ≥15 mm on echocardiography or cardiac magnetic resonance imaging in the absence of another cardiac or systemic disease capable of producing that magnitude of hypertrophy. None of the control subjects had a history of cardiovascular or other systemic diseases. Normal cardiac structure and function in healthy controls was confirmed by evaluation with a 12-lead electrocardiogram and an echocardiogram.

Gene Sequencing and Variant Classification
Genomic DNA from all HCM and control subjects was isolated from peripheral blood leukocytes and used to construct shotgun libraries of approximately 250 bp fragments with index adaptors. All coding exons of the genes encoding MuRF1, MuRF2 and MuRF3 and their adjacent 5 bp intronic sequences were enriched by using a custom designed probe library (Agilent Technologies, Santa Clara, CA, USA), and sequenced on Illumina GAIIx (Illumina Inc., San Diego, CA, USA) to generate pair-end sequencing reads of 120 bp at each end. After removal of PCR duplications with PICARD (http://picard.sourceforge.net/), sequencing reads of each individual were mapped to the human genome (GRCh37/hg19) with CLC Genomics Workbench (CLC-bio, Aarhus, Denmark). The sequencing depth was analyzed and variants in MuRF genes were called using the following filter parameters: coverage ≥25× and variant frequency ≥20%. Synonymous variants were excluded from further analysis. A rare variant was defined as one present with a minor allele frequency (MAF) of <0.5% in the studied subjects. All rare nonsynonymous variants identified by targeted resequencing were confirmed by bidirectional capillary Sanger sequencing. Eight sarcomere genes (MYH7, MYBPC3, TNNT2, TNNI3, MYL2, MYL3, TPM1 and ACTC1) were also evaluated by targeted resequencing in HCM subjects.
A rare missense variant was considered to be potentially pathogenic when predicted to be damaging by any of PolyPhen-2 [34] or SIFT programs [35]. Nonsense and frame-shift variants were considered to be pathogenic since they were expected to result in truncated proteins.

Statistical Analysis
Continuous variables are presented as mean ± standard deviation and categorical variables as frequencies (n) and percentages (%). The Student's t test was used for the comparison of continuous variables, and χ 2 or Fisher exact test for non-continuous variables. p values are two-sided and considered significant when <0.05. Calculations were performed using PASW Statistics 18 software (SPSS, Chicago, IL, USA).

Conclusions
In conclusion, rare variants in the genes encoding MuRF1 and MuRF2 increase the risk of development of HCM and lead to more severe disease phenotypes, and may act as modifier of the disease. Our study is consistent with the hypothesis that impaired ubiquitin-proteasome system contributes to the pathogenesis of HCM.