Distribution of Disease Genes, Spectrum of Mutations, and Implications for a Molecular Diagnosis Strategy
Background— Hypertrophic cardiomyopathy is an autosomal-dominant disorder in which 10 genes and numerous mutations have been reported. The aim of the present study was to perform a systematic screening of these genes in a large population, to evaluate the distribution of the disease genes, and to determine the best molecular strategy in clinical practice.
Methods and Results— The entire coding sequences of 9 genes (MYH7, MYBPC3, TNNI3, TNNT2, MYL2, MYL3, TPM1, ACTC, andTNNC1) were analyzed in 197 unrelated index cases with familial or sporadic hypertrophic cardiomyopathy. Disease-causing mutations were identified in 124 index patients (≈63%), and 97 different mutations, including 60 novel ones, were identified. The cardiac myosin-binding protein C (MYBPC3) and β-myosin heavy chain (MYH7) genes accounted for 82% of families with identified mutations (42% and 40%, respectively). Distribution of the genes varied according to the prognosis (P=0.036). Moreover, a mutation was found in 15 of 25 index cases with “sporadic” hypertrophic cardiomyopathy (60%). Finally, 6 families had patients with more than one mutation, and phenotype analyses suggested a gene dose effect in these compound-heterozygous, double-heterozygous, or homozygous patients.
Conclusion— These results might have implications for genetic diagnosis strategy and, subsequently, for genetic counseling. First, on the basis of this experience, the screening of already known mutations is not helpful. The analysis should start by testing MYBPC3 and MYH7 and then focus on TNNI3, TNNT2, and MYL2. Second, in particularly severe phenotypes, several mutations should be searched. Finally, sporadic cases can be successfully screened.
Received December 9, 2002; revision received February 12, 2003; accepted February 18, 2003.
Familial hypertrophic cardiomyopathy (HCM) is a cardiac disorder characterized by left ventricular hypertrophy (LVH), with predominant involvement of the interventricular septum in the absence of other causes of hypertrophy.1 The prevalence of the disease in the population is 0.2%.2 HCM is clinically heterogeneous, with inter- and intrafamilial variations ranging from benign forms3 to malignant forms with a high risk of cardiac failure or sudden cardiac death.4
See p 2171
HCM is characterized by an autosomal-dominant mode of inheritance. Ten genes have been identified, 9 of them encoding for cardiac sarcomeric proteins.5–7 These are the β-myosin heavy chain (MYH7), the myosin ventricular essential light chain 1 (MYL3), the myosin ventricular regulatory light chain 2 (MYL2), the cardiac α actin (ACTC), α-tropomyosin (TPM1), the cardiac troponin T (TNNT2) and cardiac troponin I (TNNI3), the cardiac myosin binding protein C (MYBPC3), and titin (TTN). The last one is PRKAG2, which encoded the γ subunit of protein kinase A, which is associated with the particular phenotype of HCM and Wolff-Parkinson-White syndrome.8,9
Numerous mutations have been described in these genes.10 However, until now, there have been no data regarding a systematic screening of them in a large panel of patients. This is a key issue in HCM because it will lead to an appreciation of the efficacy of systematic screening and, therefore, help to clarify the possibility of genotyping in clinical practice. A molecular strategy might then be proposed according to the relative frequency of the genes and mutations. In addition, the description of the spectrum of genes and mutations would facilitate presymptomatic testing and allow phenotype-genotype analyses. The aim of the present study was, therefore, to perform a systematic screening of the 9 genes associated with the classic phenotype of HCM in a large population of 197 unrelated index patients.
Patients were recruited in France, and most of them were of European origin. Clinical evaluation was performed as described elsewhere,11 ECG and echocardiography were recorded, and blood samples were collected. Informed written consent was obtained in accordance with a study protocol approved by the local ethics committee. Briefly, diagnostic criteria were defined in adults by a maximal wall thickness >13 mm on echocardiography or major abnormalities on ECG (abnormal Q waves or LVH or marked T-wave inversion). Prognosis in families was assessed at the time of genotyping and based on family history. Three groups were considered, according to the number of major cardiac events and the relation between the cardiac event and HCM (documented or highly suspected). A major cardiac event was defined as sudden death, heart failure death, stroke death, heart transplantation, or resuscitated death related to HCM, each occurring before 60 years of age. The prognosis was classified as malignant (≥2 documented major cardiac events), intermediate (one major cardiac event, documented or highly suspected), or benign (no major cardiac event in the family). The disease was called “sporadic” in patients with proven HCM but without familial history or affected relatives.
The entire coding sequences of 6 genes were systematically analyzed for the index patient, even when a mutation was identified; these included MYH7 (40 exons), MYBPC3 (35 exons), MYL2 (7 exons), MYL3 (6 exons), TNNI3 (8 exons), and TNNT2 (17 exons). When no mutation was found, analysis of TPM1 (9 exons), ACTC (6 exons), and TNNC (6 exons) genes was performed. The screening of mutations was done with a DNA single-strand conformation polymorphism analysis of each exon and flanking intronic regions, followed by sequencing each abnormal pattern on a capillary DNA sequencer (detailed methods are available on request). A variant was considered a mutation on the basis of the following 3 criteria: cosegregation with affected members in the family, absence of the mutation in 200 unrelated chromosomes of healthy adult controls, and the conservation of the mutated residue among species and isoforms.
Differences between groups were compared with the Fisher test for categorical variables and with the Mann-Whitney test for continuous variables. For all comparisons, a value of P<0.05 was considered significant.
To determine the distribution of the disease genes, 197 index cases, including 172 familial forms and 25 apparently sporadic cases of HCM, were tested for mutations in 9 genes. Disease-causing mutations were identified in 124 index patients (63%), including 4 patients with 2 mutations and 3 homozygous patients. The most frequent genes involved in the genotyped index patients were MYBPC3 and MYH7, which were mutated in 42% and 40% of cases, respectively. The others were involved in <10% of cases (Table 1). Among the 25 sporadic cases, a genetic defect was found in 15 patients (60%). Seven patients had mutations in MYBPC3 (47%), 5 in MYH7 (33%), and 3 in TNNT2 (20%).
A total of 97 different mutations, including 60 novel ones, were identified. Analysis of MYH7 led to the identification of 40 mutations, including 24 novel ones (Table 2). Most of them are located in the amino-terminal part of the protein, but 7 missense mutations were found in the rod domain of the protein (17%). Analysis of MYBPC3 led to the identification of 39 mutations (Table 3), including 26 frameshift or nonsense mutations. All were “private” mutations except an acceptor splice-site mutation (IVS21-2:a13858g), which was found in 13 families of European origin and showed a founder effect in some cases. Analysis of TNNT2 showed 5 missense mutations, one codon deletion (Del E160), and one nonsense mutation (W287ter). Analysis of TNNI3 identified 6 new mutations, 5 missense and one de novo codon deletion (Del K177). In MYL2, 3 missense and one splice acceptor site (IVS5-2:a8629g) mutations were found, and MYL3 testing revealed only one mutation (E56G). Mutations found in TNNT2, TNNI3, MYL2, and MYL3 are indicated in Table 4. Analysis of the TPM1, ACTC, and TNNC2 genes did not reveal any mutations.
Genotyping of available family members allowed us to evaluate prognosis according to the gene involved. Distribution of the disease genes varied according to prognosis in families (P=0.036; Table 5). In benign families, the prevalence of MYBPC3 and MYH7 genes was almost the same (45% and 43%, respectively). In contrast, in families who had a malignant prognosis, MYH7 was the most prevalent gene (45%), and in families with an intermediate prognosis, MYBPC3 was the most prevalent (70%). From another point of view, 90% of families related to the MYBPC3 gene were associated with a benign or intermediate prognosis, whereas 28% of families associated with the MYH7 gene were associated with a malignant prognosis. The TNNT2 gene was equally associated with a benign or malignant prognosis, as was the TNNI3 gene, but the size of the population was small.
Six families carried more than one mutation and could be classified into the following 3 groups. Group 1 included families with double-heterozygous patients who had one mutation in MYH7 and the other in MYBPC3. In the first family, the nonsense MYBPC3 E1096ter mutation was associated with the MYH7 E483K mutation.12 In the second family, 2 missense mutations (MYH7 A355T and MYBPC3V896M) cosegregated in 3 patients. Group 2 included families with compound heterozygous patients. One family had patients carrying two mutations in MYH7 (V39M and R723C) and the other in MYBPC3 (Q76ter and H257P). Group 3 included 3 families with homozygous mutated patients. Two of them were mutated in MYH7 (one with the R869G mutation13 and the other with the D778E mutation) and one in MYBPC3 (Q76ter). Phenotype-genotype analyses of these families are summarized in Table 6.
This report describes the screening of 9 genes in a population of 197 unrelated index cases with familial or sporadic forms of HCM.
The disease-causing mutation was identified in 124 index cases (63%). The lack of identification in the remaining 37% may be related to phenotypic errors, presence of mutations in nonanalyzed sequences, incomplete sensitivity of the mutation screening, or involvement of additional, as yet unidentified genes. Distribution of the disease genes of the full 197 case series was as follows: MYBPC3, 26%; MYH7, 25%; TNNT2, 4%; TNNI3, 4%; MYL2, 2.5%; and MYL3, <0.5%. These results differ from previously reported estimates in which MYH7 was the most frequent and then TNNT2 and MYBPC3.6 This difference may be related to the methods of recruitment, which were possibly based on malignant forms of HCM (with or without LVH). In contrast, we focused our analysis on a recruitment of patients with proven LVH, whatever the prognosis in these families. To test this hypothesis, we analyzed the distribution of genes according to the prognosis observed in the families. We found that MYH7 was the most frequent in families with a malignant prognosis. Because TNNT2 mutations have been reported to be associated with mild or no LVH but a high risk of sudden death,14 this may explain the low rate of TNNT2 mutations found in our population. Our approach may have some limitations due to the analysis of retrospective data and the variable number of patients per family; however, 70% of families had ≥4 genetically affected individuals.
Sporadic cases were also screened, and a mutation was found in 60% of them. Distribution of the disease genes was almost the same as for familial forms but with a higher prevalence of TNNT2 mutations. Testing the available parents revealed a nonpenetrant mutation in 4 cases and a de novo one in 2 patients. This finding has implications for clinicians; even in sporadic cases, a genetic cause should be suspected. An inquest in relatives should therefore be recommended, and information about the risk of transmitting the disease should be given.
Spectrum analysis of the mutations showed that missense, frame-shift, and nonsense mutations were identified. Most MYH7 mutations result in amino acid substitutions located in the globular head of the protein and affect the binding sites for ATP, actin, and essential or regulatory light chains. Two amino acid deletions were found in the S2 domain, and they potentially affect neck flexibility during contraction. Surprisingly, 17% of mutations were located in the rod domain of the protein. This part of the protein (LMM) is an α-helical coiled coil structure that forms the core of the thick filament. Mutations in this domain may perturb thick filament dimerisation.15 In MYBPC3, most mutations were frame shift ones, and they were predicted to lead either to a premature truncation of the protein16 or to a cellular quality control, leading to the destruction of the mRNAs that contain the premature termination codon, which results in the absence of the protein.17 Among the missense MYBPC3 mutations, only the V896M variant remains unclear. Thus, it was not considered a disease-causing mutation, but it seems to act as a modifier. TNNT2 mutations are located in regions essential for anchoring the troponin-tropomyosin complex onto the thin filament.18 In 2 unrelated patients, a termination codon (W287X) involving the last residue of the protein was identified. All TNNI3 mutations were located in the carboxy-terminus part of troponin I, which is the first binding site to cardiac troponin C. MYL2 mutations are predicted to alter the phosphorylation site and the Ca++ binding properties.19 One donor-site splice mutation (IVS5-1:a->g) is predicted to lead to a premature termination codon.
In each protein, amino acids may be considered “hot spots” for mutations.20,21 In MYH7, R403 may be mutated to L, Q, or W; R719 to Q or W; and R663 to S or H. In MYBPC3, R502 may be changed to Q or W and D778 to G or E. In TNNI3, R162 may be mutated to W or P and in MYL2, the residue R58 may be mutated to Q or E.
Unexpectedly, 6 families were characterized by a genetic status consisting of more than one mutation in 2 different genes or in the same gene. The first implication is that screening should probably not be stopped after the identification of one mutation, especially in families with a particularly severe phenotype, but should be continued on the same gene and at least on the 2 major genes. Second, in these families, the age at onset, the degree of hypertrophy, or the prognosis was related to the number of mutations. Therefore, it seems to be necessary to check for complex genetic status before establishing phenotype-genotype correlation to understand better the broad expressivity of the disease and to give better genetic counseling to these families.
In conclusion, we report a systematic molecular screening process in a large population of familial and sporadic HCM. Two genes (MYBPC3 and MYH7) account for 82% of all genotyped families. These results and their consequences on cost-efficacy relations might have implications for genetic diagnosis strategy. First, they imply that testing for already known mutations is not helpful and that systematic screening is feasible in clinical practice, despite the genetic heterogeneity of HCM. Second, they imply that these 2 genes should be systematically tested as a first approach. The development of genotyping in HCM based on this more accurate approach, along with the increasing knowledge about relations between the genotype and the phenotype, should lead to improved genetic counseling and better clinical management in families with HCM.22
We thank the family members for their collaboration. This work was supported by “Assistance Publique-Hôpitaux de Paris” and by grants from the Leducq Foundation. It is dedicated to the memory of Jean Leducq.
Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task force on the definition and classification of cardiomyopathies. Circulation. 1996; 93: 841–842.
Maron BJ, Gardin JM, Flack JM, et al. Prevalence of hypertrophic cardiomyopathy in a general population of young adults: echocardiographic analysis of 4111 subjects in the CARDIA study. Circulation. 1995; 92: 785–789.
Bonne G, Carrier L, Richard P, et al. Familial hypertrophic cardiomyopathy: from mutations to functional defects. Circ Res. 1998; 83: 579–593.
Blair E, Redwood C, Ashrafian H, et al. Mutations in the gamma(2) subunit of AMP-activated protein kinase cause familial hypertrophic cardiomyopathy: evidence for the central role of energy compromise in disease pathogenesis. Hum Mol Genet. 2001; 10: 1215–1220.
Charron P, Dubourg O, Desnos M, et al. Diagnostic value of electrocardiography and echocardiography for familial hypertrophic cardiomyopathy in a genotyped adult population. Circulation. 1997; 96: 214–219.
Richard P, Isnard R, Carrier L, et al. Double heterozygosity for mutations in the beta-myosin heavy chain and in the cardiac myosin binding protein C genes in a family with hypertrophic cardiomyopathy. J Med Genet. 1999; 36: 542–545.
Blair E, Redwood C, de Jesus Oliveira M, et al. Mutations of the light meromyosin domain of the beta-myosin heavy chain rod in hypertrophic cardiomyopathy. Circ Res. 2002; 90: 263–269.
Carrier L, Bonne G, Bährend E, et al. Organization and sequence of human cardiac myosin binding protein C gene (MYBPC3) and identification of mutations predicted to produce truncated proteins in familial hypertrophic cardiomyopathy. Circ Res. 1997; 80: 427–434.
Wagner E, Lykke-Andersen J. mRNA surveillance: the perfect persist. J Cell Sci. 2002; 115: 3033–3038.
Szczesna D, Ghosh D, Li Q, et al. Familial hypertrophic cardiomyopathy mutations in the regulatory light chains of myosin affect their structure, Ca2+ binding, and phosphorylation. J Biol Chem. 2001; 276: 7086–7092.
Moolman JC, Brink PA, Corfield VA. Identification of a new missense mutation at Arg403, a CpG mutation hotspot, in exon 13 of the β-myosin heavy chain gene in hypertrophic cardiomyopathy. Hum Mol Genet. 1993; 2: 1731–1732.
Consevage M, Salada GC, Baylen BG, et al. A new missense mutation, Arg719Gln, in the β-cardiac heavy chain myosin gene of patients with familial hypertrophic cardiomyopathy. Hum Mol Genet. 1994; 3: 1025–1026.
Charron P, Heron D, Gargiulo M, et al. Genetic testing and genetic counselling in hypertrophic cardiomyopathy: the French experience. J Med Genet. 2002; 39: 741–746.