Idiopathic dilated cardiomyopathy (DCM) is the most common cause of heart failure and cardiac transplantation in the young. The condition is characterized by unexplained left ventricle dilation, impaired systolic function, and nonspecific histologic abnormalities dominated by myocardial fibrosis. Patients may experience severe disease complications including arrhythmia, thromboembolic events, and sudden death. Recent studies have suggested familial disease in 30% to 50% of cases. Autosomal dominant transmission is most frequent followed by recessive, X-linked, and mitochondrial inheritance. Most affected families present with a “pure” cardiac phenotype while syndromic features are less frequent. Disease-causing mutations have been reported in genes encoding cytoskeletal and nuclear envelope proteins and, more recently, in sarcomeric contractile protein genes (1- 3). At present, 10 different mutations in 12 families with autosomal dominant DCM have been identified in the sarcomeric genes for alpha-cardiac actin (ACTC), alpha-tropomyosin (TPM1), cardiac troponin T (TNNT2), beta-myosin heavy chain (MYH7), and titin (TTN) (4- 11). In addition, we recently reported the first recessive mutation in cardiac troponin I (TNNI3) in a small DCM family with early and severe disease expression (10).

The fact that DCM mutations have been identified in TNNT2 and TNNI3 led us to investigate whether mutations in the remaining part of the troponin complex (cardiac troponin C [TNNC1]) were present in DCM. This study examines the relationship between genotype and clinical phenotype in families with mutations in TNNC1 and TNNT2 identified by mutation analysis of 235 consecutive DCM patients and their relatives. In addition, we report the effect of TNNC1 and TNNT2 mutations identified on inter-troponin interactions by the use of a qualitative mammalian two-hybrid luciferase assay.

The local research ethics committee approved the study, and informed consent was obtained from all participants.

The study cohort consisted of 235 consecutive and unrelated DCM probands who were evaluated in a dedicated cardiomyopathy clinic from 1995 to 2002. Relatives of DCM probands in whom a TNNT2 or TNNC1 mutation was identified were invited for clinical assessment. All participants underwent physical examination, 12-lead electrocardiogram, transthoracic two-dimensional echocardiography, and Doppler studies. Cardiac catheterization was performed in patients aged >40 years, and in those individuals with symptoms of ischemic heart disease or exercise-induced ST-segment changes consistent with myocardial ischemia. Familial disease was defined where one or more relatives had DCM identified by clinical investigation or had a family history of unexplained premature cardiac death <40 years.

Standard measurements of left ventricular end-diastolic dimension (LVEDD) and left ventricular end-systolic dimension (LVESD) were performed, and fractional shortening was calculated as (LVEDD − LVESD/LVEDD) × 100. Dimensions were corrected for age and body surface area (BSA) according to the formula of Henry [(LVEDD)(percent predicted) = measured LVEDD/predicted LVEDD × 100; predicted LVEDD = [45.3 × BSA^0.3] − [0.03 × age] − 7.2 (12).

Dilated cardiomyopathy was diagnosed in accordance with World Health Organization diagnostic criteria when echocardiography identified unexplained left ventricle dilation and impaired contractile performance: left ventricular end diastolic diameter >117% predicted for age and body surface area and a fractional shortening <25%. Relatives of DCM probands were considered to be affected if they had unexplained left ventricle enlargement >112% of predicted value or experienced unexplained heart failure or sudden cardiac death at <40 years of age (1,13). Patients with symptoms or signs of skeletal muscle disease were excluded from the study.

Genomic deoxyribonucleic acid was obtained, and protein encoding exons of TNNT2 and TNNC1 were amplified using standard protocols (primer sequences and conditions for polymerase chain reaction amplification available upon request). Mutation analysis of TNNT2 was performed by fluorescent SSCP (F-SSCP) and direct sequencing of abnormal conformers. As reported previously, the sensitivity of F-SSCP to identify sequence variations in TNNT2 was 100% (14); TNNC1 was investigated by denaturing high performance liquid chromatography (DHPLC)analysis and abnormal elution profiles subjected to direct sequencing. Initially, 50 patients were investigated for mutations in TNNC1 by direct sequencing without identifying any sequence variations. Subsequent DHPLC analysis of the remaining patients identified one sequence variation indicating a high sensitivity of this investigation as previously reported (14). When a mutation was identified, a second blood sample from the affected patient was reanalyzed to confirm the initial finding by direct sequencing, F-SSCP, or DHPLC analysis.

To delineate the effect of TNNC1 and TNNT2 mutations identified on inter-troponin interactions, we used a qualitative mammalian two-hybrid luciferase assay. The TNNT2 cDNA was cloned into the pBIND plasmid (Promega, Madison, Wisconsin) and the TNNI3 and TNNC1 cDNA were cloned into the pACT plasmid (Promega). The cloning result was confirmed by direct sequencing and compared with accession numbers X90780 (TNNI3), AY044273 (TNNT2), and M37984 (TNNC1) (15). Mutations identified in TNNT2 (R131W, R205L, δK210, D270N, and a recognized polymorphism, K253R) and TNNC1 (G159D) were introduced into pBIND-TNNT2 and pACT-TNNC1, respectively, by oligonucleotide-mediated site-directed mutagenesis. HEK293 cells were cotransfected with equimolar amounts of pBIND-TNNT2 (wild-type [wt] or mutant), pACT-TNNI3, or pACT-TNNC1 (wild-type or mutant) and the pG5Luc reporter plasmid. After 48 h, cell lysates were assayed for Photinus and Renilla luciferase using the Promega dual-luciferase assay system. Data was normalized against Renilla luciferase activity to account for differences in transfection efficiency. The values for the interaction between wild-type protein troponin T (cTnT), I (cTnI), and C (cTnC) were arbitrarily set to 100%. The values for interactions between mutant cTnT and cTnC or cTnI were expressed relative to this. All experiments were repeated 10 times. The values for luciferase activity resulting from interaction of mutated cTnT–wild-type cTnI and mutated cTnT–wild-type cTnC were compared with luciferase activities obtained from interaction of wild-type cTnT–wild-type cTnI and wild-type cTnT–wild-type cTnC by use of a t test. The values for luciferase activity were distributed in accordance with a normal distribution, and an F-test showed homogeneous distribution of variances (data not shown). Likewise, values for luciferase activity resulting from interaction of mutated cTnC–wild-type cTnI and mutated cTnC–wild-type cTnT were compared with luciferase activities obtained from interaction of wild-type cTnC–wild-type cTnI and wild-type cTnC–wild-type cTnT by use of same t test and conditions. Control experiments consisting of cotransfections of pBIND, pG5Luc, and pACT-TNNC1 or pACT-TNNI3 (wild type and mutants) as well as pBIND-TNNT2 (wild-type and mutants), pG5Luc, and pACT were performed without observing any interaction (data not shown).

A total of 43% (102 of 235) of the study cohort had one or more relatives with DCM identified by clinical investigation or a family history of premature cardiac death at <40 years of age. A total of 85% of these families (87 of 102) exhibited an autosomal dominant mode of inheritance, whereas the remaining 15% (15 of 102) transmitted the condition consistent with autosomal recessive inheritance, although autosomal dominant inheritance with incomplete penetrance could not be excluded. Inheritance patterns suggestive of maternal or X-linked inheritance were not observed.

Mutation analysis of TNNC1 in family A identified a missense mutation that was predicted to result in a G159D amino acid substitution ((Figure 1),Table 1). The mutation was present in all affected individuals and absent in 200 chromosomes from ethnically matched control individuals. The proband, IV-4, had a sudden onset of heart failure symptoms at the age of 21 years. His cardiac function deteriorated rapidly, and he underwent cardiac transplantation two months after initial presentation. Two additional individuals, III-1 and IV-2, received cardiac transplants at the age of 52 and 22 years; III-1 had been evaluated at the age of 50 years at the time of his son's transplant of which he had normal systolic function. However, two years later he developed dyspnea and edema of the lower extremities due to heart failure and required transplantation two months after onset of symptoms. The mother of the proband, III-4, who was an obligate mutation carrier, died at the age of 45 years awaiting cardiac transplant, while II-1 died of heart failure (age 62 years), after eight months of medical treatment. Postmortem examination showed dilation of all cardiac chambers, and a heart weight of 540 g; III-6 experienced unexplained sudden death at the age of 21 years; IV-7 was diagnosed with DCM at the age of 17 years and has been stabilized on angiotensin-converting enzyme inhibitors for the past year, whereas IV-5, age 36 years, has left ventricle enlargement on echocardiography and electrocardiographic abnormalities including T-wave inversion in V1 to V2 and left-axis deviation.

Mutation analysis of TNNT2 identified four different mutations in four families of which three were novel (family B, R131W; family C, R205L; family E, D270N) and one has previously been reported (family D, δK210) ((Figure 1), Table 1) (7,9). The mutations identified segregated with the disease in each family and were absent in 200 chromosomes from ethnically matched control individuals and 1,520 chromosomes from patients with hypertrophic cardiomyopathy. The disease expression in the four families with TNNT2 mutations was similar to family A in severity. Three individuals (family C, III-1; family D, III-4; family E, II-1) received cardiac transplants at the age of 16, 22, and 38 years, respectively. Three individuals died of heart failure (family B, I-2; family C, III-2; family D, III-3), three died suddenly (family B, II-2; family C, I-2; family D, II-1), whereas four have remained stable on conventional heart failure therapy.

In this study, 21 individuals were identified with mutations in TNNC1 (n = 8) and TNNT2 (n = 13) ((Figure 1), Table 1). Five died of heart failure, and six received a cardiac transplant (average age: 33 years, range: 20 to 62 years) <6 months after initial diagnosis (range: 2 to 8 months). Four experienced sudden unexplained death (average age: 24 years, range: 16 to 36 years). In total, 71% (15 of 21) experienced premature cardiac death or underwent cardiac transplantation. Six mutation carriers are currently alive (five with DCM and one with left ventricular enlargement) and have remained stable (New York Heart Association functional class I) on heart failure therapy during an average observation period of 30 months (range: 20 to 150 months) (Table 1). The average number of premature cardiac deaths per family with troponin disease (2.6 deaths/family) was significantly higher compared with the death ratio observed in the remaining families of the cohort with hereditary DCM (0.7 deaths/family) (p < 0.001, Fisher exact test). The penetrance of mutations in both TNNC1 and TNNT2 was 100%.

Myocardial tissue from eight patients was available for histologic evaluation (two biopsies obtained antemortem [D: III-3; E: II-1], two hearts from autopsies [A: III-4; C: II-2], and four explanted hearts [A: III-1, IV-2, IV-4; C: III-1]). All specimens showed varying degrees of nonspecific abnormalities including myocyte hypertrophy, increased interstitial fibrosis, and endocardial thickening with smooth muscle cells characteristic of DCM (16). There was no significant myocyte disarray characteristic of hypertrophic cardiomyopathy or features suggesting storage disease in any of the specimens evaluated (17- 18).

The mutations identified were all localized in conserved and functionally important domains of the proteins involved in the interaction with other troponin subunits and/or tropomyosin ((Table 2) and (Table 3),Figure 2) (19- 23). The effect of the mutations on cTnI, cTnT, and cTnC protein interactions was qualitatively assessed by a mammalian two-hybrid assay (Figure 3). There was a significant impairment of mutated cTnT_R131W, _R205L, _δK210, _D270N–cTnI_wt and cTnT_R131W, _R205L, _δK210, _D270N–cTnC_wt protein interactions compared with wild-type controls (t test: p < 0.001). Likewise, a significant impairment of mutated cTnC_G159D–cTnT_wt protein interaction was observed, whereas the interaction between cTnC_G159D–cTnI_wt was significantly enhanced (p values for both experiments <0.001). No significant change was observed after similar inter-troponin protein interaction studies of a recognized cTnT amino acid polymorphism (K253R), which appeared with a frequency of 2.3% in controls (Figure 3).

All mutations identified were believed to be disease-causing because: 1) they co-segregated with disease in each of the five families; 2) no sequence variations were identified in 200 ethnically matched control chromosomes investigated in TNNT2 and TNNC1 or 1,520 chromosomes from hypertrophic cardiomyopathy patients investigated in TNNT2; 3) the mutations were localized in conserved and functionally important regions of the genes; and 4) functional studies of the mutated proteins showed altered troponin protein-protein interactions.

Cardiac troponin C was identified as a novel disease gene in DCM. In addition, four disease-causing mutations were found in TNNT2, whereas our previous investigation of the same cohort of patients for autosomal dominant mutations in TNNI3 was negative (10). The overall frequency of mutations in the troponin complex in familial DCM was 6% (6 of 102) including a previously reported recessive TNNI3 mutation (10). The disease expression appeared malignant because 71% (15 of 21) of mutation carriers experienced premature cardiac death or received a cardiac transplant mostly by the fourth decade with an average duration of six months from diagnosis to event. Recent investigations of TNNT2 in DCM families have identified a δK210 deletion in three unrelated families and an R143W amino acid substitution in one family (6- 7,9). The clinical disease expression in these families was similarly severe, and a significant number of affected individuals experienced premature cardiac death/transplantation in the second or third decade of life. However, the disease expression reported was more variable, and healthy mutation carriers without signs and symptoms of disease were present even at older ages. The design of previous studies did not allow an estimate of the prevalence of TNNT2 mutations in consecutive DCM patients.

Overall, the clinical expression of mutations in the troponin complex was associated with an adverse prognosis in DCM. However, this and other studies have reported the clinical findings in patients diagnosed over decades in which advances in pharmacologic treatment have substantially improved the prognosis of patients with heart failure. Although a limited number of patients with DCM in this study are currently alive with a relatively short observation period, it is of note that their condition appears stable on modern pharmacologic therapy with angiotensin-converting enzyme inhibitors and beta-blockers. The results of previous drug trials of patients with heart failure have indicated a beneficial effect of prophylactic pharmacologic therapy in asymptomatic individuals with impaired systolic function (24- 25). Therefore, it is reasonable to assume that early diagnosis and treatment of asymptomatic patients with hereditary DCM may also improve their prognosis (26). The many premature cardiac deaths observed suggest that regular clinical investigation of relatives at risk of disease development is warranted to ensure early diagnosis. Genetic diagnosis would be helpful in identifying unaffected mutation carriers who require follow-up and would enable termination of clinical screening of relatives with a normal genotype. In addition, knowledge of the specific disease gene and mutation present in a family may allow risk-stratification, although limited data are available in relation to sarcomeric gene mutations. So far, few families have been identified, and only the genes for ACTC, TPM1, and TNNI3 have been fully investigated in large cohorts of DCM families with a frequency of mutations that account for <3% of all cases (4- 5,10). Most other patient populations studied were small in number, and our findings reemphasize the importance of designing genetic studies in larger DCM populations to describe the prevalence of these mutations. The fact that we investigated patients with a “pure” cardiac phenotype explains that we did not observe inheritance patterns suggestive of maternal or X-linked inheritance because these conditions are often associated with symptoms of systemic disease. In maternal inheritance caused by mithocondrial gene mutations, patients typically have multiorgan disease including dilated or hypertrophic cardiomyopathy, neurological symptoms, and skeletal muscle impairment. In X-linked inheritance caused by dystrophin mutations, the disease expression is most commonly associated with both DCM and skeletal muscle impairment, although a few patients with a “pure” cardiac phenotype have been reported.

Myocardial contractility is largely regulated by the troponin complex that, in turn, is influenced by the concentration of intracellular Ca²⁺. The complex is distributed along the thin filament of the sarcomere and interacts with actin and tropomyosin. During systole, Ca²⁺ binds to cTnC and introduces conformational changes of the troponin complex that attenuate the inhibitory effect of cTnI. This enables the myosin head of the thick filament to interact with actin and generate force. The Ca²⁺ concentration also influences cTnC-cTnT interaction which is important for control of sliding velocity between thick and thin filament (19- 23). Interestingly, recent studies have suggested that cTnT is essential, not only for the structural integrity of the troponin complex, but also for sarcomere assembly and cardiac contractility (27).

The amino acid substitution identified in TNNC1 (G159D) is localized in a domain of the protein constitutively occupied by Ca²⁺. This may change the affinity for Ca²⁺ and, thereby, alter the ability of the troponin complex to regulate myocardial contractility. Also, mutations identified in TNNT2 (R131W, R205L, δK210, D270N) occurred in well-conserved regions essential for interactions within the troponin complex or its interactions with tropomyosin or actin (19- 23,28- 29). The qualitative effect of the mutations on inter-troponin interactions was investigated using a mammalian two-hybrid assay, which showed significant changes for all mutant proteins compared with wild-type. A common TNNT2 polymorphism (K253R) served as a negative control in the test system and did not change inter-troponin interactions. The results indicate that this assay may be helpful in providing supporting evidence to determine if an amino acid change identified in an affected individual is a disease-causing mutation or more likely to be a polymorphism.

Previous functional studies of one of the DCM mutations identified (TNNT2, δK210) showed that mutated protein reduced the Ca²⁺ sensitivity of actomyosin ATPase activity, which resulted in decreased maximum speed of muscle contraction (30- 31). Thus, DCM mutations in the troponin complex may induce a profound reduction in force generation leading to impaired systolic function and cardiac dilation. In addition, it is possible that the myocardium of mutation carriers may be more susceptible to environmental influences such as viruses and toxic agents. It seems likely that the heterogeneous clinical appearance in DCM families with mutations in the troponin complex is a result of interplay between the specific mutation identified, modifier genes, and non-genetic factors in the environment.

In summary, mutations in the three proteins of the troponin complex were not uncommon in DCM. The disease expression was characterized by high penetrance and severe prognosis. The data suggest that genetic investigations of the troponin complex may help to identify a subset of DCM families at high risk of rapid disease development.

The authors thank the families and physicians who made this study possible, in particular Drs. Andrew Mitchell, John Dean, and Peter Townend.