Codon 102 of the Cardiac Troponin T Gene Is a Putative Hot Spot for Mutations in Familial Hypertrophic Cardiomyopathy
Background Familial hypertrophic cardiomyopathy is a phenotypically and genetically heterogeneous disease. In some families, the disease is linked to the CMH2 locus on chromosome 1q3, in which the cardiac troponin T gene (TNNT2) has been identified as the disease gene. The mutations found in this gene appear to be associated with incomplete penetrance and poor prognosis. Because mutational hot spots offer unique possibilities for analysis of genotype-phenotype correlations, new missense mutations that could define such hot spots in TNNT2 were looked for in unrelated French families with familial hypertrophic cardiomyopathy.
Methods and Results Family members were genotyped with microsatellite markers to detect linkage to the four known disease loci. In family 715, analyses showed linkage to CMH2 only. To accurately position potential mutations on TNNT2, its partial genomic organization was established. Screening for mutations was performed by single-strand conformation polymorphism analysis and sequencing. A new missense mutation, Arg102Leu, was identified in affected members of family 715 because of a G→T transversion located in the 10th exon of the gene. Penetrance of this new mutation is complete; echocardiographic data show a wide range of hypertrophy; and there was no sudden cardiac death in this family.
Conclusions The codon 102 of the TNNT2 gene is a putative mutational hot spot in familial hypertrophic cardiomyopathy and is associated with phenotypic variability. Analysis of more pedigrees carrying mutations in this codon is necessary to better characterize the clinical and prognostic implications of TNNT2 mutations.
Familial hypertrophic cardiomyopathy (FHC) is a genetically and phenotypically heterogeneous disease transmitted as an autosomal dominant trait. None of the previous hypotheses of the pathophysiological mechanisms would have predicted that defects in sarcomeric protein genes could be a possible molecular basis for the disease. The results of molecular genetic studies have nevertheless shown that many forms of the disease involve mutations in genes encoding sarcomeric proteins (for a review, see References 1 through 3): β-myosin heavy chain in the CMH1 locus, cardiac troponin T in CMH2, α-tropomyosin in CMH3, and cardiac myosin binding protein-C in CMH4.4 5 The degree and distribution of hypertrophy and the type and severity of clinical manifestations vary markedly, not only from one mutation to the other but also between individuals bearing the same mutation. Moreover, genetic analysis has revealed the presence of clinically healthy individuals carrying the mutant allele. These findings indicate that in addition to the responsible gene, environmental factors and/or other genetic factors probably play a major role in the phenotypic expression of each mutation. In this context, phenotype/genotype analysis of patients bearing different mutations in the same codon is of particular interest. Two hot spots for mutations were found in the β-cardiac myosin heavy chain gene, codon 403, and codon 719.6 7 8 9 By screening 28 French families with FHC, we found one family, family 715, linked to the CMH2 locus but not to the other loci. This report describes a novel mutation in the cardiac troponin T gene (TNNT2) in this family that could define the first hot spot for mutations in this gene. We also have partially established the genomic organization of TNNT2, allowing a more accurate positioning for this new mutation.
Clinical evaluation was performed as previously described,6 and blood samples were drawn from family 715 members. The diagnosis of all affected pedigree members was obtained by both ECG and two-dimensional echocardiography and was independently reviewed by three cardiologists who had no knowledge of the genetic diagnosis. In adults, the main criteria for FHC were the presence of left ventricular hypertrophy (end-diastolic wall thickness >13 mm) without cause, major ECG abnormalities (abnormal Q waves and/or unexplained left ventricular hypertrophy), or both. In the case of individual III-6, apical left ventricular hypertrophy was detected by MRI. The study presented here was performed after informed consent was obtained in accordance with the guidelines set down by the Comite´ d'Ethique du Centre Hospitalier Universitaire de la Pitie´-Salpe´trie`re (Paris, France).
Genomic DNA was extracted from blood samples taken from members of 28 French families. The microsatellite markers were analyzed by use of the following polymerase chain reaction (PCR; Hybaid-Omnigene PCR machine) conditions and multiplex procedure: Three markers were simultaneously amplified in a 25-μL final volume by a PCR with 35 cycles; each standard cycling program consisted of 1 minute at 94°C, 1 minute and 20 seconds at 55°C, and 1 minute at 72°C. Taq DNA polymerase and its 10× Mg-containing buffer were purchased from Boehringer. PCR products were resolved according to size by denaturing gel electrophoresis, transferred, and hybridized with an elongated (CA)15 oligonucleotide labeled with peroxidase, and the genotypes were revealed by chemiluminescence (Amersham). The previously described intragenic microsatellite markers MYOI, MYOII6 (for the CMH1 locus on chromosome 14q11), and HTMαCA10 were used and obtained from the genetic maps published by the Ge´ne´thon.11 The following list corresponds to all the markers and loci tested and to all the FHC candidate genes. For the CMH3 locus on chromosome 15q2, the loci (markers) tested were D15S153 (AFM205ye3), D15S125 (AFM214xd10), and D15S131 (AFM262xb1); for the CMH4 locus on chromosome 11p13-q13, they were D11S1763 (AFM162xg1), D11S4191 (AFM338wc1), D11S1883 (AFM039xg3), and D11S913 (AFM164zf12); and for the CMH+WPW locus on chromosome 7q3, they were D7S688 (AFM324zf9), D7S505 (AFM199zd4), and D7S483 (AFM074xg5). Fig 1⇓ shows the markers and loci corresponding to the CMH2 locus.
RNA Isolation and cDNA Synthesis
Epstein-Barr virus–transformed lymphoblastoid cell lines were prepared from family members' peripheral blood samples as described.4 Total cellular RNA was isolated with RNA Plus (Bioprobe Systems), and the TNNT2 cDNA synthesis was performed as described.4
PCR Amplification of Genomic DNA and cDNA of Cardiac Troponin T
A touchdown PCR protocol was performed between 70°C and 60°C or between 65°C and 60°C in two-degree stages, two cycles per degree. The cDNA products were amplified in a first round of PCR in a 50-μL reaction. Three overlapping fragments were obtained by a second round of PCR performed with a final dilution of 1:100 of the first-round PCR products with the following nested PCR primers: 1F and 311R (fragment A), 282F and 606R (fragment B), and 570F and 879R (fragment C). Genomic PCR was carried out on 120 ng DNA with the primers 250F and 393R (Fig 2B⇓). The primers were numbered according to the previously published cDNA sequence.12
Genomic Structure of TNNT2
The genomic organization of TNNT2 gene was partially established by use of standard procedures. This was done by screening a human genomic DNA library (Clonetech No. HL1111j) with cardiac troponin T cDNA sequences13 and sequencing selected clones.
Detection of Mutations
For single-strand conformation polymorphism analysis, PCR products were denatured for 5 minutes at 96°C in a standard denaturing buffer, kept on ice for 5 minutes, loaded onto 8% polyacrylamide gels, and then run at 7°C and 6 mA in a Hoeffer apparatus. The bands were visualized after silver staining of the gels (Bio-Rad). The sequencing of PCR fragments was performed by the dideoxynucleotide chain termination method with fluorescent dideoxynucleotides on an Applied Biosystem 373A DNA sequencer (Perkin-Elmer/ABI). Finally, for restriction endonuclease cleavage analysis, PCR products (10 μL) were digested for 4 hours at 37°C by 10 U Msp I (Boehringer) with the provided reaction buffer and the recommended reaction conditions. The digested fragments were analyzed on a 3% Nusieve (Seakem) agarose gel.
Fig 1⇑ shows the pedigree of family 715. Four members—II-4, II-7, III-5, and III-6—were affected; the Table⇓ gives their clinical status. Individuals II-4 and III-5 exhibited a large maximal left ventricular hypertrophy (19 and 35 mm, respectively) compared with individuals III-6 and II-7 (8 and 10 mm, respectively). The type of hypertrophy varied from one individual to another (septal in individuals II-4 and III-5; absent in individual II-7; apical in individual III-6). On ECG, deep Q waves were observed in leads D2, D3, VF, V4, V5, and V6 (5 to 7 mm) in individual II-7 and in leads D2, D3, and VF (4 to 5 mm) in individual III-6. Such deep Q waves are not found in normal individuals. Individual II-4 died in 1992 at 44 years of age of refractory end-stage heart failure while on a waiting list for heart transplantation. Individual II-6 died young at 36 years of age of end-stage heart failure. This woman had no major abnormalities at ECG, but echocardiography showed abnormal dilation of the left ventricle that was globally hypokinetic. Although individual I-2 had no major abnormalities on ECG, a roentgenogram record taken in 1973 showed an increased cardiothoracic index. This individual died of global cardiac deficiency at 57 years of age. Considering the family background and the clinical features presented at the time of their deaths, it is likely that these last two individuals were affected. No evidence of FHC either on ECG or echocardiography was found for individuals II-2, II-5, III-4, and III-7. Individuals III-1, III-2, and III-3 were also checked by ECG, which revealed the same normal pattern as for individual II-2.
Haplotype analysis with microsatellite markers of loci CMH1, CMH3, and CMH4 identified affected recombinant individuals (data not shown). This excluded linkage to these loci. The most likely haplotypes obtained with CMH2 microsatellite markers indicated that the haplotype 8-5-2-8-10-5 cosegregates with the disease, suggesting that the disease gene of family 715 is linked to this locus (Fig 1⇑). The haplotypes of individuals II-6 and III-8 were not determined because no blood samples were available. The small size of the family did not allow us to obtain significant logarithm of the odds [LOD] scores (simulated LOD score, 1.45). Nevertheless, on the basis of the above haplotypes, it is reasonable to assume that the disease gene of family 715 is TNNT2.
Partial Human TNNT2 Genomic Structure
Because the mutations previously found in human TNNT2 referred to an exon numbering corresponding to the rat gene,14 we have partially analyzed the human gene organization to use an appropriate numbering. We have found that the human TNNT2 gene presents an additional exon located between exons 3 and 4 of the corresponding rat genomic sequence, so exon 4 in the rat corresponds to exon 5 in the human (data not shown; manuscript in preparation).
The numbering commonly used for the amino acid sequence refers to the cardiac troponin T isoform found in the adult heart.15 We have added to this numbering the 10 amino acids corresponding to exon 5, which is alternatively spliced during development13 16 to take into account the full coding potential of the gene in the description of new mutations. The accession numbers of the sequences described here are X98478 and X98481.
Single-strand conformation polymorphism analysis of TNNT2 cDNA fragment A amplified by primers 1F and 311R (see “Methods”) detected a different pattern between affected and healthy individuals of family 715 (data not shown). The corresponding genomic DNA fragment was amplified with primers 250F and 393R (see “Methods”), and again a difference in the single-strand conformation polymorphism analysis pattern was observed in the affected compared with unaffected individuals (data not shown). Direct sequencing of the genomic PCR fragment shows a C→A mutation on the noncoding strand in affected individual II-7 (Fig 2A⇑). This mutation corresponds to a G→T transversion at the 11th nucleotide of TNNT2 exon 10 (equivalent to exon 9 in rat; Fig 2B⇑), which corresponds to nucleotide residue 287 (G287T) in the previously described adult TNNT2 cDNA sequence.12 At the amino acid level, this mutation results in an arginine-to-leucine substitution at codon 102 (Arg102Leu). This transversion fortuitously suppresses an Msp I restriction site, a feature we used to screen more individuals. Only affected members of family 715 had lost the Msp I site (Fig 2C⇑). Furthermore, on 42 probands with FHC and on 100 normal chromosomes from unrelated individuals, complete Msp I digestion of the PCR products was found (data not shown).
To give a more accurate description of the mutations that may be found on the human TNNT2 gene, we analyzed the genomic organization of the gene. From this, it is now possible to identify the position of the mutations within exons, including those alternatively spliced during development, and to use an amino acid numbering that reflects the full coding potential of this gene.
The Arg102Leu mutation we found in the TNNT2 gene is due to a G→T transversion at the second nucleotide of codon 102 located in exon 10. This mutation is associated with FHC in family 715 and is not simply a polymorphism. Indeed, the mutation was found only in the four affected individuals and not in the healthy members of family 715; the mutation was absent on 100 chromosomes of 50 healthy unrelated subjects and in 42 unrelated probands; and finally, arginine 102 is highly conserved during evolution,15 suggesting its key role in the structure and/or function of cardiac troponin T in the sarcomere.
An interesting feature of this Arg102Leu mutation is that it occurs within a site C/CGG that is cleaved by the restriction enzyme Msp I. Another mutation, at the same codon of the TNNT2 gene and thus called Arg92Gln, has been reported in three unrelated families with FHC exhibiting different haplotypes.17 This is due to a G→A transition at the second nucleotide of the codon.17 We will now refer to it as Arg102Gln. The fact that both codon 102 mutations described to date occurred in the same CpG site strongly suggests that this is likely to be a hot spot for mutation. Cytosines in the CpG sequences are frequently methylated in mammals.18 The resulting 5-methylcytosines are hypermutable because of the modification of 5′ cytosine by cellular DNA methyltransferases, which very frequently leads to spontaneous deamination of 2-methylcytosine to thymidine.19 As a consequence, mutations within CpG dinucleotides on either the coding or the noncoding strand of the gene contribute significantly to the incidence of human genetic diseases.19 Data obtained from human factor IX mutations led to the estimation that the dinucleotide CpG within human coding sequences is up to 40 times more mutable than predicted from random mutations.19 Theoretically, another C→T transition is possible at the 102 codon, but we did not find it in any of the 42 unrelated probands with FHC who were analyzed.
Comparison of the clinical features of family 715 (the Arg102Leu mutation) with those reported for the three other families (the Arg102Gln mutation) shows variations, even though in all families the incidence of disease-related deaths is high (2 of 5 in family 715 and 15 of 32 in the others). For instance, in family 715 there was no sudden cardiac death, and the two disease-related deaths were due to heart failure. In the other described families, prognosis was particularly poor, with 11 sudden deaths out of 15 disease-related deaths.17 In addition, the penetrance of the mutation was complete in family 715, whereas it was not in the other families (7 of the 32 genetically affected individuals are phenotypically healthy). Furthermore, in the three families reported previously with the Arg102Gln mutation,17 it was concluded that TNNT2 gene mutations were associated with a relatively mild left ventricular hypertrophy, even though there was a wide variation in ventricular wall thickness (15±6 mm) in the reported data. This conclusion does not apply to family 715 in which two individuals had a large degree of hypertrophy. Finally, a striking feature in family 715 was the variability of the phenotype in terms of both the extent and type of hypertrophy. Large-scale screening of probands through the simple restriction analysis described here could allow us to rapidly identify other families with mutations at codon 102 of the TNNT2 gene. This would be especially important in terms of genetic testing and the search for modifier genes in the context of the high variability of the phenotypic expression of TNNT2 gene mutations. The contribution of specific mutations to the development of FHC remains to be assessed. Considering the variability of the expression of the phenotype, it is very likely that many other genetic factors will need to be taken into consideration.
This work was supported by INSERM (Re´seau de Recherches Cliniques No. 492010), the Association Franc¸aise contre les Myopathies, the Assistance Publique-Hoˆpitaux de Paris (Contrat Emul), the Institut Fe´de´ratif de Recherches No. 14, the Fe´de´ration Franc¸aise de Cardiologie, and the British Heart Foundation. Dr Bonne is a recipient of a grant from the Fondation Bettencourt-Schueller. We are indebted to the family members for their invaluable participation. We are also grateful to M. Bennaceur for blood sample collection. We wish to thank Pascale Guicheney for her critical reading of the manuscript.
- Received July 26, 1996.
- Revision received September 26, 1996.
- Accepted September 30, 1996.
- Copyright © 1996 by American Heart Association
Schwartz K, Carrier L, Guicheney P, Komajda M. The molecular basis of cardiomyopathies. Circulation. 1995;91:532-540.
Watkins H, Seidman JG, Seidman CE. Familial hypertrophic cardiomyopathy: a genetic model of cardiac hypertrophy. Hum Mol Genet.. 1995;4:1721-1727.
Marian AJ, Roberts R. Recent advances in the molecular genetics of hypertrophic cardiomyopathy. Circulation. 1995;92:1336-1347.
Bonne G, Carrier L, Bercovici J, Cruaud C, Richard P, Hainque B, Gautel M, Labeit S, James M, Beckman J, Weissenbach J, Vosberg HP, Fiszman M, Komajda M, Schwartz K. Cardiac myosin binding protein-C gene splice acceptor site mutation is associated with familial hypertrophic cardiomyopathy. Nat Genet.. 1995;11:438-440.
Dausse E, Komajda M, Dubourg O, Fetler L, Dufour C, Carrier L, Wisnewski C, Bercovici J, Hengstenberg C, Al-Mahdawi S, Isnard R, Hagege A, Bouhour JB, Desnos M, Beckmann JS, Weissenbach J, Schwartz K, Guicheney P. Familial hypertrophic cardiomyopathy: microsatellite haplotyping and identification of a hot-spot for mutations in the β-myosin heavy chain gene. J Clin Invest.. 1993;92:2807-2813.
Consevage M, Salada GC, Baylen BG, Ladda RL, Rogan PK. 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.
Moolman JC, Brink PA, Corfield VA. Identification of a new missense mutation at Arg403, a CpG mutation hot spot, in exon 13 of the β-myosin heavy chain gene in hypertrophic cardiomyopathy. Hum Mol Genet.. 1993;2:1731-1732.
Greve G, Mares A, Bachinski L, Roberts R. A sporadic mutation in the β-myosin heavy chain gene transmits hypertrophic cardiomyopathy to the offspring of two generations. Circulation. 1993;88(suppl I):I-572. Abstract.
Mesnard L, Logeart D, Taviaux S, Diriong S, Mercadier JJ, Samson F. Human cardiac troponin T: cloning and expression of new isoforms in the normal and failing heart. Circ Res. 1995;76:687-692.