Introduction

Familial hypertrophic cardiomyopathy (FHC) is an autosomal dominant cardiac disorder characterized by unexplained myocardial hypertrophy, predominantly of the interventricular septum.1, 2 The clinical phenotype varies from asymptomatic over limited symptoms to severe heart failure or the occurrence of serious arrhythmias or sudden cardiac death.3

FHC has been associated with more than 200 mutations in 10 known genes encoding cardiac sarcomere proteins.4 Furthermore, mutations in other genes (eg mitochondrial DNA and PRKAG2) have been associated with a cardiac phenotype resembling hypertrophic cardiomyopathy.5, 6 Mutations in MYH7, encoding β-myosin heavy chain, and in MYBPC3, encoding myosin binding protein C, are the most prevalent and probably account for 20–40% of FHC cases.4, 7

We studied the occurrence of MYBPC3 mutations in a cohort of 81 consecutive patients in a nationwide Danish FHC study. We found nine mutations, six of which were novel, in 10 families. In three families, mutations in both MYBPC3 and MYH7 were found. Several of the mutations undoubtedly result in total or partial haploinsufficiency.

Materials and methods

Patients

In all, 81 unrelated FHC families were remitted consecutively from The National University Hospital, Rigshospitalet, Copenhagen, a tertiary referral center for FHC. DNA samples from 100 blood donor samples were used as controls.

Clinical evaluation

Clinical examinations were performed as previously described.8 All probands fulfilled conventional diagnostic criteria for hypertrophic cardiomyopathy.9

Mutation analyses

Mutation analyses were performed by single-strand conformation polymorphism/heteroduplex analysis (SSCP/HA)10 or capillary array single-strand conformation polymorphism (CAE-SSCP)11, 12 of amplicons covering all 34 coding exons of MYBPC3 using intronic primers from13 or defined from the gene sequence (GenBank Accession No. U91629). The sequence of primers and PCR conditions are available upon request. Genetic variants with frequencies above 1% in controls were considered polymorphisms. Seven other FHC-associated genes (ACTC, MYH7, MYL2, MYL3, TNNI3, TNNT2, and TPM1)11, 14, 15 were screened for mutations in the genotyped probands.

Messenger RNA transcript analysis

Ectopic mRNA expression of MYBPC3 in peripheral blood lymphocytes was studied.16, 17, 18 RNA was extracted from fresh blood using an RNA purification kit (Qiagen, Germany). RT-PCR was performed using MMLV Reverse Transcriptase (Stratagene, LA Jolla, CA, USA) and exon specific primers. In case of amplification with the 5′-proximal primer set, MYBPC1EF (5′-CGACCAGGGATCTTACGCATG-3′) and MYBPC12ER (5′-GGGTGCCTGCCGTAGGATCTC-3′) a second, nested PCR amplification was applied on the primary amplicon using primers MYBPC294F (5′-CGACCAGGGATCTTACGCAGT-3′) and MYBPC897R (5′-CCTCA TGGCTATCACTGATCCG-3′). For amplification of the 3′-proximal region, primers MYBPC29EF (5′-CGCTCGCC GCGTGCATTCAG-3′) and MYBPC33ER (5′-CCGTCAAAG GGGCAGGGCTTTC-3′) were applied. RT-PCR products were cloned into pCR-Script (SK) (Stratagene) and sequenced.

Results

Mutation screening

Nine mutations were detected in 10 families with one mutation present in two families (Table 1). Three of the mutations (g5256G>A (Glu258Lys; Figure 1), g16153G>A (Ala833Thr), and g15919insG) have previously been characterized.13, 19 The six novel mutations consisted of a missense mutation (g5166G>A (Asp228Asn)) and a nonsense mutation (g16182C>G; Tyr842Ter), three exonic deletions (g2432delT, g10080-96delCTGACCGTGGAACTGG, and g16086delAAG (DelLys813) and one intronic deletion (g20931-37delCCTGTCA(−8-(−)14)).

Table 1 Mutations in MYBPC3 and their tentative pathogenetic mechanism
Figure 1
figure 1

Detection of g5256G>A (Glu258Lys) mutation and consequent exon skipping. (a) Capillary array SSCP, where the upper two panels represent the wild-type chromatograms (forward (upper) and reverse (lower), respectively); the two lower panels represent the mutant chromatograms (forward (upper) and reverse (lower), respectively); (b) DNA sequence of the wild-type control (upper panel) and the heterozygote (G>A substitution) (lower panel); (c) DNA sequence of RT-PCR clone containing the splice site mutation; upper panel illustrates sequence where exon 6 has been skipped and lower panel the normal cDNA sequence.

The small size of most of the genotyped families makes it difficult to establish a causal relationship between the occurrence of a specific mutation and FHC. MYBPC3 and MYH7 mutations were found in the families E, G and XI, where the latter two carried very rare variants in MYBPC3 of unknown significance besides disease-associated mutations in MYH7.

Messenger RNA studies

In order to determine the consequences of mutations, we performed analysis of ectopically expressed MYBPC3 mRNA in the absence of cardiac tissue. Nested RT-PCR performed on DNA from a g2432ΔT mutation carrier using exon-specific primers MYBPC1EF/MYBPC12ER followed by nested PCR with primers MYBPC294F/MYBPC897R) resulted in one band on a gel, migrating as an approximately 500 bp DNA fragment. Cloning and sequencing revealed that both alleles were present.

Nested RT-PCR analysis in a g5256 mutation carrier, from family ZI, using the primer pair MYBP294F and MYBPC897R for reamplification, yielded two dominant PCR products corresponding to the wild-type allele and an allele with exon 6 skipping (Figure 1c). Exon skipping resulted in a frameshift and a premature termination codon, and, most likely as a consequence in haploinsufficiency.

The deletion in intron 31 (g20931-37delCCTGTCA (−8(−)14)) was shown by RT-PCR using MYBPC29F and MYBPC33R primers to result in both a normally spliced transcript and a transcript with exon 32 skipped.

The remaining splice-site mutations with a possible splice variation, IVS8-20C>A and IVS16-6G>A, and the insertion/deletion mutations could not be analyzed as fresh blood or cardiac tissue from the mutation carriers could not be obtained.

Clinical evaluation of patients with MYBPC3 mutations

Among the 44 carriers with mutations solely in MYBPC3, there were 25 carriers (57%) that had a maximum LVD13 mm and 14 carriers (32%) with HCM symptoms or prior myectomy. In three families, ZF, ZH, and ZP, no major cardiac events had occurred, and in the other families, one to two members had experienced a cardiac event. Three of the probands with a MYBPC3 mutation (C, ZL, and ZV) had a debut age under 18 years with severe hypertrophy. A detailed clinical description of patients with digenic inheritance is shown in Table 2. No specific phenotype could be attributed to these patients.

Table 2 Clinical/characteristics in families with mutations in both MYBPC3 and MYH7

Discussion

The prevalence of MYBPC3 mutations in Denmark, 10/81 (12.3%) is similar to the frequency of 13.6% (15/110) recently reported in a study of Caucasian patients20 and 10.8% (14/130) in a Japanese study13 but much lower than recent reports where 24% of Finnish HCM cases carried MYBPC3 mutations21 and in a French study with a prevalence of 42% MYBPC3 mutations among 124 genotyped patients.4

In agreement with other studies, we find that late onset is characteristic, but in some families, children may be affected.13, 20 The relatively low proportion of mutations leading to single amino-acid substitutions is similar to the findings in the Caucasian study,20 contrary to the findings in a Japanese and a recent French study.4, 13 The g15919insG mutation, previously described by Niimura et al,13 was found in two families (ZF and ZP) and could be a founder mutation, analogous to the Turkish, Finnish and South African founder mutations.20, 22 Haplotype analysis using STRs in the flanking regions as well as SNP analysis in the MYBPC3 shows that the two probands are not closely related.23

The finding of three families with mutations in both MYBPC3 and MYH7 is surprising, but the phenomenon of digenic inheritance in these genes has previously been described in recent French and European studies.4, 24 These findings should lead to a re-evaluation of all patients diagnosed with HCM based on MYBPC3 mutations, since they may carry the true disease-causing mutations on other genes.

Haploinsufficiency

Six out of the 10 mutations found in this study resulted in generation of a premature termination codon (Table 1). For at least three of these (g2432ΔT, g5256G>A, and g10080-96delCT…GG), and possibly also the remaining three, the translation product will be so short as to prevent effectively its incorporation in the sarcomere; thus in these three cases, there will be a quantitative rather than a qualitative deficiency of the sarcomere – haploinsufficiency – as mRNAs containing premature termination codons are often rapidly degraded through the so-called nonsense mediated mRNA decay.25 However, the splice variants that were observed may be artifacts due to ectopic expression in leukocytes, and final proof of aberrant splicing for mRNA or protein studies may only be documented in cardiac tissue, which has not been available in the present cases.

It is interesting that the most seriously affected families, ZV, ZI, and ZL are the ones where haploinsufficiency as a pathogenic mechanism is mostly likely. However, only large studies comparing clinical phenotypes with mutations and polymorphisms, in large families, and corresponding studies of transcription and translation, can provide the certain knowledge we need to have in order to use the genetic workup clinically. Presently, as our study documents, mutations in MYBPC3 should be studied carefully before being assigned as disease causing.