Mutations in the βMHC (β-myosin heavy chain), a sarcomeric protein are responsible for hypertrophic and dilated cardiomyopathy. However, the mechanisms whereby distinct mutations in the βMHC gene cause two kinds of cardiomyopathy are still unclear. In the present study we report a novel βMHC mutation found in a patient with isolated LVNC [LV (left ventricular) non-compaction] and the phenotype of a mouse mutant model carrying the same mutation. To find the mutation responsible, we searched for genomic mutations in 99 unrelated probands with dilated cardiomyopathy and five probands with isolated LVNC, and identified a p.Met531Arg mutation in βMHC in a 13-year-old girl with isolated LVNC. Next, we generated six lines of transgenic mice carrying a p.Met532Arg mutant αMHC gene, which was identical with the p.Met531Arg mutation in the human βMHC. Among these, two lines with strong expression of the mutant αMHC gene were chosen for further studies. Although they did not exhibit the features characteristic of LVNC, approx. 50% and 70% of transgenic mice in each line displayed LVH (LV hypertrophy) by 2–3 months of age. Furthermore, LVD (LV dilation) developed in approx. 25% of transgenic mice by 18 months of age, demonstrating biphasic changes in LV wall thickness. The present study supports the idea that common mechanisms may be involved in LVH and LVD. The novel mouse model generated can provide important information for the understanding of the pathological processes and aetiology of cardiac dilation in humans.

INTRODUCTION

HCM (hypertrophic cardiomyopathy) and DCM (dilated cardiomyopathy) are very common types of cardiomyopathy. HCM is characterized by abnormal cardiac hypertrophy, fibrosis and myofibrillar disarray. DCM is defined by ventricular chamber dilation and impaired contractile function. Genetic studies have indicated that approx. half of HCM is familial and caused by a mutation in sarcomeric proteins [1]. Among the causative genes, βMHC (β-myosin heavy chain) is most commonly associated with HCM [2]. Disease penetrance, severe hypertrophy and high risk of sudden cardiac death are more frequently associated with mutations in βMHC than in the other sarcomere protein genes, such as cardiac troponin T, α-tropomyosin and cardiac myosin-binding protein C genes [3]. On the other hand, approx. 25–30% of idiopathic DCM is caused by a missense mutation or deletion in cardiac genes such as βMHC, cardiac troponin T, cardiac actin, lamin A/C and dystrophin [2,4]. In DCM, βMHC mutations are also relevant to early onset and serious cardiac dysfunction [5,6]. In addition, cases in which HCM has progressed to DCM have been reported [7,8]. This progression occurs in 10–15% of patients with HCM [9]. However, the mechanisms whereby mutations of the βMHC gene lead to cardiac hypertrophy or dilation remain unclear. Moreover, it is still not clear whether there is a common aetiology for these diseases.

Transgenic mouse models expressing mutant proteins provide a means of gaining insight into the pathophysiological and clinical features of human cardiomyopathy. For example, transgenic mice carrying the missense mutation p.Arg403Gln in the αMHC gene, the murine analogue of the human βMHC gene, recapitulate the characteristics of human HCM [5,10], whereas homozygous mice for the same transgene develop DCM-like disease [11]. The homozygous mutant transgenic mice of another sarcomeric protein, myosin-binding protein C, are also affected with DCM [12]. However, the mechanisms of the primary cardiac dilation caused by the βMHC mutation are still unclear. This is because animal models bearing the analogous mutation within the sarcomeric protein genes identified in human DCM have not so far been investigated.

In the present study we explored mutations in the sarcomere proteins in a patient with isolated LVNC [LV (left ventricular) non-compaction] and found a novel mutation, p.Met531Arg, in the βMHC. We then generated the αMHC transgenic mice with a p.Met532Arg mutation corresponding with the p.Met531Arg in human βMHC. Although these transgenic mice did not develop LVNC, they showed the pathological changes from HCM to DCM. The results of our present study suggest that HCM and DCM may be closely related pathological conditions rather than independent diseases.

MATERIALS AND METHODS

Patients

The study subjects comprised 99 unrelated patients with DCM (27 familial and 72 sporadic or unknown) and five patients with isolated LVNC (one familial and four sporadic or unknown). The diagnosis of DCM was based on the criteria of the Collaborative Research Group of the European Human and Capital Mobility Project on Familial Dilated Cardiomyopathy [13], i.e. echocardiographic demonstration of depressed systolic function of the left ventricle [LVEF (LV ejection fraction) <0.45 and/or fractional shortening <0.25) and a dilated left ventricle [LVEDD (LV end-diastolic diameter) >117% of the predicted value corrected for age and body surface area) in the absence of other cardiac or systemic causes. The diagnosis of isolated LVNC was based on the following echocardiographic criteria [14] in four patients: (i) the absence of coexisting cardiac abnormalities; (ii) the presence of a two-layer structure in the myocardium, with a compacted thin epicardial band and a much thicker noncompacted endocardial layer of trabecular meshwork with deep endomyocardial spaces showing a maximal end systolic ratio of noncompacted to compacted layers of >2; (iii) the predominant localization of the non-compaction in the mid-lateral, apical and mid-inferior walls; and (iv) colour Doppler evidence of deep perfused intertrabecular recesses. One patient (with the βMHC mutation) was diagnosed by postmortem examination because echocardiographic evidence of LVNC was lacking at that time.

Informed consent was obtained from all subjects in accordance with the guidelines of the Bioethical Committee on Medical Research, School of Medicine, Kanazawa University. gDNA (genomic DNA) was purified from white blood cells [15].

Detection of mutation

Oligonucleotide primers used for the amplification of the βMHC gene exons were based on published sequences [16] and sequences obtained from GenBank®. PCR was used for amplification of gDNA, and SSCP (single-strand conformational polymorphism) analysis of this amplified DNA was then performed with a slight modification of a method published previously [17,18]. DNA fragments with abnormal SSCP patterns were sequenced by the dye terminator cycle sequencing method using an automated fluorescent sequencer (ABI Prism™ 310 genetic analyser; PE Biosystems). To increase the probability of detecting the presence of any sequence change, SSCP was carried out at two different temperatures for each exon, and the size of fragments for SSCP was kept at less than 300 bp. Sequence analysis results were validated by restriction enzyme digestion with Eco81I. To confirm the paternity of the subjects, five short tandem-repeat systems TH01, vWA, LPL, F13B and FES/FPS were investigated as previously described [19]. From the allele distributions of each short tandem-repeat locus, the probability of paternity was calculated based on the allelic frequencies in the Japanese population [20]. Screening for mutations in other genes, including dystrophin, myosin-binding protein-C, α-tropomyosin, cardiac troponin C, cardiac troponin T, cardiac troponin I, cardiac α-actin, lamin A/C, G4.5, ZASP and α-dystrobrevin was performed by direct sequencing in the proband.

Transgenic constructs

Murine αMHC (the analogous gene of the human βMHC) cDNA (5.9 kbp) and the transgenic construct, αMHC clone 918 (9.1 kbp), were generously provided by Dr J. Robbins (University of Cincinnati, Cincinnati, OH, U.S.A.). The αMHC cDNA was mutated using site-directed mutagenesis according to the manufacturer's protocol (Stratagene), which resulted in a p.Met532Arg mutation in the protein. The mutagenic primers used were 5′-CCCATGGGCATCAGGTCCATCCTGGAGG-3′ and 5′-CCTCCAGGATGGACCTGATGCCCATGGG-3′. The mutated cDNA was sequenced to confirm the presence of the correct mutation and the absence of undesired errors during mutagenesis. The mutated αMHC cDNA was subcloned into the SalI site of αMHC clone 918 between the murine αMHC promoter and the human growth hormone polyadenylation site. The transgenic construct was purified by caesium chloride ultracentrifugation and digested with EcoRI to release a 12.1 kbp fragment that was used for microinjection. This fragment was purified by agarose gel electrophoresis, dissolved in 10 mmol/l Tris/HCl (pH 7.5) containing 0.2 mmol/l EDTA and injected into the pronucleus of fertilized zygotes from BDF1 mice. The microinjections were performed at Japan SLC Inc.

Generation of transgenic mice

Founder transgenic mice were identified by hybridization of tail DNA to a 32P-labelled DNA probe corresponding to the human growth hormone 3′-untranslated region (a 630 bp HindIII/EcoRI fragment from the transgenic construct). PCR was also used to identify the transgenic mice. A forward (5′-TGCCCACCAGCCTTGTCCTAATAA-3′) and a reverse (5′-CAGGGAAGGGAGCAGTGGTTCAC-3′) primer were derived from the human growth hormone sequence; PCR with these primers produced a 411 bp fragment using DNA of mice harbouring the transgene. Stable transgenic lines were generated by mating founder transgenic mice with nontransgenic BDF1 mice. Male transgenic mice and non-transgenic male littermates were used for analysis. Experiments were conducted according to guidelines for the care and use of laboratory animals in Kanazawa University and safety guidelines for gene manipulation experiments.

RT (reverse transcription)–PCR

RT–PCR was performed to assess the amount of αMHC, βMHC and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA in wild-type and transgenic hearts. Total RNA was isolated from the heart using the AGTC (acid guanidinium thiocyanate/phenol/chloroform) method [21] and the first strand cDNA was synthesized using standard cDNA synthesis reagents (first strand cDNA synthesis kit for RT–PCR; Roche) according to the manufacturer's protocol. To assess the αMHC transgene expression, PCR cycling was performed at 94 °C for 60 s, 59 °C for 60 s and 72 °C for 60 s for 30 cycles using rTaq (Takara). The forward (5′-GCCGCGCCAGTACTTCATAGGT-3′) and the reverse (5′-TTGCGAGGCTTCTGGAAGTTGTTA-3′) primers were derived from the murine αMHC cDNA sequence. When the PCR product (351 bp) was digested with XhoI, fragments of 248 bp and 103 bp were generated from the endogenous allele, while the 351 bp fragment from the transgene was not digested because the XhoI site was abolished by site-directed mutagenesis. To determine transcript levels of αMHC and βMHC genes, cDNA products were amplified by cycling at 94 °C for 30 s, 55 °C for 30 s and 72 °C for 30 s for 25 cycles using rTaq (Takara). Sequences of primers were as follows: αMHC, 5′-ATCGCCGAGTCCCAGGTCAAC-3′ and 5′-TATTGGCCACAGCGAGGGTCTG-3′; βMHC, 5′-GTGCCAAGGGCCTGAATGAGG-3′ and 5′-AGGGCTGTTGCAAAGGCTCCAG-3′; GAPDH, 5′-ACCACAGTCCATGCCATCAC-3′ and 5′-TCCACCACCCTGTTGCTGTA-3′.

Echocardiography

Echocardiographic studies were performed using a 12-MHz phased array probe and a Sonos 5500 ultrasonograph (Philips Medical Systems). Mice were anaesthetized lightly by intraperitoneal injection of 10 μg/ml of pentobarbital sodium at a dose of 10 μl/g of body mass. Body fur of the upper sternal and subxiphoid areas was shaved and the exposed skin was moistened for better acoustic coupling. M-mode echocardiographs of the left ventricle were recorded at the middle of the left ventricle. IVST (interventricular septal thickness), PWT (posterior wall thickness), LVESD (LV end-systolic diameter) and LVEDD were measured and the percentage FS (fractional shortening) was calculated as (LVEDD−LVESD)/LVEDD.

Histological examination

Mice were anaesthetized, and the hearts were removed while still beating, rinsed in PBS, and fixed in 10% formalin before sectioning. The hearts were dehydrated through a graded alcohol series and embedded in paraffin. Longitudinal 8 μm sections were cut and stained with H and E (Haematoxylin and Eosin or with Azan and examined under an Olympus IX71 microscope. Photomicrographs were obtained with an Olympus DP70 digital camera. For electron microscopic analysis, the hearts were removed while still beating, and immersed in a cardioplegic solution (25 mmol/l KCl and 5% dextrose in PBS) to ensure complete myocardial relaxation. Blocks of 1 mm2 were dissected from the left ventricular free wall. The blocks were trimmed, fixed in 2.5% glutaraldehyde in cacodylate buffer at pH 7.4, postfixed in 2.0% osmium tetroxide, dehydrated in ethanol in propylene oxide, and embedded in EPOK812 (Oken). Sections were cut at 60 nm, stained with lead citrate and uranyl acetate, and examined with a JEM-1210 transmission electron microscope (JEOL).

Statistical analysis

Statistically significant differences between groups of non-transgenic and transgenic mice were assessed using an unpaired Student's t test. Results are expressed as means±S.D. A P value of <0.05 was considered statistically significant.

RESULTS

Baseline characteristics of the study patients

The 99 adult patients with DCM comprised 68 men and 31 women (mean age 58.1±13.1 years, range 21–82). The three adult patients (aged 29, 57 and 60 years), and two young patients (aged 10 and 13 years) with isolated LVNC comprised two men and three women. In the patient groups with DCM and isolated LVNC, the LVEDD was 64.8±7.4 mm and 62.4±12.8 mm respectively, the LVESD was 55.8±8.2 mm and 54.2±10.3 mm respectively and the LVEF was 29.1±9.8 mm and 26.2±13.7 mm respectively.

A point mutation was found in the βMHC gene of an isolated LVNC patient

SSCP analysis identified polymorphisms in the βMHC gene in 17 patients with DCM (14 with c.189C>T, two with c.732T>C, four with c.1062C>T, one with c.1128C>T and three with c.3027T>C) which have been reported previously [22]. A mutation was found in the βMHC gene derived from the proband, a 14-year-old girl (Figure 1A, II-4) with isolated LVNC. Sequence analysis of the abnormal polymorphism conformer revealed a nucleotide substitution in codon 531, resulting in substitution of a methionine residue by arginine (Figure 1B). This mutation was not detected in 200 control individuals. No other mutations in the βMHC or other genes, including myosin-binding protein-C, α-tropomyosin, cardiac troponin T, cardiac troponin I, actin, lamin A/C, G4.5, ZASP or α-dystrobrevin were identified in this proband.

A point mutation was found in the βMHC gene of an isolated LVNC patient

Figure 1
A point mutation was found in the βMHC gene of an isolated LVNC patient

(A) Upper panel: pedigree I and II indicate generations. An asterisk indicates the proband. Open circles and open boxes indicate female and male normal individuals respectively. A closed circle indicates a phenotype-positive, genotype-positive individual. A hatched circle indicates a phenotype-positive, genotype unknown individual. The forward slash indicates deceased individuals. Lower panel: PCR-restriction fragment length polymorphism analysis. Digestion of the PCR products with Eco81I generates polymorphic restriction fragments of 201 bp (wild-type allele) and/or 185 bp (mutant allele). (B) DNA sequence analysis. A single nucleotide transition from thymine to guanine at nucleotide position 1678 of βMHC was identified. This mutation leads to a missense mutation of Met531 to an arginine residue. (C) Twelve-lead ECG of the proband (II-4). (D) A photograph of the heart of the proband (II-4) at autopsy. Features of the heart were consistent with isolated LVNC.

Figure 1
A point mutation was found in the βMHC gene of an isolated LVNC patient

(A) Upper panel: pedigree I and II indicate generations. An asterisk indicates the proband. Open circles and open boxes indicate female and male normal individuals respectively. A closed circle indicates a phenotype-positive, genotype-positive individual. A hatched circle indicates a phenotype-positive, genotype unknown individual. The forward slash indicates deceased individuals. Lower panel: PCR-restriction fragment length polymorphism analysis. Digestion of the PCR products with Eco81I generates polymorphic restriction fragments of 201 bp (wild-type allele) and/or 185 bp (mutant allele). (B) DNA sequence analysis. A single nucleotide transition from thymine to guanine at nucleotide position 1678 of βMHC was identified. This mutation leads to a missense mutation of Met531 to an arginine residue. (C) Twelve-lead ECG of the proband (II-4). (D) A photograph of the heart of the proband (II-4) at autopsy. Features of the heart were consistent with isolated LVNC.

Cardiac examination of the proband (II-4) revealed left ventricular dilation and diminished contractile function like DCM (Table 1), although she was asymptomatic. Heart failure progressed and she died at the age of 14 in 1999. Figures 1(C) and 1(D) show the ECG and the photograph of the heart at autopsy. The ECG showed left ventricular hypertrophy. In the autopsy, the left ventricle was markedly dilated and prominent numerous trabeculations with intertrabecular recesses were found at the lateral wall, the inferior wall and the apex of the left ventricle. The thickness ratio between the noncompacted and compacted layer was 3–5. Other congenital cardiac malformations were not found. These findings were consistent with isolated LVNC. On microscopic examination, mild vacuolation was evident in the myocytes, and moderate subendocardial fibrosis was observed (results not shown). The endocardium was thickened by fibrous tissue, but no fibroelastosis was identified. Myocyte hypertrophy was not found. She had an identical twin sister (II-3), who had been diagnosed with DCM and also died of heart failure at the age of 13 (Figure 1A).

Table 1
Echocardiographic data in the proband and her identical twin sister
ParameterII-3II-4
Sex Female Female 
Age (years) 13 13 
LVEDD (mm) 70 70 
LVEDS (mm) 62 58 
FS (%) 11 17 
IVST (mm) 10 
PWT (mm) 10 
ParameterII-3II-4
Sex Female Female 
Age (years) 13 13 
LVEDD (mm) 70 70 
LVEDS (mm) 62 58 
FS (%) 11 17 
IVST (mm) 10 
PWT (mm) 10 

We could not find any abnormalities in the parents (Figure 1A, I-1 and I-2), elder sister (II-1) and elder brother (II-2) by clinical examination, and none of them had the p.Met531Arg mutation in the βMHC gene. The presence of this sequence variant was confirmed with restriction enzyme digest analysis. The 201 bp fragment was digested with Eco81I. The T to G transition at nucleotide position 1674 allows cleavage (yielding 16 bp and 185 bp fragments), whereas the wild-type allele is not cut. Only the proband was heterozygous for the T to G base change (Figure 1A). The allele distribution of five short tandem-repeat loci was then examined in the subject's parents to determine the paternity of the proband. The probability of paternity was 0.9873, which was considered to be highly likely. Thus we concluded that the identified p.Met531Arg mutation in the βMHC gene in the proband was de novo.

Generation of p.Met532Arg αMHC transgenic mice

To elucidate the importance of p.Met531Arg of βMHC observed in the human patient, we constructed a transgenic vector based on the mouse αMHC (the analogous gene of the human βMHC) clone 918 (Figure 2A). Nucleotides 1604 and 1614 of the coding region were mutated using site-directed mutagenesis. c.1604T>G resulted in a p.Met532Arg mutation, and c.1614C>G resulted in the deletion of the XhoI site without amino acid change and enabled us to distinguish the mutant cDNA from wild-type cDNA. PCR of gDNA revealed that we could obtain 17 transgenic lines (Figure 2B), and six independent lines expressing transgene-derived αMHC mRNA (Figure 2C). Densitometric analysis of PCR products revealed that each transgenic line had unique ratios of transgene (a 351 bp band) to endogenous (248 bp and 103 bp bands) αMHC, which could be used to distinguish each transgenic line. Among these transgenic lines, we chose line numbers 41 and 23 because they showed severe phenotypes and expressed more αMHC mRNA than the other transgenic lines.

Transgenic construct and transcript expression

Figure 2
Transgenic construct and transcript expression

(A) A construct used to generate transgenic mice. Arrow heads indicate PCR primers for detection of αMHC cDNA. The asterisk indicates the position of the mutation (c.1604T>G). UTR, untranslated region; hGH pA, human growth hormone polyadenylation signal. (B) Identification of transgenic mice by PCR analysis of gDNA. A 411 bp PCR product was amplified using transgene specific primers for hGH pA. The 411 bp products are present only in the trangenic mice. (C) Analyses of transgene RNA expression by RT–PCR using primers in (A) in the hearts of six transgenic lines. When the PCR product (351 bp) was digested with XhoI, 248 bp and 103 bp fragments were generated from the endogenous allele, while the 351 bp fragment from the transgene was not digested, because the XhoI site was abolished by site-directed mutagenesis. Densitometric analysis of PCR products revealed that six transgenic lines (the number of the transgenic line is shown on each lane) had unique ratios of transgene to endogenous αMHC mRNA.

Figure 2
Transgenic construct and transcript expression

(A) A construct used to generate transgenic mice. Arrow heads indicate PCR primers for detection of αMHC cDNA. The asterisk indicates the position of the mutation (c.1604T>G). UTR, untranslated region; hGH pA, human growth hormone polyadenylation signal. (B) Identification of transgenic mice by PCR analysis of gDNA. A 411 bp PCR product was amplified using transgene specific primers for hGH pA. The 411 bp products are present only in the trangenic mice. (C) Analyses of transgene RNA expression by RT–PCR using primers in (A) in the hearts of six transgenic lines. When the PCR product (351 bp) was digested with XhoI, 248 bp and 103 bp fragments were generated from the endogenous allele, while the 351 bp fragment from the transgene was not digested, because the XhoI site was abolished by site-directed mutagenesis. Densitometric analysis of PCR products revealed that six transgenic lines (the number of the transgenic line is shown on each lane) had unique ratios of transgene to endogenous αMHC mRNA.

The non-transgenic mice and transgenic mice seemed to grow normally. However, by 12 months of age 19% (n=33) and 20% (n=20) of mice died in transgenic mice lines 23 and 41 respectively, compared with only one death in 30 (3.3%) non-transgenic mice (Figure 3). Transgenic mice died sporadically without showing prominent organ diseases except cardiac hypertrophy. The most likely cause of death was sudden cardiac arrest.

Survival curves of transgenic and non-transgenic mice

Figure 3
Survival curves of transgenic and non-transgenic mice

Non-transgenic mice (WT), n=30; transgenic mice line 23, n=33; transgenic mice line 41, n=20.

Figure 3
Survival curves of transgenic and non-transgenic mice

Non-transgenic mice (WT), n=30; transgenic mice line 23, n=33; transgenic mice line 41, n=20.

Echocardiography

Echocardiography was performed in seven transgenic mice of line 41 and six non-transgenic littermates. All mice were 8–10 months old and the mean age was not significantly different between the groups. Representative echocardiograms are shown in Figure 4. The difference in mean values of LVEDD, LVESD and FS were not statistically significant between non-transgenic and transgenic mice of line 41 (Table 2). However, IVST and PWT were significantly greater in transgenic mice of line 41 compared with non-transgenic mice.

Echocardiographic analysis

Figure 4
Echocardiographic analysis

Representative M-mode left ventricle images in 8- to 10 month old non-transgenic mice and transgenic mice line 41.

Figure 4
Echocardiographic analysis

Representative M-mode left ventricle images in 8- to 10 month old non-transgenic mice and transgenic mice line 41.

Table 2
Echocardiographic data in 8–10 month old non-transgenic mice and transgenic mice of line 41

Values are means±S.D. HR, heart rate. *P<0.05 when compared with non-trangenic mice.

ParameterNon-transgenicTransgenic line 41
n 
HR (beats/min) 590±18 600±13 
LVEDD (mm) 3.57±0.28 3.38±0.30 
LVESD (mm) 1.97±0.18 1.84±0.25 
FS (%) 44.7±2.0 45.7±3.4 
IVST (mm) 1.02±0.08 1.16±0.13* 
PWT (mm) 1.01±0.09 1.14±0.11* 
ParameterNon-transgenicTransgenic line 41
n 
HR (beats/min) 590±18 600±13 
LVEDD (mm) 3.57±0.28 3.38±0.30 
LVESD (mm) 1.97±0.18 1.84±0.25 
FS (%) 44.7±2.0 45.7±3.4 
IVST (mm) 1.02±0.08 1.16±0.13* 
PWT (mm) 1.01±0.09 1.14±0.11* 

Myocardial histopathology and morphology showed that transgenic mice developed HCM and DCM

There were no significant differences between non-transgenic mice and transgenic mice at 1 month of age. Transgenic mice began to display left ventricular hypertrophy at 2–3 months of age, and showed left ventricular hypertrophy at about 12 months (Figure 5D). Striking histological and morphological abnormalities were observed in approx. 50% and 70% of transgenic mice of line 23 and line 41 respectively. When mice were approx. 18 months old, transgenic mice displayed enlarged atria and approx. 25% of transgenic mice progressed to exhibit left ventricular dilation compared with non-transgenic mice (Figures 5A–5C and 5E). However, transgenic mice did not show typical findings consistent with LVNC. Heart-to-body weight ratios at 12–15 months of age were significantly higher in transgenic mice compared with non-transgenic mice (Figure 6).

Histological analysis

Figure 5
Histological analysis

Hearts were obtained from a non-transgenic mouse (A) and transgenic mouse line 41 (B) at 18 months of age. (C–E) Coronal sections of hearts stained with H and E. (C) Non-transgenic mouse (18 months old). (D) Transgenic mouse line 41 with left ventricular hypertrophy (12 months old). (E) Transgenic mouse line 41 with left ventricular dilation (18 months old). Scale bar represents 1 mm. Higher magnification views of H and E-stained left ventricle sections from a 15 month old non-transgenic (F) and transgenic mouse line 41 (G). Azan-stained sections of ventricles from a non-transgenic (H) and transgenic mouse line 41 (I). Scale bar represents 50 μm.

Figure 5
Histological analysis

Hearts were obtained from a non-transgenic mouse (A) and transgenic mouse line 41 (B) at 18 months of age. (C–E) Coronal sections of hearts stained with H and E. (C) Non-transgenic mouse (18 months old). (D) Transgenic mouse line 41 with left ventricular hypertrophy (12 months old). (E) Transgenic mouse line 41 with left ventricular dilation (18 months old). Scale bar represents 1 mm. Higher magnification views of H and E-stained left ventricle sections from a 15 month old non-transgenic (F) and transgenic mouse line 41 (G). Azan-stained sections of ventricles from a non-transgenic (H) and transgenic mouse line 41 (I). Scale bar represents 50 μm.

Ratios of heart-to-body weight for non-transgenic and transgenic mice of line 23 and line 41 at 12–15 months old

Figure 6
Ratios of heart-to-body weight for non-transgenic and transgenic mice of line 23 and line 41 at 12–15 months old

Bars represent means±S.D. (n in parentheses). *P<0.05 compared with non-transgenic mice.

Figure 6
Ratios of heart-to-body weight for non-transgenic and transgenic mice of line 23 and line 41 at 12–15 months old

Bars represent means±S.D. (n in parentheses). *P<0.05 compared with non-transgenic mice.

Histological examination of transgenic hearts revealed mild relative myocyte hypertrophy and myofibrillar disarray starting at 2–3 months of age. These features became more severe with aging. Multiple pleiotropic nuclei were also observed. These histological features were scattered throughout the left ventricular free wall. H and E- and Azan-stained sections of hearts from 15 month old mice showed interstitial fibrosis especially at the endocardium. Non-transgenic mice showed regular arrangement of myofibres and no fibrotic lesions (Figures 5F–5I).

Transmission electron microscopy was performed to examine the ultrastructure of transgenic and non-transgenic mice hearts at 16 months of age. The non-transgenic mice showed normal sarcomeric structure, with regularly aligned Z-bands (Figure 7A). In contrast, transgenic mice showed an abnormal sarcomeric structure. The sarcomere lengths were greatly reduced and the myofilaments were misaligned. The M-lines were indistinct and the Z-bands were thicker than those of non-transgenic mice (Figure 7B).

Electron micrographs of longitudinal sections of left ventricular cardiac myocyte cells from a non-transgenic mouse (A) and a transgenic mouse of line 23 (B) at 16 months

Figure 7
Electron micrographs of longitudinal sections of left ventricular cardiac myocyte cells from a non-transgenic mouse (A) and a transgenic mouse of line 23 (B) at 16 months

Scale bars represent 500 nm.

Figure 7
Electron micrographs of longitudinal sections of left ventricular cardiac myocyte cells from a non-transgenic mouse (A) and a transgenic mouse of line 23 (B) at 16 months

Scale bars represent 500 nm.

The expression level of βMHC (corresponding to αMHC in human), which is associated with a cardiac stress response, was compared in the hearts of transgenic mice of line 23 and transgenic mice of line 41 with non-transgenic mice at 4 months of age. Transgenic mice demonstrated significant increases in βMHC compared with age-matched non-transgenic mice (5.6-fold and 4.6-fold in lines 23 and 41 respectively). The amount of αMHC transcripts in transgenic hearts was the same as in age-matched non-transgenic hearts (Figure 8).

Semi-quantitative RT–PCR analysis of gene expression in hearts of 4-month-old mice

Figure 8
Semi-quantitative RT–PCR analysis of gene expression in hearts of 4-month-old mice

Expression levels of αMHC and βMHC were examined. The expression level of βMHC was increased 5.6-fold and 4.6-fold in transgenic mice line 23 and line 41 respectively. GAPDH was used as an internal control.

Figure 8
Semi-quantitative RT–PCR analysis of gene expression in hearts of 4-month-old mice

Expression levels of αMHC and βMHC were examined. The expression level of βMHC was increased 5.6-fold and 4.6-fold in transgenic mice line 23 and line 41 respectively. GAPDH was used as an internal control.

DISCUSSION

Generation of a novel transgenic mouse model having a point mutation in the αMHC gene that exhibits HCM developing to DCM

In the present study we have generated the first mouse model with a point mutation in the αMHC gene exhibiting HCM that developed to DCM, and it showed a more malignant phenotype compared with other αMHC mutant mice. At first, we identified a novel de novo mutation in the βMHC gene in a patient with isolated LVNC, and generated αMHC transgenic mice with a p.Met532Arg missense mutation that is an analogous mutation of the patient. Although these mice did not develop LVNC, approx. 50–70% of them demonstrated the pathological and clinical features of human HCM after they were 2–3 months old. Moreover, 25% of transgenic mice progressed to exhibit DCM-like dilated phase HCM by 18 months of age. The phenotype of these p.Met532Arg αMHC transgenic mice is similar in part to that of previously constructed αMHC transgenic mice models of HCM. It is not appropriate to compare the consequences of mutations by the severity of cardiac impairment of each mutant mouse because the amount of mutant cDNA expression and strains of mice are different. Nevertheless, our results suggest that the p.Met531Arg mutation may have more malignant consequences for cardiac function than other mutations in αMHC. For example, the mutation, p.Arg403Gln, in human βMHC caused severe HCM associated with early disease onset and short life expectancy in humans [23]. However, the p.Arg403Gln αMHC (analogous to human βMHC p.Arg403Gln) trangenic mice displayed mild cardiac dysfunction and hypertrophy with normal survival [24]. Compared with this p.Arg403Gln αMHC transgenic model, our p.Met531Arg αMHC transgenic mice progressed to severe left ventricular hypertrophy and further to dilated cardiomyopathy with sudden death. These results indicate that cardiac impairment of p.Met532Arg αMHC transgenic mice is significantly increased. In another model, αMHC transgenic mice bearing both the p.Arg403Gln mutation and a deletion in a part of the actin-binding site of αMHC evolved from left ventricular hypertrophy to dilation at 10 months of age even though only a small amount of transgenic protein was expressed (10% of endogenous αMHC protein) [10]. However, this model does not represent the real clinical consequence caused by p.Arg403Gln mutation in humans because the compound heterozygote of this mutation has not been identified to date in individuals with HCM. These results indicate that the phenotype of cardiomyopathy in mutant αMHC transgenic mice may be milder than that in humans.

Approx. 20% of our transgenic mice died by 12 months of age. The cause of death remains unknown because we could not evaluate the transgenic mice electrophysiologically due to technical difficulties. However, a previous study has reported that the degree of ventricular hypertrophy was significantly associated with increased arrhythmia susceptibility in p.Arg403Gln αMHC transgenic mice [25]. Moreover, dead transgenic mice did not show the findings of heart failure or prominent diseases of other organs, except cardiac hypertrophy. These results suggest that these transgenic mice died of malignant arrhythmia. Further evaluation is necessary to clarify the electrophysiological abnormality in p.Met532Arg αMHC transgenic mice. In vitro functional studies may also help to elucidate the pathophysiological mechanisms affected by the p.Met531Arg mutation.

In the present study, p.Met532Arg αMHC trangenic mice did not show the phenotype of LVNC, unlike that found in the human patient. This result demonstrates that there is little impact of αMHC mutation on left ventricular morphogenesis. One explanation is that the expression pattern of αMHC in the murine heart is different from that of βMHC in the human heart. Human heart atria express αMHC and the ventricles express predominantly βMHC. In mouse, αMHC is expressed in both the postnatal atria and ventricles, whereas expression of βMHC in the embryonic ventricle is predominant over αMHC, especially when the process of ventricular myocardium compaction progresses. In mice, at 10.5 days post fertilization, αMHC gene expression begins to decrease in ventricular myocytes and trabeculations begin to form in the ventricles where βMHC is expressed dominantly [26]. Thus we speculate that the effect of βMHC mutation on left ventricle morphogenesis in humans may be much bigger than that of αMHC mutation in mice.

It is of note that the patient's monozygotic twin sister developed DCM. If the Met531 of βMHC was mutated to an arginine residue at the 1-cell stage, the twin sister had the same mutation. Although we could not examine this twin sister genetically because she died before the present study, it is possible that the p.Met531Arg mutation in βMHC caused DCM in her heart. Hence the transgenic model might reflect the pathology of human DCM rather than LVNC.

The mechanisms of onset of HCM and DCM

Met531 of βMHC is located in the actin-binding site. Replacement of a methionine residue by an arginine residue may impair the α-helix structure and disrupt the interaction between myosin and actin because methionine is a hydrophobic amino acid, whereas arginine is a basic and hydrophilic amino acid. Interestingly, other human DCM-causing mutations of βMHC are located near this region, such as p.Ser532Pro and p.Ala550Val [6,27]. For example, the p.Ser532Pro mutation which changes the charge of the amino acid at this position caused severe DCM. These results suggest that the p.Met531Arg mutation may cause catastrophic cardiomyopathy by a similar mechanism. Mutations in the genes encoding sarcomere proteins may alter contraction of cardiac muscle cells and activate multiple cellular pathways. When sarcomere proteins cannot interact sufficiently with other proteins because of the presence of mutations, cardiac remodelling may develop in order to compensate for the dysfunction, resulting in cardiomyopathy. It remains unclear why mutations of proteins with similar functions can cause two different morphologies, HCM and DCM, and whether these diseases are part of the same progressive pathology.

The results of the present study support the hypothesis that HCM and DCM reflect stages of a single progression pathway of heart disease [2]. Several studies of other mutant mice also support this hypothesis. For example, heterozygous mutant mice for the R403Q mutation developed HCM [25], whereas homozygous mutant mice developed DCM [11,12]. Furthermore, the R403Q mutation combined with a deletion in a part of the actin-binding site caused progression from HCM to DCM [26]. The fact that myohypertrophy is seen in DCM, and that HCM deteriorates into a phase that resembles DCM in human patients also supports the idea of a single pathophysiological progressive pathway.

In conclusion, the p.Met532Arg αMHC transgenic mice demonstrated a severe HCM phenotype with sudden death although they did not recapitulate the LVNC phenotype. In addition, some of the mice progressed to left ventricular dilation. These results indicate that the βMHC p.Met531Arg mutation contributes to malignant cardiomyopathy. This model would help to understand the pathological processes and aetiology of cardiomyopathy caused by MHC mutations.

Abbreviations

     
  • DCM

    dilated cardiomyopathy

  •  
  • FS

    fractional shortening

  •  
  • GAPDH

    glyceraldehyde-3-phosphate dehydrogenase

  •  
  • gDNA

    genomic DNA

  •  
  • H and E

    Haemotoxylin and Eosin

  •  
  • HCM

    hypertrophic cardiomyopathy

  •  
  • IVST

    interventricular septal thickness

  •  
  • LV

    left ventricular

  •  
  • LVD

    LV dilation

  •  
  • LVEDD

    LV end-diastolic diameter

  •  
  • LVESD

    LV end-systolic diameter

  •  
  • LVEF

    LV ejection fraction

  •  
  • LVH

    LV hypertrophy

  •  
  • LVNC

    LV non-compaction

  •  
  • MHC

    myosin heavy chain

  •  
  • PWT

    posterior wall thickness

  •  
  • RT

    reverse transcription

  •  
  • SSCP

    single-strand conformational polymorphism

We thank T. Taniguchi and Y. Kubo (Kanazawa University Graduate School of Medicine, Kanazawa, Ishikawa, Japan) for excellent technical assistance.

References

References
1
Richard
 
P.
Charron
 
P.
Carrier
 
L.
, et al 
Hypertrophic cardiomyopathy: distribution of disease genes, spectrum of mutations, and implications for a molecular diagnosis strategy
Circulation
2003
, vol. 
107
 (pg. 
2227
-
2232
)
2
Seidman
 
J. G.
Seidman
 
C.
 
The genetic basis for cardiomyopathy: from mutation identification to mechanistic paradigms
Cell
2001
, vol. 
104
 (pg. 
557
-
567
)
3
Marian
 
A. J.
Roberts
 
R.
 
Molecular genetic basis of hypertrophic cardiomyopathy: genetic markers for sudden cardiac death
J. Cardiovasc. Electrophysiol.
1998
, vol. 
9
 (pg. 
88
-
99
)
4
Ahmad
 
F.
Seidman
 
J. G.
Seidman
 
C. E.
 
The genetic basis for cardiac remodeling
Annu. Rev. Genomics Hum. Genet.
2005
, vol. 
6
 (pg. 
185
-
216
)
5
Daehmlow
 
S.
Erdmann
 
J.
Knueppel
 
T.
, et al 
Novel mutations in sarcomeric protein genes in dilated cardiomyopathy
Biochem. Biophys. Res. Commun.
2002
, vol. 
298
 (pg. 
116
-
120
)
6
Kamisago
 
M.
Sharma
 
S. D.
DePalma
 
S. R.
, et al 
Mutations in sarcomere protein genes as a cause of dilated cardiomyopathy
N. Engl. J. Med.
2000
, vol. 
343
 (pg. 
1688
-
1696
)
7
Hecht
 
G. M.
Klues
 
H. G.
Roberts
 
W. C.
Maron
 
B. J.
 
Coexistence of sudden cardiac death and end-stage heart failure in familial hypertrophic cardiomyopathy
J. Am. Coll. Cardiol.
1993
, vol. 
22
 (pg. 
489
-
497
)
8
ten Cate
 
F. J.
Roelandt
 
J.
 
Progression to left ventricular dilatation in patients with hypertrophic obstructive cardiomyopathy
Am. Heart J.
1979
, vol. 
97
 (pg. 
762
-
765
)
9
Spirito
 
P.
Bellone
 
P.
 
Natural history of hypertrophic cardiomyopathy
Br. Heart J.
1994
, vol. 
72
 (pg. 
S10
-
S12
)
10
Vikstrom
 
K. L.
Factor
 
S. M.
Leinwand
 
L. A.
 
Mice expressing mutant myosin heavy chains are a model for familial hypertrophic cardiomyopathy
Mol. Med.
1996
, vol. 
2
 (pg. 
556
-
567
)
11
Fatkin
 
D.
Christe
 
M. E.
Aristizabal
 
O.
, et al 
Neonatal cardiomyopathy in mice homozygous for the Arg403Gln mutation in the α-cardiac myosin heavy chain gene
J. Clin. Invest.
1999
, vol. 
103
 (pg. 
147
-
153
)
12
McConnell
 
B. K.
Jones
 
K. A.
Fatkin
 
D.
, et al 
Dilated cardiomyopathy in homozygous myosin-binding protein-C mutant mice
J. Clin. Invest.
1999
, vol. 
104
 (pg. 
1235
-
1244
)
13
Mestroni
 
L.
Maisch
 
B.
McKenna
 
W. J.
, et al 
Guidelines for the study of familial dilated cardiomyopathies
Eur. Heart J.
1999
, vol. 
20
 (pg. 
93
-
102
)
14
Jenni
 
R.
Oechslin
 
E.
Schneider
 
J.
, et al 
Echocardiographic and pathoanatomical characteristics of isolated left ventricular non-compaction: a step towards classification as a distinct cardiomyopathy
Heart
2001
, vol. 
86
 (pg. 
666
-
671
)
15
Haliassos
 
A.
Chomel
 
J. C.
Tesson
 
L.
, et al 
Modification of enzymatically amplified DNA for the detection of point mutations
Nucleic Acids Res.
1989
, vol. 
17
 pg. 
3606
 
16
Jaenicke
 
T.
Diederich
 
K. W.
Haas
 
W.
, et al 
The complete sequence of the human β-myosin heavy chain gene and a comparative analysis of its product
Genomics
1990
, vol. 
8
 (pg. 
194
-
206
)
17
Orita
 
M.
Suzuki
 
Y.
Sekiya
 
T.
Hayashi
 
K.
 
Rapid and sensitive detection of point mutations and DNA polymorphisms using the polymerase chain reaction
Genomics
1989
, vol. 
5
 (pg. 
874
-
879
)
18
Mohabeer
 
A. J.
Hiti
 
A. L.
Martin
 
W. J.
 
Non-radioactive single strand conformation polymorphism (SSCP) using the Pharmacia ‘PhastSystem’
Nucleic Acids Res.
1991
, vol. 
19
 pg. 
3154
 
19
Lin
 
Z.
Ohshima
 
T.
Gao
 
S.
, et al 
Genetic variation and relationships at five STR loci in five distinct ethnic groups in China
Forensic Sci. Int.
2000
, vol. 
112
 (pg. 
179
-
189
)
20
Nagai
 
A.
Yamada
 
S.
Watanabe
 
Y.
Bunai
 
Y.
Ohya
 
I.
 
Analysis of the STR loci HUMF13A01, HUMFXIIIB, HUMLIPOL, HUMTH01, HUMTPOX and HUMVWFA31 in a Japanese population
Int. J. Legal Med.
1996
, vol. 
109
 (pg. 
34
-
36
)
21
Chomczynski
 
P.
Sacchi
 
N.
 
Single-step method of RNA isolation by acid guanidinium thiocyanatephenol-chloroform extraction
Anal. Biochem.
1987
, vol. 
162
 (pg. 
156
-
159
)
22
Masami
 
S.
Hidekazu
 
I.
Toshihiko
 
Y.
, et al 
Gene mutations in adult Japanese patients with dilated cardiomyopathy
Circ. J.
2005
, vol. 
69
 (pg. 
150
-
153
)
23
Watkins
 
H.
Rosenzweig
 
A.
Hwang
 
D. S.
, et al 
Characteristics and prognostic implications of myosin missense mutations in familial hypertrophic cardiomyopathy
N. Engl. J. Med.
1992
, vol. 
326
 (pg. 
1108
-
1114
)
24
McConnell
 
B. K.
Fatlin
 
D.
Semsarian
 
C.
, et al 
Comparison of two murine models of familial hypertrophic cardiomyopathy
Circ. Res.
2001
, vol. 
88
 (pg. 
383
-
389
)
25
Wolf
 
C. M.
Moskowitz
 
I. P.
Arno
 
S.
, et al 
Somatic events modify hypertrophic cardiomyopathy pathology and link hypertrophy to arrhythmia
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
18123
-
18128
)
26
Lyons
 
G. E.
Schiaffino
 
S.
Sassoon
 
D.
Barton
 
P.
Buckingham
 
M.
 
Developmental regulation of myosin gene expression in mouse cardiac muscle
J. Cell. Biol.
1990
, vol. 
111
 (pg. 
2427
-
2436
)
27
Villard
 
E.
Duboscq-Bidot
 
L.
Charron
 
P.
, et al 
Mutation screening in dilated cardiomyopathy: prominent role of the β-myosin heavy chain gene
Eur. Heart J.
2005
, vol. 
26
 (pg. 
794
-
803
)

Author notes

1

These authors contributed equally to this study.