Expression of a Mutation Causing Hypertrophic Cardiomyopathy Disrupts Sarcomere Assembly in Adult Feline Cardiac Myocytes
Abstract Mutations in the β-myosin heavy chain (βMyHC) induce hypertrophic cardiomyopathy (HCM), cardiac hypertrophy, and sarcomere disarray, with the latter being the characteristic hallmark. Thus, we sought to determine whether expression of mutant βMyHC in adult feline cardiac myocytes, a species known to develop HCM with a phenotype identical to that in humans, induces sarcomere disarray. A full-length βMyHC cDNA was cloned from a human heart cDNA library, and an HCM-causing mutation (Arg403Gln) was induced in the βMyHC cDNA by site-directed mutagenesis using polymerase chain reaction (PCR). The normal and mutant βMyHC cDNAs were cloned into pΔE1spIB shuttle vector, downstream from a cytomegalovirus (CMV) promoter. Replication-deficient recombinant adenoviral constructs (Ad5/CMV/βMyHC-N and Ad5/CMV/βMyHC-403) were generated through homologous recombination of pΔE1spIB/CMV/βMyHC-N or Ad5/CMV/βMyHC-403 and pBHG10 after cotransfection in 293 host cells. Infection of COS-1 cells with the βMyHC construct resulted in the expression of a full-length myosin protein. Efficiency of infection of isolated adult cardiac myocytes was >95%. Expression of the βMyHC constructs into mRNA at 48 hours after infection of feline cardiac myocytes was confirmed by reverse transcription–PCR. The net total protein and β-myosin synthesis were determined by using the amount of incorporation of [3H]phenylalanine into total protein and β-myosin, respectively. Although the total amount of protein synthesis was equal among experimental groups, the net myosin synthesis at 48 hours was greater in cardiac myocytes infected with normal or mutant βMyHC constructs than control myocytes or those infected with vector alone (P<.05). Electron microscopic examination showed only minor changes in the structure of sarcomeres in all experimental groups at 48 hours after infection. However, disruption of the sarcomeric structures at 120 hours after infection with the mutant βMyHC construct was observed in ≈50% of the myocytes examined, whereas the structure of the sarcomeres remained largely intact in myocytes infected with normal βMyHC construct, adenoviral vector alone, or control cardiocytes. Similar results were confirmed by immunofluorescence using MF-20 antibody to myosin. The results of this study indicate that disruption of sarcomere assembly and myofibrillar organization due to mutant βMyHC protein is the primary defect in HCM.
- hypertrophic cardiomyopathy
- β-myosin heavy chain mutation
- sarcomere assembly
- cardiac myocytes
Hypertrophic cardiomyopathy (HCM), an inherited autosomal-dominant disease, is genetically heterogeneous and clinically characterized by sudden cardiac death and heart failure.1 The predominant cardiac pathology is myocyte hypertrophy and sarcomere disarray; the former is found in most cardiac diseases, whereas the latter is a hallmark of HCM.2 The β-myosin heavy chain (βMyHC) gene, located on chromosome 14, is the most common gene responsible for HCM, in which more than 35 missense mutations have been identified.3 Three different approaches have been taken in the structure-function analysis of the βMyHC mutations. Cuda et al4 isolated the βMyHC from the skeletal muscle of patients with HCM and showed, in an in vitro motility assay, that the rate of translocation of actin-coated filaments over mutant myosin was significantly decreased. Sweeney et al5 expressed mutant rat cardiac α-myosin heavy chain (αMyHC) meromyosin constructs containing corresponding βMyHC mutations in Sf9 cells and showed actin-activated ATPase activity and decreased actin-myosin interaction in an in vitro motility assay. Straceski et al6 expressed normal and mutant αMyHC in COS cells and showed that mutant myosin failed to form filamentous structures in ≈30% of the COS cells transfected, whereas only 2% of cells transfected with normal myosin constructs failed to form such structures.
In the present study, we specifically examined the effect of mutations in βMyHC on cardiac sarcomere assembly in a species more closely reflective of humans. Normal and mutant human βMyHC cDNA was incorporated into recombinant replication-deficient adenoviral constructs and expressed in adult feline cardiac myocytes. Replication-deficient adenoviruses provide a highly efficient method of gene transfer into a variety of cells, including adult cardiac myocytes.7 8 Adult feline cardiac myocytes offer several advantages: (1) βMyHC is the adult cardiac myosin form as in humans.9 This is in contrast to smaller rodents, such as mice and rats, in which αMyHC is the predominant myosin.9 (2) They form sarcomeres as their functioning contractile unit, which remain aligned for a prolonged period of time in culture (at least 2 weeks).10 (3) HCM is the most common cardiac disease in cats, with a phenotypic expression identical to that observed in humans.11
Materials and Methods
Cloning of Full-Length βMyHC cDNA
The full-length human βMyHC cDNA (6 kb) was cloned as a single fragment for the first time from a normal human cardiac cDNA library. The library was screened with probes constructed to hybridize to the 3′ and 5′ ends of the βMyHC cDNA. Probes were radiolabeled to a specific activity of >109 cpm/μg with [32P]dCTP (Amersham) by the random-primer procedure of Feinberg and Vogelstein.12 Two overlapping fragments of the βMyHC cDNA (one was 5.2 kb and the second was 2 kb) were excised from positive plaques, digested with Nsi I restriction enzyme at position 1019 in an overlapping region, and subsequently ligated to generate the full-length βMyHC cDNA. The full-length sequence of 6 kb was determined in both directions and found identical to that published previously.13 14
A mutant βMyHC cDNA was constructed to contain the mutation Arg403Gln, which causes HCM in humans and is associated with a high incidence of sudden cardiac death.15 16 17 The Arg403Gln mutation, due to substitution of adenine for guanine in exon 13 at coding position 1208 of the βMyHC cDNA, was introduced in the βMyHC cDNA by polymerase chain reaction (PCR)–based oligonucleotide-mediated site-directed mutagenesis.18 The incorporation of the G→A mutation at coding position 1208 was confirmed by cycle sequencing.19
Generation of Recombinant Replication-Deficient Adenoviral Vectors
To accommodate our βMyHC cDNA and promoter (7 kb), it was necessary to construct a new adenoviral vector with greater packaging capacity that combined extensive deletions in both early region 1 (E1) and early region 3 (E3). The system used for rescue of the βMyHC cDNA into the viral genome is described in detail elsewhere.20 In brief, a chimeric plasmid vector that contains the left end of the adenoviral genome up to 15.8 map units was used in which polycloning sites replace the E1 region of the adenoviral genome from map unit 1.0 to 9.8 (pΔE1spIB). The cloned normal and mutated βMyHC cDNAs were excised from pGEM4Z vector and inserted into the HindIII and Xba I sites at the polycloning region such that the 5′ end of the βMyHC cDNA was located downstream from a cytomegalovirus (CMV) promoter. The resulting chimeric constructs (pΔE1spIB/CMV/βMyHC-N and pΔE1spIB/CMV/βMyHC-403) were cotransfected along with plasmid pBHG10 (a construct that carries the adenoviral genome with E3 deletion) into 293 cells.20 Recombination of the homologous DNA sequences in pBHG10 and pΔE1spIB/CMV/βMyHC-N or pΔE1spIB/CMV/βMyHC-403 after cotransfection of human 293 cells resulted in the production of a recombinant replication-deficient virus that carries the βMyHC expression cassette in place of the original E1 region (Ad5/CMV/βMyHC-N and Ad5/CMV/βMyHC-403) as shown in Fig 1⇓. The recombinant viruses were propagated, titrated, and purified in 293 cells according to the protocol published by Graham and Prevec21 and subsequently used to infect the adult feline cardiac myocytes.
Expression of Recombinant Adenoviral Constructs Into βMyHC Protein
COS-1 cells were grown on a 150-mm plate to a confluence of 70% in the presence of 10% fetal bovine serum (FBS) in DMEM. To demonstrate expression of recombinant adenoviral constructs into protein (myosin), COS-1 cells were infected with Ad5ΔE1, Ad5/CMV/βMyHC-N, and Ad5/CMV/βMyHC-403 constructs at a multiplicity of infection (MOI) of 100:1 for 4 hours, after which viruses were removed by washing the cells in PBS. The infected COS-1 cells were cultured for an additional 48 hours. The expressed myosin was extracted using low-salt and high-salt buffers as described by Bader et al.22 In brief, cells (20 plates for each construct) as well as control COS-1 cells were washed twice with PBS and were scraped for isolation of myosin. COS-1 cells were lysed in 7 mL ice-cold low-salt buffer containing 150 mmol/L NaCl, 10 mmol/L NaHPO4, 1% Triton X-100, and 0.1 mmol/L phenylmethylsulfonyl fluoride (PMSF), pH 7.5 for 5 minutes. Cellular cytoskeleton was disrupted by using a Dounce homogenizer. The resulting supernatant was centrifuged at 16 000g for 10 minutes. The pellet was extracted for 15 minutes in high-salt buffer containing 0.6 mol/L NaCl, 10 mmol/L NaHPO4, 5 mmol/L MgCl2, 5 mmol/L ATP, and 0.1 mmol/L PMSF, pH 7.2. Extraction was terminated by centrifugation at 16 000g for 10 minutes. The supernatant was subjected to another cycle of low-salt and high-salt extraction to further enrich the myosin component. The concentration of the protein containing myosin was determined by spectrophotometry using Micro BCA protein assay kit (Pierce).
Approximately 30 μg of myosin-enriched protein extract was loaded into each lane on a 7.5% nondenaturing polyacrylamide gel and was subjected to electrophoresis for 6 hours. The separated proteins were transferred (in 25 mmol/L Tris, pH 8.3, and 192 mmol/L glycine with 20% methanol) to a polyvinylidene difluoride membrane (Bio-Rad) by using a Trans-Blot electrophoretic transfer cell (Bio-Rad). The membrane was washed twice in PBS for 5 minutes and was incubated in blocking buffer (0.1% Tween-20 and 1% nonfat dry milk in 1× PBS) at room temperature for 4 hours. The membrane was incubated with 1:100 dilution (in blocking buffer) of mouse monoclonal IgG2b-κ antibody against adult chicken pectoralis myosin (MF-20) anti-myosin antibody (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City) at room temperature for 60 minutes.22 After it was washed, the membrane was exposed to 1:5000 dilution of goat anti-mouse alkaline phosphatase conjugate for 60 minutes by using a chemiluminescent detection system per recommendation of the manufacturer (Western Exposure Chemiluminescent Detection System, Clontech).
Efficiency of Infection
To determine the efficiency of infection of adult feline cardiac myocytes with recombinant adenovirus, 104 isolated cardiac myocytes were cultured on 35-mm plates and infected with recombinant adenoviruses carrying the Lac-Z reporter gene (Ad5/CMV/Lac-Z) at MOI values of 1:1, 10:1, 100:1, 500:1, and 1000:1 for 4 hours. Cardiac myocytes were cultured for an additional 48 hours and then fixed with 0.5% glutaraldehyde in PBS (pH 7.2) solution for 10 minutes at room temperature. Cardiac myocytes were rinsed twice with PBS and stained for β-galactosidase in the buffer solution of X-gal chromogen containing 5 mmol/L each of K3Fe(CN)6 and K4Fe(CN)6, 2 mmol/L MgCl2, and 1 mg/mL 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside overnight at room temperature in the dark.
Infection of Isolated Adult Feline Cardiac Myocytes
Adult feline cardiac myocytes were isolated as we have described previously.23 24 After 24 hours in culture, the isolated adult cardiac myocytes were incubated with ≈106 plaque-forming units (pfu) of control Ad5ΔE1, Ad5/CMV/βMyHC-N, and Ad5/CMV/βMyHC-403 for 4 hours; after which, the cardiac myocytes were washed to remove unattached viruses and were cultured in medium 199 containing 0.1% human serum albumin in a CO2 incubator at 37°C for an additional 48 or 120 hours. Experiments were performed with control feline cardiac myocytes (no transduction), feline cardiac myocytes infected with Ad5/ΔE1 as a control virus, and feline cardiac myocytes infected with the Ad5/CMV/βMyHC-N and Ad5/CMV/βMyHC-403 recombinant viruses.
To determine whether the βMyHC cDNA was transcribed into mRNA, the cultured adult cardiac myocytes were collected 48 hours after infection, and total RNA was isolated by the guanidinium isothiocyanate method.25 Reverse transcription–PCR was performed by using a set of primers to amplify a 320-bp fragment unique to the human βMyHC cDNA according to a previously published protocol.26
Net Protein Synthesis
The amount of newly synthesized total protein 48 hours after infection with 106 pfu of adenoviruses was determined from the amount of incorporated radiolabeled [3H]phenylalanine. A 2-mL suspension of freshly isolated cardiac myocytes was plated at a final concentration of 1×104 cells per milliliter onto laminin-coated (20 g/mL) polystyrene Petri dishes. The control or recombinant virus (106 pfu) was added to each corresponding experimental group for 4 hours. The unattached virus was then removed by washing the cultured cardiac myocytes three times with PBS. Cardiac myocytes were cultured in medium 199 containing 0.1% human serum albumin. Medium changes were performed on the first and second days of culture. Beginning on the second day in culture, 0.4 mmol/L of unlabeled l-phenylalanine was added to the culture medium to ensure equalized specific activities of the intracellular and extracellular phenylalanine pools. After allowing the cells to equilibrate for 1 hour in 0.4 mmol/L l-phenylalanine, the cells were pulse-labeled for 6 hours with 30 Ci/mL [3H]phenylalanine. In preliminary control experiments, we established that incorporation of [3H]phenylalanine was linear for control as well as experimental groups (r=.99 for all groups; P<.001), suggesting that protein degradation was negligible during the 6-hour labeling period. At the conclusion of the study (48 hours), the incorporation of radiolabeled phenylalanine was stopped by washing the cultures three times with cold (4°C) Hanks’ balanced salt solution containing 10 mmol/L l-phenylalanine. Cardiac myocyte proteins were then solubilized with a buffered SDS sample buffer (4% SDS, 2% glycerol, and 0.125 mol/L Tris-HCl [pH 6.8]). A portion of the solubilized sample was taken for analysis of protein content by using a commercially available assay (BCA, Pierce) with bovine serum albumin (BSA) used as a standard. The extent of radiolabeling was determined after acid precipitation for 30 minutes with cold (4°C) 10% trichloroacetic acid (TCA); the precipitates were then collected on 1.6-mm glass filters, and the filters were washed sequentially with 10% TCA, 5% TCA, and 95% ethanol. The glass filters were air-dried, and liquid scintillation counting was performed. Protein synthesis was determined by using the specific activities of medium samples obtained from direct scintillation counting, as well as the concentration of phenylalanine (0.4 mmol/L) in the medium, according to the following formula: phenylalanine incorporation (nmol · g protein−1 · h−1)=[phenylalanine incorporated into total cell protein (dpm/g myocyte protein)/phenylalanine specific activity of medium (dpm/mmol)]×h−1.
Net Myosin Synthesis
For measurement of specific synthesis of βMyHC protein, cardiac myocyte cultures were labeled and prepared exactly as described above for net protein synthesis. After determining the protein content for each sample, 2-mercaptoethanol (1%) was added, the samples were heated for 3 minutes at 95°C, and the samples were layered in equal protein amounts (20 mg) onto the same gel slab. To facilitate localization of βMyHC, purified myosin (5-mg sample) and known molecular weight standards were prepared as described above and were electrophoretically separated along with the experimental samples. Electrophoresis was performed by using 7.5% SDS-PAGE. The resultant gel slab was copper-stained for 10 minutes (Bio-Rad), the stained bands of myocyte protein and comigrating standard myosin heavy chain were aligned, and the βMyHC band was excised from the gel. Destaining of the excised gel pieces was performed by using a Tris-glycine buffer (Bio-Rad), and the βMyHC protein samples were electroeluted from the gel in a running buffer consisting of 0.1% SDS, 25 mmol/L Tris, and 190 mmol/L glycine. The eluted protein samples were then precipitated on ice for 30 minutes by using 1.6 mmol/L sodium deoxycholate and 15% TCA and then centrifuged at 9000g for 20 minutes. The resulting pellet was dissolved in a buffer consisting of 0.4% SDS, 12.5 mmol/L Tris-HCl, and 15 mmol/L NaCl, and the sample was split for protein determination and scintillation counting. Results were expressed in terms of the net βMyHC protein synthesized, as described above.
Indirect Immunofluorescent Staining
Isolated adult cardiac myocytes were cultured on glass coverslips coated with laminin for 48 or 120 hours after infection with adenoviruses and were fixed with 100% methanol (chilled at −20°C) for 5 minutes. Cardiac myocytes were washed with PBS and PBS/1% BSA twice and left in the blocking buffer (5% BSA, 2% nonfat dry milk, 50 mmol/L Tris, and 0.5 mol/L NaCl) for 30 minutes. After removal of the blocking buffer, the cardiac myocytes were incubated with the anti-myosin antibody MF-20 for 30 minutes at room temperature. Samples were washed in PBS/1% BSA three times for 5 minutes each and then incubated with the rhodamine-conjugated affinity-purified goat anti-mouse IgG [F(ab′)2 fragment] (Boehringer Mannheim Co) as the secondary antibody for 30 minutes. A series of experiments with different dilutions of the primary and secondary antibodies was performed to determine their optimal concentration. After antibody treatment, samples were washed in PBS/1% BSA (with a final wash in water), then dried, and mounted with FluorSave reagent (Calbiochem).
To study the formation of sarcomeres, adult feline cardiac myocytes were cultured on glass slide coverslips coated with 1:10 dilution of Matri-gel (Collaborative Research) to increase adhesion. Cardiac myocytes were infected with 106 pfu of recombinant and control adenoviral constructs for 4 hours, washed with PBS three times, and then cultured for an additional 48 or 120 hours. Electron microscopic examination of the cardiac myocytes was performed 48 and 120 hours after infection according to the method of Brinkley et al.27
Rescue of βMyHC Expression Cassette
To confirm that the CMV/βMyHC expression cassette was rescued into the replication-deficient adenoviruses, a set of oligonucleotide primers was designed to amplify a 320-bp segment of the human βMyHC cDNA encompassed between exons 12 and 14 by PCR. The amplified fragment was identified on agarose gel electrophoresis and was sequenced by cycle sequencing.19 The results confirmed the presence of the normal and mutant βMyHC expression cassette in the recombinant adenoviral constructs (data not shown).
Efficiency of Infection of Adult Feline Cardiac Myocytes With Recombinant Adenoviral Constructs
Cardiac myocytes were infected with an Ad5/CMV/Lac-Z construct at MOI values of 1:1, 10:1, 100:1, 500:1, and 1000:1 for 4 hours. The efficiency of infection was low for an MOI of 1:1 (5%), intermediary for an MOI of 10:1 (30%), and high (>95%) for an MOI of >100:1 (Fig 2⇓).
Approximately 30 μg of myosin-enriched extract was electrophoresed on 7.5% polyacrylamide gel for 6 hours, transferred to a membrane, and probed with anti-myosin antibody MF-20. As shown in Fig 3⇓, myosin-enriched protein preparations from COS-1 cells infected with Ad5/CMV/βMyHC-N and Ad5/CMV/βMyHC-403 showed the presence of a 220-kD protein migrating at the corresponding myosin level of the protein size marker, although no myosin band was detected in lanes representing control COS-1 cells and COS-1 cells infected with Ad5ΔE1. Thus, these results indicate that the normal and mutant βMyHC constructs are expressed into full-length βMyHC protein.
Expression of βMyHC mRNA
Agarose gel electrophoresis of the PCR product showed that normal as well as mutant human βMyHC was expressed into mRNA 48 hours after infection (Fig 4⇓). Amplification of the RNA extracted from control cardiac myocytes or cardiac myocytes infected with control Ad5ΔE1 failed to show any product indicating that the primers were specific for human βMyHC. The results indicate expression of the mutant and normal human βMyHC into mRNA.
Total Protein Synthesis
There was no significant difference in the total protein synthesis in cultured control adult feline cardiac myocytes and adult feline cardiac myocytes infected with Ad5ΔE1 (control virus), Ad/CMV/βMyHC-N, or Ad5/CMV/βMyHC-403 viruses (data not shown).
Net Myosin Synthesis
In keeping with expression of the mutant βMyHC construct into the mRNA in feline cardiac myocytes and into protein in COS-1 cells, the results illustrate that βMyHC is expressed into protein. The amount of newly synthesized βMyHC protein was greater in the cardiac myocytes infected with Ad5/CMV/βMyHC-N and Ad5/CMV/βMyHC-403 than in control cardiac myocytes or those infected with control Ad5ΔE1. Net myosin synthesis was 5.2±1.0 nmol phenylalanine per gram protein in control cardiac myocytes, 5.7±1.9 nmol phenylalanine per gram protein in cardiac myocytes infected with Ad5ΔE1, 10.9±2.3 nmol phenylalanine per gram protein in cardiac myocytes infected with Ad5/CMV/βMyHC-403, and 13.42±2.8 nmol phenylalanine per gram protein in cardiac myocytes infected with Ad5/CMV/βMyHC-N (n=8, P<.05).
Indirect Immunofluorescent Staining
Final dilutions of 1:500 of MF-20 and 1:1000 of rhodamine-conjugated goat anti-mouse IgG as the primary and secondary antibodies, respectively, were used for immunofluorescent staining. Diffuse staining of the myofibrillar structures was observed in all cardiac myocytes, indicative of an abundance of myosin protein in adult feline cardiac myocytes. Only examination of cardiac myocytes under high magnification (×600) made it feasible to delineate the myofibrillar structure. Over 100 rod-shaped cardiac myocytes were examined per each group. There were no significant differences in the immunofluorescent staining pattern of the myofibrillar structure of cultured adult cardiac myocytes among experimental groups after 48 hours. Similarly, after 120 hours of culture the myofibrillar structure appeared to be intact in the control cardiac myocytes and those infected with Ad5ΔE1 or Ad5/CMV/βMyHC-N. However, in approximately half of the cardiac myocytes infected with Ad5/CMV/βMyHC-403, the structure of myofibrils showed disarray and lack of appropriate longitudinal alignment along the cell axis (Fig 5⇓).
A total of 60 viable adult feline cardiac myocytes were examined for each experimental group, and the experiments were repeated six times. Cardiac myocytes were considered viable if the plasmalemma of the sarcolemma was intact and mitochondria did not contain amorphous matrix densities. At 48 hours, no significant differences were observed in the structure of the thin and thick filaments or the sarcomeres among experimental groups. In the majority of myocytes examined, the myofibrils were aligned with the long axis of the cells, and the Z bands were in register throughout the length of the cardiac myocytes. Disruption of sarcomeric organization was observed in <10% of the cardiac myocytes in all four experimental groups. However, 120 hours after infection, in cardiac myocytes infected with Ad5/CMV/βMyHC-403, although thick filament formation appeared to be normal, their assembly into sarcomere was markedly impaired (Fig 6⇓). In control myocytes as well as in myocytes infected with Ad5ΔE1 and Ad5/CMV/βMyHC-N, <20% of the cells showed evidence of sarcomeric disarray, which was localized primarily to the ends of the cardiac myocytes and involved ≈<20% of the total myocyte sarcomeres. In contrast, approximately half of the cardiac myocytes infected with Ad5/CMV/βMyHC-403 showed disruption of the sarcomeric organization, affecting at least 50% of the sarcomeres. Bundles of thick filaments as well as clusters of Z bands with emerging rudimentary thick filaments were present in myocytes with disruption of sarcomeric structures. The percentage of cardiac myocytes with severe myofibrillar disarray was significantly greater for the cardiac myocytes infected with the mutant βMyHC construct than for cardiac myocytes infected with the normal βMyHC construct (P=.03).
This is the first expression study of a βMyHC cDNA with an HCM mutation in cardiac myocytes. The cardiac myocytes infected with mutant constructs in cell culture exhibited normal filament and sarcomere formation for 48 hours but subsequently exhibited gross impairment of sarcomere assembly despite persistence of normal-appearing filaments. Since the control myocytes and the myocytes infected with the adenoviral vector alone or the normal βMyHC construct mostly exhibited normal sarcomeres throughout this interval, the Arg403Gln mutation is strongly implicated as the cause of the abnormal sarcomere assembly. The observed normal sarcomere formation for the first 48 hours, even in the mutant infected cells, is also supportive of causality, as evidenced by the time required for turnover of the preformed endogenous normal βMyHC protein, which has a 5.6-day half-life.28 These findings indicate that the expressed mutant myosin is capable of forming thick myosin filaments but that the filaments do not assemble into sarcomeres and myofibrils. Given the relatively long half-life of βMyHC (5.6 days),28 the myofibrillar disarray most likely represents ineffective sarcomeric assembly rather than disruption of the formed sarcomeres. Altered stoichiometry of the components of the sarcomeric proteins is also a possible explanation for the observed disarray of the sarcomeric structure. However, this mechanism is unlikely, because the net βMyHC synthesis was equal in cardiac myocytes infected with normal and mutant βMyHC constructs, despite a significant disparity in the degree of sarcomeric disruption observed among these two groups of cardiac myocytes. Disruption of sarcomere assembly and selective lysis of the thick filaments are characteristic findings in patients with HCM.2 If one assumes that these findings reflect the primary lesion of HCM, impaired sarcomere assembly would be implicated in the early development of the cardiac pathology.
It is not clear whether the construct was expressed as a normal full-length protein in adult feline cardiac myocytes. The data showing increased synthesis of myosin that was of the correct molecular weight and electrophoretic migration are strongly suggestive. The increased myosin synthesis occurred in only the feline cells infected with either normal or mutant βMyHC constructs but not in those infected with the vector alone. However, this was confirmed by the protein expressed in COS cells. These cells do not normally synthesize myosin, but after infection with the adenoviral construct, they expressed a protein that was detected by an antibody specific for sarcomeric myosin. This protein had a molecular weight of 220 kD, identical to that expected for βMyHC. We have shown previously,29 from an analysis of myocardial myosin from a patient that died with the Arg403Gln mutation, that the quantity of cardiac βMyHC myosin was normal (constituting 95% of the myosin) and was normal in proportion to actin. It is not possible to determine from the present study whether the mutant βMyHC molecule is incorporated and subsequently acts as a poisonous peptide to disrupt the involved sarcomere as previously suggested30 31 or whether the sarcomere is not assembled because of some impairment, such as the lack of proper binding of myosin to other sarcomeric proteins. The inherent impaired contractility of the mutant myosin suggested by the previous studies may be secondary to sarcomere assembly. All of the observations remain compatible with our previous hypothesis26 30 and that of others31 : the hypertrophy is secondary and compensatory. It would be speculative to assume that our findings are indicative of HCM; however, the pathognomonic cardiac lesion of HCM in humans and felines is sarcomeric and myofibrillar disarray. Additionally, the results of these studies illustrate the utility of replication-deficient adenoviruses as expression vectors for the study of protein function in mammalian cells. The expression cassette cloned in the vector described in the present study, at ≈7 kb, is one of the largest inserts rescued into an Ad5 vector and is close to the predicted maximum capacity (8 kb) of the system described by Bett et al.20
In the previous attempt to assess the effect of HCM mutations on filament formation, the mutation was inserted into αMyHC cDNA and expressed in COS cells. COS cells do not form sarcomeres, but expression of the normal αMyHC did show filamentous structures in ≈98% of the transfected cells, whereas COS cells transfected with the mutant αMyHC formed nonfilamentous structures in 30% of the transfected cells. Although the studies are markedly different, the implication is similar: impaired myofibril or sarcomere formation is induced by the mutation. We did not perform functional studies, but decreased myosin-actin interaction would be expected in cells with such sarcomere disarray and would be compatible with the decreased velocity of actin-myosin interaction observed by Cuda et al4 and Sweeney et al5 in an in vitro motility assay.
In the present study, we have shown the following: (1) The adenovirus vector can be constructed to incorporate the full-length human βMyHC cDNA and its promoter. (2) The full-length β-MyHC protein is expressed. (3) The βMyHC cDNA is expressed into mRNA, resulting in increased myosin protein. (4) Expression of βMyHC with Arg403Gln mutation results in myofibrillar disarray and disruption of sarcomeric structures. This preparation provides an easily detectable and distinctive morphological phenotype in both the normal and mutant infected cardiac myocytes, which closely resemble the pathological hallmark of the disease both in felines and in humans. These features, in addition to the advantages inherent to feline adult myocytes previously outlined, make this preparation a much improved and more appropriate model for future structure-function analysis of mutations in βMyHC and other sarcomeric genes responsible for HCM, such as the recently identified troponin T and α-tropomyosin.31 A limitation of the present study is the lack of documentation of the incorporation of the mutant myosin into the filament and, second, the lack of accompanying functional studies. We specifically did not use an epitope, since the addition of even a single nucleotide may alter expression or the properties of the expressed myosin. This is particularly important because more than 36 different missense mutations have been identified in the βMyHC gene, each of which impairs cardiac function leading to HCM.3 We are currently developing combinatorial phage display antibodies31 with the hope of obtaining a species-specific antibody against mutant βMyHC. Studies such as we have previously performed23 24 are planned to measure indices of contractility and relaxation in a single-cell preparation to compare the function of normal and mutant infected cardiac myocytes.
This study was supported in part by grants from the National Heart, Lung, and Blood Institute; Specialized Centers of Research (P50-HL42267-01); the American Heart Association; the Bugher Foundation Center for Molecular Biology (86-2216); the American Heart Association, Texas Affiliate, Inc (93G-1191); the Medical Research Council and the National Cancer Institute of Canada. Dr Graham is a Terry Fox Research Scientist of the National Cancer Institute in Canada. We would like to acknowledge Drs Michael Schneider and Brent French for their critical review of the manuscript and helpful suggestions; Dr Ann Goldstein for review and interpretation of the electron micrographs; J. Rudy, Samir Kapadia, MD, Lorrie Kirshenbaum, PhD, and Dorellyn Lee-Jackson for their excellent technical assistance; and Debora Weaver and Esther Yeager for their secretarial assistance in the preparation of this manuscript and figures. Electron microscopic studies were performed at the EM Core Laboratory of the Department of Cell Biology, Baylor College of Medicine.
- Received August 19, 1994.
- Accepted March 17, 1995.
- © 1995 American Heart Association, Inc.
Wynne J, Braunwald E. The cardiomyopathies and myocarditides: toxic, chemical, and physical damage to the heart. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 4th ed. Philadelphia, Pa: WB Saunders; 1992:1394-1450.
Roberts WC, Ferrans VJ, Kaltenbach M, Epstein SW, eds. Hypertrophic Cardiomyopathy: The Therapeutic Role of Calcium Antagonists. New York, NY: Springer-Verlag; 1982:59.
Roberts R, Marian AJ, Bachinski LL. Overview: application of molecular biology to medical genetics. In: Markwald RR, Clark EB, Takao A, eds. Inborn Heart Disease: Developmental Mechanisms. Mount Kisco, NY: Futura Press; 1994:87-111.
Cuda G, Fananapazir L, Zhu W, Sellers JR, Epstein ND. Skeletal muscle expression and abnormal function of β myosin in hypertrophic cardiomyopathy. J Clin Invest. 1993;91:2861-2865.
Sweeney HL, Straceski AJ, Leinwand LA, Tikunov BA, Faust L. Heterologous expression of a cardiomyopathic myosin that is defective in its actin interaction. J Biol Chem. 1994;269:1603-1605.
Straceski AJ, Geisterfer-Lawrance AA, Seidman CE, Seidman JG, Leinwand LA. Functional analysis of myosin missense mutations in familial hypertrophic cardiomyopathy. Proc Natl Acad Sci U S A. 1994;91:589-593.
Kirshenbaum LA, MacLellan WR, French BA, Schneider MD. Highly efficient gene transfer into adult ventricular myocytes by recombinant adenovirus. J Clin Invest. 1993;92:381-387.
French BA, Mazur W, Geske RS, Bolli R. Direct in vivo gene transfer into porcine myocardium using replication-deficient adenoviral vectors. Circulation. 1994;90:2414-2424.
Swynghedauw B. Developmental and functional adaptation of contractile proteins in cardiac and skeletal muscles. Physiol Rev. 1986;3:710-760.
Cooper G, Mercer WE, Hoober JK, Gordon PR, Kent RL, Laurke IK, Marino TA. Load regulation of the properties of adult feline cardiocytes. Circ Res. 1986;58:692-705.
Atkins CE, Gallo AM, Kurzman ID, Cowan P. Risk factors, clinical signs and survival in cats with a clinical diagnosis of idiopathic hypertrophic cardiomyopathy: 74 cases (1985-1989). JAMA. 1992;201:613-618.
Liew CC, Sole MJ, Yamauchi-Takihara K, Kellam B, Anderson DH, Lin L, Liew JC. Complete sequence and organization of the human cardiac β-myosin heavy chain gene. Nucleic Acids Res. 1990;18:3647-3651.
Epstein ND, Cohn GM, Cyran F, Fananapazir L. Differences in clinical expression of hypertrophic cardiomyopathy associated with two distinct mutations in the β-myosin heavy chain gene: a 908 Leu-Val mutation and a 403 Arg-Gln mutation. Circulation. 1992;86:345-352.
Marian AJ, Yu QT, Mares A Jr, Hill R, Roberts R, Perryman MB. Detection of a new mutation in the β-myosin heavy chain gene in an individual with hypertrophic cardiomyopathy. J Clin Invest. 1992;90:2156-2165.
Bett AJ, Haddara W, Prevec L, Graham FL. A highly efficient and flexible system for construction of human adenovirus type 5 vectors with insertions or deletions in early region 1 and early region 3. Proc Natl Acad Sci U S A. 1994;91:8802-8808.
Graham FL, Prevec L. Manipulation of adenovirus vectors. In: Murray EJ, ed. Methods in Molecular Biology: Gene Transfer and Expression of Protocols. 7th ed. Clifton, NJ: Humana Press; 1991:109-128.
Bader D, Masaki T, Fishman DA. Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J Cell Biol. 1982;95:763-770.
Mann DL, Urabe Y, Kent RL, Vinciguerra S, Cooper G IV. Cellular versus myocardial basis for the contractile dysfunction of hypertrophied myocardium. Circ Res. 1991;68:402-415.
Yokoyama T, Vaca L, Rossen RD, Durante W, Hazarika P, Mann DL. Cellular basis for the negative inotropic effects of tumor necrosis factor-α in the adult mammalian heart. J Clin Invest. 1993;92:2303-2312.
Yu QT, Ifegwu J, Marian AJ, Mares A Jr, Hill R, Perryman MB, Bachinski LL, Roberts R. Hypertrophic cardiomyopathy mutation is expressed in mRNA of skeletal as well as cardiac muscle. Circulation. 1993;87:406-412.
Brinkley BR, Murphy P, Richardson LC. Procedure for embedding in situ selected cells in vitro. J Cell Biol. 1967;35:279-283.
Zak R, Martin AF, Prior G, Rabinowitz M. Comparison of turnover of several myofibrillar proteins and critical evaluation of double isotope method. J Biol Chem. 1977;252:3430-3435.
Vybiral T, Deitiker PR, Roberts R, Epstein HF. Accumulation and assembly of myosin in hypertrophic cardiomyopathy with the 403 Arg to Gln β-myosin heavy chain mutation. Circ Res. 1992;71:1404-1409.
Perryman MB, Yu QT, Marian AJ, Mares A Jr, Czernuszewicz G, Ifegwu J, Hill R, Roberts R. Expression of a missense mutation in the mRNA for β-myosin heavy chain in myocardial tissue in hypertrophic cardiomyopathy. J Clin Invest. 1992;90:271-277.