Isoform Transitions of the Myosin Binding Protein C Family in Developing Human and Mouse Muscles
Lack of Isoform Transcomplementation in Cardiac Muscle
Abstract—Mutations in the gene for the cardiac isoform of myosin binding protein C (MyBP-C) have been identified as the cause of chromosome 11–associated autosomal-dominant familial hypertrophic cardiomyopathy (FHC). Most mutations produce a truncated polypeptide that lacks the sarcomeric binding region. We have now investigated the expression pattern of the cardiac and skeletal isoforms of cMyBP-C in mice and humans by in situ hybridization and immunofluorescence microscopy using specific antibodies and probes. We demonstrate that the cardiac isoform is expressed only in cardiac muscle throughout development. The slow and fast isoforms of MyBP-C remain specific for skeletal muscle, where they can be coexpressed. Immunological evidence also suggests that an embryonic isoform of MyBP-C precedes the expression of slow MyBP-C in developing skeletal muscle. This suggests that transcomplementation of MyBP-C isoforms is possible in skeletal but not cardiac muscle.
- familial hypertrophic cardiomyopathy
- myosin binding protein C
- C-protein isoforms
- embryonic expression patterns
- cardiac muscle
Myosin filaments in striated muscle are assembled by the polymerization of myosin in association with a family of MyBPs and titin. These associated proteins1 all share the molecular building plan of the immunoglobulin superfamily.2 3 4 5 Titin is thought to fulfill the role of a molecular ruler,6 and MyBP-C might play a role as a spatially defined regulatory protein.5 7 8 9 Myosin, MyBP-C, and titin form a stable ternary complex where MyBP-C is arranged regularly in 9 of the 11 thick-filament stripes.1 10 11 12
Mutations in the gene for cMyBP-C are the cause of the chromosome 11–associated form of FHC. Most mutations identified so far lead to truncated proteins lacking the thick-filament binding region of cMyBP-C.13 14 15 16 Protein analysis of cardiac biopsies from a patient with a cMyBP-C–associated FHC has indicated that the truncated protein is not present in detectable amounts.16 A possible explanation for the cellular events in cMyBP-C–associated FHC could therefore be a haplophenotype caused by an altered protein stoichiometry. This would suggest that the missing allele cannot be complemented by any of the other known MyBP-C isoforms.
Skeletal muscle contains at least two isoforms of MyBP-C, expressed predominantly either in slow or fast muscle fibers.12 17 18 However, these isoforms can be coexpressed in single myofibrils in varying stoichiometries leading to diverse arrangements of the characteristic sarcomeric stripes.12 Therefore, skeletal muscle has a great potential to adapt in a flexible way to alterations in MyBP-C isoform stoichiometry by the modification of coexpression ratios. So far, it is unclear whether the same applies for cardiac muscle, a question of pressing importance for a more detailed understanding of the cMyBP-C–associated FHCs. Therefore, in the present study, we examine the expression patterns of all known MyBP-C isoforms in mammalian skeletal and cardiac muscle to elucidate this question and suggest that the cardiac isoform of MyBP-C cannot be transcomplemented by skeletal MyBP-Cs.
Materials and Methods
Monoclonal antibodies were raised against the recombinant N-termini of the slow and fast MyBP-C isoforms: Original λ phage isolates containing cDNAs coding for human slow MyBP-C and fast MyBP-C4 19 were used as templates for PCR amplification20 of domains 1 and 2, encompassing the MyBP-C motif. PCR products were cloned into a pET23a vector (Novagen), providing the resulting proteins with an additional histidine tag at their amino termini. Since this tag is recognized by a monoclonal anti-His tag antibody (DIANOVA), expression in Escherichia coli BL21(DE3)pLysS can be monitored by immunoblot analysis. Purification of the soluble recombinant proteins on Ni-NTA agarose columns based on their oligohistidine tags followed standard protocols (Qiagen).
Specific monoclonal antibodies were obtained from female Balb/c mice that had been immunized by a standard immunization scheme21 22 using purified recombinant proteins. Hybridomas were made monoclonal by limiting dilution and characterized further by immunofluorescence and immunoelectron microscopy21 23 and by Western blotting. This yielded the following monoclonal antibodies: α-sMyBP-C (reactive only with slow MyBP-C) and α-fMyBP-C (specific for fast MyBP-C). Antibodies were used as undiluted hybridoma supernatants.
Polyclonal antibodies were raised in rabbits against recombinant fragments of human cMyBP-C (region C0-C124) and domains 1 and 2 of human slow MyBP-C. Sera were used at dilutions of 1:500 for indirect immunofluorescence and at 1:2000 for Western blot analysis. The preimmune sera were found to be unreactive against their respective antigens in immunofluorescence and Western blotting. The monoclonal antibodies I18/23 and S54/6 against the cardiac and ubiquitous isoforms of titin were reported previously.25 Antibodies directed against sarcomeric myosin, actin, titin, and desmin were also described before.26
For in situ hybridization, the following probes were used: cMyBP-C, a fragment comprising base pairs 33 to 785 of human cMyBP-C (corresponding to the cardiac-specific region C0-C124); slow MyBP-C, complete cDNA4; fast MyBP-C, a fragment comprising base pairs 1270 to 3560 of the coding sequence amplified by reverse-transcriptase PCR using human skeletal muscle cDNA (Clontech) as a template (primer sequences were designed from the sequence published19 ); and MHC probes complementary to the 3′ untranslated regions of MHC-β/slow, MHC-2A, and MHC-2X transcripts.27 For in situ hybridization, antisense probes of these fragments were used. Sense probes of the same regions were tested by Northern hybridization on human multitissue blots (Clontech) and found to be negative, whereas the antisense probes detect their respective message (data not shown).
Mouse embryos were from spontaneous matings of Balb/c mice. The presence of a vaginal plug indicating a successful mating was regarded as gd 0. Pregnant mice were killed by cervical dislocation, and their uterine horns were removed and dissected. Further treatment of embryos was as described.28
Human tissues were obtained from 11-week-old and 19-week-old human fetuses from induced or spontaneous abortions and from adult human skeletal muscle by needle biopsy in normal volunteers.29
In Situ Hybridization and Immunocytochemistry
Cryosections of human fetal and adult muscle samples (see above) were processed for in situ hybridization and immunoperoxidase cytochemistry as previously described.30 Final concentration of 35S-labeled cRNA probes was 25 000 to 50 000 cpm/μL; slides were processed for autoradiography using Kodak NTB-2 emulsion and exposed for 3 to 15 days (generally, 7 days). The distribution of MyBP-C transcripts and isoproteins in the different fiber types present in adult human muscle was determined using the DNA probes and antibodies specific for the MyBP-C and MHC isoforms described above. Preparation of mouse embryos for immunofluorescence microscopy was as described.26
Protein extraction and blotting procedures were performed by standard procedures. Mouse embryo carcasses were dissected by decapitation and evisceration of freshly removed embryos in ice-cold Ringer’s solution under a binocular microscope at ×20 magnification. The tissues, consisting only of the rump and thus largely of somitic tissue, were transferred into liquid nitrogen immediately after dissection. The tissues were pulverized under liquid nitrogen in a mortar, transferred immediately into sample buffer,31 and subsequently analyzed on 8% SDS-PAGE. A standard high molecular weight marker (Bio-Rad) was used for reference. All DNA cloning steps followed standard protocols.32
Characterization of Antibodies
Tissue specificity of all the antibodies used in the present study was first controlled by immunofluorescence microscopy using frozen sections of mouse and baboon tissues (Bios). Western blots on total extracts of murine skeletal and cardiac muscle extracts confirmed their reactivity with specific MyBP-C isoforms. Finally, immunoelectron microscopy using Triton-extracted fiber preparations of rat slow and fast skeletal muscles and cardiac ventricle revealed the sarcomeric localization described in earlier studies (results not shown). The polyclonal antibody α-panC was raised against the recombinant MyBP-C motif of human slow MyBP-C. This protein motif is found only in members of the MyBP-C family, where it is highly conserved.5 The antibody α-panC was found to react with MyBP-C of all muscle types, including cardiac muscle (results not shown).
Expression of MyBP-C Isoforms in Embryonic and Adult Mouse Tissues
In skeletal muscle, the isoforms of MyBP-C clearly appeared in a developmentally regulated pattern demonstrating a successive expression of the slow and fast isoforms of MyBP-C (Fig 1⇓). Expression of the slow isoform of MyBP-C succeeded the expression of titin and sarcomeric myosin by 4 or 5 days, ie, at approximately gd 14, a time when primary myotubes have begun to form. Thus, slow or fast MyBP-C and titin/myosin are not coexpressed during early stages of sarcomere formation in mononuclear myoblasts. The fast isoform of MyBP-C was not detected until 4 days later and was then coexpressed in some muscles together with the slow isoform. Cardiac tissue was never stained by antibodies against the slow or fast skeletal isoforms of MyBP-C.
Antibodies directed against the cardiac isoform of MyBP-C showed strong staining of the developing heart from gd 8 onward and coincided with the staining for cardiac titin (I18/2325). However, the cMyBP-C antibody failed to detect any protein in somite-derived muscles at any developmental stage. An interesting aspect was revealed when the analysis of embryo sections was performed with the pan-MyBP-C–reactive antibody α-panC: strong staining was observed in somites from gd 9.5 onward. Cardiac muscle was stained as well. This somitic expression pattern immediately followed the staining for titin and coincided with the staining for the sarcomeric isoforms of MHC and actin (data not shown) (Fig 1⇑).
Western blot analysis of mouse embryos of gd 10.5 using the antibodies against the cardiac (region C0-C1) and slow MyBP-C isoforms and the pan-reactive antibody α-panC demonstrated a band reactive with the panC antibody of an approximate molecular weight of 150 kD in embryo carcasses, containing only somitic tissues. In the same somitic specimens, there is no detectable expression of cMyBP-C in whole embryo bodies that also contain the heart (Fig 2⇓). Reaction with the slow MyBP-C antibody is also negative in these early embryos (Fig 2⇓). We conclude that an immunologically distinct MyBP-C isoform is coexpressed in early embryonic somites with titin and precedes the expression of the slow or fast isoforms of MyBP-C in skeletal muscle. The cardiac isoform remains cardiac specific.
Expression of MyBP-C Isoforms in Human Embryonic and Adult Tissues
The expression of the cardiac and skeletal isoforms of MyBP-C was analyzed in early human fetus tissues from weeks 11 and 19 by in situ hybridization. Figs 3⇓ and 4⇓ demonstrate that in week-11 fetal muscles, the transcript of cMyBP-C was detected only in cardiac muscle but not in skeletal muscle. The fast MyBP-C was still absent in skeletal muscle at this stage but was detected in week-19 fetal skeletal muscles. The cardiac isoform of MyBP-C remained cardiac specific. The successive expression of slow, followed by fast, MyBP-C is therefore common to both mouse and human muscle development (Figs 4⇓ and 5⇓).
In adult human muscle, the transcripts of the slow and fast isoforms of MyBP-C were expressed in variable patterns. The fast MyBP-C isoform was specific for fast type 2 fibers, whereas slow MyBP-C was coexpressed with fast MyBP-C in type 2 fibers and expressed solely in slow type 1 fibers (Fig 6⇓). The distribution of the fast MyBP-C transcript and isoprotein was apparently identical in type 2A and type 2X fibers, identified by the presence of MHC-2A and MHC-2X transcripts (not shown).
Although very low levels of transcript or protein could remain undetected in the present study, our data show that the expression of skeletal isoforms of MyBP-C in cardiac muscle and the cardiac isoform in skeletal muscle does not reach the levels of the respective proper tissue-specific isoform. Thus, although coexpression of MyBP-C isoforms seems possible in skeletal muscles, the cardiac isoform of MyBP-C was never observed in conjunction with any of the other MyBP-C isoforms in either cardiac or skeletal muscle.
We have used in situ hybridization and immunofluorescence analysis to investigate the developmental tissue distribution of the known isoforms of MyBP-C in two mammalian species, mice and humans. Our data demonstrate a successive expression of the skeletal muscle–specific isoforms in an ordered way that resembles the conserved program of myogenesis observed also, although on a shorter time scale, in cultured avian skeletal muscle cells.33 Our data further demonstrate that in two evolutionarily diverged mammals, the cardiac isoform of MyBP-C is not expressed in skeletal muscle throughout development. Our data demonstrate that this is in contrast to the situation in birds, where a splice variant of cMyBP-C is transiently expressed in embryonic skeletal muscle.34 35 The embryonic splice variant of cMyBP-C expressed in avian skeletal muscles lacks the cardiac-specific phosphorylation site.5 35 It was shown that the deletion of this phosphorylation loop in the human cardiac isoform renders the protein essentially insensitive to cAMP-dependent protein kinase A.5 Although birds appear to lack a specific embryonic isoform of skeletal MyBP-C, the embryonic skeletal splice variant of the cMyBP-C gene is therefore similar in its regulatory properties to the genuine skeletal isoforms.
Most cardiac sarcomeric proteins are also expressed in skeletal muscle. Available evidence indicates that in mammals, only two myofibrillar protein isoforms, cardiac troponin I and atrial myosin light chain 2, are exclusively expressed in cardiac muscle.36 Some cardiac isoforms, such as cardiac troponin T, are transiently expressed in developing skeletal muscle both in birds37 and in mammals.38 This is also observed for cMyBP-C in birds34 35 39 but not in mammals.
In mammalian cardiac muscle, the cMyBP-C transcript and protein are detected early on, and the expression coincides with that of sarcomeric MHC in the earliest detectable stages (Fig 1⇑). Therefore, during cardiac myofibrillogenesis, both proteins are tightly linked and might be coassembled with titin and myosin. Again, the isoform specificity of MyBP-C remains tissue specific, and neither skeletal isoform is detected in cardiac muscle throughout development (Figs 1⇑ and 3⇑). Both muscle types are therefore following a defined expression program that allows no switching of isoforms between skeletal and cardiac muscle.
Our results show that slow MyBP-C is first expressed in developing skeletal muscle both in mice and humans and that fast MyBP-C is detected only at later developmental stages. This predominance of a slow isoform in developing muscle has been observed with other myofibrillar protein gene families. For example, the slow troponin I gene is predominantly expressed during development, and there is a subsequent switch to fast troponin I in fast muscles.40
In addition, our results indicate that an embryonic MyBP-C isoform, immunologically distinct from the fast, slow, and cardiac isoforms, is present in mouse skeletal muscle at very early developmental stages. Previous biochemical observations were restricted to avian muscle and suggested the existence of such an embryonic MyBP-C isoform possibly related to the cardiac isoform.33 34 35 Interestingly, the detection of the mammalian embryonic isoform coincides with the expression of titin and MHC (Fig 6⇑), suggesting that similar to the nearly synchronous coexpression of cMyBP-C and titin/myosin in the heart, a corresponding situation is found in skeletal muscle. Although one would not have assumed that the assembly mechanisms of cardiac and skeletal thick filaments would be distinctly different, the observation of the large time lag of >5 days between the expression of titin or myosin and slow MyBP-C in mice might have led to a model of sequential sarcomeric assembly of titin, myosin, and MyBP-C rather than the coassembly that our data strongly suggest.
The expression of cMyBP-C is therefore tightly restricted to cardiac muscle, which does not express the slow or fast skeletal isoforms of MyBP-C under normal circumstances. Unlike in skeletal muscle, which expresses up to three isoforms of MyBP-C in addition to MyBP-H, in cardiac muscle a transcomplementation of isoforms therefore seems to be impossible. This could plausibly explain why a dominant-negative phenotype is observed in cardiac muscle in mutations of the cMyBP-C gene, whereas no skeletal myopathies have been associated with a member of the MyBP-C family. Clearly, it will now be interesting to assay diseased myocardium from patients with MyBP-C–associated FHC for the presence of skeletal isoforms of MyBP-C in order to determine whether the expression barrier of skeletal MyBP-Cs could be overcome under pathological conditions.
Selected Abbreviations and Acronyms
|cMyBP-C||=||cardiac isoform of MyBP-C|
|FHC||=||familial hypertrophic cardiomyopathy|
|gd (with number)||=||gestation day|
|MHC||=||myosin heavy chain|
|MyBP||=||myosin binding protein|
|MyBP-C||=||myosin binding protein C|
|PCR||=||polymerase chain reaction|
This study was supported by the Deutsche Forschungsgemeinschaft (Ga 405/3–1 to Dr Gautel and Fu 339/1–3 to Prof Fürst), the European Union (BIOMED contract PL950174 to Dr Schiaffino), MURST, and ASI.
This manuscript was sent to Laurence H. Kedes, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
- Received July 1, 1997.
- Accepted November 7, 1997.
- © 1998 American Heart Association, Inc.
Einheber S, Fischman D. Isolation characterization of a cDNA clone encoding avian skeletal muscle C-protein: an intracellular member of the immunoglobulin superfamily. Proc Natl Acad Sci U S A. 1990;87:2157–2161.
Vaughan KT, Weber FE, Einheber S, Fischman DA. Molecular cloning of chicken myosin-binding protein H (MyBP) H (96 kDa protein) reveals extensive homology with the MyB-C (C-protein) with conserved immunoglobulin C2 fibronectin type III motifs. J Biol Chem. 1993;268:3670–3676.
Fürst DO, Vinkemeier U, Weber K. Mammalian skeletal muscle C-protein: purification from bovine muscle, binding to titin and the characterization of a full-length cDNA. J Cell Sci. 1992;102:769–778.
Whiting A, Wardale J, Trinick J. Does titin regulate the length of thick filaments? J Mol Biol. 1989;205:163–169.
Craig R, Offer G. The location of C-protein in rabbit skeletal muscle. Proc R Soc Lond B Biol Sci. 1976;1976:192:325–332.
Hartzell HC, Glass DB. Phosphorylation of purified cardiac muscle C-protein by purified cAMP-dependent and endogenous Ca2+-calmodulin-dependent protein kinases. J Biol Chem. 1984;259:15587–15596.
Schlender K, Bean LJ. Phosphorylation of chicken cardiac C-protein by calcium/calmodulin-dependent protein kinase II. J Biol Chem. 1991;266:2811–2817.
Dennis JE, Shimizu T, Reinach FC, Fischman DA. Localization of C-protein isoforms in chicken skeletal muscle: ultrastructural detection using monoclonal antibodies. J Cell Biol. 1984;98:1514–1522.
Okagaki T, Weber FE, Fischman D, Vaughan KT, Mikawa T, Reinach FC. The major myosin-binding domain of skeletal muscle MyBP-C (C-protein) resides in the COOH-terminal immunoglobulin C2 motif. J Cell Biol. 1993;123:619–626.
Bonne G, Carrier L, Bercovici J, Cruaud C, Richard P, Hainque B, Gautel M, Labeit S, James M, Weissenbach J, Vosberg H-P, 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.
Rottbauer W, Gautel M, Zehelein J, Labeit S, Franz WM, Grünig E, Brown BD, Vollrath B, Mall G, Dietz R Katus HA. A novel cardiac myosin-binding protein-C gene splice donor site mutation in familial hypertrophic cardiomyopathy is transcribed to mRNA in myocardium. J Clin Invest.. 1997;100:675–682.
Carrier L, Bonne G, Bährend E, Yu B, Richard P, Niel F, Hainque B, Cruaud C, Gary F, Labeit S, Bouhour J-B, Dubourg O, Desnos M, Hagège AA, Trent RJ, Komajda M, Fiszman M Schwartz K. Organization and sequence of human cardiac myosin binding protein C gene (MYBPC3) and identification of mutations predicted to produce truncated proteins in familial hypertrophic cardiomyopathy. Circ Res. 1997;80:427–434.
Yamamoto K, Moos C. The C-proteins of rabbit red, white cardiac muscles. J Biol Chem. 1983;258:8395–8401.
Dennis JE, Schimizu T, Reinach FC, Fischman DA. Localization of C-protein isoforms in chicken skeletal muscle: ultrastructural detection using monoclonal antibodies. J Cell Biol. 1984;98:1514–1522.
Weber FE, Vaughan KT, Reinach FC, Fischman DA. Complete sequence of human fast-type slow-type muscle myosin-binding-protein C (MyBP-C). Eur J Cell Biol. 1993;216:661–669.
Saiki RK, Scharf SJ, Faloona F, Mullis GT, Erlich HA. Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science. 1985;230:1350–1354.
Fürst DO, Osborn M, Nave R, Weber K. The organization of titin filaments in the half-sarcomere revealed by monoclonal antibodies in immunoelectron microscopy: a map of ten non-repetitive epitopes starting at the Z line extends close to the M line. J Cell Biol. 1988;106:1563–1572.
Fürst DO, Nave R, Osborn M, Weber K. Repetitive titin epitopes with a 42 nm spacing coincide in relative position with known A band striations also identified by major myosin-associated proteins: an immunoelectron microscopical study on myofibrils. J Cell Biol. 1989;94:119–125.
Obermann WMJ, Gautel M, Steiner F, Van der Ven PFM, Weber K, Fürst DO. The structure of the sarcomeric M band: localization of defined domains of myomesin, M-protein and the 250 kDa carboxyterminal region of titin by immunoelectron microscopy. J Cell Biol. 1996;134:1441–1453.
Fürst DO, Osborn M, Weber K. Myogenesis in the mouse embryo: differential onset of expression of myogenic proteins and the involvement of titin in myofibril assembly. J Cell Biol. 1989;109:517–527.
Smerdu V, Karsch-Mizrachi I, Campione M, Leinwand L, Schiaffino S. Type 2X myosin heavy chain transcripts are expressed in type 2B fibers of human skeletal muscle. Am J Physiol. 1994;267:1723–1728.
Hogan B, Beddington R, Costatini F, Lacy E. Manipulating the Mouse Embryo: A Laboratory Manual. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1994.
Andersen JL, Schiaffino S. Mismatch between myosin heavy chain mRNA and protein in human skeletal muscle fibers. Am J Physiol. In press.
De Nardi C, Ausoni S, Moretti P, Gorza L, Velleca M, Buckingham M, Schiaffino S. Type 2X myosin heavy chain is coded by a muscle fiber type-specific developmentally regulated gene. J Cell Biol. 1993;123:823–835.
Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K. Current Protocols in Molecular Biology. New York, NY: John Wiley & Sons Inc; 1987.
Lin Z, Lu MH, Schultheiss T, Choi J, Holtzer S, DiLullo C, Fischman DA, Holtzer H. Sequential appearance of muscle-specific proteins in myoblasts as a function of time after cell division: evidence for a conserved myoblast differentiation program in skeletal muscle. Cell Motil Cytoskeleton. 1994;29:1–19.
Schiaffino S, Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation functional significance. Physiol Rev. 1996;1996:76:371–423.
Cooper TA, Ordahl CP. A single cardiac troponin T gene generates embryonic adult isoforms via developmentally regulated alternate splicing. J Biol Chem. 1985;260:11140–11148.
Saggin L, Gorza L, Ausoni S, Schiaffino S. Cardiac troponin T in developing, regenerating denervated rat skeletal muscle. Development. 1990;110:547–554.