CELL BIOCHEMISTRY AND FUNCTION VOL.
8: 117-130 (1990)
Intracellular Assembly of Newly Synthesized Canine Cardiac Myosins YOSHINORI SEKOt, SHIGETO NAITO?, KOUJI IMATAKA?, JUN FUJIIt, PAUL K. NAKANES, FUMIMARO TAKAKU6 AND YOSHIO YAZAKIG ?The Institute for Adult Diseases, Asahi Life Foundation, Shinjuku-ku, Tokyo, fDepartment of Cell Biology, School of Medicine, University of Tokai, Isehara, Kanagawa and $Third Department of Internal Medicine, Faculty of Medicine, University of Tokyo, Bunkyo-ku, Tokyo, Japan
To investigate how newly synthesized cardiac myosins are assembled into myofilaments, we analysed the distribution of newly produced a-myosin heavy chain isozyme in sarcomeres by immunoelectron microscopy using a monoclonal antibody (CMA19), which is specific for a-myosin heavy chain. Isozymic changes in myosin heavy chains from to a type were induced in canine ventricular muscles and cultured ventricular myocytes by administration of 1-thyroxine. We incubated the glycerinated ventricular muscles or cultured ventricular myocytes with the enzyme (horseradish peroxidase) labelled Fab fragment of CMA19. After the reaction with 3, 3’-diaminobenzidine and osmification, we prepared ultrathin sections of the ventricular muscles or cultured ventricular myocytes and analysed their staining patterns by electron microscopy. There was apparent heterogeneity in the staining intensity of the myofilaments among different cells, among different myofibrils and even intramyofibrillarly. Higher magnification revealed that there were scattered foci of strong reaction which appeared to be foci of assembly of the newly synthesized a-myosin heavy chain. Immunocytochemical study also showed heterogeneous reactions within myofilaments and that there were scattered foci of myofilament assembly, which were closely associated with polyribosomes producing newly induced a-myosin heavy chain. These data suggest that newly synthesized cardiac myosins are assembled into myofilaments from the sites of synthesis, that is polyribosomes. This may explain the heterogeneity of the assembly pattern of newly synthesized cardiac myosins at the subcellular level. KEY WORDS -myosin
heavy chain isozyme; cultured cardiac myocyte; thyroxine; immunoelectron microscopy; enzyme-labelled antibody; monoclonal antibody; myofilament; polyribosome.
ELISA, enzyme-linked immunosorbent assay; MHC, myosin heavy chain; PBS, phosphate buffered saline; DMEM, Dulbecco’s modified Eagle’s minimum essential medium.
ABBREVIATIONS-
INTRODUCTION The metabolic rate of cardiac muscles, which perform continuously repetitive contraction, appears to be much higher than that of skeletal muscles. Cardiac myosin, one of major contractile proteins, is turned over very rapidly with a half-life of 5-6 This work was partly presented at the 59th Scientific Session of the American Heart Association (1986, Circulation 74, 11-82. [Abstr.]). Supported by grants-in-aid for scientific research (Nos 60480228, 60570382) from the Ministry of Education, Science and Culture and a grant on calcium signal transduction in cardiovascular system, Japan. Addressee for correspondence: Yoshinori Seko, Third Department of Internal Medicine, Faculty of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan. 0263-6484/90/020117- 14 S07.00 @ 1990 by John Wiley & Sons, Ltd.
days.’ Thus the attachment and detachment of myosin molecules must occur rapidly and extensively in myofilaments without interrupting forcegeneration. Although many researchers have studied the mechanism involved in assembly and disassembly of myosin molecules using in vitro systems, until recently, there has been no evidence showing how newly synthesized cardiac myosins are assembled into myofilaments in living cardiac myocytes. Recently, Wenderoth and Eisenberg’ showed that, by manipulating the thyroid hormone level, newly induced a-type cardiac myosin heavy chain (MHC) was incorporated into thick filaments with greater exchange at the ends than in the central
118 region in ventricular muscle of rabbits. However, they examined the distribution of newly induced aMHC in the ventricular muscle at a rather early stage of thyroid-treatment, in which newly induced a-MHC had not been fully incorporated into thick filaments. They speculated that a possible imbalance between the rates of synthesis and incorporation resulted in accumulation of a-MHC at the ends of the thick filaments. It appears that this imbalance resulted from the conditions of functioning cardiac muscle, where myofilaments are arranged densely and orderly for force-generation. Lawrence and Singer3 demonstrated that the intracellular localization of cytoskeletal proteins might reflect that of corresponding mRNAs, which were bound and translated by polyribosomes. However, it is much easier to study the spatial relationship between thick filaments and polyribosomes, which are producing MHC, in cultured cardiac myocytes than in intact functioning cardiac muscle. The purposes of this study are to investigate how newly synthesized cardiac myosins are assembled into myofilaments and to study the spatial relationship between their sites of synthesis, that is polyribosomes, and their sites of function. For these purposes, we induced isozymic changes in MHC from B to a type in canine ventricular muscles and cultured canine ventricular myocytes by administration of 1-thyroxine, and analysed the distributions of newly induced u-MHC in sarcomeres and polyribosomes producing a-MHC by immunoelectron microscopy. To enable the anti-MHC antibody to reach all parts of the intracellular structures, we used a Fab fragment of a monoclonal antibody (CMA19), which is specific for a-MHC, directly labelled with horseradish peroxidase. Our data suggest that newly synthesized a-MHC are assembled into myofilaments heterogeneously and near the sites of synthesis.
Y. SEKO ETAL.
peritoneal injections at two-week intervals with 0.1-0.2 ml of 1 mg ml-’ solutions of calf atrial myosin as a-MHC antigen. Mice were killed three days after the intravenous booster injection, and their isolated spleen cells were fused with the myeloma cells (P3X63Ag8UI). Anti-myosin activity in a medium from hybridoma colonies was screened by enzyme-linked immunosorbent assay (ELISA) using goat biotinylated anti-mouse immunogloblin and avidin D-peroxidase (Vector Laboratories Inc. Burlingame, CA), as described p r e v i~ u sly .The ~ antibody CMAl9 also reacted specifically with atrial myosin in the dog, and was selected for use in this study. To enable the anti-myosin antibody to reach all parts of the cardiac tissue, even to the level of myofilaments, we cut the antibody into Fab fragments by papain digestion, as described elsewhere.’ ELISA test showed the preservation of strong reactivity of the Fab fragment of CMA19 (CMA19Fab) with atrial myosin (Figure 1). Hor-
CMA 19Fab
E
e
0 Ventricular myosin
Ln V
Light chains ( I
d 0
MATERIALS AND METHODS
0 Atrial myosin
C
0
50i 1
+ 11)
\
Preparation of Monoclonal Antibody SpeciJc for aMHC
Myosins for immunization were prepared from calf atrial muscles by a dilution technique, as previously d e ~ c r i b e d . The ~ light chains were isolated by guanidine denaturation, as described el~ewhere.~ The preparations of hybridoma producing anti-myosin antibody specific for a-MHC have previously been de~cribed.~.’ Briefly, BALB/c mice (male, 6 weeks) were immunized by five intra-
Figure 1. Reactions of monoclonal antibody (CMA19Fab) with canine myosin by enzyme-linked immunosorbent assays. Results expressed as the percentage of maximum optical density at 540nm show specific reaction of CMA19Fab with atrial myosin ( O ) ,negligible reaction with ventricular myosin (O), and no reaction with light chains (0).
119
ASSEMBLY OF CARDIAC MYOSINS
seradish peroxidase was used for labelling of the monoclonal antibody because of its small molecular weight and availability for both light and electron microscopy. We labelled the monoclonal antibody (CMA19Fab) directly with horseradish peroxidase by the method of Wilson and Nakane.' Preparation of Cardiac Tissue Sections
Mongrel dogs weighing 8-15 kg were injected with 1-thyroxine (daily dose 300 pg kg-' subcutaneously) for 6 and 21 days. Saline was injected into the control group in the same way. Each dog was killed with potassium chloride (administered intravenously), and fresh heart muscle tissue was dissected from the left ventricular free walls in both groups and left atria in control group. After glycerination for more than one month at -2O"C, the tissues were washed with phosphate buffered saline (PBS; 0.01 M, pH 7.2) for 3 h (two changes hr-') at 4°C and cut into small blocks, then 4 pm sections were prepared by a cryostat for immunostaining. Preparation of Cultured Cardiac Myocytes
of them was kept at 37°C during Langendorff perfusion. Then, the heart was incubated in high K-low C1 solution at 4°C for more than 1 h, and was cut into small pieces of atria and ventricles. The isolation of atrial or ventricular myocytes was done simply by stirring the tissue piece in serum-free Dulbecco's modified Eagle's minimum essetial medium (DMEM) containing insulin-transferrin-sodium selenite media supplement (Sigma), penicillin and streptomycin. After the cells were cultured on slides under a humidified atmosphere of 5 per cent CO, and 95 per cent air at 37°C overnight, when most of them attached to the surfaces of slides, the culture media of ventricular and atrial myocytes were replaced with new media containing l-thyroxine (100 nM) for the thyroxine-treated group and the positive control group, respectively. The culture media of ventricular myocytes were also replaced with new media without 1-thyroxine for the negative control group. The cells were cultured in the same condition for 7 days, and were used for immunostaining. Immunostaining for Light and Electron Microscopy
The procedures used for isolating single cardiac The procedures used in immunostaining for light myocytes were essentially as described previously'o and were held aseptically. Neonatal mongrel dogs and electron microscopy by the enzyme-labelled of 1- 14 days old, weighing 200-600 g, were anesthe- antibody method were essentially those described tized with pentobarbital (25 mg kg-' body weight, by Nakane and Pierce."*12 Briefly, the cryostat administered intraperitoneally). The chest was sections were air-dried for 30 min at room temperaopened and the right ventricle was injected with ture and fixed with 100 per cent acetone for 60 min at 4°C. Washing was done three times with PBS for Ca-free Tyrode solution containing in r n ~ NaCl , 136.9, NaHCO, 11.9, KCl 5.4, MgCl, 0.53, 15 min. The sections were then treated with 1 per NaH,PO, 0.33, HEPES 5.0 and glucose 5.5 (pH cent Triton X-100 (Sigma) in PBS for 5 rnin at 7.3-7.4), to almost stop the contraction. The heart room temperature and washed three times with was then dissected out and the aorta was cannu- PBS for 15 min. After incubation with horseradish lated to perfuse the coronary arteries (Langendolf peroxidase-labelled CMA 19Fab overnight at room perfusion) with Ca-free Tyrode solution, and the temperature, the sections were washed three times heart was hung in a moist chamber where the with PBS for 15 min, then fixed with 1 per cent perfusion was continued at hydrostatic pressure of glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) 65 cm. After the blood was washed out, the perfu- for 5 min. They were washed with PBS for 5 min, sate was replaced with Ca-free Tyrode solution and reacted with 0.05 M Tris-HC1 buffer (pH 7-6) containing 0-45 mg ml- collagenase (Type I, containing 0.02 per cent (wt/vol) 3,3'-diaminobenSigma Chemical Co., St. Louis, MO), which was zidine4HCl. The reaction was carried out first with recirculated by a peristaltic pump. The perfusion of 1 per cent dimethylsulfoxide for 1 h for preincubacollagenase was continued for 30 min, then washing tion, and then with H,O, (17p1 of a 30 per cent out was done with high K-low C1 solution contain- H,O, solution per 100 ml of 0.05 M Tris-HC1 buffing in mM, taurine 10, oxalic acid 10, glutamic acid er) for 2 rnin at room temperature. For light 70, KCl 25, KH,PO, 10, HEPES 10, glucose 11 microscopy, the sections were washed with water, and EGTA 0-5 (pH 7.4). All perfusates were corn- dehydrated with ethanol and coverslips were pletely gassed with carbogen, consisting of 95 per mounted with resin in xylene. For electron microcent 0,, and 5 per cent CO,, and the temperature scopy, they were washed three times with PBS for
Figure 2. Light micrographs of canine cardiac muscle stained with CMA19Fab by the enzyme-labelled antibody method. (A) and (B) ventricular muscle of the thyroxine-treated group, for 6 day (A) and 21 day (B) treatment. (A) Weakly reactive myofibres distribute heterogeneously. (B) Many myofibres are clearly reactive. Note apparent heterogeneity in the reaction among different myofibres. Ventricular (C) and atrial (D) muscles of the saline-treated control group. (C) All myofibres are unreactive. (D) All myofibres are strongly and almost homogeneously reactive. Bar, 30 pm.
121
ASSEMBLY OF CARDIAC MYOSINS
15 min and treated with 2 per cent osmium tetroxide in 0.1 M phosphate buffer (pH 7.4) for 1 h. They were again washed with PBS, followed by ethanol dehydration and embedding in Epok 812 (Ouken Shoji Co., Tokyo, Japan) resin. Ultrathin sections were prepared with a diamond knife and examined in a JEOL JEM1200EX electron microscope. Cultured cardiac myocytes on slides were washed three times with PBS for 15 min, then fixed with 4 per cent paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4°C for 30 min and washed three times with PBS for 15 min. The subsequent procedures for immunostaining of a-MHC, without treatment with 1 per cent Triton X-100, were the same as for tissue samples. RESULTS Immunohistochemistry by Light Microscopy
As for the thyroxine-treated groups, many ventricular myofibres were weakly reactive with CMAl9Fab and there was heterogeneous distribution of
reactive myofibres in the 6 day-treated group (Figure 2A). But, most of the ventricular myofibres were clearly reactive and there was apparent heterogeneity in the reaction among different myofibres in the 21 day-treated group (Figure 2B). In the saline-treated control group, all except a few of the ventricular myofibres were unreactive (Figure 2C), all of the atrial myofibres were strongly and almost homogeneously reactive with CMA 19Fab (Figure 2D). Immunohistochemistry by Electron Microscopy
We examined the permeability of the enzymelabelled anti-myosin antibody in cardiac muscle using atrial muscle as a positive control. There were strong and almost homogeneous reactions with CMA 19Fab at the level of the sarcomere (Figure 3), showing that CMA 19Fab could penetrate throughout the cardiac tissue, even to the level of the myofilament. In the thyroxine-treated ventricles, the intensity of the reaction with CMA19Fab in the 6 day-
Figure 3. Electron micrograph of atrial muscle of the saline-treated control group stained with CMA19Fab. Strong and almost homogeneous reaction is attained throughout the tissue at the level of the sarcomere. Bar, 1 pm.
Figure 4. Electron micrographs of ventricular muscle of the thyroxine-treated (for 21 days) group stained with CMA19Fab. (A) Reactive and unreactive myofibres exist side by side. Note the heterogeneity in the reaction and foci of strong reaction within the reactive myofibril. Bar, 1 pm. (B) Reactive and unreactive cells are connected by an intercalated disc. The intensity of the reaction within the reactive cell decreases away from the intercalated disc. Bar, 1 pm.
123
ASSEMBLY OF CARDIAC MYOSINS
treated group was not strong enough for electron microscopic study. Therefore, we used a 21 daytreated group for the study. There was apparent heterogeneity in the reaction between adjacent myofibrils (Figure 4A), as well as between adjacent myofibres (Figure 4B). Such heterogeneous reaction existed even in the same reactive myofibre (Figure 4A). The distribution of reactive areas in the same myofibre did not appear to be random. Instead foci of strong intensity and a tendency of decreased intensity away from the foci were noted (Figure 4A). Almost homogeneous reaction was observed within the same sarcomere near an intercalated disc in the reactive myofibre (Figure 4B), showing that each myofilament of the sarcomere in this part of the myofibre had almost homogeneous distribution of a-MHC. Higher magnification (Figure 5A) revealed that there were scattered areas of strong reaction at the sarcomere level and that they appeared to be the foci of assembly of the newly synthesized a-MHC. Heterogeneous distribution of or-MHC was found even in the same myofilament in the course of isozymic redistribution. In contrast, there was no or negligible reaction in the ventricles of the salinetreated control group (Figure 5B), showing that there was minimal non-specific binding of CMA19Fab.
mainly consisted of a-MHC, were always closely associated with reactive polyribosomes. These scattered foci of reactive myofilaments and reactive polyribosomes, which means the foci of assembly of newly synthesized a-MHC, appeared to correspond to those seen in thyroxine treated ventricular muscle. Orderly arrangement of reactive myofilaments was seen at the tips of pseudopodia (Figure 8A, arrows). A similar pattern of the myofilament assembly was seen at the cell periphery where the myocytes touched an adjacent cell (Figure 8B). These special arrangements of myofilaments at the cell periphery appear to be related to cell motility, for which actin and myosin filaments cooperatively provide the major mechanical force. This is also supported by the findings about the intracellular localization of actin mRNA.3 Higher magnification of a focus of myofilament assembly is shown in Figure 9. Reactive and unreactive polyribosomes were closely associated with myofilaments, within which non-uniform reactions were noteworthy. The coincidental localization of reactive myofilaments and reactive polyribosomes strongly suggests that newly synthesized a-MHC were assembled into myofilaments from the nearest sites of synthesis. In this study, we could not identify any preferred region of the myofilaments for the assembly of newly synthesized a-MHC. DISCUSSION
Immunocytochemistry by Electron Microscopy
To confirm the permeability of enzyme-labelled CMA 19Fab and to evaluate the non-specific binding of it also in cultured cardiac myocytes, we examined the reactions of it with thyroxine-treated atrial myocytes and non-treated ventricular myocytes as positive and negative controls, respectively. We found again strong and almost homogeneous reactions in sarcomeres of thyroxine-treated atrial myocytes (Figure 6A) and no or negligible reaction in those of non-treated ventricular myocytes (Figure 6B). This indicates that technical artifacts due to diffusion of CMA19Fab or sectioning are minimal. We also found that most of the myofilaments were arranged in an orderly manner in both the control groups. The reaction patterns in thyroxine-treated ventricular myocytes are shown in Figures 7 and 8. As shown in Figure 7, there were scattered foci of myofilament assembly, where reactive and unreactive myofilaments were almost randomly mingled with one another. Reactive myofilaments, which
In this study, we clearly demonstrated a heterogeneous distribution of newly synthesized a-MHC among different cells, among different myofibrils, and even intramyofibrillarly. Further, there appeared to be scattered foci of assembly of newly synthesized a-MHC at the sarcomere level. Immunocytochemical study also showed that there were scattered foci of myofilament assembly, which were closely associated with polyribosomes producing newly induced a-MHC. These observations strongly suggest the close correlation between the intracellular sites of assembly and those of synthesis in myosin turnover. In human ventricular muscles, Mercardier et ~ 1 . ' have ~ quantified the amount of V1-type isomyosin which was found to range from almost 0 to 15 per cent (average, a little less than 5 per cent) of total myosin, and to vary from one heart to another. Recently, we demonstrated the distribution of ventricular myofibres containing a-MHC. Their number ranges from 0 to 15 per cent of the total myofibres but with regional ~ a r i a t i o n . 'In ~
Figure 5. Electron micrographs of ventricular muscle of the thyroxine-treated (for 21 days) group (A) and the saline-treated control group (B) stained with CMA19Fab. (A) Higher magnification shows scattered areas of strong reaction at the level of the sarcomere. Bar, 1 pm. (B) No or negligible reaction is observed in the saline-treated group. Bar, 1 pm.
125
ASSEMBLY OF CARDIAC MYOSINS
Figure 6. Electron micrographs of 7-day thyroxine-treated cultured atrial myocyte (A) and nontreated cultured ventricular myocyte (B) stained with CMA19Fab. (A) Strong and almost homogeneous reaction is attained throughout the sarcomeres. Bar, 1 pm. (B) N o or negligible reaction is observed. Bar, 1 urn. Note that most of the mvofilaments are arranged in an orderly manner in both (A) and (B).
this study, we found the proportion of ventricular myofibres, which were reactive with CMA19, to be from 0 to 10 per cent in the canine control group, which was about the same as in human. On the other hand, most of the ventricular myofibres reacted with CMA19Fab in the thyroxine treated group. Therefore, most of the a-MHC, which appeared after thyroxine treatment, must have been newly synthesized. Recent studies from our and other laboratories have revealed the presence of two isozymes in MHC, a and p types, and their relative proportion can be changed by aging, hormonal state and
pressure ~ v e r l o a d . ~ . 5-20 ~ . ~ . We ' and others have also reported the heterogeneity, at the cellular level, in the contents of the MHC isozymes as well as heterogeneous isozymic redistribution resulting from pressure overload or in response to thyroid h o r r n ~ n e . ~Mahdavi .~' et a1." identified two genes coding for two MHC isozymes and showed them to be organized in tandem. They showed that the isozymic changes in MHC during development or in response to thyroid hormone were regulated at Further, they speculated that thymRNA roid hormone may have a direct effect on transcription of the MHC genes and that one of the major
Figure 7. Electron micrograph of a 7-day thyroxine-treated cultured ventricular myocyte stained with CMA19Fab. Several foci of myofilament assembly are scattered within the cell. Note that reactive and unreactive myofilaments are randomly mingled with one another and are closely associated with reactive and unreactive polyribosomes (arrow). Bar, 1 jim.
Figure 8. Electron micrographs of 7-day thyroxine-treated cultured ventricular myocytes stained with CMA19Fab. (A) Arrows indicate orderly arrangement of reactive myofilaments at the tips of pseudopodia. Bar, 1 pm. (B) Note the orderly arrangement of reactive myofilaments at the cell periphery where the myocyte touches an adjacent cell. Bar, 1 pm.
128
Y.SEKO ETAL.
Figure 9. Higher magnification of a focus of myofilament assembly in a 7-day thyroxine-treated cultured ventricular myocyte stained with CMA19Fab. Many myofilaments show non-uniform reactions within themselves and are closely associated with reactive and unreactive polyribosomes. Arrows indicate Z-discs. Bar, 1 pm.
reasons .for the heterogeneity in the expression of MHC genes at the cellular level may be either quantitative or qualitative differences in thyroid hormone receptors in the nuclei. Until recently, no previous study has evidenced the heterogeneity in the distribution of newly synthesized myosins at the subcellular level. But, there is some information about the mechanisms of intracellular protein synthesis and assembly, which may be of great use in understanding the background of this heterogeneity. HuxleyZ4 first deduced from electron micrographs of negatively stained filaments that the thick filaments of striated muscles were remarkable bipolar structures in which myosin molecules were systematically ordered with their tails in the backbone and heads arranged along the surface. He also demonstrated that native thick filaments were dissociated in high salt into their constituent myosin molecules and again associated to form thick filaments in low salt, suggesting that the packing forces holding these structures together were mainly ionic
in nature. Saad et al.25,26demonstrated, using a fluorescence energy transfer system, that there were rapid and extensive exchanges of myosin molecules in thick filaments as well as between thick filaments and soluble myosins in uitro. They also suggested that this exchange mechanism might be a possible in-viuo mechanism for myosin isoform transitions during muscle development and myosin turnover in thick filaments. Recently, Wenderoth and Eisenberg’ studied the distribution of newly induced a-MHC in the left ventricular muscle of thyroid-treated rabbits and found that newly induced a-MHC was incorporated into thick filaments with greater exchange at the ends than in the central region of the thick filaments. They speculated that during periods of rapid protein synthesis the rates of synthesis of aMHC might be greater than the rates of incorporation of a-MHC into the thick filaments and that this imbalance would result in accumulation of new myosin at the ends of the thick filaments. They examined the distribution of a-MHC in the ventri-
129
ASSEMBLY OF CARDIAC MYOSINS
cular muscle treated with thyroid hormone for up to 4 days. This seems to be a rather early stage of thyroid-treatment to study the distribution of the incorporated a-MHC. It appears that the densely and orderly arrangement of myofilaments in functioning cardiac muscle may be one of the causes of the imbalance between the rates of synthesis and incorporation. Our present study showed that there were scattered foci of assembly of newly induced aMHC in ventricular muscle treated with thyroxine for 21 days. At this stage, it seems that most of the newly synthesized myosins were assembled into myofilaments and a balance had been reached between the rates of synthesis and of incorporation. Scattered foci of assembly of newly induced aMHC were also observed in thyroxine-treated cultured ventricular myocytes and appeared to correspond to those seen in the ventricular muscle. They were closely associated with polyribosomes producing a-MHC, suggesting a close spatial relationship between sites of synthesis and sites of assembly. Fulton et L Z ~ . ,using ~’ cultured fibroblasts, found a relationship between the distribution of polyribosomes and an associated assembly of proteins into intact cells and skeletal frameworks. Although their data were from cultured fibroblasts, the results seem highly suggestive as a model of intracellular protein assembly. They showed that newly synthesized skeletal proteins were assembled into the skeletal frameworks at or very near their site of synthesis (that is, polyribosomes) with limited subsequent exchange between the skeletal frameworks and soluble proteins. Recently, they reported cotranslational assembly of MHC in developing cultured skeletal muscle.*’ They suggested that the mode of assembly of skeletal proteins in cultured fibroblasts was also applicable to MHC, so it appears that the newly synthesized myosins are assembled into myofilaments at or near the site of synthesis with limited subsequent exchange. There have been few reports on the localization of polyribosomes in ordinary cardiac muscle. Cedergren and Harary2’ examined the ultrastructure of cultured cardiac myocytes of neonatal rats. They showed that long helical chain-like structures of polyribosomes were often located parallel and in close contact with the incomplete myofilaments. Larson et investigated the spatial relationship between polyribosomes and myosin filaments using developing and regenerating human skeletal muscle. They found that lines of ribosomes appeared along some of the myosin filaments and that the
polyribosomes appeared rarely, if ever, to extend into the pseudo-H region or into the I band. They suggested that the mRNA moves along the stationary ribosomes with the result that myosin subunits are formed in the exact locus where they will later function. Thornel131 found that myofilamentpolyribosome complexes occurred frequently in the conducting system but never in the ordinary muscles of mammalian hearts. He suggested that the polyribosomes, which were almost exclusively intermingled with thick filaments, produced myosin in situ. These reports indicate that polyribosomes are located heterogeneously in sarcomeres in close contact with myofilaments. This may explain, at least in part, the heterogeneity of the assembly pattern of the newly synthesized cardiac myosins at the subcellular level revealed in this study. ACKNOWLEDGEMENTS We wish to thank Mr Hideo Tsukamoto, Laboratory of Cell Biology, University of Tokai, for labelling the monoclonal antibody with enzyme. We also wish to thank Mr Shin-ichi Izumi, Laboratory of Cell Biology, University of Tokai, Mr Mitsuhiro Sakamoto, The Institute for Adult Diseases, Asahi Life Foundation, for taking the electron micrographs. REFERENCES 1. Everett, A. W., Prior, G. and Zak R (1981). Equilibration of leucine between the plasma compartment and leucyl-tRNA in the heart, and turnover of cardiac myosin heavy chain. Biochem. 3. 194, 365-368. 2. Wenderoth, M. P. and Eisenberg, B. R. (1987). Incorporation of nascent myosin heavy chains into thick filaments of cardiac myocytes in thyroid-treated rabbits. J . Cell Biol., 105, 2771-2780. 3. Lawrence, J. B. and Singer, R. H. (1986). Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell, 45, 407-415. 4. Yazaki, Y. and Raben, M. S. (1975). Effect of thyroid state on the enzymatic characteristic of cardiac myosin: a difference in behavior of rat and rabbit cardiac myosin. Circ. Res., 36, 208-215. 5. Yazaki, Y., Mochinaga, S. and Raben, M. S. (1973). Fraction of the light chains from rat and rabbit cardiac myosin. Biochem. Biophys. Acta, 328,464469. 6. Yazaki, Y., Tsuchimochi, H., Kuro-o, M., Kurabayashi, M., Isobe, M., Ueda, S., Nagai, R. and Takaku, F. (1984). Distribution of myosin isozymes in human atrial and ventricular myocardium: comparison in normal and overloaded heart. Eur. Heart J., 5 (Suppl. F), 103-1 10. 7. Tsuchimochi, H., Sugi, M., Kuro-o, M., Ueda, S., Takaku, F., Furuta, S., Shirai, T. and Yazaki, Y. (1984). Isozymic changes in myosin of human atrial myocardium induced
Y.SEKO E T A L .
130
8.
9.
10.
11. 12. 13.
14. 15.
16.
17.
18.
19.
by overload: immunohistochemical study using monoclonal antibodies. J. Clin. Invest., 74, 662-665. Khaw, B. A., Mattis, J. A., Melincoff, G., Straws, H. W., Gold, H. K. and Habor, E. (1984). Monoclonal antibody to cardiac myosin: imaging of experimental myocardial infraction. Hybridoma, 3, 11-23. Wilson, M. B. and Nakane, P. K. (1978). Recent developments in the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies. In Immunoyuorescence and Related Staining Techniques. Knapp, W., Holubar, K. and Wick, G., eds). Elsevier/North Holland Biomedical Press: Amsterdam, pp. 215-225. Taniguchi, J., Kokubun, S., Noma, A. and Irisawa, H. (1981). Spontaneously active cells isolated from the sinoatrial and atrio-ventricular nodes of the rabbit heart. Jpn. J . Physiol., 31, 547-558. Nakane, P. K. and Pierce, G. B. (1966). Enzyme-labelled antibodies: preparation and application for localization of antigens: J . Histochem. Cytochem., 14,929-931. Nakane, P. K. and Pierce, G. B. (1967). Enzyme-labelled antibodies for the light and electron microscopic localization of tissue antigens. 1. Cell Biol., 33, 307-318. Mercardier, J. J., Bouveret, P., Gorza, L., Schiaffino, S., Clark, W. A., Zak, R., Swynghedauw, B. and Schwartz, K. (1983). Myosin isoenzymes in normal and hypertrophied human ventricular myocardium. Circ. Rex, 53, 52-62. Kuro-o, M., Tsuchimochi, H., Ueda, S., Takaku, F. and Yazaki, Y. (1986). Distribution of cardiac myosin isozymes in human conduction system. J . Clin. Invest., 77, 340-347. Hoh, J. F. Y., McGrath, P. A. and Hale, P. J. (1978). Electrophoretic analysis of multiple forms of rat cardiac myosin: effect of hypophysectomy and thyroxine replacement. J . Mol. Cell Cardiol., 10, 1053-1076. Yazaki, Y. and Raben, M. S. (1974). Cardiac myosin adenosinetriphosphatase of rat and mouse: distinctive enzymatic properties compared with rabbit and dog cardiac myosin. Circ. Res., 35, 15-23. Yazaki, Y., Ueda, S., Nagai, R. and Shimada, K. (1979). Cardiac atrial myosin adenosine triphosphatase of animals and humans: distinctive enzymatic properties compared with cardiac ventricular myosin. Circ. Res., 45, 522-527. Chizzonite, R. A., Everett, A. W., Clark, W. A,, Jacovcic, S., Rabinowitz, M. and Zak, R. (1982). Isolation and characterization of two molecular variants of myosin heavy chain from rabbit ventricle: change in their content during normal growth and after treatment of thyroid hormone. J . Bid. Chem., 257,2056-2065. Gorza, L., Pauletto, P., Pessina, A. C., Sartore, S. and Schiaffino, S. (1981). Isomyosin distribution in normal and
20.
21.
22.
23.
24. 25. 26. 27.
28.
29. 30.
31.
pressure overload rat ventricular myocardium: an immunohistochemical study. Circ. Res., 49, 1003-1009. Lompre, A. M., Schwartz, L. K., d’Albis, A,, Lacmbe, G., Thiem, N. V. and Swynghedauw, B. (1979). Myosin isozyme redistribution in chronic heart overload. Nature (London), 282, 105-107. Nag, A. C., Cheng, M. and Zak, R. (1985). Distribution of isomyosin in cultured cardiac myocytes as determined by monoclonal antibodies and adenosine triphosphatase activity. Exp. Cell Rex, 158, 53-62. Mahdavi, V., Chambers, A. P. and Nadal-Ginard, B. (1984). Cardiac a- and p-myosin heavy chain genes are organized in tandem. Proc. Natl. Acad. Sci. U.S.A., 81, 2626-2630. Lompre, A. M., Nadal-Ginard, B. and Mahdavi, V. (1984). Expression of the cardiac ventricular a- and p-myosin heavy chain genes is developmentally and hormonally regulated. J . Biol. Chem., 259, 6437-6446. Huxley, H. E. (1963). Electron microscopic studies on the structures of natural and synthetic protein filaments from striated muscle. J . Mol. Biol., 7, 281-308. Saad, A. D., Fischman, D. A. and Pardee, J. D. (1986). Fluorescence energy transfer studies of myosin thick filament assembly. Biophys. J., 49, 140-142. Saad, A. D., Pardee, J. D. and Fischman, D. A. (1986). Dynamic exchange of myosin molecules between thick filaments. Proc. Natl. Acad. Sci. U.S.A., 83, 9483-9487. Fulton, A. B., Wan, K. M. and Penman, S. (1980). The spatial distribution of polyribosomes in 3T3 cells and the associated assembly of proteins into the skeletal framework. Cell, 20, 849-857. Isaacs, W. B. and Fulton, A. B. (1987). Cotranslational assembly of myosin heavy chain in developing cultured skeletal muscle. Proc. Natl. Acad. Sci. U.S.A., 84, 6174-6178. Cedergren, B. and Harary, I. (1964). In vitro studies on single beating rat heart cells. VI. Electron microscopic studies of single cells. J . Ultrastruct. Res., 11, 428-442. Larson, P. F., Hudgson, P. and Walton, J. N. (1969). Morphological relationship and polyribosomes and myosin filaments in developing and regenerating skeletal muscle. Nature (London), 222, 1168-1169. Thornell, L. E. (1972). Myofilament-polyribosome complexes in the conducting system of hearts from cow, rabbit, and cat. J . Ultrastruct. Res., 41, 579-596.
Received 9 October 1989 Accepted 20 November 1989
8: 117-130 (1990)
Intracellular Assembly of Newly Synthesized Canine Cardiac Myosins YOSHINORI SEKOt, SHIGETO NAITO?, KOUJI IMATAKA?, JUN FUJIIt, PAUL K. NAKANES, FUMIMARO TAKAKU6 AND YOSHIO YAZAKIG ?The Institute for Adult Diseases, Asahi Life Foundation, Shinjuku-ku, Tokyo, fDepartment of Cell Biology, School of Medicine, University of Tokai, Isehara, Kanagawa and $Third Department of Internal Medicine, Faculty of Medicine, University of Tokyo, Bunkyo-ku, Tokyo, Japan
To investigate how newly synthesized cardiac myosins are assembled into myofilaments, we analysed the distribution of newly produced a-myosin heavy chain isozyme in sarcomeres by immunoelectron microscopy using a monoclonal antibody (CMA19), which is specific for a-myosin heavy chain. Isozymic changes in myosin heavy chains from to a type were induced in canine ventricular muscles and cultured ventricular myocytes by administration of 1-thyroxine. We incubated the glycerinated ventricular muscles or cultured ventricular myocytes with the enzyme (horseradish peroxidase) labelled Fab fragment of CMA19. After the reaction with 3, 3’-diaminobenzidine and osmification, we prepared ultrathin sections of the ventricular muscles or cultured ventricular myocytes and analysed their staining patterns by electron microscopy. There was apparent heterogeneity in the staining intensity of the myofilaments among different cells, among different myofibrils and even intramyofibrillarly. Higher magnification revealed that there were scattered foci of strong reaction which appeared to be foci of assembly of the newly synthesized a-myosin heavy chain. Immunocytochemical study also showed heterogeneous reactions within myofilaments and that there were scattered foci of myofilament assembly, which were closely associated with polyribosomes producing newly induced a-myosin heavy chain. These data suggest that newly synthesized cardiac myosins are assembled into myofilaments from the sites of synthesis, that is polyribosomes. This may explain the heterogeneity of the assembly pattern of newly synthesized cardiac myosins at the subcellular level. KEY WORDS -myosin
heavy chain isozyme; cultured cardiac myocyte; thyroxine; immunoelectron microscopy; enzyme-labelled antibody; monoclonal antibody; myofilament; polyribosome.
ELISA, enzyme-linked immunosorbent assay; MHC, myosin heavy chain; PBS, phosphate buffered saline; DMEM, Dulbecco’s modified Eagle’s minimum essential medium.
ABBREVIATIONS-
INTRODUCTION The metabolic rate of cardiac muscles, which perform continuously repetitive contraction, appears to be much higher than that of skeletal muscles. Cardiac myosin, one of major contractile proteins, is turned over very rapidly with a half-life of 5-6 This work was partly presented at the 59th Scientific Session of the American Heart Association (1986, Circulation 74, 11-82. [Abstr.]). Supported by grants-in-aid for scientific research (Nos 60480228, 60570382) from the Ministry of Education, Science and Culture and a grant on calcium signal transduction in cardiovascular system, Japan. Addressee for correspondence: Yoshinori Seko, Third Department of Internal Medicine, Faculty of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan. 0263-6484/90/020117- 14 S07.00 @ 1990 by John Wiley & Sons, Ltd.
days.’ Thus the attachment and detachment of myosin molecules must occur rapidly and extensively in myofilaments without interrupting forcegeneration. Although many researchers have studied the mechanism involved in assembly and disassembly of myosin molecules using in vitro systems, until recently, there has been no evidence showing how newly synthesized cardiac myosins are assembled into myofilaments in living cardiac myocytes. Recently, Wenderoth and Eisenberg’ showed that, by manipulating the thyroid hormone level, newly induced a-type cardiac myosin heavy chain (MHC) was incorporated into thick filaments with greater exchange at the ends than in the central
118 region in ventricular muscle of rabbits. However, they examined the distribution of newly induced aMHC in the ventricular muscle at a rather early stage of thyroid-treatment, in which newly induced a-MHC had not been fully incorporated into thick filaments. They speculated that a possible imbalance between the rates of synthesis and incorporation resulted in accumulation of a-MHC at the ends of the thick filaments. It appears that this imbalance resulted from the conditions of functioning cardiac muscle, where myofilaments are arranged densely and orderly for force-generation. Lawrence and Singer3 demonstrated that the intracellular localization of cytoskeletal proteins might reflect that of corresponding mRNAs, which were bound and translated by polyribosomes. However, it is much easier to study the spatial relationship between thick filaments and polyribosomes, which are producing MHC, in cultured cardiac myocytes than in intact functioning cardiac muscle. The purposes of this study are to investigate how newly synthesized cardiac myosins are assembled into myofilaments and to study the spatial relationship between their sites of synthesis, that is polyribosomes, and their sites of function. For these purposes, we induced isozymic changes in MHC from B to a type in canine ventricular muscles and cultured canine ventricular myocytes by administration of 1-thyroxine, and analysed the distributions of newly induced u-MHC in sarcomeres and polyribosomes producing a-MHC by immunoelectron microscopy. To enable the anti-MHC antibody to reach all parts of the intracellular structures, we used a Fab fragment of a monoclonal antibody (CMA19), which is specific for a-MHC, directly labelled with horseradish peroxidase. Our data suggest that newly synthesized a-MHC are assembled into myofilaments heterogeneously and near the sites of synthesis.
Y. SEKO ETAL.
peritoneal injections at two-week intervals with 0.1-0.2 ml of 1 mg ml-’ solutions of calf atrial myosin as a-MHC antigen. Mice were killed three days after the intravenous booster injection, and their isolated spleen cells were fused with the myeloma cells (P3X63Ag8UI). Anti-myosin activity in a medium from hybridoma colonies was screened by enzyme-linked immunosorbent assay (ELISA) using goat biotinylated anti-mouse immunogloblin and avidin D-peroxidase (Vector Laboratories Inc. Burlingame, CA), as described p r e v i~ u sly .The ~ antibody CMAl9 also reacted specifically with atrial myosin in the dog, and was selected for use in this study. To enable the anti-myosin antibody to reach all parts of the cardiac tissue, even to the level of myofilaments, we cut the antibody into Fab fragments by papain digestion, as described elsewhere.’ ELISA test showed the preservation of strong reactivity of the Fab fragment of CMA19 (CMA19Fab) with atrial myosin (Figure 1). Hor-
CMA 19Fab
E
e
0 Ventricular myosin
Ln V
Light chains ( I
d 0
MATERIALS AND METHODS
0 Atrial myosin
C
0
50i 1
+ 11)
\
Preparation of Monoclonal Antibody SpeciJc for aMHC
Myosins for immunization were prepared from calf atrial muscles by a dilution technique, as previously d e ~ c r i b e d . The ~ light chains were isolated by guanidine denaturation, as described el~ewhere.~ The preparations of hybridoma producing anti-myosin antibody specific for a-MHC have previously been de~cribed.~.’ Briefly, BALB/c mice (male, 6 weeks) were immunized by five intra-
Figure 1. Reactions of monoclonal antibody (CMA19Fab) with canine myosin by enzyme-linked immunosorbent assays. Results expressed as the percentage of maximum optical density at 540nm show specific reaction of CMA19Fab with atrial myosin ( O ) ,negligible reaction with ventricular myosin (O), and no reaction with light chains (0).
119
ASSEMBLY OF CARDIAC MYOSINS
seradish peroxidase was used for labelling of the monoclonal antibody because of its small molecular weight and availability for both light and electron microscopy. We labelled the monoclonal antibody (CMA19Fab) directly with horseradish peroxidase by the method of Wilson and Nakane.' Preparation of Cardiac Tissue Sections
Mongrel dogs weighing 8-15 kg were injected with 1-thyroxine (daily dose 300 pg kg-' subcutaneously) for 6 and 21 days. Saline was injected into the control group in the same way. Each dog was killed with potassium chloride (administered intravenously), and fresh heart muscle tissue was dissected from the left ventricular free walls in both groups and left atria in control group. After glycerination for more than one month at -2O"C, the tissues were washed with phosphate buffered saline (PBS; 0.01 M, pH 7.2) for 3 h (two changes hr-') at 4°C and cut into small blocks, then 4 pm sections were prepared by a cryostat for immunostaining. Preparation of Cultured Cardiac Myocytes
of them was kept at 37°C during Langendorff perfusion. Then, the heart was incubated in high K-low C1 solution at 4°C for more than 1 h, and was cut into small pieces of atria and ventricles. The isolation of atrial or ventricular myocytes was done simply by stirring the tissue piece in serum-free Dulbecco's modified Eagle's minimum essetial medium (DMEM) containing insulin-transferrin-sodium selenite media supplement (Sigma), penicillin and streptomycin. After the cells were cultured on slides under a humidified atmosphere of 5 per cent CO, and 95 per cent air at 37°C overnight, when most of them attached to the surfaces of slides, the culture media of ventricular and atrial myocytes were replaced with new media containing l-thyroxine (100 nM) for the thyroxine-treated group and the positive control group, respectively. The culture media of ventricular myocytes were also replaced with new media without 1-thyroxine for the negative control group. The cells were cultured in the same condition for 7 days, and were used for immunostaining. Immunostaining for Light and Electron Microscopy
The procedures used for isolating single cardiac The procedures used in immunostaining for light myocytes were essentially as described previously'o and were held aseptically. Neonatal mongrel dogs and electron microscopy by the enzyme-labelled of 1- 14 days old, weighing 200-600 g, were anesthe- antibody method were essentially those described tized with pentobarbital (25 mg kg-' body weight, by Nakane and Pierce."*12 Briefly, the cryostat administered intraperitoneally). The chest was sections were air-dried for 30 min at room temperaopened and the right ventricle was injected with ture and fixed with 100 per cent acetone for 60 min at 4°C. Washing was done three times with PBS for Ca-free Tyrode solution containing in r n ~ NaCl , 136.9, NaHCO, 11.9, KCl 5.4, MgCl, 0.53, 15 min. The sections were then treated with 1 per NaH,PO, 0.33, HEPES 5.0 and glucose 5.5 (pH cent Triton X-100 (Sigma) in PBS for 5 rnin at 7.3-7.4), to almost stop the contraction. The heart room temperature and washed three times with was then dissected out and the aorta was cannu- PBS for 15 min. After incubation with horseradish lated to perfuse the coronary arteries (Langendolf peroxidase-labelled CMA 19Fab overnight at room perfusion) with Ca-free Tyrode solution, and the temperature, the sections were washed three times heart was hung in a moist chamber where the with PBS for 15 min, then fixed with 1 per cent perfusion was continued at hydrostatic pressure of glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) 65 cm. After the blood was washed out, the perfu- for 5 min. They were washed with PBS for 5 min, sate was replaced with Ca-free Tyrode solution and reacted with 0.05 M Tris-HC1 buffer (pH 7-6) containing 0-45 mg ml- collagenase (Type I, containing 0.02 per cent (wt/vol) 3,3'-diaminobenSigma Chemical Co., St. Louis, MO), which was zidine4HCl. The reaction was carried out first with recirculated by a peristaltic pump. The perfusion of 1 per cent dimethylsulfoxide for 1 h for preincubacollagenase was continued for 30 min, then washing tion, and then with H,O, (17p1 of a 30 per cent out was done with high K-low C1 solution contain- H,O, solution per 100 ml of 0.05 M Tris-HC1 buffing in mM, taurine 10, oxalic acid 10, glutamic acid er) for 2 rnin at room temperature. For light 70, KCl 25, KH,PO, 10, HEPES 10, glucose 11 microscopy, the sections were washed with water, and EGTA 0-5 (pH 7.4). All perfusates were corn- dehydrated with ethanol and coverslips were pletely gassed with carbogen, consisting of 95 per mounted with resin in xylene. For electron microcent 0,, and 5 per cent CO,, and the temperature scopy, they were washed three times with PBS for
Figure 2. Light micrographs of canine cardiac muscle stained with CMA19Fab by the enzyme-labelled antibody method. (A) and (B) ventricular muscle of the thyroxine-treated group, for 6 day (A) and 21 day (B) treatment. (A) Weakly reactive myofibres distribute heterogeneously. (B) Many myofibres are clearly reactive. Note apparent heterogeneity in the reaction among different myofibres. Ventricular (C) and atrial (D) muscles of the saline-treated control group. (C) All myofibres are unreactive. (D) All myofibres are strongly and almost homogeneously reactive. Bar, 30 pm.
121
ASSEMBLY OF CARDIAC MYOSINS
15 min and treated with 2 per cent osmium tetroxide in 0.1 M phosphate buffer (pH 7.4) for 1 h. They were again washed with PBS, followed by ethanol dehydration and embedding in Epok 812 (Ouken Shoji Co., Tokyo, Japan) resin. Ultrathin sections were prepared with a diamond knife and examined in a JEOL JEM1200EX electron microscope. Cultured cardiac myocytes on slides were washed three times with PBS for 15 min, then fixed with 4 per cent paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4°C for 30 min and washed three times with PBS for 15 min. The subsequent procedures for immunostaining of a-MHC, without treatment with 1 per cent Triton X-100, were the same as for tissue samples. RESULTS Immunohistochemistry by Light Microscopy
As for the thyroxine-treated groups, many ventricular myofibres were weakly reactive with CMAl9Fab and there was heterogeneous distribution of
reactive myofibres in the 6 day-treated group (Figure 2A). But, most of the ventricular myofibres were clearly reactive and there was apparent heterogeneity in the reaction among different myofibres in the 21 day-treated group (Figure 2B). In the saline-treated control group, all except a few of the ventricular myofibres were unreactive (Figure 2C), all of the atrial myofibres were strongly and almost homogeneously reactive with CMA 19Fab (Figure 2D). Immunohistochemistry by Electron Microscopy
We examined the permeability of the enzymelabelled anti-myosin antibody in cardiac muscle using atrial muscle as a positive control. There were strong and almost homogeneous reactions with CMA 19Fab at the level of the sarcomere (Figure 3), showing that CMA 19Fab could penetrate throughout the cardiac tissue, even to the level of the myofilament. In the thyroxine-treated ventricles, the intensity of the reaction with CMA19Fab in the 6 day-
Figure 3. Electron micrograph of atrial muscle of the saline-treated control group stained with CMA19Fab. Strong and almost homogeneous reaction is attained throughout the tissue at the level of the sarcomere. Bar, 1 pm.
Figure 4. Electron micrographs of ventricular muscle of the thyroxine-treated (for 21 days) group stained with CMA19Fab. (A) Reactive and unreactive myofibres exist side by side. Note the heterogeneity in the reaction and foci of strong reaction within the reactive myofibril. Bar, 1 pm. (B) Reactive and unreactive cells are connected by an intercalated disc. The intensity of the reaction within the reactive cell decreases away from the intercalated disc. Bar, 1 pm.
123
ASSEMBLY OF CARDIAC MYOSINS
treated group was not strong enough for electron microscopic study. Therefore, we used a 21 daytreated group for the study. There was apparent heterogeneity in the reaction between adjacent myofibrils (Figure 4A), as well as between adjacent myofibres (Figure 4B). Such heterogeneous reaction existed even in the same reactive myofibre (Figure 4A). The distribution of reactive areas in the same myofibre did not appear to be random. Instead foci of strong intensity and a tendency of decreased intensity away from the foci were noted (Figure 4A). Almost homogeneous reaction was observed within the same sarcomere near an intercalated disc in the reactive myofibre (Figure 4B), showing that each myofilament of the sarcomere in this part of the myofibre had almost homogeneous distribution of a-MHC. Higher magnification (Figure 5A) revealed that there were scattered areas of strong reaction at the sarcomere level and that they appeared to be the foci of assembly of the newly synthesized a-MHC. Heterogeneous distribution of or-MHC was found even in the same myofilament in the course of isozymic redistribution. In contrast, there was no or negligible reaction in the ventricles of the salinetreated control group (Figure 5B), showing that there was minimal non-specific binding of CMA19Fab.
mainly consisted of a-MHC, were always closely associated with reactive polyribosomes. These scattered foci of reactive myofilaments and reactive polyribosomes, which means the foci of assembly of newly synthesized a-MHC, appeared to correspond to those seen in thyroxine treated ventricular muscle. Orderly arrangement of reactive myofilaments was seen at the tips of pseudopodia (Figure 8A, arrows). A similar pattern of the myofilament assembly was seen at the cell periphery where the myocytes touched an adjacent cell (Figure 8B). These special arrangements of myofilaments at the cell periphery appear to be related to cell motility, for which actin and myosin filaments cooperatively provide the major mechanical force. This is also supported by the findings about the intracellular localization of actin mRNA.3 Higher magnification of a focus of myofilament assembly is shown in Figure 9. Reactive and unreactive polyribosomes were closely associated with myofilaments, within which non-uniform reactions were noteworthy. The coincidental localization of reactive myofilaments and reactive polyribosomes strongly suggests that newly synthesized a-MHC were assembled into myofilaments from the nearest sites of synthesis. In this study, we could not identify any preferred region of the myofilaments for the assembly of newly synthesized a-MHC. DISCUSSION
Immunocytochemistry by Electron Microscopy
To confirm the permeability of enzyme-labelled CMA 19Fab and to evaluate the non-specific binding of it also in cultured cardiac myocytes, we examined the reactions of it with thyroxine-treated atrial myocytes and non-treated ventricular myocytes as positive and negative controls, respectively. We found again strong and almost homogeneous reactions in sarcomeres of thyroxine-treated atrial myocytes (Figure 6A) and no or negligible reaction in those of non-treated ventricular myocytes (Figure 6B). This indicates that technical artifacts due to diffusion of CMA19Fab or sectioning are minimal. We also found that most of the myofilaments were arranged in an orderly manner in both the control groups. The reaction patterns in thyroxine-treated ventricular myocytes are shown in Figures 7 and 8. As shown in Figure 7, there were scattered foci of myofilament assembly, where reactive and unreactive myofilaments were almost randomly mingled with one another. Reactive myofilaments, which
In this study, we clearly demonstrated a heterogeneous distribution of newly synthesized a-MHC among different cells, among different myofibrils, and even intramyofibrillarly. Further, there appeared to be scattered foci of assembly of newly synthesized a-MHC at the sarcomere level. Immunocytochemical study also showed that there were scattered foci of myofilament assembly, which were closely associated with polyribosomes producing newly induced a-MHC. These observations strongly suggest the close correlation between the intracellular sites of assembly and those of synthesis in myosin turnover. In human ventricular muscles, Mercardier et ~ 1 . ' have ~ quantified the amount of V1-type isomyosin which was found to range from almost 0 to 15 per cent (average, a little less than 5 per cent) of total myosin, and to vary from one heart to another. Recently, we demonstrated the distribution of ventricular myofibres containing a-MHC. Their number ranges from 0 to 15 per cent of the total myofibres but with regional ~ a r i a t i o n . 'In ~
Figure 5. Electron micrographs of ventricular muscle of the thyroxine-treated (for 21 days) group (A) and the saline-treated control group (B) stained with CMA19Fab. (A) Higher magnification shows scattered areas of strong reaction at the level of the sarcomere. Bar, 1 pm. (B) No or negligible reaction is observed in the saline-treated group. Bar, 1 pm.
125
ASSEMBLY OF CARDIAC MYOSINS
Figure 6. Electron micrographs of 7-day thyroxine-treated cultured atrial myocyte (A) and nontreated cultured ventricular myocyte (B) stained with CMA19Fab. (A) Strong and almost homogeneous reaction is attained throughout the sarcomeres. Bar, 1 pm. (B) N o or negligible reaction is observed. Bar, 1 urn. Note that most of the mvofilaments are arranged in an orderly manner in both (A) and (B).
this study, we found the proportion of ventricular myofibres, which were reactive with CMA19, to be from 0 to 10 per cent in the canine control group, which was about the same as in human. On the other hand, most of the ventricular myofibres reacted with CMA19Fab in the thyroxine treated group. Therefore, most of the a-MHC, which appeared after thyroxine treatment, must have been newly synthesized. Recent studies from our and other laboratories have revealed the presence of two isozymes in MHC, a and p types, and their relative proportion can be changed by aging, hormonal state and
pressure ~ v e r l o a d . ~ . 5-20 ~ . ~ . We ' and others have also reported the heterogeneity, at the cellular level, in the contents of the MHC isozymes as well as heterogeneous isozymic redistribution resulting from pressure overload or in response to thyroid h o r r n ~ n e . ~Mahdavi .~' et a1." identified two genes coding for two MHC isozymes and showed them to be organized in tandem. They showed that the isozymic changes in MHC during development or in response to thyroid hormone were regulated at Further, they speculated that thymRNA roid hormone may have a direct effect on transcription of the MHC genes and that one of the major
Figure 7. Electron micrograph of a 7-day thyroxine-treated cultured ventricular myocyte stained with CMA19Fab. Several foci of myofilament assembly are scattered within the cell. Note that reactive and unreactive myofilaments are randomly mingled with one another and are closely associated with reactive and unreactive polyribosomes (arrow). Bar, 1 jim.
Figure 8. Electron micrographs of 7-day thyroxine-treated cultured ventricular myocytes stained with CMA19Fab. (A) Arrows indicate orderly arrangement of reactive myofilaments at the tips of pseudopodia. Bar, 1 pm. (B) Note the orderly arrangement of reactive myofilaments at the cell periphery where the myocyte touches an adjacent cell. Bar, 1 pm.
128
Y.SEKO ETAL.
Figure 9. Higher magnification of a focus of myofilament assembly in a 7-day thyroxine-treated cultured ventricular myocyte stained with CMA19Fab. Many myofilaments show non-uniform reactions within themselves and are closely associated with reactive and unreactive polyribosomes. Arrows indicate Z-discs. Bar, 1 pm.
reasons .for the heterogeneity in the expression of MHC genes at the cellular level may be either quantitative or qualitative differences in thyroid hormone receptors in the nuclei. Until recently, no previous study has evidenced the heterogeneity in the distribution of newly synthesized myosins at the subcellular level. But, there is some information about the mechanisms of intracellular protein synthesis and assembly, which may be of great use in understanding the background of this heterogeneity. HuxleyZ4 first deduced from electron micrographs of negatively stained filaments that the thick filaments of striated muscles were remarkable bipolar structures in which myosin molecules were systematically ordered with their tails in the backbone and heads arranged along the surface. He also demonstrated that native thick filaments were dissociated in high salt into their constituent myosin molecules and again associated to form thick filaments in low salt, suggesting that the packing forces holding these structures together were mainly ionic
in nature. Saad et al.25,26demonstrated, using a fluorescence energy transfer system, that there were rapid and extensive exchanges of myosin molecules in thick filaments as well as between thick filaments and soluble myosins in uitro. They also suggested that this exchange mechanism might be a possible in-viuo mechanism for myosin isoform transitions during muscle development and myosin turnover in thick filaments. Recently, Wenderoth and Eisenberg’ studied the distribution of newly induced a-MHC in the left ventricular muscle of thyroid-treated rabbits and found that newly induced a-MHC was incorporated into thick filaments with greater exchange at the ends than in the central region of the thick filaments. They speculated that during periods of rapid protein synthesis the rates of synthesis of aMHC might be greater than the rates of incorporation of a-MHC into the thick filaments and that this imbalance would result in accumulation of new myosin at the ends of the thick filaments. They examined the distribution of a-MHC in the ventri-
129
ASSEMBLY OF CARDIAC MYOSINS
cular muscle treated with thyroid hormone for up to 4 days. This seems to be a rather early stage of thyroid-treatment to study the distribution of the incorporated a-MHC. It appears that the densely and orderly arrangement of myofilaments in functioning cardiac muscle may be one of the causes of the imbalance between the rates of synthesis and incorporation. Our present study showed that there were scattered foci of assembly of newly induced aMHC in ventricular muscle treated with thyroxine for 21 days. At this stage, it seems that most of the newly synthesized myosins were assembled into myofilaments and a balance had been reached between the rates of synthesis and of incorporation. Scattered foci of assembly of newly induced aMHC were also observed in thyroxine-treated cultured ventricular myocytes and appeared to correspond to those seen in the ventricular muscle. They were closely associated with polyribosomes producing a-MHC, suggesting a close spatial relationship between sites of synthesis and sites of assembly. Fulton et L Z ~ . ,using ~’ cultured fibroblasts, found a relationship between the distribution of polyribosomes and an associated assembly of proteins into intact cells and skeletal frameworks. Although their data were from cultured fibroblasts, the results seem highly suggestive as a model of intracellular protein assembly. They showed that newly synthesized skeletal proteins were assembled into the skeletal frameworks at or very near their site of synthesis (that is, polyribosomes) with limited subsequent exchange between the skeletal frameworks and soluble proteins. Recently, they reported cotranslational assembly of MHC in developing cultured skeletal muscle.*’ They suggested that the mode of assembly of skeletal proteins in cultured fibroblasts was also applicable to MHC, so it appears that the newly synthesized myosins are assembled into myofilaments at or near the site of synthesis with limited subsequent exchange. There have been few reports on the localization of polyribosomes in ordinary cardiac muscle. Cedergren and Harary2’ examined the ultrastructure of cultured cardiac myocytes of neonatal rats. They showed that long helical chain-like structures of polyribosomes were often located parallel and in close contact with the incomplete myofilaments. Larson et investigated the spatial relationship between polyribosomes and myosin filaments using developing and regenerating human skeletal muscle. They found that lines of ribosomes appeared along some of the myosin filaments and that the
polyribosomes appeared rarely, if ever, to extend into the pseudo-H region or into the I band. They suggested that the mRNA moves along the stationary ribosomes with the result that myosin subunits are formed in the exact locus where they will later function. Thornel131 found that myofilamentpolyribosome complexes occurred frequently in the conducting system but never in the ordinary muscles of mammalian hearts. He suggested that the polyribosomes, which were almost exclusively intermingled with thick filaments, produced myosin in situ. These reports indicate that polyribosomes are located heterogeneously in sarcomeres in close contact with myofilaments. This may explain, at least in part, the heterogeneity of the assembly pattern of the newly synthesized cardiac myosins at the subcellular level revealed in this study. ACKNOWLEDGEMENTS We wish to thank Mr Hideo Tsukamoto, Laboratory of Cell Biology, University of Tokai, for labelling the monoclonal antibody with enzyme. We also wish to thank Mr Shin-ichi Izumi, Laboratory of Cell Biology, University of Tokai, Mr Mitsuhiro Sakamoto, The Institute for Adult Diseases, Asahi Life Foundation, for taking the electron micrographs. REFERENCES 1. Everett, A. W., Prior, G. and Zak R (1981). Equilibration of leucine between the plasma compartment and leucyl-tRNA in the heart, and turnover of cardiac myosin heavy chain. Biochem. 3. 194, 365-368. 2. Wenderoth, M. P. and Eisenberg, B. R. (1987). Incorporation of nascent myosin heavy chains into thick filaments of cardiac myocytes in thyroid-treated rabbits. J . Cell Biol., 105, 2771-2780. 3. Lawrence, J. B. and Singer, R. H. (1986). Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell, 45, 407-415. 4. Yazaki, Y. and Raben, M. S. (1975). Effect of thyroid state on the enzymatic characteristic of cardiac myosin: a difference in behavior of rat and rabbit cardiac myosin. Circ. Res., 36, 208-215. 5. Yazaki, Y., Mochinaga, S. and Raben, M. S. (1973). Fraction of the light chains from rat and rabbit cardiac myosin. Biochem. Biophys. Acta, 328,464469. 6. Yazaki, Y., Tsuchimochi, H., Kuro-o, M., Kurabayashi, M., Isobe, M., Ueda, S., Nagai, R. and Takaku, F. (1984). Distribution of myosin isozymes in human atrial and ventricular myocardium: comparison in normal and overloaded heart. Eur. Heart J., 5 (Suppl. F), 103-1 10. 7. Tsuchimochi, H., Sugi, M., Kuro-o, M., Ueda, S., Takaku, F., Furuta, S., Shirai, T. and Yazaki, Y. (1984). Isozymic changes in myosin of human atrial myocardium induced
Y.SEKO E T A L .
130
8.
9.
10.
11. 12. 13.
14. 15.
16.
17.
18.
19.
by overload: immunohistochemical study using monoclonal antibodies. J. Clin. Invest., 74, 662-665. Khaw, B. A., Mattis, J. A., Melincoff, G., Straws, H. W., Gold, H. K. and Habor, E. (1984). Monoclonal antibody to cardiac myosin: imaging of experimental myocardial infraction. Hybridoma, 3, 11-23. Wilson, M. B. and Nakane, P. K. (1978). Recent developments in the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies. In Immunoyuorescence and Related Staining Techniques. Knapp, W., Holubar, K. and Wick, G., eds). Elsevier/North Holland Biomedical Press: Amsterdam, pp. 215-225. Taniguchi, J., Kokubun, S., Noma, A. and Irisawa, H. (1981). Spontaneously active cells isolated from the sinoatrial and atrio-ventricular nodes of the rabbit heart. Jpn. J . Physiol., 31, 547-558. Nakane, P. K. and Pierce, G. B. (1966). Enzyme-labelled antibodies: preparation and application for localization of antigens: J . Histochem. Cytochem., 14,929-931. Nakane, P. K. and Pierce, G. B. (1967). Enzyme-labelled antibodies for the light and electron microscopic localization of tissue antigens. 1. Cell Biol., 33, 307-318. Mercardier, J. J., Bouveret, P., Gorza, L., Schiaffino, S., Clark, W. A., Zak, R., Swynghedauw, B. and Schwartz, K. (1983). Myosin isoenzymes in normal and hypertrophied human ventricular myocardium. Circ. Rex, 53, 52-62. Kuro-o, M., Tsuchimochi, H., Ueda, S., Takaku, F. and Yazaki, Y. (1986). Distribution of cardiac myosin isozymes in human conduction system. J . Clin. Invest., 77, 340-347. Hoh, J. F. Y., McGrath, P. A. and Hale, P. J. (1978). Electrophoretic analysis of multiple forms of rat cardiac myosin: effect of hypophysectomy and thyroxine replacement. J . Mol. Cell Cardiol., 10, 1053-1076. Yazaki, Y. and Raben, M. S. (1974). Cardiac myosin adenosinetriphosphatase of rat and mouse: distinctive enzymatic properties compared with rabbit and dog cardiac myosin. Circ. Res., 35, 15-23. Yazaki, Y., Ueda, S., Nagai, R. and Shimada, K. (1979). Cardiac atrial myosin adenosine triphosphatase of animals and humans: distinctive enzymatic properties compared with cardiac ventricular myosin. Circ. Res., 45, 522-527. Chizzonite, R. A., Everett, A. W., Clark, W. A,, Jacovcic, S., Rabinowitz, M. and Zak, R. (1982). Isolation and characterization of two molecular variants of myosin heavy chain from rabbit ventricle: change in their content during normal growth and after treatment of thyroid hormone. J . Bid. Chem., 257,2056-2065. Gorza, L., Pauletto, P., Pessina, A. C., Sartore, S. and Schiaffino, S. (1981). Isomyosin distribution in normal and
20.
21.
22.
23.
24. 25. 26. 27.
28.
29. 30.
31.
pressure overload rat ventricular myocardium: an immunohistochemical study. Circ. Res., 49, 1003-1009. Lompre, A. M., Schwartz, L. K., d’Albis, A,, Lacmbe, G., Thiem, N. V. and Swynghedauw, B. (1979). Myosin isozyme redistribution in chronic heart overload. Nature (London), 282, 105-107. Nag, A. C., Cheng, M. and Zak, R. (1985). Distribution of isomyosin in cultured cardiac myocytes as determined by monoclonal antibodies and adenosine triphosphatase activity. Exp. Cell Rex, 158, 53-62. Mahdavi, V., Chambers, A. P. and Nadal-Ginard, B. (1984). Cardiac a- and p-myosin heavy chain genes are organized in tandem. Proc. Natl. Acad. Sci. U.S.A., 81, 2626-2630. Lompre, A. M., Nadal-Ginard, B. and Mahdavi, V. (1984). Expression of the cardiac ventricular a- and p-myosin heavy chain genes is developmentally and hormonally regulated. J . Biol. Chem., 259, 6437-6446. Huxley, H. E. (1963). Electron microscopic studies on the structures of natural and synthetic protein filaments from striated muscle. J . Mol. Biol., 7, 281-308. Saad, A. D., Fischman, D. A. and Pardee, J. D. (1986). Fluorescence energy transfer studies of myosin thick filament assembly. Biophys. J., 49, 140-142. Saad, A. D., Pardee, J. D. and Fischman, D. A. (1986). Dynamic exchange of myosin molecules between thick filaments. Proc. Natl. Acad. Sci. U.S.A., 83, 9483-9487. Fulton, A. B., Wan, K. M. and Penman, S. (1980). The spatial distribution of polyribosomes in 3T3 cells and the associated assembly of proteins into the skeletal framework. Cell, 20, 849-857. Isaacs, W. B. and Fulton, A. B. (1987). Cotranslational assembly of myosin heavy chain in developing cultured skeletal muscle. Proc. Natl. Acad. Sci. U.S.A., 84, 6174-6178. Cedergren, B. and Harary, I. (1964). In vitro studies on single beating rat heart cells. VI. Electron microscopic studies of single cells. J . Ultrastruct. Res., 11, 428-442. Larson, P. F., Hudgson, P. and Walton, J. N. (1969). Morphological relationship and polyribosomes and myosin filaments in developing and regenerating skeletal muscle. Nature (London), 222, 1168-1169. Thornell, L. E. (1972). Myofilament-polyribosome complexes in the conducting system of hearts from cow, rabbit, and cat. J . Ultrastruct. Res., 41, 579-596.
Received 9 October 1989 Accepted 20 November 1989
