Fibre type-specific expression patterns of myosin heavy chain genes in adult torafugu Takifugu rubripes muscles

Myosin is the most abundant protein in skeletal muscles, being the primary component of thick filaments and involved in muscle contraction as well as phagocytosis, cell motility and vesicle transport in other organs of animals. The myosin superfamily consists of 18 classes of ATP-dependent motor proteins, although the best studied myosins belong to class II, which include sarcomeric and smooth muscle and non-muscle myosins (Berg et al., 2001; Foth et al., 2006). Class II myosin is a hexameric protein which consists of two heavy chains (MYHs) and four light chains. MYH possesses a globular head in its N-terminal region called subfragment-1 (S1) and a rod at the C-terminal half containing subfragment-2 (S2) and light meromyosin (LMM). S2 links S1 at its N-terminal end to LMM at its C-terminal end, and is believed to be loosely bound to the thick filament surface (Harrington and Rodgers, 1984).

Morphological, functional and metabolic characteristics of vertebrate skeletal muscles are mainly related to the cellular expression of different MYHs and muscle fibre types are characterized mainly based on the MYHs they contain (Cobos et al., 2001). On the basis of specific MYHs, mammalian myofibres are classified into types I, IIa, IIb and IId/x, with types I and IIa exhibiting oxidative metabolism, and with types IIb and IId/x being primarily glycolytic (Schiaffino and Reggiani, 1996). Type I fibres are also termed slow fibres because of slow contraction owing to low ATPase activity associated with type I MYH, whereas type II fibres, termed fast fibres, exert quick contraction and fatigue rapidly (Bassel-Duby and Olson, 2006).

In contrast to other vertebrate muscles in which different types of muscle fibres form a mosaic within anatomically the same muscle, fast and slow muscles of fish are generally separated into anatomically distinct areas. Fast muscle comprises the majority of trunk muscle, whereas slow muscle is found in a narrow mid-lateral layer just under the skin called lateralis superficialis (LS). In addition to LS slow muscle, another type of slow muscle is located at the median fins in fish belonging to the order Tetradontiformes, termed erector and depressor muscle (ED) (Winterbottom, 1974). The erectors lie superficially to the depressors and connect the anterior faces of fin ray bases to the front of the pterygiophore, the bone between vertebral spines that provides support to dorsal and anal fins.

The torafugu Takifugu rubripes Abe 1949 genome has been proposed as a model for rapid characterization of vertebrate genes due to its compact size (Brenner et al., 1993). Recently, we have investigated the genomic organization of sarcomeric and non-sarcomeric MYH genes of torafugu by in silico analysis of the total genome database, revealing that it contains 20 sarcomeric MYH genes which are organized in different clusters (Watabe and Ikeda, 2006; Ikeda et al., 2007). Common carp Cyprinus carpio is also known to possess a highly conserved MYH multigene family, although MYH gene types are much more numerous than their higher vertebrate counterparts (Gerlach et al., 1990; Kikuchi et al., 1999; Ikeda et al., 2007), probably because teleosts have undergone an additional round of genome duplication. The question, then, is why so many genes are contained in fish, even in torafugu with its small sized genome.

In the present study, the transcripts of seven sarcomeric MYH genes were identified in fast, slow and cardiac muscles of adult torafugu. While the expression of three MYH genes was restricted to fast, slow or cardiac muscles, two MYH genes were found to be expressed in both fast and slow fibres with different diameters. The functional significance of different MYH genes is discussed.

One live male specimen of torafugu (body mass about 1 kg) raised at the University of Tokyo was humanely killed and fast, slow and cardiac muscles were collected. The muscles were immediately frozen in liquid nitrogen and stored at –80°C until use for the construction of cDNA libraries, cDNA cloning and northern blot analysis. Slow muscles were dissected from two locations that included LS beneath the lateral surface of the myotome and ED for dorsal fin (Fig. 1A,B). Other laboratory-reared specimens of adult torafugu (sex unknown) were treated as above and used for histochemical analyses (body mass 290 g), for in situ hybridization (275 g), and for myosin preparation (663 g). Fast muscle from the dorsal trunk region of wild (N=3) and farm-cultured (N=3) torafugu females (body mass 0.8–1 kg) was also used for northern blot analysis.

Fast, LS and ED slow muscles were dissected from a fresh specimen of adult laboratory-reared torafugu (body mass 290 g). Muscles were snap-frozen by cooled isopentane and sections prepared by a cryostat were air-dried and stained to determine acid and alkaline stabilities for myofibrillar ATPase by the method of Johnston and colleagues (Johnston et al., 1974), after a brief preincubation (2–3 min) at pH 4.3, 4.6, 9.4, 10.0 and 10.6. NADH-diaphorase (NADH-tetrazolium reductase) staining was performed to determine glycolytic or oxidative metabolism by the method of Novikoff and colleagues (Novikoff et al., 1961).

Total RNA was isolated from various muscles using an ISOGEN solution (Nippon Gene, Tokyo, Japan). Total RNA (5 μg) was reverse-transcribed using oligonucleotide (dT)-tailed primer (supplementary material Table S1) and Superscript™ III Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. PCR amplification of MYH genes was performed using a GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, USA) with a forward primer (MYH-F) and a reverse primer (MYH-R) (supplementary material Table S1) that were designed based on a highly conserved amino acid sequence from the LMM region of sarcomeric MYHs from teleost. Amplified DNA fragments were subcloned into the pGEM-T vector (Promega, Madison, WI, USA) and 50 cDNA clones each of various muscle types were randomly sequenced with an ABI 3100 genetic analyzer (Applied Biosystems) after labelling with ABI PRISM BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems).

3′ RACE-PCR was performed to determine nucleotide sequences in the 3′ untranslated region (UTR) from each cloned MYH gene using a specific forward primer and an adapter primer (AP) (supplementary material Table S1). PCR amplifications were carried out at 94°C for 3 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s and extension at 72°C for 1 min, with a final extension step for 5 min. A single PCR reaction of 20 μl contained 1 μl each of forward and reverse primers (10 μmol 1^–1), 1 μl of cDNA template (dilution 1:100), 2 μl of 10× PCR buffer (20 mmol 1^–1 Tris-HCl pH 8.0, 100 mmol 1^–1 KCl, 20 mmol 1^–1 MgCl₂), 1 U Taq DNA polymerase and 13.8 μl of sterilized water. The amplified cDNA fragments were sequenced as described above.

The full-length cDNA encoding MYH_M86-1 was amplified using a degenerate forward primer (MYH-F1) together with a specific reverse primer (M86-1R) (supplementary material Table S1) with PrimeSTAR™ Max DNA polymerase (Takara, Otsu, Japan). PCR amplifications were carried out as described above. The entire sequence was determined by the primer walking method using designed internal forward (IFPs) and reverse primers (IRPs) (supplementary material Table S1). To determine the translation start site and 5′ UTR sequence for the full-length cDNA, 5′ RACE was performed using a GeneRacer™ kit (Invitrogen).

For DNA dot blot analysis, digoxygenin (DIG)-labelled probes were prepared by PCR for each MYH gene in the highly variable 3′ UTR region using a PCR-DIG Probe Synthesis Kit (Roche Applied Science, Mannheim, Germany) according to the manufacturer's instructions. The primers used for the preparation of DIG-labelled probes for each MYH are listed in supplementary material Table S1.

For northern blot analysis, 5 μg total RNA was prepared from three different parts of each of fast, slow and cardiac muscles from the same specimen used for cDNA cloning or from fast muscles of wild (N=3) and farm-cultured (N=3) females. These RNAs were electrophoresed on 2% agarose gels containing 2% formaldehyde and transferred to a nylon membrane (Biodyne®PLUS, Pall Corporation, Pensacola, FL, USA), and subjected to hybridization with the DIG-labelled DNA probes described above. DIG-labelled DNA probes specific to the 18S rRNA (AB437876) and α-actin (U38958) genes were used as internal controls (supplementary material Table S1).

Statistical analysis was carried out using Student's t-test to compare mRNA levels of MYH genes among fast, slow and cardiac muscles.

The deduced amino acid sequences of torafugu sarcomeric MYH genes cloned in this study were aligned using the multiple sequence alignment program CLUSTAL X (Thompson et al., 1997) along with corresponding amino acid sequences of MYHs from common carp, zebrafish Danio rerio and medaka Oryzias latipes. The phylogenetic tree was constructed by the neighbour-joining method in MEGA4 (Tamura et al., 2007) with zebrafish smooth muscle MYH as an outgroup. Bootstrap sampling analysis from 1000 replicates was adopted to evaluate internal branches.

cDNA fragments of 250–300 bp containing 3′ UTR nucleotide sequences and those encoding a C-terminal part of MYH were cloned into pGEM-T vector (Promega) (supplementary material Table S1). DIG-labelled antisense riboprobes were synthesized using T7 and SP6 RNA polymerase (Roche Applied Science). In situ hybridization was performed on transverse sections (12 μm thickness) made from adult fast and slow muscles of one laboratory-cultured torafugu (body mass 275 g) by the method of Darby and colleagues (Darby et al., 2006). Alkaline phosphatase-conjugated anti-DIG antibody (Roche Applied Science) and nitro blue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate were used for detection of riboprobes (Roche Applied Science).

Myosin was prepared from torafugu dorsal fast skeletal muscle by the method of Hwang and colleagues (Hwang et al., 1990). The chymotryptic fragments of myosin were fractionated according to Watabe and colleagues (Watabe et al., 1992), separated on SDS-PAGE and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The fragment bands were analyzed for their N-terminal amino acid sequences with an Applied Biosystems Procise 492HT protein sequencer.

ATPase was used to differentiate fibre types in fast and slow muscles of adult torafugu (Fig. 1). Fast muscle contained various fibres with different diameters (see Fig. 1D). ATPase of fast fibres with large diameters was inactivated after treatment at pH 4.6, whereas ATPase of those with small diameters was stable under this acidic pH treatment. It was noted that the stability of the small diameter fibres was more apparent the smaller the diameter. Meanwhile, most fibres in LS (Fig. 1E) and ED (Fig. 1F) slow muscle were resistant to pH 4.6, although some large sized fibres were found to be slightly acid labile. All fibres in fast muscle irrespective of their size were stained intensively for ATPase after incubation at pH 9.4, suggesting their stability at alkaline pH (data not shown). On the other hand, all fibres in fast and LS and ED slow muscles were found to be inactivated after treatment at pH 4.3, 10.0 and 10.6 (data not shown).

NADH-diaphorase staining was performed to identify oxidative fibres in skeletal muscles. All fibres in LS and ED slow muscles were stained (supplementary material Fig. S1). Importantly, fibres in LS slow muscle with large diameters adjacent to fast muscle showed a lower NADH-diaphorase reaction compared with those in a superficial region with small diameters, suggesting that the former fibres have an intermediate oxidative potential. However, none of fibres in fast muscle were stained.

Seven sarcomeric MYH genes were cloned from adult torafugu fast, slow and cardiac muscles by RT-PCR (Table 1). Three MYH genes, MYH_M86-1, MYH_M8248 and MYH_M880, were cloned exclusively from fast, slow and cardiac muscles dissected from a single laboratory-reared adult specimen, respectively, whereas two MYH genes, MYH_M2528-1 and MYH_M1034, were cloned from both fast and slow muscles and another two MYH genes, MYH_M2126-2 and MYH_M5, were cloned from both slow and cardiac muscles. The nomenclature of torafugu MYH genes found in the present study follows Ikeda and colleagues (Ikeda et al., 2007).

3′ RACE-PCR was used to determine nucleotide sequences at the UTR and those encoding a C-terminal region of the LMM domain for the above seven MYHs (supplementary material Fig. S2). These cDNA nucleotide sequences have been deposited in the DDBJ/EMBL/GenBank databases under accession numbers AB465004 (MYH_M86-1), AB465006 (MYH_M2528-1), AB465005 (MYH_M1034), AB465007 (MYH_M8248), AB465009 (MYH_M2126-2), AB465008 (MYH_M5) and AB465010 (MYH_M880).

Phylogenetic analysis was carried out to compare torafugu MYH genes with those from other fish such as common carp (Hirayama and Watabe, 1997; Imai et al., 1997; Nihei et al., 2006), zebrafish (Yelon et al., 1999; Bryson-Richardson et al., 2005) and medaka (Ono et al., 2006; Liang et al., 2007), whose expression is well known. MYH_M86-1, MYH_M2528-1 and MYH_M1034 belong to fast type as they were placed in the clade representing fast MYH genes (Nihei et al., 2006; Ikeda et al., 2007), whereas MYH_M8248 belongs to slow type and MYH_M2126-2 to cardiac type as they were monophyletic with slow- and cardiac-type MYH genes from other fish species, respectively (Fig. 2). The phylogenetic tree shows that fast-type MYH genes, MYH_M2528-1 and MYH_M1034, were placed in the same clade as fast-type MYHemb1 which is expressed in medaka from embryonic to adult stages (Ono et al., 2006), whereas MYH_M86-1 was placed in the same clade as fast-type mMYH-11 which is expressed in adult medaka acclimated to 30°C (Liang et al., 2007). MYH_M5 was a unique MYH as its phylogenetic position was independent of any other MYH genes, whereas MYH_M880 was monophyletic with zebrafish atrial MYH (Fig. 2). This phylogenetic relationship of torafugu MYH genes was in close agreement with that described by Ikeda and colleagues (Ikeda et al., 2007). MYH_M5 and MYH_M880 were found to have appeared in an early evolution of MYHs as far as the present phylogenetic tree was concerned, and these genes were regarded to belong to an ancestral slow/cardiac type (see Table 1).

Since the torafugu genome database did not contain the full-length sequence of MYH_M86-1, we determined the complete cDNA sequence of this gene, yielding a 5817 bp open reading frame (ORF) that encoded 1938 amino acid residues (supplementary material Fig. S3).

We randomly sequenced 50 cDNA clones each from the cDNA clone libraries constructed from fast, slow and cardiac muscles of adult torafugu. The clones encoding fast-type MYH_M86-1 were most abundant in fast muscle (Fig. 3A). Both LS and ED slow muscles showed an almost equal frequency of clones of five MYH genes including fast-type MYH_M2528-1 and MYH_M1034, slow-type MYH_M8248, cardiac-type MYH_M2126-2 and unique, slow/cardiac-type MYH_M5 (Fig. 3B,C). Of the three types of MYH clones from cardiac muscle, the type encoding cardiac-type MYH_M2126-2 was most abundant (Fig. 3D).

Fig. 4 shows hybridization patterns in northern blot analysis and mRNA levels relative to those of the 18S rRNA gene in three different parts of each of fast, slow and cardiac muscles from the same specimen used for the construction of cDNA clone libraries. Dot blot analysis revealed that all probes synthesized referring to fast, slow and cardiac-type MYH genes were highly specific and hybridized only with plasmids containing corresponding MYH (supplementary material Fig. S4). MYH_M880 did not show any hybridization signal, suggesting its expression was too low to be resolved by northern blot analysis. Fast-type MYH_M86-1 was expressed only in fast muscle (Fig. 4B), whereas MYH_M2528-1 and MYH_M1034 were expressed in fast as well as in LS and ED slow muscles. The relative mRNA levels of MYH_M2528-1 were significantly higher in fast than either of the two slow muscles (P<0.01; Fig. 4C). However, no significant difference in the mRNA levels was observed for MYH_M1034 in the fast and the two slow muscles (Fig. 4D).

The expression of slow-type MYH_M8248 was observed specifically in slow muscles and its mRNA levels were significantly higher in ED than in LS slow muscle (P<0.01; Fig. 4E), whereas cardiac-type MYH_M2126-2 was found to be expressed in LS and ED slow muscles, and cardiac muscles, with relative mRNA levels being significantly higher in LS slow than in ED slow and cardiac muscles (P<0.01; Fig. 4F). The unique, slow/cardiac-type MYH_M5 was expressed in the two slow muscles and mRNA levels in LS slow muscle were found to be significantly higher than those in ED slow muscle (P<0.05; Fig. 4G). On the other hand, MYH_M5 hybridization signal was only marginally detected in cardiac muscle (Fig. 4A).

Three fast-type MYH genes, MYH_M86-1, MYH_M2528-1 and MYH_M1034, were found to be expressed both in wild (N=3) and farm-cultured (N=3) fish (Fig. 5). The relative mRNA levels of MYH_M2528-1 in wild fish were significantly higher than those in farm-cultured fish (P<0.05; Fig. 5C). On the other hand, the mRNA levels of MYH_M86-1 and MYH_M1034 did not differ significantly between the two groups (Fig. 5B,D).

To identify MYHs expressed in torafugu fast muscle at the protein level, myosin was prepared from dorsal fast muscle of laboratory-reared torafugu (supplementary material Fig. S5). Proteolytic digestion of myosin produced fragments of S1, S2 and LMM, and the N-terminal amino acid sequence of S2 corresponded to that of fast-type MYH_M86-1, but not to those of other fast-type MYHs, namely MYH_M2528-1 and MYH_M1034 (supplementary material Fig. S5).

In situ hybridization was performed to localize the transcripts of the predominantly expressed MYH genes in skeletal muscles. The transcripts of fast-type MYH_M86-1 were found in all fibres of different diameters in fast muscle (Fig. 6A). On the other hand, the transcripts of fast-type MYH_M2528-1 were localized in fibres with small diameters in fast muscle (Fig. 6B), suggesting that these fibres express at least two MYH genes, MYH_M86-1 and MYH_M2528-1. Interestingly, fast fibres with smaller diameters tended to have a greater abundance of transcripts of MYH_M2528-1. The MYH_M2528-1 transcripts were also expressed in fibres of LS (Fig. 6C) and ED (Fig. 6D) slow muscles with large diameters. The fibres expressing MYH_M8248 resided a superficial part of LS and ED slow muscle with small diameters (Fig. 6E,F). Fibres expressing MYH_M2126-2 also occupied a superficial layer in LS slow muscle with small diameters. In situ hybridization for MYH_M2126-2 was not successful with fibres in ED slow muscle in the present study.

In the present study, seven sarcomeric MYH genes were cloned from adult torafugu skeletal muscles. Histochemical analysis using myofibrillar ATPase demonstrated that torafugu (290 g body mass) fast muscle contained various fibres with different pH stabilities and different diameters (see Fig. 1). These results are quite consistent with those reported for torafugu with a body mass of 154 g (Fernandes et al., 2005). Fernandes and colleagues claimed that the threshold body size for having no small fibres with the cessation of recruitment in torafugu fast muscle is about 1.2 kg and 35 cm standard length (Fernandes et al., 2005). Thus the torafugu specimens used in present study should to be at the stage of recruiting fast fibres.

MYH_M86-1 was predominantly expressed in fast muscle (see Figs 3 and 4) and such predominant expression of MYH_M86-1 in fast muscle was also confirmed by N-terminal amino acid sequence analysis on purified myosin (see supplementary material Fig. S5). In situ hybridization analysis localized the transcripts of MYH_M86-1 to all fast fibres with different diameters (see Fig. 6A). On the other hand, the transcripts of MYH_M2528-1 were found only in fast fibres with smaller diameters (see Fig. 6B). Therefore it seems that fast fibres with small diameters contain the transcripts of at least two MYH genes, MYH_M86-1 and MYH_M2528-1. Interestingly, MYH_M2528-1 was also expressed in fibres of LS and ED slow muscles with large diameters (see Fig. 6C,D). Transcripts of another fast-type MYH, MYH_M1034, were observed in fast muscle as revealed by cDNA clone library analysis (see Fig. 3). Unfortunately, in situ hybridization with a probe specific to MYH_M1034 was not successful.

In contrast, MYH_M8248 and MYH_M2126-2 were expressed only in fibres in a superficial part of LS slow muscle with small diameters, but not in fast fibres of any diameter or fibres in LS and ED slow muscles with large diameters (see Fig. 6E–G). NADH-diaphorase staining also revealed that fibres of LS slow muscle with large diameters were stained weakly compared with those in a superficial part with small diameters (see supplementary material Fig. S1). Previously, the small diameter fibres generated by hyperplasia in adult fast muscle were identified in common carp which expressed a distinct FG2MYH (Ennion et al., 1995). FG2MYH showed the highest identity to MYH_M2528-1 in deduced amino acid sequence, out of seven sarcomeric MYH genes cloned from adult torafugu in the present study (data not shown). As described in Results, MYH_M2528-1 transcripts were more abundant in fast fibres with smaller diameters. This relationship between fast fibre diameter and expression of MYH_M2528-1 agrees well with that between fast fibre diameter and ATPase staining intensity at pH 4.6. It is well known that fish muscles grow by hyperplasia even in adult (Stickland, 1983; Rowlerson et al., 1985; Weatherley and Gill, 1985; Rowlerson and Veggetti, 2001) and this feature is in marked contrast to mammals where adult muscles grow by hypertrophy (Rowe and Goldspink, 1969; Watabe, 1999). Given that such fast fibres with small diameters are generated by hyperplasia, the expression of MYH_M2528-1 is thought to be closely correlated with this fibre generation and fibres with the smallest diameter are considered to be the most recently formed. During the growth of fast fibres from small to large diameters, the expression of MYH_M2528-1 would be gradually decreased. The expression of MYH_M2528-1 in slow fibres with large diameters also indicates their involvement in muscle generation by hyperplasia. However, the nature of such fibres responsible for the process of hyperplastic growth in fish is unknown. Furthermore, the switch in expression of MYH genes during maturation of fish muscle and the mechanisms underlying such a switch are not clear. Thus it would be interesting to study the regulatory mechanisms involved in the expression of MYH_M2528-1 in muscle fibres, because such investigation will possibly clarify how these fibres are recruited in fish of indeterminate body size.

The expression levels of fast-type MYH genes were also investigated in both wild and farm-cultured torafugu individuals. Among three fast-type MYH genes, the relative mRNA levels of MYH_M2528-1 in wild fish were significantly higher than those in farm-cultured fish (see Fig. 5). The number of muscle fibres recruited to reach a given fish body size varies between species and between different strains of the same species, and is also affected by diet, exercise training and temperature (Johnston, 1999). Therefore, it is again considered that the expression of MYH_M2528-1 participates in muscle generation.

In this study we could not find transcripts of torafugu embryonic fast-type MYH_M743-2 (Ikeda et al., 2007) in cDNA clone library from fast muscle of adult, suggesting that developmental expression of fast-type MYH genes in torafugu might be tightly controlled.

In conclusion, we determined the expression patterns of seven sarcomeric MYH genes in adult torafugu fast, slow and cardiac muscles. While three MYH genes were specifically expressed in each of the muscles examined, the remainder showed a mixed expression pattern in fast, slow and/or cardiac muscle, suggesting the functional significance of each MYH. This is an important step towards understanding fibre type diversity and associated muscle growth in adult torafugu.