The fibre type composition of obliquely striated muscle of adult Hirudo medicinalis was investigated by enzyme histochemistry, by immunohistochemistry and by SDS-PAGE. The oxidative capacity of the fibres, assessed by succinate dehydrogenase activity, was similar in all three layers of body wall muscle (longitudinal, oblique and circular) and in dorsoventral muscles. Histochemical localisation of Mg2+-activated actomyosin ATPase activity gave stronger staining in the longitudinal muscle than in other layers. As muscle shortening speed is directly related to this form of ATPase activity, this suggests that the longitudinal layer fibres are faster contracting than the circular, oblique or dorsoventral muscles. Results with polyclonal antibodies specific for vertebrate myosins were consistent with the ATPase results, i.e. fibres with the lowest actomyosin ATPase activity reacted preferentially with an antibody for a slower myosin. Thus, anti-T2, selective for vertebrate tonic fibre myosin, bound preferentially to fibres in oblique, circular and dorsoventral muscles, whereas anti-S, selective for vertebrate slow twitch fibre myosin (faster than vertebrate tonic fibre myosin), bound preferentially to the bulk of longitudinal layer fibres. Whereas most of the longitudinal layer stained uniformly with the anti-S antibody, some fibres in the outermost bundles were negative for the anti-S antibody and were, therefore, different from the main mass of longitudinal fibres. SDS-PAGE analysis of contractile protein preparations from body wall muscle also revealed a difference in the composition of the oblique, circular and dorsoventral muscles compared to the longitudinal layer, supporting the conclusion that leech body wall muscle contains two fibre types.
It is well known that most skeletal muscles in vertebrates contain a mixture of muscle fibres of several differentiated types, and that the different types of fibre are contacted by motoneurones with different and characteristic properties, each motoneurone innervating only fibres of one particular type (Burke, 1981). Whether the same is true for muscles of the hydroskeleton in soft-bodied animals such as the leech is not known. Studies of muscle fibre types in invertebrates are rare: most of them are of crustacean or chelicerate muscles (e.g. Ogonowski and Lang, 1979; Levine et al. 1989) and demonstrate the existence of fast and slow types (often with different metabolic capacities as well) comparable in some respects to fast and slow types in vertebrates. Two muscle fibre types have also been found in Drosophila (Raghavan, 1981) and in the earthworm (D’Haese and Carlhoff, 1987). In both vertebrates and invertebrates, the contractile protein myosin consists of two kinds of subunit: heavy chains (relative molecular mass approximately 200×103) which form the force-generating ATPase sites, and associated light chains (two or more forms with relative molecular masses of about 20x103). Analytical techniques used to study vertebrate contractile proteins can also be applied successfully to invertebrate muscle (e.g. Costello and Govind, 1984), and invertebrate myosin light chains have received some attention because, in some species, calcium regulation of contraction is mediated by one of the light chains (Kendrick-Jones et al. 1976). By contrast, there is very little information available on invertebrate myosin heavy chains, yet studies of vertebrate myosin have shown that the heavy chain of myosin exists in multiple isoforms that are usually characteristic of particular muscle types (fast, slow, etc.) and determine their histochemical ATPase activity as well as some important functional properties (Bandman, 1985; Mascarello et al. 1986; Schiaffino et al. 1988; Sweeney et al.1988).
Body wall muscle in Hirudo medicinalis has characteristics that place it as intermediate in type between the classic smooth and striated muscles of vertebrates. Although originally classified as ‘helical smooth’ on the basis of X-ray diffraction (Hanson and Lowy, 1960), later fine structural studies showing highly ordered arrays of obliquely staggered myofilaments have led to its separate designation as ‘obliquely striated’ muscle (Rosenbluth, 1972). We were interested in whether this obliquely striated muscle, like striated muscle in vertebrates, contains a mixture of differentiated fibre types with different contractile proteins and different metabolic machinery producing energy for contraction.
To investigate fibre type composition of adult leech muscle we have used enzyme histochemistry for fibre typing of metabolic enzymes and localisation of ATPase activity, immunohistochemistry for myosin, and gel electrophoresis. Some of this work has appeared in preliminary form (Blackshaw and Rowlerson, 1989).
Materials and methods
Histochemistry and immunohistochemistry
Frozen 10 μm sections of body wall were stained for succinic dehydrogenase and alpha-glycerophosphate dehydrogenase (αGPDH). To look for differences in muscle contractile proteins we used methods developed mainly for use in mammalian muscle. For histochemical localisation of magnesium-activated actomyosin ATPase we used the method of Mabuchi and Sreter (1980), which includes magnesium as well as calcium in the ATP-containing incubation medium. The polyclonal antibodies tested against leech muscle are described in Tables 1 and 2. Antibodies were visualised by indirect fluorescence or by indirect immunoperoxidase staining of the sections.
Analysis of leech muscle myosin by SDS-PAGE
In a recent study of Limulus polyphemus L. muscle, two isoforms of myosin could be resolved by electrophoresis on pyrophosphate gels (Levine et al. 1989). However, as this method can reflect differences in either heavy or light chains of myosin, it does not provide an unambiguous demonstration of heavy chain differences. We used instead the method of Carraro and Catani (1983), in which different isoforms of the myosin heavy chain alone are separated by their differing mobility on electrophoresis in the presence of SDS in 5 % polyacrylamide gels.
Strips of body wall muscle were dissected from the leech. Two kinds of sample were removed: (1) ‘body wall’ samples, which contained longitudinal, circular and oblique muscle together with dorsoventral muscle and (2) ‘longitudinal’ samples, which were strips dissected from the muscle that constitutes the bulk of the body wall, the longitudinal layer. These samples were taken in the ventral midline from the deepest layer of muscle adjacent to the ventral nerve cord. Muscle samples were homogenised in an ice-cold low-salt buffer (0.1 mol 1-1 KC1, 5 mmol 1-1 EDTA, 0.01% 2-mercaptoethanol, pH7.5). The homogenate was centrifuged briefly and the supernatant discarded. The precipitate was resuspended in low-salt buffer, and the cycle of precipitation and resuspension was repeated three more times. This sort of procedure is often used to isolate well-washed myofibrils from vertebrate skeletal muscle and produced a sample of comparable composition (in terms of its contractile protein content) from the leech muscle, as shown in Fig. 3A. This contractile protein sample was suspended in gel sample buffer (0.075 mol 1-1 Tris-HCl, pH 6.8, 2 % SDS, 10 % glycerol, 5 % 2-mercaptoethanol, 0.001 % Bromophenol Blue) and heated to 100°C for 3 min to dissociate all protein subunits in preparation for their separation by gel electrophoresis. Conditions for electrophoresis were essentially as described by Laemmli (1970), and the separated protein bands were then visualised by silver staining. The overall composition of the leech contractile protein sample was compared with that of vertebrate myofibrillar samples on a two-stage separating gel permitting resolution of proteins covering a wide range of relative molecular masses (see Fig. 3A; sample loads 1-5 µg per lane). The myosin heavy chain composition of the contractile protein sample was analysed on 5 % separating gels, with electrophoresis conditions and staining as before, but with much lower sample loads of about 100ng. It has been shown that, despite their very similar primary structures, different isoforms of the heavy chain of myosin show small differences in migration rate in such gels (Carraro and Catani, 1983). Again the contractile protein sample was compared with reference samples of vertebrate fast and slow myosins (see Fig. 3B).
The tubular body wall of the leech contains three distinct muscle layers of differing thickness; an outer thin layer of circular muscle, one or two fibres thick, lying immediately beneath the skin; a thick inner layer of longitudinal muscle, and an orthogonal grid of oblique muscle fibres sandwiched between the other two layers. The longitudinal muscle forms the bulk of the body wall and consists of many rows of fibres arranged in bundles (Fig. 1). These three layers, together with the dorsoventral muscles that span the body cavity, are largely responsible for the posture of the leech and its locomotory activities of swimming, crawling and walking.
Histochemistry and immunohistochemistry
Succinic dehydrogenase, a mitochondrial enzyme, is an indicator of oxidative metabolism and in vertebrate muscle is strongly correlated with fatigue resistance. In the leech muscle it was found only in the central core of the fibres, where the mitochondria are located. Staining was uniform across circular and oblique layers, but in the longitudinal layer there appeared to be a gradual transition in the density of the stain, with the densest staining located in bundles at the outer edge of the layer, adjacent to the oblique muscle. αGDPH activity, an indicator of glycolytic pathways and associated with fast twitch fibres in vertebrate muscle, was not seen in any of the muscle layers, though it was detected in the skin epithelial layer on the outer edge of the body wall sections.
The methods used to look for differences in muscle contractile proteins were developed mainly for use in mammalian muscle. The method of Mabuchi and Sreter (1980) for histochemical localisation of magnesium-activated actomyosin ATPase gave stronger staining in the longitudinal muscle than in the other layers (Fig. 2A). As muscle shortening speed is directly related to this form of ATPase activity (Barany, 1967) this result suggests that the longitudinal layer fibres, which are used in swimming, are faster contracting than fibres in other layers, which are thought to have a largely postural function. The dorsoventral muscles, for example, which produce flattening of the body, are known to be tonically active throughout swimming (Ort et al. 1974).
Several polyclonal antibodies specific for vertebrate fast and slow myosins were tested against the leech. Only two of these anti-myosin antibodies gave selective staining. Their characteristics are summarised in Table 1. Anti-T2, selective for vertebrate tonic fibre myosin, also reacted against leech muscle, where it bound preferentially to fibres in oblique, circular and dorsoventral muscles (Fig. 2B). Anti-S, selective for vertebrate slow twitch fibre myosin in mammals and other vertebrates, reacted weakly with the superficial circular and oblique layers but strongly with the bulk of the longitudinal layer fibres (Fig. 2C). In vertebrates, slow twitch fibres are faster than tonic fibres. Assuming the binding specificity of anti-T2 and anti-S has the same significance in leech as in vertebrate muscle, our results are consistent with the ATPase results, i.e. the fibres reacting preferentially with the antibody selective for tonic myosin have the lower ATPase activity.
Whereas most of the longitudinal layer stained uniformly with the anti-S antibody, i.e. were anti-S positive, fibre bundles on the outer edge of the longitudinal layer adjacent to the oblique muscle contained a mixture of fibres -some of these were like the majority of the longitudinal layer fibres, but others were negative for the anti-S antibody and, therefore, different from the main mass of longitudinal fibres (Fig. 2D).
The vertebrate-raised polyclonal antibodies which were tested against the leech and did not give selective staining are listed in Table 2.
Analysis of leech muscle myosin by SDS-PAGE
The contractile protein sample derived from leech body wall was similar in composition to myofibrillar preparations from mammalian fast and slow muscles as shown by SDS-PAGE (Fig. 3A). The major components were myosin heavy chain and actin; minor components with a wide range of relative molecular masses were also present, and these included two low molecular mass bands tentatively identified as the myosin light chains (Fig. 3A). SDS-PAGE on 5 % gels revealed two clearly resolved myosin heavy chain components in the leech body wall sample (which contains muscle from all three layers and dorsoventral fibres). This is shown in Fig. 3B with mammalian fast and slow myosin heavy chains for comparison. The separation of the two bands in the leech contractile protein sample is as good as between the mammalian fast and slow isoforms of the myosin heavy chain; presumably the two leech bands are also due to different isoforms. In the muscle sample dissected from the deepest part of the longitudinal layer, only one heavy chain band was present, the more abundant component.
Unfortunately, identification of the minor heavy chain isoform as that which binds anti-T could not be done by immunoblotting, as leech myosin heavy chain (both types) consistently fails to react with any of our antibodies to myosin on both Western and dot blots. We can only conclude that the antigenic similarity between leech and vertebrate slow and tonic myosins does not survive the denaturation process used to prepare samples for gel electrophoresis.
The histochemistry and immunohistochemistry of obliquely striated muscle from leech body wall show two fibre ‘types’ with different ATPase activity and different myosin composition. The circular, oblique and dorsoventral muscles, and a small number of fibres in the longitudinal layer, are slower than the bulk of the longitudinal layer fibres, as judged by their lower ATPase activity. This conclusion is supported by their different immunoreactivity. For example, the slower (circular, oblique and dorsoventral) fibres react preferentially with an antibody that, in all vertebrate species examined, is specific for tonic fibres, which are the slowest of all. The results of the SDS-PAGE analysis of a contractile protein preparation obtained from leech body wall muscle support the conclusion from the histochemical experiments that leech body wall muscle consists of two fibre types. On SDS gels two distinct myosin heavy chain bands were found in the samples of body wall muscle that included all three layers together with dorsoventral fibres, but the weaker of these two bands was not present in samples of longitudinal muscle only. This indicates that the oblique, circular and dorsoventral muscles contain a myosin heavy chain isoform which is distinct from the isoform in thg longitudinal muscle, and most probably accounts for the different ATPase activity and immunoreactivity shown by these muscle layers.
The proposal that different myosin heavy chains are present in functionally specialised invertebrate muscles is supported by studies of obliquely striated muscle in the nematode Caenorhabditis elegans (for a review see Waterston, 1988). In C. elegans, four electrophoretically distinct heavy chain isoforms are present; two of these are constituents of body wall, vulval and anal sphincter muscles, the other two are found exclusively in the pharynx. It may be that additional myosin heavy chain isoforms are present in other muscles in the leech, such as heart or jaw muscles, which were not included in this study. The C. elegans studies, however, show that more than one isoform may be present in a single fibre. For example, both myo A and myoB are expressed in all 95 body wall muscle cells, but are confined to distinct regions of the thick filament. Their location suggests that the different isoforms may be specialised for different assembly functions, since myosin is assembled in a different way in the different regions. Our gel electrophoresis results on adult leech muscle showed only a single myosin heavy chain band present in deep longitudinal fibres; this does not exclude the possibility that it may have contained two or more unresolved isoforms, and we did not attempt to investigate this further.
Our finding that the outermost bundles of longitudinal muscle are different from the main mass of longitudinal muscle in containing a mixture of fibres is interesting. The morphological studies of Lanzavecchia and colleagues (Lanzavecchia, 1977; Lanzavecchia et al. 1977) in a different species of leech, Glossiphonia complanata, show two distinct regions of longitudinal muscle: a smaller outer zone where the fibres are arranged in ‘rosettes’ and a larger interior zone. One possibility is that different regions of the longitudinal muscle are functionally specialised, and may receive different innervation. In Hirudo, the motoneurones and modulatory neurones supplying the longitudinal layer that lie within the segmental ganglia have been identified and characterised in terms of their biochemistry and fields of innervation (Stuart, 1970; Ort et al. 1974; Kuffler, 1978; Cline, 1983); and physiological studies have shown that individual motoneurones elicit contractions with characteristic and different rise times and peak tensions (Mason et al. 1979; Mason and Kristan, 1982). It may be that in the leech the two kinds of fibre in the longitudinal muscle each have a specific and different combination of excitatory or inhibitory innervation.
An alternative possibility is that the heterogeneity in the leech longitudinal muscle is associated with the sensory rather than the motor innervation of the muscle. The longitudinal muscle is innervated by large peripheral neurones that respond to stretch and release of the muscle with hyperpolarising and depolarising potentials (Blackshaw and Thompson, 1988) and whose input influences the rhythmical contractions of the longitudinal muscle during swimming (Blackshaw and Kristan, 1990). H. medicinalis does not appear to have structurally distinct receptor muscles like those found in vertebrates or articulated invertebrates, Rather, the dendrites of the leech stretch receptors innervate specific fibres within the longitudinal layer and these fibres are located within the outermost bundles, adjacent to the oblique muscle. The association of individual sensory or motor neurones with the slower or faster fibres in the longitudinal layer could be tested by combining intracellular staining of the neurones with immunohistochemistry.