Dissolution of the contractile apparatus derived from larval muscle occurs when pupal morphogenetic movements are complete, but a residual myofibre remains. Synapses with nerves are retained by degenerating myofibres, although the nerves are not structurally normal. Within the myofibre, the appearance of several types of unit membrane aggregate is coincident with the disappearance of normal sarcoplasmic reticulum. One type of aggregate consists of stacked cisternae, separated by a gap of 300 Å which is traversed by a three-dimensional connecting lattice of regularly arranged rods. These latticed cisternae appear to be derived from sarcoplasmic reticulum, and they interconnect with smooth-surfaced cisternae. Structures akin to latticed cisternae have been described by others in a variety of cells. Of particular interest is a report of their formation as a result of denervation in rat muscle. This report, coupled with the evidence presented here, suggests that latticed cisternae are indicative of changed nervous stimuli in degenerating muscle.

Changes in the membrane systems are rapidly followed by resorption of the myofilaments. Thick myofilaments disappear first, followed by thin myofilaments, with the eventual disappearance of Z-discs completing the dissolution of the sarcomere. No evidence was obtained for the segregation of myofilaments in membrane-bound vacuoles, or for the presence of lysosomes in muscle at this time. It is suggested that these observations are consistent with theories developed by others for retention of protein structure, rather than disassembly into amino acids, at metamorphosis in insects.

Formation of surface blebs is correlated with loss of organelles to the haemolymph. Morphological evidence supports the proposition that coated vesicles are involved in exchange of proteins with the haemolymph, shortly before adult myofilaments form.

Insect myoblast fine structure is described. The bipolar shape of these cells is shown to be associated with an oriented fascicle of microtubules. Attention is drawn to the plasma membrane of the myoblast, which bears folds at the points where flexion of the cell can be demonstrated by time-lapse microscopy. Myoblasts do not contain recognizable microfilaments other than microtubules.

Fusion of free-floating myoblasts with the residual myofibre is demonstrated by light and electron microscopy, and the stages of fusion are described. Ultrastructual manifestation of a cell recognition process was not detected. The process of multinucleation of developing muscle cells in insects is directly comparable with myoblast fusion in vertebrate cells. How ever, in Calliphora the resulting myofibre contains two classes of nucleus, easily recognizable on basis of size. The structure of these nuclear classes, and the role played by each in adult myofibril formation, is investigated in an accompanying paper.

The formation of the fly puparium involves the contraction of nearly all the larval intersegmental muscles, followed by the hardening of the larval cuticle. In Calliphora erythrocephala, phagocytosis of many intersegmental muscles by haemocytes begins as soon as the puparium is hard, the process being initiated by ecdysone (Crossley, 1965, 1968). Several specific abdominal intersegmental muscles escape phagocytosis and subsequently contract, within the puparium, in morphogenetic movements that bring about the eversion of the developing adult head. Following this event, the majority of intersegmental muscles are completely destroyed by a combination of autolysis and phagocytosis. The few remaining intersegmental muscles are reconstructed to form adult muscles (Perez, 1910; Crossley, 1965). The reconstruction begins with a total loss of striated myofibrils, and accumulation of membranous structures within syncytial tubular muscle remnants. The resulting ‘residual myofibres’ then redifferentiate, forming striated myofibrils that are contractile at the time of emergence of the adult fly (Crossley, 1965). In this report the ultrastructural changes leading to the formation of the residual myofibre, and the fusion of this with myoblasts, are considered in detail.

There is an extensive literature, reviewed earlier (Crossley, 1965), on the changes in insect muscles at metamorphosis. Few authors have detected continuity of larval to adult muscle, and in view of the dimensions of the transitional fibre (width ca. 200 μm), coupled with the difficulties of investigating the anatomy of insect pupae, this is not surprising. Nevertheless, the transition phenomena now described here in detail are probably of widespread occurrence in insect metamorphosis, and may apply to other animals where muscles are reoriented without total destruction and replacement. The literature based on fight microscopy favours the idea that adult muscle nuclei arise by ‘fragmentation’ of larval muscle nuclei (Lowne, 1870; van Rees, 1888; de Bruyne, 1898; Perez, 1910; Schmidt, 1928), or by multiple divisions of dormant imaginal nuclei (Snodgrass, 1924), in the abdominal intersegmental muscles of Díptera. However, Perez (1910), Crossley (1965) and Edwards (1969) report that fusion of mononucleated cells with the myofibre introduces new nuclei into the syncytium. In a recent review of myofibrillogenesis in insects, Auber (1969) states ‘En microscopie électronique, nous n’avons pas trouvé d’aspects proubants d’une pénétration des myoblasts dan la fibre musculaire… . L’origine des petits noyaux reste à démontrer, mais il apparait qu’au moins une partie des fibres musculaires larvaires remainiées participe à la formation des fibres des muscles du vol’. It is thus necessary to establish with certainty that myoblast fusion occurs in developing insect muscles, and this is a primary aim of the first paper in this series.

An involution cycle in the intersegmental muscles of the hemipteran Rhodnius was discovered by Wigglesworth (1956), and this cycle has now been investigated with the electron microscope by several authors (Warren & Porter, 1966, 1969; Auber-Thomay, 1967; Toselli & Pepe, 1968 a, b). The transition reported here for Calliphora muscle has many structural analogies with the Rhodnius system, even though in Calliphora an involution only occurs once in the life cycle, during the fourth instar: the pupal stage. Fusion of myoblasts with ‘resting’ myofibres was not recorded by earlier workers on Rhodnius muscle (Wigglesworth, 1956; Warren & Porter, 1966; Auber-Thomay, 1967), and Auber-Thomay suggested that the increase in numbers of nuclei during muscle development was accounted for by amitosis. Recently, however, it has been suggested that this increase is brought about by fusion of spindle-shaped mononucleated cells with the muscle (Toselli & Pepe, 1968b).

There is little information on the structure and origins of insect myoblasts. The cells giving rise to the indirect flight muscles have been described for Drosophila by Tiegs (1955) and Shafiq (1963 a, b), and for Dysdercus by Edwards (1969). The mononucleated cells described by Toselli & Pepe (1968 b) can presumably be considered to be myoblasts. In Calliphora the presumptive myoblasts of the abdomen have been described on the basis of phase microscopy as ‘bipolar cells with long filamentous tail-like extensions of the cell membrane’ during a free-floating, possibly motile, stage in the haemolymph (Crossley, 1965). It is now possible to demonstrate the flexion of the membrane-bound extensions of moving myoblasts, and to describe the fine structure of the extensions. The present detailed report on the process of fusion of mononucleated cells with the myofibre of an insect allows direct comparison with the process of fusion of vertebrate myoblast and myotube.

Calliphora erythrocephala (Meig.) ( = vicina, R.-D.) larvae were reared at 25 °C and 60 % relative humidity. Under these conditions the stages within the puparium occupied 210–220 h.

Internal dorso-lateral oblique recti muscles (Hewitt, 1914), numbered 1 to 4 in an earlier anatomical study (Crossley, 1965), were obtained by dissection of the pupa under fixative. Fixatives used included 2 ·5 % glutaraldehyde solution, buffered at pH 7 ·2 with 0 ·05 M-HCl-cacodylate or phosphate, containing 0 ·15 M-sucrose; or 1 % osmium tetroxide solution, buffered at pH 7 ·2 with veronal acetate, with added sucrose. Initial fixation was at 20 °C for 10 min, with subsequent reduction to 4 °C for a further period of h. After thorough washing in buffered sucrose solution over a period of 18 h at 4 °C, glutaraldehyde-fixed material was post-fixed in 1 % osmium tetroxide, and dehydrated in a graded ethanol series. It was necessary to carry out the initial fixation of the muscles whilst these were still attached to the cuticle, since the muscles only became clearly visible after treatment of the preparation with osmium. The final dissection of muscles from the cuticle was carried out in ethanol during the dehydration sequence. Some specimens were treated with 5 % uranyl acetate in 90 % alcohol during dehydration. Dehydrated muscle was embedded in Araldite and thin sections were obtained on a Porter-Blum MT-2 or a Reichert OM-U2 microtome equipped with diamond knives. Contrast in electron micrographs was enhanced by treatment of sections with 5 % uranyl acetate in 20 % methanol followed by lead citrate, the latter according to Reynolds (1963). Sections were examined in a Siemens Elmiskop I operated at 80 kV, or in a Hitachi HU11E operated at 75 kV. In both instruments 30 μm self-heating apertures were used, and the magnifications were calibrated using a germanium shadowed carbon replica of a 463 nm diffraction grating.

Living muscles and myoblasts were dissected from pupae in saline iso-osmotic with Calliphora haemolymph (NaCl 8 ·86 g/1, KC12 ·86 g/1, CaCl21 ·13 g/1, glucose 9 ·72 g/1). Fine jets of saline were used to disperse the fat body, and to disengage muscles from the cuticle. Tissues in saline were examined on glass slides in a Zeiss Photomicroscope equipped with Nomarski interference and phase contrast optics.

(1) Formation of the residual myofibre from larval muscle

The ultrastructure of the abdominal intersegmental muscles of the larva has already been described (Crossley, 1968). In larval muscles that are to undergo transition to adult muscles this structure is unchanged during morphogenetic movements within the puparium that culminate with the eversion of the adult head, at about 24 h after puparium formation. During the next 24 h these muscles undergo drastic reorganization, beginning with the dissolution of the larval contractile elements, so that by 48 h after puparium formation only a thin strand of cytoplasm, the ‘residual myofibre’, remains. Muscles fixed 30 h after puparium formation are shown in Figs. 1–3, 5–7. Each fibre is coated with a thick (0 ·3–0 ·5 μm) extracellular basement membrane, which traces out the major folds of the sarcolemma. Electron dense plaques beneath the sarcolemma mark the sites of closest contact with the basement membrane (Fig. 2). These plaques are interpreted as hemidesmosomes, and are points of attachment, and possibly also of secretion, of basement membrane material. The peripheral cytoplasm of the myofibre is already showing signs of the fragmentation that is discussed below. Many microtubules are apparent, and in some areas their orientation is quite different from that of the fibre axis (Fig. 2). Within the myofibre two significant modifications of the larval muscle fine structure are apparent at 30 h after puparium formation. The first is the appearance of unit membrane aggregates (Figs. 3–4). These structures are made up of fragments of 75 Å width unit membrane more or less closely stacked together, and entirely enclosed by a limiting unit membrane. The edges of the membranes thus do not come into direct contact with the cytoplasm, and the entire structure is a coherent disc. In closely stacked aggregates the unit membrane surfaces are fused (Fig. 3). In loose stacks this distance between membranes is variable (Fig. 4). Similar types of aggregate were first described by Miller (1960) in the mouse kidney, and they have since been encountered in many other types of cell. In the Calliphora myofibre these structures become extremely numerous, and it is of interest to note that their appearance coincides with the disappearance of the normal organization of the sarcoplasmic reticulum, shortly before resorption of the myofilaments. A11 forms of membrane aggregate persist in the myofibre during the remainder of the transition period. The second modification of larval muscle fine structure is the appearance of peculiar fibrillar lattices on cisternae lying between myofilament bundles (Figs. 5–7). These displace normal sarcoplasmic reticulum at the level of the I-band, and are closely associated with the T-tubule system. The structure of latticed cisternae is considered in Results (2), p. 52. A nervous connexion, sometimes synaptic in nature, is retained by degenerating muscle (Fig. 1), although axons spread out over the muscle at this stage contain membranous elements reminiscent of the isolation vacuoles described by Locke (1966), which may indicate the onset of degenerative changes.

Fig. 1.

Myofilaments remain in muscle fixed 30 h after puparium formation, but unit membrane aggregates (u.m.a.) and isolated groups of myofilaments in the peripheral sarcoplasm (f) herald the onset of autolysis. Nervous connexions are present (n), although the nerve cell cytoplasm contains membranous enclosures interpreted as isolation vacuoles. An example of a neuromuscular junction with associated synaptic vesicles on muscle at this stage is shown in the inset. Lysosome-like structures are absent from nerve and muscle (b.m. = basement membrane; mc = mitochondrion; Z = Z-disc; A = A band of sarcomere; I = isolation vacuole), (× 23000; inset × 37000.)

Fig. 1.

Myofilaments remain in muscle fixed 30 h after puparium formation, but unit membrane aggregates (u.m.a.) and isolated groups of myofilaments in the peripheral sarcoplasm (f) herald the onset of autolysis. Nervous connexions are present (n), although the nerve cell cytoplasm contains membranous enclosures interpreted as isolation vacuoles. An example of a neuromuscular junction with associated synaptic vesicles on muscle at this stage is shown in the inset. Lysosome-like structures are absent from nerve and muscle (b.m. = basement membrane; mc = mitochondrion; Z = Z-disc; A = A band of sarcomere; I = isolation vacuole), (× 23000; inset × 37000.)

Fig. 2.

The thick basement membrane of 30 h pupal muscle is only applied to the plasma membrane at the sites of hemidesmosomes (hd). Primary and secondary myofilaments pass through the Z-disc, and indicate that the muscle is fixed in supercontraction. Microtubules (m) in the peripheral sarcoplasm are not oriented with respect to the fibre axis. ( × 36500.)

Fig. 2.

The thick basement membrane of 30 h pupal muscle is only applied to the plasma membrane at the sites of hemidesmosomes (hd). Primary and secondary myofilaments pass through the Z-disc, and indicate that the muscle is fixed in supercontraction. Microtubules (m) in the peripheral sarcoplasm are not oriented with respect to the fibre axis. ( × 36500.)

Fig. 3.

A unit membrane aggregate similar to that seen in Fig. 1 is here shown at higher magnification. It is composed of stacks of fused unit membranes bound by a peripheral unit membrane into a flattened disc. ( × 69000.)

Fig. 3.

A unit membrane aggregate similar to that seen in Fig. 1 is here shown at higher magnification. It is composed of stacks of fused unit membranes bound by a peripheral unit membrane into a flattened disc. ( × 69000.)

Fig. 4.

Unit membrane aggregates in autolysing muscle are not invariably fused. In this micrograph the membranes form loose and irregular stacks associated with granular material in membrane-bound enclosures. ( ×90000.)

Fig. 4.

Unit membrane aggregates in autolysing muscle are not invariably fused. In this micrograph the membranes form loose and irregular stacks associated with granular material in membrane-bound enclosures. ( ×90000.)

Fig. 5.

In this relaxed muscle cut in longitudinal section, the Z-disc (Z) is perforated and irregular. The transverse tubule system (t) is here associated with sarcoplasmic reticulum bearing an electron dense periodic surface lattice (l.c. = latticed cisternae, I = I band, A = A band). ( × 74000.)

Fig. 5.

In this relaxed muscle cut in longitudinal section, the Z-disc (Z) is perforated and irregular. The transverse tubule system (t) is here associated with sarcoplasmic reticulum bearing an electron dense periodic surface lattice (l.c. = latticed cisternae, I = I band, A = A band). ( × 74000.)

Fig. 6.

In this transverse section of 30 h muscle, periodic electron dense profiles are seen on walls of cisternae apparently derived from endoplasmic reticulum. This is a minimum latticed unit consisting of two cisternae. ( × 48 500.)

Fig. 6.

In this transverse section of 30 h muscle, periodic electron dense profiles are seen on walls of cisternae apparently derived from endoplasmic reticulum. This is a minimum latticed unit consisting of two cisternae. ( × 48 500.)

Resorption of the myofilaments proceeds rapidly, with the disappearance of the thick primary myofilaments first, leaving elongated Z-disc elements with attached thin secondary filaments. Electron dense deposits of amorphous material adhere to the thin filaments in the vicinity of the Z-disc (Fig. 8). By the time 60 h have elapsed since puparium formation contractile elements have disappeared without trace, and there is no evidence of any residual sarcomere architecture (Fig. 9). No structures reminiscent of primary lysosomes or phagosomes were encountered in the muscle at the time of myofilament dissolution. Attempts to demonstrate acid-phosphatase activity in the 30–50 h myofibre, using a modified Gomori-type medium (Gomori, 1950; Barka, 1964), gave negative results, although positive reactions were obtained in haemocytes (see Crossley, 1968). Nor is there any evidence of sequestration of muscle structures within isolation vacuoles. The dissolution of the contractile apparatus is apparently achieved in the main body of the cytoplasm.

Fig. 7.

Latticed cisternae contain amorphous material when they are first formed. The electron dense profiles rest on a unit membrane which may be indented at that point (arrows). ( × 53000.)

Fig. 7.

Latticed cisternae contain amorphous material when they are first formed. The electron dense profiles rest on a unit membrane which may be indented at that point (arrows). ( × 53000.)

Fig. 8.

A longitudinal section of a residual myofibre 50 h after puparium formation. Sarcomere organization has been disrupted by autolysis of the thick primary myofilaments, leaving thin secondary myofilaments attached to elongated Z-discs. Dense deposits of amorphous material (D) adhere to the thin secondary filaments (2f) close to the Z-disc (Z). (53 000.)

Fig. 8.

A longitudinal section of a residual myofibre 50 h after puparium formation. Sarcomere organization has been disrupted by autolysis of the thick primary myofilaments, leaving thin secondary myofilaments attached to elongated Z-discs. Dense deposits of amorphous material (D) adhere to the thin secondary filaments (2f) close to the Z-disc (Z). (53 000.)

Fig. 9.

Sixty hours after puparium formation the residual myofibre is a strand of cytoplasm without trace of contractile elements. The plasmalemma has lost its extracellular basement membrane, and bears blebs (bl) containing organelles (here believed to be ribosomes) having only a tenuous connexion with the myofibre. Channels (ch) derived from fibre clefts or T-tubules penetrate deeply into the fibre. Coated vesicles (c.v.) are present (see also inset). ( × 31000; inset × 60000.)

Fig. 9.

Sixty hours after puparium formation the residual myofibre is a strand of cytoplasm without trace of contractile elements. The plasmalemma has lost its extracellular basement membrane, and bears blebs (bl) containing organelles (here believed to be ribosomes) having only a tenuous connexion with the myofibre. Channels (ch) derived from fibre clefts or T-tubules penetrate deeply into the fibre. Coated vesicles (c.v.) are present (see also inset). ( × 31000; inset × 60000.)

The 60 h myofibre is a column of cytoplasm about 2 mm long and 50-100 μm in diameter, connected at each end to the integument (Crossley, 1965). The plasma membrane is most irregular, and portions of cytoplasm appear as blebs with only tenuous connexion to the myofibre. Since the myofibre is greatly reduced in volume in the residual condition, it is possible that some of the protruding blebs become detached and pass into the haemolymph. Profiles such as Fig. 9 are consonant with such a proposition. The nature of the contents of the blebs is variable, but granules of the dimensions of ribosomes are frequently packed in blebs (Fig. 9). Nuclei do not appear in the blebs. The basement membrane has disappeared, and coated vesicles are commonly encountered at the plasma membrane. The coated vesicles are about 1100 Å in overall diameter, and persist until about 70 h after puparium formation. Channels derived from fibre clefts or transverse tubules penetrate deeply into each residual myofibre (Fig. 9, ch). The cytoplasm contains ribosomes, mitochondria, membranous elements (discussed above) and electron lucent droplets. The latter are separated from the cytoplasm by phase boundaries or protein membranes but not by unit membranes (Figs. 10, 12), and probably correspond to the lipid-rich foci detected with the light microscope (Crossley, 1965). The myofibre ground-plasm is markedly electron dense, presumably reflecting a high concentration of dissolved protein. Microtubules traverse the cytoplasm without regular orientation with respect to each other or to the axis of the myofibre (see Crossley, 1972).

Fig. 10.

Multiple units of latticed cisternae have accumulated in the core of the residual myofibre at 140 h after puparium formation. Nuclei derived from myoblasts (N2) are smaller, and have fewer nuclear pores than those derived from larval muscle (N1). ( × 16300.)

Fig. 10.

Multiple units of latticed cisternae have accumulated in the core of the residual myofibre at 140 h after puparium formation. Nuclei derived from myoblasts (N2) are smaller, and have fewer nuclear pores than those derived from larval muscle (N1). ( × 16300.)

Fig. 11.

Latticed cisternae may bear ribosomes on the outermost membrane (arrows). Separation of internal membranes (at X) leaves dense profiles attached to each membrane, and these are presumably derived from the lattice. ( × 75000.)

Fig. 11.

Latticed cisternae may bear ribosomes on the outermost membrane (arrows). Separation of internal membranes (at X) leaves dense profiles attached to each membrane, and these are presumably derived from the lattice. ( × 75000.)

Fig. 12.

Latticed cisternae connect with smooth-walled channels extending into other parts of the cytoplasm (dr. = electron lucent droplet, ? lipid; wh. = membranous whorl; l.c. = latticed cisternae). (× 29500.)

Fig. 12.

Latticed cisternae connect with smooth-walled channels extending into other parts of the cytoplasm (dr. = electron lucent droplet, ? lipid; wh. = membranous whorl; l.c. = latticed cisternae). (× 29500.)

(2) Latticed cisternae

In the previous section attention was drawn to the formation of fibrillar lattices on cisternae at the sites formerly occupied by normal sarcoplasmic reticulum. These structures, here termed ‘latticed cisternae’, first appear about 30 h after puparium formation, and increase in number during the period of resorption of the larval contractile apparatus (30–45 h). Their number then remains static, as far as can be observed, until development of adult myofilaments is well advanced (140 h, Fig. 10) when a decline in number begins. Latticed cisternae occur, although rarely, in adult muscle (Crossley, 1972). Lattice filaments were not found in material fixed with permanganate, but were preserved by both osmium and glutaraldehyde as primary fixatives. The density of the filament lattice is particularly enhanced by treatment with uranyl ions.

The minimal unit appears to consist of two cisternae, each 300–1200 Å wide and limited by a unit membrane. The outer limits of adjacent cisternal unit membranes are separated by a gap of 320 Å ( ± 20 Å). When the muscle is sectioned transversely with respect to the myofilament array, the gap appears to be partially traversed by electron dense profiles arranged on the membranes in periodic fashion (Figs. 11, 12). Additional cisternae may be added to the minimal unit to form a stack, but each cisternum is separated from its neighbour by a ca. 320 Å gap and electron dense profiles. The peripheral elements of each stack of cisternae only bear fibrillar structures on the membrane facing the adjacent cisternum, although granules interpreted as ribosomes are present on the outer membrane (Fig. 11). The cisternae appear to contain amorphous material when they are first formed (Figs. 5, 7), but later become electron lucent (Fig. 10). The ends of newly formed latticed cisternae expand into chambers having close association with T-tubules (Fig. 5), reminiscent of lateral cisternae of normal sarcoplasmic reticulum, except that diadic contacts are not detected. When resorption of the myofilaments is complete, it can be seen that latticed cisternae interconnect with smooth-walled reticulum extending into other parts of the cytoplasm (Fig. 12).

By sectioning in various planes it can be shown (Figs. 13, 14) that the periodic electron dense profiles between cisternae are derived from elongated rods or fibrils, 70–140 Å in diameter, rather than from granules. Difficulties in alignment make measurements of fibril sizes rather imprecise, but there is an indication that two fibril populations may be involved, one having a mean diameter close to 70 Å, the other close to 140 Å. Fibrils are arranged on the surface of cisternal membranes in parallel arrays, with a centre-to-centre spacing of 310 ű 10 Å. Sections cut tangentially with respect to the cisternae reveal that the arrays of rods intersect at an angle of 60° or more (Fig. 13), giving rise to a diamond-shaped lattice. Interpretation of the structure is greatly hampered by the plication of the membranes supporting the lattice, so that fibrils are rarely maintained in the plane of section for much of their length. It was at first postulated that the array of fibrils on one membrane lay at an angle of 60° to the array on the adjacent membrane. Construction of models revealed that this structuring failed to give rise to a regular alternation of electron dense and electron lucent areas in any plane of section at right angles to the fibril axis. Furthermore, models involving helices are unable to interpret the straight fibril arrays, although it must be admitted that some profiles have a distinctly helical appearance (Figs. 5-6). Making the assumption, possibly unwarranted, that a single type of fibril relationship is involved in all cases, a model has been devised that fits all observed configurations (Fig. 15). In this model a series of secondary rods connect those fibrils directly attached to membranes. Evidence for the existence of such secondary rods can be obtained from micrographs of transversely sectioned material (Fig. 11), and is implied by the diamond-shaped profiles that appear in sections which include the entire three-dimensional lattice (Fig. 13). Nevertheless, this model should be regarded as tentative pending stereoscopic studies. Hypotheses concerning the nature of the lattice are considered in the Discussion.

Fig. 13.

In latticed cisternae sectioned obliquely it can be seen that the electron dense profiles attached to cisternae are linear filaments (arrows), arranged in rows. The rods intersect at an angle of about 60° in some areas forming a diamond-shaped lattice (asterisks). ( × 95 000.)

Fig. 13.

In latticed cisternae sectioned obliquely it can be seen that the electron dense profiles attached to cisternae are linear filaments (arrows), arranged in rows. The rods intersect at an angle of about 60° in some areas forming a diamond-shaped lattice (asterisks). ( × 95 000.)

Fig. 14.

Another oblique profile showing rods of latticed cisternae arranged in rows. The rods are laid down with a centre-to-centre periodicity of 310 Å. ( × 89000.)

Fig. 14.

Another oblique profile showing rods of latticed cisternae arranged in rows. The rods are laid down with a centre-to-centre periodicity of 310 Å. ( × 89000.)

Fig. 15.

Diagram illustrating one model of latticed cisternae that is consistent with the images obtained by sectioning in various planes. Separation of rods has been exaggerated to clarify the possible three-dimensional relationships.

Fig. 15.

Diagram illustrating one model of latticed cisternae that is consistent with the images obtained by sectioning in various planes. Separation of rods has been exaggerated to clarify the possible three-dimensional relationships.

(3) The structure of myoblasts

The origin of the mononucleated cells that unite with abdominal intersegmental residual myofibres has been considered in an earlier communication (Crossley, 1965). Evidence was presented to suggest that presumptive myoblasts are derived from imaginal discs in each abdominal segment. (Imaginal discs are discrete reservoirs of embryonic cells that are not programmed to differentiate into larval organs.) There is no information bearing on the specificity of the migration of cells from each imaginal disc to particular muscles, but presumptive myoblasts come to lie at the surface of residual myofibres between 70 and 90 h after formation of the puparium (Fig. 18). Free and attached cells are encountered. In the present work, cells are identified as presumptive myoblasts only if they are in close contact with myofibre, and remain attached during tissue processing procedures. With the exception of living presumptive myoblasts photographed in the light microscope, free-floating myoblasts are not described here, because of the difficulty of positive identification. However, the bipolar morphology of the free cell is retained for some time after the cell becomes attached to a myofibre by the tip of one pole (Fig. 16E). It is these recently attached cells that are here described as ‘presumptive myoblasts’ in electron micrographs.

Fig. 16.

A-D. Phase-contrast (after Zernicke) light micrographs of a living presumptive myoblast in tissue maintenance medium. Micrographs taken at intervals of 2 min show the flexion of the membrane close to the nucleus, and the changed orientation of the elongated pole. Movement of the entire cell with respect to the substrate is associated with these changes. E. Movement of myoblasts brings them to the surface of the myofibre where they become attached, initially by one pole (arrow). Note the very large nuclei (n) pre-existing in the myofibre derived from larval muscle (u = unattached presumptive myoblast). F. Interference contrast image of two presumptive myoblasts (magnifications A–E × 2000; F ×4000).

Fig. 16.

A-D. Phase-contrast (after Zernicke) light micrographs of a living presumptive myoblast in tissue maintenance medium. Micrographs taken at intervals of 2 min show the flexion of the membrane close to the nucleus, and the changed orientation of the elongated pole. Movement of the entire cell with respect to the substrate is associated with these changes. E. Movement of myoblasts brings them to the surface of the myofibre where they become attached, initially by one pole (arrow). Note the very large nuclei (n) pre-existing in the myofibre derived from larval muscle (u = unattached presumptive myoblast). F. Interference contrast image of two presumptive myoblasts (magnifications A–E × 2000; F ×4000).

Fig. 17.

This bipolar presumptive myoblast is attached to a myofibre by one pole. The elongated poles of the cell are packed with microtubules (m), but these diverge as they extend towards the nucleus (n). Some microtubules pass the nucleus (Fig. 19), but their number is insufficient to account for the population at the poles. It is therefore presumed that some microtubules originate in the vicinity of the nucleus. The plasma membrane of the cell is apparently discontinuous (arrows), but these discontinuities are interpreted as folds running around the cell, and may be associated with cellular locomotion. Parallel orientation of microtubules at the poles is achieved with a minimum separation distance of about 400 Å. ( × 30 500.)

Fig. 17.

This bipolar presumptive myoblast is attached to a myofibre by one pole. The elongated poles of the cell are packed with microtubules (m), but these diverge as they extend towards the nucleus (n). Some microtubules pass the nucleus (Fig. 19), but their number is insufficient to account for the population at the poles. It is therefore presumed that some microtubules originate in the vicinity of the nucleus. The plasma membrane of the cell is apparently discontinuous (arrows), but these discontinuities are interpreted as folds running around the cell, and may be associated with cellular locomotion. Parallel orientation of microtubules at the poles is achieved with a minimum separation distance of about 400 Å. ( × 30 500.)

Fig. 18.

In this very low magnification survey micrograph, a presumptive myoblast (pm) (sectioned transversely with respect to its bipolar long axis) lies alongside a residual myofibre (rm) before fusion. The size differential allows small nuclei of the myoblast and the large polyploid nuclei of the myofibre to be readily distinguished. The greater electron density of the myoblast cytoplasm is due to the relatively large number of ribosomes present in the myoblast. This is clearly shown in Fig. 19, which shows the contact zone of a similar cell association at higher magnification. ( × 7000.)

Fig. 18.

In this very low magnification survey micrograph, a presumptive myoblast (pm) (sectioned transversely with respect to its bipolar long axis) lies alongside a residual myofibre (rm) before fusion. The size differential allows small nuclei of the myoblast and the large polyploid nuclei of the myofibre to be readily distinguished. The greater electron density of the myoblast cytoplasm is due to the relatively large number of ribosomes present in the myoblast. This is clearly shown in Fig. 19, which shows the contact zone of a similar cell association at higher magnification. ( × 7000.)

The presumptive myoblast is a bipolar cell with an overall length of at least 25 μm, and a width of 5–8 μm at the equator. The nucleus is a spheroid of diameter 5–6 μm. The elongated poles of the cell are attenuated to the extent that they resemble microspikes (Weiss, 1961; Taylor, 1966). The poles are packed with 240 Å diameter microtubules running parallel to the long axis of the cell, with the exclusion of most other organelles. Parallel orientation is achieved with a minimum separation distance of about 400 Å (Fig. 17). Microtubules diverge as they extend towards the nucleus, and a few can be detected alongside the nucleus in the centre of the cell. It is possible that some microtubules may extend from one end of the cell to the other, but the number of microtubules alongside the nucleus is insufficient to account for the total population at the poles. There is no obvious proximal focus of microtubules, although many appear to fade in the vicinity of the nucleus. Many ribosomes are present in the cytoplasm of the presumptive myoblast, and small (average dimensions 0·8 ×0·2 μm) mitochondria are abundant near the nucleus (Fig. 17). Irregular membranous sacs and whorls, and electron lucent (lipid?) droplets, make up the organelle complement.

The topography of the plasma membrane of presumptive myoblasts is of considerable interest. In the vicinity of the nucleus it is unremarkable, with irregular slight depressions and elevations, but along the attenuated poles of the cell it becomes smooth. Between these two areas, where the plasma membrane expands to encompass the nucleus, apparent discontinuities are seen (Fig. 17, arrows). Discontinuities are spaced 0·5–1·0μm apart in longitudinal sections, and are interpreted as folds or waves passing around the cell at 90° to its long axis.

Examination of living bipolar cells by interference light microscopy shows that the cell flexes at the base of each attenuated membrane extension, whilst the extensions themselves remain rigid (Fig. 16A–D). It is likely that these movements form the basis of the cells locomotory activity, since ruffled membranes are entirely absent from these cells. Thus it is not surprising to record that plasma membrane discontinuities seen in electron micrographs are confined to the region at which the cell undergoes flexion. Attempts to demonstrate other morphological specializations for movement, such as filaments, fibrils, and the like, in preparations for the electron microscope have thus far been unsuccessful. Myofilaments have not been detected in myoblasts before or during fusion, and microtubules are the only demonstrable filamentous structures. An extracellular basement membrane is not present on the presumptive myoblast.

(4) Fusion of myoblasts with the residual myofibre

Three modes of contact between presumptive myoblast and residual myofibre can be recognized, and may represent a temporal sequence. One consists of apposition of plasma membranes, but with an indeterminate separation distance (Fig. 19). A second mode consists of close apposition with a gap of 150–250 Å between unit membranes. In the third mode cytoplasmic continuity between the cell systems is established in limited areas, giving fusion. Once such a fusion of myoblast and myofibre can be recognized, the two systems are termed simply ‘myoblast’ and ‘myofibre’.

Fig. 19.

Two modes of contact between presumptive myoblast and myofibre are seen in this micrograph. At (a) apposition of plasma membranes leaves a gap of about 300 Å, but elsewhere the gap is larger and variable. Note the similar alignment of many microtubules in the two cell systems. Ribosomes are less numerous in the myofibre than in the presumptive myoblast. ( ×32 500.)

Fig. 19.

Two modes of contact between presumptive myoblast and myofibre are seen in this micrograph. At (a) apposition of plasma membranes leaves a gap of about 300 Å, but elsewhere the gap is larger and variable. Note the similar alignment of many microtubules in the two cell systems. Ribosomes are less numerous in the myofibre than in the presumptive myoblast. ( ×32 500.)

Fusion of the myoblast with the myofibre takes place by dissolution of the plasma membranes separating the two systems in zones of close apposition. Images with zones of small (Fig. 20) and large (Fig. 21) area of confluence are obtained, and should dispel any doubt that may exist that such a fusion takes place. No particular organelles are specifically associated with the zones of confluence. No microfilaments traverse the zones, and no desmosomes are formed. Indeed, no anatomical basis for either a recognition or a fusion mechanism has been detected so far. One striking feature of the relationship between myoblast and myofibre at the time of fusion is the strict relative orientation of the microtubule array in each system. Thus the long axis of the myoblast always lies parallel to the long axis of the myofibre (Fig. 19). The development of microtubule orientation prior to, or immediately after, the formation of the first cytoplasmic bridges, suggests that the array of microtubules in the myofibre develops in response to the bundle of pre-existing microtubules in the myoblast, since the microtubules in the myofibre are without regular orientation until the fusion with myoblasts occurs. (Although the orientation of the myoblasts could itself be derived from the long axis of the myofibre.) This orientation is further discussed in an accompanying paper.

Fig. 20.

Fusion of myoblast (mb) with myofibre (mf) shows an early stage in the fusion at the two points is arrowed. Note the irregular outline of the nucleus of the myoblast, which has not yet passed into the myofibre. ( × 22500.)

Fig. 20.

Fusion of myoblast (mb) with myofibre (mf) shows an early stage in the fusion at the two points is arrowed. Note the irregular outline of the nucleus of the myoblast, which has not yet passed into the myofibre. ( × 22500.)

Fig. 21.

Fusion of myoblast and myofibre has occurred, producing a large zone of confluence. The cytoplasm of the two-cell systems is now indistinguishable, and the myoblast nucleus appears as a blister in the surface of the myofibre, (× 19000.)

Fig. 21.

Fusion of myoblast and myofibre has occurred, producing a large zone of confluence. The cytoplasm of the two-cell systems is now indistinguishable, and the myoblast nucleus appears as a blister in the surface of the myofibre, (× 19000.)

No distinction can be made between myoblast and myofibre cytoplasm once the systems become confluent. Observation of many sections suggests that the myoblast remains as a blister on the surface of the myofibre for a while (Fig. 21), but then with the passing of the myoblast nucleus into the body of the myofibre the blister subsides. The plasma membrane of the myoblast appears to become part of the plasma membrane limiting the myofibre, and the organelles of each system intermingle. Short chains of myoblast nuclei gather in the core of the myofibre, alongside the much larger nuclei derived directly from larval muscle (Fig. 18). Two anatomically distinct classes of myofibre nucleus are thus present from about 3 days after puparium formation onwards, and can be distinguished throughout subsequent phases of adult muscle development. The structure of these nuclei is considered in an accompanying paper (Crossley, 1972). There is no evidence that either nuclei derived from myoblasts, or from larval muscle, divide by mitosis or by ‘amitosis’.

Breakdown of the larval contractile apparatus

Although the larval and adult intersegmental muscles of Calliphora are extremely similar in fine structure (Crossley, 1968, 1971), a complete reorganization of the contractile machinery is apparently necessary to facilitate changes in the alignment and probably the functional regime of the muscles, that accompany the transition from larval to adult life. The adult muscle is also about one-third longer than its larval counterpart. Apparently elongation by addition of sarcomeres to existing structure is not practicable in this muscle during metamorphosis.

The earliest signs of reorganization within the muscle are the appearance of membrane aggregates and latticed cisternae, and their development is coincident with the loss of much of the organized muscle internal membrane systems. It seems probable that these structures represent a form of lipoprotein conservation, since they remain as tightly packed membranous sheets until the reorganization of the adult reticular systems gathers way, when the majority disappear (Crossley, 1972). The concept of conservation of macromolecules by insects at metamorphosis has been implied by the demonstration by Birt and Christian (1969) that very little protein synthesis goes on in the puparium of the blowfly Lucilia. The hypothesis that there is extensive direct incorporation of larval protein during adult development is supported not only by the demonstration of elaborate storage structures but also of coated vesicles in transitional Calliphora muscle, discussed below.

The extracellular basement lamella of early transitional muscles is closely applied to the sarcolemma only at the sites of dense plaques or hemidesmosomes. These structures thus represent sites of attachment of the basement lamella, and it would be interesting to know if they also represent the sites of secretion. Both lamellae and plaques disappear together before the period of myoblast fusion. Ashhurst (1968) has reviewed our knowledge of the connective tissues of insects, and shown that almost nothing is known about their mode of formation. Whitten (1962) showed that connective tissue membranes ceased to envelop many tissues of a blowfly in the pupal environment.

The pattern of dissolution of larval contractile elements is the inverse of the pattern of adult myofibril formation in the same muscles (Crossley, 1972). Thus the thick primary filaments disappear first, leaving Z-discs with attached thin secondary filaments. Similar results were obtained for Rhodnius by Auber-Thomay (1967). The extent of molecular demolition that myofilaments undergo in metamorphosing muscle is unknown, but it is likely that most protein molecules remain intact, as discussed above. Lysosomes and lysosomal acid hydrolases are sparsely distributed in normal muscle (Weinstock & lodice, 1969). Furthermore, there is no evidence that myofilaments are isolated within a vacuolar apparatus during autolysis, either from the present investigation, or from earlier work on other insects (Stegwee et al. 1963 ; Lockshin & Williams, 1964; Lockshin, 1969 b). Lockshin & Williams (1965a, d) postulate release of lysosomal enzymes directly into the muscle cytoplasm by rupture of vacuole membranes. Although this idea is contrary to experience in a wide range of cell systems, where enclosure in a vacuole is the prelude to catabolism (de Duve & Wattiaux, 1966), it is conceivable that a series of enzymes specific for muscle proteins could be released directly into muscle cytoplasm. Such enzymes might not be detectable by standard techniques for lysosomal enzymes and thus could account for the absence of Gomori-positive vacuoles in the present material. This area is currently under further investigation.

Recently Lockshin (1969a) has shown that protein synthesis is necessary to initiate muscle breakdown in saturniid silkmoths, and suggests that this protein is capable of ‘activating’ lysosomal enzymes. However, he points out that since cytolysomes or the like have not as yet been reported for normal insect muscles, the morphology of ‘active’ lysosomes in insect muscle is currently unknown. Randall (1970) subsequently reported acid-phosphatase positive granules in denervated Galleria muscle, but it is not known whether these granules occur in normal metamorphosing muscle.

It is emphasized that the breakdown process in Calliphora muscles, which transform ultimately into adult muscles, does not involve the haemocytes. A different process of breakdown, involving phagocytosis, applies to muscles entirely destroyed during metamorphosis (Crossley, 1968).

Two pieces of circumstantial evidence suggest some exchange of material between residual myofibre and haemolymph: the formation of blebs at the myofibre surface, and the presence of coated vesicles. Although images showing blebs do not provide conclusive evidence of loss of myofibre cytoplasm to the haemolymph, they do support a directional interpretation when considered in the context of the general reduction of myofibre volume at this time (Crossley, 1965). A similar reduction in myofibre volume occurs during degeneration of Rhodnius intersegmental muscles (Wigglesworth, 1956; Auber-Thomay, 1967). Attempts to identify the granules commonly contained within blebs using enzymes such as ribonuclease proved inconclusive. Identification of the granules as ribosomes rests only on their affinity for uranyl acetate after cacodylate buffered glutaraldehyde fixation. Glycogen shows little affinity for uranyl acetate after this method of fixation. The possibility that ribosomes may be particularly liable to expulsion is intriguing, because it suggests that they either ‘wear out’ or are unsuitable or unnecessary for the assembly of adult muscle proteins. The last possibility is consistent with the hypothesis of direct larval protein incorporation discussed above. In degenerating Rhodnius intersegmental muscle, blebs termed ‘axial vacuoles’ by Auber-Thomay (1967) contribute to the reduction of fibre volume. Such vacuoles contain myelin figures thought to be the remains of mitochondria.

Coated vesicles are not commonly encountered in muscle cells. They have been described in rat myoblasts by Heuson-Stiennon (1965), where they are disposed in relation to Z-discs in adjacent myoblasts. The coated vesicles are also seen near the Z-line in the ventricular papillary muscle of the cat myocardium (Fawcett & McNutt, 1969). In many other tissues the presence of coated vesicles has been associated with uptake of proteins by the cell (see Fawcett (1965) for review). Friend & Farquhar (1967) found that in the rat vas deferens coated vesicles may function not only as heterophagosomes for transport of absorbed protein to lysosomes, but also as transport vesicles to convey enzymes or surface coat material made within the cell to the cell surface. It seems unlikely that coated vesicles are involved in the deposition of surface coat material in the Calliphora residual myofibre, since the basement lamella is absent at the time of maximum vesicle activity. No periodicity has been detected in the arrangement of coated vesicles along the myofibre. Furthermore, since they disappear long before the Z-discs form in adult muscle, a role in the delimitation of adult sarcomeres seems improbable. Rather it seems likely that the vesicles are concerned with uptake of nutrients from the haemolymph, at a time shortly before reconstruction of the adult muscle begins.

Latticed cisternae

A survey of the literature reveals that a fibrillar lattice similar to that in Calliphora transitional myofibres is present in a variety of animal and plant tissues. The earliest reports are those of Chandler (1966) and Chandler & Willis (1966) on an intranuclear fibrillar lattice in rat neurons, in which parallel fibrils about 70 Å in diameter occur in flat or curved sheets. Two parallel layers of fibrils cross each other at an angle of 60° or more to form a lattice which strongly resembles that seen in Calliphora, except that it is not deposited on a membrane. A helical appearance is also obtained in the rat neuron, but is thought to be due to the crossed array of fibrils curving within the thickness of the section. Another somewhat similar configuration was described by Newcomb, Steer, Hepler & Wergin (1968) in bean root mitochondria as an ‘atypical crista resembling a tight junction’. Parallel rods with a separation distance of 120–170 Å overlap to form a diamond-shaped pattern strikingly similar to the configuration in Calliphora, where a comparable measurement would be 140 Å. The rods in the bean root are, however, of smaller diameter (40–50 Å) than those of Calliphora ( > 70 Å). Structures similar or identical to those of the bean root have been described in the mitochondria of liver parenchymal cells of rats kept on a vitamin-deficient diet, where filaments on the cristae are described as helical (Djaczenko, Grabska, Urbanowicz & Rezzi, 1969).

A structure closely related to that described here has been reported in one other type of muscle, the denervated skeletal muscle of the rat (Miledi & Slater, 1969). This report is of particular relevance for the present work, not only because of the anatomical comparison with Calliphora, but also because of the suggestion that the structure forms as the result of denervation. Miledi & Slater fixed their material in osmium or permanganate solutions, and found that the densities on the cisternae were not preserved by permanganate. They supposed the dense material to include a ribonucleoprotein by virtue of its affinity for uranyl ions. Both these observations are true also for Calliphora latticed cisternae. No dimensions are stated for the rat configuration, and the threedimensional structure has not been investigated, although it is termed a ‘helical complex’ by Miledi & Slater. Measurements taken from the published micrographs reveal no size differences between the rat and insect configurations, and the anatomy of the structures may prove to be identical. The possibility that the Calliphora system is itself helical has been tested with models and found to be inconsistent with many profiles obtained by sectioning (e.g. Figs. 13, 14), a conclusion also reached by Chandler & Willis (1966) for their material. The ‘tubular mass’ described by Miledi & Slater (1969) in denervated rat muscle is also found in Calliphora transitional muscle (Fig. 12).

The hypothesis that latticed cisternae are formed in response to denervation is attractive, since it has been demonstrated that breakdown of specific lepidop-teran muscles is triggered by denervation (Finlayson, 1956, 1960; Lockshin & Williams, 1964, 1965 c–d). However, in Calliphora it has been shown that total muscle breakdown by haemocyte phagocytosis cannot be artificially triggered by experimental denervation (Crossley, 1968). These experiments are not, however, directly applicable to transitional muscle, which is subject to autolysis rather than phagocytosis. Nevertheless, muscles developing latticed cisternae do retain an anatomical nervous connexion. Continuing innervation of Rhodnius muscles during involution has also been reported by Auber-Thomay (1967), and even in intersegmental muscles of the silkmoth which are totally destroyed, the presynaptic region of the neuromuscular junction does not degenerate (Lockshin & Williams, 1965 a). In Calliphora, innervation to resorbing intersegmental muscles is retained (Fig. 1), but the presence of structures reminiscent of isolation vacuoles in the axon in the vicinity of synapses is suggestive of changed nervous function. A reasonable hypothesis, capable of experimental verification, would be that transitional muscles in Calliphora fail to receive a nervous signal necessary for muscle maintenance, in spite of continuing anatomical connexion to a nerve.Cessation of the maintenance signal would then initiate transition, including formation of latticed cisternae and tubular masses. This hypothesis is supported by the work of Johnson (1959) where breakdown of aphid flight muscles and hypertrophy of fat body were shown to be initiated in response to abnormal afferent stimuli. The importance of the nervous system for the maintenance of normal muscle function in insects has been emphasized in a recent review by Nuesch (1968). Randall (1970) has shown that in Galleria, in an appropriate humoral milieu, denervation produces ultrastructural changes which resemble metamorphic degeneration.

There is no information bearing directly on the function of latticed cisternae, but formation in the sites formerly occupied by sarcoplasmic reticulum suggests that latticed cisternae are derived from this structure.

Myoblast structure

Insect myoblasts have been described using phase-contrast optics as spindleshaped mononucleate cells capable of mitosis (Cowden & Bodenstein, 1961; Crossley, 1965; Auber, 1969; Kurtti & Brooks, 1970). In vertebrate cell cultures, stable bipolar shape has been shown to be predictably associated with myogenic capability (Konigsberg, 1963). Fibroblast cells can be distinguished by the presence of a large area of ruffled membrane, and lack of inherent polarity (Abercrombie, 1961). Vertebrate myoblasts have at most a small area of ruffled membrane confined to one tip of the cell (Konigsberg, 1963), and no ruffled membrane is apparent in insect myoblasts cultured in vitro (Kurtti & Brooks, 1970). In Calliphora long membranous extensions arise from several types of blood cell (Crossley, 1964), from moving presumptive myoblasts, and from presumptive myoblasts recently attached to myofibres (Crossley, 1965). Only on the myoblasts are the membranous extensions stable. Spindle-shaped blood cells appear to be akin to fibroblasts in that their extensions can be withdrawn or converted into ruffled membranes.

In the electron microscope it can be seen that the membranous extensions of presumptive myoblasts contain a core of closely packed microtubules. The cellular asymmetry, like so many others, is associated with microtubule deposition (see review by Porter, 1966), although recent work suggests that microtubules may be essential for the development rather than the maintenance of asymmetry in cells (Tilney & Gibbins, 1969). A fascicle of microtubules has been demonstrated in microspikes from cultured vertebrate cells (Taylor, 1966), and in grasshopper embryonic cells (Kessel & Eichler, 1966).

Recent work suggests that microtubules may be polymerized at ‘organizing centres’, sometimes associated with centrioles (Tilney & Gibbins, 1969; Pickett-Heaps, 1970). Transverse sections of the Calliphora myoblast show that the number of microtubules alongside the nucleus is less than the number at the poles, indicating that microtubules do not run from end to end of the cell, and some presumably terminate in the vicinity of the nucleus. Since focal centres have not been encountered near the nucleus, and since the extreme tip of the cell has not been encountered in sections, the question of microtubule origins remains unresolved at present.

It is suggested that the wave-like folds at the base of each membranous extension are associated with the locomotory activity of the cell, which involves movement of rigid cell extensions, rather than the development of ruffled membranes. It is emphasized that no myofilaments are present in the Calliphora myoblast before fusion with the myofibre (cf. myoblasts giving rise to the flight muscles in Drosophila (Shafiq, 1963 a, b)). Nor are any other filamentous elements other than microtubules encountered in these cells. Any locomotory activity appears to be generated by deformations of the cell membrane in ways that are at present not understood. It is clear, however, that myoblasts do not develop the extensive ruffled membranes that are associated with movement of cells over a surface (Abercrombie & Middleton, 1968). The spindle shape of the myoblast would be expected to facilitate passage of the cell floating in the blood stream amidst a throng of predominantly spherical cells, at a time when the circulation is very sluggish.

As in Drosophila (Shafiq, 1963 a, b), the cytoplasm of presumptive myoblasts is filled with ribosomes, with a few small rod-shaped mitochondria. Myoblasts in Drosophila and Calliphora are a similar size and shape, with nuclei about 6 μm in diameter.

Myoblast fusion

Once myoblasts have settled on the surface of myofibres, the cell separation distances are similar to those obtained for normal adjacent cells, i.e. 100–200 Å (Robertson, 1960; Curtis, 1967), or cells agglutinated by antibodies (Easty & Mercer, 1962). A gap of 300 Å is reported by Firket (1967) to exist between myoblasts about to fuse in the chick. Some process of recognition of myoblast and myofibre must precede fusion, but ultrastructural manifestations of this process are lacking. It may be that the reduction of the thickness of the basement membrane that precedes myoblast fusion is a prerequisite for successful fusion, since in Rhodnius and chick muscle cells about to fuse the basement membrane is absent (Toselli & Pepe, 1968; Fischman, 1967). Alignment of the elongated myoblast axis with the myofibre precedes fusion in Calliphora, and in Rhodnius (Toselli & Pepe, 1968 b).

A large body of evidence supports the development of multinucleation in vertebrate myotubes by fusion of myoblasts (Lash, Holtzer & Swift, 1957; Stockdale, Okazaki, Nameroff & Holtzer, 1961; Capers, 1960; Konigsberg, 1963; Firket, 1967; Fischman, 1967). Fusion has recently been reported for insect myoblasts cultured in vitro (Kurtti & Brooks, 1970). The development of multinucleation in the Calliphora myofibre involves a very similar fusion process directly comparable with vertebrate myoblast-myotube fusion.

Larval cells in insects frequently become highly polyploid (Wigglesworth, 1966), and the nuclei derived from larval muscle present in the residual myofibre are no exception. On the other hand, the nuclei derived from imaginal discs appear to be diploid, and the resulting syncytium is therefore host to nuclei of different ploidy. There is no evidence for multiplication of nuclei within the Calliphora myofibre by either mitosis or ‘amitosis’. This accords with evidence obtained from studies of vertebrate material (Kelly & Zacks, 1969). Capers (1960) has demonstrated by means of time-lapse cinematography how misleading images suggestive of amitosis can be. Thus earlier reports of amitosis in the literature, revived by the recent work of Eigenman (1965), require confirmation by quantitative estimation of nucleic acid distribution in myoblast nuclei.

No affiliation of cytoplasm formerly associated with the myoblast nucleus can be detected in the syncytium. It is presumed that specific communication between the particular group of organelles formerly associated with the myoblast nucleus, and the nucleus itself, is lost in favour of a co-operative system. The formation of chains of myoblast nuclei, involving migration and orientation manouevres may be under the control of microtubules. The role of microtubules in the alignment of nuclei has been demonstrated in virus-induced kidney cell syncytia by Holmes & Choppin (1968). Alignment of chains of nuclei in the long axis of the muscle is the first of a series of morphogenetic changes that ultimately lead to the formation of contractile adult muscle, a process that is described in an accompanying paper (Crossley, 1972).

This study was begun during tenure of a Queen Elizabeth II fellowship at the C.S.I.R.O., Division of Entomology, Canberra, A.C.T., Australia.

Abercrombie
,
M.
(
1961
).
The bases of the locomotory behaviour of fibroblasts
.
Expl Cell Res. (Suppl
.)
8
,
188
198
.
Abercrombie
,
M.
&
Middleton
,
C. A.
(
1968
).
Epithelial-mesenchymal interactions affecting locomotion of cells in culture, ch
.
3
, pp.
56
63
. In
Epithelial-mesenchymal Interactions
(ed. Fleischmajer).
Baltimore
:
Williams & Wilkins Co
.
Ashhurst
,
D. E.
(
1968
).
The connective tissue of insects
.
Ann. Rev. Ent.
13
,
45
74
.
Auber
,
J.
(
1969
).
La myofibrillogenèse du muscle strie. I. Insectes
.
J. Microscopie
8
,
197
232
.
Auber-Thomay
,
M.
(
1967
).
Modifications ultrastructurales au cours de la dégénérescence et de la croissance de fibres musculaires chez un insecte
.
J. Microscopie
6
,
627
638
.
Barra
,
T.
(
1964
).
Electron histochemical localization of acid phosphatase activity in the small intestine of the mouse
.
J. Histochem. Cytochem.
12
,
229
238
.
Birt
,
L. M.
&
Christian
,
B.
(
1969
).
Changes in nitrogenous compounds during the metamorphosis of the blowfly, Lucilia caprina.
J. Insect Physiol.
15
,
711
719
.
Capers
,
C. R.
(
1960
).
Multinucleation of skeletal muscle in vitro.
J. biophys. biochem. Cytol.
7
,
559
566
.
Chandler
,
R. L.
(
1966
).
Intranuclear structures in neurons
.
Nature, Lond.
209
,
1260
1261
.
Chandler
,
R. L.
&
Willis
,
R.
(
1966
).
An intranuclear fibrillar lattice in neurons
.
J. Cell Sci.
1
,
283
286
.
Cowden
,
R. R.
&
Bodenstein
,
D.
(
1961
).
A cytochemical investigation of striated muscle differentiation, in regenerating limbs of the roach Periplaneta americana.
Embryology
6
,
36
50
.
Crossley
,
A. C.
(
1964
).
An experimental analysis of the origins and physiology of haemocytes in the blue blow-fly Calliphora erythrocephala.
J. exp. Zool.
157
,
375
98
.
Crossley
,
A. C.
(
1965
).
Transformations in the abdominal muscle of the blue blow-fly, Calliphora erythrocephala (Meig.), during metamorphosis
.
J. Embryol. exp. Morph.
14
,
89
110
.
Crossley
,
A. C.
(
1968
).
The fine structure and mechanism of breakdown of larval intersegmental muscles in the blowfly Calliphora erythrocephala.
J. Insect Physiol.
14
,
1389
1407
.
Crossley
,
A. C.
(
1972
).
Ultrastructural changes during transition of larval to adult muscle at metamorphosis in the blowfly Calliphora erythrocephala. II. The formation of adult muscle
.
J. Embryol. exp. Morph.
27
,
75
101
.
Curtis
,
A. S. G.
(
1967
).
The Cell Surface.
London
:
Logos Press
,
De Bruyne
,
C.
(
1898
).
Recherches au sujet de l’intervention de la phagocytose dans le development d’invertebres
.
Archs Biol.
15
,
181
.
De Duve
,
C.
&
Wattiaux
,
R.
(
1966
).
Functions of lysosomes
.
A. Rev. Physiol.
28
,
435
491
.
Djaczenko
,
W.
,
Grabska
,
J.
,
Urbanowicz
,
M.
&
Rezzi
,
R.
(
1969
).
Peculiar mitochondrial forms in the liver parenchymal cells of rats kept on a Vitamin E deficient diet
.
J. Microscopie
8
,
139
144
.
Easty
,
G. C.
&
Mercer
,
E. H.
(
1962
).
An electron-microscope study of model tissues formed by the agglutination of erythrocytes
.
Expl Cell Res.
28
,
215
227
.
Edwards
,
F. J.
(
1969
).
Development and histolysis of the indirect flight muscles in Dysdercus intermedias.
J. Insect Physiol.
15
,
1591
1600
.
Eigenmann
,
R.
(
1965
).
Untersuchungen fiber die Entwicklung der dorsolongitudinalen Flugmuskeln von Antheraea pernyi Guer. (Lepidoptera)
.
Rev. suisse Zool.
72
,
789
840
.
Fawcett
,
D. W.
(
1965
).
Surface specialization of absorbing cells
.
J. Histochem. Cytochem.
13
,
75
91
.
Fawcett
,
D. W.
&
McNutt
,
N. S.
(
1969
).
The ultrastructure of the cat myocardium. II. Atrial muscle
.
J. Cell Biol.
42
,
46
67
.
Finlayson
,
L. H.
(
1956
).
Normal and induced degeneration of abdominal muscles during metamorphosis in the lepidoptera
.
Q. Jl microsc. Sci.
97
,
215
.
Finlayson
,
L. H.
(
1960
).
A comparative study of the effects of denervation on the abdominal muscles of saturniid moths during pupation
.
J. Insect Physiol.
5
,
108
119
.
Firket
,
H.
(
1967
).
Ultrastructural aspects of myofibrils formation in cultured skeletal muscle
.
Z. Zellforsch. mikrosk. Anat.
78
,
313
327
.
Fischman
,
D. A.
(
1967
).
An electron microscope study of myofibril formation in embryonic chick skeletal muscle
.
J. Cell Biol.
32
,
557
575
.
Friend
,
D. S.
&
Farquhar
,
M. G.
(
1967
).
Functions of coated vesicles during protein absorption in the rat vas deferens
.
J. Cell Biol.
35
,
357
376
.
Gomori
,
G.
(
1950
).
An improved histochemical technic for acid phosphatase
.
Stain Technol.
25
,
81
85
.
Heusson-Stiennon
,
J-A.
(
1965
).
Morphogenèse de la cellule musculaire striee étudiée au microscope électronique. I. Formation des structures fibrillaires
.
J. Microscopie
4
,
657
678
.
Hewitt
,
C. G.
(
1914
).
The House Fly.
London
:
Cambridge University Press
.
Holmes
,
K. V.
&
Choppin
,
P. W.
(
1968
).
On the role of microtubules in movement and alignment of nuclei in virus-induced syncitia
.
J. Cell Biol.
39
,
526
543
.
Johnson
,
B.
(
1959
).
Studies on the degeneration of the flight muscles of alate aphids. II. Histology and control of muscle breakdown
.
J. Insect Physiol.
3
,
367
377
.
Kelly
,
D. E.
(
1969
).
The fine-structure of skeletal muscle triad junctions
.
J. Ultrastruct. Res.
29
,
37
49
.
Kelly
,
A. M.
&
Zacks
,
S I.
(
1969
).
The histogenesis of rat intercostal muscle
.
J. Cell Biol.
42
,
135
153
.
Kessel
,
R. G.
&
Eichler
,
V. B.
(
1966
).
Microtubules in the microspikes and cortical cytoplasm of grasshopper embryonic cells
.
J. Microscopie
5
,
781
786
.
Konigsberg
,
I. R.
(
1963
).
Clonal analysis of myogenesis
.
Science, N. Y.
140
,
1273
.
Kurtti
,
T. J.
&
Brooks
,
M. A.
(
1970
).
Growth and differentiation of Lepidopteran myoblasts in vitro.
Expl Cell Res.
61
,
407
412
.
Lash
,
J. W.
,
Holtzer
,
H.
&
Swift
,
H.
(
1957
).
Regeneration in mature skeletal muscle
.
Anat. Rec.
128
,
679
698
.
Locke
,
M.
(
1966
).
Isolation membranes in insect cells at metamorphosis
.
J. Cell Biol.
31
,
132
A.
Lockshin
,
R. A.
(
1969a
).
Programmed cell death. Activation of lysis by a mechanism involving the synthesis of protein
.
J. Insect Physiol.
15
,
1505
1516
.
Lockshin
,
R. A.
(
1969b
).
Lysosomes in insects. Ch. 13
In
Lysosomes in Biology and Pathology.
(Eds.
Dingle
,
J. T.
and
Fell
,
H. B.
)
London
:
North Holland Publishing Co
.
Lockshin
,
R. A.
&
Williams
,
C. M.
(
1964
).
Programmed cell death. -II. Endocrine potentiation of the breakdown of intersegmental muscles of silk moths
.
J. Insect Physiol.
10
,
643
649
.
Lockshin
,
R. A.
&
Williams
,
C. M.
(
1965a
).
Programmed cell death. I. Cytology of degeneration in the intersegmental muscles of the pernyi silkmoth
.
J. Insect Physiol.
11
,
123
133
.
Lockshin
,
R. A.
&
Williams
,
C. M.
(
1965b
).
Programmed cell death. III. Neural control of the breakdown of the intersegmental muscles of silkmoths
.
J. Insect Physiol.
11
,
601
610
.
Lockshin
,
R. A.
&
Williams
,
C. M.
(
1965c
).
Programmed cell death. IV. The influence of drugs on the breakdown of the intersegmental muscles of silkmoths
.
J. Insect Physiol.
11
,
803
809
.
Lockshin
,
R. A.
&
Williams
,
C. M.
(
1965d
).
Programmed cell death. V. Cytolytic enzymes in relation to the breakdown of the intersegmental muscles of silkmoths
.
J. Insect Physiol.
831
844
.
Lowne
,
B. T.
(
1870
).
The Anatomy and Physiology of the Blowfly.
London
:
van Voorst
.
Miledi
,
R.
&
Slater
,
C. R.
(
1969
).
Electron-microscope structure of denervated skeletal muscle
.
Proc. R. Soc. Lond. B
174
,
253
269
.
Miller
,
F.
(
1960
).
Hemoglobin absorption by cells of the proximal convoluted tubules in mouse kidney
.
J. biophys. biochem. Cytol.
8
,
689
718
.
Newcomb
,
E. G.
,
Steer
,
M. W.
,
Hepler
,
P. K.
&
Wergin
,
W. P.
(
1968
).
An atypical crista resembling a ‘tight junction’ in bean root mitochondria
.
J. Cell Biol.
39
,
35
42
.
Nuesch
,
H.
(
1968
).
The role of the nervous system in insect morphogenesis and regeneration
.
A. Rev. Ent.
13
,
27
44
.
Perez
,
C.
(
1910
).
Recherches histolologiques sur la metamorphoses des muscides, Calliphora erythrocephala (Meig
.).
Archs Zool. exp. gén. (5)
4
,
1
274
.
Pickett-Heaps
,
J.
(
1970
).
The evolution of the mitotic apparatus
.
Cytobios
3
,
257
280
.
Porter
,
K. R.
(
1966
).
Cytoplasmic microtubules and their functions. In Ciba Fdn. Symp
.
Principles of Biomolecular Organizations
(Ed. Wolstenholme and O’Connor), pp.
308
356
.
London
:
J. & A. Churchill
.
Randall
,
W.
(
1970
).
Ultrastructural Changes in the proleg retractor muscles of Galleria mellonella after denervation
.
J. Insect Physiol.
16
,
1927
1943
.
Reynolds
,
E. S.
(
1963
).
The use of lead citrate at high pH as an electron-opaque stain in electron microscopy
.
J. Cell Biol.
17
,
208
212
.
Robertson
,
J. D.
(
1960
).
The molecular structure and contact relationship of cell membranes
.
Prog. Biophys. biophys. Chem.
10
,
343
418
.
Schmidt
,
E.
(
1928
).
Die Histolyse und Histogenèse der Muskulatur von Psychoda alternata (Say
.).
Z. Morphol. Ökol. Tiere
13
,
117
143
.
Shafiq
,
S. A.
(
1963a
).
Electron-microscopic studies on the indirect flight muscles of Drosophila melanogaster. I. Structure of myofibrils
.
J. Cell Biol.
17
,
351
362
.
Shafiq
,
S. A.
(
1963b
).
Electron-microscopic studies on the indirect flight muscles of Drosophila melanogaster. II. Differentiation of myofibrils
.
J. Cell Biol.
17
,
363
374
.
Snodgrass
,
R. E.
(
1924
).
Anatomy and metamorphosis of the apple maggot Rhagoletis pomonella.
J. Agrie. Res.
28
,
1
33
.
Stegwee
,
D.
,
Kimmel
,
C.
,
De Boer
,
J. A.
&
Henstra
,
S.
(
1963
).
Hormonal control of reversible degeneration of flight muscle in the Colorado potato beetle Leptinotarsa decimlineata.
J. Cell Biol.
19
,
519
527
.
Stockdale
,
F.
,
Okazaki
,
K.
,
Nameroff
,
M.
&
Holtzer
,
H.
(
1961
).
5-Bromodeoxyuridine: effect on myogenesis in vitro.
Science, N. Y.
146
,
533
.
Taylor
,
A. C.
(
1966
).
Microtubules in the microspikes and cortical cytoplasm of isolated cells
.
J. Cell Biol.
28
,
155
168
.
Tiegs
,
O. W.
(
1955
).
The flight muscles of insects - their anatomy and histology: with some observations on the structure of striated muscle in general
.
Phil. Trans. R. Soc. Lond. B
238
,
221
359
.
Tilney
,
L. G.
&
Gibbins
,
J. R.
(
1969
).
Microtubules in the formation and development of the primary mesenchyme in Arbacia punctulata. II. An experimental analysis of their role in development and maintenance of cell shape
.
J. Cell Biol.
41
,
227
250
.
Toselli
,
P. A.
&
Pepe
,
F. A.
(
1968a
).
The fine structure of the ventral intersegmental abdominal muscles of the insect Rhodnius prolixus during the molting cycle. I. Muscle structure at molting
.
J. Cell Biol.
37
,
445
461
.
Toselli
,
P. A.
&
Pepe
,
F. A.
(
1968b
).
The fine structure of the ventral intersegmental abdominal muscles of the insect Rhodnius prolixus during the molting cycle. II. Muscle change in preparation for molting
.
J. Cell Biol.
37
,
462
481
.
Van Rees
,
J.
(
1888
).
Beitrâge zur Kenntnis der inneren Metamorphose von Musca vomitoria.
Zool. Jahrb. Abth. f. Anat. Ontog.
3
,
1
134
.
Warren
,
R. H.
&
Porter
,
K. R.
(
1966
).
An electron-microscope study of developing molting muscles in Rhodnius prolixus.
J. Cell Biol.
31
,
121
A.
Warren
,
R. H.
&
Porter
,
K. R.
(
1969
).
An electron-microscope study of differentiation of the molting muscles of Rhodnius prolixus.
Am. J. Anat.
124
,
1
30
.
Weinstock
,
I. M.
&
Iodice
,
A. A.
(
1969
).
Acid hydrolase activity in muscular dystrophy and denervation atrophy
. In
Lysosomes in Biology and Pathology.
(Eds.
Dingle
,
J. T.
&
Fell
,
H. B.
)
London
:
North Holland Publishing Co
.
Weiss
,
P.
(
1961
).
From cell to molecule
. In
The Molecular Control of Cellular Activity,
(Ed.
J. M.
Allen
), pp.
1
72
.
New York
:
McGraw-Hill
.
Whitten
,
J. M.
(
1962
).
Breakdown and formation of connective tissue in the pupal stage of an insect
.
Q. Jl microsc. Sci.
103
,
359
367
.
Wigglesworth
,
V. B.
(
1956
).
Formation and involution of striated muscle fibres during the growth and moulting cycles of Rhodnius prolixus (Hemiptera)
.
Q. Jl microsc. Sci.
97
,
465480
.
Wigglesworth
,
V. B.
(
1966
).
Polyploidy and nuclear fusion
.
Nature, Lond.
212
,
1581
.

Note. The number encircled in the bottom right-hand corner of each figure is the time interval, in hours, from formation of the puparium to the fixation of the tissue.