Myoblasts obtained by trypsin dissociation of 6-day chick-embryo hearts re-aggregated in rolling tubes and formed a pulsating mass within 1 h.

Newly formed intercalated discs (adhesion plaques) were the most frequent type of inter-cellular contact in the earlier stages of culture.

Desmosomes were also present.

Focal tight junctions were rare and difficult to identify.

In advanced aggregates more extended regions of very close plasma membranes apposition could be observed; lanthanum infiltration did not reveal obliteration of the intercellular gap.

The most striking feature, even after 12 days of culture, is the disorder of the contractile units within the sarcoplasm and the irregularity of the cell outlines. The ineffectiveness of the factors responsible for the orientated disposition of myofibrils in culture conditions has been emphasized.

The smooth endoplasmic reticulum never attains a regular arrangement in relationship with the myofibrillar banding. Subsarcolemmal cysterns containing a finely granular matrix (peripheral couplings) have been found after the first 24 h of culture; it has been observed that they reach their mature form only in the last stages of culture.

The T-system is lacking. Numerous pits, often complicated by beading, are present in the peripheral sarcoplasm.

All these features may affect the physiological response that can be recorded from these cells.

Trypsin-dissociated cells of an embryonic organ cluster together again when placed in a favourable medium, re-establishing a structure similar to a tissue, and reassume the characteristics of their histological development (Moscona, 1952; Stefanelli & Zacchei, 1958; Steinberg & Roth, 1964; Stefanelli, Zacchei & Ceccherini, 1961; Sheffield & Moscona, 1970; Sheffield, 1970). The massing of the cells into aggregates depends both on extrinsic factors, i.e. the experimental conditions (Curtis, 1970; Steinberg, 1970), and on intrinsic factors, namely the associative properties of the cells themselves (Lilien & Moscona, 1967; Burdick & Steinberg, 1969).

In studies on the initial stages of embryonic chick heart re-aggregation carried out by time-lapse cinematography (Stefanelli et al., unpublished) we were able to follow the constitution of a pulsating aggregate from the appearance of small cell clumps to the formation of an increasingly compact tissue. Also in the in vivo chick embryo, De Haan (1963) observed that groups of cells moving in an apparently random manner and independently of one another flow together in the cardiac region. In the initial stage, the clusters that form in vitro beat independently of one another and form a laminar structure; various ‘centres’ can still be observed, each with its own beat. Only later, when a nearly spherical mass of cells is constituted, does the beating become synchronized. As in the embryo (Obrecht-Coutris, Le Douarin & Coraboeuf, 1968) it is possible at this point to speak of a pacemaker activity of the ‘leading cells’. Investigations are now being carried out on the electrophysiological activity of the cardiac myoblasts during the various phases of aggregation.

The aim of the present work is to establish the morphological characterization of these cells and to obtain data correlating physiological and structural aspects. Hitherto the large amount of data available concerning the electrical activity of cardiac myoblasts, even in long-term cultures (Harary & Farley, 1963; Mark & Strasser, 1966; Lehmkuhl & Sperelakis, 1965; Olivo & Basa, 1967; De Haan & Gottlieb, 1968), has not been adequately correlated with the fine structure of cardiac cells in vivo, either in the fowl or mammals, or with myoblasts in short-term cultures (Cedergren & Harary, 1964; Muscatello, Pasquali-Ronchetti & Barasa, 1968; Fishman & Moscona, 1969).

Cell suspensions of heart myoblasts from 6-day embryos of white Wiandotte chickens were prepared according to the method of De Haan (1967), briefly outlined below.

After removing the blood vessels the hearts were washed in DBSS-K (potassium-free balanced salt solution), cut into fragments of 0·5–1 mm and incubated for 10 min at 37°C in a trypsin solution in CMF (calcium- and magnesium-free phosphate buffered saline) under magnetic stirring.

The supernatant was discarded and disaggregation then proceeded in three steps of 8 min each; the medium with the floating cells was collected each time and added to 10% horse serum in DBSS-K at 20°C, whereas the tryspin was renewed in the vessels with the tissue fragments.

The cell suspension was filtered through a silk sieve by applying positive pressure and centrifuged at 300 g for 10 min. The myoblasts resuspended in the culture medium were counted with a haemocytometer to obtain a final cell dilution of 6 × 105/ml.

Two culture media were employed. One, known as 629 A, consisted of a mixture of 75% DBSS-K, 20% M 199 (medium 199, Grand Island Biological Company), 4% horse serum (inactivated for 30 min at 56°C) and 1% antibiotics (penicillin and streptomycin). The other was 629 A with the addition of a 10% volume of EE (chick-embryo extract) and dialysed for 24 h against the physiological solution to stabilize the content of K-ions.

The freshly prepared media were filtered through a 0·45 μm Millipore filter. The pH of the solutions and of the media was stabilized at pH 7·3 in an atmosphere of 5% CO2. The cell suspension was placed in rolling tubes set at a speed of 10 rotations/h. The culture medium was renewed every 2 days. The beating of the cell aggregates was checked every 12 h and preparations for histological observations were made at intervals varying from 24 h to 12 days of culture.

For electron-microscopical investigations the pellets and the aggregates were fixed with 2% glutaraldehyde in 0·13M Millonig phosphate buffer pH 7·6. After an overnight washing in the same buffer they were re-fixed in 2% osmium tetroxide in Millonig buffer. Araldite was employed as embedding medium. The lead-citrate stained preparations were observed with a Hitachi HU 11 electron microscope.

Some aggregates were submitted to the lanthanum permeation according to the method of Revel & Karnovsky (1967).

In pellets obtained by centrifugation of dissociated cells from 6-day chick-embryo hearts, the myoblasts appear as rounded cells, 3–4 μm in diameter.

They have short villosities and processes and are clearly differentiated from other smaller cells which have extremely wavy outlines. These smaller cells, which lack myofilaments and glycogen granules, have fewer mitochondria and a more developed rough endoplasmic reticulum. These are presumably the fibroblast-like cells, which several authors have reported finding in the dis-sociated embryonic heart.

A certain degree of disorganization of the myofibrils is evident in the myo-blasts. Most of the myofilaments are grouped together to form short small bundles which are very occasionally crossed by opaque material of the Z-band (Fig. 1). When cut in cross-section, they do not show the characteristic hexagonal pattern. Z-material may be found in the cytoplasm isolated from the myofilaments. Associated with this material, in some cases, are fragments of plasma membrane which are invaginated in the cytoplasm and fused to form large vesicles. Those fragments of the membrane which were involved in intercellular junctions, and which have preserved the finely fibrillar substance attached to them are analogous in appearance.

Fig. 1

Pellet of myoblasts from 6-day chick-embryo heart. Roundish shape of a myoblast can be seen. The myofilaments within the sarcoplasm appear still arranged in bundles (arrows). Z-material (Z/w) separated from myofibrillar structure can be found. In the insert, a point of close apposition of plasma membranes of neighbouring cells suggesting the occurrence of focal tight junction.

Fig. 1

Pellet of myoblasts from 6-day chick-embryo heart. Roundish shape of a myoblast can be seen. The myofilaments within the sarcoplasm appear still arranged in bundles (arrows). Z-material (Z/w) separated from myofibrillar structure can be found. In the insert, a point of close apposition of plasma membranes of neighbouring cells suggesting the occurrence of focal tight junction.

Myoblasts appear as single cells but their membranes can approximate one another for a certain distance and an increased opacity of the subjacent cytoplasm may sometimes be found. The intercellular gap may be reduced or obliterated for a very short distance in correspondence to these areas (insert of Fig. 1). In addition, the cells may come into contact with short processes containing a dense thread-like material; focal tight junctions are visible at the point where the cell membranes touch.

There is no electron-dense material surrounding the plasma membranes.

The other cytoplasmatic components show no significant modification.

Helix ribosomes, which are usually fairly frequent in myoblasts at this stage of differentiation, are rarely detected whereas free ribosomes are present in large numbers.

Aggregates

Aggregates cultured for 24 h appear to consist of an outer layer of very flat cells and of an inner mass of irregularly shaped cells, with a prevailing axis (Fig. 2, insert).

Fig. 2

Three-hour aggregate of cardiac myoblasts from 7-day embryos. Inset shows a light-microscopical view of a 24 h aggregate in which external cells appear to be disposed in concentric layers surrounding the inner mass of myoblasts. In the electron micrograph the outlines of particularly well re-aggregated myoblasts are visualized by lanthanum permeation. Myofilaments are associated into more prominent bundles still not reconstituting myofibrils and are arranged without order within the cytoplasm. Where adhesion plaques are forming between cells a certain regular disposition of the bundles can be noticed as if they would converge into the membranes.

Fig. 2

Three-hour aggregate of cardiac myoblasts from 7-day embryos. Inset shows a light-microscopical view of a 24 h aggregate in which external cells appear to be disposed in concentric layers surrounding the inner mass of myoblasts. In the electron micrograph the outlines of particularly well re-aggregated myoblasts are visualized by lanthanum permeation. Myofilaments are associated into more prominent bundles still not reconstituting myofibrils and are arranged without order within the cytoplasm. Where adhesion plaques are forming between cells a certain regular disposition of the bundles can be noticed as if they would converge into the membranes.

Under the electron microscope the outer cells appear as fibroblast-like cells, which are rich in ergastoplasmic reticulum, the elements of which have a faintly opaque matrix. The Golgi complex, composed of flat and vesicular elements, also appears highly developed.

Amorphous or fibrillar extracellular material is present in the spaces between these cells.

The myoblasts in the inner part of the aggregate form a loose cellular network and where the plasma membranes run parallel and close to one another, they are separated by a space of about 20 nm, which increases slightly at the level of desmosomes.

At the ends of the cells, intercellular contacts of the type known as adhesion plaque are frequently found; in the junctional area the myofibrils are seen to attach themselves to the plasma membranes of contiguous cells by means of finely fibrillar material. These zones represent the newly forming intercalated discs (Fig. 4).

Fig. 3

Twenty-four-hour aggregate. Within the Golgi apparatus some elements show an electron-dense content. The arrow points at a subsarcolemmal element.

Fig. 3

Twenty-four-hour aggregate. Within the Golgi apparatus some elements show an electron-dense content. The arrow points at a subsarcolemmal element.

Fig. 4

Twenty-four-hour aggregate of cardiac myoblasts from 7-day embroys. A newly forming intercalated disc (id) is visible in the upper cell: it constitutes a site for insertion of myofilaments which acquire an orientated disposition, such as they do not display in other sites within the cells (arrows). In the lower cell particularly numerous tubular profiles of the smooth reticulum can be seen; the asterisk marks a distended element with a finely granular content which is similar to the subsarcolemmal elements.

Fig. 4

Twenty-four-hour aggregate of cardiac myoblasts from 7-day embroys. A newly forming intercalated disc (id) is visible in the upper cell: it constitutes a site for insertion of myofilaments which acquire an orientated disposition, such as they do not display in other sites within the cells (arrows). In the lower cell particularly numerous tubular profiles of the smooth reticulum can be seen; the asterisk marks a distended element with a finely granular content which is similar to the subsarcolemmal elements.

Focal tight junctions, when present, can be seen at intervals along the plasma membranes. The lanthanum-infiltration technique, which clearly revealed the tortuous outlines of the myoblasts, did not demonstrate the presence of any modification in the membrane structure or in the intercellular gap at these junctional areas (Fig. 2).

Even after only 24 h of culture the myofilaments in most cells have reorganized into myofibrils. Their banding is still irregular, especially in the intervals between the Z-discs. In cross-section they appear to be arranged in the typical hexagonal pattern. In most instances only 2–3 sarcomeres can be found in each section, and these longer myofibrils are usually located in close proximity to the plasma membrane and their Z-bands may be connected to this or be inserted in adhesion plaques. The contractile material in many cells, however, still appears rather disorganized even after a prolonged period of culture (Figs. 5, 6) and at a certain point the diameter of the myofibrils no longer continues to increase. ‘Z-centres’ can still be observed when the overall age of the myoblasts is 13 days.

Fig. 5

Twelve-day aggregate from 7-day embryos. The entanglement of the myoblasts and the scarce number of myofibrils lying in all directions is still remarkable after .12 days of culture. Finely fibrillar material can be seen associated with the plasma membranes, although they do not form a true intercalated disc. Arrows indicate coated vesicles, d = desmosome. Dark lines mark helix ribosomes.

Fig. 5

Twelve-day aggregate from 7-day embryos. The entanglement of the myoblasts and the scarce number of myofibrils lying in all directions is still remarkable after .12 days of culture. Finely fibrillar material can be seen associated with the plasma membranes, although they do not form a true intercalated disc. Arrows indicate coated vesicles, d = desmosome. Dark lines mark helix ribosomes.

Fig. 6

Twelve-day aggregate from 7-day embryos. The absence of a regular array of myofibrils within each cell is clearly evident. The intercalated discs mainly formed by adhesion plaques, do not have a precise relationship with the myofibrillar pattern. The arrows point to coated introflexions of the cell membranes.

Fig. 6

Twelve-day aggregate from 7-day embryos. The absence of a regular array of myofibrils within each cell is clearly evident. The intercalated discs mainly formed by adhesion plaques, do not have a precise relationship with the myofibrillar pattern. The arrows point to coated introflexions of the cell membranes.

In the 12-day aggregates the number of myofibrils parallel to one another and also parallel to the major axis of the cells is considerably increased: many contractile units, however, are found which, since they are intersected at different angles by the plane of section, show various different orientations within the cytoplasm (Fig. 6). Aggregates cultured in the presence of embryo extract showed greater cell volume when compared with those cultured alone, greater compactness of the tissue and a more advanced degree of differentiation of the cells. When the embryo extract was not included in the composition of the culture medium and the cells were more loosely packed, the myofibrils appeared even more disarranged.

The intercellular contacts, from the 24th hour onwards, increase in number on account of the greater cohesion among myoblasts. The newly formed intercalated discs hardly involve the whole ends of contiguous cells and at first they have a fairly flat shape. Later, they are more wavy in appearance and involve larger areas of the cell membranes (Fig. 6). This structure, however, is not repeated with any regularity since the tissue which is formed has no regular scheme which could be related to a geometrical model. The myoblasts are interwoven so that sometimes more than two cells contribute to the formation of an intercalated disc (Fig. 5). In addition, ‘autodesmosomes’ have been observed where the facing membranes of the same cell have closed together.

From 4–6 days of culture, short tracts of the plasma membranes of adjacent cells can be seen, with increasing frequency, to be in very close proximity to one another (Fig. 7). Only in 12-day aggregates is it possible to find more extended regions where the gap between the apposed membranes is no longer visible. They are sometimes on the surface of an intercalated disc, but are by no means a constant component of this region.

Figures 7–11

Fig. 7. Five-day aggregate. Three sites (between marks) where the intercellular gap between myoblasts in contact appear to be obliterated. These sites resemble gap junction, but the preparation procedure applied to this material does not allow their correct interpretation.

Fig. 8 Twelve-day aggregate. Flat profiles comparable with the peripheral couplings are visible in apposition to the plasma membrane. The inner dense material may appear as a dense intermediate line; an ill-defined electron-density occupies the space separating the facing membranes.

Fig. 9 Twelve-day aggregate. A coupling appears to be in connexion with a tubular profile directed toward the interior of the cell.

Fig. 10 Twelve-day aggregate. Only a segment of a tubular element shows the structural characters of the coupling. Four electron-dense bridges can be seen in the space between the apposed membranes (white marks).

Fig. 11 Six-day aggregate. Spongy appearance of the subsarcolemmal cytoplasm owing to the presence of numerous short, beaded introfiexions of the plasma membrane.

Figures 7–11

Fig. 7. Five-day aggregate. Three sites (between marks) where the intercellular gap between myoblasts in contact appear to be obliterated. These sites resemble gap junction, but the preparation procedure applied to this material does not allow their correct interpretation.

Fig. 8 Twelve-day aggregate. Flat profiles comparable with the peripheral couplings are visible in apposition to the plasma membrane. The inner dense material may appear as a dense intermediate line; an ill-defined electron-density occupies the space separating the facing membranes.

Fig. 9 Twelve-day aggregate. A coupling appears to be in connexion with a tubular profile directed toward the interior of the cell.

Fig. 10 Twelve-day aggregate. Only a segment of a tubular element shows the structural characters of the coupling. Four electron-dense bridges can be seen in the space between the apposed membranes (white marks).

Fig. 11 Six-day aggregate. Spongy appearance of the subsarcolemmal cytoplasm owing to the presence of numerous short, beaded introfiexions of the plasma membrane.

Particular attention has been given to the evolution of the internal membrane systems. The rough endoplasmic reticulum is composed of a fairly large number of elements mostly localized in the perinuclear sarcoplasm. The smooth-walled reticulum is very poorly developed after the first 24 h of culture but it later becomes more extensive and in some cases it is possible to observe networks of tubules around the myobrils. However, it often has the appearance of isolated small distended or flat profiles (Fig. 7), running parallel to the plasma membrane. Even in the 12-day aggregates a three-dimensional reconstruction of its disposition in relation to the banding of the myofibrils appears to be impossible, since the endoplasmic reticulum is not constantly associated to the Z-bands or to the other sarcomeric regions (Fig. 5). When small vesicles appear beside the Z-bands they often show a central constriction.

Cysterns were observed, close to the sarcolemma, in which ribosomes were attached only to the membrane facing the cell cytoplasm.

After the first few hours of culture, vesicular or elongated profiles containing a faintly electron-dense granular matrix can frequently be seen in the cytoplasm but only rarely in the subsarcolemmal position (Figs 3, 4).

From the third day of culture onwards, these elements, which might be compared with the peripheral couplings normally present in the cardiac muscle, become more numerous and acquire their normal subsarcolemmal location. In the early stages, they have a rather compressed shape and the inner substance is evenly distributed, whereas in 12-day aggregates they appear to be more flattened and their content has a tendency to be arranged in a dense intermediate line (Fig. 8). At this stage, they may be connected with smooth tubules directed towards the cytoplasm (Figs. 9, 10). The subsarcolemmal cysterns are separated from the plasma membrane by an intermembranous space of about 10 nm; in the younger aggregates, there are no periodic opacities bridging this space, and the apposed membranes of the couplings do not present the characteristic scalloped appearance. Facing couplings in adjacent cells are often observed.

The Golgi complex is well developed, not unlike that of the cardiac myoblasts in vivo. Elements with characteristics similar to the subsarcolemmal profiles are present in this area (Fig. 3) and can be distinguished from the dense-core granules which may also appear to be connected to the Golgi cysterns only in the more advanced aggregates.

As far as the other component of the sarcoplasmic reticulum, i.e. the transverse tubular system, is concerned, no deep introflexion of the sarcolemma was observed inside the myoblasts at any time of culture. Lanthanum permeation failed to demonstrate these tubular extensions of the extracellular space into the cells (Fig. 2), whereas it revealed fairly numerous pits and caveolae present in the peripheral sarcoplasm. In forming these invaginations, which may appear as very short beaded tubules located along the bordering cytoplasm, the surface coat follows the plasma membrane (Fig. 11). Coupling of subsarcolemmal elements with these membrane introflexions has not been observed.

Coated vesicles are a normal feature of the sarcolemmal region (Figs. 5, 6) and they are also present in the Golgi area.

Finally, a few remarks concerning the ribosomes, glycogen and mitochondria. In the first few hours of culture there is a considerable increase in the number of ribosomes and polysomes and numerous exceptionally long helix ribosomes can also be found.

Accumulations of glycogen particles occur particularly in the younger aggregates, and often at this stage of culture, they surround lipid droplets.

Mitochondria are very numerous, increased in size and sometimes branched; they have an electron-dense matrix and granules. In the more advanced aggregates, they often display a peculiar disposition of the cristae, i.e. they are concentric to the external membrane, with the matrix in the central space.

In the adult myocardium the Purkinje fibres, which form the conducting system, are easily recognized on account of their embryonic characteristics. It was therefore not possible to identify these cells in our investigations. However, we should like to draw attention to the presence of myoblasts, the cytoplasm of which appears to be empty and contains very few myofibrils and very few flat profiles of the endoplasmic reticulum. The cellular outlines are regular. At the points of contact with other myoblasts, the same junctional specializations are found, but close junctions are always present. Since this type of myoblast has come to our attention only on a few rare occasions, it has not been possible to determine whether these junctions were in fact gap or tight junctions.

We wish to emphasize three main points suggested by close examination of cardiac myoblasts which in vitro have resumed a tissue-like appearance: (1) the obvious difference between these and the embryonic myocardium in vivo, as regards the entire complex three-dimensional pattern, and its functional implication; (2) the interest of this material for the study of the assemblage of myofibrils and myofibrillar components, since these processes occur in the absence of external agents, as might be represented by the support on which monolayer cultures are grown, and also out of the influence of internal regulating factors such as nervous tissue or conducting system; (3) the endocellular organization and the types of intercellular contacts appear far from the complexity attained by the differentiated heart cells, particularly in mammal myocardium, which is the morphological model mostly referred to by electro-physiologists.

As far as the first point is concerned, lanthanum infiltration clearly demonstrated that, in vitro, even though the myoblasts assumed an essentially elongated shape, they have irregular outlines so that they form interlocking elements without following any particular schematic pattern. This condition in vivo is already lost as early as the 4th day of incubation (Caravita’& Gibertini, 1966; Manasek, 1968). The junctional contacts between cells, including regions of myofibrillar insertion, occur at almost any point along the surface of the cell so that it would be very difficult to reconcile the geometry thus resulting not only with that of mammalian cardiac muscle ‘composed of nearly cylindrical cells joined together in a single file by intercalated discs and enclosed in a basement membrane’ (Dewey, 1969) but also with that of heart muscle from lower vertebrates. It is evident that this condition implies a different spread of excitation and a different electrical behaviour. A greater interest in the electro-physiology of the myocardium of lower vertebrates will contribute to a better understanding of the problem. In fact, during the last few years the ultrastructural features of this tissue have been extensively studied in many different classes, whereas the bulk of the electrophysiological data still regard the better-known cardiac muscle of mammals, which can be more easily approximated to the theoretical models (Johnson & Sommer, 1967).

In our re-aggregated myoblasts, apart from the cell shape and the intercellular relationships the arrangement of myofibrils inside the cell differs greatly from that observed in the tissue in vivo. In fact, after 12 days of culture the number of myofibrils with a parallel disposition within the cytoplasm is still very low and they do not retain the same orientation in neighbouring cells unless they are anchored to an intercalated disc. It is possible to infer from this evident disorder that the agents responsible for the orientation of the myofibrils are less effective than in vivo. When myogenesis has been studied in vitro monolayer cultures have always been employed and even the fine-structural reports refer to myoblasts in this experimental condition (Cedergren & Harary, 1964; Muscatello et al. 1968). It can be seen that in this case the support to which the cells sediment and grow constitutes an orientating factor. Several authors (Harary & Farley, 1963; Mark & Strasser, 1966; De Haan, 1967) have reported that the disaggregated myoblasts in the suspending medium are roundish, becoming flattened, polygonal and then spindle-like only after adhesion to the glass. Later, when the striated myofibrils become visible under the light microscope, they are mostly located along the cell membranes and along the lines of tension that originate when the cell flattens and attaches itself to the glass. However, even in this condition of culture a certain disorder of the myofibrils persists and, in fact, many authors attribute the return to pace-maker activity of all the cells to this very condition.

In our re-aggregated myoblasts floating in the culture medium, the only external influence is the slow rotating motion of the roller tubes. The intrinsic factors responsible for the orienting of the myofibrillar units inside the cells should be free to act but evidently cardiac cells, when isolated from a structural and functional organization regulated by nerve fibres and guided by the system of conduction, are not capable of differentiating into a well-defined form and of constituting a unit of functional effectiveness. This behaviour seems to differ from that of the skeletal muscle fibres not having an autonomic contractile activity, which in vitro show fairly well oriented and contractile units (Shimada, Fishman & Moscona, 1967).

As far as the factors that determine the array of myofilaments in striated myofibrils are concerned, in our opinion the intercellular relationships play an important role. As soon as these are lost, i.e. after trypsin-treatment, a disorganization in the fibrillar material is observed, in agreement with the finding of Fishman & Moscona (1969). In addition the insertion of myofilaments into the plasma membrane is no longer present. However, a large number of myofilaments are still joined together in bundles; we tested the behaviour of the better differentiated myoblasts from 14-day embryos and we found that the bundles are more numerous even if the banding of myofibrils is equally lost. The action of trypsin on the Z-band material and on the contractile proteins is well known (Wollenberger, 1964), but the bonds of cohesion between myofilaments are probably less sensitive to this action. If this monodisperse condition is experimentally prolonged by preventing the re-aggregation with EDTA (Fishman, 1971), the myofibrils do not anchor to the plasma membrane. In cultures on which the cellular dilution was maintained high in order to avoid the formation of intercellular contacts, the myoblasts although they revealed an electrical activity similar to that of the interconnected cells (De Haan & Gottlieb, 1968) were far less differentiated, less vital and more sensitive to all types of treatment (Harary & Farley, 1963).

Examination of our aggregates after a few hours of culture revealed that the sarcomeric organization had been regained mainly at the level of the newly formed intercalated discs-that is to say in those regions of the plasma membranes where myofilaments find insertion sites.

The intercalated discs with a prolonged time of culture gain in extension but not much in complexity; they are slightly convoluted and often run obliquely across a myofibril not necessarily cutting a nearby myofibril at a different level. More than one contractile unit may converge in the adhesion plaque. Desmosomes may precede or follow the adhesion plaque. Very small close membrane appositions, which were observed with some frequency from the 6th day of culture, are not always found in these regions; indeed they are more often present on the lateral surface of the cells. The identification of these close junctions with nexus or gap junctions observed in mammalian heart cells is doubtful since neither permanganate fixation nor lanthanum permeation were carried out on more advanced aggregates. Gap junctions have recently been found, though in very limited number and extension in the heart of the chicken (Jewett, Sommer & Johnson, 1971) and of other vertebrates (Martinez-Palomo & Mendez, 1971) where they were thought to be absent. The tight junctions have always been regarded as the morphological expression of the intercellular pathways of low resistance through which excitation spreads making the heart tissue behave like a syncythium (for a review of the subject see Dewey, 1969; Barr, 1969). However, it has been ascertained that only in the myocardium of mammals do these junctions have an extension, a frequency and a regular distribution which enable calculations to be made on their specific resistance, found to be consistent with the hypothesis of the electrotonic spread of the action potential (Spira, 1971).

In our myoblasts which start beating within the first 24 h of culture, junctions of the type described above were not present; they are only found much later and even then rarely. Furthermore, as mentioned above, the cell characteristics are such that, as pointed out by Sperelakis (1969), these junctions are presumably not very efficient.

In the early aggregates, and also in the pellets, points of obliteration of the intercellular space between short cytoplasmic processes were sometimes present. They were comparable to the so-called focal tight junctions (Trelstad, Revel & Hay, 1966) which have been described between many types of embryonic cells known to be electrically coupled (Sheridan, 1966) and also between myoblasts in culture (Goshima, 1970; Hirakow & De Haan, 1970). In this last case they have been related to the onset of synchronized beating between cells which have just come into contact. Identification of these contacts is very difficult since they are very rare and when found are extremely small in size so that their assimilation with the tight junctions seems to us questionable. In our opinion, these transitory membrane appositions are more likely a ‘contact reaction’ of embryonic cell membranes, playing an important role in establishing cell coupling by mediating intercellular communication during development, as suggested by other authors (Bennett & Trinkaus, 1970).

Finally, a few observations on the intracytoplasmatic organization of re-aggregated myoblasts. We noticed that the sarcoplasmic reticulum is poorly developed and never constitutes a prominent and regular system with a precise distribution in relation to the banding of myofibrils, as occurs in vivo (Sommer & Johnson, 1969). It is possible that this condition depends upon the disorder of the contractile structures, since the two systems are perhaps correlated, as has been suggested by Schiaffino & Margreth (1969) for the skeletal-muscle fibres. From the 5th day of culture onwards, sarcoplasmic reticulumplasmalemmic couplings appear. At first they are characterized by a slightly flattened profile and by an intraluminar content of granular, more intensely stained material; as differentiation proceeds, they become flatter and their content assumes the aspect of a dense wavy line midway between the limiting membranes. At this stage periodic ill-defined densities can be seen bridging the space between the coupled membranes. All these characteristics are in common with the subsarcolemmal couplings described in differentiating muscle and cardiac fibres (Kelly, 1971; Sommer, 1968). Matrix material appears to accumulate parallel to the growth of the sarcoplasmic reticulum only in skeletal fibres (Schiaffino & Margreth, 1969), whereas in our myoblasts after 24 h of re-aggregation we observed that the reticulum profiles were scanty and poorly branched, but numerous isolated vesicular or oval-shaped elements with a granular matrix were present. Only later was frequent coupling with the sarcolemma observed and even later continuities with tubules of the sarcoplasmic reticulum. During their intracytoplasmatic life they were often seen in or near the Golgi region. We suggest the possibility that the junctional sarcoplasmic reticulum elements, having specific chemical and functional characters, may originate from the Golgi membrane complex and only later connect to the smooth reticulum.

The other component of the inner membrane system, the transverse tubular system, is absent at all times of culture, as it is in the heart cells of all vertebrates except mammals (Forssman & Girardier, 1970).

From the 3rd day of culture there is a steady increase in the number of short invaginations of the plasma membrane lined by its fuzzy coat; they may be belobed or present vesicles budding from them like beads; however, they remain restricted in the subsarcolemmal cytoplasm and do not have any relationship with myofibrils which may be lying in this region or with the junctional elements of the sarcoplasmic reticulum.

From our observations it may be concluded that only after several days of culture did the differentiation of the heart myoblasts attain a certain degree of structural complexity. But the endocellular organization and the intercellular relationships established in the aggregate differ from that of the tissue in vivo, displaying peculiar characters.

The electrophysiological responses that can be recorded from these myoblasts should vary significantly during the culture time parallel to the morphological changes; they should also be different from those obtained from the tissue in vivo, allowing a better understanding of the structure-function correlation.

We are greatly indebted to Mr Giuseppe Gentili and Mr Dino Scorsini for their skilful technical assistance.

The work was supported in part by a grant to the Centre of Neuroembryology from the National Research Council.

Barr
,
L.
(
1969
).
Electrical transmission between the cells of vertebrate cardiac muscle
.
In Comparative Physiology of the Heart: Current Trends
(ed.
F. V.
McCann
), pp.
102
110
.
Basel
:
Birkhäuser Verlag
.
Bennett
,
M. V. L.
&
Trinkaus
,
J. P.
(
1970
).
Electrical coupling between embryonic cells by way of extracellular space and specialized junctions
.
J. Cell Biol
.
44
,
592
611
.
Burdick
,
M. L.
&
Steinberg
,
M. S.
(
1969
).
Embryonic cell adhesiveness. Do species differences exist among warm-blooded vertebrates?
Proc. natn. Acad. Sci. U.S.A
.
63
,
1169
1173
.
Caravita
,
S.
&
Gibertini
,
G.
(
1966
).
Osservazioni sull’ultrastruttura del cuore di embrione di polio durante la miogenesi
.
Archo. ital. Anat. Embriol
.
71
,
49
68
.
Cedergren
,
B.
&
Harary
,
I.
(
1964
).
In vitro studies on single beating heart cells. VIT. Ultrastructure of the beating cell layer
.
J. Ultrastruct. Res
.
11
,
442
454
.
Curtis
,
A. S. G.
(
1970
).
On the occurrence of specific adhesion between cells
.
J. Embryol. exp. Morph
.
23
,
253
272
.
De Haan
,
R. L.
(
1963
).
Organization of the cardiogenic plate in the early chick embryo
.
Acta Embryol. Morph, exp
.
6
,
26
38
.
De Haan
,
R. L.
(
1967
).
Regulation of spontaneous activity and growth of embryonic chick heart cells in tissue culture
.
Devi Biol
.
16
,
216
249
.
De Haan
,
R. L.
&
Gottlieb
,
S. H.
(
1968
).
The electrical activity of embryonic chick heart cells isolated in tissue culture singly or in interconnected sheets
.
J. gen. Physiol
.
52
,
643
666
.
Dewey
,
M. M.
(
1969
).
The structure and function of the intercalated disc in vertebrate cardiac muscle
.
In Comparative Physiology of the Heart: Current Trends
(ed.
F. V.
McCann
), pp.
10
28
.
Basel
:
Birkhäuser Verlag
.
Fishman
,
D. A.
(
1971
).
The reversible inhibition of cardiac cell aggregation by the calcium-chelator EDTA
.
Anat. Rec
.
169
,
316
.
Fishman
,
D. A.
&
Moscona
,
A. A.
(
1969
).
An electron microscope study of in vitro dissociation and reaggregation of embryonic chick and mouse heart
.
J. Cell Biol
.
43
,
37
a.
Forssmann
,
W. G.
&
Girardier
,
L.
(
1970
).
A study of the T-system in rat heart
.
J. Cell Biol
.
44
,
1
20
.
Goshima
,
K.
(
1970
).
Formation of nexuses and eiectrotonic transmission between myocardial and FL cells in monolayer culture
.
Expl Cell Res
.
63
,
124
131
.
Harary
,
I.
&
Farley
,
B.
(
1963
).
In vitro studies on single beating rat heart cells. I. Growth and organization
.
Expl Cell Res
.
29
,
451
465
.
Hirakow
,
R.
&
De Haan
,
R. L.
(
1970
).
Synchronization and the formation of nexal junctions between isolated chick embryonic heart myocytes beating in culture
.
J. Cell Biol
.
47
,
88
a.
Jewett
,
P. H.
,
Sommer
,
J. R.
&
Johnson
,
E. A.
(
1971
).
Cardiac muscle. Its ultrastructure in the finch and humming bird with special reference to the sarcoplasmic reticulum
.
J. Cell Biol
.
49
,
50
66
.
Johnson
,
E. A.
&
Sommer
,
J. R.
(
1967
).
A strand of cardiac muscle. Its ultrastructure and the electrophysiological implications of its geometry
.
J. Cell Biol
.
33
,
103
131
.
Kelly
,
A. M.
(
1971
).
Sarcoplasmic reticulum and T tubules in differentiating rat skeletal muscle
.
J. Cell Biol
.
49
,
335
345
.
Lehmkuhl
,
D.
&
Sperelakis
,
N.
(
1965
).
Electrotonic spread of current in cultured chick heart cells
.
J. cell. comp. Physiol
.
66
,
119
133
.
Lilien
,
J. E.
&
Moscona
,
A. A.
(
1967
).
Cell aggregation: its enhancement by a supernatant from cultures of homologous cells
.
Science, N. Y
.
157
,
70
72
.
Manasek
,
F. G.
(
1968
).
Embryonic development of the heart. 1. A light and electron microscopic study of myocardial development in the early chick embryo
.
J. Morph
.
125
,
329
367
.
Mark
,
G. E.
&
Strasser
,
F. F.
(
1966
).
Pacemaker activity and mitosis in cultures of newborn rat hcirt ventricle cells
.
Expl Cell Res
.
44
,
217
233
.
Martinez-Palomo
,
A.
&
Mendez
,
R.
(
1971
).
Presence of gap junctions between cardiac cells in the heart of non mammalian species
.
J. Ultrastruct. Res
.
37
,
592
600
.
Moscona
,
A. A.
(
1952
).
Cell suspensions from organ rudiments of chick embryo
.
Expl Cell Res
.
3
,
536
539
.
Muscatello
,
LL
,
Pasquali-Ronchetti
,
I.
&
Barasa
,
A.
(
1968
).
An electron microscope study of myoblasts from chick embryo heart cultured in vitro
.
J. Ultrastruct. Res
.
23
,
44
60
.
Obrecht-Coutris
,
G.
,
Le Douarin
,
G.
&
Coraboeuf
,
E.
(
1968
).
Aspects électrophysiologiques de l’automatisme du myocarde ventriculaire chez l’embryon de poulet
.
C. r. hebd. Séanc. Acad. Sci
.,
Paris
267
,
765
768
.
Olivo
,
O. M.
&
Basa
,
M.
(
1967
).
Attività contrattile coerente di cellule cardiache embrional! dissociate e coltivate in vitro
.
Boll. Soc. ital. Biol. sper
.
43
,
1083
1085
.
Revel
,
J. P.
&
Karnovsky
,
M. J.
(
1967
).
Hexagonal array of subunits in intercellular junctions of the mouse heart and liver
.
J. Cell Biol
.
33
, C7-C12.
Schiaffino
,
S.
&
Margreth
,
A.
(
1969
).
Coordinated development of the sarcoplasmic reticulum and T system during postnatal differentiation of the rat skeletal muscle
.
J. Cell Biol
.
41
,
855
875
.
Sheffield
,
J. B.
(
1970
).
Studies on aggregation of embryonic cells: initial cell adhesions and formation of intercellular junctions
.
J. Morph
.
132
,
245
265
.
Sheffield
,
J. B.
&
Moscona
,
A. A.
(
1970
).
Electron microscopic analysis of aggregation of embryonic cells: the structure and differentiation of aggregated of neural retina cells
.
Devi Biol
.
23
,
36
61
.
Sheridan
,
J. D.
(
1966
).
Electrophysiological study of special connections between cells in the early chick embryo
.
J. Cell Biol
.
31
, C1-C5.
Shimada
,
Y.
,
Fishman
,
D. A.
&
Moscona
,
A. A.
(
1967
).
The fine structure of embryonic chick skeletal muscle cells differentiated in vitro
.
J. Cell Biol
.
35
,
445
455
.
Sommer
,
J. R.
(
1968
).
Chicken cardiac muscle. A transitional stage between amphibian and mammalian cardiac muscle
.
J. Cell Biol
.
39
,
127
a.
Sommer
,
J. R.
&
Johnson
,
E. A.
(
1969
).
Cardiac muscle
.
Z. Zellforsch. miskrosk. Anat
.
98
,
437
468
.
Sperelakis
,
N.
(
1969
).
Lack of electrical coupling between contiguous myocardial cells in vertebrate hearts
.
In Comparative Physiology of the Heart: Current Trends
(ed.
F. V.
McCann
), pp.
135
165
.
Basel
:
Birkhäuser Verlag
.
Spira
,
A. W.
(
1971
).
The nexus in the intercalated disc of the canine heart: quantitative data for an estimation of its resistance
.
J. Ultrastruct. Res
.
34
,
409
426
.
Stefanelli
,
A.
&
Zacchei
,
A. M.
(
1958
).
Sulle modalità di aggregazione di cellule embrionali di pollo disgregate con tripsina
.
Acta Embryol. Morph, exp
.
2
,
1
12
.
Stefanelli
,
A.
,
Zacchei
,
A. M.
&
Ceccherini
,
V.
(
1961
).
Ricostituzioni retiniche in vitro dopo disgregazione dell’abbozzo oculare di embrione di pollo
.
Acta Embryol. Morph, exp
.
4
,
47
55
.
Steinberg
,
M. S.
(
1970
).
Does differential adhesion govern self-assembly processes in histogenesis ? Equilibrium configurations and the emergence of a hierarchy among populations of embryonic cells
.
J. exp. Zool
.
173
,
395
434
.
Steinberg
,
M. S.
&
Roth
,
S. A.
(
1964
).
Phases in cell aggregation and tissue reconstruction. An approach to the kinetics of cell aggregation
.
J. exp. Zool
.,
157
,
327
338
.
Trelstad
,
R. L.
,
Revel
,
J. P.
&
Hay
,
E. D.
(
1966
).
Tight junctions between cells in the early chick embryo as visualized with the electron microscope
.
J. Cell Biol
.
31
, C6-C10.
Wollenberger
,
A.
(
1964
).
Rhythmic and arrhythmic contractile activity of single myocardial cells cultured in vitro
.
Circulation Res., Suppl
.
2
,
14
15
, 184-201.