ABSTRACT
In the post-gastrulation embryo of the Atlantic squid, Loligo pealei, the cells of the developing blastoderm are joined to each other by intercellular bridges which may provide a means of cytoplasmic communication between the cells. This paper describes an electron microscope survey of bridges in the developing blastoderm just prior to, and during, the onset of differentiation. The bridges are similar to those described in the gonadal tissue of many animal species and appear to result from incomplete cytokinesis followed by the disappearance of the spindle remnant. The bridges persist and chains of cells result which are generally branched and coiled. In the undifferentiated blastoderm the chains of cells show no apparent orientation to each other. However, in the apical blastoderm undergoing differentiation, chains of bridged cells appear to coincide closely with the developing mantle and shell gland primordia.
The configuration a chain of cells assumes depends upon the degree of branching (i.e. the number of cells having three bridges) and the degree of coiling of the chain. Whereas coiling is probably affected by the crowding of neighboring cells, both branching and coiling appear to be functions of spindle orientation relative to previous bridges. During mitosis the bridges appear to become occluded by systems of transverse membranous cisternae, and mitotic nuclei are thus isolated. However, the bridges apparently re-open during G1, and during periods of protein synthesis the cells within a group share a common cytoplasm. It is suggested that gene products are shared and protein synthesis of the entire bridged group may be synchronized. As the sharing of control molecules may also be facilitated, these essentially syncytial groups may respond uniformly to inducers from the yolk syncytium, or other tissues and differentiation may be synchronized within the group.
INTRODUCTION
In 1955 Burgos & Fawcett confirmed at the ultrastructural level that incomplete cytokinesis resulted in cytoplasmic bridges being formed between otherwise separate spermatids in the cat. Since then intercellular bridges have been described in developing ovaries in mammals (Franchi & Mandl, 1962), insects (Ramamurty, 1964), crustaceans (Anteunis, Fautrez-Firlefyn & Fantrez, 1966), annelids (Anderson & Heubner, 1968), amphibians (Ruby, Dyer, Skalko & Volpe, 1970b), rotifers (Bentfield, 1971), arachnids (Brinton, 1971), fish (Satoh, 1974), and cephalopods (Selman & Arnold, 1977). Similar bridges have also been described between developing sperm cells in birds (Nagano, 1961), insects (Hoage & Kessel, 1968), millipedes (Reger & Cooper, 1968), fish (Clerot, 1971), echinoderms (da Cruz-Landim & Beig, 1976), and gastropods (de Jong-Brink, Boer, Hommes & Kodde, 1977).
The primary function attributed to bridges in gonads is the synchronization of cellular events. In the higher animals, germ cells arise from a population of gonial cells that have proliferated by a series of highly synchronized mitotic divisions (Fawcett, Ito & Slautterback, 1959; Zamboni & Gondos, 1968; Gondos & Zamboni, 1969; Dym & Fawcett, 1971 ; Brinton, 1971 ; Kinderman & King, 1973; Woodruff & Telfer, 1973). Following proliferation, the germ cells undergo meiosis, also in a highly synchronized manner (Fawcett et al. 1959; Skalko, Kerrigan, Ruby & Dyer, 1972; Satoh, 1974; Filosa & Taddei, 1976; Brinton, 1971). In most male animals, meiosis is followed by synchronized maturation of the spermatozoa, during which time the cells remain bridged in clusters. This association between intercellular bridges and synchronization of maturation has been described in the testis of coelenterates, birds, insects (Fawcet et al. 1959; Dym & Fawcett, 1971), the opossum, rabbit and man (Fawcett et al. 1959), chinchilla, bull, ram (Dym & Fawcett, 1971), and the cat (Burgos & Fawcett, 1955; Dym & Fawcett, 1971). Similarly, bridges have been implicated in the synchronization of maturation in the ovaries of several animals including the tick (Brinton, 1971), the mouse (Ruby, Dyer & Skalko, 1969) and the rabbit (Gondos & Zamboni, 1969). During the development of the gonad in the above animals, the germ cells remain joined in clusters by intercellular bridges, and it is believed that the resulting cytoplasmic continuity allows for the distribution of control molecules and is responsible for the synchrony of both division and maturation. Reminiscent of the bridged cells in the testis are the developing nematocysts in Hydra (Fawcett et al. 1959). Synchronous division of somatic interstitial cells results in clusters of 14–16 cnidoblasts, linked by intercellular bridges, which then undergo synchronous maturation into nematocysts.
Regulation of the numbers of maturing germ cells by systematic cellular degeneration has been reported in female rabbits (Zamboni & Gondos, 1968; Gondos & Zamboni, 1969), mice (Ruby et al. 1969) and humans (Ruby, Dyer, Gasser & Skalko, 1970,a; Gondos, 1973a), and male monkeys (Gondos & Zemjanis, 1970; Dym & Fawcett, 1971), rats, guinea-pigs, Chinese hamsters, chinchillas, cats, cattle, and sheep (Dym & Fawcett, 1971). The mechanism for this degeneration is unknown, but an association with cell division has been noted and intercellular bridges have been implicated in synchronizing its onset (Zamboni & Gondos, 1968; Gondos & Zemjanis, 1970; Dym & Fawcett, 1971; Gondos, 1973a). Another function attributed to intercellular bridges is that of intercellular transport of nutrient materials. Evidence for this function is based primarily upon studies of the nurse cell-oocyte relationship in invertebrates (Koch & King, 1966 and 1969; Anderson & Heubner, 1968; Cassidy & King, 1969; King & Akai, 1971 ; Woodruff & Telfer, 1973) in which one cell in a cluster of bridged oogonia is destined to mature into an ovum while the others transfer nutrients and organelles to the oocyte through the bridges. Intercellular bridges are also thought to restrict mitosis in the ovary of insects. In most insects, gonial cells (cystocytes) undergo division in a precise and predictable pattern, yielding a cluster containing a species-specific number of cells attached to each other by bridges. After this pattern is achieved, mitosis stops and the cell with the requisite number of bridges begins differentiation to an ovum (King & Akai, 1971 ; Johnson & King, 1972; Kinderman & King, 1973). In his review of intercellular bridges in mammalian germ cell differentiation, Gondos (1973b) suggested that bridges serve to restrict the mobility of cells. Although groups of bridged cells migrate, bridging may retain certain essential spatial relationships.
Nagano (1961) showed that the intercellular bridges between spermatocytes in the rooster may become temporarily closed by a system of membranous cisternae forming across the channel. Dym & Fawcett (1971) described similar occluding membranes in bridges between spermatocytes in the ram, rat, guineapig, cat, chinchilla and Chinese hamster, and bridges between spermatogonia in the ram and the rat. These authors speculated that the purpose of the occluding membranes was to isolate certain cells from the germinal syncytium and the degeneration process. They also noted that when bridges became occluded synchrony of mitotic events among the cells was less apparent (Dym & Fawcett, 1971).
Recently Arnold (1974) reported that intercellular bridges occurred between the future somatic cells of the blastoderm in the embryo of the squid, Loligo pealei. The bridges appeared to be very similar in morphology to those described in the gonad of most animals, and were occasionally found to be occluded by transverse membranous cisternae in a manner similar to that described by Nagano (1961) and Dym & Fawcett (1971). The purpose of this paper is to describe the distribution of intercellular bridges in the Loligo pealei embryo and relate the resulting cytoplasmic continuity to certain aspects of the development of the organism.
MATERIALS AND METHODS
The Loligo pealei embryos used in this study were obtained by inducing adult animals to spawn (Arnold, 1962) in tanks of running sea water at the Marine Biological Laboratory at Woods Hole, Massachusetts. The strings of embryos were kept in their jelly in running sea water at 16–20 °C until they reached midblastoderm stages 16 and 17 (references to embryonic stages in Loligo pealei are from Arnold, 1965a). They were mechanically removed from the egg jelly and dechorionated with jeweler’s forceps and fine iridectomy scissors. The embryos were then fixed by a procedure modified from that of Palade (1952) in which they were immersed for 15 min in 1·0% osmium tetroxide in 0·25 molar veronal acetate buffer adjusted to pH 6·8–7·0 at 20–22 °C. This was followed by several changes of 50 % ethanol, dehydration through a graded series of ethanol into propylene oxide and embedding in Epon according to Luft (1961).
To determine the distribution of intercellular bridges, and the linkage patterns of bridged cells in the blastoderm, three-dimensional reconstructions were made of limited areas of the blastoderm from electron micrographs of serial thin sections. In the first survey, sections were cut from the undifferentiated blastoderm, just toward the animal pole from the advancing blastoderm margin, of a stage-16 embryo (Fig. 1 a). In the second survey, sections were cut from the animal pole of a stage-17 embryo (Fig. 16) where the mantle and shell gland primordia were in the early stages of differentiation (Arnold, 1971). Each embryo was trimmed and oriented in the ultramicrotome so that serial sections could be cut tangentially to the surface of the embryo. Sections were collected and mounted on formvar support films on slotted grids until both blastoderm layers and the yolk syncytial layer were sectioned completely through over a wide area. These were stained with uranyl acetate (Stempak & Ward, 1965) and lead citrate (Venable & Coggeshall, 1965) and examined with the electron microscope, and the entire area of each section was photographed. The plates were printed at a magnification of 2250 × and the prints were montaged to make single images of each section, which were then carefully examined for intercellular bridges. The location of each bridge relative to the cells was marked on each section even though the bridge did not appear in all sections. The cell outlines of every tenth section were then traced on paper (Figs. 2 through 7) and the location of each bridge was marked on each tracing. The groups of bridged cells were indicated on the tracings with transfer design patterns. Ninety-six serial sections were required for the laterally sectioned embryo, and the diagrams of the 42nd (Fig. 2), 52nd (Fig. 3) and 70th (Fig. 4) sections are presented here. One hundred and seventy-six sections were required from the thicker, differentiating blastoderm of the apically sectioned embryo. Diagrams of the 60th (Fig. 5), 100th (Fig. 6) and 176th (Fig. 7) sections are presented here. Figure 8 shows the 100th section.
Diagram showing the areas sectioned for the embryo surveys. In the laterally sectioned embryo (a) the area surveyed was just toward the animal pole from the advancing margin of the blastoderm. In the apically sectioned embryo (b) the surveyed area was at the animal pole. Scale bar : 0·5 mm.
Diagrams of sections of the laterally sectioned embryo showing the groups of bridged cells. The numbers refer to cell groups specifically referred to in the text. Groups 8 through 16 are in the middle cell layer. The others are in the outer layer. M = metaphase cell ; A = anaphase cell. Scale bars : 20 μm.
Fig. 2. Diagram of the 42nd section of the laterally sectioned embryo.
Diagrams of sections of the laterally sectioned embryo showing the groups of bridged cells. The numbers refer to cell groups specifically referred to in the text. Groups 8 through 16 are in the middle cell layer. The others are in the outer layer. M = metaphase cell ; A = anaphase cell. Scale bars : 20 μm.
Fig. 2. Diagram of the 42nd section of the laterally sectioned embryo.
Diagrams of the apically sectioned embryo showing the groups of bridged cells. The numbers refer to cell groups specifically referred to in the text. Groups 6 through 17 are in the middle cell layer. The others are in the outer layer. M = metaphase cell; A=anaphase cell; EA = early anaphase cell; LA = late anaphase cell: T = telophase cell. Scale bars: 20μm.
Fig. 5. Diagram of the 60th section of the apically sectioned embryo. Cell with four bridges is denoted (*).
Diagrams of the apically sectioned embryo showing the groups of bridged cells. The numbers refer to cell groups specifically referred to in the text. Groups 6 through 17 are in the middle cell layer. The others are in the outer layer. M = metaphase cell; A=anaphase cell; EA = early anaphase cell; LA = late anaphase cell: T = telophase cell. Scale bars: 20μm.
Fig. 5. Diagram of the 60th section of the apically sectioned embryo. Cell with four bridges is denoted (*).
Diagram of the 100th section of the apically sectioned embryo. See also Fig. 8. Cell with four intercellular bridges is denoted (*).
Diagram of the 100th section of the apically sectioned embryo. See also Fig. 8. Cell with four intercellular bridges is denoted (*).
RESULTS
The cells in both the undifferentiated blastoderm (laterally sectioned embryo) and the blastoderm undergoing differentiation (apically sectioned embryo) appeared to be joined to one or more neighbors by intercellular bridges very similar to those described in gonadal tissue (see Introduction). The cells were linked in coiled chains that in the undifferentiated blastoderm showed no obvious orientation to each other, but in the differentiating blastoderm appeared to coincide with developing organ primordia. Often the chains of cells were branched, i.e. one or more cells had three bridges. However, in neither embryo did there appear bridges joining cells of different cell layers.
There were mitotic cells in both embryos that were bridged to cells in differing stages of mitosis or to interphase cells. In each case the bridges associated with mitotic cells appeared typical except that, arranged across the channel of each one was a series of transverse membranous cisternae that appeared to interrupt the cytoplasmic continuity between the bridged cells. Therefore although mitotic cells were bridged to other mitotic cells, or to interphase cells, they appeared to be isolated cytoplasmically from those cells. Several other bridges, not associated with mitotic cells, were also of the occluded type.
Normal development of Loligo pealei has been described (Arnold, 1971; Arnold & Williams-Arnold, 1976). The surveys presented here are of embryonic stages just prior to the onset of tissue differentiation (the laterally sectioned embryo) and just after tissue differentiation has begun (the apically sectioned embryo). At these stages the embryo consists of three layers covering the apex of the large ovate mass of yolk like a cap. The innermost is a syncytial yolk epithelium that is continuous with a thin layer of cytoplasm entirely surrounding the yolk. Covering the yolk epithelium are the middle and outer cell layers, each one cell deep. As development progresses, the entire complex extends downward from the apex until the entire yolk mass is covered with cells. As the blastoderm completely covers the yolk, organ primordia begin to appear at the apex as thickenings, or invaginations, of the outer two layers.
The diagrams in Figs. 2 through 7 represent three-dimensional surveys of areas of the blastoderm of Loligo pealei embryos compiled from serial sections cut tangentially to the embryo surface. The groups of cells bridged to each other are shown as different design patterns. Both outer and middle cellular layers are included and are distinguished in the diagrams by the design pattern used. Because of the domed shape of the cell layers, the plane of section cut through the outer cell layer over an expanding area and then into the middle cell layer. As sectioning proceeded, the outer cell layer appeared as a ring of cells around the middle cell layer which appeared as a ring of cells around an area of yolk. In the case of the undifferentiated blastoderm, however, part of the tissue had pulled away from the area to be sectioned during specimen preparation, resulting in a crescent-shaped area of cells with the outer layer forming a semicircle around the cells of the middle layer (Figs 2, 3 and 4). In the diagrams in which cells appear separated by a space, or intervening cell, but lie against each other and are bridged in other sections, the linkage is indicated by a dotted line.
The area of the first survey (laterally sectioned embryo, Figs 2, 3 and 4) was in the undifferentiated region of a stage-16 embryo just behind the advancing margin of the blastoderm, about halfway down the embryo from the animal pole (Fig. 1 a). For the second survey (apically sectioned embryo, Figs 5,6 and 7), a stage-17 embryo was sectioned tangentially to the animal pole in the vicinity of the differentiating mantle such that the plane of section was perpendicular to the long axis of the embryo (Fig. 1b). Figure 8 is a micrograph showing the 100th section from the survey of the apically sectioned embryo. Figure 9 shows, at higher magnification, a chain of mitotic cells seen in Fig. 8, as well as the occluded bridges that join them. The first differentiation evident with the light microscope in the Loligo pealei embryo is the invagination of the shell gland and early differentiation of the mantle at the apical end in stages 16 and 17 (Arnold, 1971).
The 100th section in the survey of the apically sectioned embryo; higher magnification of the area outlined in Fig. 8. Six cells in different stages of mitosis are joined by occluded intercellular bridges. A pair of interphase cells is joined by an open bridge. Scale bar: 10·0μm.
Fig. 9B. Occluded bridge in the 114th section, located at (b), that joins a late anaphase cell to a telophase cell. 22900 ×.
Fig. 9C. Occluded bridge in the 93rd section, located at (c), that joins two late anaphase cells. 22900 ×.
Fig. 9D. Occluded bridge in the 100th section, located at (d), that joins an early anaphse cell to a late anaphase cell. 22 900 ×.
Fig. 9E. Occluded bridge in the 93rd section, located at (e), that joins an early anaphase cell to a metaphase cell. 22900 ×.
Fig. 9F. Open bridge in the 100th section, located at (f), joins two interphase cells. 22900 ×.
Fig. 9G. Occluded bridge in the 102nd section, located at (g), joins two metaphase cells. 22 900 ×.
The 100th section in the survey of the apically sectioned embryo; higher magnification of the area outlined in Fig. 8. Six cells in different stages of mitosis are joined by occluded intercellular bridges. A pair of interphase cells is joined by an open bridge. Scale bar: 10·0μm.
Fig. 9B. Occluded bridge in the 114th section, located at (b), that joins a late anaphase cell to a telophase cell. 22900 ×.
Fig. 9C. Occluded bridge in the 93rd section, located at (c), that joins two late anaphase cells. 22900 ×.
Fig. 9D. Occluded bridge in the 100th section, located at (d), that joins an early anaphse cell to a late anaphase cell. 22 900 ×.
Fig. 9E. Occluded bridge in the 93rd section, located at (e), that joins an early anaphase cell to a metaphase cell. 22900 ×.
Fig. 9F. Open bridge in the 100th section, located at (f), joins two interphase cells. 22900 ×.
Fig. 9G. Occluded bridge in the 102nd section, located at (g), joins two metaphase cells. 22 900 ×.
In the surveyed area of the laterally sectioned embryo, the outer cell layer consisted of cuboidal cells of uniform size and morphology averaging 15·7 μm diameter throughout the area. The cell profiles varied from round and polygonal to oblong, with few intercellular spaces. The cells of the middle cell layer appeared similar in morphology to those in the outer cell layer except that the former were larger, flatter, also undifferentiated and of relatively uniform diameter of about 21·0 μm. These cells were interspersed with intercellular spaces of varying size and shape. Cells in both layers had cytoplasmic processes that extended to other cells and to the yolk epithelium.
In the outer cell layer, there were 104 intercellular bridges linking 130 cells in 25 groups (Figs 2, 3 and 4). The largest of these groups (group 2) had 20 cells linked with 19 bridges, and the next largest (group 7) had 15 bridges linking 16 cells. The remaining groups had fewer bridges, and eight groups had only two cells joined by a single bridge. In the middle cell layer of the area surveyed, there were 42 bridges connecting 51 cells in nine groups (groups 8 through 16; Figs 2,3 and 4). One group (group 10, Fig. 4) had 15 bridges linking 16 cells. There was one group (number 8) of nine cells linked by eight bridges. The remaining groups had seven cells or less and three had only two cells. In the outer layer, groups 1 through 7 had one or more branches and in the middle cell layer, groups 9 and 10 were each branched. There were five mitotic cells in the area surveyed, three of which were bridged to other cells. In each case the bridges associated with mitotic cells appeared occluded by transverse cisternae.
In the case of the apically sectioned embryo, the area covered was larger than that of the laterally sectioned embryo and included 383 cells in both layers. The cells in this survey appeared to have a very similar morphology to those in the laterally sectioned embryo except that in the outer layer there seemed to be a gradual change in the cell diameter from one end of the surveyed area to the other (Fig. 5). The gradation in cell diameter appears to be unrelated to the bridging patterns and is probably a result of stretching of the blastoderm, or a slight misalignment of the knife during sectioning. The surveyed area was dominated in the outer layer by group number 1, consisting of 112 cells linked by 111 bridges (Figs 5, 6 and 7). The next largest group was number 3, which had 36 cells connected by 35 bridges. The remaining groups of the outer cell layer had fewer than 19 cells, and nine of these groups had only two cells and one bridge each. The middle layer in the surveyed area had 63 cells linked by 51 bridges in 12 groups (groups 6 through 17, Figs 5, 6 and 7), including five twocell groups. The largest group in the middle layer was group 6, which had 23 cells (Figs 6 and 7). The rest of the middle layer groups had fewer than ten cells. There was extensive branching in the groups of both cell layers, and one cell in group 1 had four intercellular bridges (Figs 5 and 6). As in the laterally sectioned embryo, there was no apparent bridging between the outer and middle cell layer.
In the surveyed area of the apically sectioned embryo, there appeared 21 mitotic cells, three of which were in the middle cell layer. There were 30 bridges associated with the mitotic cells and all were occluded by transverse membranes. Six of the mitotic cells appeared to be bridged only to interphase cells, while each of the others was bridged to at least one other mitotic cell. There was no apparent synchronization between mitotic cells bridged to each other. In group 3 there were six cells linked to each other (Figs 5, 6, 8 and 9) and all were in different stages of mitosis from metaphase to late telophase. Of a random sample of 71 bridges not associated with mitotic cells, 15 (21 %) were occluded, 17 (24 %) had mid-bodies and 39 (55 %) were open.
DISCUSSION
The results of these investigations indicate that intercellular bridges that result from incomplete cytokinesis, similar to those described in gonadal tissues (see Introduction) provide a possible method of communication between cells of the squid embryo. Arnold (1965b, 1968) and Arnold & Williams-Arnold (1976) have shown that in Loiigo pealei there is established a pattern of information in the egg cortex, and later in the yolk syncytium, that directs differentiation in the overlying blastoderm. They postulate that this information is possibly in the form of RNA messages which directly, or indirectly, control the selective activation of genetic material during cellular differentiation (Arnold & Williams- Arnold, 1976). Various types of cellular interactions, including gap junctions and intercellular bridges, provide possible routes for metabolic communication between the yolk syncytium and the cells of the blastoderm, and among the cells themselves. Potter, Furshpan & Lennox (1966) showed that in the embryo of Loiigo pealei, up until a few days before hatching, there is a low-resistance ionic coupling between cells of various organs and the yolk syncytium. Afterwards this coupling is lost. Although intercellular bridges could provide low-resistance channels for such coupling, it is likely that gap junctions are also involved, as there do not appear to be bridges between cell layers or between either cell layer and the yolk syncytium.
Intercellular bridges might, however, provide certain advantages in communication between cells of the same layer. The bridges would allow passage of all molecules that can pass through gap junctions, plus larger molecules, such as preformed RNA and polysomes. They could also allow passage of organelles such as elements of the cytoplasmic membrane system and mitochondria (Arnold, 1974). Although the bridges appear to be occluded during mitosis, they are apparently open in G1. and during protein synthesis and DNA replication. Gene products may then be shared between cells, and protein synthesis of the entire bridged group might therefore be synchronized. Passage of control molecules might also be facilitated and these essentially syncytial groups might respond uniformly to inducers from the yolk syncytium, or other tissues. This might allow differentiation within the group to be highly synchronized.
The cells in the differentiating blastoderm appear to be bridged in groups which closely coincide with developing organ primordia. The apically sectioned embryo was oriented so that the differentiating mantle and shell gland would occupy the center of the sectioned area (Arnold, 1965a, 1971). Cells of group 1 surround, in an interrupted ring, the cells of group 2 (Fig. 5). This configuration strongly suggests that group 2 (and only group 2), gives rise to the shell gland and that group 1 similarly gives rise to the mantle primordium (see also diagrams of Naef, 1928, Kollicker, 1844, etc.) because of the exact coincidence of the organ primordia and these groups of bridged cells. It seems likely that these groups will develop differently from each other because of their different positions relative to the yolk syncytium (Arnold, 1965b, 1968; Arnold & Williams-Arnold, 1976). However, because of the intercellular bridges, the cells within each group would differentiate synchronously into their respective organs. Because of the extremely laborious task of completely serially sectioning and analyzing the tissues, other organ primordia were not subjected to complete analysis, but enough bridged cells were found in the developing retina and the otocyst primordium to suggest a similar situation exists in these tissues also.
The configuration a group of cells assumes depends upon the degree of branching of the chain of cells (i.e. the number of cells having more than two bridges) and the degree of coiling of the chain. Whereas coiling is probably affected by the crowding of neighboring cells, both branching and coiling appear to be functions of spindle orientation relative to previous bridges. If the spindle of a dividing cell is oriented so that the plane of division lies between two previous bridges (Fig. 10-1), each daughter cell will get one of the previous bridges and the chain will remain unbranched (Fig. 10-III). Group 4 of the apically sectioned embryo is such an unbranched chain (Figs 5 through 7). If, however, the spindle is oriented so that both previous bridges lie on the same side of the division plane (Fig. 10-II), one of the daughter cells will receive both previous bridges. These two bridges, plus the one forming from the present division, will result in that particular cell having three bridges and the chain will be branched (Fig. 10-IV) such as groups 1 and 2 of the apically sectioned embryo (Figs 5 and 6). If any division within a chain occurs such that the plane of division is perpendicular to a line connecting two previous bridges, the division will result in the extension of a straight chain. Such appears to have been the case in group 4 of the apically sectioned embryo (Figs. 5 through 7) and most of the divisions in group 10 of the laterally sectioned embryo (Fig. 4). However, it appears that the end cell of a chain can divide with the spindle rotated so that the furrow occurs near the previous bridge. The close proximity of the two bridges would result in a bend or coiling of the chain. Such appears to have been the case in group 5 of the apically sectioned embryo (Figs 5 through 7), and group 10 of the laterally sectioned embryo (Fig. 4). There did appear one cell with four bridges in group 1 of the apically sectioned embryo (Figs 5 and 6). This probably resulted when a cell with three previous bridges divided in such a way that all three bridges were on the same side of the division plane.
This diagram shows how the orientation of the mitotic spindle relative to previous bridges may cause branching of a chain of cells. If a cell that has two previous bridges divides so that the division furrow lies between the previous bridges (I), each daughter cell gets one and a straight chain results (III). If the orientation of the spindle is such that the previous bridges lie on the same side of the furrow (II), one of the daughter cells gets both previous bridges. These, with the bridge from the present division, will give it a total of three (IV) and a branch in the chain results.
This diagram shows how the orientation of the mitotic spindle relative to previous bridges may cause branching of a chain of cells. If a cell that has two previous bridges divides so that the division furrow lies between the previous bridges (I), each daughter cell gets one and a straight chain results (III). If the orientation of the spindle is such that the previous bridges lie on the same side of the furrow (II), one of the daughter cells gets both previous bridges. These, with the bridge from the present division, will give it a total of three (IV) and a branch in the chain results.
The relationship between the mitotic apparatus and the plane of cleavage has been formalized as ‘Balfour’s Rules’ (cited in Wilson, 1925; Arnold, 1976) which state: (1) the plane of cleavage is perpendicular to that of the previous division, and (2) the plane of cleavage is perpendicular to the long axis of the spindle. The many cases in which bridges appear on roughly opposite sides of the cell (group 4, Fig. 6 and group 10, Fig. 4) suggest that the first rule is violated in the squid blastoderm. The second rule, however, provides the basis for the control of branching and coiling, and therefore the shape of the cell group. Despite the extensive branching and coiling that characterize the groups of cells, in neither of the surveyed embryos did there appear a cell in the outer cell layer bridged to one in the middle cell layer. That the intercellular bridge arises as a result of mitosis is supported by the observation that in neither embryo did there appear a chain of cells bridged back upon itself in a circle, and that in every cell group the number of cells exceeded the number of bridges by one.
Even though cells joined by intercellular bridges share a common cytoplasm, regional differences in morphogenic, or control substances, may occur and responses may vary within the bridged group. Moens & Hugenholtz (1975) presented evidence that in the rat, individual spermatogonia in a syncytial group may initiate differentiation at different times. In the insect ovary, one cell of a bridged group matures into an ovum while the others become nurse cells (Woodruff & Telfer, 1973). In the squid blastoderm mitosis is not synchronized. It appears that when an embryonic somatic squid cell enters mitosis, the bridges that link it to neighboring cells become occluded, thereby interrupting cytoplasmic continuity and allowing mitotic asynchrony. Throughout both embryos surveyed, cells in different phases of mitosis are bridged to each other as well as to interphase cells. An example of this can be seen in group 3 of the apically sectioned embryo (Figs 5, 6, 8 and 9) in which six cells, all in different phases of division, are evident. However, at high magnification, all of the bridges associated with the mitotic cells are seen to be of the occluded type, and there was no evidence of cytoplasmic continuity between a mitotic cell and any other cell. Dym & Fawcett (1971) made similar observations in the testis of the ram. It seems reasonable to assume that there is some advantage in asynchrony of cell division in the blastoderm and organ primordia. If divisions were synchronous, the number of cells in a group would be restricted to twice the number of cells in the previous cycle. Asynchrony would allow the cell number to be regulated to satisfy the requirements of group size and shape more easily.
The results of the embryo surveys support the hypothesis that the transverse membranes of the occluded bridge are temporary structures and that bridges can open and close repeatedly (Nagano, 1961 ; Dym & Fawcett, 1971 ; in contrast to the view of Arnold, 1974). If bridges are formed by incomplete cytokinesis during division, and every bridge associated with a cell going into mitosis becomes occluded, then three or more cells linked in a chain by open bridges indicates that the occluded bridges re-open after division. There are many such groups in both embryos surveyed (Figs 2 through 7). In group 1 of the apically sectioned embryo (Fig. 6), there are mitotic cells separated by 15 interphase cells. This means at least four cell cycles have occurred since their common parent cell first divided. In group 3 of the same embryo, there are mitotic cells separated by six interphase cells, which means they are separated by three cell cycles (Fig. 6). Since about 25 % of the total number of intercellular bridges are closed at one time, the bridges remain closed for about one-quarter of the cell cycle. Therefore, the bridges associated with these mitotic cells must have closed and opened previously during the growth of the chains.
Acknowledgement
The authors wish to thank sincerely Dr Ian Gibbons for the use of his electron microscope and Ms Frances Okimoto, Mrs Sandra Haley and Ms Concepcion Mata for their expert assistance in the preparation of the manuscript. We also wish to thank Dr Margaret Ann Goldstein and Mr David Murphy for their helpful discussions and critical review of the manuscript.