DEVELOPMENTALBIOLOGY137,207-216 (1990)

Changes in Junctional Communication Associated with Cell Cycle Arrest and Differentiation of Trochoblasts in Embryos of Pate/la vu/g&a F. SERRAS,

W. J. A. G. DICTUS, AND J. A. M. VAN DEN BIGGELAAR

Department of Experimental Zoology, University of Utrecht, Padualam 8, 358.4CH Utrecht, The Netherlands Accepted September 22, 1989 In early embryos of molluscs, different clones of successively determined trochoblasts differentiate into prototroch cells and together contribute to the formation of a ciliated ring of cells known as the prototroch. Trochoblasts differentiate after cell cycle arrest, which occurs two cell cycles after the commitment of their stem cell. To study the changes of junctional communication in embryos of Patella vulgata in relation to commitment, cell cycle arrest, and differentiation of the trochoblasts, we have monitored electrical coupling as well as transfer of fluorescent dyes. The appearance of dye coupling in embryos of Pate& occurs after the fifth cleavage (at the 32-cell stage), when the cell cycles of all embryonic cells become asynchronous and longer. At the 32- and 64-cell stages all cells are well coupled. However, after the 72-cell stage dye transfer to or from any cell of the four interradial clones of four primary trochoblasts becomes abruptly reduced, whereas electrical coupling between these cells and the rest of the embryo can still be detected. From scanning electron microscopical analysis of the cell pattern we conclude that this change in gap junctional communication coincides with cell cycle arrest and with the development of cilia in all four clones of primary trochoblasts. Similarly, after the %-cell stage the four radial clones of accessory trochoblasts stop dividing, reduce cell coupling, and become ciliated. By the formation of the prototroch, the embryo becomes subdivided into an anterior (pretrochal) and a posterior (posttrochal) domain which will develop different structures of the adult. At the 88-cell stage, the cells within each of these two domains remain well coupled and form two different communication compartments that are separated from each other by the interposed ring of uncoupled trochoblasts. The relations among control of cell cycle, changes in junctional communication, and differentiation are discussed. o ISSO Academic press, IX

INTRODUCTION

trochoblasts. One group of stem cells of these trochoblasts (the primary trochoblasts) is committed at the There is an increasing amount of evidence that inter16-cell stage. After two more cell cycles, these trochocellular channels, known as gap junctions, can facilitate blasts stop dividing and enter a phase of cell cycle ardiffusion of small molecules that carry information rest. Most of the trochoblasts follow a determinative which may lead to control of cell proliferation and cell development and they are the first cells that differendiversification (Loewenstein, 1981; Fraser et aZ., 1987, tiate, after cell cycle arrest, by developing cilia (Wilson, 1988). Junctional communication can be studied by in- 1904; Van den Biggelaar, 197’7;Janssen-Dommerholt et tracellular microinjection of fluorescent dyes or elec- aZ., 1983). Because of their changes in cell cycle period trical currents and analysis of the subsequent spread to and their early differentiation, the trochoblasts are a adjacent cells. In embryos of several species restrictions good candidate to study the significance of junctional in junctional communication between cells or groups of communication in cell cycle control and differentiation. cells with different developmental capacities have been In embryos of the mollusk Patella vulgata gap juncobserved. Such restrictions give rise to communication tions have been found in freeze-fracture replicas from compartments which may have important developmen- the 2-cell stage onward (Dorresteijn et cd, 1982). Pretal consequences (Blennerhasset and Caveney, 1984; Lo, vious works have shown that all blastomeres of the 321988; Van den Biggelaar and Serras, 1988; Warner and cell stage embryos of the P. vulgata allow gap junctional Lawrence, 1982). transfer of dyes at the 32-cell stage (De Laat et al, 1980; A suitable model system to study the relation be- Dorresteijn et aZ., 1983; Serras et al., 1988). Recently, tween changes in gap junctional communication and electrical coupling in the stages preceding the 32-cell stage has been demonstrated (Serras and Van den Bigregulation of cell proliferation and cell differentiation is the development of the prototroch of molluscs. The gelaar, 1989). In this paper we describe the pattern of prototroch is the larval locomotory organ of the intercellular communication before and after cleavage trochophora larva. It develops from different clones of arrest of developing trochoblasts of P. vulgata embryos. 207

0012-1606/90 $3.00 Copyright All rights

0 1990 by Academic Press, Inc. of reproduction in any form reserved.

208

DEVELOPMENTAL BIOLOGY

For this purpose we have intracellularly microinjected

VOLUME 13’7,199O

microelectrode was removed from the cell, and the em-

to anepifluorescence microscope, fluorescent dyesandelectrical currentsfromthe32-cell bryowastransferred stageup to the earlytrochophorestage. MATERIALS

AND

METHODS

wherethe pattern of the dye transfer (dyecoupling) wasanalyzed.Successfully injectedembryoswerephotographed in vivo on Kodak Ektachrome 400 ASA and Kodak Tri-X films.

Embryos

Adult specimens of the marine gastropod P. vulgata were collected in Roscoff (Bretagne, France). They were kept at 15°C in tanks with recirculating filtered seawater (FSW). Artificial fertilization was performed according to Van den Biggelaar (1977). Synchronously developing embryos were selected and kept in separate Boveri dishes. All experiments were carried out at 20°C in FSW. Dye Injection

Dyes were delivered into the cell by iontophoresis. Microelectrodes made from Kwik-fil capillaries with an inner filament (outer diameter, 1.5 mm; GC150-TF, Clark Electromedical Instruments, Pangbourne, UK) were pulled in a Mecanex BB-CH microelectrode puller (Geneve, Switzerland). The tip resistance was lo-20 MQ when filled with 3 M KCl. The tip was backfilled with a 3% aqueous solution of the dilithium salt of Lucifer yellow CH (MW 457 Da; Sigma, St. Louis, MO) and the remainder of the microelectrode was filled with 3 M LiCl. In some experiments a 3% solution of Fluorescein Complexon (MW 710 Da; Eastman Kodak Co., Rochester, NY) was used to fill the tip of the microelectrode. Fluoresceinated-dextrans (FITC-dextran, 41 kDa and 9 kDa, Sigma; Lucifer yellow-dextran, 10 kDa, Molecular Probes) were used to assess the existence of cytoplasmic bridges that might persist between the injected cell and its sister cell. Embryos with well-developed cilia were kept in a fixed position by gentle suction with a fire-polished pipet. The embryos were properly orientated under a Zeiss stereomicroscope. The dye-filled microelectrode was brought to the surface of the embryo with a micromanipulator (Microcontrole, Saint Guenault, Evry, France), and the desired cell was impaled. The voltage drop resulting from impalement was monitored by an oscilloscope and followed during the injection to check the stability of penetration and the condition of the injected cell. Dye was injected with rectangular 5-nA current pulses of 0.2-0.3 set in duration, applied at intervals of 2-4 set during 3-5 min. The spread of fluorescence was followed during the injection using a laser beam generated by an air-cooled argon-ion laser (162 A, Spectra-Physics, Mountain View, CA; 488 nm). The beam was reflected into the stereomicroscope, which was set up for epifluorescence. After iontophoresis, the

Electrical

Coupling

Embryos at the appropriate stage of development were transferred to a recording chamber. Microelectrodes (lo-20 MQ) were pulled as described for dye injection and mounted on micromanipulators containing a half-cell Ag/AgCl. The recording was completed with an agar 3 M KC1 bridge. Two microelectrodes were used to impale separate cells of an embryo. Entry of the microelectrode was followed by monitoring the membrane potential. Two different amplifiers were connected to each microelectrode. Both amplifiers can measure the potential and inject current through a microelectrode. Hyperpolarizing current pulses (0.5-3.0 nA in amplitude, 0.3-0.5 set in duration) were applied for a few minutes. Signals were displayed on an oscilloscope (Hameg MH2052) and a pen recorder (Gould 2200s). Only records made within the first minutes after stabilization of the membrane potential were used. By using a single microelectrode for passing current and recording potential, we avoided the need of a separate microelectrode for current injection (Purves, 1981). Using this method, the potential decline measured via the current injecting electrode has two components. One is due to the microelectrode resistance and is characterized by an abrupt decline at the onset of the current application. The second one corresponds to the cell membrane resistance and is slower due to the membrane capacitance. Compensation of the microelectrode resistance was performed with a bridge circuit. Coupling ratios were calculated as C = V/V,, where V, is the voltage transient of a cell in which current has been injected, and V, is the voltage deflection recorded in the second impaled cell. Microsurgery

Embryos at 6 and 7 hr after first cleavage were transferred to a petri dish with a wax-bottom filled with FSW. Carbon particles were mixed into the wax to provide a dark background. Embryos were orientated in small holes made in the surface of the wax-bottom. Fine glass needles pulled with the microelectrode puller were mounted on the micromanipulator and used to dissect the embryos along their anterior-posterior axis. As the trochoblasts lie equatorially, the isolated pieces of embryonic tissue always contained pretrochal ectoderm cells (anterior), trochoblasts, and post-trochal ectoderm

SERRAS, DICTUS, AND VAN DEN BIGGELAAR

cells (posterior). Each isolated piece of tissue was gently transferred to a recording chamber containing a poly-L-lysine (MW 70-150 kDa, Sigma)-coated glassbottom. Electrical coupling and dye injection were performed as described above. Scanning Electronmicroscopy Embryos at the appropriate stage were collected, fixed (2 hr, 4°C) in 2.5% glutaraldehyde in 0.1 M Nacacodylate buffer, pH 7.4, washed in buffer, and postfixed for 1 hr in 1% 090, in 0.1 M Na-cacodylate buffer (4°C). After rinsing twice in bidistilled water, the embryos were dehydrated in a graded series of ethanol. After critical-point drying with COz, the embryos were mounted on specimen holders with double-sided adhesive tape and coated with gold. Preparations were observed in a Cambridge Camscan scanning electron microscope. RESULTS

Normal Development of the Prototroch In Patella embryos, all cells cleave synchronously during the five first cell cycles; the period between the beginning of one cleavage to the beginning of the following cleavage lasts 30-35 min. However, after the fifth cleavage, i.e., at the 32-cell stage, the cell cycle of all cells becomes longer and the following cleavages become asynchronous. The 32-cell stage lasts about 90 min from the onset of the fifth cleavage to the onset of the sixth cleavage. The sixth cleavage is asynchronous and the first cells that divide are the primary trochoblasts. Between the 32-cell and 64-cell stage intermediate 40-, 56-, and 60-cell stages can be observed. Most of the prototroch cells are derivatives of the four primary trochoblast of the 16-cell stage and of the accessory trochoblasts of the 32-cell stage. After two divisions, each primary and accessory trochoblast gives rise to a clone of presumptive prototroch cells, the primary and accessory prototroch cells, respectively. In the early embryo the primary trochoblasts have an interradial position, whereas the accessory trochoblasts have a radial position. After the 64-cell stage the four clones of four primary trochoblasts do not divide further as was demonstrated in the experiments in which a primary trochoblast was injected with Lucifer yellowdextran at the 16-cell stage (n = 6; results not shown): the injected trochoblast divided only twice and gave rise to four primary trochoblasts at the 64-cell stage and the four cells contributed to the ciliated prototroch of the trochophore larva. These results are in accordance with previous observations (Wilson, 1904) and indicate that trochoblasts arrest their cell cycle after their last divi-

Intercelhdar Communication in Patella

209

sion at the 64-cell stage. According to Van den Biggelaar (1971a) the cell cycle arrest of the trochoblasts occurs at the G2 phase. The cell cycle-arrested primary trochoblasts are the first cells that will morphologically differentiate by developing cilia. Figure 1A shows scanning electron micrographs of an embryo at the end of the 72-cell stage. Small cilia in the primary trochoblasts are visible at this stage (Figs. 1A and 1B). A systematic analysis of various stages revealed that ciliation of the four clones of primary prototroch cells occurs from 60 to 80 min after their last division when the embryo has reached the 72-cell stage. The accessory trochoblasts formed at the 88-cell stage also become ciliated shortly after their last division. During the following stages of development (from 6 hr after first cleavage onward) all clones of trochoblasts shift their position so as to form a ring of ciliated prototroch cells in the 12-hr-old trochophores (Fig. 1C). This ring separates the presumptive ectodermal cells of the anterior (pretrochal) hemisphere from the presumptive domain of cells of the posterior (post-trochal) hemisphere (Fig. 1D). Transfer of Low Molecular Weight Dyes A total of 147 embryos was injected with either Lucifer yellow (111 embryos) or Fluorescein Complexon (36 embryos). As we could not find any difference in dye transfer patterns between the two dyes, we have combined the results without specifying which dye was used. From the late 32- to the &$-cell stage. All embryos that were injected between the end of the 32-cell stage and the early 64-cell stage showed rapid spread of dye from the impaled cell to the surrounding cells, independent of which cell was injected (n = 37). No indications of any restriction in dye transfer were found (Figs. 2A and 2B). Dye coupling during the ciliation of the primary trochoblasts. Dye transfer from any cell of the four clones of primary trochoblasts becomes reduced approximately 1 hr after their last cleavage cycle, i.e:, at the 72-cell stage. In five embryos injected into a primary trochoblast after the 72-cell stage no dye transfer from the impaled cell to other cells was found (Figs. 2D and 2E). In contrast, injection of dye into one of the accessory trochoblast (which have not yet gone through their last division) resulted in an extensive spread toward other adjacent cells of the embryo except to the primary trochoblasts (n = 3). Accordingly, injections into nontrochoblast cells of either the animal or the vegetal pole resulted in spread of dye to all cells of the embryo (including the accessory trochoblasts), except to the primary trochoblasts (n = 17 of 22); even 30 min

210

DEVELOPMENTALBIOLOGYV0~~~~137,1990

FIG. 1. Scanning electron micrographs of P. vulgata embryos. (A) End of the 72-cell stage viewed from the animal pole. All cells have been delineated in white to recognize the cell pattern. Notice that the four interradial clones of primary trochoblasts are ciliated. After an additional cleavage the radial cells marked “a” will develop the ciliated accessory trochoblasts. The radial accessory trochoblasts (a) have not yet gone through their last division cycle. (B) Higher magnification of a clone of primary trochoblasts of another embryo at the same stage. (C) Ciliated trochoblasts in an embryo 9 hr after first cleavage. (D) Nineteen-hour-old trochophore with a fully developed prototroch viewed from the dorsal side. pre, pretrochal domain; pro, prototroch ciliated cells; post, post-trochal domain. Scale bars in A and D, 30 pm; in B and C, 10 pm.

after injection, no spread to the primary trochoblasts could be detected (Figs. 2G and 2H). However, in three other cases dye was also restricted to the accessory trochoblasts. In these three cases the injection was performed when the accessory trochoblasts had already gone through their second division. Dye coupling during ciliation of the accessory trochoblasts. Embryos dye injected into any of the trochoblasts at this stage did not result in any detectable

spreading of dye to nonsister cells (n = 18; Figs. 3A-3D). In six of these embryos, spread to the sister cell was detected. The loss of dye coupling of the accessory trochoblasts occurs around 40 min after their last division cycle (5 hr 30 min after first cleavage). After the 88-cell stage the cells anterior from the trochoblasts consist of 12 cells from which the pretrochal region will develop. In 10 embryos injection of dye into one of these cells resulted in dye transfer to all 12 pretrochal cells.

SERRAS, DICTUS, AND VAN DEN BIGGELAAR

FIG. 2. (A) Fluorescence image of a 40-cell embryo injected with Lucifer yellow into a most animal cell. (B) Sixty-cell embryo in which Lucifer yellow was injected into a primary trochoblast. (C) Electrical coupling in an embryo at the 64-cell stage; one microelectrode was inserted into a primary trochoblast, and the other microelectrode into another cell, leaving a row of 3 cells in between. Upper trace, voltage transient of the cell where a current pulse (amplitude, 3 nA; duration, 300 msec) was applied; lower trace, voltage deflection in the second cell. (D) Fluorescence and bright-field image of an embryo injected with Lucifer yellow CH into a primary trochoblast. (E) The same embryo with only fluorescence. (F) Electrical coupling at the 72-cell stage. Upper trace, voltage transient of a trochoblast when injected with a 1-nA 300-msec current pulse; lower trace, voltage transient of the second cell. (G) Fluorescence micrograph of an embryo injected with Lucifer yellow into a most animal cell. Dye spreads to many cells, but not to the interradial primary trochoblasts. (H) Embryo injected with Lucifer yellow into a nontrochoblast of the animal hemisphere at the 72-cell stage and photographed at the 8%cell stage. (D, E) Lateral views. (G, H) Animal views. Scale bars, 50 pm; in C and F, 1 division = 10 mV. Asterisks, injected cells.

Intercellular

Communication in Patek

211

After injecting one cell of the post-trochal domain, dye spread extensively to adjacent post-trochal cells as shown in Figs. 3G and 3H. In none of the nine embryos studied was any dye spread found across the border delineated by the uncoupled trochoblasts. Analysis of these dye injection experiments demonstrates that after the 8%cell stage the embryo is subdivided into two communication compartments, the pre- and post-trochal compartments, which are separated by a ring of dye-uncoupled prototroch cells. Dye coupling in the differentiated prototroch. In contrast to the loss of dye coupling at the stages of ciliation of the primary and secondary trochoblasts (see above), trochoblasts at the end of their rearrangement into a ring-like prototroch (from 8 to 10 hr after the first cleavage) showed resumed coupling in 9 of 24 embryos. The onset of this recoupling is characterized by poor dye transfer to only a few adjacent prototroch cells. This coupling was found to occur only toward prototroch cells of different clones derived either from primary or accessory trochoblast stem cells. However, there were never more than eight labeling prototroch cells, and the spread was only detected about 5 to 10 min after injection (Fig. 4A). At later stages (from 12 hr after first cleavage onward) more extended spreading of dye to all prototroch cells was observed (n = 18; Fig. 4B), whereas transfer of dyes from prototroch cells to the adjoining pretrochal and post-trochal domains of ectodermal cells was never detected. These results are in agreement with a previous report in which it has been shown that in a fully developed prototroch of a trochophore larva, the prototroch cells mutually allow the transfer of Lucifer yellow but not of FITC-dextrans (Serras et ak, 1985). Electrical

Coupling

Measurements of electrical coupling in embryos at the 32- and 64-cell stages were performed between pairs of either nonsister cells or cells separated by interposed cells. At these stages, whatever pair of cells was chosen, we found coupling ratios close to 1.0 (Fig. 2C). After the 72-cell stage (from 4 hr 30 min to 5 hr 30 min after first cleavage) coupling ratios between two nontrochoblasts (i.e., cells which are dye-coupled) were observed to reIn four of these embryos we found that the dye was main close or above 0.9. Despite the absence of dye couimmediately transferred to all 12 cells. About 15 min pling (see above), electrical coupling between a primary after injection, labeling was also found in some of the trochoblast and another cell of the same embryo, either accessory trochoblasts. However, in none of these four another primary trochoblast of another clone or a nontrochoblast was detected (Fig. 2F) and coupling ratios embryos was any trace of dye detected in the primary trochoblasts. The other six embryos, injected about 30 ranged between 0.5 and 0.9 (Fig. 5). After the 88-cell stage (from 5 hr 30 min to ‘7hr after min later in development, showed no detectable dye coupling to any of the surrounding primary or accessory first cleavage), reduction of electrical coupling can be trochoblasts, even in later observations (Figs. 3E detected when at least one of the impaled cells is a trochoblast (coupling ratios lower than 0.5). But no and 3F).

FIG. 3. (A) Fluorescence and (B) with bright field images of an embryo injected with Lucifer yellow 6 hr after first cleavage into a single trochoblast. Notice that this embryo has been laterally photographed to show the characteristic ciliation of the injected trochoblast (arrow). (C, D) Eight-hour-old embryo. (E, F) Animal view of an 8%cell embryo injected with Lucifer yellow into a cell of the pretrochal domain. Dye spread fills this domain but is not transferred to the surrounding prototroch cells. (G, H) Lateral view of an embryo injected into a cell of the post-trochal domain. Dye is spread to other cells of the post-trochal domain but not to the prototroch. In both embryos the injected nontrochoblast cells (asterisk) are placed at the border with the trochoblasts. (E, G) Fluorescence photographs; (F, H) their respective bright field and fluorescence photographs. pre, pretrochal; post, post-trochal; dashed lines in B and D delineate the trochoblasts; dashed lines in F and H represent the border between the trochoblasts and the injected domain. Scale bars, 50 pm.

decay of the ratios is found when both impaled cells were nontrochoblasts, belonging to either the pretrochal- or the post-trochal domain (Fig. 5). In embryos of molluscs, extracellular spaces, which are electrically isolated from the bath may exist as remainders of the blastocoelic cavity present at the early stages (Ktihtreiber et al., 1986) and may generate a source of current (Dohmen et aZ., 1986). To analyze whether the gap junctional-mediated electrical coupling of the trochoblasts is overestimated due to such extracellular spaces, we used isolated fragments (monolayered) to ensure the disruption of such possible extracellular spaces and to force a direct contact between the inner side of the embryo and the incubation medium. In these fragments electrical coupling between a ciliated trochoblast and one of the surrounding cells was studied. In all cases electrical coupling was found. In 4 of 10 cases we measured coupling ratios ranging from 0.05 to 0.2 (Fig. 5). In the other six embryos it was evident that the coupling ratios were lower than 0.5. However, in these six embryos we did not consider the exact ratio because either the electrode resistance changed during the experiment or a decay of the resting

potential was observed a few seconds after impalement. This was probably due the tension exerted by movement of the cilia of the impaled cells on the intracellular microelectrode, resulting in disruption of the membrane. To study whether junctional contacts remain intact after manipulation and isolation, dye coupling was studied in nontrochoblast cells of isolated fragments. Injection of Lucifer yellow into nontrochoblasts resulted in spread to other nontrochoblasts (n = 3; Fig. 6). Input resistances were 8.5 + 0.4 MR (mean f SEM, n = 10) for prototroch cells or trochoblasts and 9 +- 0.7 MQ (n = 14) for nontrochoblasts, if measured before the 88-cell stage (before 5 hr 30 min after first cleavage), After the 88-cell stage (between 5 hr 30 min and ‘7 hr after first cleavage) the registered input resistance of the trochoblasts was 2’7 f 4.1 MQ (n = 16) and of the nontrochoblasts 11 + 1.4 MQ (n = 12). Transfer of High Molecular Weight @yes To exclude dye coupling by means of cytoplasmic bridges, experiments were performed in which FITCdextran was injected into cells dye-coupled with low

SERRAS, DICTUS, AND VAN DEN BIGGELAAR

FIG. 4. Fluorescence photographs of (A) an embryo injected with Lucifer yellow into a prototroch cell, 10 hr after the first cleavage. Dye spread to only a few other prototroch cells (arrows). (B) Embryo injected with Fluorescein Complexon from the animal pole view. Dye spread to all cells of the prototroch, but not to other regions. The dark area surrounded by the labeled prototroch is the pretrochal ectoderm (pre). The extended cilia can be seen in the injected prototroch cell. Asterisks, injected cell. Scale bars, 50 Frn.

molecular weight dyes. In embryos injected shortly after the end of the fifth and sixth cleavages (begin of 32- and 64-cell stages, respectively), dye transfer was found from the impaled cell to its sister cell only (n = 5). However, no dye transfer to the sister cells could be detected when embryos were injected late after the cleavage (n = 7). Injections of FITC-dextrans into an accessory trochoblast at the 72-cell stage (n = 6) or into a nontrochoblast at the %-cell stage (n = 4) did not result in transfer of the dye (Figs. 7A and 7B). After the injection all embryos cleaved normally and the dye was M

@-A AA AA

A

A tAA

AA

FIG. 5. Electrical coupling ratios plotted against time in hours of development after first cleavage. Black triangles represent coupling between a nontrochoblast and a trochoblast or between two trochoblasts of different clones. White triangles represent coupling ratios between two nontrochoblasts of the either pretrochal or post-trochal domain. Black boxes represent measurements in isolated pieces of ectoderm in which at least one of the impaled cells is a trochoblast.

Intercellular

Communication

in Pate&

213

FIG. 6. (A, B) Microdissected piece of ectoderm injected with Lucifer yellow. (A) Bright light and (B) fluorescence micrographs. Scale-bar, 25 pm. (C, D, E) Voltage traces of electrical coupling of three different isolated ectoderm pieces containing ciliated trochoblasts. In all three at least one microelectrode was impaled into a trochoblast. (C) Traces from an embryo 5 hr 30 min after first cleavage; (D) at 6 hr 20 min; (E) at 6 hr 40 min. Upper traces, voltage transients of the trochoblast after injection of 1 nA X 300 ms. Lower trace, voltage transient of the second cell. Scale in C, 1 division = 20 mV, scale in D and E, 1 division = 10 mV.

restricted to cells derived from the injected cell (Figs. 7C and 7D). DISCUSSION

From the results presented here we conclude that the trochoblasts follow a characteristic sequence of events after their second division, namely, (1) cleavage arrest (2) reduction of gap junctional communication, and (3) differentiation into ciliated cells. The nontrochoblast cells and also the trochoblasts before their last division, however, are characterized by (1) additional cell cycles, which after the fifth cleavage become longer and asynchronous between the different cell lines, and (2) extensive dye coupling (which appears after the fifth cleavage). Two questions may be of interest for understanding the role of gap junctional communication in Patella development: (1) Are gap junctions related with the appearance of asynchrony after the fifth cleavage? (2) Is the reduction of junctional communication of the trochoblasts important for maintaining their cell cycle arrest and the following differentiation? The results presented in this paper together with earlier results of our group (Van den Biggelaar 1971a; 1971b) suggest that the regulation of the cell cycle, including cell cycle arrest, may be related to cell coupling, as will be discussed below.

214

DEVELOPMENTALBIOLOGY

VOLUME137,199O

pend upon differences in the size of the remaining pool of maternal factors which is reflected by different lengths of the cell cycle. The degree to which the different phases contribute to the lengthening of the cell cycles has been studied in the mollusk Lymnaea. The division asynchrony between cells appears to be caused by unequal lengthening of the S and G2 phases (Van den Biggelaar, 1971a). Besides an endogenous control of the cell cycle discussed above, cell-to-cell interactions are also involved. For example, in the molluscs Patella (Van den Biggelaar, 1977) and Lymnaea (Van den Biggelaar, 1971a) after the fifth cleavage, one of the vegetal macromeres (the 3D cell) contacts the animal cells and becomes the stem cell of the mesoderm. A specific consequence of this interaction is that the duration of the cell cycles of these cells is altered. In embryos in which this contact has been prevented, this alteration fails to occur (Van den Biggelaar and Guerrier, 1979; Ktihtreiber et aZ., 1988; Ktihtreiber and Van Dongen, 1989). Van den Biggelaar and Guerrier (1979) have also studied the timing FIG. 7. (A, B) Fluorescence and dark-field images of an embryo of cell cycles of dissociated cells after the period of the injected with FITC-dextran (9 kDa) into an accessory trochoblast synchronous divisions. After the 32-cell stage the cell (asterisk) of the 72-cell stage. (C, D) The same embryo after the cycles of dissociated cells differ from the corresponding following cleavage of the accessory trochoblasts. Scale bars, 50 Frn. cells in the intact embryo. The animal cells cleave synchronously and ahead of the also synchronously diIt is known that in early embryos the cell cycle de- viding vegetal cells. The difference between the first and last dividing cell in dissociated embryos is 20 min, pends upon ooplasmic factors that are differentially distributed between the blastomeres (Clement, 1952; whereas in the normal embryo it is about 60 min. Thus, the cells of the dissociated embryo divide more Guerrier et al., 1978; Schierenberg, 1984; Schierenberg and Wood, 1985; Van Dongen and Geilenkirchen, 1974, synchronously than in the normal embryo. These ob1975). Therefore, it is not surprising that up to the 32- servations suggest that in developing Patella embryos cell stage the blastomeres of a dissociated Patella em- apart from nondiffusible cytoplasmic mitogenic factors, cell-to-cell contacts may alter the cell cycle rhythm of bryo cleave synchronously with the cells in the intact embryo (Van den Biggelaar and Guerrier, 1979). An- the different cell lines, for example, by a diffusible regother point of interest is that primary trochoblasts iso- ulatory molecule. It might be possible that such putalated at the 16-cell stage divide twice, enter a cell cycle tive molecules are transferred via gap junctions. In this respect it is worthwhile to reconsider the relation bearrest, and develop into ciliated cells as in the intact embryo (Wilson, 1904; Janssen-Dommerholt et al., tween the changes in gap junctional coupling and dif1983). This indicates that both cell cycle arrest and ci- ferentiation of the trochoblasts. liation of the trochoblasts are autonomously regulated In connection with the observation that cell contacts by endogenous factors. In analogy with the model of cell influence the duration of the cell cycle, it is important cycle control in early amphibian development (Newport to stress that once the cell cycle of the primary trochoand Kirschner, 1982a,b, 1984; Kimelman et ah, 1987), it blasts is definitively arrested, dye coupling is lost and may be assumed that also in Patella sufficient regu- electrical coupling reduced. In contrast cells which conlating factors might be distributed between the cells to tinue to divide maintain the high levels of junctional enable the first five fast and synchronous cleavages, coupling. The remaining electrical coupling between after which the initial pool of that factor is partially dye-uncoupled trochoblasts and between trochoblasts depleted and the cell cycles become longer. After the and nontrochoblast cells indicates a selective gap juncfifth cleavage, as after the midblastula transition in tional uncoupling which could inhibit the passage of the Xenopus, the cells have to synthesize new factors to presumed regulatory molecules. Therefore we suggest reach the threshold that triggers the following cell that cell cycle arrest of the trochoblasts is associated cycle. The asynchrony between different cells may de- with these changes in gap junctional communication.

SERRAS, DICTUS, AND VAN DEN BIGGELAAR

Cell cycle arrest may be an essential requirement for terminal differentiation, and gap junctional uncoupling may be significant for the isolation of the differentiating trochoblasts from the rest of the undifferentiated embryonic cells. However, it remains to be demonstrated whether changes in gap junctional communication cause or are caused by changes in the cell cycle. It is also important to speculate about the point at which the cell cycle might be arrested. As the trochoblasts in Lymnaea incorporate rH]thymidine in their nuclei (Van den Biggelaar, lgi’la), it may be assumed that at least in this species, and may be also in Patek, the trochoblasts are blocked at the transition between the G2 and the M phase. The lengthening of the G2 phase might enable the cells to express their autonomously determined developmental programme. For instance, in Xenow embryos a premature lengthening of the cell cycle results in the precocious expression of the cell differentiation (Kimelman et aZ., 1987). Also in Lymnuea embryos, lengthening of the cell cycles is associated with a precocious formation of nucleoli and cytoplasm movements (Van den Biggelaar, 1971a). Apart from a possible significance for maintaining cell cycle arrest and for differentiation, the uncoupling of trochoblasts might have a morphogenetic significance as it subdivides the embryo into developmental communication domains. The pretrochal cells located anterior to the ring of uncoupled trochoblasts will form the head ectoderm. The post-trochal domain, at the posterior side of the ring, will give rise to the rest of the body (Verdonk and Van den Biggelaar, 1983). The results presented in this paper as well as previously reported findings in the mollusk Lymnaea (Serras and Van den Biggelaar, 1987) demonstrate that the cells within both the pre- and the post-trochal domains remain highly coupled, but they are poorly coupled to any cell out of the domain. Therefore it seems that once all the trochoblasts become dye uncoupled and partially loose electrical coupling, the embryo is split into two communication compartments. This may play an important role in the independent regulation of the duration of the cell cycles and differentiation in the individual compartments. Alternatively, compartmentalization may have a role in metabolic cooperation in the differently growing tissues, which may have important consequences for the coordination of physiological activities. Uncoupling of single cells during development has also been found in the trophoblasts of the early mouse embryo (Lo and Gilula, 1979) and in the giant yolk cell of the teleost embryo (Kimmel et ah, 1984). The appearance, maintainance, and loss of coupling of Patella trochoblasts suggest, but in no way demonstrate, a role of gap junctional communication in

Intercellular

Communication in Pate&

215

trochoblast development. The resumption of mutual coupling in the differentiated prototroch, however, might have a physiological significance, rather than a developmental significance, such as the coordination of the ciliary movement. The authors thank Dr. R. Dohmen and Dr. J. E. Speksnijder for critically reading the manuscript and Gideon Zwaan for his technical assistance. The Department of Image-Processing and Design is thanked for the photographic services rendered. The anonymous reviewers which helped to improve the manuscript are warmly acknowledged. F.S. was supported by the Netherlands Organization for Scientific Research (NWO/BION). REFERENCES BLENNERHASSET,M. G., and CAVENEY,S. (1984). Developmental compartments are separated by a cell type with reduced junctional permeability. Nature (London) 309,361-364. CLEMENT,C. A. (1952). Experimental studies on germinal localization in IZ~arw,ssa.I. The role of the polar lobe in the determination of the cleavage pattern and its influence in later development. J. Exp. ZooL 121,593-626. DE LAAT, S. W., TERTOOLEN,L. G. J., DORRESTEIJN,A. W. C., and VAN DEN BIGGELAAR,J. A. M. (1980). Intercellular communication patterns are involved in cell determination in early molluscan development. Nature (i&don) 287,546-548. DOHMEN, M. R., ARNOLDS,W. J. A., and SPEKSNIJDER,J. E. (1986). Ionic currents through the cleaving egg of Lymnaea stagnalis (mollusca, gastropoda, pulmonata). In “Ionic Currents in Development” (R. Nuccitelli, Ed.), pp. 181-187. A. R. Liss, New York. DORRESTEIJN,A. W. C., BILINSKI, S. M., VAN DEN BIGGELAAR,J. A. M., and BLUEMINK, J. G. (1982). The presence of gap junctions during early Patelk embryogenesis: An electron microscopical study. Deu. BioL 91,397-401. DORRESTEIJN,A. W. C., WAGEMAKER,H. A., DE LAAT, S. W., and VAN DEN BIGGELAAR,J. A. M. (1983). Dye-coupling between blastomeres in early embryos of Patellu vulgata (Mollusca, Gastropoda): Its relevance for cell determination. Wilhelm Roux’s Arch. Dev. Biol. 192, 262-269. FRASER,S. E., GREEN, C. R., BODE, H. R., BODE, P. M., and GILULA, N. B. (1988). A perturbation analysis of the role of gap junctional communication in developmental patterning. In “Modern Cell Biology” (E. L. Hertzberg and R. G. Johnson, Eds.), Vol. 7, pp. 515-526. A. R. Liss, New York. FRASER,S. E., GREEN, C. R., BODE, H. R., and GILULA, N. B. (1987). Selective disruption of gap junctional communication interferes with a patterning process in hydra. Science 237,49-55. GUERRIER, P., VAN DEN BIGGELAAR,J. A. M., VAN DONGEN,C. A. M., and VERDONK,N. H. (1978). Significance of the polar lobe for the determination of dorsoventral polarity in Dental&m wulgare (da Costa). Dev. Biol. 63,233-242. JANSSEN-DOMMERHOLT, C., VAN WIJK, R., and GEILENKIRCHEN, W. L. M. (1983). Restriction of developmental potential and trochoblast ciliation in Patella embryos. J. EmbryoC Exp. Mcrrphol. 74, 69-77. KIMELMAN, D., KIRSCHNER,M., and SCHERSON,T. (1987). The events of the midblastula transition in Xenopu.s are regulated by changes in the cell cycle. Cell 48,399-407. KIMMEL, C. B., SPRAY,D. C., and BENNETT,M. V. L. (1984). Developmental uncoupling between blastoderm and yolk cell in the embryo of the teleost Fundulus. Dev. BioL 102.483-487.

216

DEVELOPMENTALBIOLOGY

KUHTREIBER,W. M., and VAN DONGEN,C. A. M. (1989). Microinjection of lectins, hyaluronidase, and hyaluronate fragments interferes with cleavage delay and mesoderm induction in embryos of Patella vulgata. Dev. BioL 132,436-441. KUHTREIBER,W. M., VAN DER BENT, J., DORRESTEIJN,A. W. C., DE GRAAF, A., and VAN DEN BIGGELAAR,J. A. M. (1986). The presence of an extracellular matrix between cells involved in the determination of the mesoderm bands in embryos of Patella vulgata (Mollusca, Gastropoda). Wilhelm Roux’s Arch. Dev. BioL 195,265-275. KUHTREIBER,W. M., VAN TIL, E. H., and VAN DONGEN,C. A. M. (1988). Monensin interferes with the determination of the mesodermal cell line in embryos of Patella vulgata. Wilhelm Rwuxk Arch. Dev. BioL 197,10-18. LO, C. W. (1988). Communication compartments in insect and mammalian development. In “Modern Cell Biology” (E. L. Hertzberg and R. G. Johnson, Eds.), Vol. 7, pp. 505-514. A. R. Liss, New York. LO, C. W., and GILULA, N. B. (1979). Gap junctional communication in the post-implantation mouse embryo. Cell l&411-422. LOEWENSTEIN,W. R. (1981). Junctional intercellular communication: The cell-to-cell membrane channel. PhysioL Rev. 61,829-913. NEWPORT,J., and KIRSCHNER, M. (1982a). A major developmental transition in early Xenopus embryos. I. Characterization and timing of cellular changes at the midblastula stage. Cell 30,675-686. NEWPORT,J., and KIRSCHNER,M. (198213).A major developmental transition in early Xenopus embryos. II. Control of the onset of transcription. Cell 30, 687-696. NEWPORT,J., and KIRSCHNER,M. (1984). Regulation of the cell cycle during early Xenopus development. Cell 37,731-742. PURVES,R. D. (1981). “Microelectrode Methods for Intracellular Recording and Ionophoresis.” Academic Press, London. SCHIEREMBERG,E. (1984). Altered cell division rates after laser-induced cell fusion in Nematode embryos. Dev. Biol. 101,240-245. SCHIEREMBERG,E., and WOOD, W. B. (1985). Control of cell cycle timing in early embryos of Caenorhabditis elegans. Dev. BioL 107, 337-354. SERRAS,F., BUULTJENS,T. E. J., and FINBOW,M. E. (1988). Inhibition of dye-coupling in Patella (Mollusca) embryos by microinjection of antiserum against Nephrops (Arthropoda) gap junctions. Exp. Cell Res. 179,282-288. SERRAS,F., K~HTREIBER, W. M., KRUL, M. R. L., and VAN DEN BIGGELAAR, J. A. M. (1985). Cell communication compartments in molluscan embryos. Cell Biol. Int. Rep. 9,731-736. SERRAS,F., and VAN DEN BIGGELAAR, J. A. M. (1987). Is a mosaic

VOLUME 137,199O

embryo also a mosaic of communication compartments? Dev. BioL 120,132-138. SERRAS,F., and VAN DEN BIGGELAAR, J. A. M. (1989). Progressive restrictions in gap junctional communication during development. In “Parallels in Cell to Cell Junctions in Animals and Plants” (A. W. Robards, H. Jongsma, W. J. Lucas, J. Pitts, and D. Spray, Eds.), NATO AS1 Series, Springer-Verlag, London. VAN DENBIGCELAAR,J. A. M. (1971a). Timing of the phases of the cell cycle during the period of asynchronous division up to the 49-cell stage in Lymnaea J. EmbryoL Exp. MorphoL 26,367-391. VAN DEN BIGGELAAR,J. A. M. (1971b). Timing of the phases of the cell cycle with tritiated thymidine and Feulgen cytophotometry during the period of synchronous division in Lymnaea. J, EmlnyoL Exp. MorphoL 26,351-366.

VAN DEN BIGGELAAR,J. A. M. (1977). Development of dorsoventral polarity and mesentoblast determination in Patella vulgata. J Morphol. 154,157-186. VAN DEN BIGGELAAR,J. A. M., and GUERRIER,P. (1979). Dorsoventral polarity and mesentoblast determination as concomitant results of cellular interactions in the mollusk Patella vulgata. Dev. BioL 68, 462-471. VAN DEN BIGGELAAR,J. A. M., and SERRAS,F. (1988). Determinative decisions and dye-coupling changes in the mollusean embryo. In “Modern Cell Biology” (E. L. Hertzberg and R. G. Johnson, Eds.), Vol. 7, pp. 483-493. A. R. Liss, Inc., New York. VAN DONGEN, C. A. M., and GEILENKIRCHEN,W. L. M. (1974). The development of Dentalium with special reference to the significance of the polar lobe. I. II. III. Division chronology and development of the cell pattern in Dentalium dentale (Scaphopoda). Proc. K. Ned. Akad. Wet. Ser. C 77,57-100.

VAN DONGEN, C. A. M., and GEILENKIRCHEN,W. L. M. (1975). The development of Dentalium with special reference to the significance of the polar lobe. IV. Division chronology and development of the cell pattern in Dental&m dentale after removal of the polar lobe at first cleavage. Proc. K. Ned. Akad. Wet. Ser. C 78,358-375. VERDONK,N. H., and VAN DEN BIGGELAAR(1983). Early development and the formation of the germ layers. In “The Mollusca” (N. H. Verdonk, J. A. M. Van den Biggelaar, and A. S. Tompa) Vol. 3, pp. 91-122. Academic Press, New York. WARNER, A. E., and LAWRENCE,P. A. (1982). Permeability of gap junctions at the segmental border in the insect epidermis. Cell 23, 247-252. WILSON,E. B. (1904). Experimental studies in germinal localization. J. Exp. 2001. 1,197-268.