Cell, Vol. 18. 399-409,
October
1979,
Gap Junctional Preimplantation
Copyright
0 1979
by MIT
Communication Mouse Embryo
Cecilia W. Lo* and Norton B. Gilula The Rockefeller University New York, New York 10021
Summary In this study, we examined cell-to-cell communication via gap junctional channels between the cells of the early mouse embryo from the 2-cell stage to the preimplantation blastocyst stage. The extent of communication was examined by monitoring for the presence of ionic coupling, the transfer of injected fluorescein (molecular weight 330) and the transfer of injected horseradish peroxidase (molecular weight 40,000). In the 2-cell, 4-cell and precompaction 8-cell embryos, cytoplasmic bridges between sister blastomeres were responsible for ionic coupling and the transfer of injected fluorescein as well as the transfer of injected horseradish peroxidase. In contrast, no communication was observed between blastomeres from different sister pairs. Junction-mediated intercellular communication was unequivocably detected for the first time in the embryo at the early compaction stage (late 8-tell embryo). At that stage, ionic coupling was present and fluorescein injected into one cell spread to all eight cells of the embryo. Injected horseradish peroxidase was passed to only one other cell, however, again indicating the presence of cytoplasmic bridges between sister blastomeres. Junctional communication with respect to both ionic coupling and dye transfer was retained between all the cells throughout compaction. At the blastocyst stage, trophoblast cells of the blastocyst were linked by junctional channels to other trophoblast cells as well as to cells of the inner cell mass, as indicated by the spread of injected fluorescein. In addition, the extent of communication between the cells of the inner cell mass was examined in inner cell masses isolated by immunosurgery; both ionic coupling and the complete spread of injected fluorescein were observed. Introduction The existence of an intercellular pathway capable of mediating direct cytoplasmic exchange was initially demonstrated by the observation that some excitable cells and many nonexcitable cells are in fact electrically or ionically coupled (Furshpan and Potter, 1959; Loewenstein, 1966; Furshpan and Potter, 1968); that is, after the impalement of two contacting cells with microelectrodes, the injection of current into one cell * Present address: Department of Biological Medical School, Boston, Massachusetts 02115.
Chemistry,
Harvard
in the
can be detected as a voltage deflection in the second cell. Such direct intercellular communication was further indicated by observations that cells in culture were capable of transferring low molecular weight metabolites in a contact-dependent manner (metabolic cooperation) (Subak-Sharpe, Burk and Pitt% 19691, and that the injection of a low molecular weight dye, such as sodium fluorescein, into one cell resulted in the spread of the injected dye to the surrounding cells (Kanno and Loewenstein, 1966; Furshpan and Potter, 1968; Payton, Bennett and Pappas, 1969; Johnson and Sheridan, 1971). The ability of cells to communicate, as assessed by one or more of the above criteria, was correlated morphologically with the presence of a specialized intercellular contact, the gap junction. It is now thought that gap junctions contain the membrane channels that mediate this direct cytoplasmic exchange between cells (Payton et al., 1969; McNutt and Weinstein, 1970; Johnson and Sheridan, 1971; Gilula, Reeves and Steinbach, 1972; Gilula, 1977). Almost all cells that have been examined (either in culture or in viva) can communicate, as indicated by ionic coupling, metabolic coupling or dye transfer, or as inferred from the morphological presence of gap junctional structures. Analysis of the spread of injected molecules of various sizes in communicating cells has revealed that the communicating channels behave as molecular sieves in which the major criterion for the ability of a molecule to cross appears to be molecular size; molecules of 1200 daltons or less can pass from cell to cell (Pitts and Simms, 1977; Simpson, Rose and Loewenstein, 1977). The wide prevalence of this form of direct cell-tocell exchange has led to much speculation about its potential biological function(s) in vivo (Loewenstein, 1966; Sheridan, 1968; Furshpan and Potter, 1968). One interesting possibility is that these communication channels may passively mediate the formation of morphogenetic gradients in developing embryos (Wolpert, 1978). Morphogenetic gradients are hypothesized to consist of gradients of molecules (chemical gradients) which specify the “positional information” for directing the differentiation of each cell in an embryo. Depending on the location of the cell within this gradient, it would thus be “instructed” to differentiate along a particular pathway (Wolpert, 1969). Ionic coupling has been found between cells in the early embryos of many organisms. These include starfish (Ashman, Kanno and Loewenstein, 1964; Tupper, Saunders and Edwards, 1970; Tupper and Saunders, 19721, squid (Potter, Furshpan and Lennox, 19661, Fundulus (Bennett and Trinkaus, 1970; Bennett, Spira and Pappas, 1972), Triturus (Ito and Hori, 1966; Ito and Loewenstein, 19691, Xenopus (Slack and Palmer, 1969; Sheridan, 1971; DiCaprio. French and Saun-
Cell 400
ders, 1975) and chick (Sheridan, 1968). In the case of the starfish embryo, ionic coupling was not detected until the 32-cell stage, and at that time injected fluorescein was not observed to pass between the cells (Tupper and Saunders, 1972). Similar observations have been made on the early Xenopus and Fundulus embryos, where ionic coupling has been detected without the visible transfer of injected fluorescein (Slack and Palmer, 1969; Bennett et al., 1972). Recently, however, Bennett, Spira and Spray (1978) reported that the fluorescent dye Lucifer yellow does pass between early embryonic cells of Fundulus and that, in fact, “once we expected it to be there,” fluorescein transfer could also be observed between closely coupled cell pairs. The communication properties of mammalian embryos have not yet been examined except in one study reported in a review (Powers and Tupper, 1978) in which no ionic coupling was observed between the two blastomeres of the 2-cell stage mouse embryo. This study was undertaken to characterize the cell-toceil communication properties of early mouse em-
bryos from the 2-cell stage to the preimplantation blastocyst. Ionic coupling and the cell-to-cell transfer of injected fluorescein and horseradish peroxidase were analyzed at each stage. Results As the fertilized ovum of the rodent develops its cell number is increased by successive cleavages, and at the 8-cell stage (morulae) the embryo undergoes a process called compaction. The outlines of individual blastomeres are clearly visible prior to compaction (Figures l-5); but as this process continues the individual blastomeres are no longer distinguishable, and instead the morula is outlined by a smooth surface (Figures 6 and 7). The size of the embryo is also decreased during this process, suggesting a compaction of the cell mass. Junctional complexes, including tight and gap junctions, are also formed between the blastomeres (Ducibella et al., 1975; Magnuson, Jacobson and Stackpole, 1978). As a result, a permeability seal which makes possible the subsequent ac-
b
Figure
1. Intercellular
(a) Each cell of an early
Communication 2-cell
mouse
in the Early embryo
2-Cell
(in the center)
Embryo
via a Cytoplasmic
was impaled
Bridge
with a microelectrode.
(Inset)
As a current
pulse
(top trace)
was injected
into the blastomere on the right, a voltage deflection was recorded in that cell (bottom trace) and also in the adjacent blastomere on the left (middle trace). The horizontal calibration bar = 100 msec and the vertical calibration bar = 5 nA and 20 mV. (b) HRP was injected into the blastomere on the left. The dark reaction product indicates the presence of HRP in both blastomeres. (c. d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the initiation of dye injection. (c) 4 min and (d) 16 mm after the start of Injection. The scale bar in (b) = 45.6 pm in (a-d).
Communication 401
in the Preimplantation
Mouse
Embryo
cumulation of fluid is established, leading to the formation of the blastocyst. The blastocyst is a hollow sphere consisting of a single layer of trophoblast cells which forms the wall of the blastocyst vesicle. The inner cell mass (ICM), an aggregate of smaller cells, is located inside the vesicle where it is attached to one pole of the sphere. Embryos at various stages of development were retrieved from pregnant mice, placed in culture and examined for their intercellular communication properties. The stages examined and the approximate time at which they were harvested were as follows: early 2-cell (afternoon of day 2); late 2-cell (early evening of day 2); early 4-cell (late evening of day 2); late 4cell (early morning of day 3); precompaction 8-cell (noon of day 3); late compaction 8-cell (afternoon and evening of day 3); and blastocyst (afternoon of day 4). 2-Cell Stage The early 2-cell mouse embryo is ionically coupled, as indicated by the spread of injected current from one blastomere to the other (Figure 1 a). Simultaneous
injection of fluorescein and monitoring of ionic coupling revealed that fluorescein can also pass between the two cells (Figures 1 c and 1 d). Since blastomeres of early embryos frequently undergo incomplete cytokinesis or do not complete cytokinesis for extended time periods after completion of karyokinesis, it was necessary to establish whether the observed ionic coupling and dye transfer were mediated by a cytoplasmic bridge rather than by low resistance junctional channels. For this purpose, horseradish peroxidase (HRP, molecular weight 40,000), a molecular which is above the exclusion limit of junctional channels, was injected into one blastomere while fluorescein was being injected into the other. The presence of ionic coupling was also monitored using the HRP- (or fluorescein-) filled electrode, during which time current was only pulsed from the fluorescein(or HRP-) filled electrode. After the pattern of dye spread was recorded photographically, the embryos were fixed and processed for the detection of HRP activity (see Experimental Procedures). HRP activity in the blastomeres was identified by a dark brown reaction product. Such an analysis unequivocally demonstrated the
b
Figure
2. The Absence
of Intercellular
Communication
in the Late 2-Cell
Embryo
(a) Each cell of a late 2-cell embryo was impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the left blastomere (bottom trace). no voltage deflection was detected in the adjacent cell (middle trace). The horizontal calibration bar = 100 msec and the vertical calibration bar = 5 nA and 20 mV. (b) HRP was injected into the blastomere on the right. The dark reaction product indicated the presence of HRP only in the blastomere into which HRP was injected. (c. d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the start of dye injection. (c) 4 min and (d) 19 min after the start of injection. The scale bar in (b) = 45.6 firn in (a-d).
Cell 402
b
Figure
3. Intercellular
Communication
in the 4-Cell
Embryo
via Cytoplasmic
Bridges
(a) Two cells of a 4-cell embryo were each impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the upper impaled cell, a voltage deflection was recorded in that cell (lower trace) and also in the lower impaled cell (middle trace). The calibration bars are the same as for Figure 1, except that for the lower trace the bar = 10 mV. (b) HRP was injected into the blastomere on the left. The dark reaction product indicated the presence of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c, d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the start of dye injection, (c) 4 min and (d) 14 min after the start of injection. The scale bar in (b) = 45.6 gm in (a-d)
presence of a cytoplasmic bridge, since HRP moved freely from the site of injection to the other blastomere (Figure 1 b). Thus this cytoplasmic continuity can account for the observed passage of fluorescein molecules and ionic coupling between the two blastomeres (Figures lc and Id). In the late 2-cell embryo, in contrast to the early 2-cell embryo, injected HRP and fluorescein did not pass into the adjacent blastomere and no ionic coupling was detected (Figure 2). At this stage, the cytoplasmic continuity must have been abolished as a result of the completion of cytokinesis. This observation further supports the finding that no gap junctional communication was involved in the cellto-cell permeability observed in the early 2-cell embryo. 4-Cell Stage Examination of 4-cell embryos for communication using similar methods revealed two typical patterns of transfer. In one case, ionic coupling was detected between the two impaled blastomeres, and both fluorescein and HRP passed between the same impaled cell pair (Figure 3). In the other case, ionic coupling
was not detected between the two impaled cells even though injected fluorescein and HRP were passed. In this case, the injected HRP and fluorescein spread to each of two different blastomeres, neither of which were impaled (Figure 4). These observations indicate that in the first case a cytoplasmic bridge between the impaled cell pair mediated the observed ionic coupling and dye transfer, while in the second case no coupling was detected because the two impaled cells were derived from two different mitotic sister pairs. Hence there was no junction-mediated transfer of ions or molecules in the 4-cell embryo, but rather the ionic coupling and dye transfer observed were mediated via cytoplasmic bridges that exist between sister blastomeres of the second mitotic cleavage. These two patterns were observed in both early and late 4-cell embryos. S-Cell Stage Analysis of precompaction early a-cell embryos for intercellular communication yielded results identical to those observed in the 4-cell embryos in which two distinct patterns of transfer were observed. In one
Communication 403
in the Preimplantation
Mouse
Embryo
b
Figure
4.
Intercellular
(a) Two cells of a 4-tell on the right, a voltage trace). The calibration revealed the presence the blastomere on the wrn in (a-d).
Communication
in a 4-Cell
Embryo
via Cytoplasmic
Bridges
embryo were each impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the impaled cell deflection was detected in that cell (bottom trace) but none was detected in the impaled blastomere on the left (middle bars are the same as for Figure 1. (b) HRP was injected into the impaled blastomere on the left, The dark reaction product of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c. d) Fluorescein was injected into right and fluorescence images were recorded at various times after the start of dye injection. The scale bar in (b) = 45.6
case, ionic coupling was detected together with the transfer of injected fluorescein and HRP, so that HRP and fluorescein were both present in the same two impaled cells (data not shown). In the other case, no coupling was detected between the injected cells; although the injected fluorescein and HRP were both transferred to other blastomeres, they were passed into two different ones (Figure 5). These results again indicate that the ionic coupling and dye transfer were mediated by a cytoplasmic bridge which connects sister blastomeres from the previous mitotic cleavage, and not by junctional channels. a-cell embryos which were beginning to undergo compaction were similarly examined for ionic coupling, transfer of injected fluorescein and injected HRP. Ionic coupling was observed and the injected fluorescein spread to all eight cells of the embryo (Figure 6). In contrast, the injected HRP was passed to only one other blastomere, probably the sister of the blastomere into which the HRP was being injected (Figure 6). These observations provide direct evidence that junction-mediated communication path-
ways are generated and expressed for the first time in the mouse embryo at this developmental stage. As compaction progressed to completion, ionic coupling was maintained and the injection of fluorescein into one cell of the morulae always resulted in the spread of dye to every cell of the embryo (Figure 7). Blastocyst Shortly after compaction the mouse embryo undergoes cavitation, a process whereby a cavity is generated inside the morula. The cavity enlarges in size, eventually transforming the embryo to a blastocyst with a large blastocoelic cavity. The injection of dye into a trophoblast cell of the blastocyst resulted in dye spread throughout the entire embryo, including the inner cell mass region (data not shown). Due to the three-dimensional arrangement of the blastocyst structure, however, it was difficult to discern clearly the extent of dye passage in the ICM. To determine the extent of dye spread in the ICM more precisely, ICMs were isolated from blastocysts using immunosurgery (Salter and Knowles, 1975; see
Cdl 404
b
Figure
5. Intercellular
Communication
in the Precompacted
S-Cell E:mbryo
via Cytoplasmic
Bridges
(a) Two cells of a precompaction early a-cell embryo were each impaled with a microelectrode. (inset) As a current pulse (top trace) was injected into the impaled cell on the right, a voltage deflection was detected in that blastomere (bottom trace) but none was detected in the impaled blastomere on the left (middle trace). The calibration bars are the same as for Figure 1. (b) HRP was injected into the impaled blastomere on the right, The dark reaction product revealed the presence of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c, d) Fluorescein was injected into the blastomere on the right and fluorescence images were recorded at various times after the start of dye injection. (c) 4 min and (d) 25 min afler the start of injection. The scale bar in (b) = 45.6 pm in (a-d).
Experimental Procedures). Ionic coupling in the ICM was monitored simultaneously with the injection of fluorescein. The cells of the ICM were ionically coupled and the injected dye filled the entire ICM (Figure 8), thus indicating that all the cells of the ICM were linked by junctional pathways. Discussion The observations made in this study reveal that the early mouse embryo does not possess junctional channels until early compaction at the 8-cell stage. At that time, the onset of ionic coupling is accompanied by the simultaneous spread of injected fluorescein into all eight cells of the embryo. The communication that was observed throughout the embryo at this stage was not due to the presence of cytoplasmic bridges, since injected HRP did not move into any cells other than the one sister blastomere of the injected cell. In previous studies of cell-to-cell communication in early embryonic cells of comparable stages in development, ionic coupling was usually observed, al-
though it was never determined whether cytoplasmic bridges were responsible for the observed communication in those studies (Ito and Hori, 1986; Ito and Loewenstein, 1969; Slack and Palmer, 1969; Tupper and Saunders, 1972). The one report of the absence of ionic coupling in the 2-cell mouse embryo (Powers and Tupper, 1978) is in agreement with the ObSeNStions on the late 2-cell embryo reported here. In addition, HRP injections in this study have shown that the observed ionic coupling and dye transfer in the early 2-cell embryo were not mediated by a junctional pathway but rather via a cytoplasmic bridge. It has been reported that in some early embryonic cells ionic coupling was not accompanied by the transfer of injected fluorescein (Slack and Palmer, 1969; Tupper and Saunders, 1972; Bennett et al., 1972). To explain these observations it was suggested that gap junctions of embryonic cells have channels of smaller pore size and that they would therefore, unlike the gap junctions of adult cells, allow the passage of ions but not of the larger fluorescein molecules. The present observations showed, however, that junction-me-
Communication 405
Figure
in the Preimplantation
6. Intercellular
Communication
Mouse
Embryo
in the Early Compacted
8-Cell
Embryo
via Junctional
Channels
(a) Two cells of an early compaction 8-cell embryo were each impaled with a microelectrode. (b) As current (top trace) was injected into the deflection was detected in that cell (bottom trace) and also in the impaled cell on the right (middle trace). The blastomere on the left, a voltage calibration bars are the same as for Figure 1. (c-e) Fluorescein was injected into the blastomere on the left and fluorescence images were recorded at various times after the start of injection. (c) 4 min. (d) 16 min and (e) 25 min after the start of injection. (f) HRP was injected into the blastomere on the right. The dark reaction product indicated the presence of HRP in two blastomeres. the blastomere into which HRP was injected and also an adjacent blastomere. The scale bar in (a) = 45.6 gm in (a) and (c-f).
diated ionic coupling in the mouse embryo was always accompanied by the passage of injected fluorescein. This observation is consistent with a recent report that the previous failure to observe dye transfer between Fundulus embryonic cells may be more apparent then real; with careful reexamination, both ionic coupling and dye transfer have been observed among these embryonic cells (Bennett et al., 1978).
It is interesting to consider the onset of junctionmediated communication in the early mouse embryo in the context of the development of the oocyte. Rat oocytes communicate with interacting follicle cells (cumulus oophorus) prior to ovulation (Albertini and Anderson, 1974; Gilula, Epstein and Beers, 1978). As meiotic maturation resumes at the time of ovulation, the oocyte becomes progressively uncoupled from
Cell 406
Figure
7. ionic Coupling
and Fluorescein
(a) Two cells of a late compaction 8-tell call on the right, a voltage deflection was bars are the same as for Figure 1. (c, d) after the start of injection. (c) 4 min and
Dye Transfer
in the Late Compaction
E-Cell Embryo
embryo were each impaled with a microelectrode. (b) As a current pulse (top trace) was injected into the detected in that cell (bottom trace) and also in the impaled cell on the left (middle trace). The calibration Fluorescein was injected into the cell on the right. Fluorescence images were recorded at various times (d) 13 min after the start of injection. The scale bar in (a) = 45.6 pm in (a), (c) and (d).
the surrounding cumulus cells (Gilula et al., 1978); the prefertilization oocyte is therefore a noncommunicating but previously communication-competent cell type. Morphological studies have shown that mouse embryos do not have gap junctions until the 8-cell stage (Ducibella et al., 1975; Magnuson et al., 1978). These observations are consistent with our findings, and suggest that the ability to form gap junctions in the rodent embryo is shut down at ovulation and not reexpressed until early compaction at the 8-cell stage. The time at which junctional communication is initially expressed, at the early compaction stage, is also approximately the time during which the first determination event occurs in the mouse embryo (Gardner and Rossant, 1976). This determination event appears to be strictly dependent upon the position of the blastomere, so that only the outside cells of the compacted morulae form trophoblasts, while the inside cells give rise to the cells of the inner cell mass in the blastocyst. On this basis, Tarkowski and Wroblewska (1967) have proposed an “inside-outside” hypothesis in which geographical location alone determines the
developmental fate of any blastomere. Since gap junctions are formed at approximately the time of initiation of trophoblast differentiation, it seems plausible that such “inside-outside” positional information might in fact be specified by an intercellular morphogenetic gradient formed via gap junctional channels. In addition, the formation of a permeability barrier by tight junctions (which also occurs at this time) would enhance differences between the environment of the cells buried on the inside from those exposed to the outside. The presence of gap junctions linking them could then provide a pathway for transferring and thus passively generating an intercellular gradient of ions, metabolites or other substances from the outside cells to the inside cells and vice versa. If intercellular communication is indeed involved in mediating the differentiation and development of the mouse embryo, it might be possible to detect changes in communication patterns between cells of the embryo which parallel its development. To investigate this possibility further, we have analyzed the cell-tocell communication properties of mouse blastocysts
Communication 407
Figure
in the Preimplantation
8. Ionic Coupling
and Fluorescein
(a) Two ceils of an ICM isolated was injected into the cell on the trace). The calibration bars are were recorded at various times (a). (c) and cd).
Mouse
Embryo
Dye Transfer
in the Isolate !d Inner
by immunosurgery from a blastocys left, a voltage deflection was deted the same as for Figure 1. (c. d) Fluor afler the start of injection. (c) 4 min
which have undergone an implantation-like process in vitro (see accompanying paper by Lo and Gilula, 1979). Experimental Procedures SWR (Jackson Laboratories; Bar Harbor, Maine) and ICR (Flow Laboratories, Inc.; Rockville. Maryland) female mice were superovulated with injections of 4 IU of pregnant mare serum gonadotropin (PMSG) and 4 IU of human chorionic gonadotropin (HCG) (Organon, Inc.; West Orange, New Jersey) and caged with NCS male mice (Laboratory Animal Research Center: Rockefeller University) overnight (Runner and Palm, 1953). Pregnant females were identified the following morning by the presence of a vaginal sperm plug; this was considered day 1 of pregnancy. The animals were sacrificed by cervical dislocation at various time intervals and the oviducts or uteri were removed. With the aid of a dissecting microscope and a syringe, the embryos were flushed from the excised oviduct or uterus and subsequently recovered in a petri dish. The zona pellucida or the harvested embryos were removed by a 1 O-l 5 min incubation in 0.5% pronase (Sigma Chemical Co.; St. Louis, Missouri) in Dulbecco’s phosphate-buffered saline (Mintz. 1962). The naked embryos were then placed in a 35 mm petri dish (Bio-Quest, BBL and Falcon Products; Cockeysville. Maryland) in Leibovitz medium (Grand Island Biological Co.: Grand Island. New York) and kept in a 37°C incubator. After the embryos attached to the dish (in -5-15 min). they were immediately impaled with microelectrodes. Embryos impaled after being in culture for l-4 hr gave results identical to those of embryos examined immediately after culturing. The isolated inner cell masses
Cell Mass
,t were each impaled with a microelectrode. (b) As a current pulse (top trace) ted in that cell (bottom trace) and also in the impaled cell on the right (I niddle ‘escein was injected into the impaled cell on the lefl and fluorescence itnages and (d) 10 min after the start of injection. The scale bar in (a) = 45.6 pm in
of the blastocysts were obtained after removing the trophoblasts immunosurgically (Salter and Knowles, 1975) and then the ICMs were allowed to attach to a 35 mm plastic petri dish for subsequent microelectrode impalements. lmmunosurgery The inner cell mass from blastocysts were isolated according to the method of Solter and Knowles (1975). Blastocysts were incubated in a 1 :lO dilution of rabbit anti-mouse fibroblast antiserum (provided by M. Friedlander, Rockefeller University) in Dulbecco’s modified Eagle’s medium with 10% heat-inactivated fetal bovine serum (DEM-HI) at 37°C for 45 min. The blastocysts were then washed in DME-HI and incubated in a 1:4 dilution of guinea pig complement (Beckman Instruments, Inc.; Fullerton, California) in DME-HI at 37°C. After 30 min. the embryos were removed and washed in DME-HI, and the outer layer of trophoblastic cells was stripped by repeated passage through a micropipette. Electrophysiology The methods and apparatus used for monitoring ionic coupling and iontophoretic injections have been described previously (Gilula et al., 1978). Briefly, glass microelectrodes filled with 3 M KCI with tip resistances of 40-60 MD were used for determining ionic coupling. For dye injections, microelectrodes with tip resistances of 15-25 MD (when filled with 3 M KCI) were filled with 5% sodium fluorescein (w/ v). lontophoresis was conducted with a 5 x lo-’ A current pulse of 0.5 set duration at 1 set intervals. The spread of fluorescence in the embryo due to the spread of injected fluorescein was visualized under
Cell 408
phase optics using a 100-W tungsten halogen lamp and a combination of EG 12 and 23 excitation filters, a KP 490 cut-off filter and a KP 530 barrier filter. All photographs of fluorescein injections were recorded on Kodak Tri-X Pan film with 3.5 min exposures. The film was developed with Acufine (Acufine. Inc.; Chicago, Illinois). For horseradish peroxidase injections, electrodes with tip resistances of 15-20 MS2 (when filled with 3 M KCI) were filled with 10 mg/ml of horseradish peroxidase (Type II, Sigma Chemical Co.: St. Louis, Missouri) in 0.2 M KCI. lontophoresis was conducted with a 510 x 1 Om9 A current pulse (depolarizing) of 0.5 set duration at 1 set intervals for a total of 1 O-l 5 min. After removal of microelectrodes. the embryos were fixed with 2.5% glutaraldehyde (w/v) (Polysciences Inc.; Warrington. Pennsylvania) in 0.1 M cacodylate buffer (pH 7.3) (Sigma) for 15 min and then incubated with 1 mg/ml of Hanker-Yates Reagent (p-phenylenediamine dihydrochloride and pyrocatechol; Polysciences) in 0.1 M Tris buffer (pH 7.6) containing 0.01% H202, After 5-15 min. a dark brown color product was visible in ceils containing injected horseradish peroxidase. and the reaction was terminated by rinsing the samples with the Tris buffer. The embryos thus processed were photographed with bright field optics using Kodak Tri-X pan film. Acknowledgments
J. (1976). Determination during embryin Mammals, Ciba Foundation Sympo-
Gilula. N. B. (1977). Gap junctions and cell communication, In International Cell Biology (1976-l 977). B. R. Brinkley and K. R. Porter, eds. (New York: The Rockefeller University Press), pp. 61-69. Gilula. N. B.. Reeves, coupling. ionic coupling
R. 0. and Steinbach, A. (1972). Metabolic and cell contacts. Nature 235, 262-265.
Gilula, N. B., Epstein, M. L. and Beers, W. H. (1978). Cell-to-cell communication and ovulation. A study of the cumulus-oocyte complex. J. Cell Biol. 78, 58-75. Ito, S. and Hori, N. (1966). Electrical characteristics cells during cleavage. J. Gen. Physiol. 49, 1019-l
of Trirurus 027.
Ito. S. and Loewenstein, W. R. (1969). Ionic communication early embryonic cells. Dev. Biol. 19, 226-243. Johnson, R. G. and Sheridan, cells in culture: ultrastructure 719. Kanno. Y. and large molecules.
egg
between
J. D. (1971). Junctions between cancer and permeability. Science 7 74, 717-
Loewenstein, W. R. (1966). Nature 272, 629-630.
Cell-to-cell
passage
of
Lo, C. W. and Gilula. N. B. (1979). Gap junctional communication the postimplantation mouse embryo. Cell 78, 41 l-422.
We would like to thank Theodore Lawrence for assistance and helpful discussions concerning the electrophysiology. Dr. Sidney Strickland and Mary Jean Sawey for help and advice in the mating of mice and the handling of mouse embryos, Asneth Kloesman and Kathy Wall for help in the preparation of figures and Madeleine Naylor for secretarial assistance. This research was supported by grants from the USPHS and The Rockefeller Foundation. N.B.G. is a recipient of a Research Career Developmental Award. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
Gardner, R. L. and Rossant. ogenesis. In Embryogenesis sium 40. 5-25.
May 2, 1979
References
Loewenstein, W. R. (1966). Permeability NY Acad. Sci. 737. 441-472.
of membrane
junctions.
in Ann.
McNutt, N. S. and Weinstein, R. S. (1970). The ultrastructure of the nexus. A correlated thin section and freeze-cleave study. J. Cell Biol. 4 7, 666-688. Magnuson. T.. Jacobson, J. B. and Stackpole. C. W. (1978). Relationship between intercellular permeability and junction organization in the preimplantation mouse embryo. Dev. Biol. 67, 214-224. Mintz, B. (1962). Experimental study of the developing mammalian egg: removal of the zona pellucida. Science 738, 594-595. Payton. B. W.. Bennett, M. V. L. and Pappas, bility and structure of junctional membranes apse. Science 766, 1641-l 643.
G. D. (1969). Permeaat an electrotonic syn-
Pitts. J. D. and Simms, J. W. (1977). Permeability of junctions between animal cells. Intercellular transfer of nucleotides but not of macromolecules. Exp. Cell Res. 704, 153-l 63.
Albertini, D. F. and Anderson, E. (1974). The appearance of structure and intercellular connections during the ontogeny of the rabbit ovarian follicle with particular reference to gap junctions. J. Cell Biol. 63, 234-250.
Potter, 0. D.. Furshpan. E. J. and Lennox, E. S. (1966). Connections between cells of the developing squid as revealed by electrophysiological methods. Prog. Nat. Acad. Sci. USA 55, 328-335.
Ashman. formation 604-605.
Powers, R. D. and Tupper. J. T. (1978). Intercellular in the early embryo. In Intercellular Communication, ed. (New York: Plenum Press). pp. 231-251.
R. F.. Kanno. Y. and Loewenstein. W. R. (1964). of a high resistance barrier in a dividing cell. Science
Bennett, M. V. L. and Trinkaus. J. P. (1970). between embryonic cells by way of extracellular ized junctions. J. Cell Biol. 44, 592-610. Bennett, M V. L., Spira. of electrotonic junctions Biol. 29. 419-435.
M. E. and Pappas, between embryonic
The 14.5,
Electrical coupling space and special-
G. D. (1972). Properties cells of Fmdulus. Dev.
Bennett, M. V. L.. Spira. M. E. and Spray, D. C. (1978). Permeability of gap junctions between embryonic cells of fundulus: a reevaluation. Dev. Biol. 65, 114-l 25. DiCaprio. R. A., French, A. S. and Saunders, E. J. (1975). connectivity in the E-cell Xenopus embryo. Correlation and morphological investigations. Biophys. J. 75, 373.
Intercellular of electrical
Ducibella, J.. Albertini. D.. Anderson, E. and Biggers. J. D. (1975). The preimplantation mammalian embryo: characterization of intercellular junctions and their appearance during development. Dev. Biol. 47, 231-250. Furshpan. E. J. and Potter, D. D. (1959). motor synapses of the crayfish. J. Physiol.
Transmission at the giant 745, 289-325.
Furshpan, E. J. and Potter, D. D. (1968). Low-resistance between cells in embryos and tissue culture. Curr. Topics 3. 95-l 27.
junctions Dev. Biol.
communication W. C. De Mello.
Runner, M. N. and Palm, J. (1953). Transplantation and survival of unfertilized ova of the mouse in relation to postovulatory age. J. Exp. Zool. 124, 303-316. Sheridan, J. D. (1968). Electrophysiological evidence for low-resistance intercellular junctions in the early chick embryo. J. Cell Biol. 37, 650-859. Sheridan, J. D. (1971). Dye movement and low-resistance junctions between reaggregated embryonic cells. Dev. Biol. 26, 627-636. Simpson, I., Rose, B. and Loewenstein. W. R. (1977). Size limit of molecules permeating the junctional membrane channels. Science 795, 294-296. Slack, C. and Palmer, J. F. (1969). The permeability of intercellular junctions in the early embryo of Xenopus laevis studied with a fluorescence tracer. Exp. Cell Res. 55. 416-419. Salter, D. and Knowles, B. (1975). lmmunosurgery cyst. Proc. Nat. Acad. Sci. USA 72, 5099-5iO2.
of mouse
blasto-
Subak-Sharpe. H.. Burk. R. R. and Pitts. J. D. (1969). Metabolic cooperation between biochemically marked mammalian cells in tissue culture. J. Cell Sci. 4. 353-367. Tarkowski, A. K. and Wroblewska, J. (1967). Development tomeres of mouse eggs isolated at the four- and eight-cell
of blasstage. J.
C$munication
Embryol.
in the Preimplantation
Exp. Morphol.
78, 155-l
Mouse
Embryo
80.
Tupper. J. T. and Saunders, J. W.. Jr. (1972). intercellular in the early Aster& embryo. Dev. Biol. 27, 546-554.
permability
Tupper. J. T.. Saunders, J. W.. Jr. and Edwards, C. (1970). The onset of electrical coupling between cells in the developing starfish embryo. J. Cell Biol. 46, 167-l 90. Wolpert, L. (1969). Positional information and the spatial cellular differentiation. J. Theoret. Biol. 25. l-47.
pattern
of
Wolpert, L. (1976). Gap junctions: channels for communication in development. In Intercellular Junctions and Synapses, J. Feldman, N. 8. Gilula and J. D. Pitt% eds. (London: Chapman and Hall), pp. 8394.
October
1979,
Gap Junctional Preimplantation
Copyright
0 1979
by MIT
Communication Mouse Embryo
Cecilia W. Lo* and Norton B. Gilula The Rockefeller University New York, New York 10021
Summary In this study, we examined cell-to-cell communication via gap junctional channels between the cells of the early mouse embryo from the 2-cell stage to the preimplantation blastocyst stage. The extent of communication was examined by monitoring for the presence of ionic coupling, the transfer of injected fluorescein (molecular weight 330) and the transfer of injected horseradish peroxidase (molecular weight 40,000). In the 2-cell, 4-cell and precompaction 8-cell embryos, cytoplasmic bridges between sister blastomeres were responsible for ionic coupling and the transfer of injected fluorescein as well as the transfer of injected horseradish peroxidase. In contrast, no communication was observed between blastomeres from different sister pairs. Junction-mediated intercellular communication was unequivocably detected for the first time in the embryo at the early compaction stage (late 8-tell embryo). At that stage, ionic coupling was present and fluorescein injected into one cell spread to all eight cells of the embryo. Injected horseradish peroxidase was passed to only one other cell, however, again indicating the presence of cytoplasmic bridges between sister blastomeres. Junctional communication with respect to both ionic coupling and dye transfer was retained between all the cells throughout compaction. At the blastocyst stage, trophoblast cells of the blastocyst were linked by junctional channels to other trophoblast cells as well as to cells of the inner cell mass, as indicated by the spread of injected fluorescein. In addition, the extent of communication between the cells of the inner cell mass was examined in inner cell masses isolated by immunosurgery; both ionic coupling and the complete spread of injected fluorescein were observed. Introduction The existence of an intercellular pathway capable of mediating direct cytoplasmic exchange was initially demonstrated by the observation that some excitable cells and many nonexcitable cells are in fact electrically or ionically coupled (Furshpan and Potter, 1959; Loewenstein, 1966; Furshpan and Potter, 1968); that is, after the impalement of two contacting cells with microelectrodes, the injection of current into one cell * Present address: Department of Biological Medical School, Boston, Massachusetts 02115.
Chemistry,
Harvard
in the
can be detected as a voltage deflection in the second cell. Such direct intercellular communication was further indicated by observations that cells in culture were capable of transferring low molecular weight metabolites in a contact-dependent manner (metabolic cooperation) (Subak-Sharpe, Burk and Pitt% 19691, and that the injection of a low molecular weight dye, such as sodium fluorescein, into one cell resulted in the spread of the injected dye to the surrounding cells (Kanno and Loewenstein, 1966; Furshpan and Potter, 1968; Payton, Bennett and Pappas, 1969; Johnson and Sheridan, 1971). The ability of cells to communicate, as assessed by one or more of the above criteria, was correlated morphologically with the presence of a specialized intercellular contact, the gap junction. It is now thought that gap junctions contain the membrane channels that mediate this direct cytoplasmic exchange between cells (Payton et al., 1969; McNutt and Weinstein, 1970; Johnson and Sheridan, 1971; Gilula, Reeves and Steinbach, 1972; Gilula, 1977). Almost all cells that have been examined (either in culture or in viva) can communicate, as indicated by ionic coupling, metabolic coupling or dye transfer, or as inferred from the morphological presence of gap junctional structures. Analysis of the spread of injected molecules of various sizes in communicating cells has revealed that the communicating channels behave as molecular sieves in which the major criterion for the ability of a molecule to cross appears to be molecular size; molecules of 1200 daltons or less can pass from cell to cell (Pitts and Simms, 1977; Simpson, Rose and Loewenstein, 1977). The wide prevalence of this form of direct cell-tocell exchange has led to much speculation about its potential biological function(s) in vivo (Loewenstein, 1966; Sheridan, 1968; Furshpan and Potter, 1968). One interesting possibility is that these communication channels may passively mediate the formation of morphogenetic gradients in developing embryos (Wolpert, 1978). Morphogenetic gradients are hypothesized to consist of gradients of molecules (chemical gradients) which specify the “positional information” for directing the differentiation of each cell in an embryo. Depending on the location of the cell within this gradient, it would thus be “instructed” to differentiate along a particular pathway (Wolpert, 1969). Ionic coupling has been found between cells in the early embryos of many organisms. These include starfish (Ashman, Kanno and Loewenstein, 1964; Tupper, Saunders and Edwards, 1970; Tupper and Saunders, 19721, squid (Potter, Furshpan and Lennox, 19661, Fundulus (Bennett and Trinkaus, 1970; Bennett, Spira and Pappas, 1972), Triturus (Ito and Hori, 1966; Ito and Loewenstein, 19691, Xenopus (Slack and Palmer, 1969; Sheridan, 1971; DiCaprio. French and Saun-
Cell 400
ders, 1975) and chick (Sheridan, 1968). In the case of the starfish embryo, ionic coupling was not detected until the 32-cell stage, and at that time injected fluorescein was not observed to pass between the cells (Tupper and Saunders, 1972). Similar observations have been made on the early Xenopus and Fundulus embryos, where ionic coupling has been detected without the visible transfer of injected fluorescein (Slack and Palmer, 1969; Bennett et al., 1972). Recently, however, Bennett, Spira and Spray (1978) reported that the fluorescent dye Lucifer yellow does pass between early embryonic cells of Fundulus and that, in fact, “once we expected it to be there,” fluorescein transfer could also be observed between closely coupled cell pairs. The communication properties of mammalian embryos have not yet been examined except in one study reported in a review (Powers and Tupper, 1978) in which no ionic coupling was observed between the two blastomeres of the 2-cell stage mouse embryo. This study was undertaken to characterize the cell-toceil communication properties of early mouse em-
bryos from the 2-cell stage to the preimplantation blastocyst. Ionic coupling and the cell-to-cell transfer of injected fluorescein and horseradish peroxidase were analyzed at each stage. Results As the fertilized ovum of the rodent develops its cell number is increased by successive cleavages, and at the 8-cell stage (morulae) the embryo undergoes a process called compaction. The outlines of individual blastomeres are clearly visible prior to compaction (Figures l-5); but as this process continues the individual blastomeres are no longer distinguishable, and instead the morula is outlined by a smooth surface (Figures 6 and 7). The size of the embryo is also decreased during this process, suggesting a compaction of the cell mass. Junctional complexes, including tight and gap junctions, are also formed between the blastomeres (Ducibella et al., 1975; Magnuson, Jacobson and Stackpole, 1978). As a result, a permeability seal which makes possible the subsequent ac-
b
Figure
1. Intercellular
(a) Each cell of an early
Communication 2-cell
mouse
in the Early embryo
2-Cell
(in the center)
Embryo
via a Cytoplasmic
was impaled
Bridge
with a microelectrode.
(Inset)
As a current
pulse
(top trace)
was injected
into the blastomere on the right, a voltage deflection was recorded in that cell (bottom trace) and also in the adjacent blastomere on the left (middle trace). The horizontal calibration bar = 100 msec and the vertical calibration bar = 5 nA and 20 mV. (b) HRP was injected into the blastomere on the left. The dark reaction product indicates the presence of HRP in both blastomeres. (c. d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the initiation of dye injection. (c) 4 min and (d) 16 mm after the start of Injection. The scale bar in (b) = 45.6 pm in (a-d).
Communication 401
in the Preimplantation
Mouse
Embryo
cumulation of fluid is established, leading to the formation of the blastocyst. The blastocyst is a hollow sphere consisting of a single layer of trophoblast cells which forms the wall of the blastocyst vesicle. The inner cell mass (ICM), an aggregate of smaller cells, is located inside the vesicle where it is attached to one pole of the sphere. Embryos at various stages of development were retrieved from pregnant mice, placed in culture and examined for their intercellular communication properties. The stages examined and the approximate time at which they were harvested were as follows: early 2-cell (afternoon of day 2); late 2-cell (early evening of day 2); early 4-cell (late evening of day 2); late 4cell (early morning of day 3); precompaction 8-cell (noon of day 3); late compaction 8-cell (afternoon and evening of day 3); and blastocyst (afternoon of day 4). 2-Cell Stage The early 2-cell mouse embryo is ionically coupled, as indicated by the spread of injected current from one blastomere to the other (Figure 1 a). Simultaneous
injection of fluorescein and monitoring of ionic coupling revealed that fluorescein can also pass between the two cells (Figures 1 c and 1 d). Since blastomeres of early embryos frequently undergo incomplete cytokinesis or do not complete cytokinesis for extended time periods after completion of karyokinesis, it was necessary to establish whether the observed ionic coupling and dye transfer were mediated by a cytoplasmic bridge rather than by low resistance junctional channels. For this purpose, horseradish peroxidase (HRP, molecular weight 40,000), a molecular which is above the exclusion limit of junctional channels, was injected into one blastomere while fluorescein was being injected into the other. The presence of ionic coupling was also monitored using the HRP- (or fluorescein-) filled electrode, during which time current was only pulsed from the fluorescein(or HRP-) filled electrode. After the pattern of dye spread was recorded photographically, the embryos were fixed and processed for the detection of HRP activity (see Experimental Procedures). HRP activity in the blastomeres was identified by a dark brown reaction product. Such an analysis unequivocally demonstrated the
b
Figure
2. The Absence
of Intercellular
Communication
in the Late 2-Cell
Embryo
(a) Each cell of a late 2-cell embryo was impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the left blastomere (bottom trace). no voltage deflection was detected in the adjacent cell (middle trace). The horizontal calibration bar = 100 msec and the vertical calibration bar = 5 nA and 20 mV. (b) HRP was injected into the blastomere on the right. The dark reaction product indicated the presence of HRP only in the blastomere into which HRP was injected. (c. d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the start of dye injection. (c) 4 min and (d) 19 min after the start of injection. The scale bar in (b) = 45.6 firn in (a-d).
Cell 402
b
Figure
3. Intercellular
Communication
in the 4-Cell
Embryo
via Cytoplasmic
Bridges
(a) Two cells of a 4-cell embryo were each impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the upper impaled cell, a voltage deflection was recorded in that cell (lower trace) and also in the lower impaled cell (middle trace). The calibration bars are the same as for Figure 1, except that for the lower trace the bar = 10 mV. (b) HRP was injected into the blastomere on the left. The dark reaction product indicated the presence of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c, d) Fluorescein was injected into the left blastomere of the embryo and fluorescence images were recorded at various times after the start of dye injection, (c) 4 min and (d) 14 min after the start of injection. The scale bar in (b) = 45.6 gm in (a-d)
presence of a cytoplasmic bridge, since HRP moved freely from the site of injection to the other blastomere (Figure 1 b). Thus this cytoplasmic continuity can account for the observed passage of fluorescein molecules and ionic coupling between the two blastomeres (Figures lc and Id). In the late 2-cell embryo, in contrast to the early 2-cell embryo, injected HRP and fluorescein did not pass into the adjacent blastomere and no ionic coupling was detected (Figure 2). At this stage, the cytoplasmic continuity must have been abolished as a result of the completion of cytokinesis. This observation further supports the finding that no gap junctional communication was involved in the cellto-cell permeability observed in the early 2-cell embryo. 4-Cell Stage Examination of 4-cell embryos for communication using similar methods revealed two typical patterns of transfer. In one case, ionic coupling was detected between the two impaled blastomeres, and both fluorescein and HRP passed between the same impaled cell pair (Figure 3). In the other case, ionic coupling
was not detected between the two impaled cells even though injected fluorescein and HRP were passed. In this case, the injected HRP and fluorescein spread to each of two different blastomeres, neither of which were impaled (Figure 4). These observations indicate that in the first case a cytoplasmic bridge between the impaled cell pair mediated the observed ionic coupling and dye transfer, while in the second case no coupling was detected because the two impaled cells were derived from two different mitotic sister pairs. Hence there was no junction-mediated transfer of ions or molecules in the 4-cell embryo, but rather the ionic coupling and dye transfer observed were mediated via cytoplasmic bridges that exist between sister blastomeres of the second mitotic cleavage. These two patterns were observed in both early and late 4-cell embryos. S-Cell Stage Analysis of precompaction early a-cell embryos for intercellular communication yielded results identical to those observed in the 4-cell embryos in which two distinct patterns of transfer were observed. In one
Communication 403
in the Preimplantation
Mouse
Embryo
b
Figure
4.
Intercellular
(a) Two cells of a 4-tell on the right, a voltage trace). The calibration revealed the presence the blastomere on the wrn in (a-d).
Communication
in a 4-Cell
Embryo
via Cytoplasmic
Bridges
embryo were each impaled with a microelectrode. (Inset) As a current pulse (top trace) was injected into the impaled cell deflection was detected in that cell (bottom trace) but none was detected in the impaled blastomere on the left (middle bars are the same as for Figure 1. (b) HRP was injected into the impaled blastomere on the left, The dark reaction product of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c. d) Fluorescein was injected into right and fluorescence images were recorded at various times after the start of dye injection. The scale bar in (b) = 45.6
case, ionic coupling was detected together with the transfer of injected fluorescein and HRP, so that HRP and fluorescein were both present in the same two impaled cells (data not shown). In the other case, no coupling was detected between the injected cells; although the injected fluorescein and HRP were both transferred to other blastomeres, they were passed into two different ones (Figure 5). These results again indicate that the ionic coupling and dye transfer were mediated by a cytoplasmic bridge which connects sister blastomeres from the previous mitotic cleavage, and not by junctional channels. a-cell embryos which were beginning to undergo compaction were similarly examined for ionic coupling, transfer of injected fluorescein and injected HRP. Ionic coupling was observed and the injected fluorescein spread to all eight cells of the embryo (Figure 6). In contrast, the injected HRP was passed to only one other blastomere, probably the sister of the blastomere into which the HRP was being injected (Figure 6). These observations provide direct evidence that junction-mediated communication path-
ways are generated and expressed for the first time in the mouse embryo at this developmental stage. As compaction progressed to completion, ionic coupling was maintained and the injection of fluorescein into one cell of the morulae always resulted in the spread of dye to every cell of the embryo (Figure 7). Blastocyst Shortly after compaction the mouse embryo undergoes cavitation, a process whereby a cavity is generated inside the morula. The cavity enlarges in size, eventually transforming the embryo to a blastocyst with a large blastocoelic cavity. The injection of dye into a trophoblast cell of the blastocyst resulted in dye spread throughout the entire embryo, including the inner cell mass region (data not shown). Due to the three-dimensional arrangement of the blastocyst structure, however, it was difficult to discern clearly the extent of dye passage in the ICM. To determine the extent of dye spread in the ICM more precisely, ICMs were isolated from blastocysts using immunosurgery (Salter and Knowles, 1975; see
Cdl 404
b
Figure
5. Intercellular
Communication
in the Precompacted
S-Cell E:mbryo
via Cytoplasmic
Bridges
(a) Two cells of a precompaction early a-cell embryo were each impaled with a microelectrode. (inset) As a current pulse (top trace) was injected into the impaled cell on the right, a voltage deflection was detected in that blastomere (bottom trace) but none was detected in the impaled blastomere on the left (middle trace). The calibration bars are the same as for Figure 1. (b) HRP was injected into the impaled blastomere on the right, The dark reaction product revealed the presence of HRP in two cells, the cell into which HRP was injected and an adjacent blastomere. (c, d) Fluorescein was injected into the blastomere on the right and fluorescence images were recorded at various times after the start of dye injection. (c) 4 min and (d) 25 min afler the start of injection. The scale bar in (b) = 45.6 pm in (a-d).
Experimental Procedures). Ionic coupling in the ICM was monitored simultaneously with the injection of fluorescein. The cells of the ICM were ionically coupled and the injected dye filled the entire ICM (Figure 8), thus indicating that all the cells of the ICM were linked by junctional pathways. Discussion The observations made in this study reveal that the early mouse embryo does not possess junctional channels until early compaction at the 8-cell stage. At that time, the onset of ionic coupling is accompanied by the simultaneous spread of injected fluorescein into all eight cells of the embryo. The communication that was observed throughout the embryo at this stage was not due to the presence of cytoplasmic bridges, since injected HRP did not move into any cells other than the one sister blastomere of the injected cell. In previous studies of cell-to-cell communication in early embryonic cells of comparable stages in development, ionic coupling was usually observed, al-
though it was never determined whether cytoplasmic bridges were responsible for the observed communication in those studies (Ito and Hori, 1986; Ito and Loewenstein, 1969; Slack and Palmer, 1969; Tupper and Saunders, 1972). The one report of the absence of ionic coupling in the 2-cell mouse embryo (Powers and Tupper, 1978) is in agreement with the ObSeNStions on the late 2-cell embryo reported here. In addition, HRP injections in this study have shown that the observed ionic coupling and dye transfer in the early 2-cell embryo were not mediated by a junctional pathway but rather via a cytoplasmic bridge. It has been reported that in some early embryonic cells ionic coupling was not accompanied by the transfer of injected fluorescein (Slack and Palmer, 1969; Tupper and Saunders, 1972; Bennett et al., 1972). To explain these observations it was suggested that gap junctions of embryonic cells have channels of smaller pore size and that they would therefore, unlike the gap junctions of adult cells, allow the passage of ions but not of the larger fluorescein molecules. The present observations showed, however, that junction-me-
Communication 405
Figure
in the Preimplantation
6. Intercellular
Communication
Mouse
Embryo
in the Early Compacted
8-Cell
Embryo
via Junctional
Channels
(a) Two cells of an early compaction 8-cell embryo were each impaled with a microelectrode. (b) As current (top trace) was injected into the deflection was detected in that cell (bottom trace) and also in the impaled cell on the right (middle trace). The blastomere on the left, a voltage calibration bars are the same as for Figure 1. (c-e) Fluorescein was injected into the blastomere on the left and fluorescence images were recorded at various times after the start of injection. (c) 4 min. (d) 16 min and (e) 25 min after the start of injection. (f) HRP was injected into the blastomere on the right. The dark reaction product indicated the presence of HRP in two blastomeres. the blastomere into which HRP was injected and also an adjacent blastomere. The scale bar in (a) = 45.6 gm in (a) and (c-f).
diated ionic coupling in the mouse embryo was always accompanied by the passage of injected fluorescein. This observation is consistent with a recent report that the previous failure to observe dye transfer between Fundulus embryonic cells may be more apparent then real; with careful reexamination, both ionic coupling and dye transfer have been observed among these embryonic cells (Bennett et al., 1978).
It is interesting to consider the onset of junctionmediated communication in the early mouse embryo in the context of the development of the oocyte. Rat oocytes communicate with interacting follicle cells (cumulus oophorus) prior to ovulation (Albertini and Anderson, 1974; Gilula, Epstein and Beers, 1978). As meiotic maturation resumes at the time of ovulation, the oocyte becomes progressively uncoupled from
Cell 406
Figure
7. ionic Coupling
and Fluorescein
(a) Two cells of a late compaction 8-tell call on the right, a voltage deflection was bars are the same as for Figure 1. (c, d) after the start of injection. (c) 4 min and
Dye Transfer
in the Late Compaction
E-Cell Embryo
embryo were each impaled with a microelectrode. (b) As a current pulse (top trace) was injected into the detected in that cell (bottom trace) and also in the impaled cell on the left (middle trace). The calibration Fluorescein was injected into the cell on the right. Fluorescence images were recorded at various times (d) 13 min after the start of injection. The scale bar in (a) = 45.6 pm in (a), (c) and (d).
the surrounding cumulus cells (Gilula et al., 1978); the prefertilization oocyte is therefore a noncommunicating but previously communication-competent cell type. Morphological studies have shown that mouse embryos do not have gap junctions until the 8-cell stage (Ducibella et al., 1975; Magnuson et al., 1978). These observations are consistent with our findings, and suggest that the ability to form gap junctions in the rodent embryo is shut down at ovulation and not reexpressed until early compaction at the 8-cell stage. The time at which junctional communication is initially expressed, at the early compaction stage, is also approximately the time during which the first determination event occurs in the mouse embryo (Gardner and Rossant, 1976). This determination event appears to be strictly dependent upon the position of the blastomere, so that only the outside cells of the compacted morulae form trophoblasts, while the inside cells give rise to the cells of the inner cell mass in the blastocyst. On this basis, Tarkowski and Wroblewska (1967) have proposed an “inside-outside” hypothesis in which geographical location alone determines the
developmental fate of any blastomere. Since gap junctions are formed at approximately the time of initiation of trophoblast differentiation, it seems plausible that such “inside-outside” positional information might in fact be specified by an intercellular morphogenetic gradient formed via gap junctional channels. In addition, the formation of a permeability barrier by tight junctions (which also occurs at this time) would enhance differences between the environment of the cells buried on the inside from those exposed to the outside. The presence of gap junctions linking them could then provide a pathway for transferring and thus passively generating an intercellular gradient of ions, metabolites or other substances from the outside cells to the inside cells and vice versa. If intercellular communication is indeed involved in mediating the differentiation and development of the mouse embryo, it might be possible to detect changes in communication patterns between cells of the embryo which parallel its development. To investigate this possibility further, we have analyzed the cell-tocell communication properties of mouse blastocysts
Communication 407
Figure
in the Preimplantation
8. Ionic Coupling
and Fluorescein
(a) Two ceils of an ICM isolated was injected into the cell on the trace). The calibration bars are were recorded at various times (a). (c) and cd).
Mouse
Embryo
Dye Transfer
in the Isolate !d Inner
by immunosurgery from a blastocys left, a voltage deflection was deted the same as for Figure 1. (c. d) Fluor afler the start of injection. (c) 4 min
which have undergone an implantation-like process in vitro (see accompanying paper by Lo and Gilula, 1979). Experimental Procedures SWR (Jackson Laboratories; Bar Harbor, Maine) and ICR (Flow Laboratories, Inc.; Rockville. Maryland) female mice were superovulated with injections of 4 IU of pregnant mare serum gonadotropin (PMSG) and 4 IU of human chorionic gonadotropin (HCG) (Organon, Inc.; West Orange, New Jersey) and caged with NCS male mice (Laboratory Animal Research Center: Rockefeller University) overnight (Runner and Palm, 1953). Pregnant females were identified the following morning by the presence of a vaginal sperm plug; this was considered day 1 of pregnancy. The animals were sacrificed by cervical dislocation at various time intervals and the oviducts or uteri were removed. With the aid of a dissecting microscope and a syringe, the embryos were flushed from the excised oviduct or uterus and subsequently recovered in a petri dish. The zona pellucida or the harvested embryos were removed by a 1 O-l 5 min incubation in 0.5% pronase (Sigma Chemical Co.; St. Louis, Missouri) in Dulbecco’s phosphate-buffered saline (Mintz. 1962). The naked embryos were then placed in a 35 mm petri dish (Bio-Quest, BBL and Falcon Products; Cockeysville. Maryland) in Leibovitz medium (Grand Island Biological Co.: Grand Island. New York) and kept in a 37°C incubator. After the embryos attached to the dish (in -5-15 min). they were immediately impaled with microelectrodes. Embryos impaled after being in culture for l-4 hr gave results identical to those of embryos examined immediately after culturing. The isolated inner cell masses
Cell Mass
,t were each impaled with a microelectrode. (b) As a current pulse (top trace) ted in that cell (bottom trace) and also in the impaled cell on the right (I niddle ‘escein was injected into the impaled cell on the lefl and fluorescence itnages and (d) 10 min after the start of injection. The scale bar in (a) = 45.6 pm in
of the blastocysts were obtained after removing the trophoblasts immunosurgically (Salter and Knowles, 1975) and then the ICMs were allowed to attach to a 35 mm plastic petri dish for subsequent microelectrode impalements. lmmunosurgery The inner cell mass from blastocysts were isolated according to the method of Solter and Knowles (1975). Blastocysts were incubated in a 1 :lO dilution of rabbit anti-mouse fibroblast antiserum (provided by M. Friedlander, Rockefeller University) in Dulbecco’s modified Eagle’s medium with 10% heat-inactivated fetal bovine serum (DEM-HI) at 37°C for 45 min. The blastocysts were then washed in DME-HI and incubated in a 1:4 dilution of guinea pig complement (Beckman Instruments, Inc.; Fullerton, California) in DME-HI at 37°C. After 30 min. the embryos were removed and washed in DME-HI, and the outer layer of trophoblastic cells was stripped by repeated passage through a micropipette. Electrophysiology The methods and apparatus used for monitoring ionic coupling and iontophoretic injections have been described previously (Gilula et al., 1978). Briefly, glass microelectrodes filled with 3 M KCI with tip resistances of 40-60 MD were used for determining ionic coupling. For dye injections, microelectrodes with tip resistances of 15-25 MD (when filled with 3 M KCI) were filled with 5% sodium fluorescein (w/ v). lontophoresis was conducted with a 5 x lo-’ A current pulse of 0.5 set duration at 1 set intervals. The spread of fluorescence in the embryo due to the spread of injected fluorescein was visualized under
Cell 408
phase optics using a 100-W tungsten halogen lamp and a combination of EG 12 and 23 excitation filters, a KP 490 cut-off filter and a KP 530 barrier filter. All photographs of fluorescein injections were recorded on Kodak Tri-X Pan film with 3.5 min exposures. The film was developed with Acufine (Acufine. Inc.; Chicago, Illinois). For horseradish peroxidase injections, electrodes with tip resistances of 15-20 MS2 (when filled with 3 M KCI) were filled with 10 mg/ml of horseradish peroxidase (Type II, Sigma Chemical Co.: St. Louis, Missouri) in 0.2 M KCI. lontophoresis was conducted with a 510 x 1 Om9 A current pulse (depolarizing) of 0.5 set duration at 1 set intervals for a total of 1 O-l 5 min. After removal of microelectrodes. the embryos were fixed with 2.5% glutaraldehyde (w/v) (Polysciences Inc.; Warrington. Pennsylvania) in 0.1 M cacodylate buffer (pH 7.3) (Sigma) for 15 min and then incubated with 1 mg/ml of Hanker-Yates Reagent (p-phenylenediamine dihydrochloride and pyrocatechol; Polysciences) in 0.1 M Tris buffer (pH 7.6) containing 0.01% H202, After 5-15 min. a dark brown color product was visible in ceils containing injected horseradish peroxidase. and the reaction was terminated by rinsing the samples with the Tris buffer. The embryos thus processed were photographed with bright field optics using Kodak Tri-X pan film. Acknowledgments
J. (1976). Determination during embryin Mammals, Ciba Foundation Sympo-
Gilula. N. B. (1977). Gap junctions and cell communication, In International Cell Biology (1976-l 977). B. R. Brinkley and K. R. Porter, eds. (New York: The Rockefeller University Press), pp. 61-69. Gilula. N. B.. Reeves, coupling. ionic coupling
R. 0. and Steinbach, A. (1972). Metabolic and cell contacts. Nature 235, 262-265.
Gilula, N. B., Epstein, M. L. and Beers, W. H. (1978). Cell-to-cell communication and ovulation. A study of the cumulus-oocyte complex. J. Cell Biol. 78, 58-75. Ito, S. and Hori, N. (1966). Electrical characteristics cells during cleavage. J. Gen. Physiol. 49, 1019-l
of Trirurus 027.
Ito. S. and Loewenstein, W. R. (1969). Ionic communication early embryonic cells. Dev. Biol. 19, 226-243. Johnson, R. G. and Sheridan, cells in culture: ultrastructure 719. Kanno. Y. and large molecules.
egg
between
J. D. (1971). Junctions between cancer and permeability. Science 7 74, 717-
Loewenstein, W. R. (1966). Nature 272, 629-630.
Cell-to-cell
passage
of
Lo, C. W. and Gilula. N. B. (1979). Gap junctional communication the postimplantation mouse embryo. Cell 78, 41 l-422.
We would like to thank Theodore Lawrence for assistance and helpful discussions concerning the electrophysiology. Dr. Sidney Strickland and Mary Jean Sawey for help and advice in the mating of mice and the handling of mouse embryos, Asneth Kloesman and Kathy Wall for help in the preparation of figures and Madeleine Naylor for secretarial assistance. This research was supported by grants from the USPHS and The Rockefeller Foundation. N.B.G. is a recipient of a Research Career Developmental Award. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
Gardner, R. L. and Rossant. ogenesis. In Embryogenesis sium 40. 5-25.
May 2, 1979
References
Loewenstein, W. R. (1966). Permeability NY Acad. Sci. 737. 441-472.
of membrane
junctions.
in Ann.
McNutt, N. S. and Weinstein, R. S. (1970). The ultrastructure of the nexus. A correlated thin section and freeze-cleave study. J. Cell Biol. 4 7, 666-688. Magnuson. T.. Jacobson, J. B. and Stackpole. C. W. (1978). Relationship between intercellular permeability and junction organization in the preimplantation mouse embryo. Dev. Biol. 67, 214-224. Mintz, B. (1962). Experimental study of the developing mammalian egg: removal of the zona pellucida. Science 738, 594-595. Payton. B. W.. Bennett, M. V. L. and Pappas, bility and structure of junctional membranes apse. Science 766, 1641-l 643.
G. D. (1969). Permeaat an electrotonic syn-
Pitts. J. D. and Simms, J. W. (1977). Permeability of junctions between animal cells. Intercellular transfer of nucleotides but not of macromolecules. Exp. Cell Res. 704, 153-l 63.
Albertini, D. F. and Anderson, E. (1974). The appearance of structure and intercellular connections during the ontogeny of the rabbit ovarian follicle with particular reference to gap junctions. J. Cell Biol. 63, 234-250.
Potter, 0. D.. Furshpan. E. J. and Lennox, E. S. (1966). Connections between cells of the developing squid as revealed by electrophysiological methods. Prog. Nat. Acad. Sci. USA 55, 328-335.
Ashman. formation 604-605.
Powers, R. D. and Tupper. J. T. (1978). Intercellular in the early embryo. In Intercellular Communication, ed. (New York: Plenum Press). pp. 231-251.
R. F.. Kanno. Y. and Loewenstein. W. R. (1964). of a high resistance barrier in a dividing cell. Science
Bennett, M. V. L. and Trinkaus. J. P. (1970). between embryonic cells by way of extracellular ized junctions. J. Cell Biol. 44, 592-610. Bennett, M V. L., Spira. of electrotonic junctions Biol. 29. 419-435.
M. E. and Pappas, between embryonic
The 14.5,
Electrical coupling space and special-
G. D. (1972). Properties cells of Fmdulus. Dev.
Bennett, M. V. L.. Spira. M. E. and Spray, D. C. (1978). Permeability of gap junctions between embryonic cells of fundulus: a reevaluation. Dev. Biol. 65, 114-l 25. DiCaprio. R. A., French, A. S. and Saunders, E. J. (1975). connectivity in the E-cell Xenopus embryo. Correlation and morphological investigations. Biophys. J. 75, 373.
Intercellular of electrical
Ducibella, J.. Albertini. D.. Anderson, E. and Biggers. J. D. (1975). The preimplantation mammalian embryo: characterization of intercellular junctions and their appearance during development. Dev. Biol. 47, 231-250. Furshpan. E. J. and Potter, D. D. (1959). motor synapses of the crayfish. J. Physiol.
Transmission at the giant 745, 289-325.
Furshpan, E. J. and Potter, D. D. (1968). Low-resistance between cells in embryos and tissue culture. Curr. Topics 3. 95-l 27.
junctions Dev. Biol.
communication W. C. De Mello.
Runner, M. N. and Palm, J. (1953). Transplantation and survival of unfertilized ova of the mouse in relation to postovulatory age. J. Exp. Zool. 124, 303-316. Sheridan, J. D. (1968). Electrophysiological evidence for low-resistance intercellular junctions in the early chick embryo. J. Cell Biol. 37, 650-859. Sheridan, J. D. (1971). Dye movement and low-resistance junctions between reaggregated embryonic cells. Dev. Biol. 26, 627-636. Simpson, I., Rose, B. and Loewenstein. W. R. (1977). Size limit of molecules permeating the junctional membrane channels. Science 795, 294-296. Slack, C. and Palmer, J. F. (1969). The permeability of intercellular junctions in the early embryo of Xenopus laevis studied with a fluorescence tracer. Exp. Cell Res. 55. 416-419. Salter, D. and Knowles, B. (1975). lmmunosurgery cyst. Proc. Nat. Acad. Sci. USA 72, 5099-5iO2.
of mouse
blasto-
Subak-Sharpe. H.. Burk. R. R. and Pitts. J. D. (1969). Metabolic cooperation between biochemically marked mammalian cells in tissue culture. J. Cell Sci. 4. 353-367. Tarkowski, A. K. and Wroblewska, J. (1967). Development tomeres of mouse eggs isolated at the four- and eight-cell
of blasstage. J.
C$munication
Embryol.
in the Preimplantation
Exp. Morphol.
78, 155-l
Mouse
Embryo
80.
Tupper. J. T. and Saunders, J. W.. Jr. (1972). intercellular in the early Aster& embryo. Dev. Biol. 27, 546-554.
permability
Tupper. J. T.. Saunders, J. W.. Jr. and Edwards, C. (1970). The onset of electrical coupling between cells in the developing starfish embryo. J. Cell Biol. 46, 167-l 90. Wolpert, L. (1969). Positional information and the spatial cellular differentiation. J. Theoret. Biol. 25. l-47.
pattern
of
Wolpert, L. (1976). Gap junctions: channels for communication in development. In Intercellular Junctions and Synapses, J. Feldman, N. 8. Gilula and J. D. Pitt% eds. (London: Chapman and Hall), pp. 8394.
