| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Departments of Physiology, Obstetrics and Gynaecology, and Paediatrics, The University of Western Ontario, London, Ontario, Canada N6A 5C1
b Child Health Research Institute, London, Ontario, Canada N6C 2V5
c Department of Biology, The University of York, York, United Kingdom
d Institute of Anatomy, University Hospital of Essen, Essen, Germany
e Institute of Genetics, The University of Bonn, Bonn, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
9, and connexin 57 or 10) are also transcribed in preimplantation mouse embryos. Furthermore, we show that multiple connexins are expressed in rat preimplantation embryos, indicating that multiplicity of connexin expression may be a common feature of early mammalian embryogenesis. We could detect no up-regulation of any of 3 coexpressed connexins examined in mouse embryos lacking connexin 43. Impaired intercellular coupling caused either by the loss of connexin 43 or by treatment of cultured embryos with the gap junctional coupling blocker 18-glycyrrhetinic acid (AGA) had no discernable effect on either apoptosis or glucose utilization, parameters known to be affected by gap junctional coupling in other contexts. These results, taken together with the reported inability of AGA to perturb blastocyst formation, imply that gap junctional coupling is not essential during this developmental period. We propose that connexin expression and the assembly of multiple types of gap junction channels in preimplantation embryos facilitates the diversification of communication pathways that will appear during postimplantation development. New evidence of this diversification is presented using rat blastocyst outgrowths.
developmental biology, early development, embryo, implantation, trophoblast
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The preimplantation embryo provides a good example of such compensation. One particular gap junction protein, connexin 43 (Cx43 or 1 connexin), was identified in initial studies as being incorporated into the first gap junctions that form during mouse preimplantation development [4]. After implantation, Cx43 continues to contribute to gap junctions throughout embryonic and fetal development, becoming expressed in a number of organs including the heart and ovary [1, 2]. Mouse conceptuses lacking Cx43 develop to term despite severely reduced gap junctional coupling in the preimplantation stages [5] but then die soon after birth because of a heart abnormality [6]. Thus, despite its expression in the preimplantation embryo and widespread distribution in the fetus, the absence of Cx43 does not prevent implantation or disrupt prenatal development. Recent studies, however, have begun to reveal defects in the postnatal development of reproductive organs in Cx43-deficient mice [7–9].
We have previously suggested that the capacity of preimplantation embryos to develop in the absence of Cx43 could be explained by the presence of any of at least five additional connexins, the mRNAs of which have been detected in blastocysts: Cx30.3 (ß4 connexin), Cx31 (ß3 connexin), Cx31.1 (ß5 connexin), Cx40 (5 connexin), and Cx45 (7 connexin) [10]. The intercellular channels remaining in the absence of Cx43 were shown to be permeable to dichlorofluorescein but not to carboxyfluorescein, properties that resemble those of Cx45 channels [5]. However, the permeability properties of the other coexpressed connexins are still not known; hence, it is possible that Cx45 is not the only connexin that functions along with Cx43 during preimplantation development.
In the present study, we set out to clarify the involvement of additional connexins and the role of gap junctional coupling, if any, in preimplantation development. We provide evidence that three additional connexins, Cx30 or ß6 connexin [11], Cx36 or 9 connexin [12, 13], and Cx57 or 10 connexin [14] are expressed in preimplantation mouse embryos. Furthermore, we show that multiple connexins are expressed in rat preimplantation embryos, indicating that multiplicity of connexin expression may be a common feature of early mammalian embryogenesis. We could detect no up-regulation of any coexpressed connexin when Cx43 was absent from mouse embryos. Disruption of intercellular coupling had no discernable effect on either apoptosis or glucose metabolism. We discuss published evidence from pharmacologic blockade of gap junctional coupling that preimplantation development can proceed normally in its absence. We propose that connexin expression and the assembly of multiple types of gap junction channels in preimplantation embryos facilitate the diversification of communication pathways that will appear during postimplantation development, and we present new evidence of this diversification.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Female mice (CF1 strain from Charles River Canada Ltd., St. Constant, QC, Canada) were hormonally superovulated and mated with CB6F1/J strain males (The Jackson Laboratory, Bar Harbor, ME), and preimplantation embryos for reverse transcription-polymerase chain reaction (RT-PCR) analysis were flushed from the reproductive tracts as previously described [4]. Embryo collections were timed as follows (hours post-hCG): zygotes, 24 h; two-cell, 48 h; four-cell, 60 h; compacting eight-cell, 74 h; morula, 83 h; blastocyst, 94 h. To obtain Cx43-deficient embryos for connexin immunostaining, Gja1+/Gja1- males (bred in our colony) and nonsuperovulated females (CD1 strain from Harlan Sprague-Dawley, Indianapolis, IN) were mated to generate embryo populations of which approximately 25% would be expected to be homozygous for the null mutation. Embryos were individually genotyped by PCR after immunostaining, as previously described [5]. Rat blastocysts were obtained from uteri of mated females by flushing at 4.5 and 5.5 days postcoitus (dpc) (day of vaginal plug = 0 dpc). To obtain outgrowths, the blastocysts were cultured in 200 µl of M2 medium [15] in eight-well culture chambers (Nunc Lab-Tek chamber Slide System; Nunc, Wiesbaden, Germany) at 37°C in 5% CO2 in air for 4 days. Postimplantation rat embryos were obtained by dissecting the implantation chambers at 7.5–9.5 dpc; the specimens were quickly frozen in liquid nitrogen for later cryosectioning and immunostaining. We used Sprague-Dawley rats from Zentralinstitut für Versuchtierkunde (Hannover, Germany).
Reverse Transcription-Polymerase Chain Reaction
Total RNA from mouse preimplantation embryos was isolated by centrifugation through CsCl according to the method of Valdimarsson et al. [16]. Pools of 100 embryos were used for RNA isolation, except for blastocysts, for which only 50 were used. Each RNA preparation was tested for the presence of any contaminating genomic DNA by PCR using primers that flank an intron in the ß-actin gene [4, 16]. The RNA was reverse transcribed using oligo(dT) primer, SuperScript reverse transcriptase, reaction buffer, and dNTPs supplied by Life Technologies Inc. (Burlington, ON, Canada). PCR was performed using primers specific for Cx30, Cx36, and Cx57 as summarized in Table 1. The identity of amplicons was confirmed by restriction enzyme digestion followed by direct sequencing. The PCR amplification products were visualized by electrophoresis through 3% agarose gels containing ethidium bromide, and images were captured using an Imagemaster VDS system from Hoefer Pharmacia Biotech Inc. (San Francisco, CA).
|
Rat blastocysts collected from 4 dpc uteri were immersed in 8 µl of diethyl pyrocarbonate-treated H2O. Each sample was covered with mineral oil and heated to 99°C for 1 min in a DNA thermal cycler (Biometra UNO-Thermoblock; Biometra, Göttingen, Germany) to release the total RNA and denature the proteins. The sample was cooled down to 4°C, and the RT reaction was performed using the Single-Strand cDNA Synthesis Kit (Amersham-Pharmacia, Freiburg, Germany) according to the manufacturer's instructions. The cDNA was amplified using primers for ß-actin and nested primers specific for exons 1 and 2 of the genes encoding Cx26, Cx31, and Cx43 (Table 1). The Cx30 primers used for the rat experiments did not span the intron, but the absence of contaminating genomic DNA in the samples was verified by the ß-actin PCR. A 50-µl aliquot of the reaction mixture in Taq reaction buffer contained 2 mM MgCl2, 0.4 mM dNTPs, 25 pmol of each ß-actin primer, 25 pmol of each connexin outer primer, and 2.5 units Taq polymerase (Gibco BRL, Karlsruhe, Germany). The PCR was carried out for 30 cycles of 94°C for 45 sec, 54°C for 45 sec, and 72°C for 1 min. A 30-µl aliquot of the PCR reaction was tested for the ß-actin signal by agarose gel electrophoresis. For the second PCR, 5 µl of the first-round reaction was added to 45 µl of the PCR reaction mixture consisting of Taq reaction buffer, 2 mM MgCl2, 0.4 mM dNTPs, 25 pmol of each inner primer, 2.5 units Taq polymerase, and 5 µM [32P]dCTP (3000 Ci/mmol). The program parameters were identical to those for the first-round PCR. The PCR products were separated by agarose gel electrophoresis and blotted on a nylon membrane (Amersham-Pharmacia). The nylon membrane was hybridized with connexin-specific cDNA, sealed, and exposed to Kodak XAR x-ray film (Kodak, Stuttgart, Germany) with an intensifying screen at -70°C for 1–6 h. Hybridization and labeling of the specific probes were performed as described previously [17]. The following probes were used: rat Cx26 cDNA [18], mouse Cx30 cDNA [11], mouse Cx31 cDNA [19], and rat Cx43 cDNA [20]. The blots were rehybridized with ß-actin cDNA [21] using the same conditions.
Immunolocalization
Before immunostaining of preimplantation embryos, zonae pellucidae were removed using acid Tyrode solution [15] (this and all subsequent steps were carried out at room temperature). Fixation in 1% paraformaldehyde for 20 min was followed by rinsing in PBS. The embryos were then transferred onto poly-L-lysine-coated coverslips [15] that were centrifuged for 15 min at 600 x g in chambers containing PBS to ensure that the embryos were well attached. The embryos were permeabilized with 0.25% Triton X-100 in PBS for 10 min, rinsed again in PBS, and treated with 2.8 mg/ml ammonium chloride in PBS for 10 min. Embryos were then treated with primary antibody solution for at least 1 h (all primary antibodies were rabbit polyclonals; details are summarized in Table 2). Likewise, the secondary antibody (1:50 fluorescein-conjugated goat anti-rabbit immunoglobulin G; ICN Pharmaceuticals Inc., Costa Mesa, CA) was applied for 1 h, followed by 3 rinses in PBS with 0.1% Tween 20. The specimens were viewed using a BioRad MRC 600 (BioRad Canada Ltd., Mississauga, ON, Canada) or Leica TCS NT (Leica Microsystems Ltd., Milton Keynes, UK) confocal microscope.
|
Identification of Apoptotic Cells
Apoptosis was examined in embryos resulting from a heterozygote cross (from matings of Gja1+/Gja1- males and females) and in embryos treated with the gap junctional coupling blocker 18-glycyrrhetinic acid (AGA; Sigma-Aldrich Ltd., Oakville, ON, Canada). In the latter case, eight-cell embryos were cultured to the blastocyst stage as groups of 10 in 10-µl drops of 65 µM AGA in KSOM (potassium-augmented simplex optimization medium) [32], a concentration sufficient to completely block intercellular dye transfer between mouse blastomeres [33]. Controls were cultured in KSOM without AGA but containing 0.5% dimethylsulfoxide (DMSO; Sigma), the solvent used to dissolve the AGA. To identify apoptotic cells, zona-free embryos were fixed in 1% paraformaldehyde for 20 min and attached to poly-L-lysine-coated coverslips as described previously. They were rinsed twice in PBS before being incubated in TUNEL reaction mixture (Boehringer Mannheim Canada, Laval, QC, Canada) for 60 min at 37°C in the dark. The embryos were rinsed 3 more times with PBS and treated with 50 µg/ml RNase A (Life Technologies) for 60 min at room temperature in the dark. After another PBS rinse, the embryo nuclei were stained with 10 µg/ml propidium iodide for 60 min at room temperature in the dark. Finally, they were rinsed 3 more times with PBS and sealed with a coverslip. Positive controls were prepared in which the embryos were pretreated with 100 µg/ml DNase I (Life Technologies) before TUNEL labeling to create single-stranded DNA breaks. For negative controls, embryos were incubated with the TUNEL staining reagents in the absence of terminal deoxynucleotidyl transferase. The TUNEL-stained embryos were viewed using a BioRad MRC 600 confocal microscope.
Measurement of Substrate Utilization
Day 4 mouse blastocysts were rinsed three times in modified KSOM medium containing 1 mM glucose, 0.33 mM pyruvate, and 2 mM D,L-lactate (Sigma) and cultured individually in microdrops (35–40 nl) of the same medium under mineral oil for 3 h. On completion of the culture period, the blastocysts were removed from the microdrops, and the medium was stored at -80°C for later analysis of substrate content. Blastocysts derived from matings of Gja1+/Gja1- males and females were rinsed in PBS containing 0.3% polyvinylpyrrolidone (Sigma), and genotyped by PCR as previously described [5]. In other experiments, wild-type hatched blastocysts of the strain CBA/Ca x C57BL/6 (bred in house at the University of York) were treated for 1 h with 65 µM AGA and then transferred to microdrops containing the same agent for 2 h. After the treatment, the blastocysts were removed, and the medium was frozen for later substrate content analysis. As in the apoptosis experiments, controls were treated with KSOM containing 0.5% DMSO.
After thawing, each sample of medium was assayed for pyruvate, glucose, and lactate concentrations using an ultramicrofluorescence enzymatic technique based on the formation or oxidation of NAD(P)H [34]. A 1-nl aliquot of each sample was added to 10 nl of reaction mixture, and the fluorescence of NADH or NADPH was recorded at 340 nm on a Leica Fluovert quantitative fluorescence microscope. Substrate concentrations were determined by reference to a series of standards.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous work had shown that at least six connexin genes (those encoding Cx30.3, Cx31, Cx31.1, Cx40, Cx43, and Cx45) are transcribed in preimplantation mouse embryos [10]. Since that work, several more members of the mouse connexin family, including Cx30 [11], Cx36 [12], and Cx57 [14], have been cloned and shown to give rise to functional intercellular channels. These three connexin genes are also transcribed during mouse preimplantation development (Fig. 1). The RT-PCR data indicated that all three mRNAs are present as early as the 4-cell stage and continue to be expressed into the blastocyst stage. Thus, as many as 9 connexins could potentially contribute to intercellular coupling in preimplantation mouse embryos.
|
A similar situation exists in the rat, although we did not test for as many connexins in the rat. Using RT-PCR followed by Southern blotting to verify the authenticity of the amplicons, mRNAs encoding Cx26, Cx30, Cx31, and Cx43 were identified in blastocysts (Fig. 2). Immunostaining and confocal microscopy allowed the detection and localization of all of these connexins plus Cx45 in rat preimplantation embryos (Fig. 3). Whereas Cx26, Cx31, Cx43, and Cx45 were detected in structures resembling gap junction plaques, this was not the case for Cx30. Thus, as in the mouse ([10] and Fig. 4, right column), not all connexins for which mRNAs are detectable in rat preimplantation embryos are assembled into gap junctions.
|
|
|
Previous reports have indicated that Cx26 mRNA and protein are absent from preimplantation mouse embryos [10, 35]. Our finding of Cx26 mRNA and protein in rat embryos prompted us to reexamine the situation in the mouse using the same antibody (Table 2). No immunoreactivity was seen in mouse preimplantation embryos stained with this Cx26 antibody (data not shown). Thus, the expression of Cx26 in preimplantation embryos is one feature that differs between mouse and rat.
Given the presence of multiple connexins in preimplantation embryos, we looked for up-regulation of other connexins in mouse embryos lacking Cx43. As illustrated in Figure 4, we could find no obvious changes in the expression of Cx31, Cx31.1, or Cx40 (minor quantitative changes in the expression of individual connexins would go undetected in this analysis but would not likely be of any physiologic significance). These findings are in agreement with our earlier report [5] that Cx45 expression does not change in Cx43 null mutant embryos. In the case of Cx40, we have detected only cytoplasmic immunoreactivity in preimplantation embryos, suggesting that this connexin, although it might be present, does not contribute to gap junctions [10]. The present result demonstrates that the absence of Cx43 does not cause this situation to change. We conclude that connexin genes are under independent regulatory controls in this system and that compensatory expression of other connexins is unlikely to be involved in the survival of embryos lacking Cx43 or any other single connexin.
Cellular Functions in Embryos with Impaired Gap Junctional Coupling
Knock-out mouse models have been created for several of the connexins known to be expressed in preimplantation development, including Cx43, Cx40, Cx31, and Cx45 [6, 36–40]. In all cases, the expected proportion of null mutant embryos was detected after implantation, implying a lack of significant preimplantation defects. Furthermore, the developmental progress of Cx43 null mutant embryos was examined directly in culture and was found to be normal, despite both quantitative and qualitative changes in intercellular coupling [5]. However, it remained possible that cellular functions are altered in coupling-deficient embryos, despite their normal rate of development. One such cellular function is apoptotic cell death: gap junctional coupling has been directly implicated both in the transmission of cell death signals, amplifying the effects of metabolic or oxidative stress [41], and in protecting cells from apoptotic cell death [42]. We used the TUNEL staining reaction to analyze apoptosis in individual morulae and blastocysts resulting from a heterozygote cross and then performed PCR to determine the genotype of each embryo. Although only 30% of embryos (for a total of 25) were successfully genotyped after TUNEL staining, there was no apparent difference in apoptosis frequency between genotypes (Table 3). Representative TUNEL-stained embryos are shown in Figure 5. As a further test of the possible involvement of gap junctions in apoptosis, we cultured eight-cell embryos to the blastocyst stage in the presence of 65 µM AGA, a blocker of gap junctional coupling that completely inhibits dye coupling in preimplantation embryos [33]. As summarized in Table 4, the apoptotic index (percentage of apoptotic cells) was not significantly altered by this treatment. Thus the incidence of programmed cell death is not affected in embryos with impaired gap junctional coupling.
|
|
|
Another function that could be affected by a change in gap junctional coupling is glucose utilization. Glucose consumption by the embryo increases during cavitation [34, 43]. Glucose passes freely through gap junctions, and this pathway has been identified as a potential limiting factor in glucose use by cultured cells [44]. We tested the hypothesis that the severely reduced gap junctional coupling in Cx43-deficient embryos would impair glucose use. The data in Figure 6 indicate that this hypothesis is incorrect: pyruvate and glucose consumption and lactate production were not affected by the absence of Cx43. Furthermore, when blastocysts were treated with 65 µM AGA, there was likewise no effect on substrate utilization (Fig. 7). Thus, exchange of metabolites between cells via gap junction channels is unlikely to be a rate-limiting factor in preimplantation embryo metabolism.
|
|
Connexin Expression Domains in Blastocyst Outgrowths
Previously, immunofluorescence analysis of sectioned implantation sites in the mouse and rat has revealed that expression of Cx31 and Cx43 segregate into restricted domains after implantation: Cx43 is localized in the embryo proper and visceral endoderm, whereas Cx31 is restricted to the extraembryonic ectoderm and ectoplacental cone [45, 46]. To investigate this phenomenon further, the expression of several connexins was assessed in rat blastocyst outgrowths, which mimic some of the events of attachment and implantation in the uterus. As had been found in vivo, Cx31 and Cx43 segregated into distinct domains in the outgrowths, with Cx31 appearing in the migrating trophoblast cells and Cx43 remaining in the inner cell mass (ICM) (Fig. 8). We also studied the distribution of Cx26 and Cx45. Although the immunofluorescence signals for these connexins were not as distinct, Cx26 appeared to be expressed preferentially in cells at the border between the ICM and outgrowing trophoblast cells (Fig. 9, upper). These cells could represent the primitive endoderm (hypoblast) or a distinct part of the epiblast. In contrast, Cx45 was predominantly detected in the ICM cells (Fig. 9, lower). After implantation, with ongoing development of the placenta, Cx26 is first detected in the chorionic plate region after fusion of the allantois with the chorion [47]. Immunolabeling for Cx45 became very weak after implantation, with only the embryonic ectoderm showing some minimal immunoreactivity (data not shown).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The results presented here, together with previous work [10], make it clear that the preimplantation rodent embryo is promiscuous for connexin gene expression: more than half of the 16 cloned members of this gene family are transcribed before implantation (at least in the mouse), and several of the resulting connexins are incorporated into membrane plaques, in both mouse and rat. This occurs in a developing system with at first just one and later only two distinct cell types (the trophectoderm and ICM of the early blastocyst). Furthermore, there does not appear to be any segregation of connexins between these two cell types. Gap junction channels composed of different connexins have unique properties in terms of molecular permeability, unitary conductance, and regulation of channel gating [1, 2, 48–51]. It is difficult to envision the developmental or physiologic roles that could be served by gap junctional coupling in the preimplantation embryo that would require such diversity of channel types. Indeed, when we examined two functions that might depend on gap junctional coupling, apoptosis and glucose use, we could find no differences between wild-type and Cx43 null mutant embryos. These findings raise the possibility that gap junctional coupling is not an essential aspect of preimplantation development, despite the fact that gap junction assembly is a developmentally regulated event [4].
Recently, another line of evidence was presented that suggests the same conclusion. Whereas the Cx43 knock-out line has made it possible to study the development of preimplantation embryos with severely reduced coupling, those embryos do retain some functional gap junctions with unique properties [5]. Pharmacologic inhibitors of gap junctional coupling, on the other hand, can potentially abolish coupling through whatever channels are present. This approach was taken by Vance and Wiley [33], who treated mouse preimplantation embryos with AGA (65 µM) beginning in the four-cell stage (before the onset of gap junction assembly) and continuing into the blastocyst stage. The treatment had no effect on blastocyst development, total cell number, or the ratio of cells in the trophectoderm and ICM. Correspondingly, we found that AGA treatment does not affect the use of energy substrates, the distribution of which within the embryo might be expected to depend on gap junctions. Collectively, these data strongly suggest that gap junctional coupling is dispensable for preimplantation development in the mouse. They also contradict results from earlier studies in which connexin antibodies or antisense RNA injections were used to disrupt gap junctional coupling, resulting in developmental impairment [52–54]. These discrepancies remain unresolved.
Data reported here and previously concerning rat embryos indicate that expression of multiple connexins in preimplantation development is a common feature of rodents. Although we have not examined the expression of as many connexins in rat as in mouse embryos, most of the connexins detected so far in rats, including Cx30, Cx31, Cx43, and Cx45, have also been detected in mouse. The one exception is Cx26. Whereas Cx26 mRNA and protein could not be detected in mouse preimplantation embryos, both were present in rat blastocysts. Similarly, Cx26 is not expressed in the decidua of the mouse [55] but is present in the decidua of the rat [56]. Whether these differences between mice and rats have any physiologic significance remains to be determined. On the other hand, Cx26 is expressed in the developing placenta of both species and, at least in the mouse, is required for placental function. Cx26 channels are induced concomitantly with placental labyrinth formation and are expressed between the two syncytial trophoblast layers [47, 57]. Mice lacking Cx26 die in utero around 10 dpc. because of impaired transplacental transport of glucose and perhaps other nutrients [58]. The outgrowth model does not allow precise identification of the Cx26-expressing cells, but it is likely that these cells belong to the epiblast which in part gives rise to the placenta.
Although Cx45 is expressed throughout the early preimplantation embryo, its distribution in blastocyst outgrowths suggests that it later becomes restricted to the ICM and is thus destined to be expressed mainly in the embryo proper. Indeed, recent analyses of mice heterozygous or homozygous for a targeted insertion of the LacZ coding region into the gene (Gja7) encoding Cx45 have revealed that the Gja7 promoter is strongly and nearly ubiquitously expressed in the embryo proper by Day 6.5 of gestation. In addition, expression is found in the mesodermal derivatives of the extraembryonic membranes such as the mesodermal layers of the visceral yolk sac and the allantoic mesenchyme. Mice lacking Cx45 exhibit cardiac dysmorphogenesis [39] and failure of vasculogenesis within the embryo, the visceral yolk sac, and the placenta [40]. As a consequence, these embryos die between gestational Days 9.5 and 10.5.
Why Do Preimplantation Embryos Express Multiple Connexins?
If gap junctional coupling between blastomeres can be dispensed with, why do preimplantation embryos express multiple connexins and assemble them into gap junctions? One possible explanation, supported by analyses of connexin expression domains in postimplantation conceptuses and in blastocyst outgrowths in vitro, is that expression of multiple connexins before implantation is a prerequisite for rapid segregation of gap junction channel types in the embryonic and extraembryonic regions of the conceptus that arise during and after implantation. For example, a rapid differentiation program establishes the first functional embryonic organ, the placenta. Directly after implantation, Cx31 and Cx43, which are abundantly expressed in preimplantation rodent embryos, segregate into the two major communication compartments, embryonic and extraembryonic, described by Kalimi and Lo [59]. Since Cx31 channels are apparently unable to form functional channels with Cx43 [17], this distribution pattern may establish the communication boundary between these two compartments. The outgrowth model indicates that this segregation of connexin expression domains is not under maternal influence but represents an endogenous program associated with the development of the ectoplacental cone and its derivatives. Later on in placental development, Cx31 expression is maintained in the spongiotrophoblast [33, 47]. Mice lacking Cx31 demonstrate decreased embryonic survival because of transient placental dysmorphogenesis [38]. The placentae of homozygous mice are smaller and have a reduced labyrinth and almost no spongiotrophoblast cells, which are replaced by trophoblast giant cells.
A further example of a preimplantation event that is seemingly geared to postimplantation requirements is provided by the sharp rise in glucose consumption that occurs with blastocyst formation in all mammalian species examined [60]. Although the biochemical and molecular mechanisms underlying this response have been examined in some detail, its functional significance has remained obscure. One clue as to the role of this early activation of glucose consumption and metabolism is provided by ultrastructural observations in the implanting rodent blastocyst, in which the decidual zone is found to be devoid of capillaries [61, 62] and the embryo is surrounded by a potentially anoxic zone. Leese [63] postulated that the rise in glucose consumption, and associated ability to form ATP via anaerobic glycolysis, is essential for the embryo to survive during the peri-implantation period until the maternal circulation is established. In other words, the embryo switches from a metabolism dependent on aerobic respiration to one with the potential to cope with anoxia before that need actually arises.
In conclusion, we hypothesize that the expression of multiple connexins in the preimplantation embryo allows the implanting conceptus to undergo rapid diversification of cell types required for establishment of both embryonic (i.e., fetal) and extraembryonic (i.e., yolk sac and placenta) tissues. The lack of any requirement for gap junctional communication in preimplantation development, suggested by the effects of both targeted disruption of connexin genes and pharmacologic blockade of gap junctional coupling, may simply reflect the small size of the rodent blastocyst and its capacity to move molecules into the blastocoel by transepithelial transport [64]. For example, Hewitson and Leese [65] found that uptake and metabolism of glucose differs between isolated mouse ICM and trophectoderm cells: ICM consumed 89 fmol of glucose per cell per hour and formed 193 fmol of lactate per cell per hour, whereas the figures for trophectoderm were 31 and 34 fmol per cell per hour, respectively. Since one molecule of glucose gives rise to two molecules of lactate, these figures were taken to mean that 100% of the glucose consumed by the ICM, but only 55% of that consumed by the trophectoderm, is accounted for by lactate production. Hewitson and Leese [65] therefore hypothesized that the trophectoderm acts as a transporting epithelium, transferring sufficient glucose into the blastocoel to meet the metabolic needs of the ICM. If this view is correct and applies to other metabolites, then there may be no need for gap junctions to mediate direct transfer of small molecules from cell to cell within the preimplantation embryo. We predict, therefore, that targeted disruption of the remaining connexin genes expressed during the preimplantation period, either alone or in combination, will similarly not affect blastocyst development, but will nonetheless cause defects in organogenesis. It is possible that the distinct metabolic requirements of postimplantation conceptuses have created a unique developmental situation in which other genes expressed before implantation will likewise be found to be indispensable only after implantation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada (6863-98 to G.M.K.), the U.K. Medical Research Council (G9813883 to H.J.L.), and the Deutsche Forschungsgemeinschaft (Wi 774/10-3 to E.W.). 
2 Correspondence: Gerald M. Kidder, Department of Physiology, The University of Western Ontario, Dental Sciences Building, dock 15, London, ON, Canada N6A 5C1. FAX: 519 661 3827; gerald.kidder{at}fmd.uwo.ca
Accepted: December 4, 2001.
Received: June 6, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Worsdorfer, S. Maxeiner, C. Markopoulos, G. Kirfel, V. Wulf, T. Auth, S. Urschel, J. von Maltzahn, and K. Willecke Connexin Expression and Functional Analysis of Gap Junctional Communication in Mouse Embryonic Stem Cells Stem Cells, February 1, 2008; 26(2): 431 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Huettner, A. Lu, Y. Qu, Y. Wu, M. Kim, and J. W. McDonald Gap Junctions and Connexon Hemichannels in Human Embryonic Stem Cells Stem Cells, July 1, 2006; 24(7): 1654 - 1667. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. H. Jones, T. Y. Li, E. Arystarkhova, K. J. Barr, R. K. Wetzel, J. Peng, K. Markham, K. J. Sweadner, G.-H. Fong, and G. M. Kidder Na,K-ATPase from Mice Lacking the {gamma} Subunit (FXYD2) Exhibits Altered Na+ Affinity and Decreased Thermal Stability J. Biol. Chem., May 13, 2005; 280(19): 19003 - 19011. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. D Houghton Role of gap junctions during early embryo development Reproduction, February 1, 2005; 129(2): 129 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Prindull and D. Zipori Environmental guidance of normal and tumor cell plasticity: epithelial mesenchymal transitions as a paradigm Blood, April 15, 2004; 103(8): 2892 - 2899. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||
