Expression of the gap junction proteins connexin31 and connexin43 correlates with communication compartments in extraembryonic tissues and in the gastrulating mouse embryo, respectively

Abstract

We have characterized the pattern of connexin expression in embryonic and extraembryonic tissues during early mouse development. In the preimplantation blastocyst, at 3.5 days post coitum (dpc), immunofluorescent signals specific for connexin31 and connexin43 proteins were present in both the inner cell mass and the trophectoderm, as shown by confocal laser scan microscopy. Immediately after implantation at 6.5 dpc, however, we find complete compartmentation of these two connexins: connexin31 mRNA and protein are expressed exclusively in cells derived from the trophectoderm lineage, whereas connexin43 mRNA and protein are detected in cells derived from the inner cell mass. This expression pattern of connexin31 and connexin43 is maintained at 7.5 dpc when the axial polarity of the mouse embryo is established. It correlates with the communication compartments in extraembryonic tissues and the gastrulating mouse embryo, respectively. The communication boundary between those compartments may be due to incompatibility of connexin31 and connexin43 hemichannels, which do not communicate with each other in cell culture.

Full text

INTRODUCTION
Gap junctions are tightly packed aggregates of transmembrane
channels which permit ions and small molecules to pass from
the cytoplasm of one cell to that of neighbouring cells. In many
embryonic tissues, gap-junction-mediated intercellular com-
munication was found to be organized into communication
compartments (Kalimi and Lo, 1989; Bagnall et al., 1992; for
review see Warner, 1992). Communication compartments are
groups of cells that are well coupled to each other but exhibit
restricted or no coupling at their compartment border (Lo and
Gilula, 1979a). The widespread occurrence, yet spatial restric-
tion, of gap-junction-mediated intercellular communication led
to the hypothesis that gap junctions may be involved in pattern
formation, i.e. the differentiation and segregation of embryonic
cells. It had been suggested that gap junctions mediate pos-
itional information by regulating the transfer of informational
molecules (Wolpert, 1978). In the Drosophila wing imaginal
disk, dye restriction borders coincide with those of lineage
compartments. It was proposed that communication compart-
ments may establish or maintain developmental compartments
in insects (Lo, 1988). Unfortunately, no genes coding for gap
junction proteins have been cloned from the Drosophila
genome.
In mammals, gap junctions are composed of a family of
channel-forming proteins, the connexins. Currently, 12
connexin (Cx) genes have been described in the murine
genome (Haefliger et al., 1992; White et al., 1992; Willecke et
al., 1991). Intercellular communication in the mouse embryo
is first detected at the 8-cell stage (Lo and Gilula, 1979a; Lee
et al., 1987). It was shown that intercellular communication
between blastomeres is necessary to maintain compaction (Lee
et al., 1987). Several reports have suggested that Cx43 is a
possible constituent of these first gap junctions channels
formed during mouse embryogenesis. Cx43 mRNA and
protein have been detected in the 4-cell mouse embryo and
steadily accumulate thereafter (Nishi et al., 1991; Valdimars-
son et al., 1991). Recently, it was shown that synthesis of Cx43
protein starts one cell cycle before it is detected in gap-
junction-like plaques. This delay of approximately 24 hours is
due to processing of Cx43 protein to plasma membranes
(DeSousa et al., 1993).
In the early mouse embryo, communication compartments
are first detected after implantation. In cultured blastocysts,
communication becomes gradually restricted during ongoing
development (Lo and Gilula, 1979b). Cells of the trophecto-
derm uncouple from each other as well as from cells of the
inner cell mass. In the 7.5 dpc gastrulating mouse embryo,
nine Lucifer Yellow-delineated communication compart-
ments were defined (Kalimi and Lo, 1988, 1989). They
191
Journal of Cell Science 109, 191-197 (1996)
Printed in Great Britain © The Company of Biologists Limited 1996
JCS8963
We have characterized the pattern of connexin expression
in embryonic and extraembryonic tissues during early
mouse development. In the preimplantation blastocyst, at
3.5 days post coitum (dpc), immunofluorescent signals
specific for connexin31 and connexin43 proteins were
present in both the inner cell mass and the trophectoderm,
as shown by confocal laser scan microscopy. Immediately
after implantation at 6.5 dpc, however, we find complete
compartmentation of these two connexins: connexin31
mRNA and protein are expressed exclusively in cells
derived from the trophectoderm lineage, whereas
connexin43 mRNA and protein are detected in cells derived
from the inner cell mass. This expression pattern of
connexin31 and connexin43 is maintained at 7.5 dpc when
the axial polarity of the mouse embryo is established. It cor-
relates with the communication compartments in extraem-
bryonic tissues and the gastrulating mouse embryo, respec-
tively. The communication boundary between those
compartments may be due to incompatibility of connexin31
and connexin43 hemichannels, which do not communicate
with each other in cell culture.
Key words: Inner cell mass, Incompatible connexin, Pattern
formation, Trophectoderm
SUMMARY
Expression of the gap junction proteins connexin31 and connexin43
correlates with communication compartments in extraembryonic tissues and
in the gastrulating mouse embryo, respectively
Edgar Dahl1,3, Elke Winterhager2, Bernhard Reuß2, Otto Traub1, Anette Butterweck1and Klaus Willecke1,*
1Institut für Genetik, Abt. Molekulargenetik, Universität Bonn, Römerstrasse 164, 53117 Bonn, Germany
2Institut für Anatomie, Universität-GHS-Essen, 45122 Essen, Germany
3Institut für Säugetiergenetik, GSF Forschungszentrum für Umwelt und Gesundheit, 85758 Oberschleißheim, Germany
*Author for correspondence

192
included the embryonic and extraembryonic germ layers of
ectoderm, mesoderm and endoderm. Impalements with
microelectrodes showed that low level of intercellular com-
munication exists across most of these communication com-
partments defined by dye transfer. However, a complete
absence of electrical coupling was found between cells
derived from the trophectoderm (extraembryonic cells) and
the inner cell mass (embryonic cells) (Kalimi and Lo, 1989).
At this developmental stage, Cx43 mRNA (Ruangvoravat and
Lo, 1992) and protein (Yancey et al., 1992) were detected in
all three germ layers of the embryo proper. No further
connexin has been described in the gastrulating mouse
embryo so far.
In the present study, we have investigated whether the
patterns of connexin expression in early mouse development
correlate with communication compartments, described above.
Recently, we have shown that murine connexins, when
expressed in human HeLa cells defective in gap junctional
communication, can be divided into compatible and incom-
patible connexins with respect to formation of heterotypic gap
junctions (Elfgang et al., 1995). We and others (Bruzzone et
al., 1993) have hypothesized that expression of incompatible
connexins could lead to establishment of communication com-
partments found in early mouse development. Thus, we have
analyzed expression pattern of seven connexin genes (Cx26,
Cx31, Cx32, Cx37, Cx40, Cx43 and Cx45) by in situ hybridiz-
ation and indirect immunofluorescence at different develop-
mental stages from blastocyst to gastrulating mouse embryo.
Our results show that Cx31 is expressed during preimplanta-
tion mouse development and that compartmentalized
expression of Cx31 and Cx43 may contribute to the formation
of communication compartments in the mouse conceptus at
gastrulation.
MATERIALS AND METHODS
Mating of mice and reproductive stages investigated
NMRI female mice, 6-8 weeks of age, were mated with NMRI males
overnight. The morning on which the vaginal plug was found was
considered as day 0.5 of pregnancy (Rugh, 1990). Blastocysts at 3.5
dpc were collected from mouse uteri according to standard protocols
(Hogan et al., 1986). In order to obtain postimplantation mouse
embryos, freshly dissected embryo implantation chambers were
freeze-protected on dry ice and stored at −75°C.
Immunohistochemistry
Fixation, permeabilization and indirect immunofluorescence on
whole-mount preimplantation embryos were performed as described
(DeSousa et al., 1993). The distribution of immunofluorescent signals
was analyzed with a confocal laser scanning microscope (Zeiss). For
analysis of blastocysts, entrapped in uterus or postimplantation mouse
embryos, serial sections (8-10 µm) were produced with a cryostat and
taken at equal spaces (40 µm) for Methylene Blue staining. Immuno-
fluorescence experiments were performed as described (Bastide et al.,
1993). Affinity-purified rabbit antibodies to specific C-terminal
polypeptide domains of Cx31 (Butterweck et al., 1994a), Cx40 and
Cx43 (Traub et al., 1994) were used. Rabbit affinity-purified anti-
bodies to Cx26, Cx32 (Traub et al., 1989), and Cx45 (Butterweck et
al., 1994b) have been described. Control experiments, using
preimmune sera, did not yield any punctate staining on contact
membranes characteristic for gap junction labelling (data not shown).
Sections were examined with a Zeiss Axiophot light microscope and
photographed with Fuji Neopan 1600 ASA or Ilford HP5 plus (400
ASA) film.
In situ hybridization
The protocol applied has been described in detail (Wisden et al., 1991)
and was slightly modified for the use of embryonic tissues (Dahl et
al., 1995). Connexin-specific antisense oligonucleotides (45-mers)
were derived from the C-terminal part of the coding region which is
unique to each connexin. The specificity of these oligonucleotides,
which corresponded to amino acids 200-214 (Cx31) and 325-339
(Cx43), respectively, was scrutinized by northern blot analysis using
total RNA from various stages of mouse heart development (Cx43)
or mouse skin (Cx31). The labelled antisense oligonucleotides
detected the same transcripts (and no further bands) as labelled DNA
probes (650 bp of Cx31 and 700 bp of Cx43) corresponding to the C-
terminal part of the coding region. For in situ hybridization, oligonu-
cleotides were labelled at their 3′-end with terminal transferase
(Boehringer Mannheim GmbH, Mannheim, Germany) using
[35S]dATP (10 µCi/µl; Amersham Buchler, Braunschweig,
Germany). Controls were performed on consecutive sections in the
presence of a 50-fold excess of unlabelled oligonucleotide (Dahl et
al., 1995). Microautoradiography was performed according to the
supplier’s protocol (LM-1 emulsion, Amersham Buchler).
RESULTS
In the preimplantation mouse embryo Cx31 and
Cx43 proteins are both expressed in inner cell mass
and in trophectoderm
Connexin31 is a member of the connexin gene family which
is expressed in adult murine skin (Butterweck et al., 1994a;
Hoh et al., 1991; Hennemann et al., 1992). In the present work,
we found that Cx31 is expressed in early stages of mouse
development. The distribution of Cx31 protein in blastocysts
was studied by laser scan confocal microscopy using affinity-
purified Cx31 antibodies (Fig. 1). This revealed that the Cx31
protein is ubiquitously expressed on contacting membranes of
both the inner cell mass (ICM) and the trophectoderm (TE). A
similar set of experiments confirmed the occurrence of Cx43
protein in both the inner cell mass and the trophectoderm (data
not shown) as has recently been described by several investi-
gators (Nishi et al., 1991; Valdimarsson et al., 1991; DeSousa
et al., 1993). In order to obtain blastocysts that most closely
resemble the in vivo conditions, cryosections were made from
total uteri at day 3.5 of gestation. Incubations with affinity-
purified Cx31 and Cx43 antibodies confirmed the expression
pattern obtained by laser scan confocal microscopy (data not
shown).
Cx31 and Cx43 expression is strictly
compartmentalized after embryo implantation
Before implantation of the mouse embryo, Cx31 and Cx43
proteins were detected in both principal cell layers of the blas-
tocyst, the epithelial-like trophectoderm and the mesenchymal
cells of the inner cell mass. While the inner cell mass gives
rise to the embryo proper and the yolk sac, the trophectoderm
develops into extraembryonic tissues which later will form the
placenta. Gap junctions are very abundant in decidual cells
(Kleinfeld et al., 1976; O’Shea et al., 1983). At 6.5 dpc, two
days after implantation of the mouse blastocyst, expression
patterns of Cx31 and Cx43 are dramatically changed. Fig. 2
shows a blastocyst entrapped in the uterus, forming an inter-
E. Dahl and others

193Expression of connexin31 and connexin43
stitial implantation chamber. Two consecutive sections were
incubated with either Cx31 antibodies (in Fig. 2a) or Cx43 anti-
bodies (in Fig. 2b). At this time of development, the expression
pattern of Cx31 and Cx43 protein are completely separated
from each other. Cx31 protein is exclusively expressed in cells
that are derived from the trophectoderm lineage, i.e. cells of
the extraembryonic ectoderm (EEC) and the ectoplacental cone
(EPC). Cx43 protein (Fig. 2b), in contrast, is not detected in
trophectoderm-derived tissues but only in those that are
derived from the inner cell mass. The signal is most prominent
in the extraembryonic visceral endoderm (EEN), which forms
the inner layer of the yolk sac. Thus, the expressions pattern
of Cx31 and Cx43 proteins are compartmentalized after
implantation.
Fig. 1. Expression of Cx31 protein in 3.5 dpc
mouse blastocysts. Immunofluorescence
analysis using confocal laser scan
microscopy. (a) Indirect
immunofluorescence with affinity-purified
Cx31 antibodies. (b) Indirect
immunofluorescence with Cx31 preimmune
serum. (a) Cx31 protein is detected on
contacting cell membranes of the inner cell
mass (ICM) and the trophectoderm (TE).
(b) Control, stained with preimmune serum,
exhibits no punctate signals. Bar in a, 20 µm.
Fig. 2. Immunofluorescence analysis of
Cx31 and Cx43 on two consecutive sections
of a 6.5 dpc mouse embryo. (a) Cx31 protein
is only expressed in cells that are derived
from the trophectoderm lineage, i.e. the
ectoplacental cone (EPC) and the
extraembryonic ectoderm (EEC). (b) In
contrast, Cx43 protein is only expressed in
cells that are derived from the inner cell
mass. At this developmental stage,
expression of Cx43 protein in the conceptus
is most abundant in the extraembryonic
visceral endoderm (EEN). Strong staining of
both Cx31 and Cx43 protein is seen in
decidual cells surrounding the implantation
chamber. (c,d) Phase-contrast micrographs
corresponding to a and b, respectively. Bar in
a, 40 µm.

194
Expression pattern of connexins in the gastrulating
mouse embryo
The expression pattern described above is even more pro-
nounced in the gastrulating mouse embryo. At 7.5 dpc, the
three germ layers are established and the axial polarity of the
mammalian embryo is laid down. However, there is still no
expression of Cx31 protein in cells derived from the inner
cell mass (compare Fig. 3a and b). In contrast, expression of
Cx31 protein in the ectoplacental cone and the extraembry-
onic ectoderm is enhanced (Fig. 3c). Compared to 6.5 dpc,
expression of Cx43 protein in the 7.5 dpc embryo is enhanced
in tissues derived from the inner cell mass (Fig. 3d). No Cx43
protein is seen in the region of the ectoplacental cone (EPC).
Higher magnification (in Fig. 3e) shows that, in addition to
abundant expression in ectoderm (EC), mesoderm (M) and
endoderm (EN) also exhibit distinct spots of gap junctional
fluorescence. In the visceral endoderm, there is an increasing
gradient of Cx43 protein towards the extraembryonic region.
Postimplantation expression of Cx31 and Cx43 seems to
mark cell lineages with Cx31 being restricted to cells derived
from the trophectoderm lineage, while Cx43 is detected in
cells derived from the inner cell mass. Furthermore, our
results indicate that Cx31 and Cx43 exhibit gradients of gap
junction expression within the decidua. Expression is
strongest near the implantation side and decreases toward the
myometrium (cf. Fig. 3b for Cx31). The gradient of Cx43
expression in the decidua is not evident in Fig. 3d due to over-
exposure. A similar gradient has been reported for expression
of Cx26 in rat decidua (Winterhager et al., 1993). No specific
immunofluorescent signals were detected on decidual tissue
when the corresponding preimmune sera, instead of anti-
Cx31 or anti-Cx43, were used for incubation (results not
shown).
In order to analyze whether the compartmentation of Cx31
and Cx43 expression is regulated at the level of transcription
or translation, we performed in situ hybridizations with 35S-
labelled antisense oligonucleotides. Fig. 4 shows in situ
hybridization with Cx31-specific antisense oligonucleotides
of a 7.5 dpc gastrulating mouse embryo. Comparison of the
darkfield and brightfield image (in Fig. 4a and b) reveals that
the expression pattern of Cx31 mRNA is in accordance with
the location of Cx31 protein. Cx31 mRNA is strongly
expressed in the ectoplacental cone (EPC) and weaker in the
extraembryonic ectoderm (EEC). No Cx31 mRNA is detected
in any part of the embryo proper. Fig. 4d is a high magnifi-
cation of the region boxed in Fig. 4b, which is lateral to the
ectoplacental cone. It shows that cells of the parietal
endoderm (arrowheads) do not express Cx31 mRNA. Cx31-
mRNA-expressing cells directly adjacent to the parietal
endoderm are presumably derivatives of the cone itself since
E. Dahl and others
Fig. 3. Immunofluorescence
analysis of Cx31 and Cx43 in
the 7.5 dpc gastrulating mouse
embryo using anti-Cx31 (a,b,c)
or anti-Cx43 (d,e) antibodies.
Overview (a,b) showing that
Cx31 protein is expressed in the
ectoplacental cone (star, EPC).
No Cx31 protein was detected in
any part of the three embryonic
germ layers (arrows) or the
extraembryonic visceral
endoderm (EEN), which is also derived from the inner cell mass. D, decidua. (c) The mesometrial part of the
7.5 dpc implantation chamber including extraembryonic tissues. Expression of Cx31 protein in the ectoplacental
cone (EPC) and the extraembryonic ectoderm (EEC) is more abundant than at 6.5 dpc. (d) Overview showing
that Cx43 protein is only expressed in cells that are derived from the inner cell mass. No Cx43 protein is seen in
the region of the ectoplacental cone (EPC). (e) Detail of the embryonic region (embryo proper). The most
abundant expression of Cx43 protein is found in the embryonic ectoderm (EC), but embryonic mesoderm (M)
and endoderm (EN) also express Cx43. Bars, (b,d), 250 µm; (c,e), 100 µm.

195Expression of connexin31 and connexin43
the morphology of these cells (labeled EPC) is very similar
to those found within the cone. Interestingly, at the very distal
tip of the ectoplacental cone, not all cells express Cx31
mRNA (Fig. 4c). Within this region (indicated by the arrow
above EPC in Fig. 4b) only one of two adjacent cells is
strongly labelled. Currently, we do not know whether the
unlabelled cells have not yet started expression of Cx31
mRNA or whether they express another type of connexin not
included in this investigation. As for Cx31, the mRNA and
protein expression patterns of Cx43 are very similar (Fig. 5).
Cx43 mRNA was detected in cells that are derived from the
inner cell mass and was very low or absent in trophectoderm-
derived tissues.
The expression of five further members of the connexin gene
family was investigated by indirect immunofluorescence.
Cx26, Cx37, Cx40 and Cx45 protein could not be detected in
any embryonic or extraembryonic part of the gastrulating
mouse embryo. Very weak staining of Cx32 protein was seen
only in visceral endoderm (data not shown).
DISCUSSION
Gap-junction-mediated intercellular communication is estab-
lished at the 8-cell stage of mouse development (Lo and Gilula,
1979a). It has been shown that these gap junctions channels
contain Cx43 subunits (Nishi et al., 1991; Valdimarsson et al.,
1991; Kidder, 1992; DeSousa et al., 1993). In this study, we
have demonstrated that Cx31 is abundantly expressed in both
the inner cell mass and trophectoderm of the mouse blastocyst.
Currently it is not known whether further connexins participate
in the formation of gap junctions in the preimplantation
embryo.
The 7.5 dpc mouse embryo is separated into two large com-
munication compartments (Kalimi and Lo, 1989). No intercel-
lular communication via gap junctions was detected across the
border between embryonic and extraembryonic cells. Here we
show that, within the gastrulating mouse embryo, Cx43
expression is restricted to embryonic cells, while Cx31 is
expressed in extraembryonic cells. This expression pattern cor-
Fig. 4. Expression of Cx31 mRNA in the
7.5 dpc mouse embryo. Sections were
hybridized to a mouse Cx31 antisense
oligonucleotide. (a) Darkfield micrograph
of autoradiographed and Giemsa-stained
section shown in b. (c,d) High
magnification photographs of distinct
regions indicated in b. (a,b) Cx31 mRNA is
only expressed in the ectoplacental cone
(EPC) and the extraembryonic ectoderm
(EEC). It is not expressed in the ectoderm
(EC), mesoderm (M) or endoderm (EN) of
the embryo proper. (c) Trophectoderm-
derived cells at the very distal tip of the
ectoplacental cone (for location, see arrow
in b). Interestingly, only one out of two
adjacent cells (arrows) expresses Cx31
mRNA. (d) Region lateral to the
ectoplacental cone (boxed in b). Parietal
endoderm (arrowheads) does not express
Cx31 mRNA. Cx31-mRNA-expressing
cells directly adjacent to the parietal
endoderm are derived from the
ectoplacental cone (EPC). EEC,
extraembryonic ectoderm; EEN,
extraembryonic visceral endoderm. Bars:
(a), 150 µm; (c,d), 25 µm.

196
relates with the two large communication compartments
described above. The compartmentation of Cx31/Cx43
expression could be responsible for the electrical barrier
between embryonic and extraembryonic cells. Cx31 and Cx43
are incompatible connexins in cultured cells (Elfgang et al.,
1995). Human HeLa cells, expressing either murine Cx31- or
Cx43-coding DNA, show homotypic transfer of fluorescent
Lucifer yellow between Cx31/Cx31 and Cx43/Cx43 cell pairs
but no heterotypic dye transfer via Cx31/Cx43 channels. Hela
Cx31 transfectants do not form functional gap junctions with
six other connexin transfectants but only with themselves
(Elfgang et al., 1995).
We did not detect any correlation between communication
compartments defined by dye transfer (Kalimi and Lo, 1989) and
expression of a unique type of connexin. The three embryonic
germ layers, for example, constitute distinct dye communication
compartments (Kalimi and Lo, 1988, 1989) but they all express
Cx43 protein. Physical barriers could be responsible for these
dye-defined communication compartments. The cells of the
embryonic ectoderm and endoderm are polarized and develop a
basal lamina, which contains laminin and type IV collagen
(Leivo et al., 1980). Although this basal lamina is permanently
remodeled and may have local breaks (Leivo et al., 1980), it
probably impedes formation of gap junctions between adjacent
cells. Scattered gap junctions have been detected between cells
of the embryonic mesoderm and ectoderm (Batten and Haar,
1979), and these may be responsible for the electrical coupling
measured between communication compartments defined by
dye transfer (Kalimi and Lo, 1989).
The earliest postimplantation stage investigated in this
study was 6.5 dpc, approximately 48 hours after attachment
of the blastocyst. At this stage, complete compartmentation of
Cx31 and Cx43 expression is already established. Cx31
protein is no longer detectable in cells derived from the inner
cell mass. Its expression is confined to extraembryonic tissues,
while Cx43 protein is expressed in embryonic tissues. This
pattern is maintained and strengthened in the 7.5 dpc mouse
embryo. Compartmentation of Cx31 and Cx43 expression
could be a late manifestation of diverse differentiation
programs, adopted by the inner cell mass and the trophecto-
derm. In this case, connexin genes or proteins could be very
late targets of regulator genes acting as early as the 8-cell
stage. Alternatively, the signal for downregulation of distinct
connexins in distinct cell lineages could be induced during the
implantation process. Attachment of the mammalian embryo
to the maternal uterus is accompanied by a highly regulated
program of synchronized events (Sherman et al., 1981).
Cellular signals due to embryo recognition could be responsi-
ble for differential regulation of connexin expression. Com-
partmentation may start when trophoblast cells differentiate
into the invasive ectoplacental cone. It will be interesting to
study which transcription factors regulate Cx31 and Cx43
expression in peri-implantation mouse development. A
candidate for Cx43 could be Oct4 which belongs to the family
of POU-domain transcription factors (Schöler, 1991). Oct4,
mRNA was found in the blastocyst within the inner cell mass
and less pronounced within the trophectoderm. Like Cx43,
Oct-4 was no longer detectable in extraembryonic cells after
implantation. A similar comparison can be drawn between
expression patterns of Cx31 and Mash-2. Mash-2 is a tro-
phectoderm-specific transcription factor of the achaete-scute
family (Guillemot et al., 1994). It is expressed in the embryo
throughout preimplantation development, but later in devel-
opment only in the ectoplacental cone and its derived tissues.
The skilled technical assistance of Gaby Hallas and Petra Vogel is
gratefully acknowledged. This work is part of the Ph.D. thesis
presented by E. Dahl to Bonn University in 1994. This investigation
E. Dahl and others
Fig. 5. Expression of Cx43 mRNA in the 7.5
dpc mouse embryo. Sections were hybridized
to a mouse Cx43 antisense oligonucleotide.
Darkfield (a) and brightfield (b) micrograph
of autoradiographed and Giemsa-stained
section, respectively. Cx43 mRNA is
predominantly expressed in the three germ
layers of the embryo proper (arrows). These
cells are derived from the inner cell mass.
Specific signals are low or absent in the
region of the ectoplacental cone (EPC). Bar
in a, 150 µm.

197Expression of connexin31 and connexin43
was supported by the Deutsche Forschungsgemeinschaft through SFB
284 (Projects C1 and C2) to Klaus Willecke and Otto Traub, by the
Fonds der Chemischen Industrie to Klaus Willecke and through SFB
354, project 12, to Elke Winterhager.
REFERENCES
Bagnall, K. M., Sanders, E. J. and Berdan, R. C. (1992). Communication
compartments in the axial mesoderm of the chick embryo. Anat. Embryol.
186, 195-204.
Bastide, B., Neyses, L., Ganten, D., Paul, M., Willecke, K. and Traub, O.
(1993). Gap junction protein connexin40 is preferentially expressed in
vascular endothelium and conductive bundles of rat myocardium and is
increased under hypertensive conditions. Circ. Res. 73, 1138-1149.
Batten, B. E. and Haar, J. L. (1979). Fine structural differentiation of germ
layers in the mouse at the time of mesoderm formation. Anat. Rec. 194, 125-
142.
Bruzzone, R., Haefliger, J.-A., Gimlich, R. L. and Paul, D. L. (1993).
Connexin40, a component of gap junctions in vascular endothelium, is
restricted in ist ability to interact with other connexins. Mol. Biol. Cell 4, 7-
20.
Butterweck, A., Elfgang, C., Willecke, K. and Traub, O. (1994a).
Differential expression of the gap junction proteins connexin45, -43, -31 and
-26 in mouse skin. Eur. J. Cell Biol. 65, 152-163.
Butterweck, A., Gergs, U., Elfgang, C., Willecke, K. and Traub, O. (1994b).
Immunochemical characterization of the gap junction protein connexin45. J.
Membr. Biol. 141, 247-256.
Dahl, E., Winterhager, E., Traub, O. and Willecke, K. (1995). Expression of
gap junction genes, connexin40 and connexin 43, during fetal mouse
development. Anat. Embryol. 191, 267-278.
DeSousa, P. A., Valdimarsson, G., Nicholson, B. J. and Kidder, G. M.
(1993). Connexin trafficking and the control of gap junction assembly in
mouse preimplantation embryos. Development 117, 1355-1367.
Elfgang, C., Lichtenberg-Frate, H., Butterweck, A., Eckert, R., Traub, O.,
Klein, R. A., Hülser, D. and Willecke, K. (1995). Specific permeability and
selective formation of gap junction channels in connexin transfected HeLa
cells. J. Cell Biol. 129, 805-817.
Guillemot, F., Nagy, A., Auerbach, A., Rossant, J. and Joyner, A. L. (1994).
Essential role of Mash-2 in extraembryonic development. Nature 371, 333-
336.
Haefliger, J.-A., Bruzzone, R., Jenkins, N. A., Gilbert, D. J., Copeland, N.
G. and Paul, D. L. (1992). Four novel members of the connexin family of
gap junction proteins. J. Biol. Chem. 267, 2057-2064.
Hennemann, H., Schwarz, H.-J. and Willecke, K. (1992). Characterization
of gap junction genes expressed in F9 carcinoma cells: molecular cloning of
mouse connexin31 and -45 cDNAs. Eur. J. Cell Biol. 57, 51-58.
Hogan, B., Costantini, F. and Lacy, E. (1986). Manipulating the Mouse
Embryo. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
Hoh, J. H., John, S. A. and Revel, J.-P. (1991). Molecular cloning and
characterization of a new member of the gap junction gene family,
connexin31. J. Biol. Chem. 266, 6524-6531
Kalimi, G. H. and Lo, C. W. (1988). Communication compartments in the
gastrulating mouse embryo. J. Cell Biol. 107, 241-255.
Kalimi, G. H. and Lo, C. W. (1989). Gap junctional communication in the
extraembryonic tissues of the gastrulating mouse embryo. J. Cell Biol. 109,
3015-3026.
Kidder, G. M. (1992). The genetic program for preimplantation development.
Dev. Genet. 13, 319-325.
Kleinfeld, R. G., Morrow, H. A. and DeFeo, V. J. (1976). Intercellular
junctions between decidual cells in the growing deciduoma of the
pseudopregnant rat uterus. Biol. Reprod. 15, 593-603.
Lee, S., Gilula, N. B. and Warner, A. E. (1987). Gap Junctional
communication and compaction during preimplantation stages of mouse
development. Cell 51, 851-860.
Leivo, I., Vaheri, A., Timpl, R. and Wartiovaara, J. (1980). Appearance and
distribution of collagens and laminin in the early mouse embryo. Dev. Biol.
76, 100-114.
Lo, C. W. and Gilula, N. B. (1979a). Gap junctional communication in the
preimplantation mouse embryo. Cell 18, 399-409.
Lo, C. W. and Gilula, N. B. (1979b). Gap junctional communication in the
post-implantation mouse embryo. Cell 18, 411-422.
Lo, C. W. (1988). Communication compartments in insect and mammalian
development. Modern Cell Biology, Gap Junctions (ed. B. H. Satir), pp. 505-
514. Alan R. Liss, New York.
Nishi, M., Kumar, N. M. and Gilula, N. B. (1991). Developmental regulation
of gap junction gene expression during mouse embryonic development. Dev.
Biol. 146, 117-130.
O’Shea, J. D., Kleinfeld, R. G. and Morrow, H. A. (1983). Ultrastructructure
of decidualization in the pseudopregnant rat. Am. J. Anat. 166, 271-298.
Ruangvoravat, C. P. and Lo, C. W. (1992). Connexin43 expression in the
mouse embryo: localization of transcripts within developmentally significant
domains. Dev. Dynam. 194, 261-281.
Rugh, R. (1990). The Mouse: its Reproduction and Development. Oxford
University Press.
Sherman, M. I., Sellens, M. H., Atienza-Samols, S. B., Pai, A. C. and
Schindler, J. (1981). Relationship between the programs for implantation
and trophoblast differentiation. In Cellular and Molecular Aspects of
Implantation (ed. S. R. Glasser and D. W. Bullock), pp. 75. Plenum Press,
New York.
Schöler, H. R. (1991). Octamania: the POU factors in murine development.
Trends Genet. 7, 323-329.
Traub, O., Look, J., Dermietzel, R., Brümmer, F., Hülser, D. and Willecke,
K. (1989). Comparative characterization of the 21-kD and 26-kD gap
junction proteins in murine liver and cultured hepatocytes. J. Cell Biol. 108,
1039-1051.
Traub, O., Eckert, R., Lichtenberg-Frate, H., Elfgang, C., Bastide, B.,
Scheidtmann, K. H., Hülser, D. and Willecke, K. (1994).
Immunochemical and electrophysiological characterization of murine
connexin40 and -43 in mouse tissues and transfected human cells. Eur. J.
Cell Biol. 64, 101-112.
Valdimarsson, G., DeSousa, P. A., Beyer, E. C., Paul, D. L. and Kidder, G.
M. (1991). Zygotic expression of the connexin43 gene supplies subunits for
gap junction assembly during mouse preimplantation development. Mol.
Reprod. Dev. 30, 18-26.
Warner, A. (1992). Gap junctions in development - a perspective. Semin. Cell
Biol. 3, 81-92.
White, T. W., Bruzzone, R., Goodenough, D. A. and Paul., D. L. (1992).
Mouse Cx50, a functional member of the connexin family of gap junction
proteins, is the lens fiber protein MP70. Mol. Biol. Cell 3, 711-720.
Willecke, K., Hennemann, H., Dahl, E., Jungblut, S. and Heynkes, R.
(1991). The diversity of connexin genes encoding gap junctional proteins.
Eur. J. Cell Biol. 56, 1-7.
Winterhager, E., Grümmer, R., Jahn, E., Willecke, K. and Traub, O.
(1993). Spatial and temporal expression of connexin26 and connexin43 in rat
endometrium during trophoblast invasion. Dev. Biol. 157, 399-409.
Wisden, W., Morris, B. and Hunt, S. (1991). Molecular Neurobiology, A
Practical Approach (ed. J. Chad and H. Wheal), pp. 205-225. University
Press/IRL, Oxford.
Wolpert, L. (1978). Gap junctions: channels for communication in
development. In Intercellular junctions and synapses (ed. J. Feldman, N. B.
Gilula and J. D. Pitts), pp. 83-94. Chapman and Hall, London.
Yancey, S. B., Biswal, S. and Revel, J.-P. (1992). Spatial and temporal
patterns of distribution of the gap junction protein connexin43 during mouse
gastrulation and organogenesis. Development 114, 203-212.
(Received 1 May 1995 - Accepted 24 October 1995)